As this was the first set of replicate burns, all of the procedural issues were not yet refined. The biggest procedural issue is the opening and closing of vents (i.e. doors and windows). The first test (the one taking longer to achieve flashover) initially had the outside doors closed. As we were watching the fire progress, it was determined that the fire was vitiating. Therefore, the outside door was opened slightly, which allowed the fire to grow again and transition through flashover. We planned accordingly for the second test and allowed the outside door to be left slightly open.

Since the first set of tests, we have implemented a procedure for controlling the ventilation opening by utilizing pieces of wood to hold open the doors at the same angle. A testament for this “now-refined” procedure is that the remaining set of replicate tests reveals almost identical fire growth and time/temperature curves. As I am sure you are aware, the Advanced Fire Pattern Research Project (AFPRP) co-sponsored all test burns and focused its research interests on fire pattern development and reproducibility. Therefore, it should be noted that the fire pattern development in both compartments were not affected by the difference in the time to flashover, which is a critical part of our research.

Hopefully, this answers your question regarding flashover time being variable. If you have the ISFI 2006 paper (part of the ISFI proceedings) you could see the similarities of the remaining 4 test burns. If you do not have a copy of the proceedings, you can obtain a copy through www.nafi.org.

If you have any further questions or concerns please feel free to contact any of us. Thanks for your patience in waiting for this reply...as Patrick had mentioned I was sick, but I am feeling better now.

Best regards,

Gregory E. Gorbett, CFEI, CFPS, MIFireE, IAAI-CFI
Patrick M. Kennedy, CFEI, CFPS, MIFireE
Ron Hopkins, CFEI, CFPS