|
[Sponsors] |
June 16, 2016, 06:00 |
Approx. Sim time taken by the Computer
|
#1 |
New Member
Saumik Kumar Dey
Join Date: Jun 2016
Location: Howrah, West Bengal, India
Posts: 8
Rep Power: 9 |
Hi Everyone !!
Can anyone tell me, if there is a way to estimate the approx. total simulation time (actual time taken by the computer NOT the flow time in transient case) in advance before running the simulation ?? Actually it will be helpful if I could know the expected time so that I can schedule my simulation accordingly. Thanks. |
|
June 16, 2016, 07:49 |
|
#2 |
Senior Member
Join Date: Mar 2015
Posts: 892
Rep Power: 18 |
Run your simulation for a few time steps and then predict how long the complete simulation would take. For example, say, run for 10 time steps out of 1,000 time steps. Then multiply the time taken by 100 to estimate the overall time. The CPU time can be shown via Parallel > System > CPU info.
Note: the term wall clock time (WCT) is used to describe the time taken by the computer (regardless of number of processes and simulation time). |
|
June 17, 2016, 01:14 |
|
#3 |
New Member
Saumik Kumar Dey
Join Date: Jun 2016
Location: Howrah, West Bengal, India
Posts: 8
Rep Power: 9 |
Actually I was thinking, a udf could be developed to calculate the current computer time since the sim has started and print that after every time step in the console and also the estimated time of completion as you have described above.
But I am not good at this udf programming, I have just started learning, if you kindly help me with that, it will be great. Thanks, |
|
June 17, 2016, 05:19 |
|
#4 |
Senior Member
Join Date: Mar 2015
Posts: 892
Rep Power: 18 |
You could use the clock function to calculate the CPU time (save the start time as a global variable and only initialise on the first time step). Use the DEFINE_EXECUTE_AT_END macro for calculating the current WCT at the end of each time step and print the estimate to the screen with the Message function. Note: you'll need to enter the total number of time steps either within the UDF or use rp-variables (Scheme coding).
The method I mentioned above is recommended for estimating the total WCT for your simulations but if you're keen on programming this UDF then hopefully the above instructions along with the UDF manual is useful. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[Other] Contribution a new utility: refine wall layer mesh based on yPlus field | lakeat | OpenFOAM Community Contributions | 58 | December 23, 2021 02:36 |
simpleFoam error - "Floating point exception" | mbcx4jc2 | OpenFOAM Running, Solving & CFD | 12 | August 4, 2015 02:20 |
AMI interDyMFoam for mixer nu problem | danny123 | OpenFOAM Programming & Development | 8 | September 6, 2013 02:34 |
pisoFoam with k-epsilon turb blows up - Some questions | Heroic | OpenFOAM Running, Solving & CFD | 26 | December 17, 2012 03:34 |
Orifice Plate with a fully developed flow - Problems with convergence | jonmec | OpenFOAM Running, Solving & CFD | 3 | July 28, 2011 05:24 |