[Getdp] Re-2: ... two disjoined bodies with radiation interaction ...
Matt Koch
mattkoch at alum.mit.edu
Tue Apr 11 15:26:26 CEST 2006
Hi Matthias,
attached are my test files for the problem we discussed. I have the
habit of breaking up the problem into its logical units, so that's why
there are four files (and often five, if I define non-linear materials
properties). The problem seems to calculate as is, but it does not seem
to converge. This configuration is at the end of a long line of attempts
messing with FunctionSpace, Formulation and Resolution - it still is not
the final answer. Enjoy perusing them, and if you figure out where I am
wrong, all the better.
The problem describes two disks, and the "long-distance" interaction I
am looking for is between the bottom of the top disk and the top of the
bottom disk. The top of the top disk has a heat flux imposed to
introduce some heat into the problem. That heat is to spread from the
top disk to the bottom disk across the "Interface". In earlier
configurations (don't have them around any more, but it was based on two
FunctionSpaces "TmpA" and "TmpB", rather than one "Tmp"), I achieved
convergence, but I only ever got non-zero results for the top plate, and
always zeros for the bottom plate, i.e. it did not even seem to get
initialized to 293.15 K properly.
I am at my wit's end, so whatever you can come up with might actually
break the deadlock. Ideally, I'd have each disk described by a separate
FunctionSpace TmpA (bottom disk) and TmpB (top disk) - that seems to
make the most sense to me, but then again I may not understand the true
meaning of FunctionSpace.
Regards and Thanks for looking into this,
Matt
m.fenner at omicron.de wrote:
> Dear Matt,
>
> oops, it was pure nelgigence not to respond in public. It's not fair and definitely not wise to open a side channel for communication.
>
> Sure, I am still interested. I have not been successful with integral quantities, as well. I vaguely recall trying to calculate the total heat flux through a rod with fixed tepmeratures at the ends. But I could not get it neither from global nor from integral quantities. Your problem seems to be related or even similar. If you like to share your files I would be glad to keep the discussion going. But unfortunately I have no solution ready.
>
>
> Best regards
>
>
>
> Matthias A. Fenner
>
>
> Omicron
> NanoTechnology GmbH
>
> Limburger Straße 75
> D-65232 Taunusstein
>
> +49 6128 987-277 Phone
> +49 6128 987-33277 Fax
> +49 6128 987-32277 Voice
>
> eMail m.fenner at omicron.de
> WWW.omicron.de
>
>
>
> -------- Original Message --------
> Subject: Re: [Getdp] ... two disjoined bodies with radiation interaction ... (10-Apr-2006 15:49)
> From: mkoch at gvtc.com
> To: m.fenner at omicron.de
>
>
>> Hi Matthias,
>>
>> thank you kindly for your response. It is encouraging to see interest in
>> this
>> topic. I noticed that your response did not copy getdp at geuz.org - was that
>> accidental or intentional? I would think keeping the group in the loop
>> enhances
>> chances for success. The e-mail traffic might also catch Christophe's
>> attention,
>> and his input is probably critical, I think, in these early stages?
>>
>> Anyway, to answer some of your questions, I have dealt with radiation
>> exchange
>> problems extensively in the past, and the configuration factor approach in
>> my
>> view is preferable over the ray tracing approach - if the problem is
>> sufficiently limited in scope. I am not at all concerned about this aspect
>> right now, and I am also not envisioning a grand scheme that automatically
>> considers these factors in all possible geometries.
>>
>> I am perfectly happy to hand-code the configuration factors for now. My
>> problems
>> are usually relatively confined, so this would work just fine for me.
>>
>> My true challenge has to do with GetDP and its treatment of quantities and
>> equations and such. Right now, for example, I am not even trying to do
>> configuration factors. I am just trying to have one surface heat another
>> via
>> their average temperatures (using "Integral" quantities) and a fake heat
>> transfer coefficient between them, and I can't even get that to work.
>>
>> If you are still interested, I can send you what I have thus far so you
>> can more
>> easily understand where I am stuck at. Once we cross that hurdle, I think
>> configuration factors and integrals would be the next step.
>>
>> I like your comment about setting this up as a network problem. But
>> like you, I
>> have no clue as to how to go about this. Plus, I wonder if network
>> constraints
>> are 0-dimensional, i.e. can only connect points, not surfaces?
>>
>> If you agree, I would also like to forward your response and this e-mail to
>> getdp at geuz.org.
>>
>> Thanks and Regards,
>>
>> Matt Koch
>> mattkoch at scitex.us
>> 978-726-4202
>>
>>
>> ----- Message from m.fenner at omicron.de ---------
>> Date: 10 Apr 2006 08:25:18 UT
>> From: m.fenner at omicron.de
>> Reply-To: m.fenner at omicron.de
>> Subject: Re: [Getdp] ... two disjoined bodies with radiation interaction ...
