[esp-r] Re: Convection coefficients problem

[Jon Hand]

This shows the process of debug models.  An indication that Fisher correlation might
be the cause.  This needs to be confirmed that conditions in the facade are suitable
for this correlation (velocity might be high??) and it might be that there is another
correlation that could be used.  There is also an option for the user to define hc
coefficient manually - e.g. if you know the velocity is ~1.5m/s then set a schedule
of hc coefficients that reflect that velocity within the zone hc defintion.

Certainly a short timestep - perhaps 5 minutes down to 1 minute for the building & 
flow solution. Would be more appropriate for this model.

-Jon Hand
________________________________________
From: Michael Lucchi [michael.lucchi at gmail.com]
Sent: 14 November 2013 09:34
To: Jon Hand; esp-r at lists.strath.ac.uk
Subject: Re: [esp-r] Convection coefficients problem

Thank you Jon for the precious suggestions.

The facade is connected with a HVAC system and it is a force ventilated facade. However, I also connected the three zone nodes corresponding to the ventilated facade with a flow network in order to simulate the behaviour of the facade when the plant turns off (in this period the mass flow between the three zones is due to only buoyance effects ). Generally, the simulation stops after few time-steps.  The only succeded simulation was set up with both  zones and plants time-steps of 30 minutes, but with wrong results.
Turning off the Fisher correlation and using default correlation,  the model runs.

Thank you for eventual helps.


2013/11/12 Jon Hand <jon at esru.strath.ac.uk<mailto:jon at esru.strath.ac.uk>>

The first thing that caught my attention is [ventilated facade] and [forced ventilation].  What
kind of ventilated facade is this - a rain-screen (in which case there would be low velocity
flow).  Is the facade actually force ventilated or just ventilated by differences in temperature
and thus one might describe it in via a flow network?

mzsad is in the matrix solver.  There are a few folk who might be able to address that
kind of issue.  It might be that the Fisher is causing this or perhaps not.  I would first
explore the use of the [trace] facility in bps - it can write out lots of additional information
about the heat transfer correlations.

I would suggest that you turn off the Fisher in the 3 ventilated facade zones and see
if the model runs.

And also - what is happening at 93 time-steps. (Hour 10 on year day 189)?  Does
the error happen if you run the assessment for a different day?

-Jon
________________________________________
From: esp-r-bounces at lists.strath.ac.uk<mailto:esp-r-bounces at lists.strath.ac.uk> [esp-r-bounces at lists.strath.ac.uk<mailto:esp-r-bounces at lists.strath.ac.uk>] on behalf of Michael Lucchi [michael.lucchi at gmail.com<mailto:michael.lucchi at gmail.com>]
Sent: 12 November 2013 15:43
To: esp-r at lists.strath.ac.uk<mailto:esp-r at lists.strath.ac.uk>
Subject: [esp-r]  Convection coefficients problem

Hello ESP-r users,
I’m trying to model a ventilated façade characterized by a forced convection regime using Esp-r (version 12.0 for Windows 7) for my thesis. The model is composed by one single room, and the external facades are ventilated opaque walls. The ventilated façade is composed by three thermal zones, with the aim to obtain the vertical temperature distribution. In order to reach reasonable results, I tried to impose the Fisher and Pedersen correlation for mechanically driven flows both in the room and in the ventilated façades. However, when I run the simulation, a numerical error stops any module. The following text is the error displayed by the simulator:
mzsad 1: numerical error after 93 time-steps.
(Hour 10 on year day 189).
e(iconst,n)    = -17999998.000000
a1      = -18000000.0000
I am still  a novice user of the software and any help is much appreciated.
Kind regards.
Michael Lucchi