Diffusion Bonding 1

[bmoorthy]

Understand that Diffusion bonding is not being used widely as an method of welding for fabrication of vessels and heat exchangers.

Currently we have a smally situation regarding the fabrication of Diffusion bonded item and hence i would like to know more on the application of diffusion bonding on various material and the parameters for diffusion bonding for various material.

Request some links that will give additional inputs.

 

[metengr]

Google until your fingers hurt. By the way what does this mean

Currently we have a smally situation...

http://www.google.com/search?hl=en&q=diffusion+bonding+of+metals&btnG=Google+Search

 

[bmoorthy]

Google did not help.

The parameters tht control the diffusion bonding are ot discussed.

on "smally", It was an error.

I will be more careful next time

 

[btrueblood]

Well, at a first blush:

materials in contact
cleanliness of surfaces
contact pressure
surface finish and flatness
temperature
time

We made some precision flow devices with intricate channels (think fluidics) that were stacked thin pieces of etched 304 stainless steel. The parts were first plated with a "strike" of copper, then loaded into a compression fixture and placed in a vacuum furnace for several hours at high temperature (over 1000 F). We also tried a hydrogen atmosphere and no copper coating, but had trouble with leakage using this approach.

There is information on diffusion bonding in the ASM metals handbooks, I believe in the joining volume.

 

[bmoorthy]

Btrueblood

Looks like we are taking about same component.

Ours is an exchanger (Heat Exchanger with cross flow). Thin plates are stacked oneover the other.

The plates are etched to provide fluid channel

What coating would you sugbest on 6 Mo material and which atmosphere is best suited.

I know the temperature of the soak is 0.5* Melting point nd the the process is non pressure process.

 

[unclesyd]

Here is a paper on microchannel heat exchangers that you might get some leads from.

http://www.hydrocool.com/Files/HighEffHEX.pdf

Here are many pages of reports from NASA on diffusion bonding.

http://ntrs.nasa.gov/search.jsp?N=0&Ntk=all&Ntx=mode matchall&Nt

 

[btrueblood]

Well, a similar component at least. Talk to Scarrot Metallurgical in California, they helped us with our parts. Try other brazing/heat treatment shops also, their staff will normally include metallurgists with varying degrees of knowledge on the subject. I'd suggest a copper strike plating and vacuum furnace, but YMMV, and you may or may not like the copper (even a thin line) in your application due to corrosion/reactivity concerns. When you say it's a non-pressure process, do you mean there is no compression load on the plates during the high-temperature bake?

Lastly, we had good luck with very thin foil sheets of braze alloys sandwiched between (simpler) parts. Ni/Pd and Ni/Au alloys.

 

[unclesyd]

One thing I found out about diffusion bonding, which in our case was unintentional, is that if there is a potential for two materials to form a bond it is greatly accelerated by the slightest of movement at the bond line. From empirical evidence this movement has to be measured in nanometers. This was observed in a system where we had a metal gasket with a high clamp force sitting at temperature. If the fixture just sat there was a very small amount of random bonding, less than 10% of the contact area. If we pressurized the system at temperature to say 200 psig for a leak test we got an approximate 90% bond area. As the system was designed for 2500 psig the actual relative movement of the components would have to be extremely small. We didn't test any other pressures or time at temperature under load. This was verified by several tests at 200 psig. Our problem was that the bonding caused problems later on in our process. We stopped the bonding by oxidizing the metal gasket surface and also tested a flash plating, both stopped the bonding under any conditions the components would see in our process. The whole problem went away later when we went to a self-energized gasket.

Here is another approach to getting a very thin bond line, in the order of nanometers that you may want to investigate.

http://www.rntfoil.com/technology/tech.html