Query reagrding pressure zones

[Devananda Vijayananda Vivek]

Dear friends,

I am trying to simulate a situation of backdraft in PyroSim wherein a compartment is full of oxygenn depleted methane and when there is an opening in the compartment, oxygen rushes in and cause rapid combustion and rise of pressure. 

I saw the pressure zone example cases in the FDS6 however in pyrosim the pressure zone definition does not seem to be the way it is defined in FDS6. I am getting way too high pressures than reported by experiments and other numerical experiments. I have attched few pictures of the setup as well as code below. I would really appreciate if anybody could guide me. Thank you so much.

[Devananda Vijayananda Vivek]

Dear friends,

only the chamber is having closed walls and the rest is an open boundary.  Do we need to define the pressure zone outside the chamber also? please let me know.

Thanks

[D D]

The rest of the domain (outside of the chamber) is automatically the default zone. Have you defined leakage surfaces with LEAK_PATH as the interface between the two zones?

[Devananda Vijayananda Vivek]

Hello DD,

No, I did not. The opening suddenly ruptures at t=1s. Can this be modeled as a leakage?

[Devananda Vijayananda Vivek]

Dear all,

As seen in the below image, the hole in the right-most wall activates at t = 1 sec and then air gets into the compartment igniting the methane mixture. Even if I define the compartment as a pressure zone when the hole is activated, doesn't the pressure equalize? Do you know if it is required to define the zone at all?

[dr_jfloyd]

It is not required in the current version of FDS to define the zones. In older versions you did have to define the zones. You would need to consult the User's Guide for the specific version you are using to know which applies in your case. As to why your case is not working, there is no way to know that from the images you have attached. We would need to see the .fds file to know exactly what inputs FDS is being given.

Simulating backdraft is a fairly advanced use of FDS. The default combustion model assumptions are mixed is burned and this is not the process that causes re-ignition in a backdraft. You may want to look the the NIST_Backdraft Validation cases in the Github repository. 

[Devananda Vijayananda Vivek]

Dear Dr. Floyd,

I have attached the FDS file with this message. Kindly let me know what you think. Thank you.

[Devananda Vijayananda Vivek]

Dear all, 

Below is the link to the paper on which my study is based. I am trying to recreate the conditions in the paper.

https://doi.org/10.1016/j.firesaf.2020.103270

[Kevin]

If you want to compare your simulation with an experiment, you cannot just impose your initial upper and lower layers with specified gas compositions and temperatures.  You should simulate how the experiment was conducted.

[Devananda Vijayananda Vivek]

hello Kevin,

I was recreating the simulations in the paper which in itself is a study based on an experiment.

[dr_jfloyd]

This paper isn't really attempting to replicate the experiment. It is making a rough approximation of the expeirment, and then studying the effects of changing gravity on the backdraft that results. As the modeler you would need to decide if the assumptions made in the paper for the purpose of the specific work being performed are appropriate for your specific use.

[Devananda Vijayananda Vivek]

Dear Dr. Floyd,

Yes, I am also trying to study backdraft under reduced gravity conditions so I took this paper as a reference. I am using the same assumptions to get the same results at the same time learning the FDS. 

[Devananda Vijayananda Vivek]

Dear Dr. Floyd,

I am getting abrupt pressure rise in the backdraft simulation which should not be the case. I am not able to find out why. Could you please guide me?

As you suggested, I did not define separate pressure zones. Thank you.

[dr_jfloyd]

Why should this not be the case? What is happening from 0 s to 1 s in your simulation? You have a sealed compartment with fuel, oxygen and a hot surface. What happens in a sealed volume if there is heat release?

[Devananda Vijayananda Vivek]

Dear Dr. Flyod,

The burner is only activated when the oxygen mass fraction above the burner goes above 0.2. The slot in the compartment opens at t=1sec. Burner is activated at approx t = 3.5 sec when gravity current reaches the burner. So there shouldn't be any pressure rise before 3.5 sec. Kindly correct me if I am wrong.

[dr_jfloyd]

Do you have heat release in the first second? What is the reaction rate from your A and E?

[Devananda Vijayananda Vivek]

Dear Dr. Flyod,

There is no heat release in the first second and the value of A is 1.5e12 and E is 83680 J/mol. Thank you.

[dr_jfloyd]

That curve is not exactly 0 before 1 s, the line is clearly a little above 0. There is a small amount of HRR.  Load your HRRPUV slice file. You have a small amount of HRR in the upper layer. The reaction rate in your case is d[CH4]/dt=[CH4] * [O2]^0.5 * A * exp(-E/(RT)). With your initial mass fractions, A, E, and T you are getting ~9 kW/m3 in the upper layer at t=0 or about 1.5 kW total.