Error: Zone

[mark.k...@gmail.com]

The FDS model (a test to make sure it everything works) worked initially but I had to make changes to fix leaks and is now saying Error: Zone.

My mesh boundary (3 meshes) is open and I do not have pressure zones.

The file size is big 42MB and I cannot upload it, please see link below to fds file.

https://1drv.ms/u/s!AnMDQn11O3jym8saOnqIq4-hHxBuYg?e=35ViBh

I have also attached the error code I'm getting.

I'm struggling to work out how to solve this and where to look. I did read about filling the voids but I do not believe I can create a slice through the model at the mesh interface in Pyrosim to see the voids and fill them.

I use Pyrosim to build the model which was imported from dwg.

Can you please help. Thanks Mark

[Kevin McGrattan]

FDS now looks for pressure zones, and it apparently has found many (I have to fix the ** bit). My guess is that PyroSim has created hollow boxes that are interpreted as pressure zones. My guess is also that as a hollow box goes from one mesh to another, there is some kind of mismatch.

[Kevin McGrattan]

There is a SLCF QUANTITY called 'PRESSURE ZONE' which will indicate the number of each pressure zone, and identify where they are.

[Bryan Klein]

I think there has been some recent confusion when using PyroSim and the newish auto-Zone finding method in FDS.

People often import CAD files where the geometry is represented by polygonal surfaces with nothing 'inside' of them.

In the past, 'Zones' were something that a user had to create in PyroSim, most of the time the model would have open boundaries and openings from within the building envelope to this 'ZONE0' (the infinite ambient domain) outside of the mesh.

Now, FDS is finding 'Zones' inside of walls, window frames, etc. that are modeled as hollow surface geometry.  When it does this, there are many times that this geometry might pass from one mesh resolution to another.  When this happens, if the model allows for 'thin obstructions' then smaller 'hollow' geometry might snap to be a thin OBST in one mesh and resolved as a 'hollow' zone in another.  Then all the zone errors pop-up in FDS, where the PyroSim model only has a single 'ZONE' defined.

The current advice/solution we have been giving people is to be aware of this new feature in FDS and to 'Thicken' all geometry in the model so that these thin obstructions do not create openings into these new auto-generated zones.

In PyroSim, you can thicken the geometry in the 'Obstruction Properties' UI ( https://support.thunderheadeng.com/docs/pyrosim/2021-3/user-manual/#_creating_obstructions ), or as a Global setting in the 'Angled Geometry' tab of the Simulation Parameters UI. Note, it is best to also check the 'Rasterize axis-aligned' option too, so that all obstructions, including those that are aligned with an X,Y,Z axis will be thickened.

-Bryan

[Kevin McGrattan]

By default, FDS assumes that solid walls conduct heat through to the other side if the obstruction is zero or one cell thick. Thus, the interior of a hollow duct can heat up, and if the duct is sealed, the pressure inside will rise. To handle this properly, FDS must designate this as a separate pressure zone with a different background pressure than the exterior of the duct. That is why we now automatically detect these sealed interior regions. Before we did this, we received many complaints about numerical instabilities.

[Bryan Klein]

That makes sense, and I think some of our newer code to help with manifold solids can be used to generate 'solid' obstructions that can be capped correctly at the ends if a mesh changes the alignment of the same surface as the OBST conforms to the mesh.

There is a gap between what comes out of Revit and what FDS needs and we intend to fill it. Pun intended. 😊

-Bryan

On Wed, Nov 24, 2021, 16:24 Kevin McGrattan <mcgr...@gmail.com> wrote:

By default, FDS assumes that solid walls conduct heat through to the other side if the obstruction is zero or one cell thick. Thus, the interior of a hollow duct can heat up, and if the duct is sealed, the pressure inside will rise. To handle this properly, FDS must designate this as a separate pressure zone with a different background pressure than the exterior of the duct. That is why we now automatically detect these sealed interior regions. Before we did this, we received many complaints about numerical instabilities.

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/rFgrmdcvsZA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAAJimDH8UzOkSkNoWhPZM4_tN%3DF7MGhRiLNf-WgbjBPSLJz1KQ%40mail.gmail.com.

[mark.k...@gmail.com]

I thickened all obstructions and made the changes in Simulation Parameters you suggested and it worked. Thank you.

My question then, why when I have a single mesh the simulation computes and I do not get the zone error?

Mark

[dr_jfloyd]

In Kevin's first reply to you:  " My guess is that PyroSim has created hollow boxes that are interpreted as pressure zones. My guess is also that as a hollow box goes from one mesh to another, there is some kind of mismatch."

If Kevin's guess is correct, then with a single mesh there is no possibility for a mesh-to-mesh mismatch.

[James]

Hi all,

I've run into a similar a problem with the 6.7.7 release (upgrading from 6.7.4). We import large CAD files and do not use Zone models, the checks/errors generated from 6.7.7 mean the simulations do not run. The default .fds without rasterizing axis-aligned objects is circa 30 MB, this (when rasterizing axis-aligned) becomes several gigs and we cannot import due to the size/time.

Is there a way to disable this zone checking if no zones are specified in the model or an alternative solution for larger geometries with a lot of cells?

Thanks for any information,

James

[Kevin]

I recently added a way to suppress checking for pressure zones for when one is doing a wildfire simulation. However, this flag is not directly input by the user. I can easily make it so, although this is somewhat dangerous to do. Pressure ZONEs are regions in the domain where FDS calculates a separate background pressure. If it doesn't do this, then the flow will penetrate the walls. This is why we automated the process of locating pressure zones -- they were frequently the cause of numerical instabilities. However I will add the ability to suppress ZONE finding.

Can you compile the code yourself?

[Michael]

Hello Kevin,

is there already a way to suppress ZONE finding without compiling the code? 

I have a very complex imported model and I keep having the problem with hollow zones from one mesh to another.

Best regards

Michael

[Kevin McGrattan]

If they are not important, fill in the hollow zones. FDS must account for hollow regions when it solves the pressure equation. If it did not, your calculation would be at greater risk of numerical instability or at the very least, inaccuracy.

[James]

Hi Kevin,

Yes I can/do compile the code if there is an easy way to comment this function out without breaking the rest of the code?

Many thanks,

James

[Kevin McGrattan]

I just noticed that I added a flag to the latest release that suppresses the pressure zone search. This was done to speed up outdoor fire simulations. Try adding NO_PRESSURE_ZONES=T to the MISC line and let me know what happens.