Advances in Production Engineering & Management
Volume 13 | Number 2 | June 2018 | pp 125-135 https://doi.Org/10.14743/apem2018.2.278
ISSN 1854-625G
Journal home: apem-journal.org Original scientific paper
Optimization of process parameters for machining of Al 7075 thin-walled structures
Borojevic, S.a, Lukic, D.b, Miloševic, M.b, Vukman, J.b, Kramar, D.c*
aUniversity of Banjaluka, Faculty of Mechanical Engineering, Banja Luka, Bosnia and Herzegovina bUniversity of Novi Sad, Faculty of Technical Sciences, Novi Sad, Serbia cUniversity of Ljubljana, Faculty of Mechanical Engineering, Ljubljana, Slovenia
A B S T R A C T
A R T I C L E I N F O
The aim of this paper was focused on research in order to improve the manufacturing of aluminium alloy thin-walled components through the optimization of milling process parameters. The methodology for optimization of milling parameters is developed and presented. The influence of the tool path strategy, wall thickness and feed rate on the machining time, dimensional accuracy deviation, shape and position accuracy deviation, and surface roughness in the case of line-type thin-walled parts machining were analysed. Based on the analysis of experimental results, the corresponding empirical models of responses were identified. Optimization of results was conducted using response surface methodology. Verification of optimization results was executed using two additional experiments. The results from experimental verification show a satisfactory matching with calculated optimal values. The basic scientific contribution of the paper relates to the development of a methodology for optimization of machining parameters for milling of thin-walled structures of aluminium alloy using an ANOVA method, Central Composite Design experiment and empirical modelling. Practical implications are related to the correct selection of the tool path strategy and feed rate value for machining of thin-walled aluminium components in order to achieve the required output techno-economic effects.
© 2018 PEI, University of Maribor. All rights reserved.
Keywords:
Thin-walled structures; Aluminium alloy Al 7075; Optimization;
Response surface methodology; Machining process parameters; Milling
*Corresponding author: davorin.kramar@fs.uni-lj.si (Kramar, D.)
Article history: Received 26 March 2018 Revised 28 May 2018 Accepted 29 May 2018
1. Introduction
Due to its homogeneity, corrosion resistance, and excellent ratio between load capacity and weight, thin-walled aluminium components are increasingly being used as structural parts in the aerospace, automotive, military and other branches of electrical-mechanical industry [1, 2]. New designs of these components are focused on monolithic parts rather than a larger number of components that require to be assembled after machining. In this way, the result is a design with good mechanical characteristics, better quality and accuracy, lower weight and less production time and costs [3]. The design of thin-walled structures can vary from simple line and rectangular form to more complex geometric forms [4].
The manufacturing process of thin-walled parts is mainly performed by removing the material from blanks, in a large percentage from their original volume and mass, making it a time-consuming and challenging process [5, 6]. Consequently, machining demands are directed to high productivity, causing in turn the occurrence of issues related to machining vibrations and elastic deformations of thin-walled structures that affect the surface roughness, as well as di-
125
Borojevic, Lukic, Milosevic, Vukman, Kramar
mensional and shape accuracy [7, 8]. In addition to this, permanent deformations of the structure can occur, and this may cause the incidence of rejected parts [9, 10]. Numerous factors of process planning are affected by the aforementioned issues, such as the elements of the machining system (machine tool, tool, and fixture), cutting parameters, tool-path strategy, cutting fluids, etc. [11-14].
In parallel with the increasing market demands for aluminium alloy thin-walled components, there is a number of accomplished researches focused on optimizing the structure in order to reduce mass, deformation, and vibration, i.e. increase load capacity and strength. Also, researches are focused on optimization of process planning of these components aiming to reduce the machining time and costs, and increase the accuracy of shapes and positions, and improving surface roughness, etc. [3, 4, 15, 16].
