Composites Research

The Korean Society for Composite Materials (한국복합재료학회)
권호 표지
DOI QR코드

DOI QR Code

Acquisition of Parameters for Impact Damage Analysis of Sheet Molding Compound Based on Artificial Neural Network

인공신경망 기반 SMC 복합재료의 충돌 손상 해석을 위한 파라메터 획득

  • Lee, Sang-Cheol (Department of Aerospace Engineering, Pusan National University) ;
  • Kim, Jeong (Department of Aerospace Engineering, Pusan National University)
  • Received : 2021.02.15
  • Accepted : 2021.04.02
  • Published : 2021.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

SMC(Sheet molding compound) composite is mainly used for forming of vehicle's body. Considering the car accident, it is essential to research the impact behavior and characteristics of materials. It is difficult to identify them because the impact process is completed in a short time. Therefore, the impact damage analysis using FE(finite element) model is required for the impact behavior. The impact damage analysis requires the parameters for the damage model of SMC composite. In this paper, ANN(artificial neural network) technique is applied to obtain the parameters for the damage model of SMC composite. The surrogate model by ANN was constructed with the result in LS-DYNA. By comparing the absorption energy in drop weight test with the result of ANN model, the optimized parameters were obtained. The acquired parameters were validated by comparing the results of the experiment, the FE model and the ANN model.

복합재료 중에서 SMC(sheet molding compound) 복합재료는 자동차의 차체 성형에 주로 쓰이고 있다. 자동차 산업에서는 차량 사고를 고려하여야 하므로 재료의 충돌 거동 및 특성에 관한 연구는 필수적이다. 충돌은 짧은 시간에 일어나기 때문에 육안으로 확인이 어렵다. 따라서 충돌 거동을 확인하기 위해서는 유한요소 모델을 이용한 충돌 손상 해석이 필요하다. 충돌 손상 해석을 위해서는 SMC 복합재료의 손상 모델에 대한 파라메터가 요구된다. 본 연구에서는 SMC 복합재료의 손상 모델에 대한 파라메터를 획득하기 위해 인공신경망 기법을 적용하였다. LS-DYNA에서 파라메터에 따른 결과를 이용하여 대체 모델을 구성하였다. 자유 낙하 충돌 실험에서 얻은 흡수 에너지와 인공신경망 모델을 이용한 흡수 에너지를 비교하여 최적화된 파라메터를 획득하였다. 획득한 파라메터를 유한요소 모델에 적용해 결과를 비교하여 파라메터의 신뢰성을 검증하였다.

Keywords

References

  1. Jeon, O.H., "Technology Trends of Parts and their Materials for Automotive," Auto Journal: Journal of the Korean Society for Automotive Engineers, Vol. 42, No. 9, 2020, pp. 46-49.
  2. Noh, Y.J., Lee, Y.C., and Hwang, T.W., "Synergistic Effect in Mechanical Properties of Sheet Molding Compound via Simultaneous Incorporation of Glass Fiber and Glass Bubble Fillers," Composites Research, Vol. 31, No. 1, 2018, pp. 8-11. https://doi.org/10.7234/composres.2018.31.1.008
  3. Lee, C.H., and Huh, H., "Finite Element Analysis of SMC Compression Molding Processes," Transactions of Materials Processing, Vol. 4, No. 3, 1995, pp. 204-213.
  4. Lee, B.E., Kang, D.S., Park, E.T., Kim, J., Kang, B.S., and Song, W.J., "Numerical and Experimental Investigation on Impact Performance of Fiber Metal Laminates Based on Thermoplastic Composites," Transactions of the Korean Society of Automotive Engineers, Vol. 24, No. 5, 2016, pp. 566-574. https://doi.org/10.7467/KSAE.2016.24.5.566
  5. Ha, J.S., "Test and Finite Element Analysis on Compression after Impact Strength for Laminated Composite Structures of Unidirectional CFRP," Composites Research, Vol. 29, No. 6, 2016, pp. 321-327. https://doi.org/10.7234/composres.2016.29.6.321
  6. Won, J.M., and Shin, J.U., "Development of Artificial Neural Network Model for Prediction of Seismic Response of Building with Soil-structure Interaction," Journal of the Korean Geotechnical Society, Vol. 36, No. 8, 2020, pp. 7-15. https://doi.org/10.7843/KGS.2020.36.8.7
  7. ASTM D7136/D7136M, Standard Test Method for Measuring the Damage Resistance of a Fiber-Reinforced Polymer Matrix Composite to a Drop-Weight Impact Event, 2015.
  8. Choi, S.Y., Hong, S.C., Kim, J.S., and Kim, J.S., "Development of Composite Crash Box for Automobile with Optimum Crashworthiness," The Korean Society of Automotive Engineers Spring Conference Proceedings, 2018, pp. 1022-1031.
  9. Kim, S.P., The First Step in Deep Learning : From Machine Learning to Convolutional Neural Networks, Hanbit Media. Inc., Seoul, Korea, 2016.
  10. The Arup Campus, Introduction to Composites Modelling in LS-DYNA, Composites Webinar.
  11. Livermore Software Technology Corporation, LS-DYNA® Theory Manual, Mar. 2006.
  12. Woo, M.A., Lee, S.M., Lee, K.H., Song, W.J., and Kim, J., "Application of an Artificial Neural Network Model to Obtain Constitutive Equation Parameters of Materials in High Speed Forming Process," Journal of the Korean Society for Technology of Plasticity, Vol. 27, No. 6, 2018, pp. 331-338.