change starting point, timesteps

The graphical user interface of Elmer
Post Reply
Trexabyte
Posts: 2
Joined: 14 Sep 2022, 11:57
Antispam: Yes

change starting point, timesteps

Post by Trexabyte »

Hello, I would like to resume an aborted simulation at a specific time step. How can I adjust the simulation parameters so that the simulation starts at a frequency step of 356 Hz instead of 1 Hz.
I use the scanning function:
Simulation Type = Scanning
Time Step Intervals = 1001
Timestep Sizes = 0.5
Many Thanks. If you need further information, please get in touch.
Best regards, Alex
kevinarden
Posts: 1513
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: change starting point, timesteps

Post by kevinarden »

See elmersolver manual chapter 3
https://www.nic.funet.fi/pub/sci/physic ... Manual.pdf

you would have had to ask for a restart file to be written in the first run, it is not done by default.
raback
Site Admin
Posts: 4423
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: change starting point, timesteps

Post by raback »

Hi,

If you are doing scan over frequency the benefit of good initial guess are next to nothing. The only place where the pseudo time (1,2,3,..) is seen is in the definition of Frequency. Why not have

Code: Select all

$t0=356.0
$df=1.0
Frequency = Variable "time"
   Real MATC "df*(tx+t0)"
and play with the parameters. No need for restart.

Btw, often in frequency scans you want to plot in log-log scale so having a geometric growth might be ideal, e.g.

Code: Select all

$f0=1.0
$q=1.1
Frequency = Variable "time"
   Real MATC "f0*q^(tx-1)"
-Peter
Trexabyte
Posts: 2
Joined: 14 Sep 2022, 11:57
Antispam: Yes

Re: change starting point, timesteps

Post by Trexabyte »

Thank you very much Kevin and Peter!

I first tried Mr. Raback's method. With success! My simulation is now running again. I will mention both suggestions (Kevin and Peter) in my bachelor thesis.
I am writing my bachelor thesis in the field of numerical acoustics at the TU Darmstadt. I am developing a complete workflow for simulating benchmarks for computer acoustics using Elmer.
I have already mentioned Mr. Raback several times in my draft.
Thank you both for the helpful tips. I'll get back to you if I need further assistance.👍

Best regards,
Alex
Post Reply