Answers are grounded in this paper and cite their sources.
Surface pore closure during isothermal chemical vapor infiltration (ICVI) limits precursor access to C/SiC preforms, leaving residual porosity and density nonuniformity. Selective SiC skin removal reopens gas pathways but also removes deposited matrix and requires recovery time. A material-conserving dual-porosity finite-volume model maps this trade-off. Calibrated against digitized CT density data, the model matches the density trajectory with 0.597% MAPE. A 122-case CVI-removal-CVI screen at three specimen thicknesses (5, 10 and 20 mm) provides three discrete design points rather than a continuous map. Tested schedules at 5 mm give no robust uniformity benefit: the least adverse case still widens the spread by −0.0027 g cm −3 . At 10 mm the gain is positive and grid-converged (0.0044–0.0054 g cm −3 across two independent refinement axes, ±10%) but stays below the 0.010 g cm −3 significance gate, so it is reported as a direction. At 20 mm, early partial removal at a 2 mm depth reduces density spread by 0.0269 g cm −3 on the base 40 × 20 grid and by 0.0223–0.0269 g cm −3 across both refinement axes (2.2–2.7 times the gate), at a retained-density penalty of 0.0094 g cm −3 ; an absolute-depth scan shows the gain rising monotonically with depth, so no normalized depth has special status. This 20 mm benefit is invariant across a nine-point axial-depletion and boundary-concentration set (0.02,682–0.02,691 g cm −3 ), and the schedule ranking is unchanged for rate multipliers from ×0.5 to ×1.5, though its magnitude scales with rate. The results turn SiC skin removal into a thickness-thresholded process-window rule.
No takes yet. Share an insight, caveat, or question.
Guan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: