اسامة وماجد Al-Khwarizmi Engineering Journal Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53 -61(2012) Evaluation Study of Glass Fiber Reinforced Polyester and Kevlar Reinforced Polyester by Taguchi Method Osama Sultan M.* Majid Hmeed A. Ishraq Abdul Razaq K. Department of Materials Engineering/University of Technology * Email: oalaamiri@yahoo.com (Received 27 December 2011; accepted 6 September 2012) Abstract In the present investigation two different types of fiber reinforced polymer composites were prepared by hand lay-up method using three different parameters (curing temperature, pressing load and fiber volume fraction). These composites were prepared from the polyester resin as the matrix material reinforced with glass fibers as first group of samples and mat Kevlar fibers as the second group, both with different volume fractions (4%, 8%, and 12%) of fibers. They were then tested by tensile strength and impact strength. The main objective in this study is to use Taguchi method for predicting the better parameters that give the better tensile and impact strength to the composites, and then preparing composites at these parameters and comparing them with the randomly used once. The experimental and analytical results showed that the Taguchi method was successful in optimizing the parameters that give the highest properties and it can find the most influential parameter regardless of the material used. Also it showed that the volume fraction was the most influential parameter on the tensile and impact strength. The difference between these composites was in the properties values and that the Kevlar composites have higher tensile and impact strength. Keywords: Fiber reinforced composites, Taguchi method, Tensile strength, Impact strength. 1. Introduction Composite material is well known as an excellent structural material. It consists of two or more materials (i.e. fiber and matrix) combined to give superior performance compared to the properties of the individual components. Fiber reinforced plastic composites using resin such as epoxy, polyester and vinylester are extensively used as structural materials for many applications such as automotive industry, aerospace and civil engineering structures due to their high specific stiffness, strength, and cost and weight advantages. Manufacturing of structural materials made from fiber composites calls for an improved strength to weight ratio [1,2, 3]. Glass-fiber-reinforced polymers (GRPs) have received considerable attention as alternatives to steel and aluminum as structural materials; on the other hand, Kevlar fibers as reinforcement became more popular in specific applications such as parachute webbing, rocket motor casings, jet- engine containment, aircraft seats, and automobile tires due to their improved stiffness and density properties in relative to glass and their considerably lower cost than that of boron or even graphite fibers [4, 5]. The quality of any composite material is influenced by varying processing parameters. Among these parameters, there must be one or two that have the most influence. It has been realized that the full economic and technical potential of any manufacturing process can be achieved only while the process is run with the optimum parameters. One of the most important optimization processes is Taguchi method [6]. This technique helps to study effect of many factors (variables) on the desired quality characteristic most economically. By studying the effect of individual factors on the results, the best factor combination can be determined. Taguchi designs experiments using specially constructed tables known as “orthogonal array” (OA). The use of these tables makes the design of experiments mailto:oalaamiri@yahoo.com Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 54 very easy and consistent and it requires relatively lesser number of experimental trials to study the entire parameter space. As a result, time, cost, and labour saving can be achieved. The experimental results are then transformed into a signal-to-noise (S/N) ratio. Taguchi technique recommends the use of the S/N ratio to measure the quality characteristics deviating from the desired values. Usually, there are three categories of quality characteristic in the analysis of the S/N ratio, i.e. the-lower-the-better, the-higher-the-better, and the nominal-the-better. The S/N ratio for each level of