© 2023 IIETA. This article is published by IIETA and is licensed under the CC BY 4.0 license (http://creativecommons.org/licenses/by/4.0/).
OPEN ACCESS
The deflection and deformation behaviour of a 30% weight fraction glasspolyester sandwich panel was studied as a function of graphite filler quantity and impact velocity. Experiments based on Taguchi methods were carried out in order to collect data systematically. The panel is fastened on three sides and left free for the destructive test. By applying the impact force. The vibration data collector is used to measure the deflection (TVC 200). During the destructive test, the panel is additionally fastened to a solid base for stability. The steel hammer came crashing down from above. A Vernier calliper is used to measure the distortion. The tests were planned using Taguchi's L9 orthogonal array method. Examine the effects of impact force, height, and graphite filler content on lowvelocity deflections and deformations using analysis of variance. The findings demonstrate that the mass is the primary parameter influencing the deflection, whereas the graphite filler is the primary parameter influencing the deformation. As a last step, a confirmation tests have been run to make sure the predicted experimental outcomes from those correlations were correct.
glasspolyester, sandwich panel, graphite fillers, Taguchi's approach
Composites are manmade substances made up of two or more different materials, which when combined, exhibit superior qualities than those of the individual components. The two main parts of a composite are the fibers and the matrix. The fiber reinforcement provides the bulk of the material's strength, while the matrix acts to keep the fibers in place and distribute stress across them. It is common practice to use fillers and modifiers to lower production costs, improve a product's performance, or both [1, 2]. Sandwich composites are excellent for lowweight constructions that need strong inplane and flexural stiffness [3]. Everyday occurrences like objects falling on a composite casing, cars colliding at low speeds, boulders striking a vehicle's composite bodywork, and ballistic strikes on military aircraft all result in collisions at low velocities. Biswas et al. [4] studied the effects of lowvelocity impact on composite sandwich panels made with a honeycomb core sandwiched between glass fibrereinforced face sheets. The dynamic response of the panel was studied by Chauhan et al. [5] by monitoring strain at a specific place on the panel when the panel was exposed to an impact force (both experimentally and numerically). Researchers [6] looked at how a compressive preload affected the lowvelocity impact behaviour of a carbon fibrereinforced composite plate. Modeling methodologies, with a focus on laminate delimitation and preload modeling, were developed for lowvelocity impact simulation of plates under compressive preload using LS  DYNA. Researchers [79] use the Taguchi method to show how different amounts of aluminum and graphite filler affect the dynamic behavior of a glasspolyester sandwich panel subjected to lowvelocity impact and the mechanical and tribological behaviour of a glasspolyester composite system, respectively. Highquality systems may be effectively designed using the Taguchi approach [10, 11]. Taguchi's method of experimentation offers a systematic framework for gathering, analyzing, and interpreting data in service of a given research question. Maximum insight may be gained with little effort via careful experimental design. The choice of design parameters in a Taguchi parameter design may maximize performance characteristics and lessen the system's susceptibility to the source of variation [12]. The effective use of experimental runs to the investigated variable combinations carries this out. An important part of any experimental design is identifying the variables that will be used to measure success. To determine the impact of several variables on the desired outcome, the Taguchi approach generates a standardized orthogonal array. Analysis of means and variance of the influencing variables is used to examine the experimental outcomes [13]. The primary goal of this research was to examine how the amount of graphite filler used in a composite sandwich panel affected its ability to flex and bend after being struck by a lowvelocity item. To further the investigation [8], a set of experiments based on the Taguchi method was devised to systematically collect the necessary information. The Design of Experiments (DOE) method may be easily implemented to find the optimal processing path in the least amount of time. The goal is to narrow down the number of tests and costs to just those that are necessary, out of the infinite number of potential parameter combinations [14].
The literature review shows that the research [46] investigates the behaviour of a composite plate under dynamic load, while researchers [79] demonstrate the effect of filler on the behaviour of composite material using the Taguchi technique. In this research, the authors study the influence of filler on the dynamic response of composite plates using the Taguchi technique.
In this study, (360×10^{3}) kg woven roving glass fiber with Eglass in the ((814) ×10^{6}) m size range was employed. The matrix system used was an unsaturated polyester resin, namely the commercially available TOPAZ 1110 TP medium reactive based on Phthalic Anhydride and a room temperature hardener (Methyl Ethyl Kenton Peroxide (MEKP)). Graphite granules served as the filler (no. 7782425 Merck index 10, 4410 Swiss). Injecting fillers into a polymer matrix may save production costs, increase modulus, reduce mold shrinkage, regulate viscosity, and provide a smooth surface.
All the composite honeycomb sandwich panel samples were made using the dry hand layup process described in the study [8], with a weight fraction of 30% and a graphite filler percentage of 5%, 7.5%, and 10%, respectively. Figure 1 depicts the process flow for making a composite and detailing its properties. The 54 composite honeycomb sandwich panels present with (80 mm x 40 mm x 40 mm) dimensions were manufactured.
Figure 1. Process flow chart
Finally, a band knife cutting machine was used to form ASTMcompliant cuts in the slag and coconut shellreinforced composite plates.
Tensile testing reveals the material's mechanical characteristics. Table 1 shows the results of tensile tests performed on materials following ASTMD 638 [15].
In order to find the sweet spot for deflection and deformation, three factors were investigated. Each with three levels (graphite content, mass, and height). The values for these parameters are shown in Table 2. Taguchi's method of experimentation offers a systematic framework for gathering, analysing, and interpreting data in service of a given research question. Maximum insight may be gained with little effort via careful experimental design. To model and analyse the impact of control elements on performance outcomes, Taguchi design of experiment is a strong analytic tool. The process of deciding which variables to use as controls is the most crucial part of an experiment's design. By selecting the right Taguchi orthogonal array with control parameters and levels [16, 17], the optimal level combination may be produced. Three of these characteristics have been investigated here, and their effects are summarized in Table 3 using an L9 (33) orthogonal design.
Figure 2. The honeycomb sandwich panel
Table 1. Mechanical characteristics and the amount of graphite filling
Graphite % 
Modulus of elasticity (Gpa) 
Yield stress (Mpa) 
Tesile strength (Mpa) 
5% 
2.950 
80 
151 
7.5% 
3.610 
89 
169 
10% 
3.090 
77 
158 
Table 2. Factors and their levels
a Nondestructive test 

