Community
CFD Forum
Welcome to Autodesk’s CFD Forums. Share your knowledge, ask questions, and explore popular CFD topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Simulation Results with Extreme temperature

11 REPLIES 11
Reply
Message 1 of 12
kadenlh
992 Views, 11 Replies

Simulation Results with Extreme temperature

Hi All,

 

I got some problems simulating my case for a room with floor supply and ceiling return as attached.

Here is my case:

 

1.Supply grille on floor with veloctiy = 0.124m/s, volumetric flow rate with 0.046m3/s per grille, while a strip contains 16 grilles(16NOS X 8ROWS)

2.Ceiling return grille with velocity = 0.12m/s and volumetric flow rate of 0.079 per grille (8NOS X 8ROWS)

3. some light fixtures (32NOS) suspended from the ceiling with 563W total heat dissipation per each strip

4. 12 blocks on the floor representing human and equipment load, 1880W heat dissipation per each block

 

And I have been modifying the model and these parameters for a few days, with the simplest model currently and with different parameters combinations(i.e. inlet input(velocity, volumetric flow rate, temperature); outlet input( pressure=0) or outlet input (velocity))

 

However, the major problem that I face now is the extreme temperature result inside the room.

I left the outlet and the room temperature undefined initially in order to see the real case from the simulation, but the result showed 60degC inside the room without any heat leaving the outlet. It became even worse afterwards, the temperature now rises to thousands degC. I did modelled an external volume with temperature 34degC to control the atmospheric temperature before but it also failed. This is confusing since the supply flow rate of 18degC was calculated from the cooling load within the zone, so it should be able to cater for the heat dissipation.

So my question 1 is why would this happen, did I input anything wrong? Would anyone please help me on this.

 

Also, another one, question 2 , if i model the human and equipment load as a layer of solid air on top of the supply grille but not as a block on the floor, would this affect the flow pattern like blocking the buoyancy effect?

 

Here is one more question 3 about skylight.

There is a zone with Skylight in my study, how should I model the radiation heat gain for the glazing, e.g. if I model a volume with a total heat dissipation, should I still pick the box for radiation in solver?

 

Thanks a lot for your attention.

Highly appreciate if anyone would help me one this!

 

 

11 REPLIES 11
Message 2 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

Answer 1

We need to re-think the set up here. Internal surface Boundary Conditions are not advised, so these all need to be removed.

 

Do you need that external air box modelled? I am not sure what it's purpose is?

If you do, you could still work like this but the under floor flow needs to be attached to a suppressed volume.

 

Answer 2

Yes - try to stick to modelling things realistically. Saying that - you could use a resitance region with a volumetric heat load applied to it. But does the air flow through humans and racks?

 

Answer 3

Start with just applying a heat flux to the outer surface. You would want to have radiation on.

 

Also, for this natural convection model you should not have auto forced convection on, as I think you need flow and thermal to be solved as one.

 

Hope that helps,

Jon

Message 3 of 12
kadenlh
in reply to: Jon.Wilde

Hey Jon! Thanks for your quick response!

I followed your advice and removed the external and also the bottom volume so that all BCs become external except the heat load.

The flow seemed closer to what we expected now.

I also modified the heated volume so the temperature rise is smaller now.

Result

However, there are still some problems as follows:

1. Heat Load Temperature rise: The temperature still rises to over a hundred degree celcius at the initial stage. Is this a normal phenomenon?

2. Room Temperature: We are expecting a room temperature of 22degC from the cooling load simulation before but now it's like 26~30degC at the final stage in CFD, what can I do to improve it?

3.Pressure & Vol Flow rate at the outlets: The outlets at the ceiling are set to Pressure = 0Pa but the flow currents to outlets are still not obvious. Using Bulk calculation, the volumetric flow near the oulet plane (2.Xm3/s) is half of the inlet volumetric flow rate(5.93m3/s), also the temperature becomes -14221.1, Celsius which is weird. Would you kindly suggest some improvements on this?

