|
[Sponsors] |
March 17, 2013, 10:25 |
2D engine simulation
|
#1 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hi foamers ,
I am simulating a 2D engine with sonicTurbDyMEngineFoam and simpleEngineTopoFVmesh library... my simulation is according to the 2D Tutorial case that attached by Prof.Jasak http://www.cfd-online.com/Forums/openfoam/83177-engine-simulation-mesh-motion-topological-changes-4.html Actually my result is not good.I attached the case and a picture of it at CAD=110. My problem is that velocity on the slave patch and master patch is zero(in paraview).(two boundaries that created for sliding interface.) these boundaries are valveCurtainCyl and valveCurtainPort. Boundary condition in my case for velocity on these boundaries is uniform (0 0 0) (similar to the 2D Tutorial ) But in the result of the tutorial case velocity on these boundaries is not zero(in paraview). How can I overcome this problem? Thanks and best regards, Sasan. |
|
March 20, 2013, 04:38 |
|
#2 |
Member
Laurent Fitschy
Join Date: May 2011
Posts: 40
Rep Power: 15 |
Hi,
I had a similar problem with sliding interfaces (AMI). In my case, it was a rendering artifact due to paraview. When loading internalMesh alone and choosing "cell data" instead of "point data" the solution looks good. Best regards. |
|
March 20, 2013, 06:28 |
|
#3 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hi Laurent,
I did it . But I haven't got good result... I think my boundary condition have a problem... I appreciate any help... Thanks, Sasan. |
|
March 21, 2013, 12:47 |
|
#4 |
Senior Member
Marco A. Turcios
Join Date: Mar 2009
Location: Vancouver, BC, Canada
Posts: 740
Rep Power: 28 |
Hey Sasan,
I'd have to look a little deeper at the simpleEngineTopoFvMesh mesh class, but it seems like the sliding patches are set to be walls, when they should be just general patches. Try that and see if it helps, as right now it seems like each port is separated from the rest of the cylinder. |
|
March 21, 2013, 13:31 |
|
#5 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hey Dear Marco,
Thank you very much . I appreciate your help. yes,It seems sliding patches have seperated the cylinder to five regions...I don't know where is the problem.... Thanks, Sasan. |
|
March 29, 2013, 05:14 |
|
#6 | |
Member
Martin Novák
Join Date: Dec 2012
Location: Prague
Posts: 70
Rep Power: 14 |
Hi Sasan.
Could you post the whole case somewhere here? It will be helpful. The correct setup of the case is in the ICE sim. really difficult and without the case I am not able to tell you what is wrong. However, I think that your problem is hidden in the settings of motionU boundary conditions. For the the outer sliding interface you have to have the commponentMixed bc and for the inner it has to be zeroGradient. You should probably look at the Jasak's case. ¨ I will expect something like Quote:
|
||
March 29, 2013, 06:44 |
|
#7 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hi Martin ,
Thanks for your reply. I attached the case in the first post of this thread. I attached the case again here . Actually the motionU file of my case is similar to the jasak's case but my geometry is in the X-Y plane and jasak's case is in the X-Z plane , So I changed some things in my case. The motionU file for my case: Code:
/*--------------------------------*- C++ -*----------------------------------*\ | ========= | | | \\ / F ield | OpenFOAM Extend Project: Open source CFD | | \\ / O peration | Version: 1.6-ext | | \\ / A nd | Web: www.extend-project.de | | \\/ M anipulation | | \*---------------------------------------------------------------------------*/ FoamFile { version 2.0; format ascii; class tetPointVectorField; location "0"; object motionU; } // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // dimensions [0 1 -1 0 0 0 0]; internalField uniform (0 0 0); boundaryField { piston { type componentMixed; refValue uniform (0 0 0); valueFraction uniform (0 1 0); } liner { type componentMixed; refValue