# Estimating heat transfer coeffecient

1. ### jbord39

75
Hey guys,

I am completely stumped right now (in many different ways). I am a senior in Electrical Engineering and just started an internship this summer. Sadly, they put me in a setting which involves nothing I have studied except the pure math (there is not a volt or ampere that I have seen yet).

I am trying to estimate the heat transfer coeffecient and use this in COMSOL of a hydrogen fuel cell. To be honest I have never seen any of these equations before and am really flustered. They mentioned that the h would be different for each side of the cell.

Supposing the cell is X by Y length, how would one get a ballpark estimate on the h for each side? (One side is assumed to the symmetric center, so no heat flows through here).

Any help is appreciated (as well as any tips or tutorials for COMSOFT).

Thanks,

John

2. ### Mech_Engineer

2,299
Convection is a very tricky subject, but you can make some assumptions to make things a bit more manageable. Before I can help you we're going to need a few things:

• Are you looking at cooling on the exterior of the cell, or flows inside the cell?
• Can we approximate the geometry as a simple one such as a rectangular cube?
• Is it "natural" or "forced" convection?

Given this, I can come up with an equation to help out.

3. ### jbord39

75
Thanks.

I don't have exact answers to your questions but here is the setup as I understand it.

Here is what I know:

The basic idea is there is a cylindrical shell filled with metal alloy powder which absorbs hydrogen. Hydrogen is pumped into the center of the tank. As the hydrogen is absorbed by the metal powder, heat is generated which leaves the sides of the tank.

We are simplifying the problem by using the radial symmetry of the cylinder and basically using just a square, with the left side having no flux through it (because it is the very center of the cylinder).

So I think we can approximate it as a simple cube. I am not sure what natural vs. forced conduction is. Right now, there is just air on the outside of the tank with no wind blowing on it or anything. We are mostly concerned with how the inside of the cell reacts as it releases heat.

Thanks again,

John

4. ### Mech_Engineer

2,299

I'm still not 100% clear on what you're trying to calculate- are you trying to find the temperature of the unit based on a set amount of heat generation? Or are you trying to do an axisymmetric model of the temperature distribution inside the unit?

5. ### AlephZero

7,298
It seems like you need to review some basic course material on heat conduction and convection, Newton's law of cooling, etc. At least, that will get you to the point where you can see how your heat transfer coefficient formulas fit into the "big picture", and ask some more focussed questions about your thermal model.

6. ### jbord39

75
I am in way over my head, but I didn't choose where I was put. The problem is, that I also don't know anything about fluid mechanics, mass balance, thermodynamics, or chemistry.

Does anyone know where I should start to get a very, very basic grasp of these. I don't actually have to solve a single equation, just define boundary conditions and how they couple (Darcy's Law, Dilute Species Transport Equation, and momentum equations).

Thanks,

John

(also how to calculate h based for heat flux through the surface would be nice).

7. ### jbord39

75
We're trying to include the heat flux from the metal powder inside to the air outside. Heat is generated as hydrogen is absorbed by the powder.

Thanks,

John

8. ### Mech_Engineer

2,299
If you know how much heat is being generated and you're just interested in finding the temperature at the outside surface of the unit, this would be a somewhat straightforward natural convection analysis (although it is analytically intensive, especially if you've never done the analysis before).

Solution of the problem will be as such:

• Calculate natural conveciton coefficient as a function of temperature.
• Set up thermal equivalent circuit based on know heat generation, thermal resistance of natural convection, and free-air temp.
• Calculate heat transfer coefficient based on developed equations.

9. ### Mech_Engineer

2,299
One thing you'll be happy to know is heat transfer problems are commonly set up as "thermal equivalent circuits" which parallel electrical circuits. Given Ohm's law:

$$\Delta V = I * R$$

A thermal equivalent circuit is:

$$\Delta T = \dot{q} * R_{thermal}$$

Where Delta T is the temperature difference between two points (kelvin, celcius, etc.), q dot is the heat flow (watts, btu/hr, etc.), and R thermal is the calculated thermal resistance.

So knowing this, you can calculate the outside temperature of the fuel cell if you know the thermal resistance of the natural convection around it. Unfortunately natural convection varies significantly with temperature, but the problem is at least solvable.