Thermal Conductivity and Insulation

[themaniac]

Hey people, i need some assistance.

I have a system that is exposed to a hot airflow.
We have tested the system and know the thermal time constant.
We now want to use a insulative paint to increase the time that the system can be exposed to the airflow.
I tried to approach this problem using transient conduction theory but couldn't figure out how to manipulate the time constant to take into account the paint and the paint thickness.
Has anyone has any ideas or websites ect that could help me!!

Thanks people

 

[magicme]

hi

the following thread on this forum has a similar discussion...it should help.

thread391-92345

daveleo

 

[IRstuff]

Insulation generally requires some amount of thickness, not just high resistance. rho*t/A implies that a low thickness is just as bad as a low thermal resistance.

Something like a ceramic coating with some thickness to it might be more in order.

TTFN

 

[themaniac]

Thanks for the advise.

I've got one more question.

How would you deal with a multiple material layer where the conductivity was very different.
For example

Hot Air

IIIIIIIIIIIIIIIIIIIIIIIIIIII
CCCCCCCCCAAAAAAAACCCCCCCCCCC
Cold Air

Where I = Insulation
A = Aluminium
C = Composite (such as fiberglass)

The Aluminium only accounts for approximately a third of the surface area.
I've done some calculations assuming only one of the materials is present but would like to be able to take into account both materials.

Also, whilst I'm asking questions - what would be the best approach to the following

Hot Air

IIIIIIIIIIIIIII
CCCCCCCCCAAAAAAAACCCCCCCCCCC
Cold Air

Again, thanks people for your assistance.

 

[corus]

Depending on the thickness of the materials would determine whether or not you could get away with a 1D calculation. You really need to do a 2D calculation and the best way would be to download one of the many free FE programs that are available for heat transfer.

corus

 

[magicme]

if the layers are very thin compared to the surface area, conduction in the plane of the layers may not be significant (run some numbers before deciding if thats correct)

if this is true, you can treat each region separately as 1D. for example....

hot air
IIIIIIIIIII
CCCAAAAA
cold air

becomes

hot air
IIIIIII
CCC
cold air

AND

hot air
IIIIIIIII
AAAAA
cold air

treat each as 1D case and calculate heat flux and layer temperatures.

i've seen HVAC people do similar approach in calculating heat loss through walls and windows of a building....each is treated separately.

after you get answers from this....back up and calculate the in-plane conduction heat transfer between the CCC and AAAA materials......if its significant....do what corus says....get serious about 2D analysis.

daveleo

 

[themaniac]

Thanks daveleo for the suggestion.
I'm going to get started on that type of analysis imediately

If anyone can suggest a good FREE fe 2d thermal programs it would be appreciated.

Thanks people.

 

[IRstuff]

What is the thermal conductivity of the insulator?

Essentially, for identical thickness and area of the aluminum and fiberglass, the aluminum dominates by a factor of 5000:1. This means that all the heat must try to get through the aluminum. There will be some crowding in the insulator, particularly if the aluminum area is small. This would mean that the apparent thermal conductivity of the insulator is reduced from the parallel flow condition.

TTFN

 

[AkhanThermal]

To 1st question:
The 1D heat flow across the layers is very similar to the Ohm's law: heat flux (W/m2) is a "currrent", temperature drop is like voltage drop, thermal resistance (thickness divided by thermal conductivity) is like electric resistance.
Transient processes inside the layer relax (i.e. the temperature profile becomes linear) when the Fourier number Fo=at/x2 becomes >>1, where a is thermal diffusivity (m2/sec), t is time, x is thickness (or half-thickness).
So the thin layer of paint shouldn't change significantly the inertia of the system, it just adds some additional thermal resistance.