运用神经网络设计RTM成型复合材料力学性能

Mechanical property design of RTM composites using neural networks

提出了一种设计RTM成型碳纤维织物/环氧复合材料力学性能的新方法.利用BP神经网络,以RTM成型工艺中注模压力、温度和时间为输入量,以复合材料层间剪切强度和弯曲强度为输出量,建立了反应工艺参数与力学性能内在规律的数学模型.利用此模型研究了在注模温度、时间确定的条件下注模压力与复合材料层间剪切强度的关系,网络输出的注模压力对复合材料层间剪切强度的影响规律与实验规律非常接近,说明建立的工艺参数与复合材料力学性能的关系模型是可靠的,可以用此模型对复合材料的力学性能进行设计.

A new method is presented for mechanical property design of carbon fiber braid/epoxy composite. Injection pressure, temperature and time are as input parameters and interlaminar shear strength and bend strength are as output parameters. BP neural network is used to build a model for the relationship between technology parameters and mechanical properties of composites. Based on this model, the relationship between injection pressure and mechanical properties of composites is studied when the injection temperature and time are defined, and the output regular of network is in agreement with experimental regular, which indicate that the model is reliable and can be used to design the mechanical properties of composites.