[vzooz]

Hello,

I'm wondering if there is a function (or anything else) which enables me to use the calculated value of an expression at the previous timestep in order to evaluate the value at the current timestep.

To make it more clear, the concept is that a coefficient "k" derives from a "Temperature" (among others) and is used at next for a new estimation of the "Temperature". The "Temperature" is an opening temp. at a specific boundary.

Thanks in advance!

[ghorrocks]

No there is no CEL expression to do this. I wish there was as it would make things much easier.

You can do it using user fortran of course. You can also use some undocumented features to make additional variables carry over from last time step.

And a final thing is that the CEL functions areaAve, massflow, count etc are all calculated based on the values from the previous iteration. In a steady state run that is the previous iteration, in a transient simulation it is the previous coefficient loop. Sometimes that can be useful.

[rae]

I have been struggling with variable definitions to keep the value of the previous timestep, e.g. if(mod(atstep,2)==0, <recalculate>, <keep>) but I don't know how to keep the variable unchanged, or to store its original value in another variable without cyclic issues. Is there an undocummented means like suggested or otherwise to do this? You also mention Fortran for the solver to do this but I am lost on that. Thanks.

[ghorrocks]

Search the forum for "trans_loop", that is the undocumented feature to do this. The alternate way of doing it is using fortran.