4. Room Pressure: Actually we wanted the room to be positive pressure, i.e. 10Pa so that the flow will rush through the outlets. I did input the +ve pressure in the intial conditions before, but the supply flow from the inlets seemed to be restricted. Did I do anything wrong here or is there any other way to model the positive pressure within the volume?

 

Please also find my updated model in the link below for your reference (file size too large to upload here).

http://www.mediafire.com/download/n7s7j0tal1ftb29/G1S1.rar

Sorry that I am still new to CFD. Really appreciate your help!! Thanks a lot!

 

Message 4 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

No problem at all - and please don't apologise. I was new to CFD once too 😉

 

I would not worry about adding concrete to the walls, I can't see the benefit of this.

What is the Solid Air doing? Just blocking the inlets?

 

You need to simplify the BC's.

Have velocity or flow rate at the inlet, both is overkill. A Temp is sensible.

At the outlet, just have a P=0. I would also recommend extending the outlets outside of the model if possible, this will give a more realstic flow field.

 

To respond to your questions:

 

1. Heat Load Temperature rise: The temperature still rises to over a hundred degree celcius at the initial stage. Is this a normal phenomenon?

 

I think this is down to the air flow blockagage and the BC's - not normal no 🙂

 

2. Room Temperature: We are expecting a room temperature of 22degC from the cooling load simulation before but now it's like 26~30degC at the final stage in CFD, what can I do to improve it?

 

Set the model up correctly and I suspect your results will also be in-line with expectations.

 

3.Pressure & Vol Flow rate at the outlets: The outlets at the ceiling are set to Pressure = 0Pa but the flow currents to outlets are still not obvious. Using Bulk calculation, the volumetric flow near the oulet plane (2.Xm3/s) is half of the inlet volumetric flow rate(5.93m3/s), also the temperature becomes -14221.1, Celsius which is weird. Would you kindly suggest some improvements on this?

 

As above 🙂

 

4. Room Pressure: Actually we wanted the room to be positive pressure, i.e. 10Pa so that the flow will rush through the outlets. I did input the +ve pressure in the intial conditions before, but the supply flow from the inlets seemed to be restricted. Did I do anything wrong here or is there any other way to model the positive pressure within the volume?

 

As above 😉

Message 5 of 12
kadenlh
in reply to: Jon.Wilde

Hey Jon!

I updated the model again with the following changes that you suggested:

1. Solid Air removed

2. Using just velocity or flow rate at the inlet and also Temp is sensible.

3. Using just P=0 at the outlet

4. Extending the outlets to the external surface of the model

 

Also, we are expecting displacement cooling from those 0.12m/s velocity supply grilles (i.e. the 18degC air would form a layer of cool air,heated up by the load and rise to the ceiling.) To achieve this, I made the following changes again:

1. Gravity-> Earth, with direction = 0,0,-1

2. ADV5 chosen

3. Turbulence method set to mixing

With these changes, the hot air from the load rises as expected.

 

However, there are still some problems as follows:

1. Heat Load Temperature rise: The temperature during the simulation became extreme value again (e.g. 2XX ~ 15XXX degC) and this time not only for initial stage but it kept happening. From your words"I think this is down to the air flow blockagage and the BC's", please would you suggest some solutions for this? This has been the major issue in my simulation and it became even worse Smiley Sad

 

2. Room Temperature: The room temperature is still much higher than 22degC (ard 26~30degC), what can I do to improve it?

 

3. Room Pressure: The room pressure is still not +10Pa and in the last post I mentioned about the intial conditions that i input before (the supply flow from the inlets seemed to be restricted). Would you please let me know did I do anything wrong? Is there any other way to model the positive pressure within the volume?

 

Here below are the simulation curves for your information, the major thing for problem 1 that i observed is the temperature fluctuated greatly during the simulation (i.e. 5K in this iteration and 2XX in the next one). This happened every single time and I have no idea how to control that. Please would you help me on this.Smiley Sad

SimCurve.jpg

Thanks so much again for you kind support! Really, thanks for helping me.

