Fibre-scale Modeling of C/C Processing by Chemical Vapour Infiltration using Xray CMT Images and Random Walkers

Abstract

We present a computational tool for the modeling of Chemical Vapor Infiltration of carbon/carbon composites, which is based on 3D images acquired by X-ray Computerized Micro-Tomography with a very fine resolution, such that the fibers are clearly distinguishable from each other. Preliminary image processing is necessary in order to perform segmentation between solid and void phases. Then, morphological and transport properties are computed in the images. Random walkers are used for the simulation of gas transport in continuum and rarefied regimes. The image modification under chemical deposition is handled by a specific surface discretization technique and a pseudo-VOF method. Results are presented and discussed: the notion of infiltrability is introduced as a design tool for the CVI engineer.

Modelling Infiltration of Fibre Preforms From X-Ray Tomography Data

“…This numerical computation has to be repeated in directions x, y, and z to get the full tensor. Chemical reaction with the walls is represented by the random drawing of sticking events; the probability is fixed and is a function of the molecular velocity and of the linear heterogeneous reaction rate constant [30].…”

A review of applications of CT imaging on fiber reinforced composites

“…In situ XCT during axial impregnation of unidirectional fiber beds had been used to study to produce a high-quality product in a low impregnation time. 42 136 Finally, a low impregnation time had been achieved through a detailed XCT analysis of the fibrous preform by process optimization.…”

“…Figure 26. Densification of C/C composite (a-d) Four steps of the infiltration of a 400 × 400 × 350 voxels block; (e) Rendering of the deposited matrix 136. …”

Simplified marching cubes: An efficient discretization scheme for simulations of deposition/ablation in complex media

“…in an idealized description of stacked woven fabrics [27]. More recently, the acquisition [28] and segmentations [29,30,31] of high-resolution 3D images of carbon/carbon (C/C) fibrous preforms has been made possible. The SMC algorithm has been used for the computation of the surface area and pore diameter distribution, with an excellent validation with respect to experimental data [32].…”

“…We can clearly see that the speedup gained by using the SMC scheme is more interesting when the surface area increases. [20,31].…”