[Hgholami]

Dear foamers
I modify a fsiFoam and add an object to fluidSolidInterface as:

void Foam::fluidSolidInterface::updateFluidTemperature( )
{
if (conjugate())
{
Info <<"set temperature to fluid" << endl;

const scalarField solidZoneTemperature_ =
stress().faceZoneTemperature
(
solidZoneIndex(),
solidPatchIndex()
);
const scalarField solidZoneKappaDelta_ =
stress().faceZoneKappaDelta
(
solidZoneIndex(),
solidPatchIndex()
);

scalarField nbrFluidZoneTemperature_
(
fluidMesh().faceZones()[fluidZoneIndex()].size(),
0.0
);
scalarField nbrFluidZoneKappaDelta_ =
scalarField
(
fluidMesh().faceZones()[fluidZoneIndex()].size(),
0.0
);


nbrFluidZoneTemperature_ =
ggiInterpolator().slaveToMaster
(
solidZoneTemperature_
);
nbrFluidZoneKappaDelta_ =
ggiInterpolator().slaveToMaster
(
solidZoneKappaDelta_
);
Info << "test ggiInterpolator" << endl;
flow().setTemperature
(
fluidPatchIndex(),
fluidZoneIndex(),
nbrFluidZoneTemperature_,
nbrFluidZoneKappaDelta_
);
Info << "test interface" << endl;
}
Info << "test if out of conjugate()" << endl;
}

the solver compiled with gcc5 in ubuntu18.04 (foam-extend4.0) successfully. But in running a case I see the bellow error.

Time = 0.0001, iteration: 1
Modes before clean-up : 0, modes after clean-up : 0
Current fsi under-relaxation factor: 0.05
Creating accumulated fluid interface displacement
Maximal accumulated displacement of interface points: 0
GAMG: Solving for cellMotionUx, Initial residual = 0, Final residual = 0, No Iterations 1
GAMG: Solving for cellMotionUy, Initial residual = 0, Final residual = 0, No Iterations 1
GAMG: Solving for cellMotionUx, Initial residual = 0, Final residual = 0, No Iterations 1
GAMG: Solving for cellMotionUy, Initial residual = 0, Final residual = 0, No Iterations 1
set temperature to fluid
test ggiInterpolator
test interface
free():invalid pointer
Aborted (core dumped)

I confused why this problem occur only in out of if loop. can you help me?
Thanks
Hojatollah

[Daniel_Khazaei]

take a look at these two threads:

corrupted size vs prev_size error in foam-extend4.0
*** glibc detected *** free(): invalid pointer

As stated, these kind of errors occur when the compiler has to free the same object twice or you are linking libraries that probably have classes with the exact same naming convention. Either way, this is a user coding error.

Compile and run the solver in debug mode, that will show you when and where the error occur.

Regards,

D. Khazaei

[Hgholami]

Yes. It seen, the problem can come from each code line, so I comment each line in object and found the problem comes from stress().faceZoneKappaDelta and fix it.
unfortunately, the showing error can't show correct position.

Originally Posted by Daniel_Khazaei

take a look at these two threads:

corrupted size vs prev_size error in foam-extend4.0
*** glibc detected *** free(): invalid pointer

As stated, these kind of errors occur when the compiler has to free the same object twice or you are linking libraries that probably have classes with the exact same naming convention. Either way, this is a user coding error.

Compile and run the solver in debug mode, that will show you when and where the error occur.

Regards,

D. Khazaei

[Daniel_Khazaei]

Originally Posted by Hgholami

Yes. It seen, the problem can come from each code line, so I comment each line in object and found the problem comes from stress().faceZoneKappaDelta and fix it.
unfortunately, the showing error can't show correct position.

I'm glad that you solved the problem with your code.

For detailed information you need to compile and run in debug mode and use gdb backtrace command to see which line in the code is causing any type of error. This provides step by step sequence of events that lead to the error.

[Hgholami]

Dear Khazaei
Thanks for your reply.
I have a question that not relative to this topic.
In setTemperature object. I see the reference article "Vynnycky, M., Kimura, S., Kanev, K., & Pop, I. (1998). Forced convection heat transfer from a flat plate: the conjugate problem. International Journal of Heat and Mass Transfer, 41(1), 45–59. doi:10.1016/s0017-9310(97)00113-0*".
It is for steady simulation. And also, in chtMultiRegionFoam solver, it use similar conjugate condition (mixed). Although uses temporal term in "fvm::ddt(rho, h)", but I am not sure the mixed condition for temperature is suitable for fsi problem with transient behavior. Do you sure about this condition?
Do you know, the commercial software such as ANSYS, COMSOL, use which conjugate boundary condition?
The result of mixed condition for pr >>1 for steady problem give appropriate agreement but for Pr =<1 is not good. Do you found any other reference for Pr =<1?
Thanks
Hojatollah

[Daniel_Khazaei]

Originally Posted by Hgholami

I have a question that not relative to this topic.
In setTemperature object. I see the reference article "Vynnycky, M., Kimura, S., Kanev, K., & Pop, I. (1998). Forced convection heat transfer from a flat plate: the conjugate problem. International Journal of Heat and Mass Transfer, 41(1), 45–59. doi:10.1016/s0017-9310(97)00113-0*".
It is for steady simulation. And also, in chtMultiRegionFoam solver, it use similar conjugate condition (mixed). Although uses temporal term in "fvm::ddt(rho, h)", but I am not sure the mixed condition for temperature is suitable for fsi problem with transient behavior. Do you sure about this condition?

As far as I know, there is no limitation with mixed boundary condition procedure to be used in steady or transient simulations. In chtMultiRegionFoam solver's description it is clearly stated:

Solver for steady or transient fluid flow and solid heat conduction, with
conjugate heat transfer between regions, buoyancy effects, turbulence,
reactions and radiation modelling.

Also a simple search using google brought up a few papers that use chtMultiRegionFoam solver for transient simulation. The conjugate procedure implemented in fsi toolkit is doing the exact same thing and I don't see why it can not be used in transient simulations as long as it maintains heat flux and temperature consistency on the interface for each time step.

As I didn't have enough time to check CHT + FSI cases yet, I created a new branch called feature-thermoCoupling and removed fsi related part from solids4Foam toolkit until I can make sure it is working as expected. I'm concerned about solid region formulation when dealing with CHT+FSI cases. Nevertheless, the new branch should practically work with both steady and transient CHT only simulations. Let me know if you face any problem running transient cases.

Originally Posted by Hgholami

Do you know, the commercial software such as ANSYS, COMSOL, use which conjugate boundary condition?

I don't know how these software implement conjugate boundary condition as they are closed-source, but they should all agree on two constrains:

1. heat flux consistency
2. temperature field continuity

Above conditions hold when no heat generation or temperature jump is specified on the interface.

Originally Posted by Hgholami

The result of mixed condition for pr >>1 for steady problem give appropriate agreement but for Pr =<1 is not good. Do you found any other reference for Pr =<1?

Which fluid model are you using for the simulation?
The validation test case provided with solids4Foam is using Pr = 0.01 and it is in good agreement with experimental data. Also you can see that solids4Foam and chtMultiRegionFoam produce almost identical results.

Regards,
D. Khazaei