• 沒有找到結果。

Effect of material and process variables on the performance of resin-transfer-molded epoxy fabric composites

N/A
N/A
Protected

Academic year: 2021

Share "Effect of material and process variables on the performance of resin-transfer-molded epoxy fabric composites"

Copied!
7
0
0

加載中.... (立即查看全文)

全文

(1)

Performance of Resin-Transfer-Molded Epoxy Fabric

Composites

CHANG-LUN LEE, KUNG-HWA WEI

Department of Materials Science and Engineering, National Chiao Tung University, Hsinchu, 30049, Taiwan, Republic of China

Received 10 August 1999; accepted 23 October 1999

ABSTRACT: The effects of material and process variables on glass fabric-reinforced epoxy composites by the resin-transfer molding (RTM) process were studied. It was found that the molded aged resin with 55% fiber exhibited twice the mold-filling time and caused a 7–15% deterioration in the interlaminar shear strength (ILSS) and in the flexural strength of the composites as compared to those of the composites molded with fresh resin. At a 55% fiber volume fraction, composites molded with aged resin resulted in a 35% longer filling time and a 4 –12% decreased ILSS and flexural strength as compared to those of the composites at a 44% fiber volume fraction. Moldings with a perimeter inlet exhibited a 65% shorter mold-filling time, 28% reduced void content, and 6% improved flexural strength as compared to those of the composites molded with the center inlet.© 2000 John Wiley & Sons, Inc. J Appl Polym Sci 77: 2149 –2155, 2000 Key words: epoxy; composites; material and process variables; resin-transfer mold-ing (RTM); mechanical properties

INTRODUCTION

The resin-transfer molding (RTM) process for the production of fiber-reinforced composites has at-tracted much interest from aerospace and auto-motive to power machinery industries because of the potentially high rate of production and the high quality of the finished parts.1– 4Large poly-mer composite parts with complex shapes can be efficiently fabricated by the RTM process as com-pared to the traditional prepreg layup/autoclave cure process.2,4The RTM process is a closed-mold processing technique, which involves loading the fiber preform into a mold cavity, followed by an

injection of a low-viscosity thermoset polymer resin under pressure and at elevated temperature and then curing the resin in situ to form the final parts.4The critical step in RTM is to impregnate the preform as quickly as possible while minimiz-ing undesirable features such as void content or nonuniform wetting of the preform.5

A large number of variables are related to the RTM processes: the resin characteristics, fiber preform, resin preheated temperature, mold tem-perature, injection pressure, vacuum assistance, gating method, and mold geometry.5–12 The de-sign and the control of the critical variables of the RTM process are needed in the process-develop-ment stage to maintain optimal conditions.

The Darcy law has been used to describe the resin flow in the filling stage of RTM processes. The viscosity of the resin and the permeability of the fabric preform are the two important param-eters in the Darcy law that relates to the proper

Correspondence to: K.-H. Wei.

Contract grant sponsor: National Science Council, Taiwan, Republic of China; contract grant number: NSC 87-2623-D-009-005.

Journal of Applied Polymer Science, Vol. 77, 2149 –2155 (2000) © 2000 John Wiley & Sons, Inc.

(2)

fiber wet-out and the impregnation rate of the fabric preform.13,14 The viscosity of a thermoset resin is a function of temperature and the degree of cure (chemical reaction).15Resin aging arising from an unexpected prolonged storage under am-bient conditions may result in the curing of the resin and an increase in its viscosity. Hayward and Harris6found that a high-viscosity resin will not properly penetrate the fiber and wet-out the fibers in the RTM molding with 55% fiber as com-pared to that with 40% fiber in the case of a polyester resin. Akay16 and Berglund and Ken-ny17 studied the aging effect on carbon/epoxy prepregs for an autoclave process. It was found that aging increased the activation energy of the resin but caused no deterioration in the mechan-ical strength of the resulting composites. The fi-ber volume fraction of the composites has the most significant impact on the overall strength of the composites.8However, an increase in the fiber fraction provides resistance to resin flow and af-fects the cycle time of the RTM process.8,9 Han-sen10and Abraham and Mcllhagger11found that the mold filling and the part quality by the perim-eter inlet could be controlled much better than that by the center inlet in the RTM process. Cai12 demonstrated that the mold-filling time by the perimeter inlet is reduced substantially as com-pared to that by the center inlet in a simplified RTM process analysis.