Factor and Unit 
Symbol 
Factor levels 

Level 1 
Level 2 
Level 3 

Graphite (%) 
AA 
5 % 
7.5 % 
10 % 
Mass (g) 
BB 
45 
65 
85 
Height (cm) 
CC 
90 
105 
120 
b Destructive test 

Factor and Unit 
Symbol 
Factor levels 

Level 1 
Level 2 
Level 3 

Graphite (%) 
AA 
5 % 
7.5 % 
10 % 
Mass (kg) 
BB 
12 
15 
18 
Height (m) 
CC 
1.5 
2 
2.5 
Table 3. Standard Taguchi’s orthogonal array L9
No. 
A 
B 
C 
1 
1 
1 
1 
2 
1 
2 
2 
3 
1 
3 
3 
4 
2 
1 
2 
5 
2 
2 
3 
6 
2 
3 
1 
7 
3 
1 
3 
8 
3 
2 
1 
9 
3 
3 
2 
Elastic modulus, yield stress, and ultimate tensile strength are only some of the mechanical qualities that have been measured. As can be seen in Table 1, the composite's mechanical characteristics tend to improve as the filler percentage rises. Given that the filler collaborates with the fiber and resin to provide increased strength and durability. The filler and the fiber, as well as the filler and the resin, are all compatible with one another.
4.1 S/N ratio
The signaltonoise (S/N) ratio is another fundamental component of the Taguchi technique. The S/N ratio is derived from experimental findings. Various signaltonoise (S/N) ratios may be used to achieve the required quality level. The S/N ratio "smaller is better" has been employed in calculations since both nondestructive testing (deflection) and destructive testing (deformation) are objectives of the current study. To calculate it, use the logarithmic formula [17].
$S / N=10 \log _{10}\left[\frac{1}{n} \sum_{i=1}^n y_i^2\right]$ (1)
The number of observations is denoted by n, the signaltonoise ratio by S/N, and the quantity of interest, y_{i}, by i. The experimental results for deflection and deformation are shown with the predicted S/N ratios for all 9 observations in Table 4.
4.2 Analysis of means
Each control parameter's influence is determined by averaging the S/N ratios at the same setting [16]. The optimal combination of the process parameters may be selected based on the findings of the mean analysis. A signaltonoise ratio was used to construct the analysis. Table 4 summarizes the average S/N ratio values that were obtained for each parameter across all levels in response to the average impacts of deflection and deformation. For the graphite parameter impact, level1 is the mean of the first three S/N ratio calculations shown in Table 5a (level1 equals an average of 14.582, 16.40495, and 16.5202). Based on the experimental layout (orthogonal array) given in Table 3, the first three rows of the L9 array correspond to the graphite parameter's level1. Both of the other parameters and the other two levels have undergone the same process. The optimal level is the one with the maximum signaltonoise ratio. As a result, the reaction of deflection is determined by the level combination of the three factors A2B1C1, which is comparable to (7.5% graphite), (45 g) (mass), and (90 cm) (height). Furthermore, the optimal combination level for the deformation is the same as the deflection, A2B1C1, where A2 is 7.5% graphite, B1 is 12 Kg, and C1 is 1.5 m. Table 5 shows that the parameters' relative importance varies depending on the kind of analysis being performed, with the mass parameter being more important for calculating bending moments than any of the others, while the graphite parameter is more important for calculating bending moments. This is determined by the parameter rank, which is the maximum minus minimum value of the three levels for each parameter.
Table 4. Nondestructive test and destructive test
a. Nondestructive test 