>> .
>> To: mkoch at gvtc.com
>>
>>
>>
>>> Dear Matt,
>>>
>>> since I sometimes have to deal with heat transfer problems I'd really
>>> be interested in a solution for the radiation problem. I have no idea
>>> how to attack this in a general framework in getDP just using the
>>> geometry and formulation. It sounds like a ray tracing problem.
>>>
>>> An approach from the analytical side is to calculate configuration
>>> factors between the two surfaces which are tabulated for a few
>>> spacial cases. I wonder if one could use the network constraints to
>>> couple two surfaces via configuration factors. I have not played with
>>> network constraints yet so I am not sure this is viable at all. And
>>> then it is not quite what we want since it implies special knowledge
>>> and assumptions on the geometry, homogeneous temperature of the
>>> surface for instance. Just placing the surfaces in the .geo file and
>>> solve a radiation problem would be nicer.
>>>
>>> Have you made any progress in this direction?
>>>
>>>
>>> Best regards
>>>
>>> Matthias
>>>
>>>
>>>
>>> Matthias A. Fenner
>>>
>>>
>>> Omicron
>>> NanoTechnology GmbH
>>>
>>> Limburger Straße 75
>>> D-65232 Taunusstein
>>>
>>> +49 6128 987-277 Phone
>>> +49 6128 987-33277 Fax
>>> +49 6128 987-32277 Voice
>>>
>>> eMail m.fenner at omicron.de
>>> WWW.omicron.de
>>>
>>>
>>> -------- Original Message --------
>>> Subject: [Getdp] ... two disjoined bodies with radiation interaction
>>> ... (04-Apr-2006 16:32)
>>> From: mkoch at gvtc.com
>>> To: m.fenner at omicron.de
>>>
>>>
>>>> Hello All,
>>>>
>>>> I am still trying to figure out how to formulate a "long-range" problem.
>>>>
>>>> I have two disks that are not connected, arranged much like a capacitor.
>>>>
>>>> The top disk is being heated, e.g. by a volumetirc heat source, the
>>>> bottom disk
>>>> is not. I want the top disk to transfer some of its heat to the bottom
>>>>
>> disk
>>
>>>> by
>>>> radiation.
>>>>
>>>> The typical radiation setup in the examples radiates into free space,
>>>> but not to
>>>> another body, whose temperature can also change. A body to body
>>>>
>> radiation
>>
>>>> problem instead involves configuration or view factors, as well as
>>>> temperature
>>>> integrals - the details of which are somewhat secondary until I get the
>>>> overall
>>>> structure of the problem right.
>>>>
>>>> If there is anyone out there who can help me set this up, I would much
>>>> appreciate it. If there is someone, I can share with them what I have so
>>>> far,
>>>> which will also make the problem more clear. My goal is to solve this
>>>> problem,
>>>> of course, but I think it could also make for an intersting addition to
>>>>
>> the
>>
>>>> examples.
>>>>
>>>> Regards,
>>>>
>>>> Matt Koch
>>>> mattkoch at scitex.us
>>>> 978-726-4202
>>>>
>>>>
>>>> ----------------------------------------------------------------
>>>> This message was sent using GVTC's Webmail System.
>>>>
>>>>
>>>> _______________________________________________
>>>> getdp mailing list
>>>> getdp at geuz.org
>>>> http://www.geuz.org/mailman/listinfo/getdp
>>>>
>>>>
>>>>
>>>> To: getdp at geuz.org
>>>>
>>>
>> ----- End message from m.fenner at omicron.de -----
>>
>>
>>
>> ----------------------------------------------------------------
>> This message was sent using GVTC's Webmail System.
>>
>>
>>
>>
>> To: m.fenner at omicron.de
>>
>
>
> To: mkoch at gvtc.com
> Cc: getDP at geuz.org
>
>
> _______________________________________________
> getdp mailing list
> getdp at geuz.org
> http://www.geuz.org/mailman/listinfo/getdp
>
>
>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Test-2D.sol
URL: <http://www.geuz.org/pipermail/getdp/attachments/20060411/c99d77b9/attachment.sol>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Test-2D.par
URL: <http://www.geuz.org/pipermail/getdp/attachments/20060411/c99d77b9/attachment.par>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Test-2D.pro
URL: <http://www.geuz.org/pipermail/getdp/attachments/20060411/c99d77b9/attachment.pro>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Test-2D.geo
URL: <http://www.geuz.org/pipermail/getdp/attachments/20060411/c99d77b9/attachment.geo>