CFAST: spatial segmentation of tunnel, ULT, HRR
Dominik Dvořák
Hello,
I am attempting to use the CFAST zone model for tunnel fire modelling and have encountered several issues that I can't figure out. My inquiry concerns explicitly tunnel fire modelling and recommendations regarding tunnel spatial segmentation.
I am attempting to simulate a road tunnel with a length of 500 m and a rectangular cross-section of 10 m width and 5 m height. A fire, defined by the Heat Release Rate (HRR) curve with a peak HRR of approximately 128 MW, is placed in the middle of the tunnel. A plate thermocouple was positioned under the ceiling in the middle of the tunnel (aligned with the fire source axis). Mechanical ventilation was not considered. Due to the limitations of zone models concerning the length-to-width ratio, I divided the tunnel into smaller segments interconnected by a virtual opening equivalent to the full tunnel cross-section. I considered several spatial segmentation options at 5, 10, 25, 50, and 100-metre intervals. Subsequently, I compared temperature profiles (ULT, Target) and HRR profiles (see the graphs below).
My first question relates to spatial segmentation: Is there a general recommendation for how to appropriately segment the longitudinal space? Could you provide an explanation, if possible? In a scientific article I came across, the following conclusion was reached:
“The parametric analysis demonstrated that a compartment length twice as much as the compartment width (20 m compared to 10 m) guarantees the best fitting of the experimental results, in terms of maximum temperature in each compartment.”
(BAMONTE, P., R. FELICETTI, P. G. GAMBAROVA and A. NAFARIEH. On the Fire Scenario in Road Tunnels: A Comparison between Zone and Field Models. 2011, 82, 764–769. ISSN 1662-7482. In: doi:10.4028/www.scientific.net/AMM.82.764).
My second question pertains to the temperatures of hot gases: Why do higher temperatures of hot gases occur around the "Target" thermocouple than in the upper smoke layer (ULT)? I observed this in all spatial segmentation scenarios. While seeking an answer, I came across this explanation in a support group:
"When calculating the ceiling jet temperatures, CFAST uses modified empirical correlations. This method of calculation affects the temperatures near the ceiling, which subsequently come out higher."
Nevertheless, I observed this phenomenon of higher temperatures around the "Target" thermocouple compared to ULT even when attempting to simulate a fire in a standard enclosed room (5 m width, 10 m length, and 3 m height). Is there a way to explain this phenomenon?
My final question concerns the HRR profile: In the resulting HRR profile graphs, it can be observed that in the 100-metre and 50-metre segmentation scenario, the resulting HRR profile almost matches the input HRR curve (given a bigger timestep). For other segmentation scenarios, the HRR profile fluctuates more as the segments become smaller, sometimes reaching values of 0 MW. Can this be explained?
I would be immensely grateful for any explanation regarding my inquiries. Thank you very much in advance.
Best regards,
Dominik
Reneke, Paul A. (Fed)
Hey,
It is generally not recommended to use the virtual rooms trick. There is a lot of physics going on that CFAST doesn’t consider because of simplicity. I understand the difficulties and that FDS takes a lot more time and computational resources to use but that is currently the best option. As such I can give you no guidance but one thought you might look at is doing a kind of series of case where you increase the rooms between targets you have interest in and the fire. It might tell you something it might not. I have never done it.
As to the targets. They are solved as explicit differential equations separate from the solver that is calculating the layer temperatures, layer volume and pressure difference. The targets take the layer temperatures but they also consider radiative flux to the target and as such can be hotter or cooler than the upper layer temperature. This is a studied problem with thermal couples. Usually it isn’t that big of a problem but it does exist.
I hope this helps
Paul
From: cf...@googlegroups.com <cf...@googlegroups.com>
On Behalf Of Dominik Dvorák
Sent: Tuesday, August 22, 2023 1:48 PM
To: CFAST <cf...@googlegroups.com>
Subject: [CFAST:2273] CFAST: spatial segmentation of tunnel, ULT, HRR
Hello,
I am attempting to use the CFAST zone model for tunnel fire modelling and have encountered several issues that I can't figure out. My inquiry concerns explicitly tunnel fire modelling and recommendations regarding tunnel spatial segmentation.
I am attempting to simulate a road tunnel with a length of 500 m and a rectangular cross-section of 10 m width and 5 m height. A fire, defined by the Heat Release Rate (HRR) curve with a peak HRR of approximately 128 MW, is placed in the middle of the tunnel. A plate thermocouple was positioned under the ceiling in the middle of the tunnel (aligned with the fire source axis). Mechanical ventilation was not considered. Due to the limitations of zone models concerning the length-to-width ratio, I divided the tunnel into smaller segments interconnected by a virtual opening equivalent to the full tunnel cross-section. I considered several spatial segmentation options at 5, 10, 25, 50, and 100-metre intervals. Subsequently, I compared temperature profiles (ULT, Target) and HRR profiles (see the graphs below).
My first question relates to spatial segmentation: Is there a general recommendation for how to appropriately segment the longitudinal space? Could you provide an explanation, if possible? In a scientific article I came across, the following conclusion was reached:
“The parametric analysis demonstrated that a compartment length twice as much as the compartment width (20 m compared to 10 m) guarantees the best fitting of the experimental results, in terms of maximum temperature in each compartment.”
(BAMONTE, P., R. FELICETTI, P. G. GAMBAROVA and A. NAFARIEH. On the Fire Scenario in Road Tunnels: A Comparison between Zone and Field Models. 2011, 82, 764–769. ISSN 1662-7482. In: doi:10.4028/www.scientific.net/AMM.82.764).
My second question pertains to the temperatures of hot gases: Why do higher temperatures of hot gases occur around the "Target" thermocouple than in the upper smoke layer (ULT)? I observed this in all spatial segmentation scenarios. While seeking an answer, I came across this explanation in a support group:
"When calculating the ceiling jet temperatures, CFAST uses modified empirical correlations. This method of calculation affects the temperatures near the ceiling, which subsequently come out higher."
Nevertheless, I observed this phenomenon of higher temperatures around the "Target" thermocouple compared to ULT even when attempting to simulate a fire in a standard enclosed room (5 m width, 10 m length, and 3 m height). Is there a way to explain this phenomenon?
My final question concerns the HRR profile: In the resulting HRR profile graphs, it can be observed that in the 100-metre and 50-metre segmentation scenario, the resulting HRR profile almost matches the input HRR curve (given a bigger timestep). For other segmentation scenarios, the HRR profile fluctuates more as the segments become smaller, sometimes reaching values of 0 MW. Can this be explained?
I would be immensely grateful for any explanation regarding my inquiries. Thank you very much in advance.
Best regards,
Dominik
--
You received this message because you are subscribed to the Google Groups "CFAST" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
cfast+un...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/cfast/e01c0b06-0c5e-4fd0-b487-c47e965251c2n%40googlegroups.com.