Hi All, I am introducing particle from the bottom of by bubble column filled with water.

Previously what I have done, without any problems, is to run my transient simulation till i obtain a developed folw field. I then inject my particles. I changed the size of the particle and tried to inject them as i have always done but the solver stops after only one coefficient loop, with a message that. "Error #004100018 has occured in subroutine FINMES. Message: Fatal overflow in linear solver".

Has anybody seen this type of thing? how do you solve this?

thanks Anne

try to reduce your timestep, looks like you are using quite large dtime.

Thanks for your response. I tried to reduce time steps progressively upto 10^-5 sec, just for the sake of it. But why or how does the time step cause this if the solver stops even before a step is made? I mean, either before the very first coefficient loop or just after the first coefficient loop? What does this Routine FINMES do?

Your suggestions are highly appreciated. Anne

Anne, p/s read the USER manual under SOLVER for more details

Hi KKA, Thanks for your response. When we have problems, normally the first port of call is the manual, however, given the volume of the manual, it is not uncommon that one would miss a point. I suspect that might have happened to me. Could you kindly help me narrow the scope of my search by mentioning the section or page, if you know it, that addresses this very problem.

Regards, Anne

well there can be lot of issue with solver stopping before even starting. first make sure following things 1) your mesh is all correct? I mean check the min angle, check for negative volumes and stuffs 2) check for all the boundary condition and stuffs that they are properly imposed 3) check the unit consistency with mesh and within the parameters you specify. This can create a big problem

first make sure above things, i feel there is some basic problem rather then any other things...

hi Anne

As centaur_ks, said make sure all the units are consistent, all fundamental parameters are properly set, check his reply. After all that and still the problem persit then go to the Manual, select Master Content, then choose Reference. Under reference, select FAQ, choose solver problem. I geuss you'll find alot of info over there. If it's been TASCflow then it is so easily, just go to TroubleShooting. Email me if you weren't able to locate this section in the manual.

Hi KKA and Centaur, Many thanks for your kind interest. I have indeed chaked all that you mentioned, however, what puzzles me is that the same geometry with the same physics had worked some time last year! I will again look into those areas you have suggested, just incase there is something missing something.

regards, anne

Hiya Anne

One more thing, I'm not sure if I or Cantuer mentioned in our suggestions, Kindly check your BCs. Because at times your BC might be wrong expecially when it comes to default one. Go through the state or def files to see if the BCs are place at correct regions or locations. CFX used Wall as default BC. Which means that all the regions you didn't specify the BC it'll put a wall there and that might cause a great problem. Check that one and let's see. feel free to email or post any problem at any time.

good luck!

cheers

Hi KKA and Centaur,

Many thanks for your overwhelming support.

I am going to give it a last go, print all the generous suggestions that I have received from you, go thro' them slowly as i analyse the, output files (old and new), if it does not work I will sit back and wait for the CFX5.7. If the problem persists in CFX5.7, Then I will 'shout' again, meanwhile let me 'shout' to the CFX-ANSYS to send me the CFX5.7.

Regards, Anne

Hi Anne!

I had the same problem as you few days ago. I mean I got the same error message for a simulation in which the only change I made was the mesh. So what i was found out was: overflow- is a malfunction in which the program tries to store data and there is no free space. My mesh was 300 000 cells and i was simulating combustion and co. I passed my file to someone else with a stroger computer and it was working. Worth a try! Regards,Ioana

Ioana, Among all the logical suggestions I have so far received in this posting, this is as logical as it is convising to me. This I believe is the problem. starting with a base mesh, I have verified this as follows:

1) Increase the mesh size 2) Increase the number of phases eg from liquid-solid to liquid-solid-gas 3) Set Full Reynolds model. In all the 3 cases above, the same message came up.

Conclusion: It is a data storage problem!!

Many thanks to KKA, Centeur and Iona. Kind regards, Anne