[amanbearpig]

Hello, I've found I have restart issues with solids4Foam FSI problems. For example, if I run the laminar HronTurekFsi3 tutorial problem from 0 to 6 seconds, it runs fine with no issues. However if I restart the problem from 5 seconds, it soon blows up and diverges (in my test case it gave a few material stability errors at 5.002, started giving a bunch more around 5.03 seconds, and crashed at 5.035 seconds).

I'm using foam-extend-4.0, and the master solids4Foam branch, updated as of July 14th.

Anyone else notice this issue, or have any recommendations? Thanks!

[amanbearpig]

This issue is still plaguing me, and makes it rather difficult to perform large FSI problems. Here's some data from a run of the Hron & Turek FSI3 example problem (nothing in the problem setup has been changed, other than allowing it to run beyond t=6).

Original run (no restarts) at time 18.241:

Time = 18.241

Setting traction on solid interfaces
Interpolating face values using GGI
Total force on fluid interface 0: (0.481859 -2.11122 -1.70705e-15)
Total force on solid interface 0: (-0.481865 2.11122 1.70701e-15)

Evolving solid solver
Corr 0, relative residual = 1
Corr 382, relative residual = 9.77886e-08
FDICPCG: Solving for D, Initial residual = 0.00294397, Final residual = 2.67555e-09, No outer iterations = 382
 Max relative residual = 1, Relative residual = 9.77886e-08, enforceLinear = false
Interpolating point values using GGI
Interpolating point values using GGI
FSI relative residual1 norm for interface 0: 1
FSI residual2 norm for interface 0: 0.0361545

Time = 18.241, iteration: 1
Modes before clean-up (plate): 13, modes after clean-up (plate): 0
Current fsi under-relaxation factor (plate): 0.05
Maximal accumulated displacement of interface 0: 0.0331739
GAMG:  Solving for cellMotionUx, Initial residual = 0.998813, Final residual = 5.78483e-07, No Iterations 5
GAMG:  Solving for cellMotionUy, Initial residual = 0.999831, Final residual = 5.11375e-07, No Iterations 5
GAMG:  Solving for cellMotionUx, Initial residual = 0.00904394, Final residual = 3.03658e-07, No Iterations 5
GAMG:  Solving for cellMotionUy, Initial residual = 0.00795309, Final residual = 2.59858e-07, No Iterations 5
Evolving fluid model: icoFluid
Courant Number mean: 0.0719761 max: 1.15849 velocity magnitude: 4.25753
DILUPBiCG:  Solving for Ux, Initial residual = 0.00256183, Final residual = 2.02558e-07, No Iterations 3
DILUPBiCG:  Solving for Uy, Initial residual = 0.0110399, Final residual = 1.06781e-07, No Iterations 4
GAMG:  Solving for p, Initial residual = 0.543665, Final residual = 9.29462e-07, No Iterations 68
GAMG:  Solving for p, Initial residual = 0.426793, Final residual = 7.29255e-07, No Iterations 26
time step continuity errors : sum local = 6.4016e-10, global = 1.18308e-10, cumulative = 3.74158e-07
GAMG:  Solving for p, Initial residual = 0.146667, Final residual = 8.44153e-07, No Iterations 27
GAMG:  Solving for p, Initial residual = 0.0458902, Final residual = 9.87122e-07, No Iterations 20
time step continuity errors : sum local = 1.11489e-09, global = 2.03898e-10, cumulative = 3.74362e-07
GAMG:  Solving for p, Initial residual = 0.0254093, Final residual = 7.90576e-07, No Iterations 21
GAMG:  Solving for p, Initial residual = 0.0106926, Final residual = 9.48972e-07, No Iterations 17
time step continuity errors : sum local = 1.05209e-09, global = 2.00444e-10, cumulative = 3.74562e-07
Setting traction on solid interfaces
Interpolating face values using GGI
Total force on fluid interface 0: (14.6646 -789.207 1.29213e-15)
Total force on solid interface 0: (-14.6645 789.206 -1.29209e-15)

Now from a restart at time=18.241:

Time = 18.241


    From function fluidSolidInterface::updateCoupled()
    in file fluidSolidInterfaces/fluidSolidInterface/fluidSolidInterface.C at line 62
    Enabling fluid-solid coupling
Setting traction on solid interfaces
Interpolating face values using GGI
Create GGI zone-to-zone interpolator
interface-to-interface face error: 0.0226088
calcMasterPointAddressing() const
Extended GGI, master point distance, max: 3.14671e-06, avg: -2.0144e-06, min: -7.16238e-06
interface-to-interface point error: 7.16238e-06
Number of uncovered master faces: 0
Number of uncovered slave faces: 0

Total force on fluid interface 0: (0.591128 -2.0965 -1.70314e-15)
Total force on solid interface 0: (-0.591152 2.09653 1.70321e-15)