Message 6 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

Assuming the set up is now OK (it does sound it), all of these are likely to be one issue.

Have you tried refining the mesh?

 

Also, there is no need to set initial conditions here - a steady state is essentially the room condition after an infinite time, so the starting point means very little.

The pressure is likely to be close to 0, as this is what you have at the outlet, although it would rise towards the inlet, as you would need a higher pressure to drive the flow at your prescribed flow rate.

 

Hope that helps,

Jon

Message 7 of 12
kadenlh
in reply to: Jon.Wilde

Hey Jon

 

Thanks for you reply.

I tried to refine the mesh last weekend but it didn't help at all...

Now I further simplified the model to see if this helps.

But the current result still shows a temperature rises to extreme values.

This started to happen when I tried to use gravity method for my buoyancy driven flow.

 

test.jpg

 

 

 

Attached please find the latest model for your information.

Please would you check if there are any problems and would you kindly suggest some solutions on thisSmiley Sad.

Thanks again.

Message 8 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

You still need to extend the inlets and outlets so that they are not in-line with the floor/ceiling. Add parts in CAD to achieve this.

It is also pretty difficult for the solver to figure out how to add heat to the model when the part with a heat load also has inlets. Adding heat to an air volume like this tends to be a little difficult, mainly as it does not represent reality.

What really happens here?

 

Otherwise, I think you are getting pretty close, everything else looks good.

 

Jon

Message 9 of 12
kadenlh
in reply to: Jon.Wilde

Hey Jon!

 

 

Thanks for your reply!

I also found the problem of adding heat to an air volume, this make the temperature rise like crazy.

I changed all to thermal pad now and it seems more resonable for the temperature rise.

Although in some cases the temperature rise up to thousands degC during simulation, i think this may be due to the meshing problem, am I right?

 

However, I found 2 other problems:

 

1. Actually I have 3 Zones and each of them have 3 Scenarios, thus I need to run 9 simulations here.

All the Scenarios within a zone have the same heat generation from the pad, i.e. Zone 1 heat load =52kW. The only difference is the air grilles location and number.

After simulation, the patterns for the temperature looked good, however, the heat flux values from wall calculator for Zone 1 Scenario 1 (Z1S1) to Z1S3 vary a lot, e.g. Z1S1=51kW, Z1S2=38kW,Z1S3=45kW. 

After checking, seems like some heat generated went back into the pad instead of coming out from it. 

I was wondering why this is happening as I didn't set the heat generation as a temperature-dependent function.

 

2. Also, as I mentioned, the heat load in the three scenarios for the same zone are the same, the total cooling air volumetric flow rate are also the same.

However, the zone temperature from the 'parts' result are different for the scenarios. Temp for S1 & S3 are the same, while that for S2 is 1.5degC higher. Do you know what might be the major cause of this? How may I solve it?

 

Thanks a lot again for your help!

Cheers!! Smiley LOL

 

Below is my preliminary result by the way!

 

 preview.jpg

 

Message 10 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

Glad you are making progress. Please do implement the inlet and outlet extensions I suggested too.

 

What do you mean by 'pads'. How did you apply the heat load?

 

I am not sure I follow the rest of your question, sorry! 🙂

Message 11 of 12
kadenlh
in reply to: Jon.Wilde

Hi Jon 😄

 

I did implement your suggestion on the inlets and outlets. Thanks, it worked for some cases. 

For the pads i mean thermal pads (basic conform), with total heat generation added on them, but the problem I mentioned is as follow:

3000 W was applied on the thermal pad A & B, after simulation, pad A showed 3000W heat flux while pad B showed -25XXW heat flux in the wall calculator. I was wondering why this is happening as I didn't set the heat generation as a temperature-dependent function.

 

Again, thanks a lot for your help Jon!Smiley Very Happy

Message 12 of 12
Jon.Wilde
in reply to: kadenlh

Hi,

 

What is the heat flux you are seeing? Any discrepancy could easily be down to the mesh quality.

 

No problem on the help, I am happy that it is proving to be useful!

 

Joln

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report