Novak-Marcincin et al. [17] investigated the possibility of a quick analysis of the appropriate tool path in the process of milling different features (rectangle, circle, L-shape and T-shape). Based on the obtained results, a software solution is developed to assist in the selection of optimal milling strategy on the basis of the minimal processing time. Pandian et al. [18] analysed the influence of the cutting forces on the deformation during the machining of thin-walled parts which results in different wall thickness, larger at the top and smaller in the root. Authors applied the developed Artificial Neural Network (ANN) and Particle Swarm Optimization (PSO) for prediction of the cutting force during machining of thin-walled parts. Msaddek et al. [19] represented the methodology for the optimization of machining strategies implemented for machining complex shaped pockets. For this purpose, authors developed the analytical model which is used for modelling feed rate value regarding the tool path strategy. Denkena and Schmidt [20] defined an improved simulation model which predicts a form of unsuitable deviation occurring during the milling of the ends of thin-walled structures. Their paper analyses the effect of cutting forces that causes the deflection of thin-walled structures and the occurrence of errors on machined parts. Peripheral milling of thin-walled structures is modelled in Budak and Altintas [21]. The developed model provides satisfactory predictions of cutting forces and dimensional surface defects throughout the deflection of cutting tool and thin-walled structures during the milling process. Baranek et al. [22] emphasized the influence of material removal on the surface roughness of thin-walled structures during the milling process. Experimental research on the stability of milling the thin-walled aluminium structure, as well as the measurement of cutting forces and acceleration, was conducted in Rusinek et al. [23]. The analysis was performed through the stability diagrams and recurrent quantification analysis. Kanchana et al. [24] used the finite elements method approach and the analysis of frequency response for machining of thin-walled components for aerospace industry purposes. Arnaud et al. [25] investigated the stability of machining the thin-walled aluminium structures. The system was modelled using a dynamic mechanistic model, while the dynamic system parameters were obtained using the finite element analysis. Masmali and Mathew were used an analytical approach to predict dynamic chip thickness variation during machining of thin wall workpieces [26]. Pare et al. [27] implemented me-taheuristic methods for selecting optimum process parameters in high speed CNC end-milling of composite materials. They compared Gravitational Search Algorithms (GSA) with Genetic Algorithm (GA), Simulated Annealing (SA) and Teaching-Learning-Based Optimization (TLBO) methods. Ratchev et al. [28] represented a virtual environment for the simulation and prediction of the deviation of the thin-walled structures during the machining process. The model for material removal was based on an iterative voxel transformation algorithm.
Taking into consideration the conclusion from listed literature above, it can be concluded that the field of machining the thin-walled structures has been the research subject around the world for many years now. It was identified that there is no universal model or unique approach that could be implemented in this field. The aforementioned researchers were oriented to individual input/output parameters of process plan for the machining of thin-walled parts, with different techniques applied, such as simulation methods and models, predictive models, stability diagrams, finite element method, and so on. The main goal in this research relates to the development of the methodology for optimization of machining parameters for thin-walled structures. In this way, the focus is put on the improvement of process planning function which currently
126
Advances in Production Engineering & Management 13(2) 2018
Optimization of process parameters for machining of Al 7075 thin-walled structures
represents a bottleneck in the production environment, and the integration of product design and manufacturing. The methodology is based on the simplicity and generality of application with a quick and easy way of reaching reliable data for the process planning for machining of thin-walled parts. It can easily be applied to other shapes of thin-walled structures, other materials, with the inclusion of other input/output parameters of machining process.
The primary hypothesis of this research was that the change of a tool path strategy and cutting conditions of aluminium alloy thin-walled structures affects the machining time, surface roughness and dimensional, shape and position accuracy of workpieces. According to the given hypotheses, the aim of this research was focused on the selection of an optimal tool path strategy, wall thickness and feed rate value of line type aluminium thin-walled structures for the best machining responses. The target responses which were selected include: the shorter machining time, the smallest wall thickness, perpendicularity, flatness deviation and the smallest surface roughness.
2. Materials and methods
Optimization theory as a scientific discipline is more frequently used in solving engineering problems, where, on the basis of criteria and the appropriate optimization method, the best solution is being found for a selected object of the optimization in line with machining requirements. The main phases in the proposed experimental procedure of the process parameters optimization methodology for the machining of aluminium alloy thin-walled structures are shown in Fig. 1 and described below. Optimization methodology was performed using ANOVA method within Design Expert Software [29].
Fig. 1 Flowchart for the optimization methodology
2.1 Defining the optimization task
The object of the research was the line-type aluminium alloy thin-walled structure machining, having wall thickness a from 0.5 mm to 1.5 mm, which is moderately low compared to wall height L (L:a = 30:1).
Advances in Production Engineering & Management 13(2) 2018
127
Borojevic, Lukic, Milosevic, Vukman, Kramar
v* \ Tool path strategy J>
v £
icss(a)^
Wall thickness (
- / Feed rate value {f)
\ IllAblAh
|¿7|Ac|—-
Machining time (T)
»
■MXlAblAl
-MMi
Wall thickness deviation
«O:
Perpendicularity deviation (Ab) o v a.
