Transient Modeling in Flotherm

bighead · April 14, 2009, 21:25

Hey guys, I have some queries regarding transient modeling in Flotherm. I have two objects that are placed together in the domain. However, both are at different initial conditions,ie Block 1 is at 100°C and 2 is at 25°C. The time scales for both model are of same order of magnitude, hence I cannot ignore one or the other.

Question
1) Is it even possible to model the transient heat transfer interaction ? If so, how can I do it ?

RobinB · April 16, 2009, 09:00

Certainly doable in FloTHERM.

First of all make sure that you have attached material attributes to both blocks with the correct thermal conductivity, density and specific heat values. The latter two are particularly important for a transient simulation.

Next use Initial Subdomain objects collocated with the two cuboids to define the initial temperatures.

Set up the other aspects of a transient simulation in the usual way (e.g. end time, transient grid etc.)

Then press 'Go' !

bighead · April 16, 2009, 10:02

Thxs RobinB, I will give it a try. It didnt cross my mind to use the subdomain. Just a question. How will the solver treat the interface when the two domain meet ??

RobinB · April 16, 2009, 11:09

The subdomains simply define initial conditions for the transient simulation. At the end of the first time step the temperature field in the two cuboids will have changed based on their thermal conductivity, thermal mass and surrounding air side HTCs.

bighead · April 16, 2009, 22:56

I had tried using the initial subdomain and the results are definitely much better. Also, after I initialise the solution and plot the temperature contours for both bodies, I noticed a temp gradient across a very thin layer at the interface location. Probably guess that is to ensure no temperature jump at interfaces. But thxs anyway RobinB for the help.

RobinB · April 17, 2009, 02:43

The temp gradient of the initial fields at the interface will be due to how the cell temperature scalar values are interpolated (smoothed). To see the true real values graphically select the plane plot and change the 'Fill Type' from 'Interpolated' to 'Cell Fill'

bighead · April 23, 2009, 05:33

Thxs RobinB, everything works fine now.

bighead · May 3, 2009, 23:27

Hey Robin, just a short question for you. The initial conditions for sub domain works fine. The main domain is at 90°C while the sub is at 25°C. However, at the beginning of the first time step. depending on the model (TIM thickness), I am getting large temperature changes (90 to 83°C in 0.2s). I had tried to use smaller timesteps but still the same time happen. Do you have any idea how to resolve this ?

RobinB · May 5, 2009, 06:15

The initial conditions are just that, the conditions at t=0s for a transient simulation. Once the transient simulation starts heat will start to flow from hot to cool blocks, very rapidly at first where the dT is high. As t->infinity both blocks will assume the same temperature.

In addition to this model I think you will need to set either a fixed T or fixed Q (heat source) on the side where the package sits.

I'm more than happy fielding this support issue here but really you would get as good (if not better!!) service from your local support office.

naveen kannan · May 21, 2013, 12:31

Hi,

Is there any tech note or instructions manual on how to perform transient analysis in Flotherm/Flovent?

RobinB · May 21, 2013, 12:37

Yep, the Tech Note MG245418 on SupportNet outlines a typical transient model definition process.

naveen kannan · May 21, 2013, 12:47

Thanks Robin. I'll take a look at that. I guess my keywords didn't work on support net. Will get back if I have any questions.

April 14, 2009, 21:25	Transient Modeling in Flotherm	#1
bighead New Member Melvin. Join Date: Mar 2009 Location: Singapore Posts: 27 Rep Power: 17	Hey guys, I have some queries regarding transient modeling in Flotherm. I have two objects that are placed together in the domain. However, both are at different initial conditions,ie Block 1 is at 100°C and 2 is at 25°C. The time scales for both model are of same order of magnitude, hence I cannot ignore one or the other. Question 1) Is it even possible to model the transient heat transfer interaction ? If so, how can I do it ?

