Learning Forum

[imselva]

Dear all,

The following are the images of my product. The top and bottom are two copper plates and the four bent attachments are the electrical contacts and in the middle, there is a pipe which we want to heat.
 

 

 

The problem description is as follows : 
Electrical heating: Current supplied: 6kA
                                 The voltage supplied: 2V
                                 Initial temperature: Room temperature.

The two copper flanges are water-cooled (water cooling is neglected for now). 
I want to determine the temperature distribution on the middle pipe, due to electrical heating. 

I am not sure where I should give the voltage and current values while doing the simulation. Since it is not possible to give two conditions for a single face.  

What are the other boundary conditions I should give to simulate the problem?

I am using Ansys AIM 2019 R1, a Thermal model with Electrical conduction. If I am able to give the boundary conditions in proper place I will be able to get the results and proceed further with my design optimization. 

It would be really helpful if someone could help me out simulating this problem.

Thank you

 

Yours faithfully

Selvaprakash  

[Naresh Patre]

Hello Selva prakash

Please check out below application example on Electrical heating in Discovery AIM

Discovery AIM - Structural-Thermal-Electromagnetics Simulation of LED Chip Joule Heating

Let me know if you need any further assistance.

[imselva]

Naresh Patre 
Thank you so much for your help. 
The model, for the moment is giving out a few results, but those are not accurate.

Is it possible for us to give Alternating current and voltage as input in Ansys AIM? Since we are using AC current in the actual case, we want to do the simulation using AC current and compare the experimental results and the simulation results.

[Naresh Patre]

Hello Selva prakash

I am not sure if that is possible. Let me check it and get back to you.

[Naresh Patre]

Hello Selva prakash

I am discussing this problem with the development team and they would like to know if you are interested in coupling AC electric conduction with steady-state or transient thermal.

[imselva]

Yes. As I mentioned in the problem description. I am heating the middle pipe electrically using AC current.
While the top and bottom flanges are water cooled (I set the temperatures of both the flanges to room temperature, in order to compensate for the water cooling.) and there is a heat loss in the inside of the pipe. i.e. Heat is transferred from the pipe to the specimens inside the pipe (which is neglected as of now).

Also I tried to solve for the structure, by fixing the bottom flange. The result I got for this set up while using DC current is as follows:

Electrical simulation (2,2V and 6kA) is done properly, it raises the temperature of the middle pipe to 15000°C, which is too high and the structure fails.

I have another doubt regarding the properties of the material, 
How to give "isotropic secant coefficient of thermal expansion" ? In the materials properties I got from the supplier, I have a tabular value of coefficient of thermal expansion for different temperatures but when I define the material properties in ANSYS AIM, it prompts only for "isotropic secant coefficient of thermal expansion", which I am unable to calculate/find. What value should I use for the same?

[Naresh Patre]

Selva prakash Thanks you for sharing the details. I will pass it on to the development team and will let you know their response.

Regarding your question on having a tabular value of coefficient of thermal expansion for different temperatures, I don't think it's available in AIM but let me cross check it with development and let you know.

[imselva]

Naresh Patre   

I gave the RMS Current value (after considering the losses) of 3285A as shown in the image above.

I splitted the RMS Voltage of 1.48 V into two. I gave 0V to one end face of the bottom plate and 0.74V to the other end face of the bottom plate. I gave 0V and 0.74V to the top plate but in the opposite faces, as shown in figure below
 

When I use this boundary condition, (i.e. I converted the AC current and voltage, to DC equivalent by considering the RMS values of current and voltage.), the temperature at the center pipe is around 1000°C, which seems good.

But now I have a question, Are my boundary conditions correct? Does it make sense to split and give the voltage values like this? Is the total potential of 1.48V applied to the model?

The answers looks promising, but I want to make sure that the boundary conditions are correct/making some sense. 

[Naresh Patre]

Selva prakash I will check out this boundary setup and get back to you. Meanwhile, I got the following update from development team:

From the customer's description, it seems like they are interested in the steady-state temperature.  Another option to solve this problem would be to run an AC magnetic solution and use physics coupling to transfer the ohmic loss to a steady-state thermal analysis using physic coupling.  This approach would compute the AC magnetic field around the device, which they likely don't care about, but it would also correctly compute the RMS loss in the conductor, which they do care about.  The only caveat is that they have to use a solid conductor for Maxwell to compute the ohmic loss, which from the model image it looks like they are using. 

Can you try out the above suggestions?

[imselva]

Naresh Patre Thank you so much for your reply.
Unfortunately, I do not have the License to Ansys MAXWELL. I have asked for it, but mean while I will go through the Ansys Maxwell tutorials online.
Also, if you need a proper description of the problem, feel free to ask me, I could upload a presentation file. 

[Naresh Patre]

Hello Selva prakash

Using either the RMS voltage or the RMS current should work for this case, but you should apply either a potential difference (0V and 0.74V) or apply the RMS current at one terminal (3285 A) and ground (0V) the other terminal.  If you apply both the current and the voltage to the same surface, the voltage boundary condition will override the applied current. 

[imselva]

Naresh Patre Thank you again for your response.
Okay, now  I am simulating the same problem in Ansys Workbench.

I did the Electric analysis with either of the RMS current or voltage values.
Then I gave the results to a Steady-state thermal simulation, but the temperature I am getting on the heat pipe is around 12000°C, which is so high.
But when I do a Transient thermal analysis, (say for 30seconds), the temperature is around 330°C, which is making sense according to me.

What's your take on this, Should I go for a Steady-state Thermal analysis? or Transient thermal analysis?

What are the thermal boundary conditions should I give, in general, to maintain the energy balance in the whole system?
 

[Naresh Patre]

Hello Selva prakash

Since the transient analysis are giving you sensible results, you can go with it.

In general, for thermal analysis you need to specify heat source (like temperature) and heat dissipation (like convection coefficient) terms. You can see a similar setup in the LED chip example shared earlier.

[imselva]

Naresh Patre First of all thank you.
The electrical heating simulation is working perfectly. 

I broke down the problem into electric, thermal and structural and I am simulating it in Ansys workbench. 

Here is my question, I want to calculate the life/fatigue life of this whole setup. 
I have a non-periodic load condition.  The machine operated at Temperature 1 (600°C) for a certain amount of time and at temperature 2 (900°C) for a certain amount of time, which is non-periodic. 

The pressure is also applied non-periodically,  How to simulate for these load conditions? 

I want to calculate the life of the whole setup. ? 
How to proceed with that? Is it possible in Ansys?  

[Naresh Patre]

Selva prakash Unfortunately I don't have much knowledge on Workbench simulation products. I would suggest posting questions on ANSYS Student Community or ANSYS Customer Portal based on the type of license you have.