Heat transfer simulation (air cooling and water quench)

General Questions
1

Hello all,

I would like to simulate the cooling behavior of an alloy from 1000 °C to 23 °C. A cylinder is to be cooled in air and quenched with water.

No matter what I do, the simulation always starts with 0°C and not with the starting temperature that I specified under initial conditions.

I also have the following general questions:

  • Is the Convective-Film load suitable for my purpose?
  • Which parameter can I generally use to influence whether cooling is done with air or water?

Unfortunately I can only upload one picture. I would have liked to upload more.

Thank you in advance.

Screenshot 2024-03-28 111340
Screenshot 2024-03-28 111340811×928 138 KB

2

Make sure that the initial temperature is applied to the whole part, not just to its outer surface. Selection modes will help you with that.

Yes, you can define proper sink temperature and film (heat transfer) coefficient depending on the cooling medium (to be found in literature).

3

Thank you for your answer.

That’s exactly what I was thinking.

I have assumed the following for air cooling, for example:
Sink temperature: 23°C
Film coefficient: 5 mW/(mm^2*°C)
applied to the surface

And for water quench I assumed the following:
Sink temperature: 23°C
Film coefficient: 4000 mW/(mm^2*°C)
applied to the surface

Nevertheless I get the same results. Have I missed something?

4

Can you share the .pmx file (you could upload it to some hosting website and paste the link here) ?

5
6

Why is the part split like this ? To see the temperature inside ?

Are the results the same in terms of the final temperature ? It will be the same as the part reaches steady state (sink temperature) eventually. But the cooling time should be different.

7

Exactly, I want to see the course of time in the core and at the surface until the final state is reached.

I expected the cooling time to be different, but I get the same value whether I assume a film coefficient of 10 or 4000. I took the values from the literature.

8

Compare the initial frames, long before steady state is reached. I get different results for 5 and 4000.

For node 1:

node 1
node 1703×439 13.9 KB

1 Like
9

With a different resolution in the graph and smaller inital time increment I can also see it now. Many thanks for your help!

10

You should validate your model, like with a Jominy-test or something similar.

Quenching something im water is tricky. If you just have water and you put a subject 1000°C hot, you get an instant vapor blanket - which has totally different thermo parameters.

My point is, always be extra cautious with such a simulations. If your goal is to determine some phase transformation, start with Jominy test and validate your model.

Good luck

1 Like