Hello, I would like to ask if anyone have ever met this kind of error message in CFX4 output file: oooooooooooooooooooooooooooooooooooooooooooooooooo oooooo ***** CSPACE: SINGULARITY IN COORDINATE TRANSFORMATION THE CFX-4 SOLVER HAS BEEN STOPPED BY SUBROUTINE CSPACE oooooooooooooooooooooooooooooooooooooooooooooooooo oooooo I am running the calculation of a flow with particles and it calculates and calculates and than suddenly when calculating the tracks (well it calc. the sources to the fluid phase) i got this kind of error. I got the geo imported from patran, but there is no problem with solution without particles. I just need a fast solution, and I still had not got an answer from support, so I was just wondering if any of user have met anything like this.

Well it also starts to print this error for some particles: ***** INTEG: FAILURE IN ONESTP DURING INTERPOLATION MODE IDID = -2. But not for the one where it stops the run with the previous error. Any hints?

nice summer and thanks

Matej

Hi,

Your should be more specific about your calculation- I have seen these error messages in connection with axisymmetric calculations with periodic boundaries. cspace problem) -- please give info.

It is a well known error message from CFX when doing particle calculations, so I am totaly surprised that support have not helped you yet, support should send you a object file, which solves the problem - but how is not told!

Are you using CFX4.2 ? , I have not seen the cspace problem in 4.3 or 4.4

The last error message can also be related to axisymmetric and or periodic problems - occurs frequently with turbulent dispersion close to a central axis. I did solve my problem with the use of some non standard object codes and user fortran from CFX - ask support. Regards Jan

Hi, I use CFX 4.3 The case is flow in cyklon, so there is a cyklon itself,and a tube where - in one block 2 places of particle inlet is defined and one air inlet. There's no symmetry planes no axisymetry. The cyklon has just one pressure boundary at outlet in the riser tube.

The particles are defined with the use of PARTICLE TRANSPORT BOUNDARY CONDITION in a block in computational coordinates. There is no heat transfer no chemical reaction, only turbulent dispersion (k-eps model used).

I do not know what to tell more, except to post the *.fc file )

But any hint is very usefull for me.

Thanks Matej

No hints, I have had similar problems, and for one of them the only solution was an object code from CFX. The problem was related to smal cells close to the axis of symmetry, it is not hard to imagine that problems with particle tracking can occur such a place. Specially in the case that turbulent dispersion is included, the particles gets captured in the central cells, either by crossing the symmetry 'axis' or the periodic boundaries all the time, and confuses the solver. There are different solutions to this problem but it is difficult to make something general that is not to restrictive. (the piece of object code from cfx probably did some cheating with moving the partciles away from the central axis)

But your problem is 3d in carthesian ? with a 5-block for the cylinder/cyclone? so it is a bit strange.

Have your tried a run without dispersion? not that, that is a solution.

Regards Jan