atwhaley / cfast

Automatically exported from code.google.com/p/cfast
0 stars 2 forks source link

Stop, change parameter(s), continue calculation (restart from stop point) #62

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
I have a need to start and stop a CFAST run, adjust parameters (e.g. vent 
open/close, sprinkler on/off, etc.), then have CFAST continue the calculation 
from where it stopped using the new parameter(s). Basically, have CFAST re-read 
all the input files but start calculating with the last fire, heat, smoke, etc 
values from when it stopped and then use the new input parameters as it 
continues.

This doesn't appear to be a supported function in CFAST 6.1 (unless I missed 
it). I was wondering if anyone might be interested in working on a branch to 
enable this functionality in some fashion (e.g. command line, etc). I'm happy 
to pay a bounty if anyone were willing to take on the effort.

Thanks in advance for any feedback.

Cheers,
Bill

Original issue reported on code.google.com by bill.god...@c3pathways.com on 4 Apr 2011 at 11:33

GoogleCodeExporter commented 9 years ago
The speed of CFAST is such that it is easier to make complete runs. I can send 
you a program for making variants on a base case. For a parameter study I did a 
few years ago, I generated 20,000 permutations. The total run time of the model 
was only a few hours. One of the benefits is that I did not have to keep track 
of linked files.

Having said that, I will take a look to see whether it is feasible to resurrect 
the restart capability. 

The former is the low cost option. 

Original comment by dcc...@gmail.com on 13 Jun 2011 at 4:00

GoogleCodeExporter commented 9 years ago
What is the status of this issue?

Original comment by mcgra...@gmail.com on 19 Jul 2013 at 1:38

GoogleCodeExporter commented 9 years ago
I still have the need, and I am still interested in the described solution. My 
workaround has limitations, which can impact the overall model accuracy. The 
described method is a better solution for my need. Is there some interest in 
tackling the problem? 

Thanks for following up.

Cheers,
Bill

Original comment by bill.god...@c3pathways.com on 19 Jul 2013 at 3:37