FATAL ERROR: First edge point not set
 

Hello,

I'm a new CONVERGE user and performing SI engine simulations. I'm getting 'FATAL_ERROR: DEBUG HERE, first edge point not set' as a crash file output. Also, I noticed 4 different files, as listed, in the case folder after the simulation has crashed[LIST=1][*]1. box<rank_number>.dbg[*]2. side_polygons<rank_number>.dbg[*]3. surface_polygons<rank_number>.dbg[*]4. surface_triangles<rank_number>.dbg

All these files contain co-ordinates of problem triangles, I assume.

However, there are no problem triangles flagged by CONVERGE Studio and the final validation does not yield any errors as well. But still the simulation crashed after reading data from inputs.in.

Anyone know what could cause this issue and how to prevent it? Attachment 82938

Attachment 82939

Attachment 82940

Attachment 82941

Hi,


This typically indicates issue with overlapping triangles or intersections. As you are running an engine case, please check (animation tab in STUDIO) if you have intersection as the engine parts are moving.


Thanks,
Nitesh

Hello Nitesh,

Yes, I am running an engine simulation. There are no intersecting or overlapping triangles to be found when I run the Diagnostics on my model.

Also, I'm using CONVERGE Studio v3.0 latest version and can't locate the Animate tab. Could you point out where it could be?

The animation option in the "View Options" doc and you need to activate "Check triangle intersection in each step" to check for intersection at each step (use = 1 CAD).


If you are using windows, then you might have to open STUDIO with -n flag to make animation work (or you will not see parts moving when you press "Play")


Thanks,
Nitesh

See image attached

Thanks for the animation tab, Nitesh!

I submitted the job again after verifying that there are no intersecting triangles during the entire simulation duration of 2 engine cycles. But still the simulation crashed throwing the same fatal error.

I went ahead and created some of the problematic surface triangles and there are quite a many of them. The green, unassigned are the created triangles and seem to coincide with the triangles that make up the intake valve apart from the odd one out which doesn't seem to be connected to anything. (See screenshot)

Attachment 82957

Can you check if the crash occurs near crank angles corresponding to valve open/close timing. If that is the case, check if you you have defined the events correctly.


Best,
Nitesh

The simulation starts with both the valves closed at TDC firing. Simulation crashes after reading data from inputs.in. The log file stops printing after "reading data from inputs.in" and I get the crash file in the case folder. So I can't tell you at which CAD the simulation crashes.

I have exhausted all the checks I can ask you to perform.



Please send your complete case setup and the converge.log file to support@convergecfd.com using your official email address linked to your converge license. Also add a brief summary of what all you have tried so far and the details of your hardware and CONVERGE executable version.



Thanks,
Nitesh

Hi Nitesh,

Thanks for helping out. I figured out the problem. Even though, running the diagnostics wouldn't flag/highlight any problem triangles, I displaced the intake valve up to check where the first instance was - between head and the boundary fence between valve angle and valve bottom. I modified the head triangles such that the "path" of the intake valve is clear of any potential intersecting triangles. This means that the intake port diameter on the head had to be slightly increased.

I ran the simulation for this case and it was successful.

Thanks again for everything, especially for pointing out the animation tab, I needed that.

But I think it's still weird that the solver would crash because of intersecting triangles which don't seem to exist or rather were not flagged by the Diagnostics tool in CONVERGE Studio.

Thanks for the update.


I am glad that you were able to fix the issue.

Hello
I get into trouble using converge CFD
would you please help me?

these 2 errors are:

1-1851 WARNING: The boundary type that appears to not be aligned is PERIODIC, and it is attached to boundary 1, which is moving.

2-1860 WARNING: areas of periodic boundaries 2 and 3 do not match. This may cause mass loss.

Hello Jamshid,


Looks like your boundary conditions are not set correctly and you have mismatch in areas of periodic boundaries. Please create a new post with details of your simulation (images of boundaries 1, 2 and 3) so that we can help you better. The reason I ask you to create a new post is because it might help someone else resolve similar problems.


Thanks,
Nitesh

Hi,rmotwan:

I recently encountered the same problem as you, and I have tried many ways and still can't solve it. Could you please tell me your solution? Thanks again.

Sincerely

Hongjie Zhou