Helmholtz 3D solver

[andy19191]

Elmer 6.2 (Rev: 5344)
Scientific Linux 6.1 x86_64
gcc (GCC) 4.4.5 20110214 (Red Hat 4.4.5-6)
gmsh 2.5.1

Enclose attachment (issue1.pdf and issue2.pdf) of what is intended to be a point source radiating into 1/8th free space. The expected 1D solution is largely present along the lower right hand edge but the upper edge and lower left edges have an incorrect constantish value around zero and an incorrect constantish high value respectively.

As a new user I am by no means on top of how to setup and run Elmer and so cannot rule out user/input error (or a gmsh error). However, the gross lack of symmetry and unreasonable values for some edges does suggest a bug. I note that the Helmholtz solver code was recently modified according to the thread "Re: Helmholtz Model".

Enclose the input files wg.geo and wg.sif. The solution was generated by:
> gmsh wg.geo (commands: mesh 3d, optimize 3D, save)
> ElmerGrid 14 2 wg.msh
> ElmerSolver

[andy19191]

Elmer 5.5.0 (Rev: )
Ubuntu 10.04 64 bit
gcc (Ubuntu 4.4.3-4ubuntu5) 4.4.3
gmsh 2.4.2

I have reproduced the problem using a slight older binary version suggesting it is not a recent issue (and increasing the likelihood I have not defined the problem correctly!).