[esp-r] Re: Time steps and embedded heating cooling within a construction.

[Sébastien Brideau]

Hi,

Further to what has been said already:

We have found that the radiant_floor model can have convergence issues at
very small time steps (less than 2 minutes).  This is using a concrete slab
though, and we haven't tried it with a very thin (low thermal capacity)
construction.  I have not look into the reasons why convergence issues are
happening at small time steps, but I do not see the point of modelling at
anything less than a few minutes interval for most cases anyways, so it has
worked well for us. 


Here is a paper on the development of the radiant slab model in ESP-r.
http://archive.nrc-cnrc.gc.ca/obj/irc/doc/pubs/nrcc46099/nrcc46099.pdf

Regards,

Sébastien Brideau

-----Original Message-----
From: esp-r-bounces at lists.strath.ac.uk
[mailto:esp-r-bounces at lists.strath.ac.uk] On Behalf Of Geissler Achim
Sent: December-06-12 2:50 AM
To: Jon Hand; wakana-m at ec.hokudai.ac.jp; esp-r at lists.strath.ac.uk
Subject: [esp-r] Re: Time steps and embedded heating cooling within a
construction.

Hi Wakana Matsumoto

it is not quite clear to me, what you mean by "now the convection
coefficient is equivalent to that of water". Does your model have an
explicit zone for the cooling panel and you are using an air flow network as
"heat sink"? If yes, why not simply control the zone without air flow
network? If you are referring to the cooling-panel-to-zone-air surface
convection coefficient, then this approach is not realistic, for sure.

Are you using a plant network? From the sound of it, I would guess "no". I
will send you a simple model (by PM) featuring the "radiant_floor" component
which I think Jon was referring to. This model features two simple office
zones with cooled ceiling components of type "radiant_floor" in a plant
network. I have not tested this with "very thin" constructions, so can say
nothing about the numerical stability for such a use. But the model may give
some suggestions on a possible approach.

Best
Achim


> -----Ursprüngliche Nachricht-----
> Von: esp-r-bounces at lists.strath.ac.uk [mailto:esp-r- 
> bounces at lists.strath.ac.uk] Im Auftrag von Jon Hand
> Gesendet: Donnerstag, 6. Dezember 2012 08:10
> An: wakana-m at ec.hokudai.ac.jp; esp-r at lists.strath.ac.uk
> Betreff: [esp-r] Re: Time steps and embedded heating cooling within a 
> construction.
> 
> 
> There has been recent work on a system component and control which 
> allows for pipes to be embedded in a construction layer - this is 
> normally used for heating - it might also work for a chilled fluid.  I 
> believe there are a few ESP-r users who have explored this.  I would 
> certainly like to arrange for a good example model for structural slab 
> heating as well as for thin radiant panels.
> 
> Perhaps someone in the community could offer some suggestions?  Can 
> the embedded pipe work with a thin highly conductive construction?
> 
> If we can sort a system component then short timestep for the building 
> would not be required.
> 
> Enabling a sub-minute timestep would require changes in the source 
> code and data structures within the results analysis application.  Also
testing.
> 
> -Jon Hand
> ________________________________________
> From: wakana-m at ec.hokudai.ac.jp [wakana-m at ec.hokudai.ac.jp]
> Sent: 06 December 2012 02:27
> To: Jon Hand; esp-r at lists.strath.ac.uk
> Subject: Re: [esp-r] Re: Time steps
> 
> Dear Jon Hand,
> 
> Thank you always for your help.
> 
> I study about ceiling radiation cooling system.
> My model includes a radiation panel which is very thin and the panel 
> is controlled.
> I'd like to flow water in the panel zone.
> I couldn't make water flow network, then I changed the convection 
> coefficient.
> Now the convection coefficient of panel is equivalent of water.
> The results aren't sometimes calculated.
> I think the reason of the error is that the calculation diverged.
> Therefore I tried to shorten time-steps.
> If there are other good ways, please let me know.
> 
> If I try to simulate on a second basis, what should I change?
> In simulator, there is "time-step controllers".
> Is this related to time-steps?
> 
> Best Regards,
> 
> Wakana Matsumoto
> 
> ----- Original Message -----
> >
> > The building part of the simulation is currently limited to a one
> minute frequency.  If
> > the model includes system components the system side can be assessed
> at  a higher
> > frequency up to ~1 second.  We normally assume that the building is
> not so sensitive
> > to time below a minute but the system components and electrical
> components might
> > change state more frequently.
> >
> > What (object) are you attempting to simulate where sub-minute
> timesteps are needed?
> >
> > -Jon Hand
> > ________________________________________
> > From: esp-r-bounces at lists.strath.ac.uk [esp-r-bounces at lists.strath.ac.
> uk] On Behalf Of wakana-m at ec.hokudai.ac.jp [wakana- 
> m at ec.hokudai.ac.jp]
> > Sent: 05 December 2012 09:11
> > To: esp-r at lists.strath.ac.uk
> > Subject: [esp-r]  Time steps
> >
> > Dear all,
> >
> >  Hello.
> >  I'm a graduate student in Japan.
> >  I want to know how to simulate on a second basis.
> >
> >  Until now, I've simulated on a minute basis.
> >  I've input "60/1" in the initiate simulation window.
> >  Now I try to simulate on a second basis.
> >  However the input value for 'time-steps/hour' should be less than 60.
> >  Thus, how can I do if I want to simulate the model by seconds ?
> >
> >  Any advice would be appreciated.
> >
> >  Regards,
> >
> >  Wakana Matsumoto
> >
> >
> >
> > _______________________________________________
> > esp-r mailing list
> > esp-r at lists.strath.ac.uk
> > http://lists.strath.ac.uk/mailman/listinfo/esp-r
> >
> > _______________________________________________
> > esp-r mailing list
> > esp-r at lists.strath.ac.uk
> > http://lists.strath.ac.uk/mailman/listinfo/esp-r
> >
> 
> 
> 
> _______________________________________________
> esp-r mailing list
> esp-r at lists.strath.ac.uk
> http://lists.strath.ac.uk/mailman/listinfo/esp-r

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