Re: [deal.II] Example of using MeshWorker::Assembler::MGMatrixLocalBlocksToGlobalBlocks?

2017-02-17 Thread Wolfgang Bangerth
On 02/17/2017 09:14 PM, Fred wrote: I currently have a few codes written in the meshworker framework that solve systems of PDE's. (Elasticity, Laplace/Poisson, Stokes) I am interested in experimenting with Multigrid preconditioner for these. I have been unable to get this to work using the Mes

[deal.II] Example of using MeshWorker::Assembler::MGMatrixLocalBlocksToGlobalBlocks?

2017-02-17 Thread Fred
Hello, I currently have a few codes written in the meshworker framework that solve systems of PDE's. (Elasticity, Laplace/Poisson, Stokes) I am interested in experimenting with Multigrid preconditioner for these. I have been unable to get this to work using the MeshWorker::Assembler::MGMatrix

[deal.II] Re: CXX flags for SIMD and other optimization flags (GCC@6)

2017-02-17 Thread Denis Davydov
Minor follow-up to this topic: I submitted test results to CDash with gcc6.3 on Ubuntu and extra release flags https://cdash.kyomu.43-1.org/viewTest.php?onlyfailed&buildid=4886 -march=native -O3 -funsafe-loop-optimizations -flto A few tests fail (some with binary output, some with roundoff

Re: [deal.II] Solver convergence failure with L2 residual diverging for SolverBicgstab

2017-02-17 Thread Wolfgang Bangerth
On 02/17/2017 02:28 AM, Vinayak Gholap wrote: Its clear that at t 50.2 sec I have some problem with the system of equation leading to l2 residual norm to increase. I believe the source of failure is in the assert check in SolverBicgstab.h with the L2 residual norm: // in case of failure: thr

Re: [deal.II] Neumann vector conditions

2017-02-17 Thread Jean-Paul Pelteret
Dear Franco, > My question here is this: why am I not using the normal to the face? > To really understand this you need to become more familiar how you arrive at the weak form of the govening equations that you're trying to implement. On the Neumann boundary you are prescribing a traction *v

Re: [deal.II] Neumann vector conditions

2017-02-17 Thread luca.heltai
> On 17 Feb 2017, at 14:46, Franco Milicchio wrote: > > I agree I have a translation problem here. In practice, I am integrating a > weak problem as this (in Fenics' UFL terms, but easy to recognize): > > > inner( sigma(u), eps(v) ) * dx = inner(f, v) * dx + inner(f1, v) * ds(1) I’m gue

Re: [deal.II] Neumann vector conditions

2017-02-17 Thread Franco Milicchio
On Friday, February 17, 2017 at 1:48:54 PM UTC+1, Luca Heltai wrote: > > Dear Franco, > > I think there is a problem in your formulation… > > You are integrating > > u_i n_i f_i > > I agree I have a translation problem here. In practice, I am integrating a weak problem as this (in Fenics' UF

Re: [deal.II] Neumann vector conditions

2017-02-17 Thread luca.heltai
Dear Franco, I think there is a problem in your formulation… You are integrating u_i n_i f_i and while you can certainly come up with good reasons for this to work out, I’m unsure about what you want to achieve. If your boundary conditions are ((grad(u)).n,v) = u_i,j n_j v_i and you impos

[deal.II] Re: Neumann vector conditions

2017-02-17 Thread Franco Milicchio
On Thursday, February 16, 2017 at 7:49:52 PM UTC+1, Jean-Paul Pelteret wrote: > > Dear Franco, > > Super quick answer: Step-44 demonstrates how to implement the Neumann BC > for elasticity > > . > Thanks

[deal.II] Solver convergence failure with L2 residual diverging for SolverBicgstab

2017-02-17 Thread Vinayak Gholap
Hello everyone, I am working on a code for time dependent simulation. In the part of code I have an error after a certain number of time steps for a large size problem. The error occurs when it is calculating the k1 coefficient of implicit Runge Kutta method by solving system of equations using