UDK 677.4:66.017	ISSN 1580-2949
Professional article/Strokovni članek	MTAEC9, 48(5)777(2014)
ASSESSMENT OF THE POST-IMPACT DAMAGE PROPAGATION IN A CARBON-FIBRE COMPOSITE UNDER CYCLIC LOADING
OCENA NAPREDOVANJA POŠKODBE PO UDARCU PRI PONAVLJAJOČIH SE OBREMENITVAH KOMPOZITA Z OGLJIKOVIMI VLAKNI
Daniel Kytyfi, Tomaš Fila2, Jan Šleichrt2, Tomaš Doktori, Martin Šperli
1Institute of Theoretical and Applied Mechanics, v.v.i., Academy of Sciences of the Czech Republic, Prosecka 76, 190 00 Prague 9, Czech
Republic
2Czech Technical University in Prague, Faculty of Transportation Sciences, Department of Mechanics and Materials, Konviktska 20, 110 00
Prague 1, Czech Republic kytyr@itam.cas.cz
Prejem rokopisa - received: 2013-09-30; sprejem za objavo - accepted for publication: 2013-11-11
Carbon fibre in polyphenylene sulfide composites (C/PPS) became a popular material in the aircraft industry but its fragility and low impact resistance limits its application in primary aircraft structures. This study is focused on damage propagation in the laminated composites reinforced with carbon fibres. The damage may be inflicted during the ground maintenance, by an inflight bird strike or during a flight in severe meteorological conditions (heavy storms). The initial damage was created by a drop-weight out-of-plane impact using a spherical indenter. The response of the material was analysed by monitoring the impacted zones and their propagation history. The influenced area and specimen thickness in the centres of indents were chosen as the degradation parameters. The post-impact damage propagation induced by cyclic loading was assessed using a custom-designed computer-controlled laser-profilometery device. Both the upper and lower profiles of the specimen were scanned during the interruptions of the fatigue test. Global deformation was described with an analytically determined centroidal-axis curve. Local topography changes were obtained with a subtraction of this curve. Surface-deformation maps were created and used for a demonstration of the damage propagation in the specimen.
Keywords: carbon-fibre composites, post-impact damage, laser profilometry
Ogljikova vlakna v kompozitih iz polifenilen sulfida (C/PPS) so postala priljubljen material v letalski industriji, toda njihova krhkost in slaba odpornost proti udarcem omejujeta njihovo uporabo v primarnih letalskih konstrukcijah. Ta raziskava se osredinja na napredovanje poškodbe na laminiranih kompozitih, ojačanih z ogljikovimi vlakni. Poškodba lahko nastane med vzdrževanjem na tleh, pri trčenju s ptico med letom ali med letom v hudih vremenskih razmerah (huda nevihta). Začetna poškodba je bila narejena z udarno pregibnim preizkusom s kroglastim vtiskovalcem. Odziv materiala je bil analiziran z opazovanjem območij udarca in potekom napredovanja. Prizadeto območje in debelina vzorca v področju vtiska sta bila izbrana kot parametra degradacije. Napredovanje poškodbe po cikličnem obremenjevanju po udarcu je bilo ocenjeno s po meri oblikovane računalniško vodene naprave za lasersko profilometrijo. Zgornji in spodnji profil vzorca sta bila skenirana med prekinitvami preizkušanja utrujenosti. Celotna deformacija je bila opisana z analitično določeno krivuljo težiščnice. Lokalne spremembe topografije so bile dobljene z odštetjem te krivulje. Ustvarjeni videzi deformacije površine so bili uporabljeni za prikaz napredovanja poškodbe na vzorcu.
Ključne besede: kompoziti z ogljikovimi vlakni, poškodba po udarcu, laserska profilometrija
1 INTRODUCTION	An application of a polymeric matrix lowered the
tendency towards brittle behaviour (common for car-
The design and safe operation of lightweight struc-	bon-fibre composites) and exhibited the advantages of
tures, especially in the aviation industry, is particularly	high chemical resistivity, insensitivity to moisture, good
important and challenging due to the inauspicious load	fatigue performance5,6 and recyclability.
