Online CFD Course - Recent Threads

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Colinda — Fri, 19 Apr 2024 08:32:21 GMT

A defect was identified in Fine Open 11.2 on Windows (not on Linux):

Select the EARSM turbulence model
Go to the Boundary Conditions tab
An application error appears with the error message: "Tix Error: extra switch pattern with no body"

A patch is now available.

An error about residual of Mass is Not A Number: the solver has been stopped application

Huibo — Wed, 13 Mar 2024 15:01:39 GMT

Dear,

First I use the K-w flow model stimulate my model, it could work. But I would like to test another model (S-BSL-EARSM), it didn't work. It was showing :residual of Mass is Not A Number: the solver has been stopped
application.

Could you please give me some suggestion about this issue? I have no idea about this.

Looking forward to your reply.

Best regards,

Huibo

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Huibo — Thu, 21 Mar 2024 14:42:45 GMT

Dear Colinda,

Thank you so much for your support.

Best regards,

Huibo

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Colinda — Thu, 21 Mar 2024 14:39:42 GMT

Dear Huibo,

This message indicates a defect that prevents you from entering the correct inlet boundary conditions for the EARSM turbulence model. It can be reproduced under Windows and has been submitted to the development team. We'll keep you informed of a possible workaround.

Best regards,
Colinda

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Huibo — Tue, 19 Mar 2024 09:07:58 GMT

Dear Colinda,

Could you please help me solve this problem at first? When I use the K-W flow model, I could set up the boundary condition. But when I use (S-BSL-EARSM), I couldn't set up the BC. It said : extra switch pattern with no body, this may be due to a comment incorrectly placed outside of a switch body - see the "switch" documentation. Because this issue, I thought I get the wrong result before.

I am using steady method. The maximum ratio of Expansion ratio is 2.6. The minimum is 1.And the maximum ratio of aspect ratio is 18.87, the minimum is 1.03. I used the pressure switches.

Because of the error, I can't run it now. I think I got the wrong result of EARSM flow model. So I can't share the detail of the result. I am sorry for this.

Best regards,

Huibo

RE: CFD online course 2023-2024

Colinda — Tue, 19 Mar 2024 08:16:38 GMT

The CFD Online Course is available in support.cadence.com. This does indeed require access to the software: I recommend you to follow up with Numeca Ingenieurbüro through the email address they provided to you, with academic-cfd@cadence.com in Cc.

CFD online course 2023-2024

nhatminh — Wed, 31 Jan 2024 10:19:06 GMT

Hello,

my name is Minh Tran, and I have registered for the CFD course for the 2023-2024 session. I have received the confirmation email, but as of now, I haven't received instructions for the second step, which involves installing the CFD software.

I wanted to inquire whether the new course has commenced, and if not, could you please provide an estimate of when the instructions for the next steps will be sent?

I appreciate your assistance in this matter and look forward to your response.

Thank you and best regards,

Minh Tran

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Colinda — Tue, 19 Mar 2024 08:04:56 GMT

Hi Huibo,

Could you please tell a bit more about the application; which features are you using?
For example; are you using the Non-linear harmonics method? Do you use pressure switches?

How do you decide it isn't well converged? Which criteria you use? Can you share some plots of the residuals here?
Did you have a detailed look at all the residuals: the residuals for the turbulence quantities are well converging?

Did you have a look at the pressure, velocity and temperature fields at cuts throughout the domain? Look for zones with unphysical results.

How about the other mesh quality metrics: aspect ratios and expansion ratios?

Best regards,
Colinda

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Huibo — Fri, 15 Mar 2024 10:02:17 GMT

Dear Colinda,

I used FineOpen112 to run my project. The Nb of cells is 1187334. Yesterday I have run it, it didn't have the same error. but the simulation isn't well converged. But the K-W is well converged. Today I change it again. When I set up the boundary condition, It said : extra switch pattern with no body, this may be due to a comment incorrectly placed outside of a switch body - see the "switch" documentation. And the Maximum ratio of Cell Non-Orthogonality is47.84. I think it is ok.

