Learning Forum

[VanMouses]

Hi,

I keep getting this error about my problem essentially being too complex for the student software to allow it to be solved. I have a 250 x 60 x 50 tube with 2 supports, 3 forces and a moment. The mesh size is  20mm - literally not even worth running, and apparently the problem is too complex. If i suppress the moment i can use 5mm mesh size and itll run. Any finer and i get the same error message. It is a bit of ridiculous problem really as it renders the software literally useless.

Does anyone know why I might be having this issue?

Cheers

[peteroznewman]

The free student license is limited to 32,000 nodes, which allows some very useful models to be built. Where a solid model might quickly consume the available nodes, converting the solid model to a midsurface shell model will keep the node count down.  Converting to a beam model will be another reduction in node count.

While it is easy to see how many nodes are in the mesh, connection elements used in remote boundary conditions are added to the count when the solve starts, which may explain why your mesh size limit seems to change when you include a moment.

If you reply with some images inserted into your post, or if you want to attach an archive of your model after you post your reply, we can help you get something useful out of the free student license.

[Shrikant Sahare]

"Your product license has numerical problem size limits, you have exceeded these problem size limits and the solver cannot proceed."

I am Getting this Error Again.

[VanMouses]

Thanks Peter,

I will have a look at going ahead with your recommendations.

I have attached images of my boundary conditions and mesh. I will also attach the archive

[peteroznewman]

Try using the model in the attached 2019 R2 archive that has a midsurface representation of the tube.

[VanMouses]

I initially got a geometry error which I fixed by re-associating the forces and constraints, however I then got the same issue with size limits unfortunately, even when using a 20mm mesh! I have just run this with the moment supressed, and can run at 2mm mesh, which is great. However i am not really getting the result I am after without this moment.

It seems that the torque must really eat up the allowed computation of the student licence. Perhaps its something to do with the superposition of stress states? 

Either way, thanks for the quick response and help with the issue! Please let me know if there is anything else I can do to improve this.

Also, may I ask how you managed to change the model to a sheet body? I attempted this myself but couldnt  get it to work. I was trying to use the "thin" feature when editing the model

 

[peteroznewman]

 Right mouse click on Model and Insert Mesh Numbering. Then set the Compress Node Numbers to Yes.  Finally, right mouse click on Mesh Numbering and select Renumber Mesh. That will fix your problem. 

The limit in the Student license is not that the count of total nodes is < 32,000 but that the ID of the highest node number is < 32,000.  The way this model is meshed results in high node numbers, even though the count is low.

If this solves your original problem, please mark this post with Is Solution to help other students find the correct answer.

If you have other, unrelated questions, please start a New Discussion.

[EngChristopher]

Hi. I have the same problem. My mesh has less than 32k nodes or elements (~13k). I have already applied the Mesh Numbering but the error continues to be reported. There is another solution?

[peteroznewman]

If you have a large amount of contact elements, they don't show in the Mesh Statistics, since they are created when the Solver starts, but they are counted in the limit.  Localize the contact surface area as a workaround.

[rejedejo]

There are issues with way the problem size is calculated. I was able to solve a problem with 80,236 nodes and 43,103 elements.

 

Now I'm getting this error:

"Your product license has numerical problem size limits, you have exceeded these problem size limits and the solver cannot proceed."

with a similar problem that has 11,187 nodes and 5,340 elements

This is a totally new instance of Ansys WB, subtly different geometry, and the only system in this *.wbpj file.

These two cases demonstrate there are issues with the node/element count.

[rejedejo]

showing another couple images demonstrating clearly the overnodes

[peteroznewman]

I assume you don't care if the solver runs when there are more nodes or elements than the limit.

I assume you do care when the solver doesn't run when there are fewer nodes or elements than the limit.

I have found that by inserting a Mesh Numbering item into the outline, you can force the mesh that generated an error be be renumbered and allow a mesh with a count below the limit to run.

[levent123456]

Hi i have the same problem i'm using 2020 version.I cant solve the static structural can you please help me peter

[peteroznewman]

Hi levent123456,

Please open a New Discussion for you problem. In that post, show the Error in the Message window or in the Solver Output under the Solution Information folder.

[levent123456]

[Rob]

You have too many cells for the Mech solver: Student is capped at 32k cells. There are several threads on this so please review those. 

[Fikri]

Hi i have the same problem i'm using 2020 version.I cant solve the static structural can you please help me petern

[levent123456]

hello everyone i am getting the same warning on ANSY can someone please help me about how to decrease the mesh nodesor elemnts size.nn

[Rob]

That's not the same warning. What are you trying to do?n