Problem in 3D discontinuous Galerkin computation

General Discussion
1

Dear all,
I am trying to compute discontinuous Galerkin method in 3D and trying to compute a discrete solution at certain time level. But when I use jump or mean in intallfaces for the previous time level solution in the variational form, I got this error message:
“Internal error : no jump or average in intallfaces for RHS”.
What should I do now?

2

Possible I miss some code !

I will try a add if you sane the example.

3

sample.edp (2.6 KB)
I am attaching the code. Please suggest me if there is any solution for this problem.

4

Effectively, a peace of code is missing
I try to add rapidly.

5

if it is urgent then you can compute the RHS by a matrix product

use a couple Finite element

fespace Wh(Th,[P1dc,P1dc,P1dc]; ); 
Vh [ux, uy,uz],[ uu1, uu2,uu3], [dux,duy,duz], [vx, vy, vz];
varf  vM([ uu1, uu2,uu3],[vx, vy, vz]) = -intallfaces(Th)(
				((1./dt)*k *(mean(dx(uu1))*N.x*jump(vx)+mean(dy(uu1))*N.y*jump(vx)+mean(dz(uu1))*N.z*jump(vx)+mean(dx(uu2))*N.x*jump(vy)+mean(dy(uu2))*N.y*jump(vy)+mean(dz(uu2))*N.z*jump(vy)+mean(dx(uu3))*N.x*jump(vz)+mean(dy(uu3))*N.y*jump(vz)+mean(dz(uu3))*N.z*jump(vz)
			+mean(dx(vx))*N.x*jump(uu1)+mean(dy(vx))*N.y*jump(uu1)+mean(dz(vx))*N.z*jump(uu1)+mean(dx(vy))*N.x*jump(uu2)+mean(dy(vy))*N.y*jump(uu2)+mean(dz(vy))*N.z*jump(uu2)+mean(dx(vz))*N.x*jump(uu3)+mean(dy(vz))*N.y*jump(uu3)+mean(dz(vz))*N.z*jump(uu3)
			+ (pena*NN)*(jump(uu1)*jump(vx)+jump(uu2)*jump(vy)+jump(uu3)*jump(vz))
				))/ nElementonB
				        )
matrix M(Wh,Wh); 
real[int] RHS = M*uu1[];

use RHS as you rhs of you equation !
like in Time dependent schema optimization for heat equations

6

I do the correction in version - add missing code for Discontinous Galerkin in 3d for RHS · FreeFem/FreeFem-sources@6e209d5 · GitHub

I think it work, I just do sample test.