Looking forward to your reply!

Best regards,

Huibo

RE: An error about residual of Mass is Not A Number: the solver has been stopped application

Colinda — Thu, 14 Mar 2024 11:01:43 GMT

Dear Huibo,

With which of the course exercises you experience this? Or are you running a project of your own? Would be great to know which Fidelity version, application, and solver are used.

I would recommend you to save the solution at the iteration before the solver stops. If it occurs during coarse grid initialization, make sure to activate "Save solution during FMG initialization" to force saving the solution.

The monitoring of the residuals will indicate which flow quantity diverges first and if the divergence is sudden or builds up, which will help the troubleshooting. The solution should be analyzed in the Results analysis context to point to the location of the problem.

How did you determine the boundary conditions for the turbulence quantities?
Does the field of mut/mu on cut planes throughout the domain look physical?
The mesh quality is ok?
How about the convergence of the simulation with k-w? It is well converged?

Best regards,
Colinda

RE: Size of the geometry different from what is shown in the demo video

Colinda — Mon, 18 Dec 2023 08:07:51 GMT

That is correct: as written in the challenge description, the initial cell size for the challenge is 0.004.

It is also written in the challenge instructions: “use the provided geometry and values seen below to complete the challenge, as they may differ from those in the demo!”

So, it is on purpose different from what is shown in the video. The video is merely a demo of how a case like this is done. If the challenge would be with the exact same settings, it would just be copying the steps of the video. For good learning the lessons learned need to be applied to a slightly different case. In that way you are more likely to remember.

In addition a demo with the exact same settings would allow to get answers to the quiz questions from the demo video without running the simulation.

Size of the geometry different from what is shown in the demo video

Colinda — Mon, 18 Dec 2023 08:03:27 GMT

I was following the challenge from module 2 and it seems that the geometry (*.stl) that is provided for the elbow duct is half the size of the one shown in the demo video. This becomes evident when the mesh is generated; in order to get the same mesh, I need to specify an Initial cell size = 0.004 rather than 0.008 as shown in the video.

Presentation Slides

MaxiWei — Tue, 13 Dec 2022 20:39:00 GMT

Hi all,

is it possible to get the presentation slides? It would be very helpful to learn the presented topics and have them in a structured way to look up the important points later.

Thank you very much!

Best regards

Max

RE: Presentation Slides

Colinda — Wed, 22 Nov 2023 08:19:40 GMT

The CFD Online Course is now available in support.cadence.com for your reference. Access can be provided through the professor at your university managing the license or through the individual student license as can be requested here: Contact Us | Cadence

RE: Lectures availability

Colinda — Wed, 22 Nov 2023 07:44:53 GMT

Hi Marco,
The lectures are now available in support.cadence.com.

Under Learning go to the Training Catalog and look for the CFD Online Course 2023.1.

Best regards,
Colinda

Lectures availability

Marco Ado — Mon, 13 Mar 2023 13:47:47 GMT

Hi,

I've noticed that it is no more possible to watch the recording of the lectures. Is it possible to unlock them for a few days? I would like to check for some information.

Thank you!

RE: Lectures availability

Colinda — Thu, 03 Aug 2023 07:33:35 GMT

We are migrating the lectures to support.cadence.com where they will be available for review. If specific information is needed, please ask it in this forum as a temporary solution.

RE: Challenge 7 - Simulation Setting

Manuel NIB — Mon, 06 Feb 2023 11:58:43 GMT

Hi Stefano,

to question 1: Please use as initial solution the lower value with 6 m/s (and 1m/s for vt). This is on purpose. In general it is a good idea to be as close as possible to the final expected flow field. In practice the geometry is complex and we need to find a good compromise. For this particular case 23m/s would be a very suitable initial solution for the smaller pat of the pipe. In the larger part of the pipe a much lower value is more suitable. In general it is more robust for the simulation, if we choose as initial solution a value, which is rather on the lower side. For some applications it even make sense to initialize the domain with 0m/s.

