Modeling Surface Roughness for Heat Transfer

[tibich72]

I am conducting a set of simulations in which different materials are used to spread the heat from a heat source. I have the physical measurements to compare the simulations against. Some experiments have some adhesive between the spreader and the plastic body in which the temperature is measured, and some not. For the ones that don't have, the simulations are pretty far, for the others, the results are close. I am wondering whether the simulation does not take into account the surface roughness between the spreader and the plastic body. Is there a way to model the surface roughness between the two?

I suppose I could create a special, extremely thin (in the area of 10^-5 meters) interface layer that would have the equivalent material properties of air bubbles and the material, but I'd like to avoid this.

Any hint is appreciated.

[Franz Pichler]

HI there,

have you checked the "Heat gap" functionality?
if not the nyou probably just use the heat onductivity of the materials and the interface are not modeled as additional resistances.

maybe thats what you want
best regards
Franz

[tibich72]

Hi Franz,

thanks for the suggestion. I've looked in the Elmer Models Manual, and could not find anything about 'Heat Gap" (the phrase does not exist, and gap is used only in solving thin channel flows). Could you elaborate a bit on this? Thanks.

[Franz Pichler]

HI there,

yeah there are some details that the models manual doesn't cover,
in istuations like that a forum search ("heat gap" gives lots of results hear) or a hardcore source code search can help.

But i would recommend you to start in the forum,
everything is described multiple times in here.
godd luck
best regards
Franz

[tibich72]

Hi Franz,

thanks for the tip. I could find some info on that and was able to do the experiment. Thanks again. Maybe next version of the Models Manual would include some documentation on Heat Gap.

[sebastien ROUQUETTE]

Just add a thin layer in your geometry with material properties set to Air.

If the layer is too thin.. increase it but don't forget to increase also the thermal conductivity of the air.

For example, the roughness of some steel product is about 5µm nbut it can be difficult to model as geomerty so take 50µm and multiphy the thermal conductivity of the air by 10 too... it will work. Your temperature will exhibit a difference at the two extremity of the air layer.

Good luck