Fortissimo Success Stories Fortissimòs Small & Medium, Manufacturing Enterprises 6th Edition Index 4 Chapter 1: About the Fortissimo Project 7 Chapter 2: The Fortissimo Marketplace 8 Chapter 3: Fortissimo Core Partners Fortissimo Success Stories 11 Chapter 4: Fortissimòs Small & Medium, Manufacturing Enterprises 6th Edition, January 2019 Fortissimo Experiment Partners Publishers: EPCC and Arctur 14 Chapter 5: Published: in 2019 Fortissimo Success Stories Edition: 6th Copyright © 2019 Members of Fortissimo Consortium Text Copyright © 2019 Members of Fortissimo Consortium Editor: Members of Fortissimo Consortium Number of copies: 1000 CIP - Cataloguing in publication Narodna in univerzitetna knjižnica, Ljubljana 334.012.63/.64:004.77 FORTISSIMO success stories : fortissimo’s small and medium manufacturing interprises / [editor members of Fortissimo Consortium]. - 6th ed. - [Edinburgh] : EPCC ; [Nova Gorica] : Arctur, 2019 ISBN 978-961-94587-0-9 (Arctur) 297767168 The research leading to these results has received funding from the European Union Seventh Framework Programme under grant agreement No 609029 and from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 680481. www.ec.europa.eu www.fortissimo-project.eu About the Fortissimo Project SME PARTICIPATION Fortissimo was a collaborative project that enabled European SMEs and midcaps to be more competitive globally through the use of advanced modelling, simulation and data analytics services running on an HPC Cloud-based infrastructure. Fortissimo involved SMEs, midcaps, The Fortissimo Project was funded by the European Commission within the I4MS initiative technical experts, independent software vendors (ISVs) and HPC centres. Ninety-two SMEs (ICT Innovation for Manufacturing Small & Medium Enterprises). It consisted of two phases: participated through “open calls” proposing experiments to demonstrate the benefits of the first phase ran from July 2013 to December 2016; the second phase ran from November HPC cloud-based advanced modelling, simulation and data analytics. During the 18-month 2015 to December 2018. The project was coordinated by the University of Edinburgh and duration of an experiment, the Fortissimo partners provided the participating SMEs with involved 23 core partners from different European countries. access to HPC-resources via a cloud-based infrastructure and supported them in the planning, implementation and realization of their experiments. THE AIM OF FORTISSIMO The aim of Fortissimo was to strengthen the global competitiveness of European industry. SUCCESS STORIES Fortissimo has encouraged the use of advanced modelling, simulation and data analytics The experiments from both phases of Fortissimo resulted in 79 success stories presenting by European Small & Medium Enterprises (SMEs) and midcaps involved in engineering and case studies, demonstrating the use of Cloud-based HPC. These case studies were not only manufacturing. Such use delivers improved design processes, better products and services relevant to the business needs of end users, but also showed the benefits to all involved in the and improved competitiveness. For the European Union as a whole, this means increased value chain from end users to the providers of HPC-infrastructure. This booklet presents all of employment and economic growth. the 79 success stories. THE CHALLENGE FORTISSIMO DATES AND FACTS In the past, gaining access to the High Performance Computing (HPC) resources required for advanced modelling, simulation and data analytics has often presented significant technical PROJECT NAME Fortissimo - Factories of the Future Resources, Technology, and financial challenges to SMEs. Through the use of HPC resources available via a Cloud-Infrastructure and Services for Simulation and Modelling based infrastructure, opportunities leading to business benefits can be identified and applied across the complete value chain saving companies time and money and improving their DURATION 1st phase: July 1st 2013 - December 31th 2016 competitive position. 2nd phase: November 1st 2015 - October 31st 2018 TOTAL BUDGET 1st phase: 21,7 million € 2nd phase: 11,1 million € EU CONTRIBUTION 1st phase: 16,0 million € Joint review of Fortissimo Core Partners. 2nd phase: 10,0 million € CONTRACT NUMBER 1st phase: 609029 under the 7th Framework Programme 2nd phase: 680481 under the H2020 Framework Programme NO OF PARTNERS 1st phase: 122 in total (14 Core Partners and 108 Partners in Experiments) 2nd phase: 96 in total (12 core Partners and 84 Partners in Experiments) CORE PARTNERS 1st phase: The University of Edinburgh (EPCC), Arctur, Bull, CESGA, CINECA, GENCI, Gompute, University of Stuttgart (HLRS), INRIA, Intel, Scapos, Sicos BW, SURFsara, Xlab. 2nd phase: The University of Edinburgh (EPCC), Arctur, Atos, CESGA, CINECA, Gompute, University of Stuttgart (HLRS), Intel, Scapos, Sicos BW, SURFsara, Xlab. NO OF EXPERIMENTS 1st phase: 53 in total, 43 resulted Success Stories 2nd phase: 39 in total, 36 resulted Success Stories 4 5 MAP OF SUCCESS STORIES The Fortissimo Marketplace Countries of Core Partners Countries of HPC is a proven technology in many business applications. Although HPC infrastructure Experiment Partners is affordable for large companies, the costs of owning and maintaining it are beyond the Fortissimo financial capabilities of SMEs and mid-caps. The goal of Fortissimo was to overcome this Success Story obstacle by providing services and tools running on a Cloud-based infrastructure addressing advanced modelling, simulation and data analytics. The Fortissimo Marketplace, launched in October 2016, represents an innovative and sustainable commercial ecosystem for simulation, modelling and data analytics. It simplifies access to HPC services for European companies enabling them to be more competitive. Registration for the Fortissimo Marketplace is free. It offers fast, on-demand and pay-per-use access to HPC services. It matches service providers to users` requirements, handling set up, project management and billing. Such services overcome the need to invest in HPC hardware and software leading to decreased costs and faster time to market particularly for small companies which would otherwise not be able to use HPC solutions. BENEFITS FOR SMES The Fortissimo Marketplace offers novel solutions to companies’ challenges, enables them to discover new business opportunities and brings together all the necessary actors to construct a solution that matches their business requirements. The Fortissimo Marketplace enables companies to: • Greatly reduce time-to-results using pay-per-use SaaS. • Develop complete and innovative solutions in collaboration with an easily available team with diverse expertise. • Analyse large data-sets better, pushing any limits far beyond current capabilities. • Simulate in greater detail and get more reliable results. • Be ambitious by exploring new ideas. OPPORTUNITIES AND ADVANTAGES FOR ISVS The Fortissimo Marketplace is an opportunity to offer enhanced services to customers and reach new markets. By combining software with on-demand HPC resources, Fortissimo allows an Independent Software Vendor (ISV) to offer values-added services that scale up without the need for investing in its own HPC systems. By joining the Marketplace, an ISV has the opportunity to: • Offer customers products (as a secure SaaS) on top of one of the most powerful HPC infrastructures in Europe. • Offer complete solutions working with an easily available team of diverse experts. • Gain visibility and offer products across the whole of Europe. • Focus on their core business and profit from optional services such as account management and billing. www.fortissimo-project.eu www.fortissimo-project.eu 6 7 Fortissimo Core Partners CONSORZIO INTERUNIVERSITARIO CINECA’s mission is to enable the Italian and European research community to accelerate the scientific discovery using HPC resources in a profitable way, exploiting the newest technological advances. For each Fortissimo experiment one of the Core Partners acted as a provider of High www.cineca.it Performance Computing resources and expertise, enabling the other project partners to make best use of HPC in their solution. Providing HPC resources on a pay-per-use basis GRAND EQUIPEMENT NATIONAL DE CALCUL INTEN-removes the need for the other partners to invest in their own HPC hardware. The Core SIF Partners benefit by generating income from the services via the Fortissimo Marketplace, and by developing new partnerships with users of those services. This helps keep them at the GENCI was created with the aim of placing France on the international forefront of innovative support for industry. stage in terms of HPC. Its role is to help implement the national strategy for HPC in support of scientific research within France. www.genci.fr THE UNIVERSITY OF EDINBURGH EPCC is a leading European centre of excellence in advanced GOMPUTE research, technology transfer, and the provision of High Performance Gompute is a leading HPC Cloud company, offering expert services Computing services to academia and industry. and full support to a large number of users worldwide. It delivers www.epcc.ed.ac.uk comprehensive solutions for High Performance Computing, both in-house and as a service. ARCTUR www.gompute.com Arctur is the leading service provider of supercomputing in Central Eastern Europe. Following the XaaS model, it leases the HLRS – UNIVERSITY OF STUTTGART supercomputer along with the services of system administration, The High Performance Computing Centre (HLRS) is one of three code optimization or parallelization. Arctur delivers HPC services and national computing centres affiliated to the University of Stuttgart. solutions to industrial and scientific users in various technologically It operates and provides compute resources to research groups and intensive industries. industry. www.hpc.arctur.net www.hlrs.de ATOS INSTITUT NATIONAL DE RECHERCHE EN Atos is a leader in digital services, with circa 100.000 employees in 72 INFORMATIQUE ET EN AUTOMATIQUE countries. Its deep technology expertise and industry knowledge has Over the last 40 years, Inria’s researchers have been working at the allowed it to build global client base across many different business crossroads of computer sciences and mathematics. Inria is France’s sectors. only public research body fully dedicated to computational sciences. www.atos.net/en www.inria.fr/en BULL SAS INTEL The only truly European IT company capable of designing, integrating and implementing supercomputers, Bull has made Extreme Computing Intel (NASDAQ: INTC) is a world leader in computing innovation. The one of its key strategic priorities. company designs and builds the essential technologies that serve as www.bull.com/extreme-computing the foundation for the world’s computing devices. www.intel.com FUNDACION CENTRO TECNOLOGICO DE SCAPOS SUPERCOMPUTACION DE GALICIA Scapos was founded in 2009 for the sales, marketing and support Fundación Pública Galega Centro Tecnolóxico de Supercomputación of software for technical computing and of advanced software de Galicia is a public foundation committed to the advancement solutions. Its customers range from large industrial corporations and of Science and Technical Knowledge, by means of research and research institutes to SMEs. application of HPC, communications and other IT resources for the www.scapos.com benefit of society. www.cesga.es 8 9 SICOS Fortissimo Experiment Partners Provides access to big data’s analysis, simulation and High Performance Computing. The Sicos BW GmbH, headquartered in Stuttgart, was founded in 2011 by the Karlsruhe Institute of Technology (KIT) and the University of Stuttgart in order to facilitate especially small and medium enterprises (SMEs) access to big data’s analysis, simulation and High Performance Computing. www.sicos-bw.de SURFSARA SURFsara supports research, education and industry in the Netherlands in the areas of HPC and Big Data, through the development and provisioning of advanced ICT infrastructure, services and expertise. www.surfsara.nl XLAB XLAB Research is recognized in Slovenia as one of the strongest computer science research teams outside the academic world. Its expertise lies in security and distributed systems. www.xlab.si www.fortissimo-project.eu 10 11 www.fortissimo-project.eu 12 13 Fortissimo Success Stories 56 High-Performance gear analyzer VE&D, Italy 58 Simulation of laser-based welding in the automobile and machine tool sectors Lasersystemtechnik Bollinger & Ohr, Germany FORTISSIMO SUCCESS STORIES - 1ST PHASE 60 Cloud-based simulation of target drug compounds 18 HPC Cloud-based simulation of light-aircraft aerodynamics Transisight, Germany Pipistrel, Slovenia 62 HPC Cloud-based analysis for optical control in manufacturing processes 20 HPC Cloud-based simulation of steel casting EPC, Spain Ergolines, Italy 64 Cloud-based processing of seismic data 22 HPC Cloud-based design of high-pressure vessels Seismic Image Processing, United Kingdom and Sharp Reflections, Norway Mikrosam, FYR Macedonia 66 Cloud-based simulation of pipeline components for the oil & gas industry 24 HPC Cloud-based simulation of flange tightening Dynaflow Research Group, Netherlands Texas Controls, Spain 68 Cloud-based environmental modelling 26 HPC Cloud-based design of high-voltage cables eAmbiente, Italy Prysmian Group, Italy 70 Cloud-based additive manufacturing 28 HPC Cloud-based urban planning HSL, Italy IES, United Kingdom 72 Advanced simulation for metal forming 30 HPC Cloud-based optimisation of aircraft wiring Matrici, Spain KE-Works, Netherlands 74 Optimised cutting and bending of steel reinforcement bars using Cloud-based HPC 32 HPC Cloud-based prediction of air quality Schnell, Spain Numtech, France 76 HPC Cloud-based optimisation of water turbines for power generation 34 HPC Cloud-based reduction of vehicle emissions Zeco, Italy AVL, Austria 78 HPC-based prediction of the optical properties of dyes 36 HPC Cloud-based simulation of hazardous chemicals Scriba Nanotechnologie, Italy Lonza, Switzerland 80 Multi-physics simulation of high-temperature superconducting devices 38 HPC Cloud-based design of copper-alloy moulds Oxolutia, Spain IMR, Italy 82 HPC-Cloud-based monitoring of crowds 40 HPC Cloud-based simulation of sports-car aerodynamics DFRC AG, Switzerland Koenigsegg, Sweden 84 Cloud-based design of motorcycle helmets 42 HPC Cloud-based design of centrifugal pumps NolanGroup, Italy EnginSoft, Italy 86 Simulation of High-Performance composite materials in the automobile industry 44 HPC Cloud-based simulation of drifting snow Formtech Composites, United Kingdom Binkz, Canada 88 Cerebral blood-flow simulations 46 HPC Cloud-based molecular modelling Vittamed, Lithuania Albemarle Corporation, Netherlands 90 Cloud-based simulation of complex fluids 48 HPC Cloud-based map interpolation for civil engineering Ioniqa Technologies, Netherlands Sisener Ingenieros, Spain 92 HPC Cloud-based seakeeping design 50 Advanced simulation of electromagnetic effects Wavec, Portugal and Vicusdt, Spain Seemi, France 94 HPC Cloud-based standard strength assessment of commercial ships 52 Advanced simulation of metal casting Isonaval, Spain Fundiciones de Roda, Spain 96 Cloud-based shape optimisation in manufacturing 54 Simulation of airflow in concentric chimneys Automobili Lamborghini, Italy DINAK, Spain 14 15 98 HPC Cloud-based microscopy 136 Cloud-based simulation of the hydrodynamic resistance of ships’ hulls in the presence MicroscopeIT, Poland; UZH, Switzerland; OpTecBB, Germany and Nikon, Switzerland of waves 100 HPC based Delft3D service for modelling flooding, morphology, and water quality CETENA, Italy HKV, Netherlands 138 Cloud-based micrositing of small wind turbines 102 Cloud-based simulation of marine communication buoys KLiUX Energies, Spain Alseamar, France 140 Cloud-based simulation of page curling in the copying of documents Bookscanner, Greece 142 Data analysis to improve the welfare of laboratory animals FORTISSIMO SUCCESS STORIES - 2ND PHASE Tecniplast, Italy 104 Cloud-based simulation for vehicle engineering 144 Cloud-based optimisation of a multi-body wave energy device Koenigsegg, Sweden Zyba, United Kingdom 106 Optimising gas and flame detector layouts in hazardous manufacturing and 146 CAE driven design of a water treatment plant production plants Akvola, Germany Micropack, United Kingdom 148 Earthquake resistant design using FE² technique 108 Advanced, Cloud-based HPC simulation in the foundry business Egis Industries, France ProService, Italy 150 Cloud-based multiphysics simulation for MEMS micro speaker design and development 110 HPC-based design of non-circular gears USound, Austria STAM, Italy 152 Improvement of the remote expert system based on software OTEA 112 HPC-based high-resolution modelling of magnets EcoMT, Spain Magneti Ljubljana, Slovenia 154 Digester 2.0: Optimization of the anaerobic digestion process for biogas generation 114 Cloud-based HPC optimisation of manufacturing processes Norvento, Spain EMO Orodjarna, Slovenia 156 Cross-solver Cloud-based tool for aeronautical FSI applications 116 Cloud-based HPC processing for knowledge generation in camshaft manufacture RINA, Italy EPC, Spain 158 Large-scale indie gaming analytics (LIGA) 118 Cloud-based simulation of box-framed timber beams Kumo Technologies, Spain Maderas Besteiro, Spain 160 Massively parallel virtual testing of safety-relevant driving systems 120 Near real-time analysis of airframe certification test data Spicetech, Germany Colosso, Netherlands 162 The Virtual Engineer 2 – Cloud-based optimisation of production in the dairy sector 122 Cloud-based HPC simulation of railway infrastructure for high-speed trains Moody Direct, United Kingdom Alstom, Italy 164 OptiBike robust lightweight composite bicycle design and optimization 124 Predictive diagnosis services for the automotive industry IDEC, Spain TEXA, Italy 166 HPC Cloud-based simulation of coupled electromagnetic and structural-acoustics in 126 Cloud-based simulation of ultra-high-temperature furnaces in-wheel electric motors XERION, Germany Elaphe, Slovenia 128 Cloud-based design of yacht sails 168 Multiphysics icing analysis using HPC Cloud Cape Horn Engineering, United Kingdom MT Propeller, Germany 130 Cloud-based estimation of marine freight rates 170 Cloud-based multiphysics simulation for designing highly dynamic and highly accurate Vasco Gallega Sociedad de Cartera, Spain flow controls for microfluidic applications ELVESYS, France 132 Simulation of additive manufacturing processes for the production of metal components High Performance Engineering, Italy 172 HemoSphere – prototyping of blood-processing microfluidics in the HPC Cloud Aenitis Technologies, France 134 Cloud-based design of ultra-clean containers for high-purity chemicals Rikutec, Germany 174 Cloud-based simulation of desalination systems powered by renewables Photovoltaic, Greece 16 17 EXPERIMENT #401 HIGHLIGHTS Industry Sector: Aerospace HPC Cloud-based simulation of Country: Slovenia light-aircraft aerodynamics Software used: OpenFOAM ORGANIZATIONS Pipistrel is a Slovenian SME. It was established in 1989 and is a leading designer and manufacturer of light aircraft. In order to develop its product line, Pipistrel needs to understand THE SOLUTION how air flows over its aircraft. For an SME, it is virtually impossible to use wind tunnel tests during the design The use of Cloud-based HPC allowed Pipistrel to run simulations which accurately predicted phase, because such tests are simply too expensive. how an aircraft would behave in flight. To simulate the air flow around the aircraft with The only option an SME has is to simulate the flow of the required degree of accuracy, a large computer model was used. Such a model ran in air as accurately as possible using a High Performance approximately 2 to 3 days on a Cloud-based HPC system. This enabled Pipistrel to perform Computer. To replace wind-tunnel tests satisfactorily, aerodynamic models, which accurately complex simulations in a reasonable time and at an affordable cost. simulate real airflows, need to be deployed. Such models require significant compute cycles and memory. End User HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.pipistrel.si www.xlab.si www.arctur.si Pipistrel needs to simulate the flow of air over the body of • 10 times cheaper to an aircraft only occasionally during the design process. It use Cloud-based HPC THE CHALLENGE estimates that it is 10 times cheaper to use Cloud-based simulations to having HPC simulations than have a suitably powerful in-house a powerful in-house The challenge facing Pipistrel was to perform simulations of the flow over its aircraft in system which is only used for part of the time. The indicative system. sufficient detail to model real physical effects accurately. Such simulations require expensive annual costs of using Cloud-based HPC simulations are computer resources which are normally beyond the means of an SME. However, the use of • Higher resolution of approximately €30.000 compared with an in-house costs of Cloud-based HPC offers the possibility of running such simulations on a pay-per-use basis simulations by HPC €300.000, which shows that this saving is considerable. which is financially viable for an SME. provide more quality This case study allowed Pipistrel to use HPC for the data. first time and to learn about its capabilities. Pipistrel ran • Design time more demanding, higher fidelity simulations. It gained is reduced by considerable experience in the use of HPC-based simulation. incorporating This experience will help Pipistrel to estimate the time and simulations in each the cost of such simulations better. This will help them to design phase. decide if the use of HPC is justified or not in future projects. Pipistrel learned that the use of HPC will be very valuable during a design phase of future aircraft. HPC can be used to run much more demanding simulations that improve the fidelity of results. The time needed for such simulations running on an HPC system is roughly the same as the coarser simulations currently run on Pipistrel’s in-house cluster. The higher-resolution simulations give more and better data that can be incorporated into each design phase. This both accelerates the design phase and reduces the number of the design cycles. 18 19 EXPERIMENT #402 HIGHLIGHTS Industry Sector: Casting HPC Cloud-based simulation of Country: Italy steel casting Software used: OpenFOAM ORGANIZATIONS Ergolines is an Italian SME. They are world leaders in manufacturing products used in the production of speciality steels, including special instruments used in continuous casting. Ergolines’ THE SOLUTION goal is to develop equipment which supports production of flawless steel alloys with properties that satisfy an HPC simulations gave new insights into the physics of the problem and different ladle-increasingly quality-oriented market. Ergolines routinely uses emptying mechanisms. The discovered link between the shroud vibrational signal and the in-house computational resources during development to fluid dynamics of the system forms the basis of an innovative monitoring technology, which simulate the flow of liquid steel as it solidifies. This case will significantly improve occupational safety and productivity in steel plants. Ergolines had study addresses the problem of slag carry-over from the ladle to the tundish, which can lead previously used simulation in its design process, but this was their first experience of using impurities in steel or poor ladle yield. This is a complex phenomenon which cannot be observed full-scale HPC and seeing its potential benefits. directly, and simulating it requires the use of HPC, which Ergolines has not previously used. End User HPC Expert & Centre BUSINESS IMPACT BENEFITS www.ergolines.it www.arctur.si To get accurate results in complex slag carry-over simulations, • The computational it was necessary to use very fine-grained simulations for times significantly THE CHALLENGE both geometry and time. This required powerful computing reduce due to the resources that Ergolines did not have. Access to a Cloud-Cloud-based HPC In continuous steel casting it is crucial to prevent slag (a by-product) from transferring to the based HPC system thus allowed Ergolines to reduce services. tundish. This causes breakouts, resulting in hot liquid steel spraying out – with serious safety computational times without incurring the high costs of and economic implications. An automated system for ladle-slag monitoring is thus in high • Occupational safety HPC ownership. The results from the fluid-dynamic analysis demand, one which matches the vibration of the slag with the fluid dynamics of the system. and productivity enabled development of a new automatic slag detection This requires detailed simulation, which can only be done using HPC. improve with the technology for steel continuous casting, which could bring new automatic slag significant improvements to occupational safety and detection technology productivity of steelworks. developed during the The ability to detect slag while it is passing through the experiment. shroud enables a steel plant to control the closing of the ladle • 60% less steel is more finely. For an average ladle size of 100 tons, 0,5 - 1% of wasted with the steel usually remains in the ladle. With the new monitoring new monitoring technology, 60% of the lost steel can be saved. On an average technology, which production of 1 million tonnes of per year, a medium-size represents energy and factory could avoid re-melting 6.000 tonnes of steel. Re-monetary savings of melting 6.000 tonnes of steel at approximately €70 to €100 up to €670.000 per per tonne represents a saving of €420.000 to €600.000. The year. loss of a further 300 tonnes of steel could be also avoided, saving 70.000€, for total savings of up to €670.000 per year, per medium-sized plant. Casting is a high-energy process, so on top of the monetary savings, energy savings for steel plants equipped with the monitoring technology would be significant. 20 21 EXPERIMENT #403 HIGHLIGHTS Industry Sector: Composites HPC Cloud-based design of Country: high-pressure vessels FYR Macedonia Software used: OpenFOAM ORGANIZATIONS Founded in 1991, Mikrosam is an SME, which manufactures winding machines for the construction of composite pressure vessels such as natural gas containers in automobiles. It THE SOLUTION is involved in the design of pressure vessels and the development and adjustment of its machines to produce A HPC-based computer model was developed to simulate the behaviour of composite such designs. The biggest advantage of modern composite materials using an open-source software package. This model needed significant computing materials is that they are many times lighter than and yet resources and ran on an HPC system. Such simulations reduce the amount of physical as strong or stronger than widely used metals. Composite testing needed in the design of composite materials. Previously Mikrosam was using materials have the potential to replace widely used steel and aluminium, often with better simulation in its design process. However this case study was their first experience of HPC performance. Composite components can lead to weight savings of between 60% to 80% by and the benefits it could bring. replacing steel components and between 20% to 50% by replacing aluminium components. End User HPC Provider BUSINESS IMPACT BENEFITS www.mikrosam.com www.arctur.si For each filament winding machine used in the production • Reduced design time of high-pressure vessels, different combinations of by 30%. THE CHALLENGE materials and winding angles for the composite, need to • Reduced number of be considered. With the simulation code developed in this Previously for Mikrosam, design involved physical tests. To replace physical tests with physical tests and case study it is possible to shorten the design time and numerical simulation requires significant computing resources. An HPC-based system is prototyping costs. to reduce the number of physical tests and prototyping necessary to get results in a reasonable time frame. The challenge was to develop a model costs. The case study has shown that using simulation • Reduced production for the simulation of composite materials and to implement it on an HPC system which on an HPC system can reduce time for composite costs by about would enable accurate results to be obtained within an acceptable development time. design by about 30% and reduce the cost of design and €30.000 per year for testing by 10%. As a result of the Fortissimo experiment, the next 5 years. Mikrosam will be able reduce its production costs by about €30.000 per year for the next 5 years. More importantly its product offer (production machines for high-pressure vessels) will give each of its customers an advantage in production costs of around €300.000 per year per machine for machines that are currently on the market. The considerable advantage of Mikrosam’s products over those of its competitors will, as a conservative estimate, lead to an increased revenue for the company of around €400.000 per year for the next 5 years. This is a significant amount for this SME. The experience gained in this experiment together with reduced production costs for both Mikrosam and its customers constitutes a base for further growth of the company and the resultant creation of new jobs. 22 23 EXPERIMENT #404 HIGHLIGHTS Industry Sector: Mechanical Engineering HPC Cloud-based simulation of Country: Spain flange tightening Software used: code_aster ORGANIZATIONS Texas Controls is a Spanish SME offering tightening and sealing solutions to large industrial facilities in a number of sectors. These are especially important to customers in the oil and gas industry, THE SOLUTION where it is imperative to avoid leaks in pipes, pressure vessels, or reactors that are under extreme pressures and temperatures. A computer model was developed which represented all the functional parts of a flange, In these situations, preventing leaks is much less costly than including the gasket and the tightening bolts. This model was driven by a user interface, dealing with consequences of one. It is crucial to be able to which enabled different tightening scenarios to be evaluated. Using the model, Texas Controls predict the behaviour of flanged joints, and to understand could accurately simulate and improve the design of the tightening process. The model was elastic interactions between them. Texas Controls has used computer simulations previously, implemented using both open-source and proprietary simulation codes. Several sizes of case however, the computational demands of modelling the behaviour of flanges were high, and so an studies were run. HPC-based solution was necessary. This was the first time they had used such technology. End User Domain Expert HPC Centre & Expert BUSINESS IMPACT BENEFITS www.texascontrols.com www.aimen.es www.cesga.es Based on previous experience in the field (when no • 33% time saving simulations were carried out), a non-optimised tightening of per flange for Texas THE CHALLENGE a 24 stud bolt flange took 108 man-hours; using simulation, Controls. Texas Controls reduced the process to 72 man-hours To seal a joint properly, a gasket is placed inside a groove located on both flanges. The • Optimization of flange - a 33% time saving per flange. Whilst this represents challenge here was to simulate and optimise tightening of the flanges - during the closing and tightening process. considerable savings in labour costs, the most important tightening process, the gasket and the flanges may be damaged through deformations and outcome is the reduction in downtime of industrial • Increased revenue high levels of stress. If the joint is damaged, it could fail, which could seriously harm workers, installations such as refineries. The cost of “down time” for of €2 million over the the surrounding community, and the environment. a medium-sized hydrocracker is about €21.000 per hour next 3 years. (€500.000 per day). Using advanced simulation, flange tightening can be reduced from 27 hours to 18 hours. This means a saving to the end-user of ~€180.000, because the shutdown path is shortened by the same amount that the tightening process is optimized. HPC-based simulation also allows technicians to avoid damage to flanges during the tightening, which is not possible using the usual experience-based method. This can also have significant cost implications. These benefits give Texas Controls a significant competitive advantage in a highly technical industrial sector, which should result in winning major, international commissioning contracts. This is expected to result in an increase in revenue of €2 million over the next 3 years, and a related 15% increase in staff employed. Texas Controls has a range of other services where the use of cloud-based HPC tools have the potential to similarly enhance Texas Controls’ competitive advantage. 24 25 EXPERIMENT #405 HIGHLIGHTS Industry Sector: Electrical Engineering HPC Cloud-based design of Country: Italy high-voltage cables Software used: COMSOL ORGANIZATIONS Prysmian Group is a world leader in the development and supply of energy and telecom cables and associated systems. Prysmian Group develops underground and submarine power cables in the energy THE SOLUTION sector, and medium- and low-voltage cables for the construction and infrastructure industry, and so need to model the physical Both the third-party and open-source simulation packages were ported to a Cloud-based behaviour of high-voltage power-transmission cables. Prysmian HPC system. The resultant simulations enabled Prysmian to improve calculation of losses has a clear business requirement to scale-up its current inside energy cables, especially those induced by the magnetic field due to the load current. simulations of energy cables and systems to a larger refinement Prysmian is also able to better model the heat transfer from the cable to the environment. and size, which exceeds the computing power available in-house. Prysmian would also like to This has improved their design capability significantly – 2D simulations have been replaced undertake more ambitious simulations, requiring new competencies and tools. The demonstration by much more realistic 3D simulations on the HPC system. of a successful, cost-effective cloud-based HPC simulation would be a breakthrough for Prysmian. End User HPC Expert & Service Provider BUSINESS IMPACT BENEFITS www.prysmiangroup.com www.cineca.it The business benefits from the use of a Cloud-based • Move from 2D model HPC system arise from several sources. The move from to 3D. THE CHALLENGE the previous 2D model to an advanced 3D model enables • Better and more much better cables to be designed. This allows Prysmian Prysmian has used a standard modelling package for several years as its main tool for detailed simulations to retain the competitive edge needed to remain market electromagnetic simulations, using 2D models on a few high-end workstations. However, finer in a shorter time. leader with a time to market independent of the increase simulations were needed, which were achieved using Cloud-based HPC to create a 3D model, in complexity of the design process. The need for Cloud- • Savings around 30% and addressing how improved simulations could be exploited in a future business model. based HPC simulation is of the utmost importance in per annum in costs Another goal was to use open-source software in these simulations, to avoid expensive licences. this case study. Cloud-based HPC modelling has enabled for cycle alone. better, more detailed simulations to be made in a shorter time. Running such simulations would not be feasible on a network of workstations, so previously Prysmian did not have this capability. Prysmian estimates that the use of a Cloud-based HPC system instead of investing in an in-house system can provide savings of around 30% per annum in costs for cycles alone. This is based on the costs of the cycles needed to run the simulations in the HPC-Cloud compared with the costs of a suitable HPC computer system which would cost €150.000, amortised over 3 years with additional annual operational costs of €40.000. Further savings were also identified through the use of expertise on-demand at the computer centre rather than through the employment of a member of staff with the necessary expertise in simulation. 