process parameters is computed based on the S/N analysis. Regardless of the category of the quality characteristic, a greater S/N ratio corresponds to better quality characteristics. Therefore, the optimal level of the process parameters is the level with the greatest S/N ratio. Furthermore, a statistical analysis of variance (ANOVA) can be performed to analyze which process parameters are statistically significant. With the S/N and ANOVA analyses, the optimal combination of the process parameters can be predicted. Finally, a confirmation experiment may be needed to verify the optimal process parameters obtained from the parameter design [7,8]. Taguchi method provides a simple efficient and systematic approach to optimize design for performance, quality and cost. The methodology is valuable when design parameters are qualitative and discrete. Taguchi parameter design can optimize the performance characteristic through the setting of design parameters and reduce the sensitivity of the system performance to source variation [9,10]. The Taguchi approach reduces the experimental trials to a minimum number and it is a multi – step process which follow a certain sequence for the experiments to yield an improved understanding of product or process performance [11]. 2. Materials and Methods In this study, general purpose unsaturated polyester resin is used as the matrix and two types of reinforcing fibers (i.e. glass fibers and Kevlar fibers). Table (1) shows the mechanical properties of these materials. Two types of composites were prepared by the hand lay-up method (i.e. glass fiber reinforce polyester GFRP and Kevlar reinforced polyester KFRP). At first a glass mold of the dimensions (30*30*5 cm) was used with clean and regular inner surfaces, and then the resin was mixed with 2% methyl-ethyl-ketone peroxide (MEKP) as a hardener. After a short time, the mixture became as a gel which is then poured into the mold. The fibers in a mat shape were put between two layers of the matrix. The layered structure was put under different loads for about 24 hours for proper curing at room temperature and allowed to harden on cure. It was cured at room temperature for 24 hours and followed by oven cure at different temperatures so that the matrix completely seeps in and become dry. Three volume fractions (4%, 8% and 12%) of reinforcements were used for each type of composites. The resulted composites were then cut into appropriate specimens for the tensile and impact tests. After some preliminary tests, the experimental conditions shown in Table (2) were chosen to study the effects of processing parameters (Curing temperature, pressing load and volume fraction) on the tensile and impact strength of the composites. Table 1, General Properties of the Used Materials. Property U-Polyester E-Glass fibers Kevlar fibers (49) Tensile strength (MPa) 40-90 1500 2800 Modulus of elasticity (GPa) 2- 4.5 76 125 Density (gm/cm3) 1.2-1.5 2.59 1.44 Elongation (%) 2 3.8 2.1 Linear coefficient of thermal expansion (*10-6 K-1) 75 5 -2 Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 55 Table 2, Control Factors and Their Levels. Factor Control factor Level 1 Level 2 Level 3 A Curing Temperature (ºC) 20 40 60 B Pressing load (gm) 100 150 200 C Volume Fraction (%) 4 8 12 3. Application of Taguchi Method The Taguchi method of design of experiments is a statistical tool based on the systematic approach of conducting minimal number of experiments using a mathematical instrument called orthogonal arrays. Traditionally, the method has been used to predict the significance of contribution of each design variables and their level to achieve optimum combination by conducting a real time experiment. In our work, an orthogonal array of the type (L9 33) was chosen since we have three factors (variables) and three levels. Table (3) represents the standard orthogonal array of type (L9 33). During the composites preparation method, three process parameters for each type of composites were considered. These are: (1) curing temperature; (2) pressing load, and (3) fibers volume fraction. Each is at three levels as listed in Table (2). The degree of freedom for three parameters in each of the three levels is calculated as follows [6]: Degree Of Freedom (DOF) = number of levels -1 …(1) For each factor, DOF equals to: For (A); DOF = 3 – 1 = 2 For (B); DOF = 3 – 1 = 2 For (C); DOF = 3 – 1 = 2 Table 3, The