[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [TRNSYS-users] Problem with radiative Heater (in Heating Type Manager of PREBID)
Hi Florain,
I had a similar problem. I found a trick to avoid the problem. You can use
an input for the radiative part of the heating power (you have to define
this in Prebid).
For example: the name of the input is RADFRAC and the fraction you wish is
0.5.
In your *.dck file you have to add the following statement:
RADFRAC=0.5*GT(TIME,1)
This statement means that if time >1 hour the radiative part will be 0.5. If
you add RADFRAC to the inputs of type 56, you can run your model.
I also asked questions about this to the TRNSYS hotline. It has something to
do with the Trnlib.dll file that you use. You have to compile the Trnlib.dll
file in release mode. I attached my questions and the answers about this
topic to this mail and maybe this will help you as well.
Regards,
Maurice Jong
ECN-DEGO
The Netherlands.
-----Original Message-----
From: Florian Keppler [mailto:flo_usa@gmx.de]
Sent: 13 October 2004 00:31
To: trnsys-users@engr.wisc.edu
Subject: [TRNSYS-users] Problem with radiative Heater (in Heating Type
Manager of PREBID)
Dear Trnsys-User,
hello Solar Energy Lab,
this is Florian from the University of Stuttgart. Right now I'm working at
my diploma thesis using TRNSYS and a problem occured we were not able to
solve. I hope that you might have a solution or an idea how to fix it.
In PREBID I'm using the Heating Type Manager (Ideal Heater) with limited
power to heat a room to a certain temperature (20°C). So far I was only
using convective heat, means the radiative part of the heating power was
equal to zero. In this case everything was fine and i didn't have any
problems running the simulation. However, when i changed the radiative part
of the heating power to any value (0.000001 to 0.99), the simulation stops
at the very first timestep showing an error message that the pane
temperature of the zone window is out of bounds. This happens with any
radiative part of the heating power (except 0), no matter how big the
heating power or the radiative part actually is.
We varied many parameters (heating power, radiative part, window type)
always getting the same error message. It would be great if you had an idea
how to fix that or if you could identify the problem. Thank you very much in
advance!!
I attached the files I used in case you have time to have a look at them
(PREBID files, DCK file, LIB file...).
Thanks again and kind regards,
Florian (Keppler)
--
+++ GMX DSL Premiumtarife 3 Monate gratis* + WLAN-Router 0,- EUR* +++
Clevere DSL-Nutzer wechseln jetzt zu GMX: http://www.gmx.net/de/go/dsl
From: hotline [hotline@transsolar.com]
Sent: 13 November 2003 09:47
To: Jong, M.J.M.
Subject: AW: Bug in TRNSYS ?
Dear Maurice,
normally you want to work with the Release Mode, if you are just interested to get a working and fast trnlib.dll to simulate with TRNSYS only.
If you are writing e.g. your own types and want to check different values in your Fortran Code during the simulation, you need to to build a trnlib.dll in DEBUG Mode. This trnlib.dll is a little bigger than the Release One and hence little slower.
regards,
Matthias
-----Ursprüngliche Nachricht-----
Von: Jong, M.J.M. [mailto:jong@ecn.nl]
Gesendet: Donnerstag, 13. November 2003 09:32
An: hotline
Betreff: RE: Bug in TRNSYS ?
Hi Matthias,
I did not know that there are 2 modes of compiling (debug and release mode) but you were right, I compiled the Trnlib.dll in the debug mode. I changed the settings to the release mode and now I can run my simulation without any problems. The internal radiative gain at t=0 is very, very little (-1.32E-36). This seems to be OK.
What is the difference between the debug and release mode in the compiler ? What is the best way to compile a new Trnlib.dll file ? In the debug or the release mode ?
Regards,
Maurice.
-----Original Message-----
From: hotline [mailto:hotline@transsolar.com]
Sent: 12 November 2003 19:57
To: Jong, M.J.M.
Subject: AW: Bug in TRNSYS ?
Dear Maurice,
what I ment by "Debug" is, wether you build your trnlib.dll with your extended CFV66A.dsw in "Debug" or in "Release"-Mode. If you are not working with the fortran Code while simulating the "Release" Mode of the Compiler settings are the one to use. My guess is that you are working with a trnlib.dll which is created in the "Debug"-Mode. Can you check please?
regards,
Matthias Rudolph
-----Ursprüngliche Nachricht-----
Von: Jong, M.J.M. [mailto:jong@ecn.nl]
Gesendet: Mittwoch, 12. November 2003 09:14
An: hotline
Betreff: RE: Bug in TRNSYS ?
Hello Matthias,
What is a "Debug" trnlib.dll file ? I work with an "own" trnlib.dll file. I added some TRNSYS types we bought in the past with the CVF66A.dsw file in the Fortran compiler. I did not try my model with the original trnlib.dll file. Is this maybe the problem ?
Regards,
Maurice Jong.
-----Original Message-----
From: hotline [mailto:hotline@transsolar.com]
Sent: 10 November 2003 11:27
To: Jong, M.J.M.
Subject: AW: Bug in TRNSYS ?
Hello Maurice,
sorry for the delay....good job!
For the reason of this: first of what i am thinking of is the question wether you are working with a "Debug" trnlib.dll - is this correct?
In case of yes - I guess that the initialization of the radiative gains is not defined, hence there are some billions of kJ/hr as you experienced.
regards,
Matthias Rudolph
Hello Torsten or Matthias,
I found the "solution" for my TRNSYS problem with the error message as shown below.
***** ERROR ***** TRNSYS ERROR
TYPE 56, ZONE 2, SURFACE 17 PANE TEMPERATURE OUT OF BOUNDS AT TIME
1.00000
I discovered that this only appears when you have heating in a zone. If you set the radiative part of the heating to 0, everything is OK. If the radiative part of the heating is > 0 then you have the error message as shown above.
I checked the solar radiaton on the window and the gains in the zone and they were both normal. I also checked the radiative gains in the zone with N-type 13 (Internal radiative gain). On t=0 (start of the simulation), the value of this N-type is 431602000.00 kJ/hr, while I defined the radiavive gain on 0 for t=0. After t=0, the value of the N-type is exactly the same as I defined the radiave internal gain in Prebid. Is this a bug in TRNSYS (version 3.07, 04/11/03) ?
To solve my problem, I defined the heating power as an input in Prebid. In the *.dck file I added the following statement:
Heatingpower=360000*GT(TIME,1)
With this statement, I can run my model without any problems. Although my model runs fine now, I think this is not the normal way to solve this problem.
With kind regards,
Maurice Jong.
P.S. The Transsolar website is still inaccessible from my PC. Do you know this ?
============================================================================
======
Maurice Jong
ECN - Energy research Centre of the Netherlands Department: DEGO
P.O. Box 1 1755 ZG Petten Country: The
Netherlands
Telephone: +31 (0)224 564938 Fax: +31
(0)224 568966
E-mail: jong@ecn.nl Internet:
http://www.ecn.nl/dego
============================================================================
======