MG - test3 - is it okay?

[jothurgood]

MG - test3 - is it okay?

[JOThurgood]

its because you hacked them to be zero gradient at some earlier stage apparently

532 if (mg_state%eta_bc_xmin == fixed) then
533 ! this%eta(0,:) = -this%eta(1,:)
534 this%eta(0,:) = this%eta(1,:)
535 endif
536 if (mg_state%eta_bc_xmax == fixed) then
537 ! this%eta(this%nx+1,:) = -this%eta(this%nx,:)
538 this%eta(this%nx+1,:) = this%eta(this%nx,:)
539 endif
540 if (mg_state%eta_bc_ymin == fixed) then
541 ! this%eta(:,0) = -this%eta(:,1)
542 this%eta(:,0) = this%eta(:,1)
543 endif
544 if (mg_state%eta_bc_ymax == fixed) then
545 ! this%eta(:,this%ny+1) = -this%eta(:,this%ny)
546 this%eta(:,this%ny+1) = this%eta(:,this%ny)
547 endif
548

[JOThurgood]

...so that is equivalent to eta=1 Dirichlet in this case

[JOThurgood]

so should add inhomo dirichlet to eta_bcs and check same answer is given