Exp. No. 
Combinations 
Deflection (δ µm) 
S/N ratio (dB) 

A 
B 
C 

Graphite (%) 
Mass (g) 
Height (cm) 

1 
0.05 
45 
90 
5.3592 
14.582 
2 
0.05 
65 
105 
6.6107 
16.40495 
3 
0.05 
85 
120 
6.699 
16.5202 
4 
0.075 
45 
105 
4.7673 
13.56545 
5 
0.075 
65 
120 
5.9432 
15.48041 
6 
0.075 
85 
90 
6.12974 
15.74884 
7 
0.1 
45 
120 
6.50386 
16.26342 
8 
0.1 
65 
90 
6.422766 
16.15444 
9 
0.1 
85 
105 
6.703766 
16.52638 
b. Destructive test 

Exp. No. 
Combinations 
Deflection (Δ mm) 
S/N ratio (dB) 

A 
B 
C 

Graphite (%) 
Mass (Kg) 
Height (m) 

1 
0.05 
12 
1.5 
9.7 
19.73543 
2 
0.05 
15 
2 
12 
21.58362 
3 
0.05 
18 
2.5 
14.1 
22.98438 
4 
0.075 
12 
2 
6.3 
15.98681 
5 
0.075 
15 
2.5 
7.5 
17.50123 
6 
0.075 
18 
1.5 
6.8 
16.65018 
7 
0.1 
12 
2.5 
11 
20.82785 
8 
0.1 
15 
1.5 
9.4 
19.46256 
9 
0.1 
18 
2 
11.3 
21.06157 
Table 5. Mean S/N ratio for each factor level of
a. Nondestructive test 

Factor 
Symbol 
Average of levels for S/N ratio 
MaxMin 
Rank 

Level 1 
Level 2 
Level 3 

Graphite (%) 
A 
15.83572 
14.93157 
16.31475 
1.38318 
2 
Mass (g) 
B 
14.80362 
16.01327 
16.26514 
1.46151 
1 
Height (cm) 
C 
15.49509 
15.49893 
16.08801 
0.58908 
3 
b. Destructive test 

Factor 
Symbol 
Average of levels for S/N ratio 
MaxMin 
Rank 

Level 1 
Level 2 
Level 3 

Graphite (%) 
A 
21.43448 
16.71274 
20.45066 
4.72174 
1 
Mass (Kg) 
B 
18.85003 
19.51580 
20.23204 
1.38201 
3 
Height (m) 
C 
18.61606 
19.54400 
20.43782 
1.82176 
2 
The underlined value represents the optimum level (A2B1C1)
Using the information supplied in Table 5, the primary impact of the S/N ratio is shown visually in Figure 3. For both destructive and nondestructive testing, the figures depict the correlation between the S/N ratio and parameter values. Graph 2 shows that the S/N values consistently decrease as the mass increases. As a result, the smaller mass will result in the smallest bending moment. The graph also reveals that the greatest mean S/N value is found at graphite level 2. In the case of mass, the largest S/N mean is found at level 1. Levels 1 and 2 seem to have the same impact and are closer to a straightline connection when the height parameter is included. This indicates that the deflection is not significantly affected by whether using level 1 or level 2. However, Figure 3 demonstrates that the mass and height parameters are related to deformation, suggesting that these two factors share the same impact. To rephrase, more deformation occurs as height or mass increases. In addition, the best level is the one that produces the greatest S/N ratio on the graph. Therefore, the parameters and their values of A2B1C1 are the best possible combination.
Figure 3. Representation of S/N ratio versus levels of each parameter for the destructive test
4.3 Analysis of variance (ANOVA)
Table 6. ANOVA
a. Deflection δ (nondestructive test) 