uniform (0 0 0); valueFraction uniform (1 0 1); } cylinderHead { type fixedValue; value uniform (0 0 0); } intakePort { type fixedValue; value uniform (0 0 0); } presin { type fixedValue; value uniform (0 0 0); } exhaustPort { type fixedValue; value uniform (0 0 0); } valveStem1 { type fixedValue; value uniform (0 0 0); } valveStem2 { type fixedValue; value uniform (0 0 0); } presout { type fixedValue; value uniform (0 0 0); } valveCurtainCyl1 { type componentMixed; refValue uniform (0 0 0); valueFraction uniform (1 0 1); } valve1 { type fixedValue; value uniform (0 0 0); } valve2 { type fixedValue; value uniform (0 0 0); } valveCurtainPort1 { type zeroGradient; } valveDetachCyl1 { type zeroGradient; } valveDetachPort1 { type zeroGradient; } valveCurtainCyl2 { type componentMixed; refValue uniform (0 0 0); valueFraction uniform (1 0 1); } valve3 { type fixedValue; value uniform (0 0 0); } valve4 { type fixedValue; value uniform (0 0 0); } valveCurtainPort2 { type zeroGradient; } valveDetachCyl2 { type zeroGradient; } valveDetachPort2 { type zeroGradient; } defaultFaces { type empty; } } // ************************************************************************* // Thanks and best regards, Sasan. |
|
March 29, 2013, 10:06 |
|
#8 | |
Member
Martin Novák
Join Date: Dec 2012
Location: Prague
Posts: 70
Rep Power: 14 |
Quote:
motionU seems fine, but you have made mistake in the naming of the sliding interafaces. valveCurtainCyl is outer (without "hole") and valveCurtainPort is inner (with "hole"). If you take a look in meshModifiers, you will see that just 5 of them have been created insteed of 9. So change the naming and the case should be fine. |
||
March 30, 2013, 03:55 |
|
#9 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hi Martin ,
Thank you very much.. I corrected the name of valveCurtainCyl and valveCurtainPort and I created all mesh modifier : Code:
/*--------------------------------*- C++ -*----------------------------------*\ | ========= | | | \\ / F ield | OpenFOAM Extend Project: Open source CFD | | \\ / O peration | Version: 1.6-ext | | \\ / A nd | Web: www.extend-project.de | | \\/ M anipulation | | \*---------------------------------------------------------------------------*/ FoamFile { version 2.0; format ascii; class polyTopoChanger; location "constant/polyMesh"; object meshModifiers; } // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // 9 ( valveSlider1 { type slidingInterface; masterFaceZoneName curtainPortZoneV1; slaveFaceZoneName curtainCylZoneV1; cutPointZoneName cutPointsV1; cutFaceZoneName cutFaceZoneV1; masterPatchName valveCurtainPort1; slavePatchName valveCurtainCyl1; typeOfMatch integral; coupleDecouple true; projection visible; attached false; active true; } valveAttachDetach1 { type attachDetach; faceZoneName detachFaceZoneV1; masterPatchName valveDetachCyl1; slavePatchName valveDetachPort1; triggerTimes 0(); manualTrigger true; active true; } valvePoppetLayer1 { type layerAdditionRemoval; faceZoneName poppetZoneV1; minLayerThickness 0.0005; maxLayerThickness 0.001; oldLayerThickness -1; active true; } valveBottomLayer1 { type layerAdditionRemoval; faceZoneName bottomZoneV1; minLayerThickness 0.0005; maxLayerThickness 0.001; oldLayerThickness -1; active true; } valveSlider2 { type slidingInterface; masterFaceZoneName curtainPortZoneV2; slaveFaceZoneName curtainCylZoneV2; cutPointZoneName cutPointsV2; cutFaceZoneName cutFaceZoneV2; masterPatchName valveCurtainPort2; slavePatchName valveCurtainCyl2; typeOfMatch integral; coupleDecouple true; projection visible; attached false; active true; } valveAttachDetach2 { type attachDetach; faceZoneName detachFaceZoneV2; masterPatchName valveDetachCyl2; slavePatchName valveDetachPort2; triggerTimes 0(); manualTrigger true; active true; } valvePoppetLayer2 { type layerAdditionRemoval; faceZoneName poppetZoneV2; minLayerThickness 0.0005; maxLayerThickness 0.001; oldLayerThickness -1; active true; } valveBottomLayer2 { type layerAdditionRemoval; faceZoneName bottomZoneV2; minLayerThickness 0.0005; maxLayerThickness 0.001; oldLayerThickness -1; active true; } pistonLayer { type layerAdditionRemoval; faceZoneName pistonZone; minLayerThickness 0.0005; maxLayerThickness 0.001; oldLayerThickness -1; active true; } ) // ************************************************************************* // Code:
