About setting the 'autoignition temperature' in the reaction section of pyrosim

[jianyiping730-cyber]

I am conducting open-space n-heptane simulations using different mesh sizes. I found that when the auto-ignition temperature (AIT) in the reaction settings is set to -273°C (the default value), the mass loss rate and HRR vary considerably with mesh size, and the model runs relatively fast. When the auto-ignition temperature is set to 215°C (the default auto-ignition temperature of n-heptane reported in the SFPE Handbook), the differences in mass loss rate and HRR among the mesh sizes become smaller, but the simulation takes significantly longer. Why does this happen?Please check the attached document for the detailed comparison chart.
question.docx

question.docx

[mcgratta]

The one thing worse than posting one MS Word document into the Discussion is posting two identical Word documents. Do not assume that everyone has Word installed.

As far as your case study, you need to first look at the relative size of the time steps in the "fast" and "slow" simulations. For two simulations using the same mesh and end time, the time step determines the overall amount of time the case takes to run. The time step is inversely proportional to the maximum flow velocity. Look for where this occurs in both cases and ask yourself why that might be.

[jianyiping730-cyber]

Thank you for the suggestion. I compared the time steps in the two cases.

In the AIT=215 °C case, the time step decreases rapidly during the initial ignition stage, from about 0.039 s to roughly 0.002–0.005 s within the first few seconds, whereas the default-AIT case maintains a much larger time step initially.

I also noticed that the 215 °C case uses a 5000 °C ignitor for the first 10 s. My current interpretation is that this very hot ignitor may generate a high local gas velocity near the ignition region, which reduces the time step through the CFL constraint. After the fire is established, the time steps in the two cases become much more similar.

In my long-tunnel simulations, the HRR and mass loss rate are also very similar for AIT=215 °C and the default AIT, while the default-AIT cases run noticeably faster.

Since my subsequent work is mainly concerned with the established fire behavior (HRR, mass loss rate, temperature and smoke development), rather than the ignition process itself, would it be acceptable to leave AUTO_IGNITION_TEMPERATURE at its default value? Or would you recommend keeping the literature value of 215 °C for n-heptane?

Thank you for your help.

[mcgratta]

Most users only set an AIT when they are interested in flame extinction. By default, FDS assumes that fuel and oxygen burn on contact. You can think of it as piloted ignition even though there is no actual pilot. You are correct about the 5000 C causing trouble. Keep in mind that an arc fault in an electrical wire might generate such high temperatures, or even higher, but that would be over a tiny volume for a very short time. In your case, the entire cell volume achieves this temperature and maintains it for seconds.

If you are modeling a tunnel fire, there is no need for AIT. Just assume that the fire is not going to completely extinguish. If the oxygen in a region drops too low to sustain combustion, the fuel gases will be transported somewhere else and then burn on contract with oxygen. This is not what would really happen, so you have to decide if you can live with this assumption.