Vol.15 No.3

CHINESE JOL RNAL OF AERONALTICS

August 2002

Studies on Rheological Behaviors of Bismaleimide Resin System for Resin Transfer Molding

LU Yao, DUAN Yue-cin, LIANG Zhi-yong (School of Material Science & Engineering, Beijing University of Aeronautics and A stronautics,

Beijing 100083, Ch ina)

Abstract: The rheological behavior of bismaleimide resin for resin transfer molding(RTM) was studied with DSC analysis and viscosity experiments. A rheological model based on the dual-Arrhe-nius equation w as est ablished and used to simulate the rheological behavior of t he resin. The model predictions determined from the dual-A rrhenius equation were in good agreement with experimental data. T he processing window of the resin system can be well determined based on the developed model. T he rheological model is important for processing simulation and quality control of RTM processing for high performance composites-

Key words : composites; RTM; bismaleimide resin; rheological models; curing reaction kinetics

. KKW, ^^ . 2002,15

( 3) : 181- 185.

rn w «isMM^rn^nrnM^RTM

%Mft^M^ttaTTm%,aadsc w^mmimmmm^

SfiS® RT M IX

« p,^a®^ RTM I.

RTM IX; «IS;

1000-9361(2002) 03-0181-05

The RTM process has been recognized as a very high potential process for advanced composites manufacturing, and has been widely used in many industry fields such as aerospace, mobile and civil architecture. The RTM process must depend on the special low viscosity resin, whose rheologi-cal behavior contains a low viscosity area meeting the demand of the RTM. This means, the viscosity of the resin should maintain 50-200cP when injected in molding for complete fill and fiber wet-[12]

ting ' . Therefore, through establishing the chemorheology model of the resin-viscosity function about temperature and time(or cure degree), the resin viscosity at any temperature and time(or cure degree) can be calculated, and the low viscos-

ity area can be defined precisely. Hence, one can optimize the RT M process parameters and improve the quality of the products[3].

QY8911-4 BMI resin is a suitable resin for RTM. In order to optimize the RTM process parameters and improve the quality of the products, the chemorheological behavior of the QY8911-4 BM I resin is studied, its rheological model is established, and a processing window is predicted[4-8]

1 Viscosity Experiment

QY8911-4 BMI resin is prepared for determination of viscosity. Viscosity determinations are carried out at dynamic temperatures and selected

Received date: 2002-03-20; Revision received date: 2002-03-21

Foundation items: N ational N atural Science Foundation of C hina( 59833110) and N ational Defence Foun dation( 00j00. 5.3. hk 0144) ^ide LRL ■ http:/,'ww w . hkxb. ret. cri qa' 2°°2' 03,0181/

LU Yao, DUAN Yue-xin, LIANG Zhi-yong

temperatures, using NDJ-1 rotating viscometer,

according to GB7193. 1.

1. 1 Viscosity at heating condition

The viscosity curve of QY8911-4 BMI resin is shown as Fig- 1.

ig. 1 Viscosity-temperature curve of BMI resin system

For QY 8911-4 BMI resin, its viscosity will descend with the heating condition- On the other hand, it will ascend because of resin cross-linking. During heating the resin, viscosity descending is primary at lower temperatures; otherwise ascending is primary at higher temperatures. The viscosity will be a minimum when the two effects keep balance- QY8911-4 BMI resin will be melted at 80 C. So viscosity measurement begins at 80 C. Viscosity maintains under 800cP between 80 C and 140 C. Therefore, isothermal viscosity measurement is selected in the range of 90-140 C. 1. 2 Isothermal viscosity measurement

According to the dynamic viscosity curve, isothermal viscosity measurements are carried out at 90C, 126 C and 135 C. The curves of isothermal viscosity are shown as Fig. 2. The viscosity changing rate is improved with time and temperature.

Temperature/ "C

thermal viscosity-time curv

2 Dual Arrhenius Model

In the case of isothermal cure, the viscosity, which is affected by time and temperature, can be described into the dual Arrhenius model given

ln T, t) = ln ^ + E o/( R T) +

k -texp[ Ek/ (RT)] (1)

where and kro are the Arrhenius preexponential factors, respectively, while Ei and Ek are respectively the activation energies for the flow and the curing reaction. R is an ideal gas constant. T he model can be simplified

ln A = ln Oo + E o/( RT) (2)

ln B = ln kro + Ek /(RT) (3)

So the model can be described as

ln n( T, t) = ln A + Bt (4)

The curves of lnH-t at the isothermal temperature are shown as Fig. 3.

Fig. 3 ln^ vs t curves at isothermal temperature According to Fig. 3, the model parameters lnA and B can be calculated; the results are presented in T able 1.