spectra composed of a large number of low-amplitude	Micromechanical modelling of the composites with
cycles and sudden impacts1. Low-amplitude cycles are	imperfections7 sufficiently describes the degradation
caused by aerodynamic loads and engine vibrations.	process. However, the material models based on the
Wayward strikes may be inflicted during the ground	X-ray computed tomography of the specimen represent-
maintenance, by inflight collisions (bird strikes) or	ing the material at the macroscopic level including a
severe meteorological conditions (heavy storms).	complex microstructure could not be evaluated using the
The damage-tolerance approach commonly used in	finite-element simulations with the plasticity applied due
aerospace engineering requires a comprehensive know-	to the computational complexity and enormous memory
ledge of the material-degradation process and a reliable	requirements8. The presented work aimed to extend the
prediction of a structure safe life2. The thermoplastic	range of non-destructive testing (NDT) techniques com-
composites commonly used for these purposes allow an	prising the lock-in thermography9 or the modified-im-
application of an optimised manufacturing technology3 4.	pulse excitation technique10.
2 MATERIALS AND METHODS
2.1	Specimen description
The base material, a carbon-fibre/polyphenylene sulphide (C/PPS) composite manufactured by Letov le-tecka vyroba, s. r. o., was delivered as plates with a thickness of (2.5 ± 0.05) mm. The material consists of quasi-isotropic 8-ply carbon fabric with its volume fraction higher than 90 %, bonded with a thermoplastic matrix. The surface is covered with a thin glass-fibre cloth protecting the core against mechanical and chemical influences. The final specimens with a rectangular shape with the dimensions of 250 mm x 25 mm were cut from the plates using a water-jet cutter.
2.2	Initial damage
The first step of the experimental procedure was to inflict the initial damage to the specimens under controlled conditions. A drop tower designed within project SGS12/163/OHK2/2T/16 with the maximum impact energy of 50 J was used. The strike was carried out using a spherical indenter with a diameter of 20 mm and the energies of (10, 20 and 15) J on (30, 50 and 70) % of the length of the samples. The imprints of the diameter in the range of millimetres and the depth in the range of tens of micrometers then occurred.
2.3	Fatigue loading
For a life-cycle assessment the specimens were cyclically loaded using a Mikrotron (Russenberger Prüfmaschinen, AG) resonant testing machine (Figure 1). To ensure the loading at the chosen stress level (33 % of the tensile strength) the mean loading-force value of 6 kN
and the amplitude of 5 kN were set. A sinusoidal force was applied in the force-driven experiments.
Due to a relatively high testing frequency (approximately 75 Hz), the experiment was monitored with a thermal imaging camera SC7600 (FLIR Systems, Inc.). To prevent exceeding 50 % of the glass transition temperature the specimen temperature was held at maximally 60 °C. At the same time, the lower frequency limit was set in order to avoid a specimen rupture11. The fatigue experiment was interrupted six times at the predefined numbers of cycles to perform profile scanning.
2.4	Profile measurement
To obtain the information about damage propagation during the life cycle, a set of profilometery experiments was performed. A custom-designed scanning device equipped with laser scanner ScanControl LLT2600-25 (Micro-Epsilon Messtechnik) depicted in Figure 2 was used for this purpose. The device allowed us to measure the line profiles with the length of 20-40 mm, defined by 1024 measured points. The altitude resolution of the scans was 4 ^m. The scanner was mounted on a motor-ised computer-controlled single-axis linear stage with the minimum incremental motion of 10 ^m and the on-axis accuracy of ± 0.5 ^m. One scanning sequence took approximately 15 minutes.
2.5	Damage-propagation assessment
The changes in the impact depth, the sample thickness and the area of influenced zones were chosen as the degradation parameters. The automatic procedure for a surface reconstruction (Figure 3) and profile-change assessment was carried out using the tools developed in the MATLAB (Mathworks, Inc.) computational environment. The variable position of the samples in the scanning area required the use of the corner detection
Figure 1: Experimental device for dynamic loading
Slika 1: Eksperimentalna naprava za dinamično obremenjevanje
Figure 2: Custom-designed computer-controlled profilometery device
equipped with a ScanControl LLT2600-25 laser scanner
Slika 2: Po meri oblikovana računalniško vodena naprava za profilo-
metrijo, opremljena z laserskim optičnim bralnikom ScanControl
LLT2600-25
Figure 3: Reconstruction of the sample surface based on laser triangulation Slika 3: Rekonstrukcija povr{ine vzorca, ki temelji na laserski triangulaciji
Figure 4: a) Increase in the influenced zones, b) the maximum depth of the impact depression and c) the thickness of the sample plotted against the number of loading cycles
Slika 4: a) pove~anje obsega prizadetega obmo~ja, b) maksimalne globine udrtine ter c) debelina vzorca glede na {tevilo ciklov obreme-
algorithm based on the altitude threshold to detect the specimen boundaries in the captured data. Transformation functions were obtained and the objects were transformed into a unitary coordination system.