Although various studies have attempted to characterize the RTM process, most of the previ-ous studies on RTM have been concentrated on flow modeling.5,13,18,19,20 Studies concerning the effects of material and processing variables on high volume fraction fiber composites by RTM are lacking. The effects of process variables on the physical and the mechanical behavior of an one-part high-performance epoxy composite produced by the RTM process were investigated.21 Vari-ables studied include injection pressure, injection temperature, and fabric structure.

In this study, the effect of the material and process variables on the performance of glass fab-ric-reinforced composites based on a two-part ep-oxy resin were investigated. The composites were produced by the RTM process under constant pressure and isothermal filling conditions. Vari-ables studied include resin aging, fiber volume fraction, and gating arrangement. The mold-fill-ing time, the void content, and the matrix-domi-nated properties, such as the interlaminar shear strength (ILSS), the flexural strength, and the

dynamic mechanical properties of RTM compos-ites are presented.

EXPERIMENTAL

Materials

The resin used was a two-part epoxy/amine resin LY564/HY2954 supplied by Ciba–Geigy (Haw-thorne, NY). The mixing ratio of the base resin LY564 to the hardener HY2954 is 100:35 by weight. Detailed information pertaining to this resin can be found elsewhere.22 Both the fresh and the aged resin were used for the processing of RTM moldings. The aged resin is designated as the matrix resin and the hardener, which has been stored in a separate container under ambi-ent conditions of 70% relative humidity and 25°C temperature up to 18 months. The fabric rein-forcement used was a highly permeable glass fab-ric from Brochier (Decines Cedex, France), Injectex EF420-E01-100, with a surface density of 420 g/m2. Preparation of RTM Composites

A flat rectangular aluminum mold with the di-mensions of 0.325⫻ 0.325 ⫻ 0.003 m was used for the preparation of the RTM composites. It com-prised a top and a bottom half as well as a picture frame with a center gate or a perimeter line gate at the corner on the top half, clamped together with 12 evenly distributed bolts around the mold edge. Both the center inlet and the perimeter inlet, as shown in Figure 1, were used in this study. A rubber gasket was placed around the perimeter of the mold halves to provide a proper seal. Heat blankets were used for the heating of the mold and the resin pot. Ten and eight layers of glass fabrics of EF420 were used to produce Figure 1 Different gating arrangements in the RTM process.

(3)

3-mm-thick plates of of a 55 and 44% fiber volume fraction, respectively. The resin was preheated and degassed at 35°C for 20 min. Then, the resin was injected into the mold maintained at 70°C under 147 kPa. The injection pressure was mon-itored with a Flutterless威 gauge by adjusting the air pressure in the resin pot. The mold was vented to the atmosphere once the resin injection started. Two J-type thermocouples neighboring to the center inlet and the perimeter inlet were used to monitor the temperature throughout the whole process. After the resin flowed out from each out-let, the outlets were opened and closed alter-nately several times to eliminate any entrapped air in the mold cavity. Upon completion of the mold filling, the mold temperature was increased to 140°C for 4 h to cure the plates. The cured plates were then cooled, demolded to the tradi-tional, followed by examining for quality by ultra-sonic C-scanning. The average thickness varia-tion from the center to the edge of the molded samples was 0.14 mm, which indicates that the injection pressure exerted on the mold caused the mold to open slightly. Figure 2 displays the ex-perimental setup for the RTM process. Except when a specific variable was studied, the RTM composites were produced under the standard settings of the process variables as given in Ta-ble I.

Measurements of Viscosity and Dynamic Mechanical Properties

The resin viscosity and the dynamic mechanical properties, including storage modulus (G⬘), loss modulus (G⬙), and loss tangent (tan␦), of the RTM composites were measured with a Rheometrics mechanical spectrometer Model RMS-605 in ac-cordance with the specification of ASTM D 4473 and ASTM D 4065, respectively. Viscosity

mea-surements were performed with disposable paral-lel plates of 25 mm in diameter, which were sub-jected to the forced oscillations with a gap of 0.5 mm. A time sweep was conducted to obtain the viscosity–time profile at 70°C. The composite specimens were subjected to the forced oscilla-tions at a frequency of 1 Hz and 0.08% strain over the temperature range of 25–250°C. The maxi-mum value of G⬙/G⬘, that is, tan␦, was taken as the glass transition temperatures of the compos-ites.

Testing of Physical and Mechanical Properties Ultrasonic C-scan inspection and optical image analysis were used to investigate the void content in the composites. The composite plates were first inspected using the ultrasonic C-scan test. The C-scan analysis provided basic information of the

Figure 3 Microphotograph of the cross-sectional view of EF420/LY564 RTM laminates.

