Steady State Heat Transfer (FEA)

[Anonymous]

I am trying to simulate a system consisting of an empty steel box with a 150W lightbulb hanging from the ceiling.  When the light bulb is turned on, the air inside the box heats to an average temperatue of 38.8 C.  In theory, I believe that if my setup approximates reality then when I find the average vector component of heat flux normal to one of the box's exterior surfaces then this value multiplied by the box's total extermal surface area will yield a heat loss rate of ~150W.  So far, I have two ideas of how to do this, but neither are working:

1)  Add a brick type element inside (filling) the box and assign this element the material "Air" from the Autodesk library and a constant temperature of 38.8C.  If the air is a brick type element, though, I suspect that applied convection loads inside the box will not work properly and that the only heat flow will come from conduction through the brick "air" element.  Since air has a low conductance, the heat loss obtain will be way off.   Also, I suspect that ignoring the lightbulb and the radiation it emits is a mistake as well.

2)  Add a small metal object inside the box rather than the air element.  Make this a heat source with total integrated output of 150W.  Add convection to its surfaces and form a radiation enclosure between it and the inside surfaces of the box.  The first difficulty arises in that the autodesk material library does not contain tungsten.  The second difficulty is: if I apply both a radiative load (representing the bulb's tungsten fillament) and a convective surface load (representing the blub's glass globe) then which surface temperature do I assign to the object (fillament ~ 2500 C, glass ~ 250 C)?  Could I solve this problem by making a two layered object?

Thus far, my attampts at both approaches have yielded heat loss values smaller than reality.  Could my ignoring the conductive contact between the box and the tile floor it sits on be the reason?  How could I fix this?

Even a reply that only answers part of my post would be appreciated.

[Joey.X]

Your initial statement is correct,  as balanced thermal problem, 150W lightbulb will have total 150W heat loss rate for an enclosure if it is not completely thermal isolated.

However,  to prove this simple statement in simulation, here are two different thoughts. 

- You can either model a system close to the physical system as a bulb in room, which brings many complicated factors such as Joule heating(electricity to heat), radiation, combined convention (free+forced), boundary conditions.

  Or you can make a much simplified system to verify above statement, i.e, make a heat source with 150W,  perform steady state conduction heat transfer only in 2D domain, verify if the summation of total heat rate in outer surfaces is 150W.

-  Notes: 

(a) The material (with different thermal properties) should not impact the final thermal balance in steady state. 

(b) The conclusion depends on how did you setup the simulation models and how did your verify result in both of approaches, please post your archive models and illustrate how did you verify the inputs and outputs.