-
Filter by Topic
Release
Popular Topics
- List all discussions
LinkedIn
Facebook
Twitter
Delicious
Digg
Stumbleupon
More services...Understanding the time step increment in the transient model of fuel cell
|
Thread index | Previous thread | Next thread | Start a new discussion |
RSS feed |
Turn on email notifications |
1 Reply
Last post: October 4, 2011 6:09pm UTC
October 4, 2011 3:00pm UTC
Understanding the time step increment in the transient model of fuel cell
Dear all
I would like to introduce myself as a Graduate student working on fuel cells. I have made a transient fuel cell model with an sinusoidal voltage function as an input. I run the model at the initial state first in order to getIn the study 3 (transient) , i gave the starting point as 4.8, an interval of 0.25 secs and ending point as 15 seconds through range(4.8,0.25,15) with a tolerance of 0.05 .
The program ran successfully and gave me results. However, in the results , there are a lot of data points instead of the specific points 0.25 seconds apart. I am sending the excel sheet with all the specified data points (the left most column in the data sheet) on which the program is calculating the results. I want to attach the model but it is well beyond the attachment limit.
Can anyone explain me on what basis is the COMSOL software selecting the data points and how is it related to my input.
Regards
Abhishek
Reply | Reply with Quote | Send private message | Report Abuse
October 4, 2011 6:09pm UTC in response to Abhishek Raj
Re: Understanding the time step increment in the transient model of fuel cell
Hi
first to upload a model, you shold clear the mesh and the solutions, then it will be small enough, as in V4 all model steps will regenerate OK
check the sub-nodes of the time stepping solver as there are a few settings therein, the solver will take intermediate steps if required not all are storeed (except if you decide so). Alos with a sinus excitation, use rather a strict or an intermediate stepping scheme rather than the automatic, as the latter might jump over some of the steps (from my understanding it's optimised for diffusion equations typically with exp or log responses )
--
Good luck
Ivar
Reply | Reply with Quote | Send private message | Report Abuse
Rules and guidelines
