Subject: RE: standard packages for 1076.1
From: Tiller, Michael (M.M.) (mtiller@ford.com)
Date: Fri Oct 12 2001 - 05:12:49 PDT
While I would agree that one can always pose the problem such that the product of the through and across variables has units of power, I see many practical examples of why one would not do this. In order to adhere to the convention, your heat transfer examples reference a series of "entropic" properties like "entropic specific heat" and "entropic capacitance". To be frank, I've been working in heat transfer and thermodynamics for 10 years and I've never heard of these properties. I don't claim to be an expert, so I decided to look into this a little deeper. A search of Yahoo! turns up a total of 3 hits (two of which are on your pages). As a result, I would argue with your contention that this issue was "solved a long time ago". In my opinion such a formulation is not only more complicated, it is less intuitive.
The standard that we adhere to in our modeling is that the through variable be the first derivative, with respect to time, of some conserved quantity. For hardware description languages this is a very natural approach since the problem formulation will result in all of the through variables being summed to zero at each junction. Trying to adhere to the "power convention" seems silly to me when it isn't really necessary for any practical reason (at least in VHDL-AMS, Modelica, MAST, etc...).
As far as I can tell, there would be nothing incorrect about using the conserved quantity itself (as opposed to the first derivative with respect to time) as the through variable, which is what the magnetic discussion is about if I understand correctly. It seems a tad inconvenient to me but it seems to still obey the semantics of what a through variable should be.
-- Mike> -----Original Message----- > From: Herman Mann [mailto:mann@vc.cvut.cz] > Sent: Friday, October 12, 2001 6:41 AM > To: vhdl-ams@eda.org > Subject: Re: standard packages for 1076.1 > > > > Ladies and Gentlemen, > > I thought that the problem of association of through and across > variables with physical quantities has been solved a long time ago. > > We give our students a course where the variables are defined in > the following way: > http://virtual.cvut.cz/course/node67.html > This system of variables is consistent, i.e. the product of variables > in each through-across-variable pair corresponds to power. > > The consistency becomes important when multi-domain systems > are considered. E.g., in thermodynamic systems not only heat > transfer, but also fluid flow must be considered at the same time. > > Basics on modeling magnetic phenomena are on > http://virtual.cvut.cz/course/node42.html > > whereas the thermal phenomena are on > http://virtual.cvut.cz/course/node60.html > > Some examples of electro-magneto-mechanical systems that can > be simulated interactively across the Internet are on > http://virtual.cvut.cz/dyn/examples/ > > With many apologies for entering your discussion, > > Herman Mann > > > > ---------------------------------------------------------------- > Herman MANN, DrSc., Associate Professor > Computing Centre, CZECH TECHNICAL UNIVERSITY > Zikova 4, CZ-166 35 Prague 6, Czech Republic > Tel:+420-2-2431-0349 Mobil:+420-(0)605-586485 > Fax: +420-2-2431-0271 Home:tel/fax:+420-2-3333-7904 > Personal WWW: http://virtual.cvut.cz/mann/ > Projects: http://virtual.cvut.cz/course/ > http://www-er.df.op.dlr.de/cacsd/ > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >
This archive was generated by hypermail 2b28 : Fri Oct 12 2001 - 05:31:21 PDT