How to apply variable heat loads to surface in transient heat analysis?

[Anonymous]

Hi all,

How do I go about adding a variable heat load (Solar Radiation, calculation done manually) to a surface in transient heat analysis? I've tried heat source command with negative values (Heat Sink) but have been getting odd results.

Or is there a better way to add solar radiation heat loads?

Thank you in advance.

ZH

[AstroJohnPE]

Hi ZH,

Your choice of load is probably correct, so I'm guessing that your odd results are related to the input. According to the documentation, a negative heat source will remove heat from the model. Are you saying that your model is heating the Sun instead of being heated by the Sun? That would create some unexpected results!

I suggest that you explain what you are trying to model (and how the load changes in reality), how you applied that load to the model (applied to surfaces? type of load? input for the load? load curve?), and what the odd results are, then someone may have a better suggestion.

Also, note that the load curve "factor" has different interpretations depending on the type of load. Some loads are multiplied by the factor, and some are added by the load curve factor. Search the documentation for the text "transient heat analysis load curve" to find the page with the table that gives the explanation of the factor for each type of load.

[Anonymous]

Thank you for the swift reply john,

I am trying to look at how environmental heat affects a certain machine foundation and was trying to add convection and solar radiation loads. I applied the load to the top surface of the foundation (a steel plate). The bottom surface is air conditioned so I applied a surface controlled temperature. It's odd as the minimum value drops to -200 degrees when my heat load is approximately 950w/m^2 and controlled temperature at 24 degrees.

The heat source command seems to be a generating one and if I apply that to the surface with a positive value, wouldn't that mean it is giving off heat when it is in fact absorbing heat from the sun?

Could I also consult you on how whether the radiation function would be better in this case and how do I goo about using the radiation function.

Cheers,
ZH

[AstroJohnPE · Accepted solution]

Hi ZH,

I’m making a lot of assumptions since we do not have any images of your model. For future reference, a hand sketch of the problem, images of the results, etc. would be helpful.

It's hard to say whether the calculated temperature of -200 is odd or not without being able to look at the results. (Ok. The -200 doesn't make sense from a reality point of view, but simulations don’t always represent a physical situation.) I can think of two reasons why unexpected temperatures occur:

1) Based on your description, you are removing heat with the 950 W/m^2 heat flux, and therefore supplying heat with the 24 degree controlled temperature. So, the calculated temperature must be less than 24 degrees. This could be the case if the -200 is occurring on or near the surface with the heat flux. (This ignores the convection that you mentioned which could be supplying or removing heat depending on the input.)

2) The applied temperatures can give inaccurate results if the stiffness of the applied temperatures is too large, if the mesh is distorted, and/or if the load is applied near a corner. This would be the case if the low temperature occurs at a few points on or near the surface with the applied temperature. A stiffness around 100 to 1000 times larger than the conductivity of the material is usually sufficient to control the temperature and avoid numerical problems.

* * *

About the sign of the heat flux, it doesn't matter whether it makes sense or not. What matters the most in this case is that the documentation says "A positive value will cause heat to be supplied to the model. A negative value will cause heat to be taken away from the model." In other words, the load is representing what the environment does to the model. In your case, the Sun is providing heat to the model, so the magnitude should be positive.

* * *

In my opinion, the heat flux is a good way to represent the heat from the Sun.

1. a) The surface radiation load requires just as many assumptions (or even more) than the heat flux.
2. b) If the surface sees any other parts of the model, or any sources other than the Sun (like the rest of the environment), then the surface radiation ignores the radiation to these other sources. That may not be a good assumption in your model.
3. c) Finally, radiation requires an iterative solution, so the analysis runtime would be "much" longer than using heat flux.

[Anonymous]

Thank you very much john!
That cleared up everything on thermal analysis 🙂