An analysis of autoignition temperature round robin test data

Autoignition temperature (AIT) round robin tests were conducted by seven laboratories on eight polymeric materials. A statistical analysis was conducted that focused on investigating data variability both within and between test methods. With some exceptions, the methods ranked the materials consistently and as expected from their chemical makeup. There was no consistent difference between AIT values determined at 6.9 and 10.3 MPa. BAM and ASTM G72 consistently provided lower, more conservative AITs at both 6.9 and 10.3 MPa. Within an estimated accuracy obtained from all methods from this round robin, data from the following methods appeared to be more accurate than the rest: NF E 29763/88 and Parr Bomb at 6.9 MPa; automated ASTM G72 at 10.3 MPa. The following methods provided lower within-method variability than the average for all methods: BS 3N 100 and ASTM G72 at 6.9 and 10.3 MPa; Parr Bomb at 6.9 MPa. At 6.9 MPa the BAM method provided data with greater within-method variability than the average for all methods. For most materials, the automated ASTM G72 method at 10.3 MPa provided data with greater within-method variability than the average for ail methods. Two data points for this method, Zytel(R)(6) and Teflon(R)(6) FEP, exceeded critical k. At 6.9 MPa, the largest within-method variability resulted from NF E 29763/88 on Zytel, the only data that exceeded critical k. Two extreme results were caused by significant effects of method-specific test parameters: BAM on Zytel at 6.9 MPa and BS 3N 100 on Viton(R)(6) at 10.3 MPa. Statistically these results had a probability of less than 5 percent of occurring by chance.