26 27 EXPERIMENT #406 HIGHLIGHTS Industry Sector: Urban Planning HPC Cloud-based urban planning Country: United Kingdom Software used: In house code ORGANIZATIONS Founded in 1994, IES is an SME based in Scotland. IES develops the world’s leading integrated building-performance modelling software system, designed to enhance building performance and create THE SOLUTION more sustainable buildings. The IES‘s Virtual Environment (VE) is a suite of tools based around a single integrated data IES’s planning tool addresses large simulations quickly and effectively through Cloud-based model. IES provides leading-edge support for the design, HPC, but uses a familiar workstation for data display. This case study has proved to be very construction and operation of some of the largest, most significant for IES. Simulations run from personal workstations or company servers were challenging buildings in the world. Based on its VE, IES has often long and tedious, with the consultant having to wait for results to be available. A cloud-developed a planning tool for cities, which will enable stakeholders to assess the energy efficiency based HPC approach significantly decreased the run-time of simulations whilst substantially of a city, quality of living, etc. The tool can be used in cities just beginning the journey towards a increasing the number of buildings per simulation. ‘smart’ city, or those that are well on their way towards integration with ‘smart’ technologies. End User, Application Expert & ISV HPC Provider & Expert BUSINESS IMPACT BENEFITS www.iesve.com www.epcc.ed.ac.uk In this case study, model simulations ranging from the • Ability to model much small to the very large were tested. Typical speedups larger structures. THE CHALLENGE (comparing the workstation to the HPC system) were • Ten-fold speed-up in between 5 and 10 times faster for the HPC system. The The purpose of IES’s VE is to provide the information required to design, build and operate turnaround time for major benefit of this is that simulations, which previously better performing, more sustainable communities. The planning tool based on the VE relies simulation jobs. had unrealistically long compute times of days or weeks heavily on HPC cloud-based simulation, because of the amount of data generated by multiple on the workstation, could be run in a few hours or days • Much greater buildings. A major objective here was to enable VE desktop installs and web-based interfaces using the Cloud-based HPC system. As IES’s VE is used productivity for to access calculation resources hosted on an HPC Cloud infrastructure. today by many of the world’s leading architectural and consultants working engineering practices, this has clear commercial benefits. in design projects. Indeed, based on the successful experiment, IES is now offering an HPC-based service to its customers. This service embodies a pay-as-you-go approach which is underpinned by HPC-systems available from EPCC, the HPC centre of the University of Edinburgh. A 64-core in-house system costing £33.000 would have an annual running cost of around £22.000 (support, maintenance, electricity, housing, etc.). Amortised over three years, this gives a total cost per core hour of £0,10, compared with a cost of £0,05 for Cloud-based HPC cycles – already a 50% saving. Cloud-based HPC looks even more cost effective when it is considered that an in-house system would never be used continuously, and so would be even less competitively priced than this calculation indicates. 28 29 EXPERIMENT #408 HIGHLIGHTS Industry Sector: Electrical Wiring HPC Cloud-based optimisation Country: Netherlands of aircraft wiring Software used: MORE ORGANIZATIONS KEW, an SME founded in 2008, specialises in the optimisation of engineering-intensive projects in the manufacturing industry. In particular, KEW is active in the aerospace sector and has THE SOLUTION developed an application to optimise the routing of wiring within an aeroplane. This is an important issue in the The solution has involved porting the KEW optimization software to run on an HPC system design of aircraft, which traditionally has been addressed and developing the necessary “glue” software to bring all the necessary software components by trial and error. Although this problem can be tackled together taking account of any software licensing issues. The successful implementation using computer models, applications to do this are very of this solution has enabled typical optimisations to be run on an HPC system much more computationally intensive. SMEs generally do not have the available resources to buy and quickly and effectively. It should be noted that this case study was the first time that KEW maintain the large computer systems needed to perform the necessary optimisations, nor do had used HPC in its wiring optimization. they have the required expertise to use such systems. End User Domain Expert HPC Expert HPC Centre BUSINESS IMPACT BENEFITS www.ke-works.com www.noesissolutions.com www.cineca.it www.gompute.com By using advanced simulations, KEW improved the quality • Reduction of about of its designs, and reduced costs by 2,5% per design. Using 90% in the lead-time THE CHALLENGE Cloud-based HPC instead of in-house resources contributed for a single (non-to a reduction of 90% in the lead-time for the design of a recurring) electrical The challenge addressed by this case study was to adapt a wiring optimisation application single Electrical Wiring and Interconnection System, and design. from KEW to run on a Cloud-based HPC system, so that wiring layouts could be optimised in an eight- to ten-fold reduction in computational costs. The a feasible length of time, at an acceptable cost. This solution would involve computationally • Reduction of impact of a single simulation may be very significant, as a intensive simulations that could be run on a pay-per use basis, with significant savings over computational costs single optimised wiring design may be used in hundreds owning a system, and would offer sufficient resources to satisfy demand. of 8 to 10 times by of aircraft. The automated, optimised process running on using Cloud-based a Cloud-based HPC system gave a 2,5% reduction in cost HPC vs in-house and weight of the wiring system. This is a recurring benefit, resources. as all aircraft of the same type will utilise the same wiring design. As the aerospace industry has very fine margins, • Average saving of a 2,5% saving may increase profit margins by 50%. Saving costs and weight of one kilogram in the wiring may enable a 20kg reduction in wiring system of 2,5%, overall aircraft weight, which will reduce fuel consumption meaning an important over the aircraft’s life. The cost of computation on a Cloud-recurring benefit and based HPC system is about €660 for a single design run, profit margins of up compared to an annual cost of an in-house HPC system of to 50% in aerospace about €61.000. Furthermore, a Cloud-based HPC system industry. is much more flexible if more computational power is required. This demonstrates the feasibility and cost-effectiveness of using Cloud-based HPC for engineering simulations. SMEs are much more able to afford to use Cloud-based HPC, allowing them to compete better with larger organisations. 30 31 EXPERIMENT #410 HIGHLIGHTS Industry Sector: Urban Planning HPC Cloud-based prediction Country: France of air quality Software used: ADMS Urban ORGANIZATIONS Numtech is a French SME specialised in the development and use of innovative digital tools for air-quality and meteorological simulations. They are the market leader in France for modelling the weather and THE SOLUTION atmospheric dispersion. Its customers are mainly large companies, local and regional authorities, and research institutes. This case study used the ADMS-Urban software from CERC running on the Extreme Numtech makes substantial use of the ADMS code from CERC Factory HPC offering from BULL. ADMS-Urban was adapted to run on a Cloud-based HPC in its business. European regulations on air-quality now require system. The results of the simulations were then made available via a familiar workstation more and more testing and evaluation of pollution adaptation environment. In doing this, an evaluation of the viability of this service on commercial HPC and reduction scenarios. Rather than evaluating 2 to 3 scenarios, consulting companies and Clouds has been carried out and possible business models for such a service have been regional air-quality agencies now need to evaluate tens of scenarios. This requires an increase in proposed. their capacity for computing, beyond what they can manage with in-house resources. End User Software & Domain Expertise Provider HPC Providers & Experts BUSINESS IMPACT BENEFITS www.numtech.fr www.cerc.co.uk www.bull.com To offer a simulation service, CERC needs to source • A Cloud service on a computer cycles, which means it either needs to own pay-for-use basis is THE CHALLENGE and maintain a sufficiently powerful HPC system, or it the pricing solution needs to buy cycles from an HPC centre. In the former for the infrequent The challenge here was to demonstrate the use of Cloud-based-HPC services to investigate case, a powerful enough in-house server would need to users of HPC. air-quality at city-scale. Running simulations using Cloud-based HPC would help to increase be purchased. Investigation shows that if the average the numbers of scenarios which could be feasibly simulated in a given time, and reduce the • Production savings time a server is in use falls below 40%, a pay-on-demand computational time needed for simulations. The outcomes of using this system include are severely lowered Cloud service becomes a more economically viable option, shorter times for simulations and cost reductions, with resultant competitive advantages. in case of Cloud- compared with the costs to acquire and maintain that based HPC services server. This depends on the mode of use, but buying cycles use. on demand offers considerable flexibility to SMEs looking to set up a service. As a result of this study, CERC can • Possibility to launch now offer the ADMS-Urban software as a cloud service on multiple scenarios a pay-for-use basis, rather than requiring a customer to at a reasonable purchase an annual licence and run the software locally on cost using SaaS workstations. This allows for an attractive pricing option constitutes a new for customers needing an infrequent use of the model. business model that Using Cloud-based HPC services, Numtech can save further increases the production costs of €125.000 over 5 years, which allows turnover. them to offer more competitive services. By recruiting new customers, Numtech’s turnover is expected to increase by a total of €750.000 over the same time. Finally, the new capacity to launch multiple scenarios at reasonable cost using SaaS constitutes a new business model that will further increase NUMTECH’s turnover by €150.000 up to 2020. 32 33 EXPERIMENT #412 HIGHLIGHTS Industry Sector: Automotive HPC Cloud-based reduction of Country: Austria vehicle emissions Software used: AVL CRUISE ORGANIZATIONS AVL is the world’s largest independent company in the development of powertrain systems for internal combustion engines and associated instrumentation and test systems, and has been working THE SOLUTION in partnership with companies all over the world for more than 60 years. The need for CO2 reduction, the increasing Providing sufficient computational resources to accomplish optimization tasks in an complexity of new powertrain systems, and a requirement to acceptable time-frame is a struggle faced not only by SMEs, but even by larger companies. achieve the highest possible level of process efficiency are The outcome of this case study has been to demonstrate the viability of on-demand some of the key challenges facing the automotive industry computing resources in the design of powertrains with specific emphasis on the reduction now and for the foreseeable future. AVL provides its customers, many of which are SMEs, with of CO2 emissions. This solution involves the running of AVLs simulation codes on a Cloud-a set of comprehensive simulation tools in a flexible and open environment enabling multi-based HPC system where computer resources are made available on-demand. disciplinary solutions as an integral part of the powertrain development process. End User & Code Owner HPC Centre & Expert BUSINESS IMPACT BENEFITS www.avl.com www.hlrs.de/home/ The most clear cost benefit of using HPC Cloud resources • Possibility to is the possibility to lease a powerful computing cluster lease a powerful THE CHALLENGE for single projects, instead of acquiring and maintaining computing cluster in-house computational resources. These would likely be for single projects This case study addresses the use of on-demand, Cloud-based HPC resources to tackle underutilized for most of the time, and may even not be instead of acquiring the important requirement for the reduction of CO2 emissions in the design of vehicles. sufficiently utilised to justify their cost when in use. Using and maintaining The majority of projects in the area of vehicle optimization involve studies with large-scale a Cloud-based solution, taking into account all additional computational variations in parameter and components on a limited palette of base vehicle models. These cloud overheads, short-term projects running millions resources. studies require high levels of CPU cycles on-demand. of simulations on 400 cloud CPU cores for a period of a • The use of a Cloud- couple of weeks, several times a year, would run with costs based solution can reduced by up to 90% when compared to the total cost of lead to reduced costs ownership of a dedicated in-house system. This is the cost by up to 90%, when range where it becomes attractive for SMEs to participate compared to the total in projects which require high CPU power for only a short cost of ownership of time. a dedicated in-house The results of using a Cloud-based approach mean system. AVL are better able to offer their powerful simulation platforms to its clients. Given the breadth of the product space AVL and its clients occupy – among them diesel engines, electric drives, alternative fuels, control software, transmissions, and batteries - this should have significant economic and competitive advantages for AVL. AVL is now better able to go about its business, tackling the development of highly creative, mature and application-specific solutions for its customers in order to meet their market challenges. 34 35 EXPERIMENT #414 HIGHLIGHTS Industry Sector: Chemical Engineering HPC Cloud-based simulation of Country: Switzerland hazardous chemicals Software used: ms2 ORGANIZATIONS Founded in 1897 in Switzerland, Lonza is one of the world’s leading and most trusted suppliers to the Pharma & Biotech and Specialty Ingredients markets. A key part of Lonza’s business is in distillation THE SOLUTION columns, whose operation requires a detailed knowledge of the thermodynamic properties of the target compounds. The chemical A detailed molecular-simulation code has been implemented on an HPC system driven by industry usually measures the required data experimentally, and a simple, web-based user interface. Multiple simulations of state points can be initiated this was previously how Lonza operated as well. However, when through this interface enabling the complete thermodynamic properties of a compound to be determining the properties of hazardous substances (explosive, determined in a reasonable length of time. For example, whilst the calculation of a complete toxic or mutagenic), computer-based simulations are a very attractive alternative to dangerous and set of physical properties would take ~6 months on a 16-core cluster, the calculation time expensive physical tests. Powerful predictive methods now exist that calculate the properties of can be reduced to below 20 hours on an HPC system. compounds using simulation, but these require significant computing power. End User Domain Expert HPC Centre & Expert BUSINESS IMPACT BENEFITS www.lonza.com www.uni-paderborn.de www.hlrs.de/home The use of simulation can bring massive savings to • Massive savings Lonza’s production process. Shortening the time needed to in the production THE CHALLENGE determine a compound’s properties brings clear benefits process of chemical to the design process. It is also clear that the cost of intermediates In this case, the goal was to demonstrate the benefits of porting an existing third-party code cycles is much less than that of owning and maintaining manufacturers. to an HPC system. To determine a full set of physical properties for one compound, around a large HPC system in-house. The cost to determine a 200 state points need to be calculated. Currently, a single state point calculation takes 20 • €1.100 savings single property of a single compound experimentally is hours on a 16-core workstation, so the total calculation would take up to 4.000 hours - almost on the costs of approximately €2.700. For a mixture of compounds, this six months. Access to an HPC system would decrease this to a reasonable level. experimentally cost increases significantly - 60 gas solubility data points determining pure of a binary mixture can cost up to €50.000. Compared to component densities that, 60 molecular simulations covering the entire fluid for a substance and region up to arbitrary high pressures will cost around ~€47.000 savings €1.600, while for the mixture the cost usually doubles for mixtures of compared to a pure component, so €3.200. Not only does compounds. this represent a huge reduction in costs for Lonza, this also covers conditions that are difficult or impossible to test • €86.400 overall saving experimentally. A typical distillation column designed by in the design process Lonza costs around €1,5 million. This includes the design of a distillation of a system to obtain all required physical properties column costing €1,5 through experimental measurements, costing €100.000 million. (100 staff days at €1.000). Using a modelling simulation, this cost would be only €13.600, saving €86.400 for a single distillation process, of which Lonza designs in excess of five per year. Given these figures, it is clear that HPC cloud-based simulation has considerable benefits for Lonza. 36 37 EXPERIMENT #415 HIGHLIGHTS Industry Sector: Casting HPC Cloud-based design of Country: Italy copper-alloy moulds Software used: Elmer FEM ORGANIZATIONS Founded in 1959, IMR is an SME which designs and manufactures foundry equipment for brass alloys and bronze; the company also dedicated consulting services for designing THE SOLUTION moulds and other manufacturing processes. It is essential to ensure a laminar flow of liquid metal through the mould A computer model was developed, based on an open-source software package, which would and a uniform cooling gradient, in order to avoid cracks simulate the flow of copper alloys. This model produces a simulation of the filling of a mould and defects in the final piece. The success of this depends by the molten copper alloy at low pressure, and the reliability of the results were validated by on the design of the mould; the quality of which usually comparison with physical tests. Using a Cloud-based HPC system, the time for simulations depends on the experience of mould makers. Currently, the mould is often modified several could be reduced from one day to 3 hours, which is a more acceptable time frame. times before committing to production. There are several commercial packages for the simulation of casting processes, but they have not been widely used in this sector. End User HPC Provider & Expert BUSINESS IMPACT BENEFITS www.imr.it www.arctur.si Although IMR had used simulation in the past, this was • With HPC the their first experience of using an HPC Cloud-based system development time THE CHALLENGE for simulations. The use of HPC reduced the number of is faster by 20% and changes required to the mould prototype during its design. with that 20% of the In the past, IMR has tried to conduct simulations with commercial software, but never with This reduced by 20% the time for development of the testing costs are HPC. The costs associated with purchasing dedicated casting software, the necessary mould and saved 20% of the cost of testing the mould saved. hardware, and the training required are not viable for SMEs like IMR, whose normal activities before mass production can begin. do not require such an investment. They therefore wanted investigate how Cloud-based • Cloud-based HPC The average cost for the design and testing of a set simulations could improve the time to market and productivity. simulation method of moulds for a new product based on the traditional, saves about € experience-based, trial-and-error method is currently 8.000 per set and 3 about €41.000. The use of a Cloud-based HPC simulation weeks of testing and method saves about € 8.000 per set and 3 weeks of modifications. testing and modifications. This includes all costs, such • Fast and accurate as set-up times and computing costs, not just design design ability without and testing. IMR has about 8 sets of moulds per year the wastage of to develop, so this represents a total annual saving materials inherent of €64.000. It aslo reduces the time to market of the in the trial-and-error moulds, so IMR may be able to produce more moulds per method amounts to year, increasing their profitability. The ability to quickly annual savings of and accurately design a mould, without the wastage of €64.000. materials inherent in the trial-and-error method, will further increase their competitiveness in their business sector. Furthermore these simulations create the opportunity for IMR to offer its customers a new, HPC-based design service. 38 39 EXPERIMENT #417 HIGHLIGHTS Industry Sector: Automotive HPC Cloud-based simulation of Country: Sweden sports-car aerodynamics Software used: iconCFD ORGANIZATIONS Koenigsegg is a Swedish SME. Established in 1994, they are a leading designer and manufacturer of high-performance sports cars. In the development of these cars, intensive CFD simulations THE SOLUTION are carried out to reduce the cost of wind tunnel testing. Models are created which accurately replicate real-life cars, Before the start of this case study, Koenigsegg had only limited in-house computing including all geometric details - such as rotating wheels resources and no experience in HPC-based CFD. Access to Cloud-based HPC allows and integrated components (including heat exchangers, simulations to run on a pay-per-use basis, which makes access to powerful computing fans and condensers). These models can be very large resources financially viable for an SME, as well as reducing hardware and maintenance costs. and complex, so the use of HPC can make a significant difference in how accurately a The use of ICON simulation software on a Cloud-based HPC system has enabled Koenigsegg simulation can be conducted and how long it takes to complete. This case study was the first to reduce or even, in some circumstances, avoid wind tunnel testing. time that Koenigsegg had used HPC in the design of a hypercar, the One:1. End User ISV HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.koenigsegg.com www.iconcfd.com www.dit.hua.gr www.cineca.it In this case study, 100% of the aerodynamic development • 5% savings in of the Koenigsegg One:1 has been conducted using operational costs. THE CHALLENGE HPC-based CFD simulations. Koenigsegg were able to • 30% savings in design completely eliminate expensive wind tunnel tests thanks There are two ways to determine how air flows over a car: wind tunnel testing and simulating costs. to the high fidelity of HPC simulations, allowing them to air flow using Computational Fluid Dynamics (CFD). Wind tunnel tests are expensive and have confidence that the simulation results would transfer • 50% reduction in wind time-consuming, so are only used infrequently, but simulations can be used throughout the into real-world gains. In less than eight months, hundreds tunnel and physical design phase. The challenge here was to show how Cloud-based HPC resources can reduce of simulations to test various configurations have been testing. the amount of required wind tunnel testing and show the cost-effectiveness of this approach. carried out. The results were an impressive 250% increase in down-force with only a 15% increase in drag at 250km/h and with a 50% higher down-force at 440km/h, the vehicle’s maximum speed. Tests have shown that HPC-based simulation, supported by external software and expertise, led to a return on investment in less than three months for the production of a new car configuration. Significant costs can be saved and transferred to other critical parts of the development and production process. The benefits obtainable by the use of the Fortissimo HPC-Cloud can be quantified as a 5% saving in operational costs, a 30% saving in design costs, a reduction of 50% in wind tunnel and physical testing, a 60% saving in prototyping costs, and a 30% shortening of the time to market. Furthermore, savings in development were about €90.000 per year on the design process, a 1,5% reduction in overall development costs. These calculations take account of a computing cost on the Cloud-based HPC system of around €100.000. 40 41 EXPERIMENT #418 HIGHLIGHTS Industry Sector: Turbomachinery HPC Cloud-based design of Country: Italy centrifugal pumps Software used: ANSYS CFD ORGANIZATIONS Founded in 1984, EnginSoft is a consulting SME operating in the field of computer-aided engineering, virtual prototyping and advanced simulation, including computational mechanics and THE SOLUTION fluid dynamics, numerical crash testing, and environmental engineering. EnginSoft has around 160 employees, 6 sites A simulation model has been implemented on a Cloud-based HPC system for a centrifugal in Italy and 5 branch offices in Europe. In this case study, pump using a commercially available software package. This represents an attractive Enginsoft addressed the design of centrifugal pumps using solution in terms of cost, effectiveness and relevance for those SMEs which do not have the advanced HPC-based simulation. Centrifugal pumps are resources to perform the necessary simulations on their own. The benefits of simulation widely used in many industrial applications, from oil & gas to water treatment, automotive using Cloud-based HPC system has been demonstrated through a series of experimental and home appliances. Such devices may be required to operate over a wide flow range and runs. the prediction of operating characteristic curves is essential for a designer. End User Computer Centre & HPC Expert BUSINESS IMPACT BENEFITS www.enginsoft.com www.cineca.it The test runs have shown that the use of HPC-based • Timeframe for a new simulation using a combination of Cloud infrastructure hydro plant reduced THE CHALLENGE and external expertise results in a return on investment from 2-3 years to 6 in less than six months. The simulation of centrifugal months. Numerical simulation has become an important and common tool for pump designers. Many pumps is not easy due to a number of challenges: complex tasks can be solved faster and cheaper numerically than by means of experiments and, • Customized workflow geometries, unsteady flows, turbulence, secondary most important, the complex internal flows in water pump impellers can be predicted well. for pump design flows, flow separation, boundary layers and so on. Performing the numerical simulations required would not be possible for a typical SME, which optimization. Simulating these requires a high-fidelity CFD model, very has neither the technical expertise nor the computing resources to carry out such a simulation. fine computational grids and the analysis of transient • Increase of market flows. The design and optimisation of a single pump can share by 1% with thus take 2 to 3 years to complete. With the dvantage resultant profit of of access to Cloud-based HPC, this can be reduced to 6 €100.000. months. The improved design process using simulations can give Enginsoft a significant commercial advantage, allowing them to design and test up to six different pump designs in the time it would take a competitor to perfect a single design. It can also offer these pumps at an extremely competitive cost thanks to the money saved: not only do EnginSoft not need to purchase and maintain costly dedicated computational resources, the increased efficiency of HPC-based simulations saves time, and therefore money, throughout the design process. Due to this improvement in the design process, Enginsoft expects to increase its market share by at least 1% with a resultant profit of €100.000 per year. 42 43 EXPERIMENT #419 HIGHLIGHTS Industry Sector: Civil Engineering HPC Cloud-based simulation Country: Canada of drifting snow Software used: OpenFOAM ORGANIZATIONS Founded in 2005, Binkz is an SME whose business is consultancy, specialising in single and multiphase flows. Binkz provides state-of-the-art consultancy services using Computational Fluid THE SOLUTION Dynamics (CFD) for applications such as wind engineering, process technology and aircraft icing. Binkz has developed This case study has shown that the simulation of drifting snow using snowFoam is feasible the CFD program snowFoam. This program allows an using a Cloud-based HPC system. Analysis can be completed within a few weeks, which accurate assessment of snow loads on buildings. The fits well with the timescales for the design of buildings. In the solution developed here, the maximum snow load that may be accumulated on a building user has access to computing resources, storage, and visualization facilities from a desktop rooftop is an essential parameter in assessing the safety and stability of a building. When environment. The required computational resources needed and their costs are appropriate compared to existing alternatives, snowFoam is more accurate, more reliable and more considering those for the overall design of a building. versatile, but it requires the computational resources that only an HPC system can provide. End User & Code Developer HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.binkz.ca www.vortech.nl www.surfsara.nl Every year, roof collapses due to accumulated and drifting • The simulation of snow are responsible for losses of hundreds of millions drifting snow requires THE CHALLENGE of Euros as well as bodily injuries and loss of life. This is significant compute a problem for all countries in Northern Europe and more resources, which can The overall challenge was to study the commercial feasibility of a CFD consultation service generally in the Northern hemisphere. The maximum snow only be provided by a for assessing snow loads on buildings, employing snowFoam, to civil engineering firms on load that may be accumulated on a building rooftop is an large HPC system. the Fortissimo HPC Cloud infrastructure. For the viability of such a consultancy service, it is essential parameter in assessing the safety and stability essential that both the simulation time and the cost of the computation are acceptable within of a building. It is, however, hard to predict the maximum the framework of a typical CFD consultation project. snow load when designing a new building. This leads to a costly over-design of the structure, which could be avoided if the snow load could be predicted with sufficient accuracy. The simulation of drifting snow requires significant compute resources, which can only be provided by a large HPC system. A typical simulation of drifting snow takes 50.000 CPU hours. This equates to 150 CPUs for 14 days. Furthermore, at a cost of €0,2 per CPU hour, this represents a cost of €10.000. A small consultancy at Binkz would not be able to afford the capital cost of a system containing 150 CPUs, neither could it use a smaller system in-house because the computation time would be much longer than the target two weeks. Even if Binkz were to buy a suitable system, then it would only be used for a fraction of the time and its overall costs would be much greater than the use of a Cloud-based system. Consequently, there is a clear benefit for Binkz in the use of a Cloud-based HPC system. 44 45 EXPERIMENT #420 HIGHLIGHTS Industry Sector: Chemical Engineering HPC Cloud-based molecular Country: Netherlands modelling Software used: ReaxFF ORGANIZATIONS The Albemarle Corporation is a global leader in the development, manufacture, and distribution of highly engineered speciality chemicals. It serves customers in approximately 100 THE SOLUTION countries in a wide range of sectors, including petroleum refining, automotive, transportation, pharmaceuticals, Molecular modelling is a proven powerful tool, providing key information for the design of and food safety. Albemarle already uses HPC in the new chemicals and materials. However, accurate molecular modelling requires significant development of its products, however, it wants to improve computing power that even an SME with experience in HPC, such as Albemarle, would not its capability in this area through the use of CPU-GPU have access to. The solution involved not only the porting of the simulation code for the hybrid HPC platforms which offer significant benefits in terms of price-performance and target computer system, but also the development of a simple user interface to prepare the power-performance. In order to take advantage of this, the simulation codes in use by models and their submission to the HPC system. Albemarle need some reprogramming. End User ISV HPC Expert & Service Provider BUSINESS IMPACT BENEFITS www.albemarle.com www.scm.com www.surfsara.nl The software for modelling large-scale molecular systems • The use of a Cloud- has applications in sectors such as electronics, organic based HPC system THE CHALLENGE chemistry, food, paints, dyes, adhesives and alloys and on a pay-per-use ceramics for the aerospace industry. Modelling these basis is up to 50% The challenge in this case was to port an existing simulation code for molecular modelling systems is complex, and in this case required the use cheaper than owning so that it would run on a hybrid HPC platform. To demonstrate the successful porting of of a CPU-GPU hybrid system to provide the necessary and maintaining a this code and the benefits of using a hybrid HPC system, a test case was chosen from the computational power. This case study demonstrated sufficiently powerful petroleum refining sector which involved the use of catalysts in the removal of sulphur from a successful port of a molecular modelling software in-house system, vehicle fuels. package to a hybrid HPC system, with resultant cost representing a yearly benefits. It was determined that, in this case, the annual saving of €38.000. costs for the use of a Cloud-based HPC system on a • Albemarle has pay-per-use basis were approximately half that of owning allocated a and maintaining a sufficiently powerful in-house system, significant budget representing a yearly saving of €38.000. The ability to for Cloud-based HPC access powerful computing resources on a pay-per-use computing for its next basis offers significant flexibility to SMEs who would not business year. be able to access HPC in other ways. This allows them to develop and refine their product lines more efficiently, giving them a competitive advantage in their sectors. The results of this case study will benefit many SMEs, not just Albemarle, as the code developed in this case study can be offered to other companies wishing to perform detailed molecular simulations. As a result of this case study, Albemarle has allocated a significant budget for Cloud-based HPC computing for its next business year. 46 47 EXPERIMENT #501 HIGHLIGHTS Industry Sector: Renewable Energies Cloud-based map interpolation Country: Spain for civil engineering Software used: In house code ORGANIZATIONS Sisener Ingenieros, a Spanish SME, is an engineering company whose main market is in energy projects, especially in the field of renewables. Projects undertaken by Sisener need accurate THE SOLUTION topographical information not only to determine the earthworks to be performed (which depends on intrinsic Ingecon has ported its software to run on an HPC system via a PC-based GUI. The use of information, such as levels, slopes and volumes) but also HPC has reduced the processing times for wind farms to a few minutes and for other projects for the overall design of the installation. Ingeniería y Control requiring greater resolution to a few hours. For the software vendor this is a scalable way to Electrónico (Ingecon), a Spanish SME, develops and sells offer a new service to customers. Civil engineering companies save time and money because software for managing cartographic data. The objective of this experiment is to demonstrate how data acquisition times are reduced and expensive data gathering can be replaced by cheaper, Cloud-based HPC can be used to convert and manage cartographic data across a range of formats quicker numerical interpolation. and resolutions in order to improve the design process of wind farms and to reduce overall costs. End User HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.sisener.com www.ingenieriaycontrol.es www.cesga.es Ingecon sees this as a major opportunity to provide a new • Reduction in the cost service not only in the design of wind farms, but also in of map data. THE CHALLENGE other application areas where cartographic data is used. • Cost of typical design For a company like Sisener, the costs of a yearly software Civil engineering projects need accurate cartographic data, requiring a map resolution of 1 can be reduced from licence and a powerful enough computer to perform metre or finer. Maps at this resolution are often unavailable, but can be created by applying €4.000 to €2.000 the necessary computations are prohibitive. However, a Kriging interpolation to existing maps. However, this is numerically intensive and typically taking in account staff a pay-per-use service where computing costs and requires HPC resources which are unavailable to SMEs. The challenge here was to create a effort and computing software licensing are available, would be a very attractive solution using Ingecon’s existing software that would be viable for Sisener. costs. proposition for all parties. • Annual savings up to The major benefit for Sisener is a reduction in the cost of €50.000. map data. Through the use of interpolation, as opposed to buying raw data, the cost of a typical design can be • New potential reduced from €4.000 to €2.000 taking staff effort and customers for HPC computing costs into account. A company like Sisener will Expert from this new be involved in around 25 such projects per year, so annual service. savings amount to €50.000. Ingecon already has two further companies testing this solution and expects to engage 10 companies by the end of 2016. Based on data from Eurostat, there are over 350.000 potential customers across the EU for the proposed service from Ingecon of which 40.000 are based in Spain. Ingecon estimates a yearly income from this service at €200 per customer with 200 customers by 2018 and a 1% share of the European market (3.500 customers and a revenue of €700.000) by 2020. 