VAPOR-GROWN CARBON-FIBERS - MODELING OF FILAMENT LENGTH DISTRIBUTIONS

A mechanistic model has been used to predict filament length distributions for vapor grown carbon fibers during the lengthening stage. Optimal values for the two parameters, alpha and Lt, in this model are obtained by curve-fitting one set of experimental results, giving excellent agreement. The model can easily be used to predict length distributions for cases where alpha and L vary with time in any prescribed manner. Results are obtained for the case where alpha undergoes a step increase after some time, and it is found that the filament length distributions are quite sensitive to the history, alpha(t). In fact, situations can easily be found for which the mean lengths are identical, but the distributions differ significantly.