Unexpected wall-temperature patches and apparent dead cells in an HT3D pyrolysis solid

[Hallucination-lee]

Hello,

I am conducting a simple test involving a solid obstruction that combines three-dimensional heat conduction (HT3D) with solid-phase pyrolysis.

I have attached the minimal input file, HT3D_PYRO.fds, and two Smokeview images showing the wall-temperature distributions on the +X and -X faces.

HT3D_PYRO.txt

FDS versions tested

I reproduced the same behavior using:

FDS 6.11.0
FDS 6.11.1

Therefore, the issue does not appear to be specific to FDS 6.11.0.

The test case contains:

One mesh
One rectangular HT3D obstruction
SOLID_PHASE_ONLY=.TRUE.
A spatially uniform volumetric internal heat source : 200,000
One pyrolysis reaction
A solid char residue
No shrinking or swelling
No BURN_AWAY
No dynamic creation or removal of obstructions

The gas-phase cell size is 1 mm, and the obstruction dimensions are approximately 44 mm × 44 mm × 200 mm.

The char density was calculated as 2325.600 × 0.8631 = 2007.22536 kg/m³
so that the char residue should maintain approximately the original solid volume. The material should therefore not disappear or undergo significant geometric shrinkage.

Expected behavior

Because the internal heat source is uniform throughout the obstruction and the thermal properties are also uniform, I expected the solid temperature to increase relatively uniformly.

Some temperature gradients near the surfaces may be expected, but I did not expect large irregular hot and cold patches on individual faces.

I also expected every wall cell to respond to the internal heating.

Actual behavior

The wall-temperature results show highly irregular local temperature distributions.

In particular:

Large irregular hot and cold patches appear on the surfaces.
The lower region of the +X face remains at approximately 20 °C.
The upper region of the -X face also remains at approximately 20 °C.
These regions do not heat up even though a uniform INTERNAL_HEAT_SOURCE is applied to the entire obstruction.
The affected regions appear similar to inactive or “dead” wall cells.
Other parts of the same obstruction reach temperatures of several hundred degrees Celsius.

The attached figures show the wall-temperature distributions at approximately 9.2 s:

The behavior is visible with:

&BNDF QUANTITY='WALL TEMPERATURE',
CELL_CENTERED=.TRUE. /

I also checked the temperature using a wall-temperature device. Therefore, this does not appear to be only a Smokeview interpolation or visualization issue.

(DEVC : -X Wall temperature at z = 0.024 m)

Comparison with a non-pyrolyzing material

I repeated the same test using a non-reactive material with no pyrolysis reaction.

With the non-pyrolyzing material:

The obstruction heated normally.
The surface-temperature field remained spatially consistent.
No regions remained fixed at 20 °C.
The irregular hot and cold surface patches disappeared.

The abnormal behavior therefore appears only when a pyrolyzing material is combined with HT3D.

Questions

1. Is this behavior a known limitation or bug involving the coupling of HT3D and solid-phase pyrolysis?

2. Why do some wall cells remain at the ambient temperature of 20 °C even though the obstruction contains a uniform volumetric internal heat source?

3. Could these regions be wall cells that are no longer correctly connected to the HT3D internal nodal structure after the pyrolysis calculation is initialized or updated?

4. Is any additional material or surface definition required when using pyrolysis with HT3D, such as specifying the reactive MATL_ID on both the SURF and OBST lines?

5. Is there a recommended workaround that preserves fully coupled HT3D heat conduction and Arrhenius pyrolysis?

Thank you for your help.

[mcgratta]

I'll take a look at this.

[mcgratta]

I ran you case with the latest source code. I do not see what you are seeing:

[mcgratta]

I am running this on RH Linux 9

 Fire Dynamics Simulator

 Current Date     : July 21, 2026  09:38:38
 Revision         : FDS-6.11.1-26-g69a8114-master
 Revision Date    : Mon Jul 20 05:58:02 2026 -0400
 Compiler         : Intel(R) Fortran Compiler for applications running on Intel(R) 64, Version 2025.3.2 Build 20260112
 Compilation Date : Jul 20, 2026 15:20:24


 MPI version: 4.1
 MPI library version: Intel(R) MPI Library 2021.17 for Linux* OS

 Hypre library version: v3.0.0
 Sundials library version: v7.5.0
 Consult FDS Users Guide Chapter, Running FDS, for further instructions.

I will try your case on my Windows desktop.

[mcgratta]

I am running the case on my Windows 11 desktop with the official 6.11.1 release and I do not see the problem that you see.