48 49 EXPERIMENT #502 HIGHLIGHTS Industry Sector: Electrical Engineering Advanced simulation of Country: France electromagnetic effects Software used: Gramat ORGANIZATIONS SEEMI is a French SME which develops solutions for product packaging and transport. The devices it builds are often found in environments with significant electromagnetic fields, which can lead THE SOLUTION to serious malfunctions in equipment. Electrical devices play a major role in all types of automated and embedded systems. The solution developed here provides seamless access to their code, running on an HPC Cables, both shielded and non-shielded, have become a machine, from a PC. Cloud-based HPC has dramatically reduced the computation time for major issue in terms of safety, weight, performance, power complex electromagnetic simulations from hours to seconds, and is now an essential tool consumption, cost and reliability. It is essential to verify during for simulating the complex electromagnetic behaviour of equipment. Algo’Tech now offers a the design stage that cables are not susceptible to external electromagnetic effects, and to shield pay-per-use electromagnetic-design service to its customers which enables them to carry out from any potential interference. Simulation has become mandatory in making such decisions. large-scale simulations as needed. The customers for products designed by SEEMI are major national and international groups. End User HPC Expert ISV HPC Provider BUSINESS IMPACT BENEFITS www.seemi.com www.inria.fr www.algotech-informatique.com www.bull.com The cost-benefits of simulating the electromagnetic • For the end behaviour of devices depends on where they are deployed. users, the cost of THE CHALLENGE For example, an electromagnetic issue in manufacturing electromagnetic equipment has cost implications, but may present no issues reduces from Algo’Tech has developed an electromagnetic simulator running on a PC to simulate small and danger to life. In contrast, unwanted electromagnetic hundreds to tens of medium-sized problems for customers. This approach allows SMEs, such as SEEMI, to reduce effects in a vehicle may constitute a significant danger. In thousands €. the design time and costs of electrical networks used in the packaging and other sectors. For the former case, the value of a simulation can be estimated larger installations, computing on a PC becomes too time-consuming to meet user requirements, • Due to the new at 1% of the cost of the risk. In the latter case, the value as thousands of calculations are needed to cover a range of common frequencies. service offered an of a simulation can be much greater because the cost increase of 10% in of the risk is much higher. The market price for a small-licence revenue is scale simulation of the electromagnetic behaviour of a expected in Algo’Tech. device is around €240. This cost can be broken down as follows: €200 for Algo’Tech to cover licence fees; and €40 • Due to the new HPC- to the computer centre for the cost of cycles. For a larger based simulations simulation the market price is €1.750: €1.500 in licence an annual growth in fees; and €250 for the cost of cycles. revenue of around 8% is expected. For the end users, the cost of electromagnetic issues, detected during the installation phase on customer premises, would be between tens to hundreds of thousands €. This compares favourably with the corresponding costs and benefits of simulation. As a result of the new service it can offer, Algo’Tech expects to see an increase of 10% in licence revenue and for HPC-based simulations to constitute 20% of its overall business. Overall, it expects an annual growth in its revenue of around 8% as a result of its new HPC-based, electromagnetic-simulation. 50 51 EXPERIMENT #503 HIGHLIGHTS Industry Sector: Casting Advanced simulation of Country: Spain metal casting Software used: Click2Cast ORGANIZATIONS Fundiciones de Roda is a Spanish SME specialising in the casting of grey and ductile cast iron. To save time and money, foundries need advanced simulation to detect metal casting defects as THE SOLUTION soon as possible, but, like many other SMEs in this sector, Fundiciones de Roda finds the advanced simulation of its Click2Cast has been ported to a High Performance Computer system available via a Cloud casting process prohibitively expensive because of the costs infrastructure. Access to this application has been made available as a pay-per-use service of software licences and of computer hardware. Quantech which enables even inexperienced users access to advanced simulation via a simple click-is an ISV which develops and markets the software package based interface. Click2Cast tackles 90% of all casting techniques including High Pressure Die Click2Cast, which simulates the casting of a range of metals including aluminium, steel, brass Casting, Gravity Casting, Low pressure Die Casting and Tilt Pouring. and copper. The software is able to simulate a range of components such as automobile and aerospace components, hydraulic valves, turbine disks, impellers and flanges. End User HPC Expert ISV HPC Provider BUSINESS IMPACT BENEFITS www.fundicionesderoda.es www.cimne.com www.quantech.es www.cesga.es The use of Click2Cast simulation service supported by • Design time can be an HPC Cloud enables a foundry to determine the most reduced by up to 60% THE CHALLENGE efficient casting technique quickly and optimise its - from a week to a configuration. In this way, the weight of casting systems couple of days. Many SMEs in the casting industry are reluctant to use advanced simulation software can be reduced as much as possible. In particular, the use because of the high costs of software licences and of access High Performance hardware. • Reduction of end of a service that simulates the completed metal casting Here, a service was developed based on Quantech’s Click2Cast package which offers SMEs user’s annual design processes can significantly save time and money in the such as Fundiciones de Roda an affordable, pay-per-use option for simulating casting costs by €3.200. development of new types of moulds, because defects in processes accurately and quickly and without the need for detailed specialist knowledge. the casting process can be detected at design time and • Increased annual before expensive prototyping. revenue by €20.000 as a result of A pay-per-use service such as that now offered is a very increased in attractive option for SMEs in this sector. Through the productivity. use of an HPC Cloud-based service, design times can be significantly reduced by up to 60% - from a week to • New potential a couple of days. Through the use of the Click2Cast customers and the service Fundiciones de Roda can reduce its annual design business growth costs by €3.200. Furthermore, because of an increase in estimated up to 40%. productivity, it expects to increase its annual revenue by €20.000. Quantech now offers the only casting design and simulation pay-per-use service in the market. Over the next 5 years, the growth in its business is estimated to be 40%, with a total of approximately 500 new customers by the third year. The additional profit for Quantech per customer is estimated to be €1.000 per annum. 52 53 EXPERIMENT #505 HIGHLIGHTS Industry Sector: Energy Simulation of airflow in concentric Country: Spain chimneys Software used: OpenFOAM ORGANIZATIONS DINAK is a Spanish SME that specialises in designing, manufacturing and installing domestic and industrial chimneys and ventilation systems. In order to improve its capabilities and be THE SOLUTION competitive in a global market, DINAK needs to optimize its chimney designs to reduce CO2 emissions and be more A reliable HPC-based simulation, based around the open-source Open Foam package, was energy efficient. The design process involves many variables, implemented to allow DINAK to shorten the design process down to a single week. A user so physical testing is not feasible on the grounds of cost and interface provides the ability to explore the design space in a systematic way, improving the time. DINAK thus needs a simple and powerful tool to analyse quality of the resulting designs. This experiment, successfully validated against commercial and test the design of exhaust chimneys, help them to improve the design of concentric chimneys, software and experimental data, has demonstrated that the design of concentric chimneys and gain understanding of the exhaust processes. To reduce design times to reasonable levels, using HPC-based simulation resources is feasible and accurate. HPC has become an essential component in the development of such a design tool. End User HPC Expert Application Expert HPC Provider BUSINESS IMPACT BENEFITS www.dinak.com www.gac.udc.es www.aimen.es www.cesga.es HPC simulation has enabled DINAK to accelerate and • HPC simulation optimise the design of concentric chimneys, allowing them reduced design and THE CHALLENGE to increase their competitiveness and enter new markets. In testing costs, which the past, the first company to market with a new chimney resulted in savings In a building with stoves, chimneys need to be installed to ensure proper ventilation, and design could expect to increase their market share by up to over €12.250 for end poorly designed chimneys can cause incomplete combustion. Reducing emissions requires 10%. Currently, DINAK develops around 3 to 4 new products user. designing new, optimised chimneys, and simulation plays an important role in the design per year, and this advantage would increase its turnover by process. There are many complex variables involved in the optimization of the design of a • Optimized design approximately €100.000. Before this experiment, DINAK chimney, and physical tests are generally too expensive and time-consuming to conduct. enables DINAK required approximately a month for the design and testing products to pass CE of a new chimney. The cost of a single chimney design Mark tests with a was €13.400 - 3 specialist engineers for 1 month (€9.300), zero-failure rate. 1 craftsman for 1 month (€2.100) and prototype costs of €2.000. HPC simulation reduced this time to 1 week, • Potentialities to and costs reduce to €1.140 - 1 week of an engineer’s time increase sales of (€900) and computing costs of €240. This gives DINAK computer cycles by a saving of over €12.250 per design, and the optimised around 10.000 CPU design enables DINAK products to pass CE Mark tests h/year for CESGA. with a zero failure rate. Previously there were 1 to 2 failures per year resulting in additional operating costs of around €6.000. Based on experience gained in this experiment, UDC is planning to offer, via the Fortissimo Marketplace, a consultancy service to develop web interfaces for Cloud-based applications. AIMEN will be able to offer consultancy services in the Marketplace based on OpenSource modelling. CESGA will increase its sales of computer cycles by around 10.000 CPU hours per year. 54 55 EXPERIMENT #506 HIGHLIGHTS Industry Sector: Automotive High-Performance gear analyzer Country: Italy Software used: MSC Marc ORGANIZATIONS VE&D is an SME which has been working in the field of automotive engineering for over 50 years, providing design services to its industrial partners. In particular, it has been involved THE SOLUTION in designing gearboxes for large companies such as Piaggio. Italy has a large number of gear manufacturers, The research group of UNIMORE, the ISV partner in this experiment, developed several many of whom are SMEs, operating in the areas of gear computational tools for gear analysis and optimization. It became clear that HPC Cloud production, gear design and gearbox manufacture. The technologies would be needed to create a tool for design engineers with an appropriate objective of this experiment was to develop an HPC Cloud-time to solution. The HPGA software appears as a simple user interface with standard gear based High Performance Gear Analyzer (HPGA) using advanced software and High datasheets, menus for the creation of gear geometry and checking for data coherence, Performance Computing resources which would allow SMEs such as VE&D to benefit from performance analysis and optimization, and the presentation of results. the most advanced methods for gear analysis and design. End User HPC Expert & Provider Application Expert & ISV BUSINESS IMPACT BENEFITS www.veandd.eu www.cineca.it www.unimore.it The current yearly costs for VE&D to simulate gears in- • Annual savings of house are approximately €160.000: €10.000 for software €17.000 for end user THE CHALLENGE licences; €88.000 for staff costs, computer hardware VE&D as a result costs of between €20.000 to €40.000 and other costs of using HPGA via A typical SME gear designer is generally an expert in gears and traditional design tools. (administration, personnel and security) of €32.000. The Cloud-based HPC. They usually have only a basic knowledge of software for structural analysis, and limited yearly costs for VE&D to simulate gears using HPGA via awareness of specialized applications. The challenge here was to support the design of • Increased market Cloud-based HPC are approximately €143.000: €10.000 for reliable and efficient gears using Cloud-based advanced simulation, without the need for share. the use of HPGA (licence and computing costs included); significant investments in computational resources, expensive software licences and training. €25.000 for computer hardware, €79.000 for staff costs • Increased reputation. and other costs of €28.000. VE&D will see an annual saving of €17.000 as a result of using HPGA via Cloud-based HPC. Given that there are a significant number of SMEs, which design gears, this represents a substantial saving across this sector. A further benefit is that HPGA enables tooth contact analysis, which is more refined than that currently performed by VE&D. The market is going to increasingly require this kind of analysis. Having this capability is expected to increase VE&D’s market share. For the ISV UNIMORE the benefits of HPGA will be twofold. Firstly, there will be a direct increase of contracts related to the design of mechanical transmissions, starting from the actual turnover of about €20.000 per annum in 2017 rising to €100.000 per annum after 5 years. Secondly, the reputation of UNIMORE will rise leading to a subsequent increase in collaborative research projects. 56 57 VE&D’s Core Business Powertrain, E-Drive Systems & Vehicle Dynamics Engineering www.veandd.com EXPERIMENT #507 HIGHLIGHTS Industry Sector: Laser Welding Simulation of laser-based welding Country: Germany in the automobile and machine Software used: tool sectors Abaqus ORGANIZATIONS Lasersystemtechnik Bollinger & Ohr (LBO) is an SME founded in 1999, specialising in laser-welding technology. Many of the components of automobiles and machine tools THE SOLUTION are welded, and these welds can suffer thermally-induced stress during manufacture. This can result in flaws which Lauer & Weiss (L&W), an expert in the development of software solutions for the automobile may affect the durability of components. Traditionally, and machine tool industry, has developed a simulation of laser-welding processes based testing of welds involves the cutting, polishing and on the commercially available ABAQUS package. The accuracy of this methodology has sanding of samples to obtain micro-sections, but this is been positively assessed through the physical inspection of welding samples in cooperation expensive and destroys the piece being inspected. Simulation of the welding process on an with LBO. This confirms that simulated welds accurately model real-world cases and that HPC system would dramatically reduce the time to assess and optimise a particular weld, expensive physical prototypes can be eliminated. thus avoiding expensive prototypes. End User HPC & Application Expert, ISV HPC Provider, Expert & Host Centre BUSINESS IMPACT BENEFITS www.lasersystemtechnik.de www.lauer-weiss.de www.hlrs.de The cost of a physical prototype at LBO is approximately • Reduced computing €3.300, and testing takes around one week. In physical time to less than THE CHALLENGE prototyping, multiple trials are usually needed, costing 24 hours for L&W, in total €13.000 and 3 weeks of time. In comparison, a through the use of Simulations allow companies avoid costly physical prototypes and the speed-up of the computer-based analysis costs L&W €6.000. Results are HPC Cloud resources. development cycle. However, there are few, if any, simulation tools to model welding available within two weeks. Once a running FEM model processes, because they are highly complex and difficult to observe. Accurate simulations • The costs for an HPC- has been created, modifications to certain parameters are require significant computing power due to the non-linear behaviour of materials and the based evaluation can made very quickly, thus a second simulation loop is less highly transient conditions, and SMEs cannot afford the necessary computer hardware. be reduced to about expensive. The in-house cycle and licence fees remains €5.500. at €2.000, but the staff cost reduces to ~€500. Assuming that three simulations runs are necessary to find suitable • Cost reduction of welding parameters, the costs for the evaluation of a weld ~€40.000 per year at L&W using HPC-based simulation are about €11.000, for L&W’s welding which is almost the same as the physical weld trials at trials, through the use LBO with roughly the same development time. However, a of Cloud-based HPC complex calculation of a welding process would block all simulations. licences at L&W for seven days. This means that no other projects could be processed during this time, which is not acceptable. In recent years, LBO has conducted four to five expensive welding trials per year. Using Cloud-based HPC simulations, L&W can generate a significant cost reduction of ~€40.000 per year for these trials. For each new client like LBO, L&W sees a further benefit of €20.000 to €30.000. In Germany alone there are several tens of such companies which gives L&W a significant potential for new business. 58 59 EXPERIMENT #508 HIGHLIGHTS Industry Sector: Pharma Industry Cloud-based simulation of target Country: Germany drug compounds Software used: In house code ORGANIZATIONS Transinsight is a German SME which develops software products in the area of bioinformatics where it analyses high-throughput data. This case study addresses the identification THE SOLUTION of existing drugs to treat illnesses other than those for which they are currently prescribed. This has the potential The use of an HPC-Cloud infrastructure combined with algorithmic improvements to make a significant impact in drug discovery where enabled substantially better computational performance. This was achieved through the the costs of developing new treatments are becoming parallelisation of the algorithms used combined with the more efficient use of memory. prohibitive. The assessment of target compounds requires This resulted in a significant reduction in the time and cost of the evaluation of a single the use of Cloud-based HPC because the search space is so large and complex. A new Cloud-compound. The Cloud-based approach enabled significant computational resources to be based HPC service will be offered by Transinsight to support drug discovery both by SMEs deployed without the need to purchase and maintain expensive hardware. and by larger organisations. End User Domain Expert HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.transinsight.com www.biotec.tu-dresden.de www.dit.hua.gr www.epcc.ed.ac.uk The journey for a drug from invention to market is a long • Reduced cost of drug one. There are many challenges to overcome, and many discovery for end THE CHALLENGE reasons that development of a promising compound may users due to pay-per-fail. The time required to develop a new drug de novo use query service. In the field of drug discovery, there is great interest in investigating unknown drug-target ranges between 10 and 17 years; that is, if it ever makes relationships of existing compounds. Investigating these requires significant computational • 3% increase in profit it. The chance for a new drug to actually make it to market resources. The use of Cloud-based computing can speed up drug development and reduce its for service provider. is only 1:5.000. These slim chances are accompanied costs by uncovering off-target effects, and thus causes of adverse drug reactions, early in the by the high cost for developing a new drug, which may development pipeline. reach an average of US$ 403 million. These rising costs threaten to make the development of new drugs increasingly unaffordable for both companies and patients. Repositioning existing drugs for new diseases could deliver the productivity increases that the industry needs. A prerequisite for drug repurposing is drug promiscuity, a drug’s ability to bind to several targets. Here, a HPC Cloud infrastructure demonstrates the viability of a system for investigating previously unknown binding capabilities of existing compounds. Because existing drugs have already undergone extensive safety and bioavailability studies, the cost and time to market of one of these compounds may be significantly reduced. Transinsight estimates that there are hundreds of potential users of its proposed service. Each user represents a potential profit of €2.000 per annum comprising around 4.000 queries regarding protein matching. For Transinsight this represents a potential increase in profits of around 3% per annum. 60 61 EXPERIMENT #509 HIGHLIGHTS Industry Sector: Camshaft Production HPC Cloud-based analysis for optical Country: Spain control in manufacturing processes Software used: M3 ORGANIZATIONS EPC is a Spanish SME specialising in the manufacture of camshafts. These are critical components in high power engines, where tolerances are very small. For this reason, EPC is always THE SOLUTION working to improve its quality-control process, integrating the latest innovations in hardware and software. The Cloud-based HPC resources were used, enabling new methods of analysis such as large-integration and exploitation of 3D optical scanning systems scale geometry extraction and temporal analysis of the dimensional quality of large batches. for dimensional quality control in manufacture results Moreover, traditional tasks such as reverse engineering are now optimised enabling the in significant benefits in terms of time and dimensional much faster generation of control data for the production line. This allows the generation information generated compared to traditional tactile technologies. The use of Cloud-based of additional information concerning the manufacturing quality at both the part and system HPC has the potential to provide a detailed and quick analysis of manufacturing processes levels. enabling significant improvements in the control of quality in production processes. End User Domain Expert ISV HPC Provider BUSINESS IMPACT BENEFITS www.engpwr.com www.unimetrik.es www.datapixel.com www.cesga.es A typical file size for manufacturing data is around 300 • Reduction in the Megabytes, representing 15 million points. This means that a time needed to THE CHALLENGE single company generates several Terabytes of information extract dimensional in short periods of time to be processed by a service provider information from 5 The challenge of this case study was to develop a service which enables companies, such as Unimetrik. At the same time, software developers minutes to 1 minute. particularly SMEs, to control manufacturing processes with very high accuracy using optical such as Datapixel have to work on the optimisation of their scanning techniques. This requires intensive analysis of the parts to be manufactured using • Unimetrik will processing algorithms to enable quicker access to the 3D digital information. The combination of scanning, measurement and analysis can identify increase its information contained in the data collected. The use of the HPC any problems in the manufacturing process early enough for necessary corrections to be made. services by 30% Cloud-based service enables a reduction in the time needed to and its portfolio extract dimensional information from 5 minutes of international to 1 minute. Due to this time reduction: customers by 20%. • Unimetrik will increase its service business by 30% and its • Datapixel expects an portfolio of international customers by 20%. This is increase of 25% in expected to lead to an increased turnover of €400.000 over new licenses. the next five years due to the commercialisation of these advanced services and the creation of two new positions • EPC expects to within the company. improve the quality of its manufacturing • Datapixel, expects an increase of 25% in new licences, process reducing representing an additional turnover of €750.000 over the next the production of five years, due to the commercialisation of the optimized defective parts to 0%. data-processing software developed in this case study. • The end user, EPC, expects to improve the quality of its manufacturing process reducing the production of defective parts down to 0%. This would result in a cost reduction of €1,5 million over the next five years. 62 63 EXPERIMENT #511 HIGHLIGHTS Industry Sector: Oil & Gas Cloud-based processing of Country: Norway & seismic data United Kingdom Software used: Pre-StackPro ORGANIZATIONS Seismic Image Processing (SIP) and Sharp Reflections are premiere suppliers of geological and geophysical services, with a strong reputation for integrated processing, depth imaging, and THE SOLUTION rock physics. SIP and Sharp Reflections offer a wealth of experience and proprietary technologies which provide Significant changes to the underlying software architecture of Pre-StackPRO have been clients with unique and innovative solutions. SIP is the end made, including fully decoupling it from a specific hardware infrastructure, so end users can user of the Pre-StackPRO software tool, developed by Sharp choose the datacentre and hardware that best fits their needs. With the implementation Reflections. Sharp Reflections is an innovative software of Pre-StackPRO in the HPC Cloud it is now available from any remote location through company bringing fast, full-survey pre-stack computing to the interpreter’s desktop. The application remote desktop connections, which enables new business models and opportunities for Pre-StackPRO harnesses the power of many-core CPUs to deliver visual, real-time affordable collaboration on seismic datasets. processing via the Cloud. Sharp Reflections defines a leading edge in seismic data analysis. End User ISV HPC Expert HPC Provider BUSINESS IMPACT BENEFITS www.seismicimageprocessing.com www.sharpreflections.com www.itwm.fraunhofer.de www.bull.com www.sharpreflections.com A common configuration for in-house seismic processing • Savings, over the by SMEs using Pre-StackPRO comprises a perpetual in-house solution, of THE CHALLENGE software licence costing €87.000 plus 20% annual between €20.000 to maintenance. The software runs on a 2-node in-house €30.000 for each SME As the Oil & Gas industry has had to deal with more and more complex geological targets, system costing €40.000, with annual maintenance costs are expected. high-resolution processing of seismic data and interpretation of results has become of €4.000. Over 4 years, this gives an annual cost of strategically important. Pre-StackPRO addresses this need by taking advantage of innovative • The flexible pay- ~€55.000, regardless of the number of hours of processing and powerful computation and visualization capacities. Oil & Gas SMEs cannot afford HPC per-use approach on the system. These costs are an obstacle to SMEs, and clusters, and therefore do not benefit from the software’s full scaling capabilities. ranging from 1 week this setup has obvious limitations - when the processing to 3 years enables requirements are high, they cannot be met by the in-house larger data-sets to be system, and when they are low, expensive hardware and processed by scaling software are not being used. Based on the outcomes of up the hardware as this experiment, Bull is offering a compute node with Pre-needed. StackPRO installed and licensed as a Cloud-based service. Such nodes can be purchased on a pay-per-use basis • Provision of such a ranging from 1 week to 3 years. Clearly the advantage of service is expected such an offering is that variable workloads can be easily to increase Sharp and cost effectively accommodated. Savings, over the Reflection’s total in-house solution, of between €20.000 to €30.000 for each revenues up to 4 SME can be expected. Furthermore, the flexible pay-per-use million € and 23% by approach enables much larger data-sets to be processed 2022. by scaling up the hardware as needed. It is expected that the provision of such a service will increase Sharp Reflection’s total revenues significantly - about 1 million €, equivalent to 10% of total revenues in 2017, gradually increasing to 4 million € and 23% of total revenues in 2022. 64 65 EXPERIMENT #512 HIGHLIGHTS Industry Sector: Oil & Gas Cloud-based simulation of pipeline Country: Netherlands components for the oil & gas Software used: industry HELYX ORGANIZATIONS The Dynaflow Research Group (DRG) is an SME that has provided engineering consultancy services to globally leading oil & gas companies since 1983. This work often requires multidisciplinary THE SOLUTION simulations encompassing static and dynamic analysis of fluids and mechanical components. To satisfy their advanced ENGYS developed and tested a novel client-server framework for their CFD software HELYX® modelling requirements, DRG relies on the open-source based with the help of NAG and DRG. The new client-server framework allows end users to perform CFD software solution HELYX® developed by the ISV ENGYS®. CFD simulations on remote HPC hardware directly from a desktop Graphical User Interface. In the past, such simulations were mainly performed on The new technology facilitates access to and effective use of remote HPC resources from DRG’s local systems with a small-scale parallel capability. The objective of this experiment was to a local desktop, such as those employed by DRG. ENGYS could also offer this product on a enable DRG to run CFD simulations using HELYX® on a Cloud infrastructure, via a familiar desktop short-term licence as an on-demand service. environment, which would significantly enhance DRG’s simulation capabilities in this area. End User HPC Expert Domain Expert HPC Provider BUSINESS IMPACT BENEFITS www.dynaflow.com www.engys.com www.nag.co.uk www.epcc.ed.ac.uk A simple calculation based on the work in this experiment • Potential savings of showed that the costs to DRG of buying and maintaining €2.000 per simulation THE CHALLENGE a computer cluster in-house for performing these CFD per day for end simulations could be up to 5x the costs of an equivalent users due to faster For many SMEs in the engineering and manufacturing sectors, including DRG, in-house Cloud-based HPC solution. There are also clear benefits simulations and pay-computing hardware is usually insufficient for solving large problems. On-demand Cloud-based in turn-around times for simulations, as well as the per-use model. HPC solutions, combined with open-source software, can offer a more cost-effective alternative. opportunity to perform much larger computations using The main challenge faced by DRG and its partner ENGYS was to create a new methodology to • Capability to perform a Cloud-based HPC platform. The simulations are up perform oil & gas CFD simulations using on-demand Cloud-based HPC solutions. simulations on a to 10 times faster, with potential savings of €2.000 per scale previously not simulation per day (based on standard industry rates). The possible. new client-server approach in HELYX® is also expected to increase sale opportunities for ENGYS by as much as 20%, • Increase of 20% in thanks to the introduction of short-terms licenses for on-business for the demand usage of the software in the Cloud. Furthermore, service provider. the possibility of providing these resources via the Fortissimo Marketplace should lead to increased revenue by exposure to a wider market. EPCC will see an increase in its commercial activities through extended use of its commercial supercomputer platforms. NAG will create case studies based on this experiment to showcase its capability in software engineering and High Performance Computing. Such marketing materials will be distributed to potential customers to attract new HPC consulting business. Current estimates are that this would result in an additional revenue for NAG in excess of €150.000 over the next three years. 66 67 EXPERIMENT #514 HIGHLIGHTS Industry Sector: Environmental Cloud-based environmental Assessment modelling Country: Italy Software used: Aermod ORGANIZATIONS eAmbiente is an SME operating in the area of environmental consulting. It provides its services to architects and designers involved in the design of large factories and industrial plants, characterized by THE SOLUTION significant environmental impacts. eAmbiente’s mission is to reduce these impacts to acceptable levels. Since 2002, This issue has been addressed through the development of a customized platform to eAmbiente has used computer simulation to model emissions integrate environmental software including a single easy-to-use GUI available to potential end and water flow to predict the risk to soil and groundwater. users. This GUI gives one-stop access to Cloud-based HPC resources for the end user. The During this period, environmental modelling has become an use of this solution enables an 80% reduction in both simulation time and time to result with important tool during the planning phase of buildings, factories and public infrastructures. Although consequent cost benefits. Furthermore, the use of open-source software for the simulations the required simulations are well understood, their performance and ease of use are limited. In has resulted in a significant reduction in costs. particular, models may take prohibitively long run-times on conventional computing resources. End User HPC Expert Domain Expert HPC Provider BUSINESS IMPACT BENEFITS www.eambiente.it www.progesi.eu www.t2i.it www.cineca.it The total cost of running a single traditional Environmental • The experiment can Impact Assessment (EIA) on a powerful in-house system generate an increase THE CHALLENGE is ~€30.500. The same assessment, using open-source of 28% in yearly software, Cloud-based HPC resources and reduced staff revenue. The challenge of this Case Study is to overcome the current limitations of environmental effort is ~€6.500, a saving of ~€24.000 per EIA. eAmbiente modelling tools. The main aim is to set up and test an innovative service for SMEs, public • Increase in expects to carry out an additional 4 EIAs yearly over each sector and private stakeholders through a single access point. This includes a Cloud-based competitiveness at of the next five years resulting in an overall cost saving of service that will launch and aggregate the results of different models in parallel and an offline the European level. ~€64.000 per year. Furthermore, eAmbiente expects to service used to evaluate and interact with the results of the simulations. increase its revenue with around €374.000 of additional • 5% increase in business over the next five years, corresponding to about commercial revenues €75.000 per year. eAmbiente has a revenue of about in next two years. €270.000 per year for the EIA sector. The experiment can generate an increase of 28% in that yearly revenue. As a result of the experiment, two new employees over the next five years is expected. Finally an increase in competitiveness at the European level is anticipated since the workflow is easily customizable to European environmental legislation. T2i will develop a brand new EIA-related service starting from scratch. There is no such (digital) service in T2i’s portfolio. T2i estimates a 5% increase in commercial revenues in the next two years corresponding to an expected increase in sales of €5.000 to €10.000 per annum. Finally, this experiment offers a success story for CINECA. It is estimated that this will bring at least one new customer per year, with an approximate 5% increase in revenues from commercial services. 68 69 EXPERIMENT #515 HIGHLIGHTS Industry Sector: Automotive HPC Cloud-based additive Country: Italy manufacturing Software used: RBF Morph ORGANIZATIONS HSL is an SME and a market leader in both Additive Manufacturing and the rapid production of prototypes. In recent years Additive Manufacturing has emerged as a viable mainstream THE SOLUTION production technology. Overcoming technical and bureaucratic obstacles has allowed 3D printing to grow as a cost-effective Numerical grid parameterisation using a mesh morpher avoids the time consuming task option for small and medium scale production together with of mesh generation (that can take up to 70% of the total analysis cost). Access to CFD the ability to produce complex shapes not achievable by simulation through the Fortissimo HPC Cloud allows a speed-up in calculation times reducing standard manufacturing processes. This flexibility in design the time to market and to return on investment. Using a collaborative, interactive, Cloud enables the optimisation of components, a reduction in manufacturing time by almost a third interface helps analysts and clients to work together, increasing customer satisfaction and and a halving of production costs by reducing the waste of materials and energy. building better products more effectively. End User Domain Expert ISV HPC Provider www.hsl-italia.it www.web.uniroma2.it www.rbf-morph.com www.cineca.it BUSINESS IMPACT BENEFITS HSL expects that the new service will see the development • 5% increase of a business department with two to three staff, in revenue for THE CHALLENGE potentially growing to five to six staff after three to four commercial service. years. For UTV it is estimated that thanks to the new Computer-aided engineering (CAE) tools are able to suggest new shapes and accurately • New business service know-how acquired the relevant department will increase predict the behaviour of components, making them a natural choice in the design chain, but for optimisation of its industrial research services. The economic benefit is can be prohibitively expensive for SMEs. This is especially true when dealing with complex airbox components estimated to be between €40.000 and €70.000 per year Computational Fluid Dynamic (CFD) simulations. The goal is to demonstrate the validity of an total revenue growth in a five-year plan. Finally, this case study comprises a optimisation service using CAE by optimising a Lamborghini 12-cylinder airbox. of 16% per year in that success story for CINECA, in the application field of high sector. fidelity CFD, that is estimated to bring in at least two new SME customers per year, with an approximate 5% increase in revenue for commercial services. Having access to the RBF Morph morphing tool combined with CFD analysis powered by HPC opens a wide range of business opportunities. In parallel with existing rapid prototyping services, HSL can now propose to its clients alternative component designs corresponding to appropriate performance indicators. For HSL, offering a shape optimisation service in parallel with its existing core business of rapid prototyping activities, represents an opportunity to establish customer activity over a range of key R&D areas. HSL is ready to offer the developed tool to a range of existing clients in the automotive industry, anticipating for the next two years, a total revenue growth of 16% per year in that sector. 70 71 EXPERIMENT #516 HIGHLIGHTS Industry Sector: Automotive Advanced simulation for metal Country: Spain forming Software used: Stampack ORGANIZATIONS MATRICI is a Spanish SME founded in 1964, specialising in the production and design of complex metal panels in the automobile and aerospace industries. The sheet metal forming industry is THE SOLUTION important in the development of the world’s economy. The use of HPC-based simulation can have a significant financial impact A Cloud-based version of Stampack (StamHPC) has been adapted, developed and validated, on the manufacturers of such panels. However, despite recent and its performance and usability evaluated in an industrial setting. A graphical interface significant advances in computing hardware and software, offers easy and intuitive use in the metal forming industry. This allows any end user, even high-end computer simulation and engineering design tools without HPC expertise, to launch a calculation and get results with an appropriate response are often unaffordable for small companies because of the large capital investment in computing time. Furthermore, an appropriate licensing server has been developed which supports the power required. This experiment demonstrates the benefits of advanced Cloud-based HPC tools in availability of StamHPC on a pay-per-use basis. design and how these can be delivered as a pay-per-use service, affordable by small companies. End User Technology Expert ISV HPC Provider, Expert & Host Centre BUSINESS IMPACT BENEFITS www.matrici.com www.scai.fraunhofer.de www.quantech.es www.cesga.es MATRICI estimates that the use of the Cloud-based • A minimum additional StamHPC in its design processes is worth an additional revenue of €200.000 THE CHALLENGE €200.000 in revenues per year. QUANTECH and CESGA per year is estimated expect around 300 new metal forming SMEs to use by the end user The challenge here was to develop a software solution that allows industry to simulate metal StamHPC over the next 5 years, leading to a potential (through the use forming with higher efficiency and ease of use than possible using today’s state-of-the-art return on investment of €60 million. A typical simulation to of a Cloud-based commercial codes. Stampack, a simulation code available from QUANTECH, models the support the design of a metal forming process takes around StamHPC in its design forming of sheet metal panels. However, the application would need to be modified to be able 120 iterations, each of which needs 20 computing hours processes). to run effectively in a Cloud of HPC resource on a pay-per-use basis. on a standard workstation (8 cores, 32 GB of memory). • Using the StamHPC Such a simulation represents 10 weeks of work involving a solution within an workstation and an engineer. Using the StamHPC solution HPC Cloud makes within an HPC Cloud it is possible to run, at the same time, possible to run several options of the feasibility design concepts, reducing simultaneously the time required to design a prototype. The reduction of several options of time to solution is about 50%, with similar cost reductions. the feasibility design A company like MATRICI performs 200 Feasibility design concepts. studies per year in order to make offers to customers, the reduction of design costs, plus the accuracy of the • The reduction of time new StamHPC software would represent savings around to solution is about 1,5 million €/year (2,5% of current company turnover for 50%, which means an MATRICI). CESGA estimates that it will see, due to the sale overall reduction of of computer cycles as a result of the Cloud-based StamHPC, engineering costs of an additional annual revenue of €45.000 in 2017 growing 50%. to €365.000 in 2021. SCAI will use the success of this experiment as a reference to support the sales of its software products, particularly those relevant to HPC-based Clouds. 72 73 EXPERIMENT #517 HIGHLIGHTS Industry Sector: Building Industry Optimised cutting and bending of Country: Spain steel reinforcement bars using Software used: Cloud-based HPC Optimo ORGANIZATIONS Schnell Software, a Spanish SME, designs specialized CAD-CAM software for companies which cut and bend steel, particularly the cutting of iron bars for reinforced concrete. THE SOLUTION SCHNELL already has optimisation software running on a PC; however, execution take several hours if high levels The optimization software has been adapted to run in a distributed Cloud-based HPC of optimization are required, which is impractical in a infrastructure. Through the use of a graphical user interface it is easy to setup and optimise production environment.The optimisation of cutting and bar cutting and bending optimisations. This hides all the complexity of the Cloud. Launching bending of iron bars is based on complex algorithms which optimisations is as simple as sending some input files, with appropriate parameters and are numerically intensive. Better optimisation would enable SCHNELL’s customers to plan their receiving the results after a manageable processing time. schedules to obtain the best possible combination of orders, minimising wastage in the cutting and bending process, and saving time in the analysis of production orders and raw material requirements. End User & ISV HPC & Application Expert HPC Provider, Expert & Host Centre BUSINESS IMPACT BENEFITS www.schnellsoftware.com www.bifi.es www.cesga.es A large foundry produces 2.000 tonnes of steel bars • Steel waste reduction per month. Through the use of Cloud-based high-level by 2%, which THE CHALLENGE optimisation, it can reduce waste steel by 2% (480 tonnes represents 480 tons per annum). The average price of steel is €500 per tonne, so per year. The challenge facing SCHNELL was to modify its software to run on a Cloud-based HPC this represents a saving of €240.000 per annum. A lower-system rather than a local PC. The objective was to reduce the compute time from hours • Reduction in staff level optimisation offers savings of only around €60.000 down to minutes for even large-scale optimisations, and develop a graphical user interface costs by €30.000. per annum. The higher-level optimisation also results in for users. The modified software would serve as a database portal and clients would be able a €30.000 reduction in staff costs. For small companies, • Over five years, to carry out a real simulation of a cutting process with minimal cost and effort. acquiring an in-house optimization system that guarantees Schnell estimates a reliable cut of high quality requires significant investment a total profit of in hardware and software. Not all small and even large €440.000 from the companies are willing to do this because they are not provision of its Cloud-sure of the benefits. This barrier can be overcome using a based optimisation Cloud-based, pay-per-use approach as companies can test service. and evaluate the service without a large capital investment. A simple analysis of the costs of performing the high-level optimisation in-house on a sufficiently powerful system results in a yearly expenditure of €80.000. The corresponding computing costs of performing the same optimisation in the Cloud are only €7.500 per annum. Over 5 years SCHNELL estimates a total income of ~€480.000 from the provision of its Cloud-based optimisation service to steel foundries, including Cloud-computing and SCHNELL licence costs, resulting in a total profit of ~€440.000. In addition to this, CESGA will increase sales of computer cycles by around 175.000 CPU hours per year. 74 75 EXPERIMENT #519 HIGHLIGHTS Industry Sector: Hydraulic Turbines Cloud-based optimisation of water Country: Italy turbines for power generation Software used: ANSYS CFD ORGANIZATIONS Zeco is an Italian SME in the renewable energy sector. It specialises in the production of different types of water turbine. SMEs like Zeco must develop and innovate their products to THE SOLUTION remain competitive. High-fidelity simulation using CFD has become an essential tool for turbine designers because A CFD-based HPC application has been developed which enables the design of a small hydro it results in better designs for less effort and lower cost. power plant in a very fast and reliable way, compared to current practices. The use of this However, for Zeco and, in general for SMEs, full exploitation application can contribute significantly to savings in time and money in the development of of CFD tools is often not possible as they lack the necessary new water-turbine systems. High-fidelity simulations and the availability of HPC significantly computing power, and the skills to exploit it effectively. The objective here is to demonstrate reduce the development costs of prototypes, so the time to market is also significantly lower. how all the necessary resources can be assembled to give ZECO a one-stop-shop for the simulation of turbines leading to business benefits across the whole value chain. End User HPC Expert ISV HPC Provider BUSINESS IMPACT BENEFITS www.zeco.it www.enginsoft.com www.ansys.com www.cineca.it HPC-based CFD calculations have reduced the design time • Reduction of of a turbine from 1 year to 3 months. As manufacturing the designing time from 1 THE CHALLENGE turbine takes 8 months, the time to market can be reduced year to 3 months. from 20 to 11 months. Without the use of HPC, the Current practice in the design of hydro-power plants is to determine the most suitable • Reduce operational development process could take up to two years, which design in a series of time-consuming experiments. However, SMEs in this sector face stiff costs by €350.000 is no longer a competitive time frame in this sector. Using completion and tight deadlines to sell their turbines in both national and global markets. The per installation over 2 such turbines, a medium-sized hydropower plant costing challenge facing Zeco is to improve its design processes by the use of HPC-based high-years. 1,5 million € can reduce operational costs by €350.000 per fidelity simulations of flow in its turbines through the use of CFD-based tools. installation over two years. Furthermore, the optimization • 1% increase in plant through HPC leads to a 1% increase in plant efficiency, efficiency, 50% with a 50% reduction in the total number of days required reduction in total for maintenance. This means an increase of the revenue number per days of related to energy production of up to 40.000€ per year per maintenance. installation. Due to these improvements, Zeco expects to • Increase by at least increase its market share by at least 5% with an additional 5% in market share. profit of €50.000 per year. As a result of the increased market for advanced simulation using HPC, EnginSoft expects a growth of 10% in business related to the turbo-machinery market sector, which means an additional profit of around €50.000 per year. CINECA estimate potential revenues for the HPC service of €100.000 per year. The workflow developed here is applicable to other sectors as well, so there is a large potential market. CINECA’s target is to acquire two customers for this service per year for the next three years, with an estimated increased revenue of €900.000 and a profit of around €100.000. 76 77 EXPERIMENT #520 HIGHLIGHTS Industry Sector: Pharmaceutical HPC-based prediction of the optical Engineering properties of dyes Country: Italy Software used: Perturbed Matrix Method (PMM) ORGANIZATIONS Scriba Nanotechnologie is an SME active in the fields of smart packaging, health, and machine vision. In particular, Scriba is involved in the development of Time Temperature Integrators THE SOLUTION (TTIs), which are devices that record the thermal history of products. Placed close to perishable products, they can A computational pipeline for the simulation of optical properties of dyes in different be used to monitor the thermal history of packaging and environments has been set up and a Cloud-based web portal has been developed to manage storage conditions. TTIs form an important part of cold the simulations. This portal has been specifically designed for users with no expertise in chain logistics in a world where production is global. TTIs simulation - the user only needs to select the material of interest and the experimental use dyes which may change if there is a change in temperature. This experiment uses conditions to be simulated, and default options are presented to help the user proceed molecular dynamics to simulate dyes used in the manufacture of new and innovative TTIs. through the steps. Simulations reduce the time to customise these devices, thus lowering the overall cost. End User HPC Expert & Provider Domain Expert BUSINESS IMPACT BENEFITS www.scriba-nanotec.com www.cineca.it www.tcpc.univaq.it www.unimore.it Through the use of the advanced simulation package, • Annual savings of Scriba expects to see a saving of €90.000 per simulation around €400.000 per THE CHALLENGE in the first year of use, with further savings of €70.000 in year. subsequent years. This can be broken down as savings The challenge faced by this experiment is to use HPC to simulate candidate dyes, based • Value of the external in personnel costs of €45.000, in infrastructure costs of on molecular dynamics, to be used in the fabrication of innovative TTIs. This requires a consultancy to the €10.000, in material costs (including computing costs) multidisciplinary approach using molecular dynamics-based simulations of dyes. The aim two domain experts of €25.000, and optical readout and further development is to reduce the time required for customisation of TTIs, lowering the overall cost of these is estimated to be of €15.000 in the first year. Scriba estimates that it will products. The goal is to develop a new tool in cold chain logistics. €10.000. be involved in approx. 6 simulations per year resulting in an annual saving of around €400.000. CINECA plans • Development of to use the success of this experiment to target SMEs in virtual tool. the European market who have the need for high-fidelity chemical and molecular dynamics simulations. The target is to acquire two customers of this type per year in each of the following three years. UNIMORE and UNIVAQ expect to see an increase in their consultancy business as a result of this successful experiment. Both partners will be further involved in the enhancement of the software, operating as external consultants. The value of the external consultancy to the two domain experts is estimated to be ~€10.000 per year each. The number of potential users to run simulations explicitly for the development of temperature sensors is about 5 and for the design of optical (bio)sensors about 100. Scriba will make available the developed virtual tool in the Fortissimo Marketplace, subject to appropriate terms and conditions. 78 79 EXPERIMENT #521 HIGHLIGHTS Industry Sector: Elecrtomagnetic Multi-physics simulation of Components Industry high-temperature superconducting Country: Spain devices Software used: FEMPAR ORGANIZATIONS Oxolutia is a technology-based Spanish SME specialising in thin-film oxide architectures. These are deposited by industrial inkjet printers using special inks. Oxolutia has investigated the THE SOLUTION high-temperature superconductor (HTS) space as a potential new market for them to move into. These In order to reduce computation times to reasonable levels, a multiphysics-based HPC materials can be used to construct generators, motors application has been developed which enables the behaviour of HTS devices to be simulated. and superconducting magnets, for applications such as The use of this application can contribute significantly to savings in time and money in power cables, energy storage and magnetic resonance the development of devices adapted to meet customer demands. It enables Oxolutia to imaging (MRI) devices. However, they are complex and difficult to work with, exhibiting strongly understand the performance and characteristics of its products, and to increase their non-linear hysteretic behaviour and time-dependence. This requires a robust, fast and powerful competitiveness by allowing them to enter a new business area. computing environment to obtain solutions consistent with a productive design cycle. End User HPC Expert ISV HPC Provider BUSINESS IMPACT BENEFITS www.oxolutia.com www.icmab.es www.cimne.com www.cesga.es High-fidelity simulations and the availability of HPC can • A new business significantly reduce the development costs of prototypes. turnover of €25.000 THE CHALLENGE In this experiment, this has opened up a new market for over the next year is the SME involved, Oxolutia. Through the use of this service, expected by the end The challenge was to develop a design tool that could calculate the required properties of Oxolutia expects to see an increase in business turnover user, of that €13.000 HTSs - magnetic, electric and current density fields, and mechanical and thermal properties of €25.000 over the next year. This is based on the sales of of net profit. including temperature, stress and strain fields – in a reasonable time frame. This was done HTS devices - production costs, licence fees and computer using FEMPAR, an electromagnetic software package from the ISV CIMNE. The results of the • ISV CIMNE cycles would cost €12.000, giving Oxolutia a net profit of modified FEMPAR package were verified by comparison with other commercial solvers. anticipates a net €13.000 in the first year. Looking over the next 3 years, profit of €54.000 from Oxolutia expects to see an increase in business turnover of related new business €500.000, spending €150.000 on production costs, licence over the next 3 fees and computer cycles - a net profit of €350.000. years due to the CESGA and CIMNE plan to offer a service enabling SMEs provision with CESGA to simulate the electromagnetic and physical behaviour of of the pay-per-use HTS devices. This service will be offered on a pay-per-use simulation service. basis including licence fees and the costs of computing cycles. CIMNE anticipates a net profit of €54.000 from related new business over the next 3 years due to the provision of service. Over this period CESGA expects to increase its provision of CPU cycles to CIMNE by 32.000 core hours. ICMAB, the HPC application expert, expects a profit over this period of €45.000 based on increasing of consultancy and research contracts using the pay-per-use service. 80 81 EXPERIMENT #522 HIGHLIGHTS Industry Sector: Smart Cities HPC-Cloud-based monitoring Country: Switzerland of crowds Software used: LBASense ORGANIZATIONS DFRC is a leading centre for knowledge, research, and development of geospatial data. Its expertise lies in crowd monitoring and smart cities. DFRC delivers powerful end-to-end solutions using data about THE SOLUTION the location and the movement of people using a suite of tools, software platforms and applications. DFRC operates Existing simulation tools have been ported to enable them to run on an HPC system with LBASense, a service which anonymously measures crowd many processors. Detailed tests have been made to verify the correctness, accuracy and behaviour on a city-wide scale. This information is highly stability of the ported algorithm and to determine the speed up of the implementation. The valuable for services such as transportation planning, the outcome of the experiment has been a High Performance combined simulation tool that promotion of tourism and public safety applications. LBASense, which currently operates in reduces the deployment time of LBASense. This allows DFRC to finish a single iteration of the Bern, Zug, Skelefteå, Barcelona, Prague, and Singapore, requires an extensive “fingerprinting” of simulation within hours as opposed to days. the cellular network radio signals in order to calculate the location of mobile phones. End User HPC Provider BUSINESS IMPACT BENEFITS www.dfrc.ch www.hlrs.de The ability to access a cost-effective, one stop-shop, • Reduced deployment Cloud-based HPC service has enabled DFRC to costs for LBASense THE CHALLENGE significantly reduce the deployment costs of LBASense. by ~€25.000. This saving is as a result of the reduction in time to The challenge of this experiment was to improve the process of “fingerprinting” a city, • DFRC is able to offer deployment, allowing DFRC to save the costs of two by offering a near-real-time simulation to determine the best locations at which to take a cheaper, faster and weeks each for two engineers per deployment, which is measurements. Doing this was expected to reduce the time to map a city from weeks to more focused service approximately €25.000. The results show that DFRC is days. To respond to this challenge, the mapping algorithm, which determines the best and hence, be more able to speed up the whole process of mapping a city locations for measurement points, would need to be ported to a Cloud-based HPC system. competitive with (i.e. 1.000 base stations – 10x10 km) by a factor close to respect to cellular the number of processing units available. That is, with 32 operators. processing units it is possible to speed up the computation process by a factor of almost 32. This allows DFRC to offer • Expected growth in a cheaper, faster and more focused service, and hence to DFRC’s business from be more competitive with respect to cellular operators. 20 installations in DFRC expects to gain new market share, especially by 2017 to 200 in 2021, being able to reduce time to market. Based on the overall with an overall cost costs of installing LBASense in a city, and the speed with saving of ~€8 million. which it can be done, DFRC expects to see a growth in its business from 20 installations in 2017 to 200 in 2021 with an overall cost saving of ~€8 million. HLRS, the HPC provider in this experiment, benefits from the experiment through an increased knowledge about commercially relevant scenarios leading to improved offerings for a wider range of customers, which will increase its competitiveness. 82 83 EXPERIMENT #601 HIGHLIGHTS Industry Sector: Automotive Cloud-based design of motorcycle Country: Italy helmets Software used: In house code ORGANIZATIONS The NolanGroup is a Mid-Cap Italian company that are one of the leading manufacturers of motorcycle helmets worldwide. Nolan produces helmets for professional, leisure and racing activities, THE SOLUTION participating in the main international motorcycle competitions. Nolan wants to enhance its helmet design capabilities to The CASCo Multiphysics platform has been ported to an HPC system. This enables detailed reduce costs and time to market through the use of advanced models to be simulated in a simple way by those not expert in the underlying physics. This HPC-based simulation. Moxoff is an Italian SME developing has involved the enhancement and customisation of GUIs, workflows and file formats, and mathematical modelling software, and has been in partnership the development of the interface with an HPC scheduler system. These enhancements have with Nolan since 2010. Moxoff’s projects cover a wide range of applications, including multiphysics achieved the aim of allowing non-experts to perform detailed simulations and to make their modelling, numerical simulations, statistics and big data analytics. CASCo is a dedicated results easily accessible and in a much shorter timescale. multiphysics platform developed by Moxoff to support the simulation of Nolan’s motorcycle helmets. End User Technology & Domain Expert HPC Expert & Provider BUSINESS IMPACT BENEFITS www.nolan.it www.moxoff.com www.cineca.it In this experiment, a powerful simulation platform has • Platform will increase been developed by Moxoff which can be easily configured its turnover by 10%. THE CHALLENGE to provide Simulation-as-a-Service using massive HPC • Reduce software computing resources, even across totally different Nolan have only recently introduced a simulation-based approach using CASCo to simulate development costs by industries. Nolan can either use its own IT resources or, external aerodynamics, thermal effects, acoustics, and impacts of its helmet designs. The €50.000 per year. for time sensitive simulations, use a pay-per-use service choice of computational platform here is critical, as the right choice can reduce compute hosted in the CINECA HPC facilities. Moxoff estimates that • 5% increase in times by orders of magnitude. The aim of this experiment is to enable advanced and fine its exploitation of the SaaS and the platform will increase commercial revenues. detail simulations with feasible runtimes through the use of HPC via the CASCo platform. its turnover by 10% and will significantly contribute to its planned growth for the coming 3 years. Furthermore, through the wider development of this platform, Moxoff will reduce its software development costs by €50.000 per year. Nolan estimates that, for each new helmet developed, the use of HPC simulations will save money in physical prototype testing and final product tuning. This new approach will provide a net saving of €52.000 per helmet once the cost of compute cycles and software licences are deducted. Furthermore, advanced simulations will lead to a 3-month reduction in the current 15-month development cycle which can have important implications for Nolan’s market impact. Finally, this experiment comprises a success story for CINECA, in the application field of integrated multiphysics workflows, that is estimated to bring at least two new SME customers per year, creating an approximate 5% increase in commercial revenues. 84 85 EXPERIMENT #602 HIGHLIGHTS Industry Sector: Manufacturing Simulation of High-Performance Country: composite materials in the United Kingdom automobile industry Software used: In house code ORGANIZATIONS Formtech Composites specialises in the design, engineering and manufacture of lightweight composite structures and components, using carbon, glass, aramid, and other High Performance fibres. THE SOLUTION It collaborates with leading automotive, motorsport, military and aerospace partners to take forward composite research, KIT, the domain expert, working with Fraunhofer SCAI, has developed a prototype simulation engineering, prototyping and serialised manufacture. The platform which allows the user to plan manufacturing processes and predict the final industrial use of carbon-fibre-reinforced plastic (CFRP) is structural performance of a material. An integrated workflow to optimise the design of CFRP being driven by the increasing use of composites to reduce components has been developed, which uses the prototype from KIT and Fraunhofer SCAI. weight in the automotive industry, where cost and performance are major factors. With world-It is accessible through a web based interface and runs on a Cloud-based HPC system at wide tooling budgets running into hundreds of millions of Euros, simulating processes and Gompute, which can easily satisfy the necessary compute requirements. manufacturing times are central to delivering high-quality products at an affordable price. End User ISV Domain Expert HPC Provider BUSINESS IMPACT BENEFITS www.formtech-composites.co.uk www.scai.fraunhofer.de www.fast.kit.edu/lbt www.gompute.com The platform developed here increases the quality of design • Increases quality of composite materials, adds functionality, and reduces of designs, added THE CHALLENGE the engineer’s work load. It shortens simulation times from functionality and days to hours whilst offering more detailed simulations. reduced workload on To make High Performance CFRP economically viable for large scale production, it is essential Using simulation leads to less material usage, due to the engineerś side due to to reduce overall development and production costs. The big advantage of a continuous reduced need for physical prototyping and mechanical developed platform. virtual CAE chain would be the acceleration of the development loops. The objective of this testing. Formtech has already exploited the HPC capacity of experiment was to develop a CAE chain which could reduce the overall development and • Shorter simulation Gompute to meet a project deadline where a 50% reduction production costs through the implementation of the associated development loops. times from days to in compute time was imperative, so using advanced hours. simulation has enabled Formtech to maintain a competitive edge over other companies world-wide. As a result of using • Less material usage advanced simulation, Formtech anticipates an increase in and fewer mechanical revenue per annum of ~€100.000. KIT estimates that, due tests. to time saving through using HPC systems, it will increase its annual revenues by around €30.000. Additionally, annual personnel costs will be reduced by €10.000. Fraunhofer SCAI expects a strongly growing demand for customised and integrated CAE development. Fraunhofer SCAI estimates that, during the next 2 to 3 years, it can increase its regular annual licence revenues by ~€60.000. The outcome of this experiment has given valuable feedback to Gompute. The success of this CAE chain implementation has already brought new business with annual turnovers of over €2.000 for the sale of compute cycles. KIT and Fraunhofer SCAI plan to offer a CFRP simulation service via the Fortissimo Marketplace starting in Q3 2017. 