Couple LES-Boussinesq

[POSThu]

Dear all,

I am trying to simulate a non-isothermal turbulent airflow field, and refering to the paper “OpenLB—Open source lattice Boltzmann code”, I think the couple of LES model with Boussinesq assumption can be achieved by the class “SmagorinskyBoussinesqCouplingGenerator3D”. And refering to the official thermal example case “squareCavity2d”, I wrote the code like this:

  SuperLattice3D<T, NSDESCRIPTOR> NSLattice( superGeometry_room );
  SuperLattice3D<T, TDESCRIPTOR> ADLattice( superGeometry_room );

#ifdef TauEffLES
  ExternalTauEffLESForcedBGKdynamics<T, NSDESCRIPTOR> bulkDynamics(
    converter.getLatticeRelaxationFrequency(),
    instances::getBulkMomenta<T,NSDESCRIPTOR>(), 0.1);

  ExternalTauEffLESBGKdynamics<T, TDESCRIPTOR> TbulkDynamics(
    converter.getLatticeThermalRelaxationFrequency(),
    instances::getAdvectionDiffusionBulkMomenta<T,TDESCRIPTOR>(), 0.1);

#elif defined(SMAGORINSKY)
  SmagorinskyForcedBGKdynamics<T, NSDESCRIPTOR> bulkDynamics(
    converter.getLatticeRelaxationFrequency(),
    instances::getBulkMomenta<T, NSDESCRIPTOR>(),0.1);

  SmagorinskyBGKdynamics<T, TDESCRIPTOR> TbulkDynamics(
    converter.getLatticeThermalRelaxationFrequency(),
    instances::getAdvectionDiffusionBulkMomenta<T,TDESCRIPTOR>(), 0.1);
#else
  ForcedBGKdynamics<T, NSDESCRIPTOR> bulkDynamics(
    converter.getLatticeRelaxationFrequency(),
    instances::getBulkMomenta<T,NSDESCRIPTOR>());

  AdvectionDiffusionBGKdynamics<T, TDESCRIPTOR> TbulkDynamics(
    converter.getLatticeThermalRelaxationFrequency(),
    instances::getAdvectionDiffusionBulkMomenta<T,TDESCRIPTOR>());
#endif

  std::vector<T> dir{0.0, 0.0, 1.0};

  T boussinesqForcePrefactor = 9.81 / converter.getConversionFactorVelocity() * converter.getConversionFactorTime() *
                               converter.getCharPhysTemperatureDifference() * converter.getPhysThermalExpansionCoefficient();

#if defined(TauEffLES)
  SmagorinskyBoussinesqCouplingGenerator3D<T, NSDESCRIPTOR>
  coupling(converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2*converter.getConversionFactorLength()), converter.getLatticeLength(4+2*converter.getConversionFactorLength()),
           boussinesqForcePrefactor, converter.getLatticeTemperature(Tinlet), 1., dir, 0.85, bulkDynamics.getPreFactor());
#elif defined(SMAGORINSKY)
  SmagorinskyBoussinesqCouplingGenerator3D<T, NSDESCRIPTOR>
  coupling(converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2*converter.getConversionFactorLength()), converter.getLatticeLength(4+2*converter.getConversionFactorLength()),
           boussinesqForcePrefactor, converter.getLatticeTemperature(Tinlet), 1., dir, 0.85, bulkDynamics.getPreFactor());
#else
  NavierStokesAdvectionDiffusionCouplingGenerator3D<T,NSDESCRIPTOR>
  coupling(converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2-2*converter.getConversionFactorLength()), converter.getLatticeLength(2+2*converter.getConversionFactorLength()),
           converter.getLatticeLength(-2*converter.getConversionFactorLength()), converter.getLatticeLength(4+2*converter.getConversionFactorLength()),
           boussinesqForcePrefactor, converter.getLatticeTemperature(Tcold), 1., dir);
#endif

  NSLattice.addLatticeCoupling(superGeometry_room, 1, coupling, ADLattice);

After compiled the code, I tried to use different turbulent model by defining TauEffLES or SMAGORINSKY. When I defined TauEffLES, the program can output some results, and it seems that the two distribution functions have been coupled. However, when I defined SMAGORINSKY, the program can not be excuted and some errors have been outputed in the console, like this:

==== backtrace (tid:  12187) ====
 0 0x000000000004cb95 ucs_debug_print_backtrace()  ???:0
 1 0x000000000042f1f3 olb::SmagorinskyBoussinesqCouplingPostProcessor3D<double, olb::descriptors::D3Q19<olb::descriptors::FORCE> >::processSubDomain()  ???:0
 2 0x0000000000427404 olb::SmagorinskyBoussinesqCouplingPostProcessor3D<double, olb::descriptors::D3Q19<olb::descriptors::FORCE> >::process()  ???:0
 3 0x000000000040ed3c main()  ???:0
 4 0x0000000000022555 __libc_start_main()  ???:0
 5 0x000000000040f7df _start()  ???:0
=================================
[n1902:12187] *** Process received signal ***
[n1902:12187] Signal: Segmentation fault (11)
[n1902:12187] Signal code:  (-6)
[n1902:12187] Failing at address: 0x86e00002f9b
==== backtrace (tid:  12140) ====
 0 0x000000000004cb95 ucs_debug_print_backtrace()  ???:0
 1 0x000000000042f1f3 olb::SmagorinskyBoussinesqCouplingPostProcessor3D<double, olb::descriptors::D3Q19<olb::descriptors::FORCE> >::processSubDomain()  ???:0
 2 0x0000000000427404 olb::SmagorinskyBoussinesqCouplingPostProcessor3D<double, olb::descriptors::D3Q19<olb::descriptors::FORCE> >::process()  ???:0
 3 0x000000000040ed3c main()  ???:0
 4 0x0000000000022555 __libc_start_main()  ???:0
 5 0x000000000040f7df _start()  ???:0

I have tried some measures but the problem is still here. I hope someone can give me some advices. Thank you.

Best Regards,
Mengqiang

[Adrian]

If you post your full code I’d be happy to take a closer look.
Did you already try running the code in unoptimized debug mode and checking the backtrace there?

For a starting point the actual types of NSDESCRIPTOR and TDESCRIPTOR would be of interest.

[Author]

Posts