Evolving solid solver
Corr 0, relative residual = 1
Corr 400, relative residual = 9.96212e-08
FDICPCG: Solving for D, Initial residual = 0.0985019, Final residual = 3.8842e-07, No outer iterations = 400
 Max relative residual = 1, Relative residual = 9.96212e-08, enforceLinear = false
Interpolating point values using GGI
Interpolating point values using GGI
FSI relative residual1 norm for interface 0: 1
FSI residual2 norm for interface 0: 7.59435

Time = 18.241, iteration: 1
Modes before clean-up (plate): 0, modes after clean-up (plate): 0
Current fsi under-relaxation factor (plate): 0.05
Maximal accumulated displacement of interface 0: 0.908185
GAMG:  Solving for cellMotionUx, Initial residual = 0.99996, Final residual = 8.82619e-07, No Iterations 5
GAMG:  Solving for cellMotionUy, Initial residual = 0.999994, Final residual = 3.1945e-07, No Iterations 6
GAMG:  Solving for cellMotionUx, Initial residual = 0.00959629, Final residual = 4.53848e-07, No Iterations 5
GAMG:  Solving for cellMotionUy, Initial residual = 0.00936811, Final residual = 7.7219e-07, No Iterations 5
Evolving fluid model: icoFluid
Courant Number mean: 0.0724723 max: 0.718575 velocity magnitude: 4.25766

PISO: Operating solver in PISO mode

DILUPBiCG:  Solving for Ux, Initial residual = 0.00310856, Final residual = 3.53159e-07, No Iterations 2
DILUPBiCG:  Solving for Uy, Initial residual = 0.0107548, Final residual = 6.75168e-07, No Iterations 3
GAMG:  Solving for p, Initial residual = 0.523643, Final residual = 9.0237e-07, No Iterations 88
GAMG:  Solving for p, Initial residual = 0.387484, Final residual = 7.55657e-07, No Iterations 26
time step continuity errors : sum local = 7.59467e-10, global = 1.19647e-10, cumulative = 1.19647e-10
GAMG:  Solving for p, Initial residual = 0.0990405, Final residual = 9.37202e-07, No Iterations 42
GAMG:  Solving for p, Initial residual = 0.0431658, Final residual = 7.1427e-07, No Iterations 22
time step continuity errors : sum local = 8.89019e-10, global = -1.37628e-10, cumulative = -1.79811e-11
GAMG:  Solving for p, Initial residual = 0.0186159, Final residual = 9.35416e-07, No Iterations 23
GAMG:  Solving for p, Initial residual = 0.00802076, Final residual = 7.12265e-07, No Iterations 19
time step continuity errors : sum local = 8.63516e-10, global = -1.3629e-10, cumulative = -1.54271e-10
Setting traction on solid interfaces
Interpolating face values using GGI
Total force on fluid interface 0: (-37.2983 -941.572 4.14577e-16)
Total force on solid interface 0: (37.2977 941.579 -4.14509e-16)

The non-restart run converged in 15 FSI iterations.

The restart run began getting "material linearity enforced for stability!" warnings at iteration 4. It did however 'converge' after 20 iterations, though with a maximum fluid velocity of over 12. In the non-restarted run, the maximum fluid velocity was 4.25.

Obviously something isn't working quite right here, and I'm not quite sure where the issue lies.

[salachnaj]

Hi. I found a similar problem. I validated the Hron Turek FSI case without any problem on a single run. Now I am working now on different 3D case, and I get a strange results of a structure deformation

First I restarted my case after 2,3s. Here, I report x displacement(violet) and ydispl against experiment (dotted)
https://ibb.co/n1KFDP2

Here I restarted my case after around 1,5s and I see strange shift of displacement
https://ibb.co/mGLdqgt

I see the similar problem was reported here
https://sciendo.com/pdf/10.2478/caim-2021-0002

'The restart introduces a larger error at the FSI interface with respect to the error of an uninterrupted simulation.'

Have you found a fix for the issue? Let me know,

[anaspauzi]

Hi,
In the bitbucket solids4Foam, Dr Cardiff mentioned the problem with extrapolatedPressure BC, and the solution is to change to zeroGradient

https://bitbucket.org/philip_cardiff...-restart-gives

[anaspauzi]

Dear respected FOAMers,

I managed to restart for the first time but failed for the 2nd time.

A large lift force was present at the start of the restart and later changed the solid deformation.

In the time folder, there seems to be missing epsilon_0, and also occurs with other RANS models, omega_0 and R_0 after the first restart.

I tried copying the epsilon, omega, and R from the previous time-step to the newer one, rechanged the file name, and heading in the consequence time step to epsilon_0, omega_0, and R_0, but failed to run, missing ")" at zeroGradient;

Can anyone help?

Thank you

ANAS