[esp-r] Re: CFD results and one minute time step

[Geissler Achim]

Dear Ian,
as already outlined originally in this same context, we do not have the time / capacity to maintain a developer branch (I am aware of the organization). 
A well maintained code is definitely in our interest, of course. However, for changes as the ones done, we found checking results against the unchanged code for a few cases sufficient and then used the code. Generally, we will never be able to really dig into the functionality / model level of the code. So, I see three possibilities here:
	(1) If at any time in the future we are again able to do some small changes and they seem to work and somebody in the list may be able to use them, in future we do not share them
	(2) We share them in the manner I did with the community ... which can decide what to do with them
	(3) ESRU or NRC or whoever creates some kind of "basket" for such minor code changes - to which we could / would forward any such small changes we may be able to do - and they are checked / integrated by somebody who does extensive development, anyway, and has all the procedures for validation and whatnot up and running.

Best regards,
Achim


> -----Original Message-----
> From: esp-r-bounces at lists.strath.ac.uk [mailto:esp-r-
> bounces at lists.strath.ac.uk] On Behalf Of Beausoleil-Morrison, Ian
> Sent: Thursday, February 15, 2007 3:01 PM
> To: ESP-r mailing list
> Subject: [esp-r] Re: CFD results and one minute time step
> 
> Dear Achim and other ESP-r users/developers,
> 
> Last year we created ESP-r Central, a source code repository for ESP-r.
> This was created to allow developers to manage their source code changes
> to ESP-r and to facilitate the contribution of their code into the
> general release of the program.  Please refer to
> http://www.esru.strath.ac.uk/Programs/ESP-r_central.htm for more
> information.
> 
> We can create developer-specific sub-branches in ESP-r for the type of
> code contribution outlined below.  And when this code is ready to be
> merged into the general release of ESP-r, there are a set of procedures
> in place to handle this.
> 
> Regards,
> - Ian
> 
> 
> 
> ________________________________
> 
> 	From: esp-r-bounces at lists.strath.ac.uk
> [mailto:esp-r-bounces at lists.strath.ac.uk] On Behalf Of Geissler Achim
> 	Sent: February 15, 2007 08:50
> 	To: giulio benincasa; esp-r at lists.strath.ac.uk
> 	Subject: [esp-r] Re: CFD results and one minute time step
> 
> 
> 
> 	... some additional information (to my earlier post ... if the
> list moderator approves of it) ...
> 
> 
> 
> 	A short description of some changes I did. All changes but one
> concern CFD, the main ones concern output to TECPlot.
> 
> 	General:
> 
> 	esrudfs/cfcalc.f: two issues, one purely cosmetic, in lines 331
> / 332 we prefer NPLT=MAXIT/100 and MAXIT,100, respectively. For
> calculations spanning more than one time step, the minimum iteration
> number can be lower in our opinion: Line 519 NITER.LT.5 (or even LT.2).
> 
> 
> 
> 	TECPlot:
> 
> 	Attached are changed files esrudfs/cfutil.f and dfs.f and
> esrures/mocfd.f. My changes are headed with "Cag at 241106" so can be
> easily found.
> 
> 	The change in tecplotend() is addition of parameter "Timestep",
> elimination of entry of file name for each time step (the automatically
> generated file name is not nice and should be rewritten to avoid
> whitespace in the filename ...) and, most important, eliminate the extra
> cells that the TecPlot output somehow has. The changes are admittedly
> somewhat 'quick and dirty'. However, if the mesh size is reasonably
> small near interesting surfaces, no relevant information should be lost,
> usually.
> 
> 
> 
> 
> 
> 	________________________________
> 
> 		From: esp-r-bounces at lists.strath.ac.uk
> [mailto:esp-r-bounces at lists.strath.ac.uk] On Behalf Of giulio benincasa
> 	Sent: Thursday, February 15, 2007 1:52 PM
> 	To: esp-r at lists.strath.ac.uk
> 	Subject: [esp-r] CFD results and one minute time step
> 
> 
> 
> 	Hi,
> 	I have make a simulation with conflation  method  of double skin
> facade mechanically ventilated;
> 
> 	the time period is ten hours and the time step is one minute;
> 
> 	the results seem to be satisfy (..for me ) but I have a notice a
> unexpected difficulty to store CFD results:
> 
> 	the folder CFG of my project store at most automatically 100
> graphs  then ask me "if I want to overwrite on foregoing graph" ;
> 
> 	-- if I dont overwrite then I must write another 500 name of
> graph.. multiply for the number  interest variable ...multiply for vary
> grid position ....it become hard working
> 
> 	--if I store the results from one hour to one hour  ( I wish
> convey really this):
> 	  I have notice that lose one's way the simulation results ( all
> same graph) except in the first 100 minutes;
> 
> 	There is anybody that has notice the same difficulty  ? And
> there is any solution ?
> 
> 	Thank you very munch in advance,
> 	Best regards Giulio
> 
> 
> 
> 
> 
> 
> _______________________________________________
> esp-r mailing list
> esp-r at lists.strath.ac.uk
> http://lists.strath.ac.uk/mailman/listinfo/esp-r