Effect of sizing agent on carbon fiber density measurements using a floatation method

The density of T300 carbon fiber, as well as the amount of sizing and micro-structure were measured, before and after removing the sizing. The results showed that there was apparent stratification when measuring the density of the carbon fiber by the sink/float method, without removing the sizing. The upper sample layer was seriously flocculated with an excessive sizing amount of 2.17%, while the bottom-layer also flocculated. Two methods including acetone extraction and pyrolysis, were conducted to removing the sizing. It was shown that the measured amounts of sizing for the two methods were 1.20% and 1.53%, respectively. For the sample after immersion in acetone, there was still stratification in the sink/float test, with the upper sample layer flocculated because of the remaining sizing; while samples at the bottom dispersed relatively well. However, for the thermally treated samples the sizing was adequately removed, which resulted in the uniform density and dispersion during the test. This indicates that the non-uniform distribution of sizing agent on the surface of the CFs is the main reason for stratification during the density measurement. Therefore, the sizing must be fully removed in order to obtain an accurate density for the carbon fibers. Moreover, it can be seen that pyrolysis is an effective way to remove the CF sizing agent.