86 87 EXPERIMENT #603 HIGHLIGHTS Industry Sector: Health Cerebral blood-flow simulations Country: Lithuania Software used: In house code ORGANIZATIONS The non-invasive accurate and quick measurement of the intracranial pressure (ICP) is of paramount importance for the diagnosis and treatment of neurological diseases, brain injuries and other THE SOLUTION neuro-pathologies. The world’s first accurate, non-invasive ICP measurement device, based on two-depth ultrasound An internal Carotid/Opthalmic Artery model has been developed, based on MRI imaging, Doppler technology, has been developed by a Lithuanian SME, and used to simulate blood flow. This model has been tested and optimised to run on a Vittamed. The Simula Research Laboratory was founded in multi-processor HPC system. It will be used by Vittamed in the future development of its Norway in 2001. Its main objective is to create knowledge non-invasive ICP measurement technology. Simula has written and tested the necessary about fundamental scientific challenges of genuine value for society. Simula has developed leading-software tools needed to implement the required simulations. These tools enable the efficient edge blood flow models. The use of these models in the simulation of ICP requires the use of HPC. implementation of the model on an HPC system. The implementation of these models on a Cloud-based HPC system is the objective of this experiment. End User HPC & Application Expert Provider & Host Centre BUSINESS IMPACT BENEFITS www.vittamed.com www.simula.no www.cineca.it Vittamed can now reduce their time to market due • Shorter simulation to shorter simulation times. This will also allow them time. THE CHALLENGE to design more accurate products based on blood • Design more flow simulations - Vittamed is targeting entry to the The challenge facing the partners in this experiment was to create mathematical models accurate products ophthalmological market for glaucoma diagnostics, where and develop the necessary software tools to enable simulations of cerebral blood flow in the based on blood flow more precise and accurate ICP measurements are needed. ophthalmic artery. The computational requirements of this made it necessary to adapt the simulations. In glaucoma management, improved measurements software tools to run on an HPC system. The goal was to demonstrate the feasibility and of intraocular and intracranial pressure would enable • Development benefits of such simulations via a pay-per-use Cloud-based HPC solution. the development of innovative screening technologies research projects and treatment methods. Vittamed can gain a significant with industrial, commercial benefit resulting from the global expansion of clinical and scientific the market for ICP diagnostic devices in ophthalmology. communities. The global market for ophthalmic diagnostic equipment is estimated to reach $947M in 2017. The new market niche for SME in ophthalmology will create a commercial opportunity with an estimated potential of more than €100 million per year. Simula Research Laboratory will exploit the results of this experiment by developing research projects with industrial, clinical, and scientific communities across Europe and beyond. The open-source solver developed is an attractive alternative to commercial solvers because there are no associated licence fees. The results of the experiment will support the HPC Cloud provider, CINECA, in offering its services to the biomedical market, enabling SMEs to benefit from HPC-based simulation. 88 89 EXPERIMENT #604 HIGHLIGHTS Industry Sector: Manufacturing Cloud-based simulation of Country: Netherlands complex fluids Software used: RheoCube, SuSi ORGANIZATIONS Ioniqa, a Dutch SME, specialises in the development of MSMs (Magnetic Smart Materials) used in various mechanical, chemical processes and a range of applications including automotive and seismic shock absorbers. THE SOLUTION Ioniqa has developed a PET-recycling process involving MSMs to degrade coloured PET waste into ‘virgin quality’ raw materials, A user interface has been developed within the RheoCube environment, which emulates competing with oil-based PET in both quality and costs. With a “wet lab” familiar to material scientists. It provides a workflow that enables the user to this game changing technology Ioniqa has found a profitable design materials and run experiments on the HPC Cloud backend. Objects such as materials, solution for almost a quarter of all plastic waste in the world. The samples, states and experiments can be intuitively managed and shared with other project modelling of Magnetic Smart Materials and Processes is very demanding involving the simulation of members. Data analysis and visualisation tools have been integrated in the Cloud-based post-the microstructure of complex fluids and its effect on transport and rheology. Electric Ant Lab, a Dutch processing supporting the interpretation of simulation results by the user. SME, has developed a software package called SuSi which allows to perform such simulations. End User ISV HPC Expert & Provider BUSINESS IMPACT BENEFITS www.ioniqa.com www.electricant.com www.surfsara.nl RheoCube as a virtual complex-fluids rheometry solution • EAL estimates a has the proven potential to significantly reduce R&D costs total net profit of THE CHALLENGE and time-to-market for complex-fluids and smart material ~€400.000 over products. With lab researchers as users in mind the total the next five years Electric Ant Lab (EAL) has developed HPC software for the detailed simulation of complex fluids. addressable market has a size of €2 billion. based on income The high-fidelity simulation of an MSM, as needed by Ioniqa, requires a resolution of smaller from consulting As a result of this experiment, RheoCube can be offered than 1μm in space and 0.1μs in time. Simulations of a material sample of 1mm3 and a physical and the provision as a SaaS (Software-as-a-Service), making high-fidelity time span of 1 second with this resolution would take around 10 years on a single CPU-core. of RheoCube as an simulations of complex fluids available to SMEs in an State-of-the-art HPC resources with high-end GPGPU accelerators are required to bring the total SaaS. affordable way, thus giving them a valuable tool for runtime of complex-fluid simulations down to an acceptable timescale of 1 or 2 days. material innovation. Using RheoCube, Ioniqa is able to • SURFsara has a reduce the costs of their MSM development by up to 90%. projected revenue EAL estimates a total net profit of ~€400.000 over the stream through its next five years based on income from consulting and collaboration with providing RheoCube as an SaaS, leading to a yearly profit Electric Ant Lab of of ~€550.000 from 2022 onwards. ~€13.000 in 2017 with a doubling of As resource provider for RheoCube, SURFsara has a revenue each year for projected revenue stream of ~€13.000 in 2017 with a the following 4 years. doubling of revenue each year for the following 4 years. • Both, RheoCube and Both, RheoCube and EAL’s consultancy on HPC EAL’s consultancy will simulations of complex fluids and smart materials will be be offered through offered through the Fortissimo Marketplace. the Fortissimo Marketplace. 90 91 EXPERIMENT #605 HIGHLIGHTS Industry Sector: Maritime HPC Cloud-based seakeeping Country: design Portugal & Spain Software used: SeaFEM, In house code XFire ORGANIZATIONS Seakeeping is the study of a ship or marine structure that is subject to the action of waves. WAVEC is a private non-profit association active in applied research and consulting. VICUSDT THE SOLUTION operates in the shipbuilding and shipping industries, providing hydrodynamic engineering services. It also provides The relevant software packages have been ported to the HPC Cloud-based system and hydrodynamic analysis capabilities to the offshore and integrated into an overall simulation package. An interface between the end user and the HPC marine energy sector, carrying out advanced simulations resources has been implemented, so simulations can be run from a familiar desktop system of structures for the oil and gas industry. COMPASSIS is an whilst using the full capabilities of the HPC system. The simulations have been benchmarked SME ISV which provides simulation software for a variety of different engineering fields including using a model of an off-shore floating wind platform and a model of a stern trawler, which multiphysics simulations and structural and seakeeping analysis. CIMNE, a research organisation, showed a speed-up by a factor of 45 on the HPC system. will also provide some of the software components required for this simulation. End User ISV Domain Expert HPC Provider BUSINESS IMPACT BENEFITS www.wavec.org www.vicusdt.com www.compassis.com www.cimne.com www.cesga.es The use of Cloud-based HPC simulations enables cases • Due to speed up to be analysed far more quickly than was previously in calculation THE CHALLENGE possible. Due to the speed up in calculation time, it also time, previously allows previously infeasible cases to be analysed, and infeasible cases can The challenge addressed in this experiment was demonstrate how advanced computing can solutions devised. The ability to take on seakeeping be analysed and optimise seakeeping design. Seakeeping simulations are large and complex, and so require cases that other companies cannot gives WAVEC and solutions devised. significant compute power to be completed in a reasonable time. To achieve this, existing VICUSDT a significant competitive advantage in the seakeeping software would need to be adapted to run on remote HPC resources. This would • Over the next four sector. HPC Cloud-based simulation of marine structures give WAVEC and VICUSDT a powerful design tool and a significant competitive advantage. years WavEC and has the potential to expand WAVEC’s activities as a VICUSdt expect a consulting company specialized in services for the notable increase in offshore renewable industry and other related industries its total profit based (aquaculture, oil and gas). Over the next four years, from on a revenue due 2017 to 2020, WAVEC expects an increase in its total profit to an increase in its of ~€550.000, based on a revenue of €1,8 million, due to consultancy activities. an increase in its consultancy activities. Over the same four-year period, VICUSDT expects an increase in its total profit of €480.000, based on a revenue of €1,4 million, due to an increase in its consultancy activities. Over the same four year period, COMPASSIS expects an increase in profit of €2 million based on an increase in its licence sales of €4,5 million. A service for seakeeping assessments based on the results of this experiment is planned to be available in the Fortissimo Marketplace. Furthermore, CESGA proposes to offer a service in the Marketplace bases on the benchmarking of key applications and their performance and scalability. 92 93 EXPERIMENT #606 HIGHLIGHTS Industry Sector: Maritime HPC Cloud-based standard strength Country: Spain assessment of commercial ships Software used: RamSeries ORGANIZATIONS ISONAVAL, an SME working in the areas of naval architecture and marine engineering services, is specialized in the generation of analysis and production information for ships, yachts and naval artefacts. THE SOLUTION Merchant ships (such as tankers and container ships) must undergo a standard strength assessment as a statutory requirement, The relevant software packages, including RamSeries, have been ported to an HPC Cloud-and advanced simulation could be useful in standard strength based system and integrated into an overall simulation package. An effective interface assessment calculations. COMPASSIS is an SME ISV which between the end user and the HPC resources has been implemented which integrates the markets simulation software, RamSeries, in different engineering various software components and the HPC system. The simulations have been benchmarked fields including multiphysics simulations and structural analysis. In this experiment the expertise using a model of the full 3D hull structure of a merchant ship. These demonstrated a and software of COMPASSIS will be complemented by FNB-UPC, a University research centre significant speed-up by a factor of 42 through the use of an HPC system. developing innovative simulation tools and implementing them on HPC systems. End User ISV Domain Expert HPC Provider BUSINESS IMPACT BENEFITS www.isonaval.net www.compassis.com www.upc.edu www.cesga.es A standard strength assessment study of a tanker requires • Significantly reduced more than a week on a desktop system. Using RamSeries computation time. THE CHALLENGE with HPC resources reduces this time to less than 6 hours. • Increased market, This significantly reduced compute time fits much better to The challenge addressed in this experiment was to demonstrate the use of advanced for Compassis, by the design cycle of companies. COMPASSIS will increase its simulation in standard strength assessments of merchant ships. Large amounts of computing introducing the use of market by introducing the use of the RamSeries software power are required to complete these simulations in a reasonable time. The objective here was RamSeries software. for the direct strength assessment of a complete ship to adapt standard strength assessment software, RamSeries, to run on remote HPC resources, structure. This assessment requires large computational • Additional annual and demonstrate the benefits of advanced simulation using Cloud-based HPC for this purpose. and data storage resources. COMPASSIS estimates an revenue for end user additional annual revenue of €24.000 in 2017 growing ISONAVAL, due to to €120.000 in 2020, due to the sale of direct strength the sale of direct assessment of complete ship structures using RamSeries strength assessment integrated with a Fortissimo HPC infrastructure. ISONAVAL of complete ship estimates an additional annual revenue of €15.000 in 2017 structures. growing to €60.000 in 2020, also due to the sale of direct strength assessment of complete ship structures. CESGA will offer new HPC added-value services for SMEs such as benchmarking to analyse performance of HPC applications, including multi-core scalability and its dependency on different parameters such as size of the problem and processor frequency. It expects a consequent increase in its HPC services and customers. New alliances with ISV and application experts have been formed during this experiment. Revenues based on 3 benchmarking studies and an annual fee for hosting the ISV software and for infrastructure maintenance will be around €35.000 over a 4-year period. 94 95 EXPERIMENT #607 HIGHLIGHTS Industry Sector: Automotive Cloud-based shape optimisation Country: Italy in manufacturing Software used: Camilo ORGANIZATIONS Automobili Lamborghini is an Italian manufacturer of luxury sports cars and SUVs. Founded in 2006, OPTIMAD is an Italian SME active in shape and topology optimisation in the engineering THE SOLUTION sector. The objective of this experiment was to demonstrate that by combining High Performance Computing (HPC) The SOUTH platform has been developed. This provides an integrated environment with and optimisation software, there is an opportunity to access to optimization software and both in-house and Cloud-based HPC computing change current design practices and bring automatic shape resources. There is a high degree of automation, but users can interact with the optimisation optimisation (ASO) to a wider audience of manufacturing process and investigate partial solutions. Users experienced in simulation but inexperienced SMEs. The goal was to develop and implement an HPC Cloud-based platform that gave the in optimization can easily set up runs. This is an innovative and cost-effective approach to inexperienced user the possibility to optimise automatically the shape of a particular mechanical making shape-optimisation software available to SMEs. system. The aim was to text this solution using practical test cases from the end-user Lamborghini. End User ISV Application Expert HPC Provider BUSINESS IMPACT BENEFITS www.lamborghini.com www.optimad.it www.strath.ac.uk www.cineca.it Lamborghini has its own HPC resources, but having its • Reduction in environment duplicated by a supercomputing-class HPC development time of THE CHALLENGE centre has several benefits. These include a reduction in 40%. development time of 40%, and the ability to dealing with There are several issues associated to the use of automatic shape optimization. It takes • Reduction of peaks in the demand for computer resources by using significant computational power, because many simulations need to be evaluated, and only simulation time from Cloud-based HPC. The use of SOUTH reduces the effort specialised engineers can set up a viable ASO procedure. The challenge facing this experiment 3 months to 10 days. for a typical optimisation from 2,25 person months to 2,0 was therefore to develop a platform which combined Cloud-based HPC resources with in-days and the simulation time from 3 months to less than house computing capabilities, and develop an intuitive user interface for non-expert users. 10 days. Although the number of CPU hours required rises from 120.000 to 160.000, the overall reduction in cost is ~ €18.000. Generally, considering the average design activity in the company, yearly cost savings are expected to range from €140.000 to €200.000. Until now, OPTIMAD has provided automatic shape optimization mainly as a consultancy service with typical cost to the end user of ~ €20.000 to €40.000 per optimisation. Through the SOUTH platform, the same service may be delivered with typical costs of ~ €8.000 to €12.000 whilst keeping the same profit margins. This will permit OPTIMAD to offer a more competitive and attractive service to the market. It is estimated that in the first year through the use of SOUTH in its consultancy service, the increase in revenue for OPTIMAD will be ~ €75.000 with a net profit of ~ €22.000. By deploying SOUTH as a scalable SaaS using Cloud-based HPC resources, OPTIMAD expects further revenues in the first year of operation to be €150.000 with a net profit of €37.000. 96 97 EXPERIMENT #608 HIGHLIGHTS Industry Sector: Health HPC Cloud-based microscopy Country: Poland, Switzerland & Germany Software used: In house code ORGANIZATIONS MicroscopeIT is a Polish technology-based SME company, founded in 2012, which has introduced an internet service called VIRTUM. VIRTUM provides computations in a Software-as-aTHE SOLUTION Service (SaaS) model for use in microscopy. Before this experiment, VIRTUM comprised a client-server architecture The VIRTUM service has been adapted to work on an external Cloud-based HPC using in-house computer resources. UZH is a University infrastructure. Through the use of HPC, the time taken to process test cases has been laboratory making significant use of microscopic data. In reduced from one day to a few hours, a typical reduction of 5 times. A cross-platform, simple this experiment UZH provided a test case in the area of user-interface has also been developed. This supports the visualisation of data and its fluorescent microscopy. OpTecBB is the competence network for optical technologies and management from almost any device. Furthermore, the VIRTUM service provides flexible micro-system technology in the region of Berlin-Brandenburg. licensing models allowing end users to optimize their costs. End User ISV HPC Provider & Expert Hardware Vendor www.interfacegroup.ch www.microscopeit.com www.arctur.si www.opta-tech.com BUSINESS IMPACT BENEFITS www.optecbb.de www.nikon.com Clinical laboratories and R&D departments produce and • High cost savings in analyse huge amounts of microscopic image data. This is the IT infrastructure THE CHALLENGE used in medical diagnosis, to accelerate drug discovery, are possible due and for biomedical research. To support the analysis of to Cloud-based The challenge here was to adapt the VIRTUM service to an HPC Cloud-based infrastructure this data, there is a significant SME-based industry for the processing that is and to test the new service on four numerically intensive test cases. The test cases were: manufacturing of microscopes and spectroscopes serving more cost-effective. analysis of images from remotely controlled fluorescence microscopy; processing data from the biotech and material science industries. In many super-resolution microscopy; management and visualization of spectroscopic data; and • Reduction in staff cases the analysis of this data can be computationally whole slide imaging and analysis in digital pathology and high-throughput screening. costs by 50% through intensive requiring the use of powerful computers. In this increased efficiency. experiment, the expertise and software of MicroscopeIT has been complemented by the end user requirements of • The time taken to UZH and OpTecBB and the HPC expertise of Arctur. This process test cases has greatly enhanced VIRTUM’s availability, capabilities has been reduced by and affordability, particularly to SMEs. VIRTUM-DP has the 5 times, from one day potential to remove obstacles and bottlenecks in current to a few hours. oncological diagnostics. Its main benefit is a significant improvement in clinical diagnosis. This is due to the increased speed and quality of diagnosis, an increased throughput of diagnoses, and more accessible storage of samples. VIRTUM-DP can result in a reduction in staff costs by 50% through increased efficiency, which in the USA alone results in an overall saving per annum of $1,7 billion. Furthermore, there is a significant cost saving in IT infrastructure through the use of Cloud-based processing as in most cases, computer resources available via the Cloud are more cost-effective. 98 99 EXPERIMENT #610 HIGHLIGHTS Industry Sector: Civil Engineering HPC based Delft3D service for Country: Netherlands modelling flooding, morphology, Software used: and water quality Delft3D ORGANIZATIONS Deltares is an independent non-profit institute for applied research in the field of water and subsurface. They have five areas of expertise: flood risk, environment, water and subsoil resources, THE SOLUTION delta infrastructure, and adaptive delta planning. They have developed a modelling suite, Delft3D, to tackle their challenges. The official Deltares version of Delft3D has been installed and certified at the HPC systems One of the SMEs that use Delft3D is HKV Consultants. They of SURFsara. A portal for DaaS has been developed, with a command-line interface with provide consultancy services and conduct research in the fields extended options and flexibility available for experienced users, and a simpler, web-based of safety, drought and flood risk analyses for rivers, coasts, interface available for newer users. Delft3D has been optimized for large-scale HPC, which is estuaries and urban water systems. Its clients include private businesses, governments, research necessary for the coupled modelling involving different Delft3D modules, and for the highly institutes, and regional water authorities worldwide. For their consultancy services and research, detailed modelling increasingly required by SMEs. at the moment, HKV runs Delft3D on local desktop machines and clusters. End User Domain Expert & ISV HPC Expert & Provider BUSINESS IMPACT BENEFITS www.hkvconsultants.com www.deltares.nl www.surfsara.nl DaaS opens new possibilities for SMEs such as HKV, • End user expects because it offers quick, efficient, and flexible access to additional revenues THE CHALLENGE HPC hardware resources for modelling and simulation of about €70.000 using Delft3D. This Fortissimo experiment served as a per annum due to an Delft3D has been open source since 2011 and is used worldwide. It consists of modules starting point for gradual transition for current Delft3D increase in the use of for modelling hydrodynamics, waves, morphology, water quality, and ecology. However, to users and growth model for new Delft3D users worldwide. (certified) Delft3D and improve the quality of its work, HKV needed easy access to a version of Delft3D that could Deltares and HKV themselves are already involved in of about €500.000 scale to available hardware resources in a flexible way. The objective was to develop a pilot further projects using Delft3D running on remote HPC per annum due to an version of Delft3D-as-a-Service (DaaS) that could run for remote use by HKV. systems. Deltares expects additional revenues of about increase in Delft3D - €70.000 per annum due to an increase in the use of related advisory and (certified) Delft3D and of about €500.000 per annum due research projects. to an increase in Delft3D-related advisory and research • With DaaS, the end projects by Deltares itself. With DaaS, Deltares expects an user expects an increase of business opportunities in data and forecasting increase of business services based on Delft3D and projects related to large opportunities in scale computing with Delft3D, for example for dike safety data and forecasting assessments. Due to the new possibilities of DaaS for services. large ensemble modelling and highly detailed modeling, HKV expects additional revenues of about €100.000 • HKV expects and €30.000 per annum, respectively. With DaaS, HKV additional revenues of expects an increase of business opportunities in add-on about €100.000 and services based on Delft3D. Furthermore, DaaS will reduce €30.000 per annum. risks in projects with a lot of Delft3D modelling, with the opportunity to spend more project time on creating added value for the client. 100 101 EXPERIMENT #611 HIGHLIGHTS Industry Sector: Maritime Cloud-based simulation of marine Country: France communication buoys Software used: K-FSI ORGANIZATIONS ALSEAMAR is a French SME, specialised in the design and manufacturing of radio communication systems for both surface ships and submarines. ALSEAMAR wanted to develop a releasable, THE SOLUTION communication buoy for submarines. This experiment aimed to design such a device using HPC-based CFD. K-Epsilon is a K-FSI and FINE™/Marine (Numeca) software has been ported to the BULL eXtreme factory French SME which offers CFD consulting and the development HPC-as-a-Service, to create an appropriate simulation model. An online solution monitor of customised software. In particular K-Epsilon develops has been integrated with the eXtreme factory web interface. Because end users such and markets the software package K-FSI which formed the as ALSEAMAR do not possess the skills to perform simulations of such complex cases, basis for the simulation model used in this experiment. In addition to this, the role of K-Epsilon K-Epsilon makes use of the pay-per-use HPC Cloud on their behalf. Specific workflows have in this experiment was to support the partners in the simulation of a releasable communication been developed to reduce engineering time and to feed into the ALSEAMAR design cycle. buoy, using Cloud-based HPC, and to demonstrate the benefits of advanced CFD simulations. End User Application Expert & ISV HPC Expert & Provider BUSINESS IMPACT BENEFITS www.alseamar-alcen.com www.k-epsilon.com www.bull.com Computational Fluid Dynamics (CFD) has demonstrated • Two new jobs its ability to model detailed scenarios with great positions were THE CHALLENGE robustness. However, some complex applications are created during the still challenging. For such applications, HPC is required Fortissimo project, ALSEAMAR needs to investigate the behaviour of a radio communication buoy released from to perform the required computations in a reasonable and K-Epsilon expects a submarine. Numerical simulation is the only way to do this, because tests involving real timeframe. In this experiment, ALSEAMAR has developed to create another submarines are almost impossible. This type of simulation involves a strong fluid structure a releasable communication buoy for submarines. The three positions in a interaction coupling and the use of large and fine overset meshes. Computations are required resultant revenue is estimated to be €500.000 in 2017 five-year plan. to run for multiple days to converge. To compute these cases in a reasonable timeframe, and €1 million in 2018. ALSEAMAR estimates that 3 jobs simulation software must necessarily be run on an HPC system. • K-Epsilon’s annual will result and its turnover will increase by 20%. These turnover is expected figures are based on the sale of 125 units in 2017 and to increase by 65% 250 in 2018. Through this experiment, ALSEAMAR has over the next 5 years. gained confidence in using CFD simulations to improve their design process, offering an alternative to physical • Bull will earn 20% tests. This experiment enabled K-Epsilon to integrate its of the turnover in-house tool K-FSI on a HPC Cloud, with a pay-per-use from simulations by service. The results have demonstrated the capabilities of K-Epsilon, through the K-Epsilon to address very complex cases with K-FSI. 2 jobs provision of cycles via were created during the Fortissimo project, and K-Epsilon eXtreme Factory. expects to create another 3 positions in a five-year plan. K-Epsilon’s annual turnover is expected to increase by 8% in 2017, and by 65% over the next five years. For Bull, this experiment has been an opportunity to reach new potential customers in the maritime market. Bull will earn 20% of the turnover from simulations by K-Epsilon, through the provision of cycles via eXtreme Factory. 102 103 EXPERIMENT #701 HIGHLIGHTS Industry Sector: Automotive Cloud-based simulation for vehicle Country: Sweden engineering Software used: iconCFD® ORGANIZATIONS Koenigsegg Automotive AB is a Swedish SME whose core business is the development and production of high-performance, high-quality, limited-edition motor vehicles THE SOLUTION – so called hyper-cars. ICON Technology & Process Consulting (ICON) is a UK head-quartered company which The solution developed by ICON is to provide cost-effective and easy-to-use access to High specialises in the delivery of advanced Computation Performance Computing. The ease of use is achieved by providing browser accessible ‘apps’ Fluid Dynamics (CFD) products and related services for which provide the necessary features to set up and analyse simulations. The apps connect industry. to HPC-systems which provide pay-per-use access; capital costs are avoided altogether, as is the need for experienced personnel to run the HPC systems. The result for the end user End User HPC Expert & Provider ISV is a feature-rich interface connected to computing resources capable of rapidly returning simulation results. The entry cost and operational costs are very attractive for small companies. www.koenigsegg.com www.epcc.ed.ac.uk www.iconcfd.com THE CHALLENGE Analysing how air flows around a vehicle is essential to making it faster and more efficient. BUSINESS IMPACT BENEFITS Physical testing is expensive due to the costs of models and test facilities. Computer simulation removes the need for these, and enables the engineers to make changes to the Obtaining an in-house capability for rotating wheel • €100.000 annual design more easily. However, simulating airflow with the required degree of accuracy requires aerodynamic simulation requires an initial investment of saving on design computing power beyond the reach of most designers, mainly due to the capital costs of the €200.000 - €300.000 including hardware, and software costs – saving equipment. licences allowing approximately 50 simulations per year. 30% over previous In addition, an annual maintenance and operational cost solutions. of approximately €100.000 including hardware warranties, electricity, software licences, etc. is required. This is a • High-end simulations prohibitive cost for SMEs. for as little as €900 per simulation. In comparison, a streamlined process accessed on-line and benefitting from on-demand HPC and ISV expertise • No need for SME to could cost as low as €900 per simulation to the end user, purchase expensive therefore making high-end simulation affordable to SMEs. equipment. This can reduce or even eliminate the need for physical • Significant time testing, saving further time and money on models and wind saving as there is tunnels. no need for physical The experience of Koenigsegg is that around €100.000 can testing. be saved annually on design costs by using this technology that was previously only affordable to large companies. For ICON, the apps they have developed will allow them to offer their services to new clients in the automotive SME sector. 104 105 EXPERIMENT #702 HIGHLIGHTS Industry Sector: Oil & Gas Processing Optimising gas and flame detector Country: layouts in hazardous manufacturing United Kingdom and production plants Software used: FLACS Risk, FLACS Cloud ORGANIZATIONS Micropack is a British SME which designs, manufactures and installs its own range of detectors for use in fire and gas safety. It also consults on fire and gas safety design THE SOLUTION and gas dispersion. Gexcon AS is a Norwegian company and the market leader in dispersion, fire and explosion The goal of the Fortissimo experiment has been to provide FLACS as Software as a Service modelling software for hazardous production areas. (SaaS) offering, with the usage of HPC fully integrated in the FLACS GUIs, under the new Gexcon develops the multi-physics modelling application product name “FLACS-Cloud”. The new service enables Gexcon’s customers to tap into FLACS, a simulation tool used worldwide for safety studies. powerful HPC Cloud resources seamlessly from the GUI, so that it appears no different to the user compared to accessing a local cluster. The key difference is that the users only End User HPC Expert & Provider ISV pay for the time that they use, making it much more cost-effective. The access to HPC resources available through the FLACS-Cloud service renders it feasible to apply proper CFD-based optimisation to the placement of gas detection systems. This is expected to result in reduction of the costs for the installation and maintenance of the detection system, and www.micropack.co.uk www.epcc.ed.ac.uk www.gexcon.com lowering the risk of severe accidents in a facility. THE CHALLENGE Gas detectors are commonly installed in manufacturing and production facilities where BUSINESS IMPACT BENEFITS hazardous substances are processed (such as oil and gas installations), to automatically alarm and trigger safety measures in response to leak events. Gexcon, the software provider, Using FLACS-Cloud, Micropack and other users can • Potential savings produces FLACS, a CFD modelling software used for safety studies. To date, the amount of improve safety and reduce the costs of detector into the millions of computing power required for a CFD-based optimisation of gas detector layouts has been installations and ongoing operational expenses in industrial Euro for a typical prohibitive in terms of both costs and time. Gexcon and Micropack joined forces to develop and manufacturing facilities that incur potential risks of gas installation. the next-generation methodology for gas detection system design. leaks, explosions, fires, and other hazards. • The realisation of Optimisation means that fewer detectors are needed to FLACS-Cloud, creating cover the same space. The cost savings of one or more revenues for Gexcon. detectors per sector of the detection system can quickly add up to figures in excess of €100.000. Typical gas • Computational detection systems can easily comprise multiple tens of speedup, up to a detectors, so the savings can run into millions of Euro per factor of 10, using installation. HPC over traditional resources. The use of on-demand resources is attractive to Micropack, as they save tens of thousands of Euro on computing • Improved placement equipment. Using HPC, a project’s computing phase of detectors, therefore can also typically be shortened by up to a factor of ten improving safety. (compared to using in-house resources), which will allow to service more clients, and offer a better client experience. Gexcon will offer FLACS-Cloud via the Fortissimo Marketplace and continue to develop Cloud-based industrial safety software and business models. 