Flatness deviation (Ac)
Surface roughness (Ra)
Fig. 2 Scheme of input parameters and responses
Basic optimization task refers to the determination of the influence of tool path strategy, wall thickness a and feed rate value f on machining time T, wall thickness deviation Aa, perpendicularity deviation Ab, flatness deviation Ac, as well as surface roughness Ra. This task was solved using the experimental optimization method, which scheme is shown in Fig. 2. The parameters T, Aa, Ab, Ac and Ra can be observed as measured outputs, i.e. responses.
2.2 Setting the experimental plan
The stage of setting the experimental plan was carried out using the Design-Expert Software. The software supports a large number of experimental designs, while for the present case the Central Composite Design was selected. Three factors, two numerical and one categorical all on three levels generated the total of 33 experiments. The first numerical factor was the thickness of the line-type thin-walled structure a, with the lowest limit of 0.5 mm, and the highest limit value of 1.5 mm. Second numerical factor was milling feed rate f, with the lowest limit at 150 mm/min, and the highest limit value at 350 mm/min. For the third, categorical factor, three different tool path strategies were chosen. Tool path strategies for machining the thin-walled structures were selected based on the recommendations from the literature [4, 18, 21]. These tool path strategies are as follows:
Tool path strategy No. 1 (Parallel spiral) I Tool path strategy No. 2 (Zigzag)
• Tool path strategy No. 3 (True spiral) ¡HI 2.3 Machining strategy selection
This phase encompasses the selection of the machining strategy in the terms of machining passes sequencing during machining process of thin-walled structures. Selected machining passes sequence for this experiment was adopted from [30]. Selected machining passes sequence implies that the first machining pass was executed at the left side of thin-walled structure, then the second and the third machining pass were executed at the right side of thin-walled structure. Next, the fourth and the fifth were executed again at the left side, and so on up to the twentieth pass at the bottom of the structure as shown in Fig. 3. Machining of thin-walled structure was performed with the following fixed parameters: depth of cut ap = 3 mm, step over distance ae = 8 mm for each machining pass and spindle speed n = 2000 min1.
K
Fig. 3 Selected machining passes sequence for the machining of thin-walled structures

128
Advances in Production Engineering & Management 13(2) 2018
Optimization of process parameters for machining of Al 7075 thin-walled structures
2.4 Experimental procedures
Realization of the experiments
Block samples (33 pieces) with dimensions of 70x40x40 mm from Al 7075 alloy (AlZnMgCu1.5) were chosen and prepared. This Al-alloy has very good mechanical properties, as well as high resistance to fatigue and corrosion, which makes it suitable for use in the milling of thin-walled structures. The chemical composition of this alloy is Al (87.1-91.4 %), Cr (0.18-0.28 %), Cu (1.22 %), Fe (<0.5 %), Mg (2.1-2.9 %), Mn (<0.3 %), Si (<0.4 %), Ti (<0.2 %), Zn (5.1-6.1%). Basic mechanical characteristics are: tensile strength (560 MPa), Rp0.2 (500 MPa), yield strength (7 %), and hardness (150 HBW).
CNC machining centre - EMCO Concept Mill 450, and the milling cutters with the identification R216.32-10025-AK32 from the Sandvik Coromant were used for the realization of the experiments. The positioning and clamping of the samples in the milling process were carried out using a pneumatic clamp with pre-set limiter.
Measuring of the responses
According to the experiment plan, the following responses of the thin wall structure machining process were measured:
•	Machining time - T,
•	Wall thickness deviation - Aa,
•	Perpendicularity deviation - Ab,
•	Flatness deviation - Ac,
•	Surface roughness - Ra.
The measurements of the dimensional parameters were performed using the coordinate measurement machine Zeiss Contura G2 (Fig. 4a) together with sensor touch probe Zeiss Vast XXT shown in Fig. 4b. The Maximum Permissible Error MPE = 1.9 |j.m for size measurement according to ISO 10360-2:2009 is specified for this machine.
Surface roughness measurements were conducted with the INNOVATEST contact device, in direction perpendicular to the cutter movement on both sides of the wall. The device has the ability of measuring the parameters Ra and Rz. The roughness measurements were performed with a stylus radius of 2 |j.m and resolution of 0.01 |j.m.
Fig. 4 Measuring of parameters Aa, Ab and Ac on machined parts by CMM Analysis of the responses and control of the limit values
At this phase, the intervals of responses values obtained from the measurements of all three tool-path strategies were analysed (Fig. 5). Limit values of responses were predefined as follows: maximal wall thickness deviation maxAa = 0.2 mm, maximal perpendicularity deviation maxAb = 0.1 mm, maximal flatness deviation maxAc = 0.1 mm and maximal surface roughness
Advances in Production Engineering & Management 13(2) 2018
129
Borojevic, Lukic, Milosevic, Vukman, Kramar
maxRa = 1.6 |j.m. The analysis of the limit value intervals of all responses shows that the values of wall thickness deviation Aa, perpendicularity deviation Ab, flatness deviation Ac and surface roughness Ra for tool path strategy No. 2 do not fall into the predefined limit values, causing the tool path strategy No. 2 to be excluded from further research. Tool path strategies No. 1 and No. 3 satisfied the limit values, and therefore both were taken into consideration for determining the optimal strategy.
