[esp-r] Re: Problem with airflow network

[Aizaz Samuel]

Liam,

This type of known pressure node is in fact equivalent to a boundary type 
node. The problem is that you have internal nodes from which you add and 
subtract volume (not mass) when you use a fixed volume flow rate component. 
The solver imposes a mass balance and when temperature changes, the mass of 
air blown by the fixed volume flow rate component changes. So the solver 
blows up. 

Suggestions for your network:
1. Try using fixed mass flow rate components
2. Do not put in more of such components than are absolutely necessary e.g. 
just put in mass flow components to blow air from mixing to north/south, the 
return path from north/south to mixing will have the correct flows due to 
continuity and you can get away by just putting an air flow opening type 
component.

Regards,
Aizaz

On Friday 10 July 2009 14:15, Liam O'Brien wrote:
> Thanks, Jon and Monika.  Does the "known: boundary" type of node not count
> as a boundary node?  I specified its ambient temerpature and pressure as
> 20C and 0 Pa, repsectively.  Both unknown [zonal] nodes have connections to
> this node.
>
> Liam
>
> > From: jon at esru.strath.ac.uk
> > To: obrien_liam at hotmail.com; esp-r at lists.strath.ac.uk
> > Date: Fri, 10 Jul 2009 08:14:27 +0100
> > Subject: [esp-r] Re: Problem with airflow network
> >
> >
> > It is essential that there be at least one connection to a boundary node.
> > The solver really needs this rule to be followed.
> >
> > For more information on defining air flow networks there is a chapter in
> > the current version of the ESP-r Cookbook (ESRU web page -> software
> > downloads -> esp-r -> beginners guide
> >
> > -Jon Hand
> > ________________________________________
> > From: esp-r-bounces at lists.strath.ac.uk [esp-r-bounces at lists.strath.ac.uk]
> > On Behalf Of Liam O'Brien [obrien_liam at hotmail.com] Sent: 09 July 2009
> > 20:24
> > To: esp-r at lists.strath.ac.uk
> > Subject: [esp-r] Problem with airflow network
> >
> > Hello everyone,
> >
> > I'm setting up a fairly rudimentary airflow network, in which two zone
> > air nodes exchange air with a third node that is intended to mix the air
> > and send it back to the nodes. A fixed volume rate component moves the
> > air. The "mixing" node was assigned to known boundary conditions of P=0
> > Pa, T=20C. However, after several tries, I continue to get the following
> > message for each timestep:
> >
> > Please check (in the following order) that:
> > - each node has some flow path to a boundary pressure and
> > - iteration convergence control is not too strict.
> > Day 2 Month 1 Hour 24.000: 100 iterations; flows solution failed
> >
> > I assume the first issue is triggering the message. I would have thought
> > that having the mixing node at known conditions would alleviate this
> > error.
> >
> > Any thoughts? (I realize this could be accomplished in the operations
> > file, but I want something more sophisticated eventually.)
> >
> > The afn file is below.
> >
> > Thanks very much!
> >
> > Liam O'Brien
> >
> > ----------------
> >
> > 3 1 4 1.000 (nodes, components, connections, wind reduction)
> > Node Fld. Type Height Temperature Data_1 Data_2
> > south 1 0 2.5000 20.000 0.0000 250.01
> > north 1 0 2.5000 20.000 0.0000 250.01
> > Mixing 1 2 20.000 0.0000 0.0000 2.0000
> > Component Type C+ L+ Description
> > Circulation 30 2 0 Constant vol. flow rate component m = rho.a
> > 1. 0.200000003
> > +Node dHght -Node dHght via Component
> > south 8.750 Mixing -8.750 Circulation
> > north 8.750 Mixing -8.750 Circulation
> > Mixing -8.750 south 8.750 Circulation
> > Mixing -8.750 north 8.750 Circulation
> >
> > ________________________________
> > Windows Live helps you keep up with all your friends, in one
> > place.<http://go.microsoft.com/?linkid=9660824>
> >
> > _______________________________________________
> > esp-r mailing list
> > esp-r at lists.strath.ac.uk
> > http://lists.strath.ac.uk/mailman/listinfo/esp-r
>
> ________________________________
> Windows Live helps you keep up with all your friends, in one
> place.<http://go.microsoft.com/?linkid=9660824>