Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
How to store final time of simulation in a variable ?
Posted 09.12.2023, 05:46 GMT-5 Studies & Solvers Version 6.1 3 Replies
Please login with a confirmed email address before reporting spam
I'm chaining two time-dependent simulations, the first one finishing on a stop condition and yielding the solution, say 'sol 1'.
I know how to create and start a new simulation starting from the final state of the latter.
What I don't manage to do is to record the value of the last time of simulation 1 and use it as the starting time of simulation 2. I've tried do define a variable tlast1 with the expression :
withsol( 'sol1', t, setind(t, 'last') )
It yields the correct value of final time, but repeated at each output time of 'sol1'. Thus, no surprise that when I try to use this quantity as the starting time of the second simulation, I get the error :
Unknown model parameter. - Parameter: tlast1 - Property: tlist (Output times) - Time dependent: Time Dependent
because tlast1 is not a scalar (by the way, why ?)
I've been deeply through the documentation and can't manage to find a way to do that. If anyone can help, thank you very much by advance.
Please login with a confirmed email address before reporting spam
Sorry, I forgot to join a basic MPH file reproducing the problem. Here it is.
Attachments:
Please login with a confirmed email address before reporting spam
Hi Olivier,
Did you mamage to solve this problem? I am facing similar situation.
Thank you!
Please login with a confirmed email address before reporting spam
Olivier,
your issue is that the time list in the time dependent study is not accepting variables, only parameters. This is for a reason because variables might produce cumbersome conditions for a solver. A workaround might be to run a model method at the end of study 1 that sets a parameter in the parameter list to the tlast1 value and you can use this parameter then in study 2. I never did that by myself in such a case but I think it should be feasible. I successfully did write methods that create and set parameters for other purposes, e.g. geometry where variables can't be used as well.
Cheers Edgar
-------------------Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com