To question 2: You can choose either of these. We can do this because we specify a constant value. Here it does not matter, if we specify in the cartesian system the constant value or in the cylindrical system the constant value. But your thought is right!

Best regards
Manuel

Pipe with varying cross-section

Commodaro — Mon, 06 Feb 2023 11:08:18 GMT

Hi,

Question 1

in instructions for Challenge 7, we have

Initial Solution: Vz=6 m/s

Boundary Condition - inlet: Vz=23 m/s

Should't they be equal to 23m/s ?

Question 2

Boundary Conditions - Outlet: averaged static pressure imposed.

Should we specify it as "Cartesian - Static pressure imposed" or as "Cylindrical - Static pressure imposed" (we used cylindrical for the Inlet)?

I see in the demo video that Cartesian is used.

thanks

Stefano

Pipe with varying cross-section: change of tables in simulation context

Konra — Tue, 31 Jan 2023 19:44:38 GMT

Hallo,
ich habe seit der letzten Challenge das Problem, dass sich die Tabelle im Simulationskontext nicht ändern lässt. Egal wie oft ich die Simulation neu durchlaufen lasse, ich kann mir nicht den Totaldruck und y+ anzeigen lassen. Auch im Convergence monitor kann ich unter Physical quantities nur noch Lift, Drag und Moment auswhälen. Massenstrom ist nicht mehr vorhanden. Unten in der Tabelle werden verscheidene Werte angezeigt, aber nicht der Totaldruck und y+, wie ich es im Simulations output eingestellt habe.

So kann ich die Challenge leider nicht weiterbearbeiten.

RE: Challenge 7: Simulations Auswertung fehlerhaft seit letzter Challenge

Manuel NIB — Mon, 06 Feb 2023 11:32:32 GMT

Hi Konra,

please excuse my late response.

The y1 simulation should reach the convergence criterion of -8. The simulation y50 should not. I would suggest let's first check, if your grid seems to be alright. Do you also have the following cell count?

Best regards
Manuel

RE: Challenge 7 - Simulation Setting

Commodaro — Mon, 06 Feb 2023 11:12:21 GMT

Question 1, also Vt is different, 1m/s in initial solution, 3 m/s in boundary conditions. Shall we use 3 overall?

RE: Challenge 7: Simulations Auswertung fehlerhaft seit letzter Challenge

Konra — Wed, 01 Feb 2023 15:30:50 GMT

Hi Manuel,

thanks that solved my problem with the convergence Monitor.

I still have a problem with the simulation: None of the two simulations is even close to reaching the convergence criterion of -8.0. The y1 Simulation has a mass flow error of 2.95% when finished. y50 has a mass flow error of 0.003081%. Is this to be expected?

RE: Challenge 7: Simulations Auswertung fehlerhaft seit letzter Challenge

Manuel NIB — Wed, 01 Feb 2023 09:43:10 GMT

Hi Konra,

I hope it is okay that I answer your question in English, so that also some other international students can follow this topic.

You mentioned that you only see lift, drag and momentum. That sounds to me that the flow configuration of the simulation is set to external. Please set it internal.

Then the physical quantities of the convergence monitor will switch to mass flow, pressure ration and efficciency. I assume that this should also solve the problem with the tabulated data.

Please let know, if this fixed the problem.

Best regards
Manuel

Calculation change in lift and drag.

AnasY — Tue, 24 Jan 2023 12:10:23 GMT

Hello,

In the questionnaire the formula to calculate the change in lift (analogical to drag) is given as followed: (lift_refined - lift_unrefined) / lift_unrefined. As the answers to choose from are given in percentage, am i correct in converting the result from the formula to percentage by moving two decimal places of the result to the right ?

Best regards.