2.000-I 1 77G~
lo 1.556-1 <
o 1.333-
% 0 889-œ
I 0.667 f 0.444-
CO
5 0.222-
0-
Tool path strategy
1.800E 1.671-Ê 1.543-< 1.414-g 1.2861 1.157 5 1.029-"S, 0.9001
g 0.643-=5 0.514-I 0.386 fe 0.25711 0.1290
1 2 3
—r 2 3
0.7000.651-
r-. 0.601-
i 0.552-Ü 0.503-^ 0.454-g 0.4041 0.355-■o 0.306$ 0.256-
I 0.207£ 0.1580.1090.0590,010-
□ D
1 2 3
3,00 2.80 Ç 2.60
2.40 ^ 2.20 OJ 2.00-
œ 1.S0
£ 1.60 ra
3 1.40 -O
1.20
0)
ü 1.00 (Ü
t 0.30 -W 0 60 0.400 20 -
Fig. 5 Intervals of the values T, Aa, Ab, Ac and Ra
3. Results and discussion 3.1 Empirical modelling
After manual elimination of inadequate tool path strategy No. 2, responses of both other strategies, No. 1 and No. 3 were empirically modelled. Results from this stage are empirical models of responses for both tool path strategies, which are listed in Table 1. All models were obtained using quadratic polynomial modelling with some transformations for better model results.
Table 1 Empirical models of responses
Tool path	Empirical model				Responses
No. 1	InT = 4.89861 + 2.14456 • 10"3 • a - 7.7971 • 10"3 • f + 7.84629 • 10	n6	•f2		Machining time
No. 1	Via = 0.66044 - 0.32615 • a - 2.34118 • 10"3 • f + 1.68538 • 10"3 ■	• a •	f		Wall thickness deviation
No. 1	InAb = 1.51485 - 5.24519 • a - 0.020474 • f + 2.29188 • a2 + 4.12498 •	10	-5 •	f2	Perpendicularity deviation
No. 1	1 -= = 3.20174 + 3.36622 • a Vic				Flatness deviation
No. 1	VRa = -0.80634 + 1.80076 • a + 5.71813 • 10"3 • f - 1.67948 • 10"3 - 0.44182 • a2 - 8.04146 • 10"6 • f2	•a	•		Surface roughness
No. 3	InT = 5.437 - 0.02208 • a - 8.04091 • 10"3 • f + 7.84629 • 10"6 •	f2			Machining time
No. 3	Via = 0.85989 - 0.44928 • a - 1.96566 • 10"3 • f + 1.68538 • 10"3 ■	• a •	f		Wall thickness deviation
No. 3	InAb = 1.70287 - 5.24519 • a - 0.020474 • f + 2.29188 • a2 + 4.12498 •	10	"5 •	f2	Perpendicularity deviation
No. 3	1 -= = 2.52262 + 3.74762 • a Vic				Flatness deviation
No. 3	VRa = -0.78597 + 1.57972 • a + 5.71813 • 10"3 • f - 1.67948 • 10"3 - 0.44182 • a2 - 8.04146 • 10"6 • f2	•a	•		Surface roughness
130
Advances in Production Engineering & Management 13(2) 2018
Optimization of process parameters for machining of Al 7075 thin-walled structures
Tool path strategies ■ C1 Tool path No-1 ♦ C3 Tool path No 3
Wall thickness a [mm]
Fig. 6 Machining time T for tool paths
ï.®
Fig. 7 3D surface diagram for the thickness deviation Aa with tool path strategy No. 1
Based on the carried experiments and empirical modelling, graphical results, i.e. 3D surface diagrams were obtained and discussed below.
As it is well-known, machining time is influenced by feed rate value. While wall thickness, on the other hand, does not show almost any significance on machining time. This refers to both tool path strategies. It was recorded that machining time for tool path strategy No. 1 was significantly shorter than for No. 3 for same feed rate values f as can be seen in Fig. 6.
The influence of feed rate f and wall thickness a on wall thickness deviation Aa for tool path strategy No. 1 is shown in 3D surface diagram in Fig. 7. It can be seen that maximum wall thickness deviation for this tool path strategy was lower than 0.05 mm.