Figure 2 Experimental setup for the RTM process.

Table I Standard Settings of Process Variables for EF420/LY564 RTM Composites

Variables Settings

Transfer pressure 147 kPa

Mold temperature 70°C

Resin temperature 35°C

Cure condition 140°C, 4 h

Fabric temperature 70°C

Resin viscosity at 25°C 617 mPa s

Pack pressure 196 kPa

Fiber volume fraction 55⫾ 2%

Nominal thickness of panels 3.0 mm

(4)

void distribution in the composites. After the C-scan analysis, the void content was determined from polished cross sections of the plate with a computerized image analyzer coupled with a ZEISS optical microscope. The detailed proce-dures of the ultrasonic C-scan and image analysis are described elsewhere.21A typical microphoto-graph of the cross-sectional view of EF420/LY564 RTM laminates is shown in Figure 3. The voids were identified as the black dots in the figure. The ILSS was tested in accordance with ASTM D 2344 using a three-point loading fixture with the span-to-depth ratio of 5. The three-point flexural tests were run in accordance with ASTM D790 at a span-to-depth ratio of 32. Seven specimens were measured for each ILSS and each flexural test.

RESULTS AND DISCUSSION

Effect of Aging

The resin flow in the RTM process depends strongly on its viscosity, which is a function of temperature and the degree of cure.15The evolu-tion of viscosity during mold filling at 70°C for the aged and the as-received (fresh) LY564/HY2954 resin is shown in Figure 4. In Figure 4, the aged resin exhibits a higher viscosity than that of the fresh resin. Hence, the time period allowed for the mold filling, at which the viscosity of the aged

resin is below 1000 mPa s, was reduced. The increase in the viscosity of the aged resin is prob-ably due to the formation of a polymer network arising from its prolonged storage. In RTM, mold filling can be treated as a pseudo-steady flow of Newtonian fluids through porous media obeying the Darcy law. Hence, the average resin velocity is proportional to the average pressure gradient, but is inversely proportional to the resin viscosity. An increase in the resin viscosity leads to a de-crease in the resin flow rate and an inde-crease in the mold-filling time under the same injection pressure. Aged resin exhibited almost twice as much mold-filling time and void content as that in the case of the as-received resin, as given in Table II. The influence of resin aging on the mechanical properties of the EF420 glass fabric composites are also given in Table II.

In Table II, the interlaminar shear strength (ILSS) and the flexural strength of the plates molded from aged resin were 7–15% inferior to that of the plates molded from the fresh resin. The decrease in ILSS and flexural strength was caused by an increase in the void content in the plates of the aged resin, which led to a reduction of the cross-sectional area and an initiation of failure from large discrete voids.23 Voids ap-peared because the penetration of the fabrics and the wet-out of the fibers were hindered by the higher viscosity of the aged resin as discussed by Figure 4 Viscosity profiles as a function of time for as-received and aged LY564/

(5)

Chen et al.24 The ILSS often decreases with in-creasing void content.25Flexural strength usually serves as an indicator of the reinforcing efficiency of the plate.6 The difference in the flexural mod-ulus of the composites molded from the aged resin and the fresh resin was very small because the modulus is largely controlled by the fibers. The aged resin also resulted in the deterioration in the storage modulus (G⬘), the early peak of loss mod-ulus (G⬙), and the lower glass transition temper-ature (Tg), as shown in Figure 5 and given in

Table II, due to the poor wetting and the poor quality in the resulting composites.

Effect of Fiber Volume Fraction

The fiber volume fraction is one of the most im-portant factors affecting the mechanical strength

of high-performance composites. However, more fibers can retard the flow of resin in the RTM process. The aged resin was used to produce RTM composites of different fiber volume fractions for comparison. The mold-filling time increased by 35% when the fiber volume fraction increased from 44 to 55% as given in Table III. This can be attributed to more resistance to the resin flow in the fabric reinforcement owing to the decreased permeability of the fabric preform. The leading of either the macroflow fronts around the fiber bun-dles or the microflow fronts within a fiber bundle is likely to form the voids during the filling stage.18 –20Thus, voids in the 44% fiber fraction composite were 66% less than that of the 55% fiber fraction composite as given in Table III. The difference of the impregnation rates between the Table II Effect of Aged Resin on the Mechanical Properties and the Filling Time of EF420/LY564 Composites Resin Conditions ILSS (MPa) Flexural Strength (MPa) Flexural Modulus (GPa) Void Content (%) Tg (°C) Filling Timea (s) As-received 50.3 (4.8) 393.7 (12.4) 22.8 (1.4) 0.83 (0.49) 153 193 Aged 42.8 (1.4) 365.4 (15.8) 22.1 (0.7) 1.60 (1.10) 139 405

( ) denotes standard deviation.

aFilling time obtained as the slowest resin flow-out.

