JEC COMPOSITES MAGAZINE - Issue #67 - August/September 2011 - 52

simulation

the composite. Then, the demoulding process is represented with the help of the boundary conditions illustrated in Figure 4 (b), allowing bending of the L-shaped composite with respect to the Z axis.

Numerical results
The simulations show the presence of residual stresses in the composite structures during the manufacturing process. The degree of cure at the end of manufacturing is 98%, which represents a fully cured composite. Figure 5 shows the final deformed shape and the initial mesh. Obviously, the angles are smaller than 90° at the end of the composite elaboration.
Fig. 5: Distortion of part 1 at the end of the cure cycle

obtain the targeted accuracy. The difference between numerical and experimental results may be largely induced by material property data, as shown by several complementary sensitivity studies. The mean value of the chemical shrinkage coefficient for carbon/epoxy materials is taken from the literature.

Conclusions
Due to the coupled phenomena, the macroscopic thermo-chemicomechanical model is used, taking into account the existence of the rubbery and glassy states of the resin during the composite manufacturing process. This model can predict the magnitude of residual stresses and shape distortions of the composite part. The degree of cure reaches 98% at the end of manufacturing, i.e. a fully reticulated matrix, with homogeneous curing in the composite thickness. This model is then validated by experimental measurement confrontations. The L-shaped composite parts manufactured comply with the Hexcel Composites porosity requirements (max. 1%). The distortions are clearly demonstrated by angular measurements. The measured spring-in is about 1°. The sensibility study of the numerical model shows that a careful knowledge of the material is required and that the mechanical boundary conditions and the absence of part/mould contact in the simulation can cause additional numerical errors. n
More information: klinkova@enise.fr, drapier@emse.fr, bergheau@enise.fr References

The predicted values of and are dissimilar, but they differ from the measured values. However, the observed trend is the same: < (see Figure 3 (a), Figure 3 (b). The values of the predicted spring-in angles for both manufactured composite parts are = 0.87° and = 0.90° and the relative angular difference, dr, is 3.4%. The inner and outer spring-in angles are slightly different than expected. The angular variation of the L-shaped composite part is about one degree, which is consistent with literature [1, 6, 7]. Table 1 synthesizes the experimental and numerical results and their comparison. It can be concluded from these results that the model is capable of predicting the angular and thickness variation trends caused by the composite manufacturing process, Table 1: Differences between simulation 1 and but one must be careful experiment 1, where is the mean value of with the material the respective measurements, σ the standard deviation and Cv the variation coefficient properties in order to

Discussion

jec composites magazine / No67 August - September 2011

[1] L.K. Jain, M. Hou, L. Ye, and Y.W. Mai. Spring-in study of the aileron rib manufactured from advanced thermoplastic composite. Composite Part A, 29(82):973–979, 1998. [2] O. Klinkova, S. Drapier, and J.M. Bergheau. Simulation des contraintes résiduelles dans les structures composites : déformée après cuisson et usinage. In 19 ème Congrès Français de Mécanique, pages 1–6, 2009. [3] J. Lange, S. Toll, J-A.E. Manson, and A. Hult. Residual stress build-up in thermoset films cured above their ultimate glass transition temperature. Polymer, 36(16):3135–3141, 1995. [4] P . Parlevliet, H.E.N. Bersee, and A. Beukers. Residual stresses in .P thermoplastic composites – a study of literature. Part III: Effects of thermal residual stresses. Composites: Part A, 38:1581–1596, 2007. [5] M. Shimbo, M. Ochi, T. Inamura, and M. Inoue. Internal stress of epoxide resin modified with spiro ortho-ester type resin. Journal of Materials Science, 20:2965– 2972, 1985. [6] M.J. Svanberg and J.A. Holmberg. Predictions of Manufacturing Induced Shape Distortions. PhD thesis, Lulea University of Technology, Sweden, 2002. [7] Q. Zhu, P Geubelle, M. Li, and C.L. Tucker. Dimensional accuracy of .H. thermoset composite: Simulation of process-induced residual stresses. Journal of Composite Materials, 35(24):2171–2205, 2001. [8] ESI Group. SYSTUS Analysis reference manual. www.esi-group.com, Nov. 2008.

http://www.esi-group.com

JEC COMPOSITES MAGAZINE - Issue #67 - August/September 2011

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #67 - August/September 2011