Standard (L9 33) Orthogonal Array [9]. Experiment number P1 P2 P3 1 1 1 1 2 1 2 2 3 1 3 3 4 2 1 3 5 2 2 1 6 2 3 2 7 3 1 2 8 3 2 3 9 3 3 1 * (P) represents factors. In this research nine experimental trails were conducted at different parameters, for each type of composites and they were cut and tested by tensile and impact tests. Figure (1) shows some of the specimens. The tensile tests were performed by using the Time testing machine according to (ASTM D638- 87) standard in which the gage length is (60mm). The impact tests were performed by Time testing machine (Izod) type XJU-22 according to (D256- 87) standard with the dimensions (55 * 10 * 4mm). Tables (4) and (5) indicate the used parameters and the result values of tensile and impact strength for (GFRP) and (KFRP) respectively. A standard three level (L9 33) orthogonal array with nine experimental runs was selected. The total degree of freedom is calculated from the following [6]: Total DOF = no. of experiments – 1 …(2) The total DOF for the experiments is: Total DOF = 9 – 1 = 8 Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 56 Fig. 1. Some of the Prepared Specimens (upper: GFRP, lower: KFRP). Table 4, Experimental Runs of the OA (L9 33) for the GFRP. Experiment number Control factor Response factor A B C Tensile strength (MPa) Impact strength (MPa) 1 20 100 4 56 49.7 2 20 150 8 62 61.9 3 20 200 12 78 76.4 4 40 100 12 74 71.4 5 40 150 4 58 53.3 6 40 200 8 65 63.2 7 60 100 8 69 67.1 8 60 150 12 91 78.4 9 60 200 4 60 56.1 GFRP: Glass Fiber Reinforced Polyester. A: Curing Temperature (ºC); B: Pressing Load (gm); C: Volume Fraction (%) and E: Error. Table 5, Experimental runs of the OA (L9 33) for the KFRP. Experiment number Control factor Response factor A B C Tensile strength (MPa) Impact strength (MPa) 1 20 100 4 59 68.6 2 20 150 8 81 87.4 3 20 200 12 107 101.4 4 40 100 12 101 98.7 5 40 150 4 64 70.5 6 40 200 8 92 89.8 7 60 100 8 93 90.2 8 60 150 12 119 105.1 9 60 200 4 77 73.9 KFRP: Kevlar Fiber Reinforced Polyester. Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 57 Taguchi method stresses the importance of studying the response variation using the signal – to – noise (S/N) ratio, resulting in minimization of quality characteristic variation due to uncontrollable parameter. Since the tensile and impact strength were required to be high, the concept of "the larger the better" is considered. The S/N ratio used for this type of response is given by [6]: …(3) Where i: experiment number. u: trail number. Ni: number of trails for experiment i. yu: the response (property) value for a trial condition repeated (u) times. The composite preparation parameters, namely: (A) curing temperature, (B) pressing load and (C) particles volume fraction were assigned to the 1st , 2nd and 3rd column of the (L9 34) array, respectively. The 4th column was assigned as error (E). The S/N ratios were computed for tensile strength and impact strength in each of the nine trial conditions for each type of composites and their values are given in Tables (6) and (7). Computation scheme of pareto ANOVA (Analysis Of Variance) for three level factors is shown in Table (8). In order to study the contribution ratio of the process parameters, pareto ANOVA was performed for tensile and impact strengths. The details are given in Tables (9) and (10) respectively for the GFRP and Tables (11) and (12) for the KFRP. Table 6, S/N Ratio of Tensile and Impact Strengths for the GFRP. Experiment number A B C E S/N Ratio (Tensile strength) S/N Ratio (Impact strength) 1 1 1 1 1 34.96 33.93 2 1 2 2 2 35.85 35.83 3 1 3 3 3 37.84 37.66 4 2 1 3 2 37.38 37.07 5 2 2 1 3 35.27 34.53 6 2 3 2 1 36.26 36.01 7 3 1 2 3 36.78 36.53 8 3 2 3 1 39.19 37.89 9 3 3 1 2 35.56 34.98 Table 7, S/N Ratio of Tensile And Impact Strengths for the KFRP. Experiment number A B C E S/N Ratio (Tensile strength) S/N Ratio (Impact strength) 1 1 1 1 1 35.42 36.73 2 1 2 2 2 38.17 38.83 3 1 3 3 3 40.59 40.12 4 2 1 3 2 40.09 39.89 5 2 2 1 3 36.12 36.96 6 2 3 2 1 39.28 39.07 7 3 1 2 3 39.37 39.1 8 3 2 3 1 41.51 40.43 9 3 3 1 2 37.73 37.37 Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 58 Table 8, Pareto ANOVA for Three Level Factors [6]. Factors A B C E Total Sum. at factor level A1 B1 C1 E1 T A2 B2 C2 E2 A3 B3 C3 E3 Sum. of squares of differences SA SB SC SE ST Degree of freedom (Contribution ratio)/100 SA/ ST SB/ ST SC/ ST SE/ ST 1 T = A1 + A2 + A3 SA = ( A1 - A2)2 + ( A1 - A3)2 + ( A2- A3)2 SB = ( B1 - B2)2 + ( B1 - B3)2 + ( B2 - B3)2 SC = ( C1 - C2)2 + ( C1 - C3)2 + ( C2 - C3)2 SE = ( E1 - E2)2 + ( E1 - E3)2 + ( E2 - E3)2 ST = SA + SB + SC + SE Table 9, Pareto ANOVA of Tensile Strength for the GFRP. Factors A B C E Total Sum. at factor level 108.65 109.12 105.79 110.41 329.09 108.91 110.31 108.89 108.79 111.53 109.66 114.41 109.89 Sum. of squares of differences 15.226 2.13 114.385 4.105 135.846 Degree of freedom 2 2 2 2 8 (Contribution ratio)/100 11.208 1.568 84.202 3.022 1 Optimum level 2 A3 60 3 B2 150 1 C3 12 Table 10, Pareto ANOVA of Impact Strength for the GFRP. Factors A B C E Total Sum. at factor level 107.42 107.53 103.44 107.83 324.43 107.61 108.25 108.37 107.88 109.4 108.65 112.62 108.72 Sum. of squares of differences 7.161 1.933 126.64 1.5 137.234 Degree of freedom 2 2 2 2 8 (Contribution ratio)/100 5.218 1.409 92.28 1.093 1 Optimum level 2 A3 60 3 B3 200 1 C3 12 Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 59 Table 11, Pareto ANOVA of Tensile Strength for the KFRP. Factors A B C E Total Sum. at factor level 114.18 114.88 109.27 116.21 348.28 115.49 115.8 116.82 115.99 118.61 117.6 122.19 116.08 Sum. of squares of differences 31.075 11.485 252.766 0.0734 295.4 Degree of freedom 2 2 2 2 8 (Contribution ratio)/100 10.52 3.888 85.567 0.025 1 Optimum level 2 A3 60 3 B3 200 1 C3 12 Table 12, Pareto ANOVA of Impact Strength for the KFRP. Factors A B C E Total Sum. at factor level 115.68 115.72 111.06 116.23 348.5 115.92 116.22 117 116.09 116.9 116.56 120.44 116.18 Sum. of squares of differences 2.506 1.071 135.102 0.03 138.71 Degree of freedom 2 2 2 2 8 (Contribution ratio)/100 1.81 0.77 97.4 0.02 1 Optimum level 2 A3 60 3 B3 200 1 C3 12 4. Results and Discussion Tables (4) and (5) represents the values of tensile and impact strengths for both types of composites and it shows that Kevlar composites gave higher tensile and impact than the glass fiber composites. This is true since the Kevlar fibers have more toughness. From Table (9), it can be seen that the third level of factor (A) give the highest summation (i.e. A3, which is 60ºC curing temperature). The highest summation for factor (B) is at the second level (i.e. B2, which is 150 gm) and the highest summation for factor (C) is at the third level (i.e. C3, which is 12% volume fraction). These predicted parameters are already used in the GFRP composite preparation as indicated in Table (4). It can be seen from Table (10) for the impact results of GFRP that the optimum levels were A3, B3 and C3 (i.e. 60ºC, 200 gm and 12%). These parameters were not used in the composite trails as indicated in Table (4). An experiment was conducted at the predicted parameters (A = 60ºC, B = 200 gm, and C = 12 % volume fraction), and the resulted specimen was tested by impact. The resulted impact strength was (82 MPa) which is greater than the impact strength values in Table (4). The optimum levels of Kevlar composites parameters for the tensile and impact strengths are similar to the results of impact of the glass fiber composites in which A3, B3 and C3 (i.e. 60ºC, 200 gm and 12%) had gave the highest contribution ratio. These parameters were also not used in the composite trails as indicated in Table (5). Two additional specimens of Kevlar fiber reinforced polyester were prepared at the predicted parameters (A = 60ºC, B = 200 gm, and C = 12 % volume fraction), and were tested by tensile and Osama Sultan M. Al-Khwarizmi Engineering Journal, Vol. 8, No.4, PP 53- 61 (2012) 60 impact. The resulted tensile strength was (121 MPa) while the impact strength was (109 MPa) which are greater than the strength values in Table (5). These results proved the success of Taguchi method in the prediction of the optimum parameters for higher tensile and impact strengths. In Tables (9, 10, 11 and 12), it was found that the fiber volume fraction contributes a larger impact on tensile and impact strength of the composites followed by curing temperature and then finally pressing load. 5. Conclusions In this research Taguchi's off – line quality control method was applied to determine the optimal process parameters by which a glass fiber reinforced polyester and Kevlar fiber reinforced polyester were prepared. For this purpose, concepts like orthogonal array, S/N ratio and ANOVA were employed. After determining the optimum process parameters, a confirmation experiments were conducted. In light of our analysis the following conclusions were drawn: 1. The optimum level of process parameters to obtain good tensile and impact strengths for the composites prepared by the hand lay-up method are 12% volume fraction of fibers, 60ºC curing temperature, and 150 gm pressing load for tensile strength and 200 gm for impact strength for the glass fiber reinforced polyester. While it is 12% volume fraction of fibers, 60ºC curing temperature, and 200 gm for tensile and impact strengths for the kevlar fiber reinforced polyester. 