Source 
Sum of Squares SS 
Degree of Freedom d.f. 
Mean Squares MS 
F value (MS/error) 
Contribution (%) 
Graphite (%) 
2.96014 
2 
1.48007 
3.25 
35.9543 
Mass (g) 
3.66271 
2 
1.83135 
4.02 
44.4878 
Height (cm) 
0.69858 
2 
0.34929 
0.77 
8.48506 
Error 
0.91162 
2 
0.45581 

11.0726 
Total 
8.23305 
8 


100 
b. Deformation Δ (destructive test) 

Source 
Sum of Squares SS 
Degree of Freedom def. 
Mean Squares MS 
F value (MS/error) 
Contribution (%) 
Graphite (%) 
37.2347 
2 
18.6174 
336.65 
82.3951 
Mass (Kg) 
2.8662 
2 
1.4331 
25.91 
6.3425 
Height (m) 
4.9788 
2 
2.4894 
45.01 
11.0178 
Error 
0.1106 
2 
0.053 

0.2447 
Total 
45.1904 
8 


100 % 
The purpose of the ANOVA statistical approach is to investigate the effect of various design elements on the overall variation of the outcomes [18]. Deflection and distortion ANOVA findings are shown in Table 6. Mass, as can be shown in Table 6a, is the most important factor, accounting for 44.48% of the total variation in the process. Therefore, it is the most important factor in determining deflection (by 35.95%). Height has less of an impact than width and depth. It is also evident from the relative contributions of the three factors that the mass parameter is the most important one, followed by the graphite% and the height parameter. Table 6–b reveals that the graphite parameter contributes 82.39% to the deformation, making it the most important one. The height parameter comes in second, followed by the mass. The findings of the analysis of variance and the results of the analysis of mean are identical.
4.4 Confirmation test
Confirmation tests are performed as the last stage of experimental design. Predicting and verifying inference derived during the analysis state and improvements in observed values by using the optimum level of process parameters [17, 19, 20] are the goals of this exercise. The optimal values of deflection and deformation are used in the confirmation experiment. Consequently, the formula for the deflection is graphite (7.5%, 45 g, 90 cm) while the formula for the deformation is graphite (7.5%, 12 Kg, 1.5 m). This equation [17] may be used to determine the ideal S/N ratio for a given set of process parameters.
$\eta_{o p t}=\eta_m+\sum_{i=1}^k\left(\eta_i\eta_m\right)$ (2)
where:
η_{opt}: predicted S⁄N ratio.
η_{m}: Total mean of S⁄N ratios.
η_{i}: mean S⁄N ratio at optimum levels.
k: number of main design parameters that affect the quality characteristics.
In order to compare the anticipated S/N ratio with the observed value, an experiment was performed using the combination A2B1C1, for both, deflection and deformation. Based on the variables listed in the last column of Table 4, a deflection value of (_m=15.69401 dB) for the overall mean S/N ratio were obtained. the value for distortion of (_m=19.53263 dB). The cumulative mean of the S/N ratio have been utilized in conjunction with the highlighted variables in Table 5 to get the optimal anticipated value of S/N. The confirmation test findings are summarized in Table 7.
Table 7. Confirmation results
a. Nondestructive test 

Optimum Process Parameters Graphite (7.5%), Mass (45g), Height (90cm) 

Experiment 
Prediction 

Levels 
A_{2}B_{1}C_{1} 
A_{2}B_{1}C_{1} 
Deflection (µm) 
4.565 
4.921 
S/N (dB) 
13.188 
13.84226 
b. Destructive test 

Optimum Process Parameters Graphite (7.5%), Mass (12Kg), Height (1.5m) 