Wall thickness deviation decreases with higher feed rates. Model also shows that wall thickness deviation is proportional to the wall thickness, and that with the higher wall thicknesses, feed rate has almost no influence on wall thickness deviation.
3D surface diagram for wall thickness deviation Aa using tool path strategy No. 3 is shown in Fig. 8. In this case maximum wall thickness deviation of 0.17 mm and min. deviation of 0.08 mm was achieved. Wall thickness deviation for tool path strategy No. 3 is also strongly influenced by feed rate with lower wall thickness, while less influence of feed rate is noticed with higher wall thickness. It can be seen that tool path strategy No. 1 gives much better results when thickness deviation is considered.
3D surface diagram for the perpendicularity deviation Ab using tool path strategy No. 1 is shown in Fig. 9. Models for perpendicularity deviation Ab for both tool path strategies show the same trends, i.e. perpendicularity deviation rises with wall thickness reduction, while feed rate does not have significant influence. Best results, i.e. the lowest deviation can be achieved with middle values of feed rate. Maximum perpendicularity deviation Ab for tool path strategies No. 1 of 0.04 mm is predicted and it is lower than for tool path strategies No. 3 (0.05 mm).
Fig. 8 3D surface diagram for the thickness deviation Aa with tool path strategy No. 3
Fig. 9 3D surface diagram for the perpendicularity deviation Ab with tool path strategy No. 1
Advances in Production Engineering & Management 13(2) 2018
131
Borojevic, Lukic, Milosevic, Vukman, Kramar
Influence of feed rate and wall thickness on flatness deviation Ac for tool path strategy No. 1 is shown by 3D surface diagram in Fig. 10. Flatness deviation for both tool path strategies (No. 1 and No. 3) shows the same dependence, i.e. flatness deviation is significantly influenced by wall thickness, while feed rate almost does not have any influence.
Flatness deviation is inversely proportional with wall thickness. Maximum flatness deviation Ac for tool path strategies No. 1 has value of 0.035 mm and it was just a bit lower than for the tool path strategies No. 3 (0.04 mm), both for the same wall thickness a = 0.65 mm.
3D surface diagram for the surface roughness Ra for tool path strategy No. 3 is shown in Fig. 11. Also in this case, surface roughness models for both tool path strategies show the same dependence, i.e. surface roughness is proportional to the wall thickness, while feed rate affects roughness inversely proportionally only with higher wall thicknesses. Maximum surface roughness achieved with tool path strategy No. 1 was Ra = 1.41 |j.m and it was higher than for tool path strategy No. 3 (Ra = 0.83 ^m).
Taking into consideration empirical models and 3D surface diagrams, it can be concluded that tool path strategy No. 1 has advantages over tool path strategy No. 3 in the following responses: machining time, thickness deviation, perpendicularity deviation and flatness deviation. Tool path strategy No. 3 has only an advantage in surface roughness. Based on these reasons, tool path strategy No. 1 can be declared as the one which generates the best responses for previously adopted conditions of experiment. For those reasons, the optimization of process parameters for thin wall structures machining was executed only for tool path strategy No. 1, i.e. parallel spiral.
Fig. 10 3D surface diagram for the flatness	Fig. 11 3D surface diagram for the surface
deviation Ac with tool path strategy No. 1	roughness Ra with tool path strategy No. 3
3.2 Evaluation of the experimental results and optimization
Empirical models were then used for machining optimization. As concluded above, the optimization of machining parameters was performed for the tool path strategy No. 1 only. The main objective of optimization was the selection of optimal feed rate values for specific wall thickness between 0.5 mm and 1.5 mm, using the step of 0.1 mm, which would generate the smallest values of responses, i.e. the smallest deviation in thin wall structure machining.
Criteria and goals of optimization process are given in Table 2. As it can be seen, the highest weight coefficient was assigned to wall thickness deviation, because the scope of the experiment was related to machining of thin wall structures, where the main geometrical feature of thin wall structure is its thickness. Other weight coefficients have the same values i.e. other responses have the same level of importance.
The optimal feed rate values for different wall thickness values and corresponding predicted values of responses for the tool path strategy No. 1 are shown in Table 3. Results show that except for the thickest wall, the feed rate should be set at maximum value f = 350 mm/min.
