Font Size: a A A

Research On Scale-span Numerical Simulation Techniques For CFRP Drilling

Posted on:2021-01-29Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:1522306800476804Subject:Aviation Aerospace Manufacturing Engineering
Abstract/Summary:PDF Full Text Request
Because the homogenization modeling method is adopting for the simulation of drilling carbon fiber reinforced plastic/polymer(CFRP),it is difficult to balance the modeling sacle and computational efficiency of finite element(FE)model by using the conventional FE numerical analysis method in ABAQUS/Explicit environment.The real damage behavior of drilling CFRP laminates is difficult to simulate because of the composition of CFRP has multi-scale characteristics.For instance,burrs,interlamination delamination phenomenon.etc.So that it cann’t provide a good technical support for the optimization research of drilling process parameters and bit structure design,etc.,and it is not reliable enough to put forward a reasonable optimization of the hole making process plan to satisfy the high precision assembly requirements of structural parts.To cure the above problems,this paper follows the research idea of” elastic property characterization,damage constitutive modeling,numerical simulation,experimental validation and engineering application” to research the span-scale numerical simulation techniques for CFRP drilling.Based on the composition form of CFRP and the material properties of fiber and matrix,the deeply research on the elastic properties characterization,dynamic progressive damage model,the evaluation of the drilling numerical simulation of CFRP and the application of scale-span CFRP model are carried out via the combination of the scale-span FE method,intelligent algorithm.The main contents and results of research are as follows:(1)An elastic properties characterization method for CFRP considering the fiber random distribution is proposed based on the scale-span modeling method.Meanwhile,the macroscopic elastic properties of UD-CFRP are characterized and the change relation of the ealastic properties of multidirectional CFRP(MD-CFRP)with the change of the angle and the different stacking sequence of UD-CFRP are predicted.A structural unidirectional representative volume element(UD-RVE)model which considering the fiber random distribution is established according to the geometrical structure of unidirectional CFRP(UD-CFRP)at the microscopic level,and the corresponding periodic boundary conditions is imposed.The elastic properties of UD-CFRP are precisely characterized through the optimum global mesh size of UD-RVE model based on asymptotic homogenization theory.The structural multidirectional RVE model is simplified to characterize the elastic properties of multidirectional CFRP with the change of the angle and the different stacking sequence of UD-CFRP and their change relation are predicted as well.A variety of elastic properties test experiments are carried out to vertify the elastic property characterization results of different types of CFRP via the elastic property test standard.(2)A dynamic progressive damage evolution model of CFRP is established based on micro mechanics of failure theory.Meanwhile,the corresponding VUMAT user-defined subroutines are developed to simulate the intra-laminar and inter-laminar damage failure behavior of CFRP.According to the simulation analysis of UD-RVE model,a method to calculating the microscopic stress of UD-CFRP and achieving the stress amplification factors are proposed,and the numerical analysis of CFRP laminates under dynamic loading condition is converted from the macroscopic scale to the microscopic scale via SAFs.Based on the micro mechanics of failure theory,a dynamic progressive damage evolution model of CFRP are established by introducing the damage state variables at the microscopic level,and the corresponding VUMAT code is developed for simulating the intra-laminar damage behavor of CFRP.An inter-laminar failure model of cohesives element based on hybrid failure criteria is established to simlulate the inter-laminar damage failure behavior of CFRP under dynamic loading.(3)A three-dimensional scale-span model of drilling CFRP laminate with a tapered drill-reamer(TDR)is established.Meanwhile,the dynamic mechanical response and progressive damage behavior of CFRP are carried out in drilling,and the analysis results are evaluated from the angle of accuracy and efficiency by comparing with the FE model which adopting the conventional macroscopic constitutive model.Based on the established dynamic progressive damage constitutive model of CFRP,a three-dimensional scale-span FE model of drilling CFRP laminate with a TDR is established by the FE analysis software of ABAQUS/Explicit.The drilling behavior of CFRP is simulated when different process parameters is adopted.The correspongding drilling experiment testing and damage observation system are established to verify the accuracy of the simulation results of the scale-span FE model under the same drilling conditions,and the precision of the analysis results is evaluated by using the precision hole-making quality evaluation index.A comparative analysis is carried out on the prediction of thrust force and torque and material damage behavior when the macroscopic constitutive damage model of Hashin,Chang-Chang and Tsai-Wu are adopted in the drilling FE model of CFRP.The computation efficiency and simulation accuracy of the scale-span FE model are evaluated from the aspects of calculation time,simulation results and experimental test accuracy.(4)A method of predicting thrust force is proposed based on the combination of scale-span FE model and artifitical neural network model in drilling,and the thrust force of drilling CFRP laminate with a TDR is predicted high-efficiency and accurately.The mesh size of the initial scale-span FE model is optimized to obtain optimum computing time on the premise of ensuring the thrust force simulation accuracy.Based on the pseudo-random sequence algorithm,an order-driven FE calculation method is proposed to obtain the training sample set of the drilling scale-span FE model with different material properties of CFRP and process parameters.A feature selection model based on the samples of the material properties of CFRP,process parameters and thrust force is established to obtain the weight ratio of the material properties of CFRP,process parameters for the influence of thrust force.A multi-layer revised back propagation artificial neural network model is established by using Python software.The thrust force of the new types of samples are predicted based on the training completion model,and the prediction accuracy is evaluated via experiments with the equal experimental condition.All the analysis results of the scale-span FE models and experimental results showed that the global mesh size of the UD-RVE model considering the fiber random distribution approximately0.7μm is the most optimal mesh,which has high accuracy(the maximum deviation is only 4.235%,the minimum deviation is 0.6558%)and high simulation efficiency of characterizing the elastic properties of CFRP.Meanwhile,the scale-span drilling FE model of CFRP can simulate the damage phenomenon including the tearing damage at the entrance of the hole,the pit and delamination damage of the hole-wall,the burr,tearing and delamination damage at the exit of hole.Compared with the drilling FE model which adopting the conventional macroscopic damage constitutive model,it has higher accuracy in the terms of predicting the thrust force,torque and delamination coefficient in exit of the hole.Compared with the experimental results,the maximum deviations are 3.37%,7.69% and 4.28%,respectively.In addition,the established scale-span drilling FE model with different material properties of CFRP and process can be applicable for the training sample for the prediction of thrust force in the artificial neural network model.The thrust force of drilling CFRP can be predicted quickly and accurately via the well-trained neural network model,and the maximum absolute deviation is only 4.56% for the prediction of new sample set.
Keywords/Search Tags:CFRP, Elastic property characterization, Micro mechanics of failure theory, Scale-span numerical simulation, Drilling, Thust force
PDF Full Text Request
Related items