Divergence of the laser beam was taken into account for the real-altitude matrix estimation and the blur of the edges caused by the same effect was reduced with gradient filters. The curvature of the surfaces was not caused only by the local impact zones but also by the overall bending of the samples due to a combination of the initial impact damage and cyclic loading. A piece-wise continuous second-order curve (the centroidal axis) was fitted and set as a new reference level. Then the altitude matrices were updated. On the straightened surfaces, the local impacted zones were quantified (area, maximum depth) using the data-registration procedure. From the subtraction of the upper and lower profile, the change in the sample thickness was obtained.
3 RESULTS
Based on the reconstructed profiles from the laser measurements, the influenced zones were identified on the basis of thresholding. In the areas of interest, the impact depression depth and the local thickness were assessed. Propagation of the chosen degradation parameters on two selected samples for several distinct impact levels is depicted in Figure 4.
Damage propagation exhibits similar evolution on different tested samples. The most significant parameter was the maximum depth of the impact on the impacted side. The initial depth corresponds to the strike energy, while later the depth decreases with the increasing number of the loading cycles. The area of influenced zones grows with the number of the loading cycles but, surprisingly, the initial areas were not proportional to the strike energy. The area of damaged zones inflicted by lower energy impacts also showed a faster increase. The changes in the thickness of the samples due to the
influenced zones were negligible as the differences in the thickness were only two or three times higher than the noise.
4 CONCLUSIONS
The presented study describes the possibility of a time-lapse profilometery measurement for an evaluation of the post-impact damage propagation in a C/PPS composite under cyclic loading. The chosen parameters (the area of impacted zone, the maximum depth and the sample thickness) provide the information about damage accumulation in the material. Generally, laser profilo-metery is a suitable method for the NDT testing and evaluation of the surface damage. The described modified method is applicable to bigger components and structures. With respect to our measured data, the reliability of the method was reduced by the resolution of the available laser scanner.
Acknowledgements
The research was supported by Technology Agency of the Czech Republic (grant No. TA03010209), Grant Agency of the Czech Technical University in Prague (grant No. SGS12/205/OHK2/3T/16), research plan of the Ministry of Education, Youth and Sports
MSM6840770043 and by institutional support RVO: 68378297.
5 REFERENCES
1	R. Aoki, J. Heyduck, An Experimental Study of Impact-Damaged Panels under Compression Fatigue Loading, In: J. Füller et al. (eds.), Developments in the Science and Technology of Composite Materials, Elsevier Science Publishers Ltd., 1990, 633-642 1990, 633-642
2	J. P. Gallagher, USAF damage tolerant design handbook, Flight Dynamics Laboratory, Air Force Wright Aeronautical Laboratories, 1984
3Z. Padovec, M. Ružička, Mechanics of Composite Materials, 49 (2013) 2, 221-230
41. C. Finegan, R. F. Gibson, Composite Structures, 44 (1999) 2-3, 89-98
5	J. Minster, O. Blahova, J. Hristova, J. Lukes, J. Nemecek, M. Sperl, Journal of Applied Polymer Science, 123 (2012) 4, 2090-2094
6	J. Minster, O. Blahova, J. Lukes, J. Nemecek, Mechanics of Time-Dependent Materials, 14 (2010) 3, 243-251
7	M. Sejnoha, J. Zeman, International Journal of Engineering Science, 46 (2008) 6, 513-526
8	J. Vorel, J. Zeman, M Sejnoha, International Journal for Multiscale Computational, 11 (2013) 5, 443-462
9R. Montanini, Infrared Physics & Technology, 53 (2010) 5, 363-371
10	D. Kytyr, T. Fila, J. Valach, M. Sperl, UPB Scientific Bulletin, Series D: Mechanical, 75 (2013) 2, 157-164
11	M. Pirner, S. Urushadze, International Applied Mechanics, 40 (2004) 5, 487-505