[esp-r] Re: CFD - bouyancy modelling has no effect?

[Aizaz Samuel]

Andrew,

I ran the CFD exemplar with and without buoyancy and found differences in the results.

The dfd file used is room.dfd available within the air movement exemplar set. The module tested is dfs.

I first ran the dfd file without modification (no buoyancy) and then extracted the monitored cell information (*mon file) from within dfs by choosing to not delete temporary files when dfs exits. 

I then changed the buoyancy line in the dfd file from "0" to "1 0.5" (1 sets the buoyancy calculation flag and 0.5 is the default under-relaxation) Then I ran the model through dfs again and extracted the mon file.

The two mon files are different which implies the overall solution is different. Furthermore because dfs supplies results to bps during all conflation methods, bps based simulations will be different.

Regards,
Aizaz
________________________________________
From: esp-r-bounces at lists.strath.ac.uk [esp-r-bounces at lists.strath.ac.uk] on behalf of Andrew Cowie [cn06arc at leeds.ac.uk]
Sent: 26 March 2014 15:53
To: esp-r at lists.strath.ac.uk
Subject: [esp-r]  CFD - bouyancy modelling has no effect?

Hi all,

I've been running some simulations with the CFD in ESP-r, and I've noticed something that I consider a little odd.  Simulations run with bouyancy modelling (full ideal gas law calcs) and local heat sources in the CFD domain (ie. casual gains assigned to source-type cells) give results which are identical to simulations run without these elements (ie. no bouyancy modelling at all and complete removal of the source-type cells).  When I say identical, I mean that in every sense of the word; the unix command "diff" gives no output when comparing the CFD results output as ParaView files.

I am curious as to why this may be; I would expect to see at least minor numerical differences in the CFD results with these features.  Are these features automatically encapsulated within the adaptive conflation algorithms, overriding the CFD parameter file?  Are they no longer supported?  Do we know for a fact that these input parameters actually have any affect on the CFD calculations?  If anyone can shed any light on this I'd be greatly appreciative.

Cheers,
Andy Cowie
_______________________________________________
esp-r mailing list
esp-r at lists.strath.ac.uk
http://lists.strath.ac.uk/mailman/listinfo/esp-r