[renyun0511]

Hi, all,

Recently, I try to simulate the transient inner flow of centrifugal pump by PimpleDyMFoam on OpenFOAM-2.1.0. But it always ended with the floating point exception error after some steps. I looked into the internal Field of my case by Ensight software and I found that there was a wrong pressure distribution in this centrifugal pump, although the directions of both velocity at inlet and velocity at outlet are right and the direction of impeller rotation is also right. I don't know why it was this? Can someone give me some hints? I will be very appreciated it!

In order to explain my questions better, I give the pressure distributions of this pump with steadystate and transient simulation at the first output time step, respectively. Here are:

Attachment 21641,Attachment 21642

The boundary file:
boundary.txt
The fvSchemes file:
fvSchemes.txt
The fvSolution file:
fvSolution.txt
The controlDict file:
controlDict.txt

The initial BCs:

FoamFile
{
    version     2.0;
    format      ascii;
    class       volVectorField;
    location    "3500";
    object      U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
dimensions      [0 1 -1 0 0 0 0];
internalField   nonuniform List<vector> 
1196068
(
(0.716798 0.840057 -7.4744)
(0.638485 0.861783 -7.48279)
(0.681562 0.835391 -7.58056)
...  ...
    wall
    {
        type            fixedValue;
        value           nonuniform List<vector>
88020
(
(0 0 0)
(0 0 0)
(0 0 0)
...  ...
    wallimp4
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);
    }
    wallimp3
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);
    }
    wallimp2
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);
    }
    wallimp1
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);
    }
    outlet
    {
        type            zeroGradient;
    }
    inlet
    {
        type            fixedValue;
        value           uniform (0 -2.708 0);
    }
}

FoamFile
{
    version     2.0;
    format      ascii;
    class       volScalarField;
    location    "3500";
    object      p;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
dimensions      [0 2 -2 0 0 0 0];
internalField   nonuniform List<scalar>
1196068
(
0.38443
0.248261
0.21847
...  ...
    wall
    {
        type            zeroGradient;
    }
    wallhgb
    {
        type            zeroGradient;
    }
    wallqgb
    {
        type            zeroGradient;
    }
    interafcevol1
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
1446
(
5.50122
5.60699
5.11278
...  ...
    interfacevol
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
22196
(
-92.0152
-109.644
-123.262
...  ...
    interfacesuc
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
2368
(
-675.839
-676.411
-675.763
...  ...
    interfaceimp1
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
1120
(
-676.74
-675.488
-676.987
...  ...
    interfaceimp
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
8680
(
-105.911
-104.788
-102.254
...  ...
    interfaceckyc
    {
        type            cyclicAMI;
        value           nonuniform List<scalar>
748
(
-6.02945
-7.33232
-6.99204
...  ...
    outlet
    {
        type            fixedValue;
        value           uniform 0;
    }
    inlet
    {
        type            zeroGradient;
    }
}

Thank you very much!


Yun Ren

[renyun0511]

Some description for the above attached pictures:
the pressure distribution by steady-state simulation, that is, Time=0s:
Attachment 21641
the pressure distribution by transient simulation, that is, Time=0.00005s:
Attachment 21642

[linnemann]

Try with these

[renyun0511]

Hi, Linnemann,

Thank you very much for your reply, but I'm afraid I can only open the fvSolution.txt file. There is something wrong with fvSchemes.txt. Can you uploaded it again?

Regards,

Yun Ren

[linnemann]

Here we go again

[renyun0511]

Hi, Linnemann,

I'm afraid that " Gauss faceMDLimited 1" is not a valid scheme in Version-2.1.0. When I added it in my original fvSchemes file, there is an error occured:

[0] --> FOAM FATAL IO ERROR:
[0] Unknown discretisation scheme faceMDLimited
Valid schemes are :
59
(
Gamma
GammaV
LUST
MUSCL
MUSCLV
Minmod
MinmodV
OSPRE
OSPREV
Phi
QUICK
QUICKV
SFCD
SFCDV
SuperBee
SuperBeeV
UMIST
UMISTV
biLinearFit
blended
clippedLinear
cubic
cubicUpwindFit
downwind
filteredLinear
filteredLinear2
filteredLinear2V
filteredLinear3
filteredLinear3V
fixedBlended
limitWith
limitedCubic
limitedCubicV
limitedLinear
limitedLinearV
limiterBlended
linear
linearFit
linearPureUpwindFit
linearUpwind
linearUpwindV
localBlended
localMax
localMin
midPoint
outletStabilised
pointLinear
quadraticFit
quadraticLinearFit
quadraticLinearUpwindFit
quadraticUpwindFit
reverseLinear
skewCorrected
upwind
vanAlbada
vanAlbadaV
vanLeer
vanLeerV
weighted
)

Regards,

Yun Ren

[linnemann]

Hi

Im sorry was writing based on memory and the memory was wrong :-)

here is the correct syntax

faceMDLimited Gauss linear 1;

[renyun0511]

Hi, Linnemann,

I changed the fvSchemes and fvSolution as you uploaded, but there is error about "mpi". So, I changed little by little, and finally, I found that it is wrong with the relaxationFactors:
Here is yours:

relaxationFactors
{
    fields
    {
 "p.*"  0.3;
    }
    equations
    {
        "U.*"           0.7;
 "k.*"  0.5;
 "omega.*" 0.5;
    }
}

and when I return to my original setup, the case can going. The difference between your fvSolution and mine is only the relaxationFactors, and here is mine:

relaxationFactors
{
    fields
    {
    }
    equations
    {
        "U.*"           1;
    }
}

I think in the transient simulation, there is not a relationship with the relaxation factors any more.

At last, when I use your fvSchemes instead of mine and my fvSolution, the result didn't changed a little.


Regards,

Yun Ren

[ARohit]

Dear all,

I am having fundamental question regarding steady and transient simulation.

In steady sim, it takes around 100 to 150 iterations to solve a perticular problem. any problm for instance.

But while doing transient sim, we give maximum coefficient loops i.e max iterns per timestep as 5 or 6..

How it is possible for computer to get soln at that timestep in such small no of iterns.?

thanks in advance.