Did you find the heat flux column that matches the plot in the CFAST Validation Guide?

Kevin,
I did find it thank you for sending that. So, I was looking at the wrong data. The curve that peaks at about 3 kW/m2 is from the column Target Total Gauge Flux for Target 1 and is only available in the output validation mode. It appears to line exactly up with the Long Wall Heat Flux data curve on page 262 of the V&V Guide.

I was looking at the Target Incident Flux which is defined in the CFAST User’s Guide (pg. 56) as the "total heat flux striking the front surface of the current target." The User’s Guide defines the Target Total Gauge Flux as the "total net heat flux to the front surface of the current target assuming the target radiative losses are at ambient temperature (validation output only)." My first inclination would be to assume that incident flux (total flux in) would always bound total gauge flux (total flux in less radiant flux out). But that’s obviously incorrect.

Regardless of why it’s incorrect, it leads me to ask how we should be using the validation output data. Until now, I considered it inappropriate to use that data for fire modeling purposes. However, it’s clearly being used in the V&V Guide to make a comparison against empirical test data. If I’m trying to model postulated real world fire scenarios, this data indicates I should be using the validation output results in my analysis. For this particular case, if I didn’t use the validation output data, I would be under-reporting flux to the wall.

Can you provide some additional information on the validation output data and whether or not it can/should be used for analysis purposes as appropriate output data for realistic fire modeling?

Thanks
Ray

I should note that all the validation results were run with the -v option in cfast which sets the net heat flux calculations to be consistent with the room temperature water-cooled heat flux gauges used in the experiments

Let me try to re-state in different words.
In the validation mode, heat flux to a target is calculated, based on a wall surface temperature assumed to be ambient. This results in an increased flux value because heat flux is dependent on the temperature difference. When not in the validation mode, heat flux to a target uses a calculated wall surface temperature (higher than ambient), so the heat flux value is lower, and more physically accurate.

So, that means I should NOT use validation mode when analyzing postulated real-world fire scenarios. Is that correct?

If you want to know the flux of energy absorbed at the solid surface, you do not want the "cold gauge" result. However, I am not sure if CFAST adds extra output quantities in "validation" mode, or if it redefines existing quantities.

Yes, but whether CFAST adds extra output quantities in "validation" mode, redefines existing quantities, or modifies empirical data for comparison, it makes no difference. The point is that the validation mode results are (or at least might be) modified and therefore less accurate at modeling actual flux values than the output quantities provided in the non-validation mode.

Flux levels delivered to a target are often an important parameter in determining whether a postulated fire might cause excessive damage as in say for example, exceeding a damage threshold of 6 kW/m^2 for thermoplastic electrical cables (NRC Inspection Manual, Chapter 0609, Appendix F Attachment 6, "Guidance for the Identification of Targets and their Ignition and Damage Criteria," [https://www.nrc.gov/docs/ML1808/ML18087A410.pdf]). It's very important for users to understand what CFAST output parameters should or should not be used in performing thermal analysis.

It sounds as if you're agreeing that CFAST users should NOT employ spreadsheet output results identified in Table C.1 of the CFAST User's Guide as "validation output only" without at least addressing the modifications that may have been made to those outputs. Is that right?

By default, CFAST outputs the "net" and "incident" heat flux. For regulatory purposes, you probably want to use the incident because that would be consistent with a simple "point source" model in which the temperature of the target is not taken into account and assumed to be ambient.

Yes, but whether CFAST adds extra output quantities in "validation" mode, redefines existing quantities, or modifies empirical data for comparison, it makes no difference. The point is that the validation mode results are (or at least might be) modified and therefore less accurate at modeling actual flux values than the output quantities provided in the non-validation mode.

Flux levels delivered to a target are often an important parameter in determining whether a postulated fire might cause excessive damage as in say for example, exceeding a damage threshold of 6 kW/m^2 for thermoplastic electrical cables (NRC Inspection Manual, Chapter 0609, Appendix F Attachment 6, "Guidance for the Identification of Targets and their Ignition and Damage Criteria," [https://www.nrc.gov/docs/ML1808/ML18087A410.pdf]). It's very important for users to understand what CFAST output parameters should or should not be used in performing thermal analysis.

It sounds as if you're agreeing that CFAST users should NOT employ spreadsheet output results identified in Table C.1 of the CFAST User's Guide as "validation output only" without at least addressing the modifications that may have been made to those outputs. Is that right?