Table 1 Parameters of the viscosity model

T emperature/ C lnA B

90 5. 97 0. 008

126 4. 43 0. 039

135 3.44 0. 060

The curves of ln A -1/ T and lnB -1/ T are presented in Fig. 4 and Fig. 5 respectively.

According to Fig. 4 and Fig. 5, one can obtain the following relation about the parameters lnA and lnB

lnA = - 14.91 + 7603/ T (5)

lnB = 12.98 - 6465/ T (6)

©l1 Flg'2) 1 Publishing • ei • , , , ■ t

Note that the time unit should be changed

August 2002

Studies on Rheological Behaviors of Bismaleimide Resin System for Resin Transfer Molding

Fig. 4 lnA v.s 1/ T of the dual-Arrhenius equation

Fig. 5 lnB vs 1/ T of t he dual-Arrhenius equation from minute to second, and the rheological model of Q Y8911-4 BM I resin can be presented as lnn = - 14. 91 + 7603/T +

texp( 17. 07 - 6465/ T) (7)

For nonisothermal cure, where the resin temperature history is given by T = f (t), the viscosity can be presented as an integral formation as

ln n = - 14.91 + Jexp 17.07-

T (t) 6465

T (t )J

3 Reaction Kinetics

The curing reaction kinetics of the resin is studied by DSC analy sis[9-11J. In this paper, a kinetic model based on Kissinger function is established. The parameters in the kinetic model preexponential factor and the activation energy for the curing reaction are compared with the parameters in the dual Arrhenius model, so the rationality of the dual Arrhenius model and the physics definitions of the model parameters are verified. 3. 1 DSC analysis

made in American Rheometric Scientific Company is used in this study. The resin samples are scanned at different heating rates as 5C/min, 10 C /min and 20 C/min. And the results are shown in Fig. 6.

Fig. 6 Dynamic DSC curves at various heating rates

3. 2 Data analysis

The Kissinger function is performed to describe the curing reaction kinetics of the thermalset resin. The relation between the temperature at maximum rate o f the DSC curve and the constant

heating rate is shown as follow s

T 2 T m

(lnA + lnR - lnE) -

where ^is the constant heating rate; Tm is temperature at maximum rate; A is preexponential factor, and E is the activation energy of the curing reaction.

According to Eq.(9) , plot ln(4/ T2.) vs 1/ Tm as Fig. 7. The slope is - E/R, and the intercept is lnA + lnR - lnE.

Fig. 7 ln(<b Tm) vs 1/ Tm The activation energy for QY8911- resin system is 98486J/mol, and the preexponential factor is 21.61.

The reaction order n can be determined by the €Differential Scanring Calorimetry (DSC) peak

shape factor s of the DSC curve. S is defined

LU Yao, DUAN Yue-cin, LIANG Zhi-yong

as the absolute ratio value of the tangent slopes at the curve inflection point. Therefore, n can be obtained as the following equation[13]

n ~ 1.26S1/2 (10)

The reaction order is 2. 076 from Fig. 6. So the curing reaction kinetic model of QY8911-4 resin system is shown as

dtt 1 - 11846/ N 2.076 , 1 1 N

— = 21. 61e (1- Of) (11)

Eq. 11 can be presented as an integral form a-tio n as

a= 1- [ 1 - (1 - 2. 076)21. 61te(- 11846/^jH^

The curing degree vs time at different isothermal temperatures is shown as Fig. 8.

Table 2 The parameters of the two models

Fig. 8 Isothermal curing degree vs time curves According to Fig. 8, the curing degree vs time shows the linear relation on the initial stage of the curing reaction, and the time maintaining the linear relation accords with the time maintaining the low viscosity in the dual Arrhenius model. It is certain that the assumption for the dual Arrhenius model(the curing degree vs time shows the linear relation in the low viscosity area) is right.

The parameter relationship between the dual Arrhenius model and the Kissinger equation is presented as the following experience equation

E = En-1 and k = k—1 (13)

where E and k are the activation energy and the preexponential factor from the Kissinger equation; Ek and k— are from the dual Arrhenius model; n is the reaction order.

The results of the parameters of the dual Ar-rhenius model compared with the parameters of the

para dual Arrhenius Kiss inger Calc. error / %

k 17. 07 21. 61 21. 18 1. 96

E/ R 6465 11846 12594 6. 32

According to the DSC analysis, one can conclude that the resin viscosity at the low curing degree area can be calculated by the dual Arrhenius model, and the parameters in the dual Arrhenius model and in curing reaction kinetics have the corresponding relation.

4 Model Application

4.1 Prediction of processing window

The resin viscosity at a certain temperature and time can be calculated by the dual Arrhenius model. The viscosity map is shown as Fig. 9.

