UDF Error message
 

Hi,

I wrote an udf to modify the turbulent viscosity for the k-e model. I compiled the udf and everything seemed fine. However, when I tried to run the case using the udf, I got the following error message :

" Floating point error: invalid number "

Any suggestions to solve the problem.

thanks

Re: UDF Error message
 

As a first step you could program the standard formualtion for mu_t an see if to get the same results as fluent before trying to implement own models. Maybe this will help to find errors in the udf. Maybe you udf tries to access values which are not available during the first iteration. Run the case without udf for some iterations and hook the udf later.

RoM

Re: UDF Error message
 

It seems that you have a division by zero....

It comes probably from epsilon which should never be equal to zero, so I suggest you replace epsilon by (epsilon + a very small number)....

please try

Re: UDF Error message
 

Thanks a lot for the answers.

I 've tried to run few iterations without the udf before I switch it on but still I got the same error message.

I checked the UdF there is no division by zero.......

Re: UDF Error message
 

Can you post you source?

RoM

Re: UDF Error message
 

Hi Tania,

Be careful sometimes if you have not followed the units properly in your UDF then also this error comes.I will suggest check units properly. Bye

Arvind Phate