May 21, 2013, 12:31	Transient analysis	#10
naveen kannan New Member Naveen Join Date: May 2013 Posts: 2 Rep Power: 0	Hi, Is there any tech note or instructions manual on how to perform transient analysis in Flotherm/Flovent?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
transient modeling	venkatesh4386@gmail.com	CFX	1	August 8, 2008 07:46
Transient vs Steady State	Adam	CFX	1	April 12, 2007 11:34
Modeling Flow/Saturation/Absorption in Fibers	Gene Dougherty	Main CFD Forum	0	June 6, 2003 14:49
Turbulence model for Transient Pipe Flow Modeling	Brian	FLUENT	0	June 7, 2002 12:15
difference between false and true transient	mahesh prakash	Main CFD Forum	1	January 21, 1999 13:45

April 16, 2009, 09:00		#2
RobinB New Member Robin Bornoff Join Date: Mar 2009 Posts: 23 Rep Power: 17	Certainly doable in FloTHERM. First of all make sure that you have attached material attributes to both blocks with the correct thermal conductivity, density and specific heat values. The latter two are particularly important for a transient simulation. Next use Initial Subdomain objects collocated with the two cuboids to define the initial temperatures. Set up the other aspects of a transient simulation in the usual way (e.g. end time, transient grid etc.) Then press 'Go' !

April 16, 2009, 10:02		#3
bighead New Member Melvin. Join Date: Mar 2009 Location: Singapore Posts: 27 Rep Power: 17	Thxs RobinB, I will give it a try. It didnt cross my mind to use the subdomain. Just a question. How will the solver treat the interface when the two domain meet ??

April 16, 2009, 11:09		#4
RobinB New Member Robin Bornoff Join Date: Mar 2009 Posts: 23 Rep Power: 17	The subdomains simply define initial conditions for the transient simulation. At the end of the first time step the temperature field in the two cuboids will have changed based on their thermal conductivity, thermal mass and surrounding air side HTCs.

April 16, 2009, 22:56		#5
bighead New Member Melvin. Join Date: Mar 2009 Location: Singapore Posts: 27 Rep Power: 17	I had tried using the initial subdomain and the results are definitely much better. Also, after I initialise the solution and plot the temperature contours for both bodies, I noticed a temp gradient across a very thin layer at the interface location. Probably guess that is to ensure no temperature jump at interfaces. But thxs anyway RobinB for the help.

April 17, 2009, 02:43		#6
RobinB New Member Robin Bornoff Join Date: Mar 2009 Posts: 23 Rep Power: 17	The temp gradient of the initial fields at the interface will be due to how the cell temperature scalar values are interpolated (smoothed). To see the true real values graphically select the plane plot and change the 'Fill Type' from 'Interpolated' to 'Cell Fill'

April 23, 2009, 05:33		#7
bighead New Member Melvin. Join Date: Mar 2009 Location: Singapore Posts: 27 Rep Power: 17	Thxs RobinB, everything works fine now.

May 3, 2009, 23:27		#8
bighead New Member Melvin. Join Date: Mar 2009 Location: Singapore Posts: 27 Rep Power: 17	Hey Robin, just a short question for you. The initial conditions for sub domain works fine. The main domain is at 90°C while the sub is at 25°C. However, at the beginning of the first time step. depending on the model (TIM thickness), I am getting large temperature changes (90 to 83°C in 0.2s). I had tried to use smaller timesteps but still the same time happen. Do you have any idea how to resolve this ?

May 5, 2009, 06:15		#9
RobinB New Member Robin Bornoff Join Date: Mar 2009 Posts: 23 Rep Power: 17	The initial conditions are just that, the conditions at t=0s for a transient simulation. Once the transient simulation starts heat will start to flow from hot to cool blocks, very rapidly at first where the dT is high. As t->infinity both blocks will assume the same temperature. In addition to this model I think you will need to set either a fixed T or fixed Q (heat source) on the side where the package sits. I'm more than happy fielding this support issue here but really you would get as good (if not better!!) service from your local support office.

May 21, 2013, 12:37		#11
RobinB New Member Robin Bornoff Join Date: Mar 2009 Posts: 23 Rep Power: 17	Yep, the Tech Note MG245418 on SupportNet outlines a typical transient model definition process.

May 21, 2013, 12:47		#12
naveen kannan New Member Naveen Join Date: May 2013 Posts: 2 Rep Power: 0	Thanks Robin. I'll take a look at that. I guess my keywords didn't work on support net. Will get back if I have any questions.