106 107 EXPERIMENT #703 HIGHLIGHTS Industry Sector: Automotive Advanced, Cloud-based HPC Country: Italy simulation in the foundry business Software used: Optimus, ITACA, elasticLM ORGANIZATIONS ProService s.r.l. is an SME based in Italy. Founded in 2002, it provides machines, tools and consultancy services for the metal casting industry, focusing on iron foundries. Noesis Solutions is a THE SOLUTION Belgian system integrator and independent software vendor. Its research areas include integrated simulation frameworks, Noesis and ProService collaborated to develop a new adaptive process for controlling the large-scale optimization and High Performance Computing. liquid-iron stage of the foundry-casting process, based on Noesis’ Optimus software and Fraunhofer SCAI is the Institute for Algorithms and Scientific ProService ITACA technologies. The two companies worked together to create a model of Computing of the Fraunhofer Society, based in Germany. the process based on historical foundry data, and provided the required expertise to deploy the solution on a Cloud-based HPC platform. Thermal analysis software is used to generate End User HPC Expert ISV & Domain Expert HPC Provider production data, which is sent to the Fortissimo Cloud-based HPC infrastructure. The correction model is updated by the HPC resources based on the production data, and then returned to the foundry system, where the correction is calculated and applied. www.proservicetech.it www.noesissolutions.com www.scai.fraunhofer.de www.hlrs.de THE CHALLENGE The goal of this experiment was to reduce the scrap rate of the foundry process. A service BUSINESS IMPACT BENEFITS was needed which could predict the behaviour of the current batch of metal to reduce the amount of waste. If it were possible to adjust the process in real time, the foundry could lower For ProService, the tools developed in the experiment mean • Decrease in foundry energy consumption and production costs. Using historical data from previous production it can realistically expect to reduce the scrap rate of a scrap rate by 0,5%, runs was seen as the key to this problem, but this entails data analytics requiring computing foundry customer from a current rate of around 3% to 2,5%. saving up to €500.000 resources far in excess of those available to foundries or consultants like ProService. For a small foundry, this translates to a saving of €50.000 per year. per year – with the number growing to €500.000 for the • Increase in revenue largest facilities. The ability to deliver this kind of saving to a of €236.000 per year customer puts ProService at a significant advantage over its for ProService and competitors. ability to offer service The experiment partners plan to offer the foundry via the Fortissimo optimisation tools as a service, marketed to end users via Marketplace. the Fortissimo Marketplace. This is expected to provide an • Massive potential annual revenue of €236.000 for ProService. saving for EU foundry The EU is one of the largest steel producers in the world, industry. second only to China in terms of output. Around 11% of steel worldwide comes from EU foundries. With increasing demand for cast metal products from developing countries, ensuring the competitiveness of European foundries is key to tapping into a global market. Europe has more than 1.900 foundries. If just 10% of them were to use this system, with each saving the minimum potential of €50.000, this would mean savings for the sector in excess of 9,5 million € per year. The potential of this service is therefore extremely high. 108 109 EXPERIMENT #705 HIGHLIGHTS Industry Sector: Mechanical Engineering HPC-based design of non-circular Country: Italy gears Software used: Optimus, Maple, Matlab ORGANIZATIONS STAM is an Italian engineering and manufacturing firm. Since 2011, it has pioneered the development of Non-Circular Gears, which are used in many sectors such as THE SOLUTION automation & robotics, defence & security, aerospace, transport, and energy. Noesis solutions is a Belgian The partners developed an HPC-based engineering workflow for NCG design. The workflow, system integrator and independent software vendor. CloudGear, allows STAM to optimise the design of an NCG in less than 3 minutes, providing fast feedback concerning potential modifications and improvements. Previously, just the computation phase of the workflow would take 25 minutes, not including set-up and results End User ISV HPC Provider delivery. Due to the near-real-time delivery of results, STAM is able to significant speed up its internal design process. Using Noesis’ Optimus software, the design process was automated and the most computationally expensive part of the process sent to an HPC system. CloudGear will also be offered by Arctur through the Fortissimo Marketplace. www.stamtech.com www.noesissolutions.com www.arctur.si THE CHALLENGE Non-Circular Gears generate a prescribed motion with great precision, regardless of external BUSINESS IMPACT BENEFITS factors. This is a highly desirable feature in advanced transmissions. Designing these gears is a complex process that requires a significant amount of expertise and expensive, Customers of the CloudGear workflow will be able to • Increase in turnover, computationally intensive iterations. In-house computing systems are not powerful enough design better NCG-based solutions at lower costs. This in rising to 20% after 7 to perform these calculations. A cloud-based engineering workflow would significantly turn benefits users of their machinery, plants, prototypes years, for STAM, due shorten the time required to design and deliver an NCG. A key element of this is cost-effective and other products. Five potential customers for to sales of NCG- access to the necessary computing power. CloudGear have already been identified. based solutions and STAM sells a custom NCG solution for approximately software. €15.000 each. The CloudGear approach will allow them • NCGs can be to sell 8 new solutions in the first year, 10 in the second, designed faster, and and 12 in the third. The reduction in development time with more advanced and lower manufacturing costs give STAM a significant features, that would advantage. not be possible to Noesis will benefit through an increase in sales of the simulate without HPC. Optimus software directly related to the sales of the • Noesis expects a CloudGear solution, as CloudGear is built on the Optimus 3% increase in total platform. The Optimus optimization tool is extremely sales, worth a total of flexible and can easily capture the simulation needs of end 1,6 million € over five users in a number of different industries. years. 110 111 EXPERIMENT #706 HIGHLIGHTS Industry Sector: Automotive HPC based high-resolution modelling Country: Slovenia of magnets Software used: ELMER, ParaView, NetGen ORGANIZATIONS Magneti Ljubljana is a Slovenian SME that has produced permanent metallic and systems magnets for the European market for over 60 years. These have many THE SOLUTION uses in a variety of sectors. XLAB is a Slovenian R&D company with a strong research background in the To tackle the problem of pressing tool optimization, XLAB developed a set of software fields of distributed systems, Cloud computing, system services based on open-source solutions. XLAB built a computer model of the pressing security, information visualization and image processing. tool and its behaviour during the compaction process. This model is highly configurable, so Magneti can reuse it for other applications. The optimization service runs as a web End User HPC Expert HPC Provider application, which provides an easy-to-use interface. The application connects to Arctur’s HPC system and submits an HPC job according to configuration and input parameters from the end user. This means that even inexperienced users are able to design and run experiments using HPC resources, avoiding the need for costly training. Magneti only needs to pay for the computing resources it uses, providing the company with a cost-effective solution. www.magneti.si www. xlab.si www.arctur.si THE CHALLENGE Magneti produces its magnets through a process called compaction, which uses a hydraulic BUSINESS IMPACT BENEFITS press to apply pressure to magnetic powders until they solidify. The hydraulic press is made up of several very expensive parts which regularly wear out and must be replaced. The The partners in this experiment have derived a number of • Design of a better pressing tool needed to be optimized, so it could be used for longer and with lower material benefits from their involvement in the Fortissimo project. pressing tool that costs, but doing this requires the ability to automatically detect yielding of the tool under a saves money, does given pressure. This requires many iterations of computer simulation and post-processing, For Magneti, there are several benefits. Due to the not need to be which exceeds in-house capabilities. optimized geometric properties of the pressing tool, the replaced as often, and quantity of excess material in an existing tool was reduced uses less energy. by around 32%, reducing material costs. This has reduced the cost of making the pressing tool by 27%, which • Savings of over represents an annual saving of €87.000. The pressing tool €100.000 per year for is also of a higher quality, containing narrower coils which Magneti and ability to consume less power. Assuming Magneti replaces all of create new services their pressing tools with those designed by the HPC tool, based on improved this will save another €16.200 annually. pressing tool. XLAB has broadened its software development expertise • Magneti and to the field of magnet production, and gained knowledge XLAB both gained about developing complex software with the help of open-experience in source tools. The existing software developed for Magneti simulation. can be extended and modified to potential new customers coming from the same or similar industries. Arctur has increased its reputation in the research community, potentially allowing it to attract new customers from the magnet production industry. 112 113 EXPERIMENT #707 HIGHLIGHTS Industry Sector: Automotive Cloud-based HPC optimisation of Country: Slovenia manufacturing processes Software used: TensorFlow ORGANIZATIONS EMO is a Slovenian SME specialising in the production of tools and dies for stamping, particularly sheet metal, in the automobile and aerospace industries. EMO utilises THE SOLUTION laser metal deposition (LMD) technology. AIMEN is a Spanish not-for-profit organization with expertise EMO currently uses LMD. As part of this process, EMO needs to gather and analyse in laser technologies, manufacturing processes and significant amounts of data. A more efficient workflow would enable it to complete more the development of monitoring systems for industrial projects without extra investment. A new system has been developed in this experiment that applications. exploits recent advances in AI for image analysis and in data acquisition from images of the process. The new technology, called CyPLAM, is a novel approach to the online monitoring End User Domain Expert HPC Provider & Expert of LMD. It uses deep-learning principles, working on the Fortissimo infrastructure, to enable online and real-time quality control and monitoring of key features such as dilution and clad height. CyPLAM has been validated by testing on a repair application using LMD. www.emo-orodjarna.com www.aimen.es www.cesga.es THE CHALLENGE LMD is an additive manufacturing technology that enables the generic 3D printing of large BUSINESS IMPACT BENEFITS metal parts. Additive manufacturing is a rapidly growing sector, as it allows for complex components to be produced with short lead times. However, the lack of sufficient control CyPLAM addresses the quality issues of LMD-created • Using HPC, EMO can remains a barrier, as it can result in unnecessary reworking, waste and an increase in 3D parts, putting EMO at an advantage as they are first to use reduce operational printing times. This reduces both profits and efficiency. This experiment aimed to overcome the technology. Future CyPLAM users can expect to cut costs by 20% and the current deficiencies in online monitoring and control of laser processing, so that EMO can production times by over 30%, as well as producing a higher-save over 2.000 realise the full benefits of additive manufacturing and create better quality products. quality product needing less reworking. Overall, end users machine hours per can expect a 20% saving in operational costs and a 30% year. reduction in lead-time, compared to traditional approaches. • Users of CyPLAM can AIMEN will use CyPLAM technology to support its recently stay competitive in launched CLAMIR system, a commercial process control the global Additive system for Laser Additive Manufacturing. Manufacturing The experience and knowledge acquired during this market. experiment have allowed CESGA to obtain new projects • As a result of the and contracts within the industry. They have also created a expertise gained training course on Machine Learning. in this experiment 3D printing and other additive manufacturing technologies have CESGA is offering had a major impact on the European manufacturing industry, a new Machine allowing fast and flexible prototyping and part creation. This Learning service industry is home to many SMEs due to the comparatively (SaaS) based on low cost of entry, and is growing fast. The worldwide 3D TensorFlow. manufacturing industry is growing at a rate of 25% per year, and is expected to be worth 6,5 billion USD in 2019. 114 115 EXPERIMENT #708 HIGHLIGHTS Industry Sector: Automotive Cloud-based HPC processing for Country: Spain knowledge generation in camshaft Software used: manufacture M3 ORGANIZATIONS EPC Group Europe is headquartered in Spain. It specialises in the manufacture of large camshafts (over 2 meters long) for use in the naval industry, agriculture, military, generators, THE SOLUTION and other sectors. TRIMEK is a Spanish SME founded in 1993. TRIMEK is one of the main manufacturers of metrological TRIMEK developed a metrological module of its M3 software, which enables the use of systems and solutions worldwide, and is the 4th worldwide big data resources for data analysis. The new module is able to analyse scanning and provider and the leading company in the Basque Country and measurement data and compare parts much faster than before by using Cloud-based HPC. Spanish markets of Coordinate Measuring Machines (CMM). The production process is therefore more time-efficient. The time savings enables more of the produced parts to be checked in the same amount of time, increasing the overall End User ISV HPC Expert production quality. Reducing the processing time for data analysis allows a higher knowledge and control of the quality of the production process and of its needs, such as maintenance requirements. By using this system, the end user receives a higher-quality product, saving time and money, leading to benefits for their own client. www.engpwr.com www.trimek.com www.cesga.es THE CHALLENGE EPC manufactures more than half a million camshafts every year. The company needs BUSINESS IMPACT BENEFITS to ensure that it produces high-quality parts. Monitoring of the manufacturing process generates a massive amount of data which could be used as a basis for machine learning TRIMEK, as service expert, developed a new software • 20% increase in to improve the production process. However, current software solutions are not able to module for statistical analysis. Due to the optimized data-turnover over five fully process the data fast enough, so the potential knowledge that may be generated from processing software developed in this project, TRIMEK years for TRIMEK. it cannot be exploited. In order to function alongside a busy production line, the solution expects an increase of 20% in new licences sales. requires significant computational power, together with the application of novel machine • EPC can save 1,5 learning technologies. EPC has tested and used the new M3 capacities within million € over five their production site, proving its ability to improve years using the new production capacity and speed. In a short time, the solution. expected benefits to the internal production process will reduce the number of defective parts to almost zero, • The number of resulting in a cost reduction of 1,5 million € over the next defective parts is five years. reduced to almost zero. CESGA will exploit the experiment results in its big data platform, which has been analysed and improved to fulfil the requirements of its commercial customers. As initial step, CESGA has created a white paper to show the capabilities of the infrastructure to use as marketing tool to attract other companies to use it. CESGA expects to be able to increase its SME projects and services in the next years to reach the strategic indicators for 2020. 116 117 EXPERIMENT #709 HIGHLIGHTS Industry Sector: Construction Cloud-based simulation of box Country: Spain framed timber beams Software used: ANSYS Mechanical ORGANIZATIONS Maderas Besteiro is a Spanish SME that specializes in timber solutions for structural applications, such as house-building. It has recently taken an interest THE SOLUTION in offering a new product line of box framed timber beams made from Eucalyptus globulus. USC-PEMADE A virtual engineering workflow was designed and validated. The workflow combines the is the Structural Timber Engineering Department of the design, calculation and virtual validation stages of the process, through access to a Cloud-University of Santiago de Compostela. based HPC platform provided by Fortissimo. Using this solution has allowed Madera Besteiro to reduce their development costs and design times. A pay-per-use HPC model significantly End User Domain Expert HPC Provider reduces the computational costs compared to using in-house resources, and results in a better, more refined end product. The company can now optimise Eucalyptus-based solutions at minimum 19% faster than before the experiment. This will allow them to offer a better product at a lower cost than their competitors. www.mbesteiro.com www.pemade.com www.gompute.com THE CHALLENGE Eucalyptus is ideal for structural purposes because of its high quality and resistance. Using BUSINESS IMPACT BENEFITS this wood can allow construction companies to build homes faster, with less waste and less energy, in line with European Union housing guidelines. Thanks to the virtual design process offered by • Maderas Besteiro Designing and testing a new structural product is an extremely costly process. As an SME, the Fortissimo Marketplace, Maderas Besteiro has can improve its Maderas Besteiro faces a challenge to develop new products due to its small number of staff significantly reduced their cost per m², when comparing overall productivity and lack of sufficient financial resources. A cost-effective computer simulation process would the new box framed timber beam with the traditional by 23%, thanks to help to solve this problem. framing system. Using eucalyptus instead of other optimisation of just hardwoods, their turnover is increased €40.000 per year. one of its products. More importantly, Maderas Besteiro is now able to easily • Increase in turnover optimize new designs, meaning that it has a competitive of €40.000 per year advantage in its industry. and ability to easily From now on, PEMADE will be able to simulate more optimise new designs. complex models than before, including complete • PEMADE have engineering structures. Moreover, the ability to simulate gained the ability hardwood species such as eucalyptus will make it the to simulate whole leading academic group in Spain in this field. complex structures, End users of this workflow can reduce their computing as opposed to their costs by up to 81%. From a business perspective, it has previous capabilities. an enormous potential as the timber industry has over 400.000 active companies. Gompute is now ready to offer a complete solution in the Fortissimo Marketplace for customers who need to simulate timber structures in a fast, easy and affordable way. 118 119 EXPERIMENT #710 HIGHLIGHTS Industry Sector: Aeronautics Near real-time analysis of airframe Country: Netherlands certification test data Software used: Pivotal ORGANIZATIONS Colosso is a Dutch SME specialized in the design, analysis and physical testing of high-tech materials and structures, such as those used in the aerospace industry. THE SOLUTION Within this context, the key focus of Colosso is on the improvement of its processes and products. KE-Works One of the ways of reducing the cost and lead-time of the material qualification process is to is a Dutch software company. It develops KE-chain, a improve the predictive capability of material models. Improved models lead to a reduction of modelling platform for engineering applications. the amount of testing required to achieve a reduction in the overall weight of the airframe. The approach taken in this experiment was to use KE-chain, together with Colosso’s data End User & Domain Expert ISV HPC Provider & Expert analysis and storage framework to calibrate a new algorithm to model materials based on data from fatigue tests. An HPC environment provided by Gompute was used to provide on-demand computing resources. This resulted in an improved ability to predict crack propagation in airframe components and a reduction in the required amount of fatigue www.ke-works.com www.gompute.com testing. THE CHALLENGE The development of aircraft for civil aviation is driven largely by the economics of the BUSINESS IMPACT BENEFITS materials constituting the airframe. Improvements in strength and durability can reduce aircraft weight and allow regulators to increase the inspection intervals. The results of this experiment directly contribute to the • Improvements to There is a continuous demand for better materials and a greater understanding of how these further development of a Colosso’s advanced physical the material model materials perform in aircraft components. However, introducing a new structural material and virtual test data monitoring and analysis tool (Pivotal). for testing airframe for an airframe is costly and takes several years, so there is a significant need for better KE-works will integrate the developed features and components allowed certification processes. infrastructure deployment options in the offerings of KE-better components to chain. be manufactured. Using HPC will enable (near) real-time data processing and • Development of evaluation. A Cloud-based, pay-per-use HPC system means a new, promising that using Pivotal and KE-chain is much more cost-effective software product, than using in-house resources, making these offerings Pivotal, which can be more competitive. offered together with The experiment, supported by the Cloud infrastructure KE-chain. provided by Gompute, strengthened the competitive • Increase in revenue position of both KE-works and Colosso. Because of the of up to €100.000 per increased potential and capabilities of KE-chain and Pivotal, year. an increase in yearly revenue of €50 - 100.000 is predicted. 120 121 EXPERIMENT #713 HIGHLIGHTS Industry Sector: Mechanical Engineering Cloud-based HPC simulation of Country: Italy railway infrastructure for Software used: high-speed trains Siemens NX ORGANIZATIONS Alstom Ferroviaria S.p.A. is the Italian branch of Alstom Transport, a multinational leader in the railway transportation sector. Alstom Transport has a presence THE SOLUTION in over 60 countries and employs over 28.000 people. A promoter of sustainable mobility, Alstom Transport This experiment focused on predicting the effects of failures in different components of the develops and markets the most complete range of turnout (or set of points), one of the basic part of railways infrastructure, with high accuracy. equipment and services in the railway sector. HPC was necessary as the turnout is a complex and large-sized system, which became even clearer during the course of the experiment. The dynamic interactions between the basic End User Application Expert HPC Provider & Expert components of the system were studied by taking into account the dynamic load due to the transit of high-speed trains and the failure modes of the turnout. Hypertec Solutions, an Italian SME, was responsible for the development of the model, adapted by CINECA into the Fortissimo Cloud environment. www.alstom.com www.hypertecs.it www.hpc.cineca.it THE CHALLENGE Alstom is investing huge effort in creating new services for the railway industry and other BUSINESS IMPACT BENEFITS transportation fields. One of the main areas of investment currently is the development of a diagnostic service to automatically schedule maintenance intervals. The major benefits reported by the partners are: • 50% decrease in time A diagnostic service would need to: understand how parts degrade and determine which For Alstom, replacing the system based on field required to create would need to be serviced at each interval; automatically plan availability of spare parts or measurements with an effective diagnostic monitoring a new diagnostic consumables; and quickly report to engineers what needs to be done in order to service the based on simulation has resulted in up to a 50% decrease system for a railway train and the network, as efficiently as possible. in time-to-results (from between 16 & 22 months to 8 component. months), with a corresponding cost saving of between • Savings up to €50.000 and €70.000 per application. In the best case €280.000 per year scenario, this will result in a saving of €280.000 per year. for Alstom, based For Hypertec, the experiment allowed it to increase its on optimising four engineers’ skills, enabling new services for Hypertec’s applications per year. industrial customers in many areas of CAE Simulation. It • A predictive service is estimated that, as a consequence, HPC-related revenue developed which will increase 87% in 2017 to €45.000. By 2020, this is will more accurately expected to reach €100.000 per year. monitor the railway network and schedule service interventions when necessary. 122 123 EXPERIMENT #714 HIGHLIGHTS Industry Sector: Automotive Predictive diagnosis services for Country: Italy the automotive industry Software used: R, R Shiny ORGANIZATIONS TEXA, an Italian SME founded in 1992, designs, produces and sells diagnostic instruments for cars, motor bikes, and other vehicles. This allows for services such as THE SOLUTION real-time monitoring of fleet vehicles. T2I is an Italian research organisation that supports companies through During the experiments, four Data Analytics prototype services were defined. These are based the design, development and testing of new products information gathered from on TEXA’s On-Board Diagnostics (OBD) systems. These services and services. cover areas that may affect the reliability, condition, or service needs of a vehicle - such as how it is driven, failure patterns, and overall health of the vehicle. End User Application Expert HPC Provider & DA Expert A Cloud HPC-powered workflow was developed. This was designed to easily integrate into TEXA’s existing automotive Data Analytics services. A service architecture has been defined that connects the existing TEXA infrastructure, equipped to collect data from installed black boxes, to an HPC Cloud provider. www.texa.com www.t2i.it www.hpc.cineca.it THE CHALLENGE Vehicle manufacturers generally have limited knowledge of a vehicle’s life once it leaves them. BUSINESS IMPACT BENEFITS A service that can predict failures, mechanical problems or damage at the component level, and offer detailed information on these components, would be extremely valuable, saving For TEXA, the Net Present Value of these new services is • Development of a manufacturers and fleet managers time and money. This service would gather and analyse estimated to reach an overall value of 1,2 million € over the service using TEXA’s data from TEXA’s sensors, which could be used to redesign parts and modify maintenance next 3 years. The ability to use an HPC-enabled workflow ‘black boxes’, which schedules. This type of analysis requires significant computing power. to analyse data from their diagnostics systems will enable can monitor and better oversight of fleet vehicles and predict failures in predict failures in fleet time for these to be addressed quickly. vehicles. This has T2I will apply the principles developed here to offer value for both TEXA HPC for predictive maintenance in other similar fields and its customers. (mainly highly “sensorized” products in sectors like home • New services have appliances, smart buildings, and energy). T2I estimates a an overall value of 1,2 5% increase in commercial revenues from related services, million € over three in years 2018 – 2020, worth up to €105.000. years for TEXA. • T2I increase revenue by 5% per year, worth up to €105.000 over three years. 124 125 EXPERIMENT #801 HIGHLIGHTS Industry Sector: Manufacturing Cloud-based simulation of Country: Germany ultra-high-temperature furnaces Software used: OpenFOAM ORGANIZATIONS XERION (Germany) specializes in the development and production of ultra-high-temperature furnaces and producing new materials for high-tech industry applications. THE SOLUTION NAVASTO (Germany) is the application expert responsible for the development of the method for the simulation of a XERION initially designed and manufactured a prototype furnace. After that, XERION and high-temperature industry furnace. The HPC provider of NAVASTO conducted tests on the furnace at various temperatures to acquire pressure and the experiment is the HPC Center Stuttgart (HLRS). temperature data, which enabled simulations of the furnace (carried out using the open source software OpenFOAM on Cloud-based HPC) to be validated. Based on the results End User Application Expert HPC Provider of the validated model, the companies identified areas where the furnace design could be improved, which was a direct outcome of the simulation results created by using cloud-based HPC resources. The tools developed in the experiment will eventually be integrated into the standard design tools of XERION for the design of high-temperature furnaces. www.xerion.de www.navasto.de www.hlrs.de THE CHALLENGE Ultra-high-temperature furnaces are one of the critical parts of the manufacturing chain of BUSINESS IMPACT BENEFITS electronic components. Knowing the temperature distribution and flow inside the furnace is essential for high quality production. As the temperatures can reach up to 3.000° physical Only a few companies provide services to high-temperature • Improved furnace measurements are costly. Hence, simulations that provide such data improve the design furnace customers and this experiment will allow XERION design process process. The simulations deal with the interdependencies of convection, conduction, to compete in this area. This has already generated and better use of radiation, Joule heating and a complex multi-material geometry, which requires significant considerable interest in XERION’s target markets including resources associated computing power. research institutes and companies using furnaces in their with modifications production or research and development activities. performed prior to As a result of the experiment, furnaces can be created with delivery and after improved thermal design. Thanks to advanced modelling, installation resulted in the properties of these furnaces are better understood. savings up to €25.000 This prevents issues such as unnecessary oversizing of for on-site repairs. components (e.g. heating power) and reduces the cost of • The price of the the furnace. This can further result in the reduction of the furnace can be sales price of the furnace system of around 20%. reduced by about 20% For NAVASTO, the experiment has resulted in increased (€40.000) in case of expertise in the field of multi-region simulation, high-large furnaces. temperature simulations and simulation process design. • There is expected the It plans to offer this knowledge to its customers via the ROI to be tripled by Fortissimo Marketplace. 2023. Also HLRS has gained valuable knowledge that will allow it to attract more customers from this domain and use this knowledge to provide platforms and services that satisfy their needs. 126 127 EXPERIMENT #802 HIGHLIGHTS Industry Sector: Maritime Cloud-based design of yacht sails Country: United Kingdom Software used: AeroSimPortal ORGANIZATIONS Cape Horn Engineering is a UK-based Computational Fluid Dynamics (CFD) consultancy company specializing in the design of racing yachts. PLUS Solutions is an Italian THE SOLUTION company specializing in tailored software solutions and virtual prototyping in multiple sectors such as Computer-The outcome of this experiment is the development of AeroSimPortal, an easy-to-use tool Aided Engineering (CAE). HPC resources and expertise for sail design studies. This tool removes previously-existing barriers that prevented small were provided by the Italian HPC center CINECA. companies from using CFD services. Through the use of an HPC web- and Cloud-based simulation platform, all of the necessary tools to analyse sails are made easily available to all End User Application Expert HPC Provider companies. The simple graphical user interface (GUI) allows the user to launch complex CFD/FSI simulations by only uploading a geometry and sailing conditions as inputs. Behind the GUI, there is a fully automated simulation workflow. Therefore, even inexperienced users can now www.cape-horn-eng.com www.plusolutions.it www.hpc.cineca.it access design tools that they were previously unable to use. THE CHALLENGE Cape Horn uses computer simulation to design the best sails for its boats. However, sail BUSINESS IMPACT BENEFITS design is a very computationally intensive process. Sail design workflows need to analyse the sail’s “flying shape”, which affects the airflow and pressure. This is achieved with a complex The AeroSimPortal platform constitutes an innovative and • Additional revenues Fluid-Structure Interaction (FSI) coupling between CFD and Finite Element Model (FEM) sustainable approach to deploying specialized sail design from consulting are software packages. workflows. A Cloud-based solution is especially useful estimated at about 97% of businesses in this area are SMEs. In this sector, there are still technological and cost to SMEs, which could otherwise not be able to use these €300.000 over the barriers to accessing HPC-based simulation. services regularly enough to justify the cost of purchasing next three years. their own HPC hardware. The revenue streams resulting from this experiment include both consultancy services and • Additional income SaaS fees. from the introduction of a SaaS platform is Through the AeroSimPortal platform, Cape Horn is able to estimated at approx. reduce the cost of its design and consultancy services by €200.000 in the next around 50% while keeping the same profit margins. This three years. allows the company to have extremely competitive prices and to increase its market share. Beyond the consultancy • Significant work time service, Cape Horn will deploy AeroSimPortal through reduction for sail a scalable pay-per-use platform. The AeroSimPortal is designers since they expected to be available on the Fortissimo Marketplace in can concentrate on early 2019. design issues instead of on fluid dynamics. PLUS and CINECA also benefited from expertise gained through the experiment. They will be able to target customers in the maritime engineering sector and will be able to better compete on the market thanks to this experience. 128 129 EXPERIMENT #803 HIGHLIGHTS Industry Sector: Logistics Cloud-based estimation of marine Country: Spain freight rates Software used: TensorFlow ORGANIZATIONS IMATIA Innovation is a Spanish SME that provides Data Analytics, Artificial Intelligence design and development services. VGSC (Vasco Gallega Sociedad de Cartera, THE SOLUTION Spain), a company specialising in R&D was the end user of the experiment. They work with companies, who need to A software-based solution has been developed, which enables an estimation of the freight ship products by sea and need to find the optimal way of cost of any particular shipment. The core of this solution is a neural network (NN) that uses transportation for the lowest cost. CESGA provided HPC VGSC’s historical data, historical information of the market and other external parameters. In expertise and resources for the experiment. order to obtain an optimal NN, a large number of parameters have been studied by building and training numerous specific NN models. HPC proved to be the best tool for creating and End User Application Expert HPC Provider, Expert & Host Centre training the neural network due to the large number of parameters that need to be included for the best estimation. The solution is delivered as a secure SaaS. Its performance will be improved in the future by retraining when more data becomes available. www.kaleidologistics.com www.imatia.com www.cesga.es THE CHALLENGE The logistics chain is an important part of the manufacturing process. Logistics can account BUSINESS IMPACT BENEFITS for 20% of the final cost of a product for a manufacturing company. What is more, in cases where products need to be shipped by sea, this can rise to 90% of the final product coast. IMATIA will offer a service for freight cost estimation. The • Better cost Therefore, finding an optimal transportation option is an important but also complex task. solution is based on the outcomes of this experiment and predictions and faster The tools currently available on the market are too slow compared to the fast decision-is published and promoted as a consulting service via quotations, improving making that the process requires due to the great number of variables that must be the Fortissimo Marketplace. Furthermore, the solution is the performance of considered. combined with several other capabilities that the company VGSC by 5%. provides, such as data processing, multiple data sources information analysis and integration, as well as NN • Return of €75.000 knowledge (especially the TensorFlow). during a two-year amortization period The SaaS platform developed for the experiment was for VGSC. tested by VGSC workers in order to demonstrate that it is able to interact with their internal software and can be • Better logistics know- used without any issues. Additionally, VGSC has acquired how enables IMATIA high-value knowledge about the influence of several to understand the parameters from their historical data on the freight rate. problems of potential This enabled VGSC to improve their internal QA processes clients from the and software solutions. maritime sector. CESGA has improved their HPC machine learning (ML) services and developed a HPC Python toolkit to perform hyper-parametric search on ML problems using the TensorFlow. 