Experiment 
Prediction 

Levels 
A_{2}B_{1}C_{1} 
A_{2}B_{1}C_{1} 
Deformation (mm) 
5.323 
5.6963 
S/N (dB) 
14.523 
15.11357 
Following are some inferences that may be made:
This study has been carried out in the laboratory of the University of TechnologyIraq.
[1] Alpaydin, N.K., Türkmen, H.S. (2009). The dynamic response of the sandwich panel subjected to the impact load. RAST 2009  Proceedings of 4th International Conference on Recent Advances Space Technologies, Istanbul, Turkey, pp. 176180. https://doi.org/10.1109/RAST.2009.5158191
[2] Alwan, H.L. (2014). Application of Taguchi approach to study the effect of filler type on tribological behavior of polymer composite under dry conditions. Engineering and Technology Journals, 32(8): 19161930.
[3] Basavarajappa, S., Chandramohan, G. (2005). Wear studies on metal matrix composites: A taguchi approach. Journal of Materials Science and Technology, 21(6): 845850.
[4] Biswas, S., Patnaik, A., Kumar, P. (2011). Silicon carbide filled polymer composite for erosive environment application: A comparative analysis of experimental and FE simulation results. Silicon Carbide  Materials, Processing and Applications in Electronic Devices. https://doi.org/10.5772/21547
[5] Chauhan, S.R., Kumar, A., Singh, I., Kumar, P. (2010). Effect of fly ash content on friction and dry sliding wear behavior of glass fiber reinforced polymer composites  A Taguchi approach. Journal of Minerals and Materials Characterization and Engineering, 9(4): 365387. https://doi.org/10.4236/jmmce.2010.94027
[6] Long, A.C. (Ed.). (2014). Composites forming technologies. Elsevier.
[7] Daniel, M.I., Abot, L.J. (2000). Fabrication, testing and analysis of composite sandwich beams. Composites Science and Technology, 60(1213): 24552463. https://doi.org/10.1016/S02663538(00)000397
[8] Erickson, M.D., Kallmeyer, A.R., Kellogg, K.G. (2005). Effect of temperature on the lowvelocity impact behavior of composite sandwich panels. Journal of Sandwich Structures & Materials, 7(3): 245264. https://doi.org/10.1177/1099636205048636
[9] Eweed, K.M., Sabri, S.Z. (2015). Effect of graphite content on the dynamic behavior of sandwich panel under low velocity impact. Journal of Engineering and Sustainable Development, 19(1): 113. https://www.iasj.net/iasj/download/f211afe134a3e95e
[10] Heimbs, S., Heller, S., Middendorf, P. (2008). Simulation of low velocity impact on composite plates with compressive preload. LSDYNA Anwenderforum, Bamberg, pp. 1124.
[11] Kiran, G., Suman, K., Rao, N., Rao, R. (2011). A study on the influence of hot press forming process parameters on mechanical properties of green composites using Taguchi experimental design. International Journal of Engineering, Science and Technology, 3(4): 253263. https://doi.org/10.4314/ijest.v3i4.68557
[12] Mahapatra, S.S., Patnaik, A. (2009). Study on mechanical and erosion wear behavior of hybrid composites using Taguchi experimental design. Materials and Design, 30(8): 27912801. https://doi.org/10.1016/j.matdes.2009.01.037
[13] Mahel mohammed, F., Mershed, K.A., Zuhair Sabri, S. (2013). Aluminum filler content effect on the dynamic behavior of sandwich panel subjected to impact load. Engineering and Technology Journals, 31(14): 27162726.
[14] Hodgkinson, J.M. (1990). Advanced Composites Bulletin. Mechanical Testing of Advanced Composites, 11: 13.
[15] Mohammadi, I., HaghighiYazdi, M., Safarabadi, M., Yousefi, A. (2023). Crashworthiness analysis of a composite guardrail under impact loading. Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications, 146442072311524. https://doi.org/10.1177/14644207231152432
[16] Roy Ranjit, K. (1990). A primer on Taguchi method. VNR Publication, New York, 710.
[17] Shaker, D.H., Mohammed, F.M., Smait, D.A. (2021). Acoustic isolation behavior of composite materials reinforced by natural filler using robust design. Solid State Technology, 64(2): 16581666.
[18] Taguchi, G. (2011). Taguchi on robust technology development. Taguchi on Robust Technology Development, 115(3): 336337 https://doi.org/10.1115/1.800288
[19] Vasiliev, V.V., Morozov, E.V. (2018). Mechanics of a composite layer. In Advanced Mechanics of Composite Materials and Structures, 75189. https://doi.org/10.1016/B9780081022092.000025
[20] Yeh, M.K., Chiu, Y.W. (2013). Bending strength analysis of centrallydebonded composite sandwich beam using taguchi method. Proceedings of the 13th East AsiaPacific Conference on Structural Engineering and Construction, EASEC 2013, Sapporo, Japan, pp. 15.