[khattaksadia]

I need to input the following time varying heat flux at a particular spot in a geometry:-

Time(s) Heat Flux(W/mm2)
0 0
0.01 4.14
0.02 4.14
0.03 0
0.51 4.14
0.52 0
1.01 4.14
1.02 0
1.51 4.14
1.52 0
2.01 4.14
2.02 0
2.51 4.14
2.52 0
3.01 4.14
3.02 0
3.51 4.14
3.52 0
4.01 4.14
4.02 0
4.51 4.14
4.52 0
5.01 4.14
5.02 0
5.51 4.14
5.52 0
6.01 4.14
6.02 0
6.51 4.14
6.52 0
7.01 4.14
7.02 0
7.51 4.14
7.52 0
8.01 4.14
8.02 0
8.51 4.14
8.52 0
9.01 4.14
9.02 0
9.51 4.14
9.52 0
10.01 4.14
10.02 0
10.50 0


Kindly tell me how to do it.

P.S.: I have asked this question after going through all the manuals and examples. One way I came across was to apply udf. I have gone through various udf coding but I couldnt figure out how should I do the udf coding for my particular case. I have tried uploading heat flux profile for above stated table but no different results were observed.

[KevinZ09]

Use a profile file for your boundary condition. See section 6.6 of Fluent's User Manual for details and examples.

[KevinZ09]

I'm replying again because I noticed you editted your post since my first reply. I do wonder what you mean with "no different results were observed" and "couldn't get answer of my interest"? I mean, if you want a time dependent boundary condition, and you have tabular data, using a profile file is, by my knowledge, the way to go. You could use a UDF, but that's complicating things unnecessarily. So if you did not find improved results, either I misunderstand what you're trying to do or you did something wrong. In either case, I'd need more information to better help you, if you still want help....

Do you want to apply this heat flux over a whole boundary or just a tiny part of a boundary/face? And is the heat flux every time on for only 0.01s? And you said you didn't observe any improved results. Improved compared to what case? Please elaborate.

[khattaksadia]

The "P. S." describes what I've done so far yet I couldn't succeed in achieving it that makes me confused. The Geometry for which I am using is as follows:




The elliptical profile is made by "Boolean (Imprint)" command, hence similar profile in the inner face of this hollow cylinder. This elliptical imprint in inner face of the hollow cylinder signifies the profile made when a pulsed laser beam strikes the hollow cylinder's inner face. The laser beam strikes it for 0.02 seconds and the similar pulsed process for complete cycle of 10.5 seconds. The tabular depicts the Heat Flux of the laser beam for first 0.022 seconds, and then strikes with heat flux upon 0.51 seconds, then upon 1.01 seconds and so on. Meanwhile, the coiled grooves between the outer and inner walls signifies cooling channels in which water is flowing. Hence when the heat flux isn't there, say between 0.03 and 0.5 seconds, the component gets cooled.
I'm working upon Proton Beam Dump which has to be thermally and structurally stable. For such a case I need to input this tabular data in fluent's setup. I tried inputting it under "Profiles" but there isn't any noticeable change in temperature. I need to couple fluent along with transient thermal so as to study the cooling effects of the fluid when pulsed laser beam is striking. For example, if I change the velocity of water at inlet, what change in thermal analysis would be observed etc.

[KevinZ09]

Thanks for the detailed description. It helps in getting to understand better what you're trying to do. Some observations/suggestions/questions:

1: Is the elliptical imprint that represents the laser beam present as a separate face on the inner face of the cylinder? I just wonder what face of the inner cylinder looks like? Are the faces overlapping (the laser beam and inner wall of cylinder), or are they joint/connected faces?
2: In your table I noticed it says W/mm^2. Does it really represent Watt per squared millimeter? If so, do note that Fluent asks for W/m^2, so you'd need to convert it to that unit. If you actually ended up using values of 4.14, it's no surprise you hardly find a difference, as 4.14 W/m^2 is very little.
3: I'm not totally getting the pulsation of the laser beam. It's on for the first 0.02 seconds. Then off till 0.51s, when there's a pulse again, till let's say 0.52s. Then it goes off again? So basically:
0 - 0.02 = on
0.02 - 0.51 = off
0.51 - 0.52 = on
0.52 - 1.01 = off
1.01 - 1.02 = on
etc.
Is that how it works?
4: Then as a follow up on 3, do note that Fluent will linearly interpolate between the data. So if your table reads
0.52 = 0
1.01 = 4.14.
Then at intermediate values, like at 0.765, it has a value of 2.07. So make sure your table actually represents what you're trying to model. The value it uses obviously also depends on your time step. So using time steps of 0.5 will give very low values of energy input. But I'm assuming you're using a time step of 0.01 or smaller?
5: Since the pulses are so short, again, a very little amount of heat is added in your time step, so you might not find much of a difference in temperature, especially compared to the cooling of the water. So I'd suggest, if you get the table loaded correctly, to first run tests without the water cooling.
6: Is there actually a fluid flowing through the hollow cylinder?

