[deal.II] Re: Assembly of material forces

Dear Seyed, On top of what Wolfgang has already said, I have a few more direct comments about what you posted a few days ago. > The computation of configurational forces is accomplished by the following > formula: > > > >

Re: [deal.II] Re: Add heterogeneous parameters controlled by function on cells

Dear Wolfgang and Jean-Paul Thank you for your deeply supports and I am clear now I would close this topic. On Thursday, January 26, 2017 at 3:33:15 PM UTC+1, Wolfgang Bangerth wrote: > > On 01/26/2017 01:20 AM, Lam DANG wrote: > > > > But I still a concern that the function class (the Coeffic

Re: [deal.II] Re: Add heterogeneous parameters controlled by function on cells

On 01/26/2017 01:20 AM, Lam DANG wrote: But I still a concern that the function class (the Coefficient class in step 6) will be called in many time in the main code. Yes, but so are many other functions (e.g., FEValues::shape_grad). Is this function *particularly* expensive in your case? I'

Re: [deal.II] Re: Add heterogeneous parameters controlled by function on cells

Dear Lam, Great, I'm glad that worked for you. Do we have another approach, for example store the heterogeneous > permeability belonged to each cell, to avoid wasted computational cost? You can looks at step-18 and step-44, which both store local data associated with the quadrature points. B

Re: [deal.II] Re: Add heterogeneous parameters controlled by function on cells

Dear Wolfgang, Jean-Paul and all I followed your suggestions and the code run. Thank for your helps. But I still a concern that the function class (the Coefficient class in step 6) will be called in many time in the main code. Do we have another approach, for example store the heterogeneous per