2. Taguchi method was successful in predicting the parameters that give the highest properties and it can find the most influential parameter regardless of the material used. 3. From the Pareto analysis it was evident that the volume fraction is a major contributing factor for improving tensile and impact strengths. 4. Taguchi method proved its success in predicting the optimum parameters to reach the best properties. 6. References [1] Zulkifli R., "Surface Fracture Analysis of Glass Fiber Reinforced Epoxy Composites Treated with Different Type of Coupling Agent", European Journal of Scientific Research, Vol.29, No.1, (2009), pp.55-65. [2] Iwaya T. et. al., "Recycling of fiber reinforced plastics using depolymerization by solvothermal reaction with catalyst", J Mat. Sci., Vol.43, (2008), pp.2452–2456. [3] Kang G. et. Al., "Uniaxial ratchetting of polymer and polymer matrix composites: Time-dependent experimental observations", Mat. Sci. Eng. A, Vol.523, (2009), pp.13–20. [4] Bagherpour S. et. al., "Effects of concentrated HCl on the mechanical properties of storage aged fiber glass polyester composite", Materials and Design, Vol.30, (2009), pp.271–274. [5] Wallace M. M., Bert C. W., "Experimental determination of dynamic young's modulus and damping of an Aramid-fabric/polyester composite material", Proc. Okla. Acad. Sci., Vol.59, (1979), pp.98-101. [6] Muhammed O. S., Saleh H. R., Alwan H. L., "Using of Taguchi Method to Optimize the Casting of Al–Si /Al2O3 Composites", Eng. & Tech. J., Vol.27, No.6, (2009), pp. 1143 – 1150. [7] Kamaruddin S., Khan Z. A., Foong S. H., "Application of Taguchi Method in the Optimization of Injection Moulding Parameters for Manufacturing Products from Plastic Blend", IACSIT International Journal of Engineering and Technology, Vol.2, No.6, (2010), pp. 574 – 580. [8] Mehravar R. et. al., "Applying the Taguchi Method for Optimized Fabrication of Lactalbumin Nanoparticles as Carrier in Drug Delivery and Food Science", Iranica Journal of Energy & Environment, Vol.2, no.1, (2011), pp.87-91. [9] Ross Phillips J., "Taguchi technique for quality engineering", New York: McGraw – Hill, (1988). [10] Roy Ranjit K., "A primer on Taguchi method", New York: Van Nostrad Reinhold, (1990). [11] Basavarajappa S., Chandramohan G., Paulo Davim J., "Application of Taguchi techniques to study dry sliding wear behaviour of metal matrix composites", Materials and Design, Vol. 28, (2007), pp. 1393 – 1398. )2012( 53- 61، صفحة 4، العدد8مجلة الخوارزمي الھندسیة المجلد أسامة سلطان محمد 61 دراسة تقییم ما بین البولي استر المقوى بالیاف الزجاج والبولي استر المقوى بالیاف الكفلر بواسطة طریقة تاكوجي اشراق عبد الرزاق كاظم عبد المجید میدحماجد *أسامة سلطان محمد الجامعة التكنولوجیة /دقسم ھندسة الموا oalaamiri@yahoo.com : االلكتروني البرید* الخالصة درجة (ات ھي تم في ھذا البحث تحضیر مادتین متراكبتین مختلفتین من البولیمر المقوى بااللیاف بواسطة طریقة التحضیر الیدوي وباستخدام ثالثة متغیر كمجموعة ف الزجاج تم تحضیر المواد المتراكبة من راتنج البولي استر كمادة اساس مقوى بالیا). حرارة المعالجة، حمل الكبس و الكسر الحجمي لاللیاف تم بعدھا اجراء اختباري %). ١٢، %٨، %٤(اولى من العینات والیاف الكفلر كمجموعة ثانیة من العینات وبكسور حجمیة مختلفة لكل نوع من االلیاف ھي لمتغیرات التي تعطي افضل مقاومة ان الھدف االساسي من البحث ھو استخدام طریقة تاكوجي للتكھن بافضل ا. مقاومة الشد ومقاومة الصدمة على العینات اظھرت النتائج . شد وصدمة للمواد المتراكبة ومن ثم تحضیر مواد متراكبة عند ھذه المتغیرات ومقارنتھا مع المتغیرات المستخدمة المختارة عشوائیا الخواص وكذلك یمكنھا ایجاد المتغیر االكثر تاثیرا على التجریبیة والتحلیلیة بان طریقة تاكوجي كانت ناجحة في التكھن وایجاد المتغیرات التي تعطي اعلى كذلك ان الفرق . اظھرت النتائج كذلك بان الكسر الحجمي كان المتغیر االكثر تاثیرا على مقاومة الشد والصدمة. الخواص بغض النظر عن المادة المستخدمة .ت الیاف الكفلر اعطت مقاومة شد وصدمة اعلىبین المادتین المحضرتین كان في قیم الخواص وان المادة المتراكبة ذا mailto:oalaamiri@yahoo.com