That's it for now. Hope it helps a bit. Once I get some more feedback I'll try to help you along further. Good luck for now!

[khattaksadia]

Thanks for giving me the insight upon my problem.

1: Yes. I have already stated in my previous reply that the elliptical imprint represents the region where laser beam (l.b.)strikes. As per already known fact from engineering drawing, when an inclined circular beam hits any surface, the profile that we will get will be elliptical. Similar concept has been used for the modeling here. I had modeled an inclined l.b. striking inner face of the hollow cylinder. Since I require only elliptical region of l.b.to apply boundary condition of heat flux, I have used Boolean (Imprint Faces) command in Design Modeler of ANSYS Workbench. I have already attached the jpeg file in my previous reply, kindly right click and view it in new tab. Inform me if you require any other view of geometry, I'll attach another one.

2: I have gone through the units properly. Before inputting the values I have made all the necessary conversions already.

3: Yes. I have already described it's working in previous reply. 0.01-0.02 sec l.b. strikes so heat flux (hf) = 4.14 W/mm^2 (heating), 0.03-0.50 sec no l.b. so hf = 0 (cooling), 0.51-0.52 l.b. strikes so hf= 4.14 W/mm^2, 0.53-1.00 no l.b. so hf = 0, and so on till 10.5 sec.


4: The solution was consuming very high GB so I did not write intermediate points and just relevant input. In reality the time span is 0-10.5 sec with time step of 0.01 sec and the values of heat flux follows the pattern as I've stated.

6: As per my previous reply, the geometry has grooves cut inside the hollow cylinder along which water actually flows to carry away the heat. Please refer both jpegs attached in my previous reply.

Kindly let me know if you need any other details.

[KevinZ09]

Thanks for the info. You say you don't find a noticeable change in temperature. The temperature of what? And what is the time-step you're using?

And as mentioned earlier, the way you've currently set up your table, it won't replicate your laser beam's behaviour. Fluent will interpolate between intermediate points. So I think you'll have to include more data points. Or go the UDF-route and write an UDF with a function profile.

[khattaksadia]

As mentioned earlier I'm working upon beam dump upon which a circular laser beam is striking (since its going to be inclined hence an elliptical profile). Upon this elliptical profile I wish to input Variable Heat flux in the pattern as stated above. And as stated earlier, the total time span is 10.5 sec and the time step = 0.01.

Being a beam dump, it needs to be thermally and statically stable. Reason of coupling Fluent with the Transient Thermal is to regulate the Heating (i.e., temperature) of the beam dump through flowing Water in the cooling channels (in the coiled grooves as shown in the picture) so as to ensure the thermal stability of the beam dump when the beam actually strikes the inner surface of the beam dump.

When I'm Coupling Fluent with Transient Thermal through System Coupling, what actually should happen is that as I increase the Velocity of Water, Temperature in Transient Thermal should decrease and as I reduce the Velocity of Water, Temperature should increase. However, no matter what velocity I'm giving as input in Fluent, there's no change in the Temperature in Transient Thermal.

If you wish I'll save my analysis and will give you the link so that you can view it which might help you in giving me the insight.

Thanks in Advance.

[cfd335]

Originally Posted by khattaksadia

As mentioned earlier I'm working upon beam dump upon which a circular laser beam is striking (since its going to be inclined hence an elliptical profile). Upon this elliptical profile I wish to input Variable Heat flux in the pattern as stated above. And as stated earlier, the total time span is 10.5 sec and the time step = 0.01.

Being a beam dump, it needs to be thermally and statically stable. Reason of coupling Fluent with the Transient Thermal is to regulate the Heating (i.e., temperature) of the beam dump through flowing Water in the cooling channels (in the coiled grooves as shown in the picture) so as to ensure the thermal stability of the beam dump when the beam actually strikes the inner surface of the beam dump.

When I'm Coupling Fluent with Transient Thermal through System Coupling, what actually should happen is that as I increase the Velocity of Water, Temperature in Transient Thermal should decrease and as I reduce the Velocity of Water, Temperature should increase. However, no matter what velocity I'm giving as input in Fluent, there's no change in the Temperature in Transient Thermal.

If you wish I'll save my analysis and will give you the link so that you can view it which might help you in giving me the insight.

Thanks in Advance.

Hi Kevin,

Can you please suggest me how to provide gaussian heat source during TIG welding of inconel 600. For current = 140 A, voltage = 12V, weld speed = 2 mm/sec. If I need to use a UDF, where should it be given, compiled and loaded into Ansys Fluent?