Learning Forum

[nguy3394]

Hello, when I model a heating element (heat-wire) with a volumetric heat source, Ansys is consuming much less power than the experimental results to achieve the same highest temperature.

Is there a way to fix this issue? Does Ansys Fluent accounts for the electrical resistance & contact resistance of Joule heating? Thank you.

[Rob]

Are you just adding a heat source directly, or are you doing something else? 

[nguy3394]

just adding a heat source directly, heating up a water volume

[Rob]

So, where does electrical resistance come into the model? You're adding thermal source in W/m3 to a solid(?) that then heats the surrounding fluid. What are the external boundaries, and where are you measuring temperature? 

[nguy3394]

water column is encased in steel wall, with convective air outside. Heatsource at the bottom column, similiar to boiling a kettle. Measuring max temperature of the water column at the bottom. I am trying to raise the power consumption of the solid heatsource by changing the material properties: but the options are limited to density, heat capacity and thermal conductivity.

[Rob]

What's the water density? 

[nguy3394]

water density is changing with temperature, boussinesq model. I am only looking at the heater material at the moment

[Rob]

But the water is moving? 

[nguy3394]

yes, the water is circulating inside the cylinder. But I am not sure how this affect the Joule power consumption?

[Rob]

If the water is not moving (or is moving too quickly) it'll alter the temperature gain in the solid. 

[nguy3394]

I am trying to understand this: when you multiply the volumetric heat source (W/m3) with the heat source volume (m3) what do you get in term of Wattage? It is not the same as power consumption in Joule heating correct?

[Rob]

The heat source multiplied by the volume is the number of Watts added, correct. The electrical input to the element won't all be turned into usable heat, but it's probably a fair assumption. 

So, if the heat input is correct, what is different to the experiment? Near wall mesh can be important. Similarly, how thick is the steel casing, and how did you calculate the external temperature & HTC?

[nguy3394]

I believe Fluent alone does not simulate the Joule heating effect, heating a say NiCr wire is different than heating a platinum wire, one consumes more power than the other. I believe Fluent must be paired with the thermal-electric module to simulate joule, where we can see voltage, current and resistance, but I am not sure if it can be done for 2D axisymmetric case

[Rob]

If you want to model the electrical heating part have a look at Maxwell. I'd usually just work out the heat flux and apply that.