Re: [deal.II] assertion failed: be_value_from_constant: unknown constant kind (shared/cfe/edgglue/edg_decl.c, line 1374)

Hi Wolfgang, thank you for the answer. The issue persist with a recent compiler version (17.0.5.20170817), and I can't use gcc because most of the libraries installed on the cluster were built with icc. I tried commenting out functions and found out that the error persists with only the follow

Re: [deal.II] Re: Problem with the use of FE_Nedelec with a mesh file generated from Gmsh

Dear professor Bangerth, Thanks for the reply. As I said, I will first try the new implementation myself to see if it works. Provided that it solves my problem correctly, I may create a small testcase that can be added into deal's test suite. Hope I can make some contribution in this regard. Best

Re: [deal.II] Re: Problem with the use of FE_Nedelec with a mesh file generated from Gmsh

Thanks for your suggestions. I think I will first try the new implementation to see if it works. Later I may create a small testcase which can be added in the test suite. Jianan 2017-12-28 10:58 GMT-05:00 Bruno Turcksin : > Jianan > > 2017-12-28 10:40 GMT-05:00 Jianan Zhang : >> >> So, can you

Re: [deal.II] How to realize a moving frame method?

On 12/21/2017 05:54 AM, Tongzhao Gong wrote: I'm using the deal.ii library to solve a phase-field model to simulate columnar dendrite growth during solidification, and several references proposed a "moving frame method" to improve the computational efficiency. For example, I employed a 100

Re: [deal.II] Re: Problem with the use of FE_Nedelec with a mesh file generated from Gmsh

On 12/28/2017 08:40 AM, Jianan Zhang wrote: To illustrate, I tried to solve the problem with Ross's exact solution (but with a larger value of omega) on the mesh generated by Gmsh, Then the results obtained are obviously wrong. But when I try the same exact solution on the mesh created by dea

Re: [deal.II] Re: Problem with the use of FE_Nedelec with a mesh file generated from Gmsh

Jianan 2017-12-28 10:40 GMT-05:00 Jianan Zhang : > > So, can you help me contact the people working on the PR to try this out > to see if my code works with the new implementation? > If you just want to try the new implementation you can take Ross' branch

Re: [deal.II] Computing Overlap Matrix

On 12/27/2017 10:15 PM, Phani Motamarri wrote: I have a STL array of parallel distributed vectors *std::vector X;* This unrelated to your problem, but storing (linear algebra) vectors in std::vector is awkward because it may require copying vectors around when you resize the std::vector. A

Re: [deal.II] Re: Problem with the use of FE_Nedelec with a mesh file generated from Gmsh

Hi, Bruno, Hope you had a good holiday. I have done more checks on my code these days and found that the problem truly lies in the use of FE_Nedelec for non-rectangular meshes. To illustrate, I tried to solve the problem with Ross's exact solution (but with a larger value of omega) on the me