Turbulent Flow Simulations with Elmer - A Request

Numerical methods and mathematical models of Elmer
annier
Posts: 1168
Joined: 27 Aug 2013, 13:51
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by annier »

Dear All,
I will also try to see some of the models in OpenFOAM, and try to see if there are ways how we can implement newer tests on existing turbulence modeling using CFD in Elmer.


Yours Sincerely,
Anil Kunwar
Anil Kunwar
Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice
RaJa
Posts: 77
Joined: 22 Oct 2014, 09:48
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by RaJa »

Hello,

may I ask if there is any progress on this topic?

Right now, I have exactly this use case: a complex mesh that relies on prims (due to extrusion) and turbulent flow. Right now I am getting the "706 bubbles not implemented" error (even if I switch off stabilization in the k-solver). I could really need this feature.

I would offer help, if I could code Fortran.

Thanks,
Rainer
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by kevinarden »

have the same error, Elmer suggested to use P elements, but when I add P elements it says the same.

KESolver:
KESolver:
KESolver: -------------------------------------
KESolver: KEpsilon iteration: 1
KESolver: -------------------------------------
KESolver:
KESolver: Starting Assembly...
KESolver: Assembly
ERROR:: ElementInfo: Bubbles for element: 605 are not implemented.
ERROR:: ElementInfo: Please use p-element basis instead.
STOP 1
case.sif
(2.94 KiB) Downloaded 310 times
mika
Posts: 230
Joined: 15 Sep 2009, 07:44

Re: Turbulent Flow Simulations with Elmer - A Request

Post by mika »

If the mesh contains prisms or pyramids, at the moment the value of the keyword "Stabilization Method" must be defined to be some other string than "bubbles". If bubble functions were wanted for stabilization purposes, one could try to use for example

Stabilization Method = String "p-bubbles"
Element = "p:1 -tetra b:1 -pyramid b:1 -prism b:1 -brick b:1"

A basic p-element definition without additional bubble functions could be

Stabilization Method = String "none"
Element = "p:2"

-- Mika
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by kevinarden »

Thanks Mika, that does indeed allow the simulation to proceed.

Kevin
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by kevinarden »

https://www.openfoam.com/documentation/ ... ulent-flow

I have built three of the turbulent flow cases from the openfoam user guide in Elmer. They have working meshes and sif files, but I have not confirmed the answers yet. Two of them have links to NASA test data cases.

I plan on building more over time.

The finished cases are here.

https://github.com/mrkearden/turbulent

It contains the sif, meshes, and the Salome study. The cube case has two meshes, one with pyramids, using p-elements and one with all cube elements.

Kevin
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by kevinarden »

I have created 5 of turbulent test cases using Elmer in the github repository.
annier
Posts: 1168
Joined: 27 Aug 2013, 13:51
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by annier »

Dear Kevin,
Thank you very much for developing test cases for Kepsilon solver of Elmer software.


Yours Sincerely,
Anil Kunwar
Anil Kunwar
Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Turbulent Flow Simulations with Elmer - A Request

Post by raback »

Hi All,

I've been working a little with this turbulence stuff the last two days. Some baby steps taken. So far I'm taking some ideas from:
http://num.math.uni-goettingen.de/bail/ ... mierka.pdf

The work happens on branch "KESolverRevisit". What I've been trying to do is to gradually adapt some improved strategies without breaking the code for the old functionality. For example,
  • Already earlier in the summer Juha has parallelized the Flux Corrected Transport (FCT) shceme that we hope could be used to transport the turbulent fields.
  • The transient simulation may now be used as pseudo-transient to drive the simulation forward in time until convergence is found. For this purpose there are many new transient level keywords for convergence. Typically just replace "Steady State Keyword" with "Transient Keyword". For example "Transient Convergence Tolerance". Going transient is needed if we want to utilize FCT.
  • As in the paper it is now possible to use a auxiliary field eps/k which has some favourable numerical properties. At least for the eye it would seem that the convergence of the turbulent quantities looks better.
  • The k-eps solver has been monolithic i.e. the fields are solver from 2x2 block matrix. It is now possible to run the solver with just either component. This has not yet been much tested. This is also needed for FCT.
The current state does not probably offer much benefits over the previous version.
There are further steps to be taken:
  • Ensure the blockwise operation of KESolver.
  • Take FCT into use (instead of Bubble stabilization)
  • Further optimize assembly and operation of KESolver once it is proven.
  • Implement turbulence models to the new generation IncompressibleNSVec solver. This will provide robust block precondioning technique and faster assembly.
  • etc.
Attached is a sif file that works with the test case "Step_ke".

-Peter
Attachments
test.sif
Tentative sif file for test case Step_ke
(4 KiB) Downloaded 271 times
kishpishar
Posts: 54
Joined: 17 Jun 2015, 10:04
Antispam: Yes

Re: Turbulent Flow Simulations with Elmer - A Request

Post by kishpishar »

Hi Peter,

Many thanks for your efforts towards the new solver and sorry for not being able to respond earlier. I will do some tests with both the old and new versions and post the results. Thanks again,

-kumar
Post Reply