132
Advances in Production Engineering & Management 13(2) 2018
Optimization of process parameters for machining of Al 7075 thin-walled structures
Table 2 Criteria and goals of optimization process					
Responses	Goal		Lowest limit	Highest limit	Weights
Wall thickness	Equal to specific value		G.S	l.S	l
Feed rate	in range		1SG	BSG	l
Machining time	minimize		1G	9G	l
Thickness deviation	minimize		G.GGS	l	lG
Perpendicularity deviation	minimize		G.G1	l.8S	l
Flatness deviation	minimize		G.G1	G.7	l
Surface roughness	minimize		G.2	B,2	l
Table 3 The optimal parameter settings and predicted responses for different wall thickness (tool path strategy No. 1)					
a [mm] f [mm/min]	T [min]	Aa [mm]	Ab [mm]	Ac [mm]	Ra [um]
G.S BSG	22.9	G.G7	G.G42	G.GS	G.76
G.6 BSG	22.9	G.GS4	G.GB7	G.G6	G.77
G.7 BSG	22.9	G.GGl	G.G4B	G.GB	G.7
G.8 BSG	22.9	G.GGB	G.GB6	G.G29	G.8
G.9 BSG	22.9	G.GG6	G.GB1	G.G26	G.9
l BSG	22.9	G.Gll	G.G29	G.G2B	l.G
1.1 BSG	22.9	G.Gl7	G.G27	G.G21	l.G
l.2 BSG	22.9	G.G2S	G.G28	G.Gl9	l.l
l.B BSG	2B	G.GB4	G.G29	G.Gl7	l.l
l.4 BSG	2B	G.G44	G.GB2	G.Gl6	l.l
l.S 29B	2B	G.GS6	G.GB7	G.G1S	1.1
3.3 Verification of the optimization results
The verification of optimization results were conducted on two samples with wall thickness of a = 0.7 mm and a = 1.2 mm. Feed rate f was set according to the optimization at 350 mm/min and tool path strategy No.1 was selected. The other cutting conditions of experiments were kept the same as in the main experiment. After machining of the samples, the process of measurement of responses was conducted in the same manner and with the same equipment.
Verification results are shown in Table 4. It can be seen that results of verification tests deviate from the predicted values for less than 10 %. These results can be accepted as satisfactory, considering the complexity of the machining process of thin-walled structures, the conditions of machining, machine rigidity, the occurrence of vibration, and other influencing factors.
Table 4 Results of verification test
Sample	Tool path strategy	a [mm]	/[mm/min]		T [min]	Aa [mm]	Ab [mm]	Ac [mm]	Ra [um]
1	1	G.7	350	Predicted	22.9	G.GG1	G.G4B	G.GB	G.7
2	1	1.2	350	value	22.9	G.G2S	G.G28	G.G19	1.1
1	1	G.7	350	Measured	2B	G.GG1	G.G41	G.GBB	G.76
2	1	1.2	350	values	2B.B	G.G27	G.G27	G.G2G	1.18
1	1	G.7	350	Deviations	G.4%	G %	4.6 %	1G %	8.S %
2	1	1.2	350	of values	1.7 %	8 %	B.6 %	S.B %	7.2 %
4. Conclusion
The optimization methodology for thin wall structures machining was presented in this research. Significant savings in machining time were achieved by selecting the optimal tool path strategy and feed rate value, and at the same time meeting the quality factors (i.e. responses) in terms of accuracy of wall thickness deviation, perpendicularity deviation, flatness deviation and surface roughness. The determined tool path strategy and feed rate value are suitable for machining the line-type aluminium alloy thin-walled structures. They are also suitable for structures with similar configuration, such as the rectangular-type structures.
Based on these results, it can be concluded that the tool path strategy No. 1, i.e. parallel spiral provides the most satisfactory response values for wall thickness from 0.5 mm up to 1.5 mm. It
Advances in Production Engineering & Management 13(2) 2018
9
Borojevic, Lukic, Milosevic, Vukman, Kramar
is also important to highlight the significance of generated empirical models, which give insights into machining process within the considered research.
The verification of the optimization results shows the maximum deviation of 10 % in measured and predicted values, which represents a significant achievement in the field of thin-walled structures machining. The obtained results confirmed the defined research hypotheses. The represented methodology gives reliable data on the input/output parameters of machining process, which can be used for developing a knowledge base within the process planning system. It was proved that the optimization of machining parameters has an influence on improving the manufacturing of thin-walled components, primarily in terms of machining time, machining accuracy and quality itself.
In addition to the presented technological effects, realized optimization has a significant impact on the economic effects, which are easily recognizable here. Based on obtained results, it can be seen that the main machining time for the tool path strategy No. 1 is shorter than the machining time for tool path strategy No. 3, thereby it affects the increase of the manufacturing productivity. On the other hand, using the proposed optimization methodology, it is easy to identify the values of the input parameters that provide the required parameters of accuracy and surface quality. This reduces the possibility of scrapping and additional milling operations, which results in a reduction of production costs.
