Hey Brandon -
Two things:
1) did you know that there is a thermal simulation for the Fusion Electronics workspace? I don't think you used it here. That's just fine - but I thought I would let you know. That Electronics Cooling workspace is simplified for electronics workflows.
2) I know what's going on here. What you've done is a very common beginner's mistake for thermal analysis. Easy to fix though.
The problem is that 20 C boundary condition you defined. While saying "this surface is in contact with the air, so it is 20 C," sounds reasonable, it doesn'T work the way most people expect. If you say that surface is 20 C - then it is ALWAYS going to be 20 C. NO MATTER WHAT!
In effect - if that surface begins to heat up due to the thermal loads around it, it will keep pulling more energy out of the system until it reaches 20 C. If it starts to cool down - it will start pumping energy INTO the system to keep it 20 C!
Using temperatures as boundary conditions is tempting, because we can so easily measure them - but a thermal analysis should be thought of more like an energy balance problem. You should not define states of temperature, you should define heat sources, and heat sinks.
So - chips and stuff like that are easy. They put out so and so many Watts of energy. This is easy to define. If that was ALL you did, you couldn't really do a thermal steady state analysis though, because you would be defining a heat source, but nowhere for the heat to leave the system. Anything you do not define is considered adiabatic. The model doesn't know about air around it, or anything else! In the case I just described - the chip would keep heating the system up without end, so there would be no steady state result possible. Although a transient analysis COULD be done.
You'll need to define areas of conductive heat loss to get energy out. (if you have any) This is harder to do. You'll need to know how much energy your steel screws pull out and enter this. Its gonna be minus Watts/sq m or something like that.
Even this is usually not enough! There is the odd chance that your heat source and your conductive heat loss line up juuuuuuuust right. All the heat going in, goes out. This will give you a steady state. But it is highly unlikely. To get everything balanced up - you are going to have to account for convective losses too.....and this, on every surface where they are occuring.
This will square up the circle for you because convection is special. As a part gets hotter, convection pulls ever more energy out of the system. Eventually - your part gets so hot that it will pull all the energy you are entering into the system out, and you will reach a steady state! This is the trick!
The units on that coefficient are Watt per sq cm per degree Kelvin, or something like that....see? The other thing you will have to define is the ambient temperature. If you put this thing in a 500 degree oven, it would actually pump energy back INTO the system...get it?
Where do you get all these numbers?
Well - there are handbooks full of them. My Marks Mechanical Engineers Handbook has whole chapters on this. Google can help. Try looking for "convection coefficient for ambient air at 20 C with 60% humidity" and "thermal conductivity of steel" for examples.
This is a long explanation - but this happens so much to so many people, that I thought it might be worth it to describe it in some detail. Hope it helps.
Mickey Wakefield
Fusion 360 Community Manager
