Correctly Using the FDS Validation Guide

[Tom]

Hi All,

I wish to use FDS to model building occupant safety. I have read the Validation Guide and require some clarification on section "1.2 How to Use this Guide". Please let me know which option I can use. If both options are fine, then that's great. But if neither of them are, please explain to me what I should do.

How I interpret the use of the Validation Guide is as follows:

Option A

1. I have scenario that I wish to model with FDS.
   
2. In this scenario I know my desired peak HRR, the diameter of my fire, what my compartment dimensions are along and my ventilation requirements.
   
3. I calculate what my various non-dimensional quantities are (as specified in Section 3.58) and proceed to Table 3.15 to select the appropriate experimental data set.
   
4. I use the FDS simulation, that was created for that experimental data set, for my intended scenario.
   
5. I then determine the accuracy of the model outputs using Table 16.1

OR

Option B

1. I have a scenario that I have simulated in FDS.
   
2. I calculate what my various non-dimensional quantities are (as specified in Section 3.58) and proceed to Table 3.15 to select the appropriate experimental data set.
3. I then determine the accuracy of the model outputs using Table 16.1

Regards,

Tom

[Kevin]

Take a look at this document

https://www.nrc.gov/docs/ML1231/ML12314A165.pdf

We worked with US NRC on this. Your Option B is closer to what we have done. That is, we take a hypothetical fire scenario, say a fire within a particular room of a power plant, we calculate the non-dimensional quantities, Q*, r/D, etc, and if our hypothetical scenario is within the ranges, we use the accuracy metrics. There are no separate accuracy metrics for each experimental data set within the guide. You need only show that you are using FDS within the parameter space for which it was validated.

[Tom]

Thank you very much for the feedback and the attached document.

[Tom]

Hi

Is it correct to use Table 16.1 from the FDS Validation Guide if I have formed a validation range based on specific validated experiments that match my scenarios? I have this question because I may not be using all of the validated experiments used in the formation of Table 16.1.

[Kevin]

Table 16.1 contains statistics for each output quantity based on all the points. If you want to use a subset of the data, you would need to generate new values for bias and standard deviation.

[Tom]

Okay. Is the intention to use Table 3.15 entirely as your validation range and then Table 16.1 as it is? For example, is the validation range of Fire Froude Number from the minimum number to the maximum number (0.1-10000).

[Kevin]

The Heskestad flame height correlation is for fires that are completely unconfined. So if you were modeling a jet flame in the open, you might invoke the comparison of FDS vs Heskestad flame height correlation. But if you were interested in the heat flux to targets, you would want your fire scenario to be similar to those listed on the scatter plot for Target Heat Flux.

How you use the FDS validation results is up to you. The various standards documents dealing with fire model validation ask that you compare the model with experiments, but they are not much more specific. Use common sense. Look at the experiments that are used for comparison for a particular output quantity and make the case that these experiments are comparable to your hypothetical fire scenario.

[Tom]

I'm interested in using the fire modeling process outlined in NUREG 1934, Fig. 2-1. The applicable scenarios are defined, and validated experiments are selected on a basis that they match the elements of the scenarios. Each validated experiment has its own non-dimensional quantities, and the validation range is formed from the total selection of non-dimensional quantities from all of the selected validated experiments. My own scenarios must then fall within this validation range for my model to be considered validated. 

My question is that I then can't use table 16.1 from the FDS Validation Guide if I do this.

[Tom]

Is it better to create a validation range from the validated experiments that used the same output quantities in your own scenarios? Then you are able to use table 16.1.

[Kevin]

Yes. For each output quantity in Table 16.1, the data is from a small subset of all the validation experiments. You want to make sure that your hypothetical fire scenarios is within the parameter range of that subset of experiments.

[Tom]

Thank you for the help, it is much appreciated.