In order to obtain the generality of the optimization methodology for thin-walled structures process planning, it is necessary to expand the research to other shapes, such as triangular, hexagonal and other more complex shapes.
Researching should also be expanded on machining of other light alloys and composites, as well as thin-walled structures with wall thickness up to 3 mm, and larger height-to-thickness ratio (> 30:1). Monitoring of the cutting forces, occurrence of vibrations in machining and possible deformations of workpiece structures should be taken into consideration as well.
Acknowledgement
This paper was created as part of cooperation between Faculties of Mechanical Engineering in Banjaluka, Novi Sad and Ljubljana in the field of production engineering and machining process planning.
References
[1]	Jambor, A., Beyer, M. (1997). New cars - new materials, Materials & Design, Vol. 18, No. 4-6, 203-209, doi: 10.1016/S0261-3069(97)00049-6.
[2]	Hirsch, J., Al-Samman, T. (2013). Superior light metals by texture engineering: Optimized aluminium and magnesium alloys for automotive applications, Acta Materialia, Vol. 61, No. 3, 818-843, doi: 10.1016/j.actamat.2012. 10.044.
[3]	Smith, S., Dvorak, D. (1998). Tool path strategies for high speed milling aluminum workpieces with thin webs, Mechatronics, Vol. 8, No. 4, 291-300, doi: 10.1016/S0957-4158(97)00058-5.
[4]	Pompa, M.G.R. (2010J. Computer aided process planning for high-speed milling of thin-walled parts: Strategy-based support, Ph.D. Thesis, University of Twente, Enschede, Netherlands, doi: 10.3990/1.9789036530408.
[5]	Scippa, A., Grossi, N., Campatelli, G. (2014). FEM based cutting velocity selection for thin walled part machining, Procedia CIRP, Vol. 14, 287-292, doi: 10.1016/j.procir.2014.03.023.
[6]	Huang, X., Sun, J., Li, J. (2015). Effect of initial residual stress and machining-induced residual stress on the deformation of aluminium alloy plate, Strojniški vestnik - Journal of Mechanical Engineering, Vol. 61, No. 2, 131-137, doi: 10.5545/sv-jme.2014.1897.
[7]	Kopač, J., Kržič, P. (2008). CAM algorithm as important element by achieving of good machined surface quality, Strojniški vestnik - Journal of Mechanical Engineering, Vol. 54, No. 4, 280-287.
[8]	Dai, B., Yu, G.-B., Guan, Y.-Q., Shao, J.-P., Guan, Y.-Q., Wu, X.-M., Liu, Y.-X. (2015). Machining surface quality analysis of aluminum alloy thin-walled parts in aerospace, International Journal of Security and Its Applications, Vol. 9, No. 11, 201-208, doi: 10.14257/ijsia.2015.9.11.19.
[9]	Herranz, S., Campa, F.J., López de Lacalle, L.N., Rivero, A., Lamikiz, A., Ukar, E., Sánchez, J.A., Bravo, U. (2005). The milling of airframe components with low rigidity: A general approach to avoid static and dynamic problems, Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, Vol. 219, No. 11, 789-801, doi: 10.1243/095440505X32742.
134
Advances in Production Engineering & Management 13(2) 2018
Optimization of process parameters for machining of Al 7075 thin-walled structures
[10]	Zhou, X., Zhang, D., Luo, M., Wu, B. (2014). Toolpath dependent chatter suppression in multi-axis milling of hollow fan blades with ball-end cutter, The International Journal of Advanced Manufacturing Technology, Vol. 72, No. 5-8, 643-651, doi: 10.1007/s00170-014-5698-6.
[11]	Rai, J.K., Xirouchakis, P. (2008). Finite element method based machining simulation environment for analyzing part errors induced during milling of thin-walled components, International Journal of Machine Tools and Manufacture, Vol. 48, No. 6, 629-643, doi: 10.1016/j.ijmachtools.2007.11.004.
[12]	Aoyama, T., Kakinuma, Y. (2005). Development of fixure devices for thin and compliant workpieces, C1RP Annals, Vol. 54, No. 1, 325-328, doi: 10.1016/S0007-8506(07)60114-0.
[13]	Izamshah, R., Mo, J.P.T., Ding, S. (2011). Hybrid deflection prediction on machining thin-wall monolithic aerospace components, Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, Vol. 226, No. 4, 592-605, doi: 10.1177/0954405411425443.