Figure 5 Storage modulus (G⬘), loss modulus (G⬙), and loss tangent (tan ␦) as a function of temperature for EF420/LY564 composites produced with both as-received and aged resin.

(6)

macropores of fiber bundles and the micropores among individual fiber filaments was diminished by having a lower fiber fraction. The 44% fiber fraction composite had a higher flexural strength and a higher ILSS than that of the 55% fiber fraction composite as a result of a reduction of voids in the resulting composites, as listed in Ta-ble III. However, the decreased fiber fraction leads to a lower flexural modulus and a lower Tg,

as given in Table III, due to the increased resin content and the significant loss of stiffness in the composites. In general, glass reinforcement im-posed the additional constraint/stiffness on the resin matrix in a similar way as does increased crosslinking.26 Crosslinking increases the glass transition of a polymer by introducing restrictions on the molecular motions of a chain.27

Effect of Gating Arrangement

The center inlet and the perimeter inlet for the moldings were used to assess the effect of differ-ent gatings on the performance of the RTM pro-cess and the resulting composites. The fill-ing time was dramatically reduced for the mold-ings with the perimeter inlet as compared to that with the center inlet, as given in Table IV. The reduction in the mold-filling time in the perimeter inlet case can be explained by the fact that the much higher flow rate in the perimeter line gate

and a relatively low resin viscosity at the initial stage of mold filling resulted in such a large re-duction of the mold-filling time. For the molding with the center point inlet under the same pres-sure, the inlet flowrate is relatively small and decreases with time in an outward direction.11,12 Therefore, the mold-filling rate slows down and the resin viscosity increases with time, and voids are, therefore, formed. It is evident that the dif-ference in the impregnation rates between the macropores and micropores facilitated the forma-tion of voids in the moldings with the center inlet. As shown in Table IV, moldings in the center inlet case exhibited a much higher void content than that in the perimeter inlet case. Additionally, de-creases in the ILSS and the flexural strength were observed for the moldings in the center inlet case as compared to those in the perimeter inlet case. As expected, the two moldings exhibited about the same glass transition temperatures (Tg). Evidently, the perimeter inlet is a better

gating arrangement for mold filling in the RTM process as compared to the center inlet.

CONCLUSIONS

Resin-transfer molded glass fabric composites based on a two-part epoxy were produced under Table III Effect of Fiber Volume Fraction on the Mechanical Properties and the Filling Time of EF420/LY564A Composites Fiber Volume (%) ILSS (MPa) Flexural Strength (MPa) Flexural Modulus (GPa) Void Content (%) Tg (°C) Filling Timea (s) 44 48.3 (2.8) 382.0 (13.8) 19.3 (0.7) 0.54 (0.38) 134 300 55 42.7 (1.4) 365.4 (15.8) 22.1 (0.7) 1.60 (1.10) 139 405

LY564A: aged resin. ( ) denotes standard deviation. aFilling time obtained as the slowest resin flow-out.

Table IV Effect of Different Gating on the Mechanical Behavior and Filling Time of EF420/LY564 Composites Gating Arrangement ILSS (MPa) Flexural Strength (MPa) Void Content (%) Tg (°C) Filling Time (s) Center inlet 50.3 (4.8) 393.7 (12.4) 0.83 (0.49) 153 193a Perimeter inlet 51.7 (3.7) 419.2 (15.8) 0.60 (0.14) 152 68

( ) denotes standard deviation.

(7)

various material and processing conditions. Vari-ables investigated included resin aging, the fiber volume fraction, and the gating arrangement. Aged resin resulted in a twofold mold-filling time, 7–15% reduced mechanical properties, and a lower glass transition temperature in the compos-ites as compared to those of the composcompos-ites molded from fresh resin. The increase in the glass fiber content from 44 to 55% led to a 35% longer filling time and a 4 –12% decrease in the mechan-ical properties because of the increased void con-tent. Molding in the perimeter inlet case exhib-ited a 65% less mold-filling time, a 28% lower void content, and a 6% better flexural strength as com-pared to the molding in the center inlet case. This work was funded by the National Science Council, Taiwan, Republic of China (Contract NSC 87-2623-D-009-005).

