Induction Heating Process Characterization

[Engineer16]

Hi All,

I'm working on a process involving induction heating. I'm heating up an irregular shape ss die to make a bond. My dilemma is in my understanding of where the eddy currents will be that generate the heat (based on skin effect). The clam shell die is pictured below. The right hand image is the orientation that the die will be in, centered in the coil. The interface of the clamshell halves does not seem to impact the heat profile as compared to a one-piece die. I'm concerned with skin effect- the indentation is where I need heat to be generated. Will eddy currents be generated on this surface and therefore heat the "internal surface"? Or does the peripheral surface prevent the current from generating current in the indented area and the heat observed is from dissipation from the outer surface heating? This is time sensitive so any and all answers are appreciated. Thanks in advanced!

 

[JBA]

For some information about your question see the "The Characteristics of the Part" section on the below website

http://www.gh-ia.com/induction_heating.html

 

[Aaron Crowl]

Let's look at the induction law:

$$ ∫E⋅dl=-∫∫\frac{∂B}{∂t}⋅dA$$

The first part is an integral around some loop and the second part is a double integral on the surface created by that loop.

Draw any circle on the surface of the indented part. If you apply flux to the surface of that circle then there will be an induced field and current. So yes, it will heat up. However, notice that the right side of the equation depends on the flux captured by an area. The calculated current will be higher when that area is bigger so the currents will be stronger as you move outward. In other words the highest temperatures are going to be on the periphery of your die.

Do you have to split your die? Could you use a spiral heater like a stove top burner? That would give you better control.

 

[Nidum]

I'm heating up an irregular shape ss die to make a bond.

What temperature is required ?

 

[Engineer16]

What temperature is required ?

Minimum temp is 160.

 

[Engineer16]

Let's look at the induction law:

$$ ∫E⋅dl=-∫∫\frac{∂B}{∂t}⋅dA$$

The first part is an integral around some loop and the second part is a double integral on the surface created by that loop.

Draw any circle on the surface of the indented part. If you apply flux to the surface of that circle then there will be an induced field and current. So yes, it will heat up. However, notice that the right side of the equation depends on the flux captured by an area. The calculated current will be higher when that area is bigger so the currents will be stronger as you move outward. In other words the highest temperatures are going to be on the periphery of your die.

Do you have to split your die? Could you use a spiral heater like a stove top burner? That would give you better control.

Thank you Aaron, this helps a lot! Unfortunately the die must be split in order to remove it from the system once the bond is made, and we are constrained to the current orientation due to other processes on the same machine.

Another problem the I'm running into is that the halves of the die are held together by a PTFE clamshell (slip fit). It is embedded so that only the indented portion is exposed. Prior to fabricating the die in two pieces I had tested a single die and it created sufficient reflow for the bond (no PTFE holder). I've done an experiment where I know the PTFE reduces the amount of heat generated. I believe it is acting as a heat sink, which would be in line with your equation and statement- that the periphery generates more heat. I'm thinking that the PTFE sucks the heat out of it, preventing the indented surface from heating due to the radiant heat from the periphery. I'm wondering if exposing the die so that it is not fully embedded, or changing the PTFE for a different material may help. Do you have any thoughts on that?

 

[Aaron Crowl]

Could you try something ceramic for the clam shell? It's a great insulator so in theory it will help trap the heat in the die. The heat will flow from high to low on your die and the temperature will become more uniform.

Edit: You could make a mockup out of plaster of Paris. People make do-it-yourself crucibles with it. However, my quick research says that plaster of Paris emits toxic fumes under extreme heat. I can't find any good data on safe operating conditions for plaster so don't take that route without some research. I don't think your temp would count as extreme but I'm not confident enough to say it's safe. Anyway, a fired ceramic would surely be safer and stronger than plaster.

 

[Aaron Crowl]

Ahh, here you go. Insulative epoxy compounds used by Nasa. The potting compound (link here) looks like it will do the job for you. That compound can be cast 2 to 3 inches thick and you can do it all at room temperature.

 

[Aaron Crowl]

I got back from a job and I looked up some stats. It turns out that PTFE has a lower thermal conductivity than ceramics. Air has a lower conductivity than PTFE. You may be better off suspending the die.

Also, one last thing. How close is the tolerance of the inside of your die to whatever structure it clamps on to? Maybe the whole root of the problem is that the two halves are not pressed together hard enough. That would prevent a good electrical contact. If that's the case then you are limiting the path of the eddy currents you desire.