[14]	Mwinuka, T.E., Mgwatu, M.I. (2015). Tool selection for rough and finish CNC milling operations based on tool-path generation and machining optimization, Advances in Production Engineering & Management, Vol. 10, No. 1, 18-26, doi: 10.14743/apem2015.1.189.
[15]	Lazarevic, D., Nedic, B., Marušic, V., Mišic, M., Šarkocevic, Ž. (2017). Regenerating the NC code in order to improve the surface quality, Tehnički vjesnik - Technical Gazette, Vol. 24, No. 2, 355-362, doi: 10.17559/TV-20161018122324.
[16]	Kuczmaszewski, J., Piesko, P., Zawada-Michaiowska, M. (2017). Influence of milling strategies of thin-walled elements on effectiveness of their manufacturing, Procedia Engineering, Vol. 182, 381-386, doi: 10.1016/ j.proeng.2017.03.117.
[17]	Novak-Marcincin, J., Janak, M., Novakova-Marcincinova, L., Fecova, V. (2012). Possibility of quick check on suitability of milling strategy, Tehnički vjesnik - Technical Gazette, Vol. 19, No. 4, 959-964.
[18]	Pandian, P., Prabhu, R., Sakthimurugan, K. (2013). Surface error compensation in HSM of thin wall structures, International Journal of Engineering Science Invention, Vol. 2, No. 2, 1-11.
[19]	Msaddek, E.B., Bouaziz, Z., Dessein, G., Baili, M. (2012). Optimization of pocket machining strategy in HSM, The International Journal of Advanced Manufacturing Technology, Vol. 62, No. 1-4, 69-81, doi: 10.1007/s00170-011-3801-9.
[20]	Denkena, B., Schmidt, C. (2007). Experimental investigation and simulation of machining thin-walled workpieces, Production Engineering, Research and Development, Vol. 1, No. 4, 343-350, doi: 10.1007/s11740-007-0017-9.
[21]	Budak, E., Altintas, Y. (1995). Modeling and avoidance of static form errors in peripheral milling of plates, International Journal of Machine Tools and Manufacture, Vol. 35, No. 3, 459-476, doi: 10.1016/0890-6955(94)P2628-S.
[22]	Baranek, I., Buransky, I., Peterka, J. (2013). Influence of material removal way on thin-walled part quality by milling, MM Science Journal, 414-417, doi: 10.17973/MMSI.2013 06 201306.
[23]	Rusinek, R., Zaleski, K. (2016). Dynamics of thin-walled element milling expressed by recurrence analysis, Mec-canica, Vol. 51, No. 6, 1275-1286, doi: 10.1007/s11012-015-0293-y.
[24]	Kanchana, J., Prabhu Raja, V., Prakash, R., Radhakrishnan, P. (2002). Dynamics of high-speed machining of aerospace structures using finite-element analysis, Defence Science Journal, Vol. 52, No. 4, 403-408.
[25]	Arnaud, L., Gonzalo, O., Seguy, S., Jauregi, H., Peigné, G. (2011). Simulation of low rigidity part machining applied to thin-walled structures, The International Journal of Advanced Manufacturing Technology, Vol. 54, No. 5-8, 479488, doi: 10.1007/s00170-010-2976-9.
[26]	Masmali, M., Mathew, P. (2017). An analytical approach for machining thin-walled workpieces, Procedia CIRP, Vol. 58, 187-192, doi: 10.1016/j.procir.2017.03.186.
[27]	Pare, V., Agnihotri, G., Krishna, C. (2015). Selection of optimum process parameters in high speed CNC end-milling of composite materials using meta heuristic techniques - A comparative study, Strojniški vestnik - Journal of Mechanical Engineering, Vol. 61, No. 3, 176-186, doi: 10.5545/sv-ime.2014.1914.
[28]	Ratchev, S., Nikov, S., Moualek, I. (2004). Material removal simulation of peripheral milling of thin wall low-rigidity structures using FEA, Advances in Engineering Software, Vol. 35, No. 8-9, 481-491, doi: 10.1016/ j.advengsoft.2004.06.011.
[29]	Bose, G.K., Mahapatra, K.K. (2014). Parametric study of die sinking EDM process on AISI H13 tool steel using statistical techniques, Advances in Production Engineering & Management, Vol. 9, No. 4, 168-180, doi: 10.14743/ apem2014.4.185.
[30]	Raja, I.R.A. (2011). Hybrid deflection prediction for machining thin-wall titanium alloy aerospace component, Ph.D. Thesis, School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Melbourne, Australia.
Advances in Production Engineering & Management 13(2) 2018
135