REFERENCES

1. Pilato, L. A.; Michno, M. J. Advanced Composite Materials; Springer-Verlag: Berlin, Heidelberg, New York, 1994.

2. Hasko, C.; Dexter, H. B.; Loos, A.; Kranbuehl, D. J Adv Mater 1994, Oct., 9 –15.

3. Rosenwasser, S.; Nagy, G.; Mehle, G. IEEE Trans Magnet 1999, 35, 307–311.

4. Beckwith, S. W.; Hyland, C. R. SAMPE J 1998, 34, 7–19.

5. Chan, A. W.; Larive, D. E.; Morgan, R. J. J Compos Mater 1993, 27, 996 –1008.

6. Hayward, J. S.; Harris, B. SAMPE J 1990, 26, 39 – 46.

7. Mychajluk, G.; Manoochehri, S.; Parnas, R. S. J Adv Mater 1996, 28, 9 –18.

8. Spoerre, J.; Zhang, C.; Wang, B.; Panas, R. S. J Compos Mater 1998, 32, 1244 –1272.

9. Rudd, C. D.; Owen, M. J.; Middleton, V. Mater Sci Technol 1990, 6, 656 – 665.

10. Hensen, R. S. RTM Processing and Applications; Society of Manufacturing Engineers (SME) Tech Paper, 1990; pp 214 –225.

11. Abraham, D; Mcllhagger, R. Composites A 1998, 29, 533–539.

12. Cai, Z. J Compos Mater 1992, 26, 1310 –1338. 13. Bruschke, M. V.; Advani, S. G. SAMPE Q 1991,

Oct., 2–11.

14. Shin, C. H.; Lee, L. J. Polym Compos 1998, 19, 626 – 639.

15. Yousefi, A.; Lafleur, P. G.; Gauvin, R. Polym Com-pos 1997, 18, 157–168.

16. Akay, M. Compos Sci Technol 1990, 38, 359 –370. 17. Berglund, L. A.; Kenny, J. M. SAMPE J 1991, 27,

27–37.

18. Parnas, R. S.; Salem, A. J.; Sadiq, T. A. K.; Wang, H. P.; Advani, S. G. Compos Struct 1994, 27, 93– 107.

19. Parnas, R. S.; Phelan, F. R., Jr. SAMPE Q 1991, Jan., 53– 60.

20. Chan, A. W.; Morgan, R. J. SAMPE Q 1992, Apr., 48 –52.

21. Lee, C. L.; Wei, K. H. Polym Eng Sci, accepted. 22. Lee, C. L.; Wei, K. H. J Appl Polym Sci, accepted. 23. Wisnom, M. R.; Reynolds, T.; Gwilliam, N. Compos

Sci Technol 1996, 56, 93–101.

24. Chen, Y. T.; Davis, H. T.; Macosko, C. W. AIChE J 1995, 41, 2261–2273.

25. Mallick, P. K. Fiber-Reinforced Composites Mate-rials, Manufacturing, and Design; Marcel Dekker: New York, 1988; p 214.

26. Akay, M Compos Sci Technol 1993, 47, 419 – 423. 27. Nielsen, L. E. Mechanical Properties of Polymers

and Composites; Marcel Dekker: New York, 1974; Vol. 1, p 23.

數據

Table I Standard Settings of Process Variables for EF420/LY564 RTM Composites
Figure 5 Storage modulus (G ⬘), loss modulus (G⬙), and loss tangent (tan ␦) as a function of temperature for EF420/LY564 composites produced with both as-received and aged resin.
Table IV Effect of Different Gating on the Mechanical Behavior and Filling Time of EF420/LY564 Composites Gating Arrangement ILSS (MPa) Flexural Strength(MPa) Void Content(%) T g (°C) Filling Time(s) Center inlet 50.3 (4.8) 393.7 (12.4) 0.83 (0.49) 153 193

參考文獻

相關文件

6 《中論·觀因緣品》,《佛藏要籍選刊》第 9 冊,上海古籍出版社 1994 年版,第 1

The first row shows the eyespot with white inner ring, black middle ring, and yellow outer ring in Bicyclus anynana.. The second row provides the eyespot with black inner ring

 develop a better understanding of the design and the features of the English Language curriculum with an emphasis on the senior secondary level;..  gain an insight into the

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

Hope theory: A member of the positive psychology family. Lopez (Eds.), Handbook of positive

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

Monopolies in synchronous distributed systems (Peleg 1998; Peleg

Corollary 13.3. For, if C is simple and lies in D, the function f is analytic at each point interior to and on C; so we apply the Cauchy-Goursat theorem directly. On the other hand,