--> FOAM FATAL ERROR: Error in point ordering: mixed used and unused points at the end of point list. Last used point: 2176 (-0.013 0 0.00107703) First unused point: 2177 (-0.013 -0.000900796 0.00107703) and point 2178 (-0.037 -0.0100088 0.00107703) is used by a live face. Face 1901 4(2180 2178 2198 2200) with points 4((-0.037 -0.00899545 0.00107703) (-0.037 -0.0100088 0.00107703) (-0.037 -0.0100088 -0.00107703) (-0.037 -0.00899545 -0.00107703)) Done. From function void polyMesh::initMesh() in file meshes/polyMesh/polyMeshInitMesh.C at line 174. FOAM aborting Aborted (core dumped) I attached a picture from grid around the intake valve at this time.The picture shows that the grid is degenerating ..... Also I attached the case.. Can you help me for overcoming this error? Do you know what is the origin of this error? I appreciate your help. Thanks and best regards, Sasan. |
|
March 30, 2013, 16:32 |
|
#10 | |
Member
Martin Novák
Join Date: Dec 2012
Location: Prague
Posts: 70
Rep Power: 14 |
Quote:
I am on vacation, so I can not test it by myself. Do you keep me informed about your progress? Martin |
||
March 30, 2013, 16:47 |
|
#11 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Dear Martin ,
Thank you very much.. I appreciate your help.. I am very glad that you help me. I will try to do it and I will report my result and progressing to you. Thanks and best regards, Sasan. |
|
March 31, 2013, 09:12 |
|
#12 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
Hi Martin ,
I hope you have a happy vacation. I played with min layer thickness for the valves (several times) and I changed the deform angle to 0.. I corrected the clearance (=0.01) and the bore(=0.1) and I checked the valve diameters (they were correct). unfortunately I couldn't overcome the error... I found a thread [ http://www.cfd-online.com/Forums/ope...t-working.html ] that they have discussed about this error for engine simulation they have said this error occurs when all points on the sliding interface are not on the same radius and they attached a utility that can repair this problem. I compiled the utility but I don't know that how can I use this utility to solve my problem.. I appreciate any help from you. Thanks and best regards, Sasan. Last edited by sasanghomi; March 31, 2013 at 16:17. |
|
April 4, 2013, 10:57 |
|
#13 |
Senior Member
Sasan Ghomi
Join Date: Sep 2012
Location: Denmark
Posts: 292
Rep Power: 15 |
I created a small change in geometry and I increased the gap between valve and valve seat and I refined the mesh at this gap and I could run the case...
Thanks, Sasan. |
|
April 30, 2015, 15:00 |
|
#14 |
New Member
BO
Join Date: Dec 2014
Posts: 21
Rep Power: 12 |
Hi guys,
Sorry for hacking into this threads. However, when I tried to run the case shared in the very beginning, some error showed up: Unknown solver type laplaceFaceDecomposition Code:
Valid solver types are: 9 ( RBFMotionSolver displacementComponentLaplacian displacementInterpolation displacementLaplacian displacementSBRStress laplace pseudoSolid velocityComponentLaplacian velocityLaplacian ) |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
diesel engine combustion simulation | kebi112358 | OpenFOAM | 4 | September 25, 2012 14:16 |
Engine Simulation in parallel | Peter_600 | OpenFOAM | 0 | July 26, 2012 07:02 |
help need for my project: design n simulation of ic engine | sid87 | ANSYS | 5 | February 27, 2012 23:03 |
Simulation of a four stroke engine cycle | Alfred Herman.S. | Siemens | 1 | May 27, 2002 08:05 |
Question concerning about the flow simulation of a fan in an engine room. | ghlee | Main CFD Forum | 3 | October 21, 1998 08:05 |