130 131 EXPERIMENT #804 HIGHLIGHTS Industry Sector: Manufacturing Simulation of additive manufacturing Country: Italy processes for the production of Software used: metal components ABAQUS ORGANIZATIONS HPE (Italy) provides engineering solutions and technology to the automotive, motorsport, automation solution and defence sectors. Their major focus is on powertrain THE SOLUTION technologies. EXEMPLAR was founded by the Politecnico di Torino University and provides services for Computer This experiment provided a numerical analysis environment, which enables users to forecast Aided Engineering projects. The HPC expertise and the results of the AM process using Cloud-based HPC simulation. The forecasting tool is able resources were provided by the Italian HPC centre CINECA. to predict the stress state of the final component as well as how the product might deform due to differential cooling during the manufacturing process. The multi-scale and multi-End User Application Expert HPC Expert & Centre physics nature of the problem has been captured in the tool, allowing engineers to better understand the manufacturing process. This approach has created a robust solution by identifying which parameters most impact the quality of the final result before the production itself begins. Hence, it is clear which parameters need to be tightly controlled. www.hpe.eu www.exemplar.com www.unimore.it www.hpc.cineca.it THE CHALLENGE Additive Manufacturing (AM, also called 3D printing) technologies are generating huge BUSINESS IMPACT BENEFITS interest in all industry sectors because they enable the rapid design and manufacture of prototypes and end-products. However, there are challenges associated with the use of The impact of this experiment for HPE has been the • HPE reduced these technologies such as component failure due to issues with layer deposition. There are increase of the effectiveness of its optimized AM solutions. time-to-market of multiple points during the fabrication process when the manufacture of a component may Among the major benefits are the reduced time-to-market their products. fail. In addition, post-production failures of complete parts can also occur. and the reduced number of defective parts due to a closer match between the initial design and the manufactured • The number of part. What is more, this has resulted in lowered production defective parts is costs. reduced. In terms of the reduction of time required for tuning the • Savings in the amount AM process, there were two main outcomes. Thanks to of up to €90.000 per this experiment, the build failures and multiple design year. iterations can now be avoided. The benefits for HPE could • EXEMPLAR expects actually be replicated for any SME that purchases the an annual turnover service via the Fortissimo Marketplace. Furthermore, the of €125.000 from the results of the experiment are expected to increase HPE’s SaaS offer. market share by 50% and reduce costs by €90.000 per year, thanks to the lower levels of material wastage. EXEMPLAR is able to offer the technology developed here as a commercial service under a SaaS model, with a potential annual turnover of €125.000. 132 133 EXPERIMENT #805 HIGHLIGHTS Industry Sector: Manufacturing Cloud-based design of ultra-clean Country: Germany containers for high-purity chemicals Software used: MpCCI ORGANIZATIONS RIKUTEC is a leading developer and manufacturer of molded containers. Fraunhofer SCAI (Germany) is a non-profit R&D institute providing software development. Hagen Stiftung THE SOLUTION and Hagen Engineering (Germany) are domain experts in plastic parts and blow molded articles. Accuform (Czech The partners have developed a standardized and automated workflow incorporating different Republic) provided the process simulation software B-SIM. simulation tools and therefore eliminating the need for manual intervention. This involved Gompute (Sweden) provided the HPC resources. extension to the multi-physics interface software MpCCI from Fraunhofer. Through using Cloud-based HPC and parallelization of the simulation, the time required for a new design End User HPC Expert ISV Application Centre & HPC Provider was reduced significantly. The combination of this new workflow and the use of HPC means Engineering Service Provider the automated simulation of blow molded parts is possible for the first time. As a result, the development times have been shortened and more designs can be tested at the same time. Additionally, expensive physical molds are used less often, which reduces the costs of manufacture. www.rikutec.eu www.scai.fraunhofer.de www.t-sim.com www.hagen-stiftung.de www.gompute.com THE CHALLENGE Extrusion blow molding is the standard method for manufacturing hollow plastic parts such BUSINESS IMPACT BENEFITS as bottles, cans and large containers. A specific example is ultra-clean containers for the microelectronic device industry. Companies manufacturing these products need faster, more RIKUTEC benefits from this experiment through access • Reduced costs for reliable and efficient design methods. Prior to the experiment, the automated optimization the to more reliable and efficient design methods with faster each design loop individual design and simulation had to be carried out manually. The existing process used a response times. The number of design iterations was amounting to more standard workstation system, with one single design loop taking up to 2 days. reduced and the quality of products is enhanced. Through than € 20.000. the application of simulation methods, substantial cost savings were achieved. Hagen Engineering is using the • Optimization of item experiment results to expand its market position as a weight reducing the provider of simulation services. More companies are total costs by several expected to use CAE-methods in the design process thousand euros per of plastic parts in the future because the access to an year. automated, integrative CAE work-flow makes them more • Due to the use of competitive. The extension of its multi-physics software HPC the computation solution will help Fraunhofer to find new customers. Over times are shortened the next 2-3 years Fraunhofer expects 5-10 new users with by a factor of 4 - 5. license revenues of €30.000 to €60.000 per year. In the medium term, Fraunhofer expects to see a rapid increase in the market for integrated manufacturing simulation solutions. Interface tools like the MpCCI Mapper will enable Fraunhofer to exploit that market growth. 134 135 EXPERIMENT #806 HIGHLIGHTS Industry Sector: Maritime Cloud-based simulation of the Country: hydrodynamic resistance of ships’ United Kingdom hulls in the presence of waves Software used: HELYX ORGANIZATIONS The Numerical Algorithms Group (NAG) is a UK HPC expert and software supplier. ENGYS is a UK ISV specialising in open source CFD solutions including THE SOLUTION the HELYX software used in this experiment. CETENA is an Italian company specialising in the maritime field. Performing the simulations in-house with the required degree of accuracy is not a viable EPCC, the supercomputing centre at the University of approach for a SME as it takes too long. A single wave and speed condition simulation takes Edinburgh, was the HPC provider. multiple days to complete on a medium size HPC cluster with 80 to 100 cores. Methods that need less computing time exist, but do not produce results of a sufficiently high quality. The End User ISV HPC Expert HPC Provider & Host Centre solution is to use a Cloud-based HPC approach, in which multiple jobs can be submitted to a large on-demand HPC resource. The HELYX client has been adapted to be able to submit jobs to the HPC service seamlessly. The software that runs on the HPC system has been optimized to run efficiently on multiple cores and return the results quickly to the user. www.cetena.it www.engys.com www.nag.co.uk www.epcc.ed.ac.uk THE CHALLENGE Developing effective methods to achieve accurate predictions of additional ship resistance BUSINESS IMPACT BENEFITS in waves is currently the priority for ship designers and maritime consultancy companies. Physical testing is time consuming and expensive. Hence, accurate and reliable numerical Potential savings for CETENA can amount to more than • Lower computational methods need to be developed as a viable alternative to traditional experimental tests. In €100.000 per year by employing the HPC numerical cost and no need for this respect, the potential savings for a shipbuilding company could amount to more than solutions developed. The pay-per-use business model expensive physical €100.000 per year by reducing the need for experimental models in favour of an HPC-based is highly attractive for SMEs, as the purchase and model tests. simulation analysis. maintenance of an in-house cluster are too expensive. • Higher quality of The results of the experiments will enable ENGYS increase results due to the use the sales of both HELYX and HELYX-EcoMarine products. of Cloud-based HPC. State-of-the-art CFD software solution for advanced ship resistance combined with on-demand HPC will offer a • Severely reduced time competitive advantage for users. The target market is huge, in the design cycle with over 830 active shipbuilders and shipyards operating due to being able worldwide and a predicted global revenue of €240 billion to analyse multiple by 2019, representing a five year projected 1,2% annualized scenarios in parallel. growth. In Europe alone, there are over 150 large shipyards • Potential savings that with a global market share of around 6% in terms of tonnage could amount to more and 35% for marine equipment, totaling €60 billion in annual than €100.000 per turnover according to official EU figures from 2012. As a year for a typical user. first step towards exploiting this large market, the partners are currently targeting a small group of companies in the maritime industry located in Italy, Japan and South Korea to take-up the technology developed in Fortissimo. 136 137 EXPERIMENT #807 HIGHLIGHTS Industry Sector: Energy, Civil Enginering Cloud-based micrositing of small Country: Spain wind turbines Software used: OpenFOAM ORGANIZATIONS KLiUX Energies (Spain) manufactures vertical-axis wind turbines for urban and residential environments. The University of Zaragoza (UNIZAR-BIFI, Spain) hosts its own THE SOLUTION computing centre and provided application expertise. nablaDot (Spain) provided CFD consultancy services for the A tool was developed that could integrate all of the necessary factors required to calculate experiment. Gompute (Sweden) is a HPC solution provider. the optimal placement of turbines. This tool includes GIS (Geographic Information System), It provided the HPC expertise in this experiment. NWP (Numerical Weather Prediction), CFD (Computational Fluid Dynamics) and analytics software to accurately simulate the urban environment. Due to the complexity of the End User Application Expert Host Centre HPC Provider & Expert calculation, HPC is the most suitable solution. A Cloud-based solution is optimal as it is the most cost-effective way of accessing HPC resources. Significant savings are possible by using the developed tool, with which an urban area of approximately 2 km2 can be modelled costing of around €1.500 – €2.000 and requiring approximately 15 days of computing calculation. www.kliux.com www.bifi.es www.nabladot.com www.gompute.com THE CHALLENGE Wind flow over urban areas is very difficult to characterize. This is due to a number of factors, BUSINESS IMPACT BENEFITS such as high turbulence and other strong effects created by complex urban geography. When placing wind turbines, it is critical that the detailed flow pattern of the area is well understood. The main output is the tool developed for the evaluation • Yearly saving of Current urban wind assessment tools are costly in relation to the cost of a small wind turbine. of turbine placement in urban environments. Although €80.000 to €120.000 It can take up to a year and cost €12.000 to carry out the necessary physical measurements, currently the tool still requires some manual intervention, for the end users of which can easily double the cost of placing a single turbine. the goal is to eventually offer it as an online pay-per-use KLiUX. service. Meanwhile, nablaDot offers it as a consultancy-based service. Estimating the average wind speed, wind • Additional sales of its direction, air density and wind hours available at every site products are expected is crucial for KLiUX to determine which wind turbine size to increase revenues model is most appropriate for each installation site and by 20% to 40%. significantly increased KLiUX’s ability to compete. • nablaDot expects an nablaDot plan to use the technology developed in this increase in turnover experiment commercially (initially as a consultancy, of €150.000 annually moving to SaaS later) with an expected increase in in the medium term, turnover of €150.000 annually in the medium term and the increasing this to creation of two new job posts. When the number of users twice as much in the increases a higher turnover of extra €300.000 is expected longer term. and the creation of three more job posts. This experiment has allowed UNIZAR-BIFI to experience working with multi-workflow solutions. This experience proved to be extremely useful as it is beneficial to other clients, which may have complex computational needs. 138 139 EXPERIMENT #808 HIGHLIGHTS Industry Sector: Digital, Data Analytics Cloud-based simulation of page Country: Greece curling in the copying of documents Software used: CURLO ORGANIZATIONS Bookscanner is a Greek company that specializes in digitizing printed material. Vertoyo is a Greek SME, which specializes in technology development and software THE SOLUTION service provisioning in the digitization field. HPC expertise were provided by The Laboratory of Robotics and A Deep Neural Network (DNN) has been trained using simulated page curling of 1 million Automation, Democritus University of Thrace. Arctur, the images. The algorithm takes two already cropped pages from a book image and outputs Slovenian HPC centre, was the HPC provider. an artificially curled book page. Using the artificially curled book pages, a state-of-the-art deep Convolutional Encoder-Decoder (CED) Neural Network was trained in order to apply End User ISV HPC Expert HPC Provider & Host Centre the de-curling process. After training, the DNN can de-curl newly scanned pages with very good success rate. Evaluation of the page de-curling problem showed accuracy of curling correction in over 90% in most cases. With the computational power provided by HPC, the training procedure of the DNN is at least 30 times faster than using a typical workstation. www.bookscanner.gr www.vertoyo.com www.duth.gr www.arctur.si THE CHALLENGE Digitization of books is an important process, both for commercialization and for preservation BUSINESS IMPACT BENEFITS of older texts. The Bookscanner© product can automatically and physically turn pages and automate the scanning process. However, this process results in a ‘page curling’ effect where The only current method for page curling correction is • Automatic de-curling the pages are attached to the spine. The scanned pages need to be digitally flattened, a based on a projected laser grid that requires each page to of pages and time tedious and expensive process that this experiment aimed at improving and commercializing. be scanned twice. The solution created by the experiment, savings. which is called CURLO, removes the need for additional laser grid projection equipment and provides a 50% • Increase of 10-15% improvement on the standard curling correction procedure. in revenues for participating SMEs. As a result of this experiment, the CURLO solution can be offered as a post-processing service to accompany • Payback period for the the Bookscanner© product. The collaboration with Arctur CURLO investments is has allowed an improvement in the quality of batch-mode 3-4 years. scanning. This will be offered as a Software as a Service • The excess (SaaS) framework for scanned page de-curling. revenue after 3 In addition to the digital content market, the partners years is estimated in this experiment are ready to address the needs of at approximately digitization and the paperless economy, e.g. insurance and €300.000 - €350.000. paperless banking. Especially regarding for the banking sector, the most recent studies estimate that productivity improves up to 39% when electronic forms replace paper and workflows are used to streamline processes. 140 141 EXPERIMENT #809 HIGHLIGHTS Industry Sector: Manufacturing, Data Data analysis to improve the welfare Analytics of laboratory animals Country: Italy Software used: In house code ORGANIZATIONS Tecniplast is an Italian company, a worldwide leader in the laboratory animal industry, providing solutions and services to support research. Moxoff is an Italian SME with THE SOLUTION a deep knowledge in data intelligence and mathematical/ numerical modelling. CINECA is one of the largest HPC A measurement system using a variety of sensors to provide real-time tracking of lab centres in Europe, eager to facilitate external access to its animals has been developed. Moxoff developed a High Performance Data Analytics (HPDA) systems for industry and SMEs. platform on an HPC infrastructure, implementing algorithm which process all the raw data recorded by the sensors of Digital Ventilated Cages (a Tecniplast product). It can be used for End User HPC Expert & ISV HPC Provider automatic event, anomaly detection, interactive data exploration and analysis. As the system incorporates potentially thousands of sensors, depending on the number of animals in the study, cloud-based HPC infrastructure was necessary to cope with the large amount of data. The real-time nature of the system is another reason why HPC is considered necessary as other computing solutions would be too slow. www.tecniplast.it www.moxoff.com www.hpc.cineca.it THE CHALLENGE In addition to errors of measurements, any form of human inspection of laboratory animals BUSINESS IMPACT BENEFITS used for research disturbs their environment and causes stress, which can potentially affect the results of the research. There are two types of measurements: research measurements The offering of this new unique measurement system • Tecniplast expects and measurements made to care for the animals. A system of quantitative measurements of product allows Tecniplast to increase its current market to increase revenues animal health status and activity would allow accurate information of the overall health of all share and to expand its customer base. As there is a by over €800.000 animals to be seen, and issues dealt with as they arise to support the ongoing research. very large potential market, this could be one of the most per year with a gross important data analysis services ever provided by Moxoff. margin of 90%. By continuing their partnership, Moxoff and Tecniplast will be able to offer new services to Tecniplast customers as • Tecniplast expects to well. increase their market share as their service The platform developed in the experiment will be extended offer is now much in the future in order to be applied to many IoT contexts more competitive in and markets. Automatic identification of states and trends, the market. user behavior profiling, predictive maintenance etc. are applications that are increasingly important in the market • Moxoff and CINECA due to the exponential growth of connected devices in the have the opportunity recent years. to enter a new market. In addition to acting as the provider of choice for the platform when the workflow will be implemented at a production-ready level, the HPC centre CINECA will benefit from the success of this experiment by aquiring numerous SMEs in the HPDA market as future customers. 142 143 EXPERIMENT #810 HIGHLIGHTS Industry Sector: Maritime, Civil Cloud-based optimisation of a Engineering multi-body wave energy device Country: United Kingdom Software used: OpenFOAM ORGANIZATIONS Zyba is a UK registered SME with expertise in digital and numerical modelling. Since 2014, Zyba’s primary goal is to engineer simple systems that work with nature. Together THE SOLUTION with BioRock Technology, Zyba is harnessing wave energy to protect coasts from erosion, enhance coral reefs and Simulations of the CCell device in different incoming wave conditions were carried out using deliver sustainable power. HPC resources and expertise in the OpenFOAM software package on an HPC machine. In addition, an easy-to-use GUI was this experiment were provided by Arctur. developed, which allows simulations to be set up quickly and to show a series of scripts and tools written to streamline of the workflow on the HPC system. The increased computation End User & ISV HPC Expert & Provider power allowed the whole CCell system to be modelled for the first time, including power-take-off hardware, software, and the intended control logic. This information provided a completely new insight into how the different pieces of the system work together. www.ccell.co.uk www.arctur.si THE CHALLENGE Zyba’s core product, CCell, is a curved wave energy converter designed to be efficient and BUSINESS IMPACT BENEFITS simple to assemble and operate. It moves with the waves to simultaneously extract their energy for electricity and reduce their impact on the beach. BioRock reefs use the low-voltage As a result of the experiment, Zyba has been able to • Reduced simulation charge from CCell to form limestone rock from seawater minerals that serve as habitat for increase its productivity and deploy a series of pilots set up time from 2 corals to grow at an accelerated rate, creating an active breakwater to protect shores from offshore. The automated and streamlined design process hours to less than 1 erosion. Optimization of the CCell for each site is heavily dependent on the local environment enabled it to rapidly optimize each design as well as minute. and is a computationally intensive task, which is too expensive for most SMEs’ budgets as is reduce design costs and development time. With the new physical testing. GUI, less experienced users can now complete design • Mitigated use of tasks on their own, which empowers individuals within physical modelling, the team and increases the rate of innovation. Overall the providing a nine-fold new system has enabled Zyba to significantly speed up reduction in cost. its market entry with saving around 9-12 months and will • Reduced HPC costs continue to facilitate the evolution of future products and from 0,09€/core-hour service offerings. to 0,05€/core-hour as Provision of resources from an HPC centre transforms an all software used was imposing capital expense to a more manageable operating open source. cost. Administrative tasks associated with IT management • Ability to concurrently and upgrades are also effectively outsourced, alleviating run simulations pressure on a small SME team and allowing them to focus improved productivity on their design work. by a factor of seven. 144 145 EXPERIMENT #901 HIGHLIGHTS Industry Sector: Wastewater Treatment CAE driven design of a water Country: Germany treatment plant Software used: OpenFOAM ORGANIZATIONS AKVOLA TECHNOLOGIES provides cost-effective solutions to clean hard-to-treat industrial wastewater containing high concentrations of oil and suspended solids. NAVASTO THE SOLUTION is a SME, which offers engineering services in the field of flow dynamics for the development of complete product During the experiment, investigations were conducted on a prototype plant to enable a solutions. ARCTUR, the Slovenian HPC centre, is the HPC calibration of the simulation process. After developing a reference cases for several aspects Provider. of the simulation, the resulting flow fields match very well with the experimental data. The validated and fully automated numerical workflow then could be used to identify regions End User Application Expert HPC Provider inside the wastewater plants where the flow structures are unfavourable for the floatation-filtration process. Subsequent improvements can be evaluated with very short turnaround times. This speedup in the design process is enabled by the combination of the open-source software OpenFOAM in combination with Cloud-based HPC. www.akvola.com www.navasto.de www.arctur.si THE CHALLENGE Enabling efficient and sustainable water cleaning is one of the main challenges of the 21st BUSINESS IMPACT BENEFITS century. To tackle future challenges, Akvola as a first time HPC resource user, wanted to research the viability of devising a water treatment plant entirely based on CAE technology. Identifying the shortcomings of akvoFloat (a hybrid water • The advanced The goal of this experiment was to gain the ability to simulate an industrial-water treatment treatment process consisting of flotation and filtration) hydrodynamic design plant with sufficient accuracy. This would enable Akvola to improve their design process and their mitigation through Computational Analysis of the wastewater-from an experiment-based workflow to a more economic CAE-based design process. The saves time and energy for plant optimisation. This results treatment plant simulations required significant computing power and a validation of the numerical model in lowering the price of the plant and reduces the capital increases the filtration against experimental data. investment needed by end users for new technology. efficiency by 20%. Also, CFD can be used in adapting the plant to the type of • The smaller size the water, since every industry creates water of different of the plants and qualities. Additionally, CFD-induced changes in the plant the reduced energy design reduce chemical and energy consumption and, and chemical in that way, reduce operational costs. These reductions consumption translate into shorter payback times, which is the biggest shortens the time driver for a sale. The increased expertise in the field of to positive return on multiphase simulation, hydrodynamic simulations and investment. simulation process design will help Navasto to showcase their competencies and capacities at conferences, trade • The development fairs and on the company’s website in order to increase costs and the visibility of the company. This can be used to reach development time out to new customers and increase business. Potential for new products are end users are manufacturers of water treatment devices reduced. and companies that design or manufacture devices with complex internal. 146 147 EXPERIMENT #902 HIGHLIGHTS Industry Sector: Civil engineering Earthquake resistant design using Country: France FE² technique Software used: Digital-Structure ORGANIZATIONS EGIS INDUSTRIES is a French Consultancy firm undertaking civil engineering projects. STRAINS is a French SME specialized in the development of civil THE SOLUTION engineering software. It offers consultancy to a wide variety of clients. EPCC is a UK HPC centre at the Digital-Structure is a cloud platform for structural analysis. The product was enhanced University of Edinburgh. during the Fortissimo experiment to address non-linear dynamics involving the behaviour of non-linear materials and resultant friction. This product has a highly robust convergence in comparison with traditional Finite Element Software. The data structure is optimised for End User ISV HPC Provider parallel computing: a local scale assesses the plastic behaviour of local steel joints and a global scale evaluates the dynamic behaviour of the structure. These computations require a huge amount of computing power. Since HPC infrastructure is expensive for SMEs, a Cloud-based solution on a pay-per-use basis offers the best way to control costs. www.egis-group.com www.strains.fr www.epcc.ed.ac.uk THE CHALLENGE Evaluating the reliability of civil engineering structures under extreme loads (earthquake, BUSINESS IMPACT BENEFITS explosions, flooding, etc.) requires non-linear static and dynamic analyses. To be accurate, these analyses must pay attention to structural details which are more likely to contribute Using Digital-Structure with its CAD or data import tools, • Engineers’ to failure. The demand for these kinds of analysis is increasing and is becoming mandatory. users can design a complete steel structure and its productivity doubles These analyses are mostly done by SMEs, but the computing resources needed to generate joints on the web platform. Improving the resistance and using Digital-exploitable results within a reasonable time and cost clearly exceed in-house capabilities. behaviour of the structure is easy: loads can be defined Structure. and elastic, elastoplastic, yield, and dynamic analyses • The value of can be launched even when a significant level of compute consultancy projects resources is needed. Digital-Structure significantly reduces increases by at least engineering time thanks to fast computations, complete 20% thanks to finer workflows (from CAD to building standards verifications) security margins and and a user-friendly interface dedicated to civil engineering. optimised designs. The software is very intuitive with no maintenance required (installations, updates, license, etc.). Computing times • Effective control of are much faster thanks to the use of HPC. The reduction computing resource in cost of designs for an SME is estimated to be between costs and risks. 40% and 60%. Overall these factors can double the productivity of an engineer. As a consequence, Digital-Structure enables users to optimise designs, determine finer and more accurate security margins and reduce environmental impact and production costs. Its use can result in a significant increase in the value of an SME’s structural designs. 148 149 EXPERIMENT #903 HIGHLIGHTS Industry Sector: Electrical Engineering Cloud-based multiphysics simulation Country: Austria for MEMS micro speaker design Software used: and development COMSOL Multiphysics ORGANIZATIONS USOUND GMBH based in Graz, Austria, is a fast-growing audio company, developing and producing the most advanced audio systems for personal applications based THE SOLUTION on MEMS (Micro Electro Mechanical System) technology. HLRS is the High-Performance Computing Center The USound in-house multi-physics simulation, running on the commercial tool COMSOL Stuttgart of the University of Stuttgart in Germany. Multiphysics, has been used to predict the Sound Pressure Level produced. This approach had restrictions and simplifications due to limited computational power available. The calculation time for this model is 10 to 20 hours, which slows down the development process End User & ISV HPC Provider and does not allow fast parameter variation. HPC offers the possibility to use a more detailed geometry without compromises, analyze and consider acoustical losses with thermos-viscous methods and speed the development process. The COMSOL Multiphysics tool was adapted to run on the HLRS cluster system directly over the native user interface. www.usound.com www.hlrs.de THE CHALLENGE USound GmbH develops unique MEMS miniature speakers based on the piezoelectric effect. BUSINESS IMPACT BENEFITS Such speakers have all the advantages of MEMS technology; automated production and assembly, a smaller form factor, negligible heat generation and improved linearity compared USound is pioneering MEMS speakers. In the area • Ability to use highly to the classical loudspeaker. A MEMS production run is expensive and time consuming. Very of microphones, MEMS have replaced traditional detailed models in powerful multi-physics simulation tools are needed as an alternative to physical prototypes to microphones almost completely by now. USound is fast simulations. predict the acoustic performance of the speaker and to reduce the number of production runs expecting the same to happen to micro-speakers. To needed. • Using exact models, address the mobile communications market, further €400.000 can be technology and product development is needed. The saved by reducing the development for the next-generation mobile speaker number of prototyping product will be 50% faster using the cluster computing runs from 4 to 2 (one methods developed in this experiment. It will therefore run is in the order of allow USound GmbH to gain market share and react to €200.000). customer needs in a timely manner. In the growing sector of MEMS devices, powerful multi-physics simulations • Faster time to market are mandatory because experimental evaluation is very by up to 1,5 years. expensive. As devices become more complex, all sorts of mechanical, electrical and thermal influences have to be considered. These lead to very demanding simulation models. MEMS developers are limited by their conventional workstations. The experiment shows that for a demanding FEM (Finite Element Method) MEMS model, COMSOL cluster computing offers a step forward in increasing the complexity and reducing computation times of the FEM model. 150 151 EXPERIMENT #904 HIGHLIGHTS Industry Sector: Energy Improvement of the remote expert Country: Spain system based on software OTEA Software used: Cassandra, R, In house code ORGANIZATIONS EcoMT, Spain, is an ICT company that develops and integrates remote control and monitoring of facilities. CESGA, Spain, is the centre for computing, high THE SOLUTION performance communications systems and advanced services. GOMPUTE, Sweden, a leading simulation and EcoMTś intention is to address maintenance management by using ML, through the HPC solution provider. ITMATI, Spain, is a Technological analysis and processing of data from hundreds of devices and field sensors. This enables Institute for Industrial Mathematics. the anticipation of failures and the scheduling of predictive maintenance. These lead to the optimization of processes, the minimisation of downtime and costs and the more efficient End User Application Expert HPC Expert HPC Provider use of energy. Due to these needs, the OTEA remote expert system (OTEAres) was developed as an expert remote management system with an aim for energy optimization through the premature detection of incidents based on historical records. HPC infrastructures offer the required capabilities to support these services and to provide scalable resources when required. www.ecomt.net www.itmati.com www.cesga.es www.gompute.com THE CHALLENGE Heating and cooling account for a half of the EU’s energy demand. New Heating, Ventilation BUSINESS IMPACT BENEFITS and Air-conditioning machines produce huge amounts of data. A deterministic system with brute force conditions was applied to detect anomalies and reduce energy consumption in The savings of clients connected to the OTEAres platform, • EcoMT clients can more than 3.000 installations worldwide, with 600.000 variables monitored resulting in 3 would be immediate, it enables decision support in real expect to save up to trillion records for 5 years. The challenge of this experiment was to replace the brute force time offering a 24/7 service through its control centre, 7% on their electricity solution with a superior machine leaning (ML) model which is numerically intensive and which is the department in charge of controlling the bills. needs HPC. proper functioning of the installations, remotely resolving incidents in the shortest time possible. • Implementation of control panels in The previous version is updated into a new ML model other facilities, with capable of predicting and organising a wider range of an estimation of €15 incidents. It is estimated an energy saving of 7% per million of revenue in 5 installation, a reduction of 30% in preventive maintenance years. visits and a reduction of 20% in corrective maintenance. • Managing current More than 20% of the incidents registered could be installations with corrected remotely without the need for a technician. OTEAres doubles the Predictive maintenance will prolong the life of customers’ annual income to HVAC machines and reduce the number that need to EcoMT. be replaced. EcoMT expects to double the number of controlled facilities in the short term by extending this service to all actual connected installations and to quadruple this number in the long term, with an estimation to produce 1 million euros of total annual revenue. 