Learning Forum

[AnnaYang]

Hi everyone,

I was trying to solve a steady-state temperature distribution problem in ANSYS Workbench 17.2. 

However, there are some confusing definitions that I am facing:

 

1. Assume heat flow = Q (W) and the subjected area = A (m^2)

I was wondering what's the difference between applying a "Heat flow (Q)" and applying a "Heat flux (Q/A)" on the same surface since I found that the temperature distribution by applying the two method above were different.

 

2. If a body part was heated uniformly by a external heat source, how can I apply this kind of boundary condition?

I've tried to use the internal heat generation option but the result was not even close to the theoretical solution.

The way I applied the heat generation is taking the heat flow value over the volume of subjected body.

 

3. As I realized, the coefficient of free convection in air surrounding,h (W/m^2-C) ,can be obtained by using empirical formula.

If the result from simulation of free convection is not as good as expectation, is that reasonable to enlarge the value of h?

If so, what's the limitation of increment ( like h should be lower than 25 or something)?

 

Any suggestion would be appreciated, thank you!

[Rob]

1)  If I apply  10W to a 2m^2 area my W/m^2 is 5.  So, if I apply 10W or 10W/m^2 I'm adding a different amounts of energy. 

2) If something is heated uniformly from an external source, I'd apply a boundary (surface) effect: not a volume source. 

3) Think about what you're doing: if I don't like a value is it reasonable to guess a different number?  Why do you think the result is not as good as expected? 

 

The ANSYS tools are best used with a thorough understanding of the problem you're trying to solve. Otherwise you won't know if there's something wrong (or missing) from the set up or solution. We set boundary conditions based on our understanding: this needs to be correct. 

[AnnaYang]

Hi rwoolhou,

1) I understood what you expressed, but what I applied to the surface is 10W and 5W/m^2 respectively, and the result is different.

Sorry that I did not describe it clearly.

2) You mean, if a cube body ( Volume = 8 m^3)  got a heat flow = 24 W, you will apply 24 W on each face of this body instead of applying 3 W/m^3 to the whole body?

3) The expectation comes from the result of a practical experiment that I've conducted. In this experiment, there are multiple temperature measuring points on the structure, and the measurements of experiment is much lower than the result of simulation. So I was wondering if there is something wrong with the  convection coefficient.

 

Thank you for reply!

 

[Rob]

1) Odd, what do you get when you report fluxes on that surface for a converged solution?

2) Depends: what are you trying to mimic? Remember you're applying a total heat load so always need to account for volume or surface area when setting boundary conditions. 

3) Nothing wrong with the coefficient function in the solver, whether the value is correct is an entirely different question. 

I think it'll help if you post images of what you're setting up & what you're expecting. From your questions it's likely you've missed a setting somewhere, or there's a definition issue. 

[AnnaYang]

Hi rwoolhou,

1) I've checked the setting and found the reason is wrong unit transformation , thank you.

2) The situation is: a volume of oil flow was heated up by the pump power loss of oil supply system, so I adopted internal heat generation to simulate heated oil flow volume. However, the temperature rise of the oil volume is over 100 Celsius degree and the one of experiment is only 3 Celsius degree.

3) Maybe the difference comes from neglect of the forced convection effect, I will evaluate the significance of my case.

 

Thank you very much !