Fig. 9 Viscosity map The temperatures, at which the resin viscosity is less than 800cP and the time maintaining the low viscosity is more than 40 minutes, are 90 C, 100 C , 110 C and 120C. Therefore, the processing windows of the QY8911-4 BMI resin system for the RTM processing are the temperature range from 90 to 120 C. 4. 2 Viscosity simulation

For nonisothermal cure, where the resin temperature history is given by T = f (t) , the viscosity can be presented as the integral formation as Eq. 8. So the viscosity can be simulated at the given condition of the temperature as a function of time, and the results are shown in Fig. 10.

The resin is heated at a constant rate, maintained at an assigned temperature for about 100

Kissinger equation are shown in Table 2. [ectronicPul minutes, and then heated again to cure complete—

August 2002

Studies on Rheological Behaviors of Bismaleimide Resin System for Res in T ransfer Molding

ig. 10 Calculated viscosity during processing procedure ly. From Fig. 10, one can get the same trends at different conditions: At the first heating stage, the viscosity drops continually, at isothermal stage, the viscosity rises gradually, and at the second heating stage, the viscosity drops momently and rises quickly to a gel point.

Compared with different viscosity curves, at 110C the viscosity can maintain less than 300cP fo r 55 minutes. So the resin can be suitable to manufacture the medium or small sized structures as the RTM processing. One can confirm that the temperature at 110C is the suitable injection temperature. If the parts are relatively large and the injection time is long, the temperatures at 90100 C are the suitable injection temperature.

5 Conclusions

QY8911-4 BMI resin system is the suitable resin for the RTM processing. The dual Arrhenius model can be used to present the rheological behavior of the QY 8911-4 BMI resin at low viscosity areas. And the model predictions are in good agreement with experimental data. The processing window for the RTM processing can be predicted precisely by the dual Arrhenius model, and the viscosity can be simulated during the processing procedure. The rheological model is important for processing optimization and quality control of RTM processing for high performance composites.

References

[1] ^S,g», . rtm

^S®^) [ RJ . ^«ffiSIK^ 104 ^W^, 1994.

(Liang Z Y, Duan Y X, Zhang Z G. Report on RTM pro© ^¿"I te hlj'ol^gy ^ijpi^tWR] Beijing ^"^c polishing House. All rights reserved, http://www.cnki.net

versity of Aeronautics and Astronautics 104 Staff Room,

1994.)

Beckwith S Wi Craig R H- Resin transfer molding , a decade of technology advances [J] . SAMPE Journal, 1998, 34(6): 297- 311.

Gutowski T G. Advanced composites manufacturing [M] . New York: John Wiley &Sons Inc, 1997. 393- 457. Rudd C D, Long A C, Kendall K N, et al. Liquid molding technologies [ M] . London: Woodhead Publishing Ltd, 1997. 23- 28.

Loos A C, Springer G S. Curing of epoxy matrix composites [J] . Journal of composite materials, 1983, 17 (3): 135167.

Halley P J, Mack ay M E. Chemorheology of thermoset - an overview [ J] . Polymer Engineering and Science, 1996, 36 (5) : 593- 608.

Panagiotis I K Ivana K P, Attwood David. Modeling the cure of a commercial epoxy resin for applications in resin transfer molding [J]. Polymer International, 1996, 41 (2): 183- 191.

gJRft . EPON862

[J]. Ä^WM^fö, 2001, 18 ( 1): 16- 19. (Liang Z Y, Duan Y X, Lin Y. Studies on rheological behaviors of EPON862 epoxy resin system for resin transfer mold-ing[ J]. Acta Materiae Compositae Sinica, 2001, 18 ( 1) : 1619.)

[9] Eckold C. Design and manufacture of composite structure [M]. New York: McGraw-Hill Inc, 1994. 271- 277.

[10] Kamal M R, Sourour C. Kinetics and thermal characterization of thermoset cure [ J] . Polymer Engineering and Science. 1973 ( 1) : 58- 77.

[11] Lipshitz S D, Macosko C W. Rheological changes during a urethane network polymerization [J]. Polymer Engineering and Science, 1976 (16) : 26- 38.

[12] Shin D D, Hahn H T. A consistent cure kinetic model for AS4/3502 graphite/epoxy [J]. Composites Part A, 2000, 31: 991- 999.

[13] ffi^Ä, ft®: dtBffi^ffi^

1994.44- 45.

( Gao J W. Modern testing technology on polymer material [ M] .Beijing: Beijing University of Aeronautics and Astronaut ies Publishing House, 1994.44- 45. (in Chinese))

Biographies:

LU Yao Born in 1977, he received B. S. and M.S. from Beijing University of Aeronautics and Astronautics in 1999 and 2002 respectively. Tel: (010) 82317127, E-mail: luyao2000 @263.net