152 153 EXPERIMENT #905 HIGHLIGHTS Industry Sector: Energy Digester 2.0: Optimization of the Country: Spain anaerobic digestion process for Software used: biogas generation OpenFOAM ORGANIZATIONS NORVENTO is a company focused on renewable energies and manufacturing of wind turbines. ENERGYLAB is a Spanish (non-profit) private THE SOLUTION technology centre. UNIVERSITY OF MINHO is Application Expert and CESGA from Spain is the Before this experiment, there was no integrated model for the design of anaerobic digesters HPC Provider. considering flow and biological aspects, able to take advantage of HPC systems. A new open-source based solution that allows the simulation of flow (fluid dynamics) and biological aspects of an anaerobic digester was developed. Due to the scales involved in the process End User Domain Expert Application Expert HPC Provider (300 m3 for anaerobic digester and particles around 0,01 cm3), the numerical solution of this coupled problem is very complex. Highly refined meshes are needed, which results in large computation times. Consequently, the solution requires HPC to provide results in an affordable time which model the engineering processes. www.norvento.com www.energylab.es www.uminho.pt www.cesga.es THE CHALLENGE Anaerobic Digestion is the process by which microorganisms degrade organic matter in the BUSINESS IMPACT BENEFITS absence of oxygen, producing a methane-rich biogas stream that can be used for energy generation. Industrially this process takes place in large full tanks called anaerobic digesters. Developed tools allow Norvento to reduce volume and cost • 5% increase of The development of a computational modelling solution to simulate digesters and support of industrial anaerobic digesters, increasing the amount thermal and electrical the design and optimization of the energy balance was seen as a potential solution to the and quality of the produced biogas while reducing the production of development of advanced digesters. energy consumption. As a result, Norvento renewable Norvento’s biogas energy power plants are able to increase their electrical plants. and thermal production. For Norvento, a cost reduction in investment and maintenance as a direct outcome • More efficient of reduced size digesters, will improve and support a digesters will profitable and sustainable business, and favour new contribute to business lines, such as the implementation of on-site the reduction of digesters for cattle raising installations. EnergyLab greenhouse gas obtained a valuable knowledge of fluid-dynamics, HPC emissions. and simulation models. Potentially, new services focused • New services in the on energy efficiency, based on the HPC simulation tools EnergyLab portfolio developed, can be set up and included in the EnergyLab and growth expected service portfolio. The University of Minho developed the by 30% in total sales. simulation solver based on OpenFOAM and acquired valuable insights in a completely new area for them: the simulation of processes involving chemical reactions. The knowledge obtained and the developed tools may be relevant for other industrial applications. 154 155 EXPERIMENT #906 HIGHLIGHTS Industry Sector: Aerospace Cross-solver Cloud-based tool for Country: Italy aeronautical FSI applications Software used: SU2, RBF Morph, OpenFOAM ORGANIZATIONS RINA CONSULTING SPA is an Italian engineering consulting company providing high-quality tailored solutions. University of Rome and the National Technical University of Athens provided THE SOLUTION expertise in numerical optimization and mesh morphing. CNR, the largest public research organization in Italy, provided The RBF4AERO platform is a toolbox which tackles all the aspects related to aircraft expertise in cloud technology, with the IAC research group. numerical design and optimization. It does this by making the CFD model parametric through CINECA is an HPC provider and host centre. an innovative shape optimization tool based on a high-performance meshless morphing technique. End User HPC & Application Expert Application Expert HPC Provider The experiment performed the successful adaptation of the RBF4AERO platform to work on Cloud-based HPC. This adaptation was applied to the optimization of a component of a real aircraft using computational models relevant to the aeronautical sector. The Consortium benefitted from the technical collaboration with Piaggio Aerospace, which provided the www.rina.org www.cnr.it www.epu.ntua.gr www.web.uniroma2.it www.hpc.cineca.it models to set-up the optimization of the shape of the P180 Avanti EVO vehicle. THE CHALLENGE For an aircraft production cycle, the ability to reduce production costs decreases rapidly BUSINESS IMPACT BENEFITS during the first phases of the project. In fact, 80% of the development cost and 70% of the life cycle cost of a product are determined during its conceptual design phase. Considering Improving aircraft performance has a clear economic • RINA CONSULTING this, the need to reduce the research and development (R&D) investigation time is significant. value. In this experiment the RBF4AERO platform was estimates an Consequently, powerful design tools are essential to shorten the time to acquire physical used on an industrial case demonstrating its capability to increased turnover of knowledge about the aircraft components and so to reduce production costs. solve a wide range of real world aeronautical aero-elastic €100.000 one year optimizations in a reliable and cost-efficient manner. after the finalization This achievement was realised through the generation of the experiment. of a computational mesh aligned with the requirements dictated by a worldwide mid-size business jets stakeholder, • The service will allow the generation of consistent shape modifications of RINA to lower its industrial significance and the demonstration of the current hourly rate by feasibility of a fully coupled fluid-structure interaction up to 10%. computation with CAE models characterised by non- • NTUA estimates an matching external surfaces. increase by 5% in the RINA CONSULTING estimates an increased turnover of revenues. €100.000 one year after the finalization of the experiment and a total accumulated of at least €400.000 over three years. RINA estimates an increase of at least 2 qualified jobs over three years. The service will allow RINA to lower its current hourly rate by up to 10% and so reach new clients. NTUA estimates an increase by 5% in the revenues coming from FSI related projects in the next three years. 156 157 EXPERIMENT #907 HIGHLIGHTS Industry Sector: Data Analytics Large-scale indie gaming analytics Country: Spain (LIGA) Software used: In house code ORGANIZATIONS KUMO is a design, engineering, and innovation-led SME applying disruptive technologies across both hardware and software in the areas of 3D technologies and digital THE SOLUTION asset creation and management. CNR is the largest public research organization in Italy and a top-level R&D LIGA has provided KUMO with two advanced services: Smart Retail Recommendation performer in Europe. CINECA is the HPC provider and Engine (REC-ENG), that analyse user preferences and activities and 3D Content Analytics host centre. Services & Dashboard (3D-DASH) that allows content providers and game developers to understand better trends in games. LIGA helps KUMO to meet its goals by monitoring the End User Domain Expert HPC Provider popularity of game entities on social networks. Currently, LIGA infers entities’ popularity by hourly monitoring of STEAM, a digital distribution platform which offers online multiplayer gaming. At the end of July 2018, LIGA stored 25 million entries in its database, describing the popularity of game entities among players. HPC techniques are needed to store and process efficiently this huge amount of data in a timely manner. www.3dkumo.com www.cnr.it www.hpc.cineca.it THE CHALLENGE The videogaming market is a complex arena with an increasingly crowded marketplace where BUSINESS IMPACT BENEFITS dozens of games are released every day. The motivation for this experiment is that exploiting the huge amount of data available in gaming web portals and social networks may lead to The Smart Retail Recommendation Engine REC-ENG is • Increased KUMO value added services for the benefit of game developers that need to understand market meant mainly for internal KUMO use. The main benefit for turnover in the range trends, and to companies targeting gamers. KUMO, end user in this experiment, is interested the customers of KUMO making use of REC-ENG is that of 75.000 - 150.000€ in particular in increasing its penetration into the gaming-related market, with a focus on 3D the new tool will support their decision-making process in per year. printable models. terms of models to be collected and shared. The tool will reinforce the links between models and users’ likes, and • Approximately 10% of also among models themselves. yearly growth rate in next 2 years. The 3D Content Analytics Services & Dashboard 3D-DASH is meant both for internal use and as a service to be sold • Increased number of to videogame-producing brands. The main benefit for technological transfer the brands customers of KUMO making use of 3D-DASH contracts of CNR. is that the new tool will support their decision-making • New customers in process in terms of designing or characterising new the HPDA market for models and 3D characters in their games. CINECA. The experiment will also help KUMO to expand in other sectors interested in or dealing with 3D models, like engineering (automotive, energy, etc.), culture (museums, artists, etc.), multimedia (animation, film producers, etc.). 158 159 EXPERIMENT #908 HIGHLIGHTS Industry Sector: Automotive Massively parallel virtual testing of Country: Germany safety-relevant driving systems Software used: OpenDrive, TensorFlow, OpenScenario, OpenCV ORGANIZATIONS SPICETECH GMBH focuses on data-centric IT solutions, machine learning and data science. XLAB d.o.o. is focusing on distributed systems, Cloud computing, THE SOLUTION information visualization and image processing. HLRS is the HPC centre of the University of Stuttgart, Germany, Test and validation aspects will have to be solved in multi-dimensional parameter spaces. and was the HPC provider in this experiment. The proposed approach allows the user to freely choose dimensions and parameter-spaces, which describe the test scenarios. For example, for a car arriving at a road intersection, a potential validation task would be to ensure a correct classification of objects (cars, End User & ISV HPC Expert HPC Provider pedestrians) interacting at the intersection. Just positioning them and adjusting their speeds yields 10-15 dimensions of test space. The VALICY framework uses High-Performance Computing (HPC) and Artificial Intelligence (AI) in order to quickly scan these dimensions for object classification errors. www.spicetech.de www. xlab.si www.hlrs.de THE CHALLENGE Various branches of the high-technology industry are experiencing strong increase in the BUSINESS IMPACT BENEFITS complexity of their products. In particular, for products and systems, which act autonomously in safety-critical environments, e.g. for autonomous cars in urban traffic scenarios, the cost The VALICY framework, developed in the course of • Parameter variation for validation and certification is rising sharply. Virtual validation will be a key methodology this experiment, offers the opportunity to integrate a runs can include to face such challenges, due to the fact that enormous amounts of test cases must be highly complex virtual test and validation task into the 10-15 testing considered covering a huge testing and validation space. product development process. VALICY can be integrated dimensions. into classical milestone-based development but, more interestingly, into iterative and incremental processes. • Thousands of Hence, high opportunities for fast test and validation validation runs can feedback arise in fields, where software and hardware run in parallel by development move forward together. using HPC processing power. The proposed approach samples high-dimensional test spaces efficiently, due to distributed parallel computations. • VALICY decreases the Furthermore, the Machine Learning (ML) component validation cost by an ensures that after a short period of adjustments to identify estimated factor of critical cases, only the scenarios that ML expects to be of hundred. high relevance are tested and validated. The VALICY framework is valuable for industry sectors that struggle, not only with high system complexity but also with new norms and regulations. Appropriately designed test and validation spaces can be created to support early and rapid knowledge of the certification process over a broad range of variations. 160 161 EXPERIMENT #909 HIGHLIGHTS Industry Sector: Food Manufacturing The Virtual Engineer 2 – Cloud-based Country: optimisation of production in the United Kingdom dairy sector Software used: The Virtual Engineer #2 using MATLAB ORGANIZATIONS MOODY DIRECT is a British SME that manufactures bespoke components both manually and with the latest CNC machinery to support servicing of OEM THE SOLUTION machines in a range of industries. THE UNIVERSITY OF PORTSMOUTH was the domain expert in this experiment. This experiment developed predictive maintenance software as a service solution to Moody EPCC provided HPC resources and expertise. for machines at Müller UK & Ireland Ltd. Data collection, management and transfer was developed by the University of Portsmouth and supported by EPCC for full analysis in near real-time direct from the production site. A GUI client was developed for Moody Direct Ltd’s End User Application Expert HPC Provider mobile engineers that connects seamlessly to the HPC system showing real time information to aid decision-making. HPC was exploited for the high-performance data analytics, resulting in a predictive maintenance platform (Software as a Service) for users in a Cloud-based online system. www.moodydirect.com www.port.ac.uk www.epcc.ed.ac.uk THE CHALLENGE The Dairy industry is a low-margin, high turnover industry. Each machine can break down BUSINESS IMPACT BENEFITS up to 3-5 times per year with a breakdown lasting up to 6 days at a time, and penalty fines can cost the customer up to €380.000 per day. Moody Direct delivers a maintenance service Moody Direct now has access to a working online platform • Savings for dairy into production sites. Current predictive maintenance measures are not comprehensive and that gives an oversight of the machines at the Müller manufacturers in not multi-machine, multi-site. Moody wants to offer enhanced Service Agreements with UK site. Currently the system is only streaming data from 1 excess of €50 million customers to reduce downtime and achieve increased sales through supply and manufacture machine, but there are 3 other machines onsite that can as penalty fines are of spares. in future share their data with the online platform. Moody avoided. Direct can offer an enhanced service-based business model which is anticipated to bring better management of • Increased productivity the scheduling of engineers and control of stocks of spare of approximately parts. The platform will also improve the capability and 200.000 bottles per knowledge of the Engineers and increase their productivity machine per year. as a result. • Reduction in For Müller, the benefits are increased productivity and environmental energy minimisation due to improved machine health. contamination and There will also be a reduction in food waste as machines food waste, up to continue to run successfully without the need to dispose 200.000 bottles of of lost milk during machine failure. milk saved from disposal per machine. For the University of Portsmouth, the experiment has led to the generation and transfer of knowledge within its Operational Research and Artificial Intelligence academic community and an improved knowledge of big data analytics and parallel processing through HPC. 162 163 EXPERIMENT #910 HIGHLIGHTS Industry Sector: Maufacturing OptiBike robust lightweight Country: Spain composite bicycle design and Software used: optimization Optimus, Code_Aster, MxNet ORGANIZATIONS IDEC is an SME, focused in composites and new materials. Its main strength is the development and industrialization of Advanced Composite Structures. THE SOLUTION UNITO is an Italian university from Turin, that provided HPC expertise together with NOESIS, who is also an The experiment focuses on providing SMEs with a user-friendly service to optimize the engeneering partner. ARCTUR, the Slovenian HPC centre, configuration of the layers of a composite material part, returning, in a reasonable timeframe, is an HPC Provider. the best performing orientation of the layers, that respects manufacturing constraints. In order to speed up the process, this service is deployed on an HPC Cloud infrastructure and it End User HPC Expert HPC Provider leverages machine learning algorithms to exploit and enhance the expert’s knowledge about performance and manufacturability. Once defined the best-performing configuration, the service is able to analyse the robustness of the optimal design. The approach is validated on a use case that involves the simulation of stresses and deformations on a composite bike frame. www.idec.aero www.noesissolutions.com www.en.unito.it www.arctur.si THE CHALLENGE Composite structures, especially those incorporating carbon fibres, are much more complex BUSINESS IMPACT BENEFITS than isotropic metal alloys, as they require lengthy development, significant knowledge and fine tuning. Building prototypes to test various configurations may take years to reach an The solution leads to a reduction of 80% in the time to • The end user optimal structure. Numerical simulations significantly reduce the time and effort required, but design and optimise a bicycle that can currently take up has reduced the modelling software tools and HPC infrastructures represent a large investment and solving to 8 months. In addition, it reduces the number of the development time by optimization problems with hundreds of parameters require highly skilled engineers. manufactured prototypes by 75%. Having this optimization 80% and number of workflow available, means that IDEC is capable of staying physical prototypes ahead of the competition despite its limited resources. by 75%. This is particularly relevant in volatile sectors such as competition bike manufacture. The reduced development • €45.000 saved by time, will also allow IDEC to react more quickly to the end user per bike modifications of the design. frame model. On the other hand, Noesis will be able to provide this • 9 new customers composite optimization service to customers looking expected by the end for a general-purpose solution for their composite parts. user within 5 years Moreover, the methodology on which the service is based, after the end of the enriches Noesis’ offer of engineering services and is project. meant to attract customers that are looking for a tailored solution for very high-end composite parts. Presence in the Fortissimo Marketplace will allow Noesis to reach out to a number of customers that are not yet in its sales network. 164 165 EXPERIMENT #911 HIGHLIGHTS Industry Sector: Automotive HPC Cloud-based simulation of coupled Country: Slovenia electromagnetic and structural- Software used: acoustics in in-wheel electric motors Altair Flux, GNU Octave, Altair Hyperworks ORGANIZATIONS ELAPHE PROPULSION TECHNOLOGIES LTD is a Slovenian SME. It is among the most experienced companies in the field of development of in-wheel motors THE SOLUTION for electric and hybrid vehicles. CINECA is participating as an HPC provider and host centre in this experiment. The main result of this experiment is the workflow developed for an optimization and automated coupled multiphysics computation of a detailed acoustic noise map of in-wheel motors which was developed from scratch. This new methodology, which benefited from the Cloud-based HPC was implemented and evaluated on the next-generation prototype of End User HPC Provider Elaphe M700 and L1500 in-wheel motors. The overall noise level of the motors improved by a minimum of 20 dB(A) and up to 60 dB(A) over the whole in-wheel motor operating range. www.in-wheel.com www.hpc.cineca.it THE CHALLENGE Despite delivering several benefits, the unsprung electric in-wheel propulsion technology BUSINESS IMPACT BENEFITS is still being considered as inferior to chassis mounted electric propulsion. This is true in particular for design process issues related to the optimisation of noise, vibration and The market preference for green products is growing • Estimated yearly harshness (NVH) characteristics, where complying to EU regulation No. 540/2014 on the and environmental policies are increasingly in favour savings of €120.000-sound level of motor vehicles is crucial obtaining an optimized NVH-acceptable design for of sustainable, highly-efficient electric propulsion. With €150.000 for end the product’s industrialized version is also critical to reach a production process with less lower noise levels, electric vehicles will improve the user in development/ physical prototype assembly and measurement. urban soundscape for many people, mitigating sound design process. pollution at all hours. On the other hand, lighter weight and the race for efficiency improvements under strict • End user expects also conditions is driving an increase in the efforts of vehicle up to 80% shorter designers and manufacture in studying different vehicle time-to-market and concepts, with NVH characteristics being one of the at least 80% faster most important aspects. A holistic approach requiring introduction of NVH - coupled electromagnetic (EM) and structural-acoustic related improvements (SA) simulations is necessary to optimize sound levels to products. while preserving the required performance and other • The optimized imposed design features. During this experiment, the use design is expected to of HPC substantially decreased the simulation time and increase Elaphe sales improved the quality of obtained results. This resulted in a by 20%. substantially reduced development cycle. The process was completed 10-times faster and at the same time enabled more complex simulations, directly resulting in improved products for Elaphe. 166 167 EXPERIMENT #912 HIGHLIGHTS Industry Sector: Aeronautical Multiphysics icing analysis using Country: Germany HPC Cloud Software used: UderEis ORGANIZATIONS MT PROPELLER is a German SME and a leading manufacturer of propellers for small aircraft. NORTHERN NUMERICS LTD is a UK SME specializing in simulations THE SOLUTION of freezing phenomena for the aeronautical, automotive, naval, construction and process industries. ARCTUR is a The Northern Numerics UderEis software offers some of the most advanced physical models Slovenian SME, participating as an HPC provider. for simulating ice accretion on aircraft components. UderEis is HPC capable and cut down the MT Propeller simulation times to less than a day. UderEis also allowed the simulation of ice accretion on rotating propellers. Using the computational resources of the Arctur HPC End User ISV HPC Provider Cloud it was possible to perform ice accretion analysis on a rotating propeller as installed on a small aircraft. A user-friendly GUI was developed, with detailed online tutorials for prospective users of UderEis to facilitate software adoption. The UderEis on-demand flexible licensing scheme is very competitive compared to licensing schemes of larger vendors. www.mt-propeller.com www.northernnumerics.com www.arctur.si THE CHALLENGE MT Propeller aims to improve the design of ice protection systems for its propellers, shorten BUSINESS IMPACT BENEFITS simulation times from one week to less than a day. This allows MT Propeller to incorporate the ice accretion simulations more efficiently in its design cycles. Increased computational The dramatic reduction of ice accretion simulation • Reduced costs resources through HPC uptake would allow MT Propeller to design more complex propellers times and the ability to analyse ice protection system for simulations of for larger aircraft and analyse ice protection systems performance as installed on the aircraft. performance on rotating propellers as installed on aircraft multiple €10.000. increase the reliability of the ice protection system and result in multiples of €10.000 savings for MT Propeller. • Reduced simulation The shorter turnaround times on HPC platforms and times from one week correspondingly more efficient design cycles lead to to less than a day. a substantial reduction in the labour costs for design. • Use of much more Moreover, the new ability to analyse larger and more realistic physical complex cases for larger aircraft allow MT Propeller to models by the end explore a new market and potentially increase their sales. user. As for Northern Numerics, the availability of its products • Reduced HPC costs on the Fortissimo Marketplace and increased visibility for the HPC expert will allow it to attract more business. Northern Numerics by outsourcing conservatively expects to double its sales volume. computational Both MT Propeller and Northen Numerics benefit from resources to the outsourcing computational resources and management Cloud. to the Fortissimo Cloud on demand, with corresponding savings in the higher margin. Arctur aims to increase its SAAS and PAAS sales by offering a new product in its portfolio, with annual sales increase. 168 169 EXPERIMENT #913 HIGHLIGHTS Industry Sector: Microfluidics Cloud-based multiphysics simulation for Country: France designing highly dynamic and highly accurate Software used: flow controls for microfluidic applications OpenFOAM, OpenModelica ORGANIZATIONS ELVESYS, from France, is an SME that commercializes the world’s widest brand of microfluidic flow control products. GOMPUTE, from Sweden, is a leading THE SOLUTION simulation and HPC solution provider. ITAINNOVA from Spain, is a public non-profit Technology Centre, working The solution is based on the development of simulation models that can be used both for in innovation and technological development. testing the virtual system and for controlling the real one. These models are used in two ways: as advanced control strategies to anticipate the behaviour of the system, mitigating the effect of sensitivity and dead volumes and as a soft sensor, providing flow characteristics not End User ISV & Application Expert HPC Provider & Expert available by sensing, such as flow depths, interface location or shear stresses. The development must involve HPC for Computational Fluid Dynamics (CFD) simulations of the microfluidic chip in service. Batches of CFD simulations need to be run in the cluster to build real-time simulation tools based on Reduced Order Models (ROMs). www.elveflow.com www.itainnova.es www.gompute.com THE CHALLENGE Microfluidics is both the science that studies the behaviour of fluids through sub-millimetre BUSINESS IMPACT BENEFITS micro-channels, and the technology for manufacturing its components. The control of complicated as the chips are sensitive to external variations; systems have dead fluid The main business achievement is a methodology to • Improved presales volumes slowing down the reaction time; and the most convenient sensors are not always develop accurate model-predictive-control strategies in argumentation available. The experiment aims to generate real-time computer simulation that mimics the virtual and real microfluidics systems. The information yielding a 10% yearly physical device in order to allow the best dynamic controllability, both in models and real provided by the simulation workflow will allow a increase in pressure world. continuous improvement of the value chain. controller sales. Elvesys will offer easy access to microfluidics simulation • Reduce the presale for their customers by a Cloud-based virtual lab that cost associated with will facilitate the selection and purchase of microfluidic device lending and devices. Also, Elvesys gets a deep insight into the demonstration. performance of the products at a component and system level. An improvement from 5 to 10% of the profit margin • Reduce the discounts per offer and access to a greater share of the Total thanks to increased Addressable Market (TAM) (projected to reach $112 billion understanding of the by 2023) are expected. product line. ITAINNOVA will enhance and renew the portfolio of services for its clients. It demonstrated the impact of real-time simulation tools on manufacturing sectors beyond this experiment. Itainnova will offer s development of new similar simulation procedures as a consultancy service for SMEs, focusing on mechatronic setups. 170 171 EXPERIMENT #914 HIGHLIGHTS Industry Sector: Microfluidics HemoSphere – Prototyping of Country: France blood-processing microfluidics in Software used: the HPC Cloud HemoSphere ORGANIZATIONS AENITIS TECHNOLOGIES S.A.S. is a French microfluidic company that is developing a blood sorting device. ELECTRIC ANT LAB B.V. is a Dutch software company THE SOLUTION that specialises in complex fluid simulations. SURFSARA is the operating company for the Dutch national super A simulation model of the microfluid device of Aenitis was developed which is much truer computer, HPC centre Cartesius. to reality than previously used models. This was achieved by explicitly incorporating the generation of a soundwave in the system. Rather than assuming that the soundwave induces certain forces, the model allows the acoustic forces to develop naturally from the interaction End User ISV HPC Provider of the soundwave with the particles (blood cells) present in the microfluidic channel. In this way, more insight and understanding of the processes in the microfluidics device could be generated. www.aenitis.fr www.electricant.com www.surfsara.nl THE CHALLENGE Electric Ant Lab (EAL) has developed HemoSphere for the detailed simulation of complex BUSINESS IMPACT BENEFITS fluids. This technology can be applied to prototyping microfluidic chips. The microfluidics industry is a young medical technology and equipment field where small-scale phenomena The use of simulations in the development of • Savings between are difficult to observe. In this field there is a large demand for more cost-effective and acoustophoretic microfluid devices as developed at 20.000€ and 25.000€ methods as opposed to trial-and-error process for prototyping new chips. The challenge was Aenitis can significantly reduce development costs. In for a single prototype. to replicate the behaviour and small-scale phenomena of an acoustophoresis chip into a the laboratory, a single microfluidics geometry needs to browser-based simulation tool. undergo a large series of tests to assess its performance. • Time saving: With the same approach in HemoSphere, these experiments in the development iterations can be run faster. More information physical lab can take about the fluid behaviour will also be generated, as these about 4-6 months, phenomena are not easy to research visually. In the where the virtual laboratory only one experiment iteration at the same time experiments can be can be achieved. In HemoSphere multiple experiments can performed in a matter be executed at the same time. of days or weeks. For microfluidic companies this approach can create a tremendous impact on the time spent in the development and prototyping stage of the microfluidic chip design process. It also requires less resources in terms of produced physical prototypes and the usage of human blood samples. 172 173 EXPERIMENT #915 HIGHLIGHTS Industry Sector: Renewable Energy Cloud-based simulation of Country: Greece desalination systems powered Software used: by renewables TRNSYS, GenOPT ORGANIZATIONS PHOTOVOLTAIC is a Greek SME, providing innovative solutions in water desalination coupled with renewable energy. THE AGRICULTURAL UNIVERSITY ATHENS focuses research THE SOLUTION in agro-nutrition, environmental ecology and agricultural economics. ARCTUR participated as an HPC Provider. The major innovation introduced by the experiment is the simultaneous optimization of VERTOYO from Greece, is an expert for implementation of both the desalination system and the renewable energy system, along with the tuning of integration of advanced technology services. the control system. Currently, the companies operating in the design and implementation of RE-Desalination facilities need to address requirements of both the desalination process End User Domain Expert ISV HPC Provider and the RE supply. This is a very broad spectrum of required skills and results in installations that operate in a highly unoptimized way. The main contribution of the experiment to the engineering and manufacturing actors, is the development of a design and sizing platform based on optimizations through simulations for RE-Desalination systems leading to systems providing lower cost of water and improved efficiency. www.photovoltaic.gr www2.aua.gr/en www.vertoyo.com www.arctur.si THE CHALLENGE Based on the current industrial state of the art, the design process of RE-powered BUSINESS IMPACT BENEFITS desalination systems takes place in two distinct steps. Step 1 includes the design of the reverse osmosis desalination system and Step 2 includes the design of the renewables The solution (named DESAL) is expected to tap into • Depending on the system based on the desalination unit designed in Step 1. This approach using two distinct the growing desalination market and manage to get an individual case steps leads to non-optimized systems. important share as far as design and sizing software is the new approach concerned. PHOTOVOLTAIC will be using DESAL in order leads Photovaltaic to provide cost-effective turn-key solutions to customers to the design of worldwide and also through the network of the PHAESUN systems featuring associated companies, to which PHOTOVOLTAIC belongs a decrease in the and which has presence worldwide across four continents. water production cost Moreover, the expected collaborations with 3rd party ranging from 5% to desalination and renewables SMEs, as well as private end over 40%. users and public sector users will further increase DESAL’s utilization and market share. • The new approach can increase production of water by more than 40% without changing core properties of the system. 174 175 ICT Innovation for Manufacturing SMEs www.i4ms.eu Seventh Framework Programme www.cordis.europa.eu/fp7 Horizon 2020 www.ec.europa.eu/programmes/horizon2020/ European Commission www.ec.europa.eu The research leading to these results has received funding from the European Union Seventh Framework Programme under grant agreement No 609029 and from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 680481. This publication does not represent the opinion of the EC and the EC is not responsible for any use that might be made of information appearing herein. For further information, please visit: www.fortissimo-project.eu Follow us on: www.linkedin.com/groups/5096901 www.twitter.com/fortissimopro www.fortissimo-project.eu 176