Furnace brazing (silver solder) Cracking 303SS 1

[srivince]

I'm experiencing an intermittent problem with cracking on a 2x2x.750" block of 303 stainless during furnace brazing. The part is machined to have two .257" ports parallel to each other & equally spaced on the edges. The ports have a 7/16-20 thread for a gland nut at each end. Intersecting the ports (on the face of the piece)are four .062 holes into which a stainless tube is brazed. The cracks are appearing adjacent to the tube & outward to the edge & into the thread. The part is brazed at 1300F in nitrogen atmosphere. They are introduced to the oven at room temp. & cooked for twenty minutes then removed & allowed to cool to room temp. with no quenching. If anyone could shed some light on my problem I'd be greatly appreciative. As I said this is an intermittent condition. Can it be related to a bad bar of stock or is there something in the process I could change? Thank you in advance for any help. Vince

 

[unclesyd]

A little more information is needed.

What is the brazing alloy and form?

Are you using a flux and if so what is it?

You cooling in air from 1300° F

Do you have to use 303 SS?
1300°F is awful hot for 303 SS.

 

[metengr]

Could be "liquid metal embrittlement" with the braze metal. I ran into this situation when we were brazing copper leads to 3XX series stainless steel. The Zn in the copper-base braze metal caused the problem.

 

[swall]

Is the cracking occurring on first time through parts or on re-brazed parts?

 

[srivince]

The alloy is 505 (Handy-Harmon) & the flux is Wolverine Ultra-Flux. The silver solder is .032 wire formed into 1/8" OD rings & placed over the 1/16" tubes. We get an excellent joint & even though cracked none of the cells have leaked in test (gas-tight leak-down). In regard to making these parts out of 303, we do this to simplify machining.

 

[srivince]

The alloy is 505 (Handy-Harmon) & the flux is Wolverine Ultra-Flux. The silver solder is .032 wire formed into 1/8" OD rings & placed over the 1/16" tubes. We get an excellent joint & even though cracked none of the cells have leaked in test (gas-tight leak-down). In regard to making these parts out of 303, we do this to simplify machining. Cracking occurring on first run through.

 

[unclesyd]

The brazing material H&H 505 (BAg-24) has 28% Zn + 2% Sn so metengr could be on the right track. Also the flux you are using is a very reactive flux and contains KF and H3BO4 could be a contributor.

It has been my experience that 303 neither likes to be brazed or welded. We never to attempt either process on 303 SS.

Also I seen several instances where the heads of 303 SS screws and bolts pop off after a short exposure to 1000°F.

They are a lot of other alloys that have reasonable machining properties that could replace 303 SS and be better candidate for the brazing operation.

I would get with someone Carpenter and discuss you problems with machining and subsequent brazing operation.

http://www.cartech.com

 

[tomwalz]

First thing we would try is a lower temp alloy with a longer cooling range and a more yielding metallurgical structure. Maybe BAg-1 or BAg-7. (BAg is the American Welding Society designation for silver based braze alloys. 505 is BAg-24)

The thinking is that the difference in coefficients of expansion is ripping the steel. If you use a lower temp alloy there is not so much difference in movement. If you use an alloy that is liquid over a greater range then the materials have more time to adjust. If you use a softer braze alloy then the stress is absorbed into the braze alloy instead of into the steel.

We have a lot of info on the effect of different braze alloys at

http://www.carbideprocessors.com/Brazing/book/12.htm

This focuses on brazing carbide to steel but it will give you some idea of the differences a braze alloy switch can make.

Other thoughts
1. Why use flux in a Nitrogen atmosphere? Can you just clean the parts and not use flux?

2. If the braze alloy is shrinking and pulling the metal apart then you could try a much softer braze alloy. The 505 you are using now is a BAg-24 which is a 50% silver with 2% nickel. This sets up pretty hard. The softest alloy is Bag-1 which is 45% silver with 24% cadmium. This is really soft and the strain caused by cooling and the differences in coefficients of expansion will be absorbed more in the braze alloy.

Cadmium has health problems associated with it but you are using an enclosed furnace.

3. The 505 has a solidus to liquidus range of 1260 to 1290 so it sets up pretty fast. An alloy such as BAg-3 with range of 1170 – 1270 may help. The BAg-1 is 1125 - 1145
The lowest temperature common cad free alloy is BAg-7 which is 56% silver with tin and is 1145 – 1205. The alloy may be weaker than the 505 you are using now but I don’t see where you need much strength in the joint. (Again our reference for stength is ceramic tipped saw blades.)

4. 505 has 28% zinc. There are alloys with less zinc.


Tom Walz



Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessor.com

 

[metengr]

tomwalz;
Liquid metal embrittlement (LME) of stainless steel can occur from cadmium or zinc containing braze metals. These alloys form low melting eutectic compounds that easily penetrate grain boundaries on the surface of an austenitic stainless steel. In some cases, intergranular fissures will occur creating the perception that you are pulling apart the stainless base metal.

Based on experience, I would have a metallurgical examination performed on one of these cracked pieces to either confirm LME or some other mechanism.

 

[tomwalz]

Metengr:

I thought about posting a P.S. saying I was a brazing guy and not nearly the steel guy you are. Then the phone rang and I didn’t get to it. Anyway it is true. I approached it from what I knew which is the brazing side.

I was sure the rest of you knew what you were talking about on the steel but I didn't know the degree to which it affected the problem. I thought a switch in braze alloy would be a simple and easy test to run and might help with the problem. And we have solved similar problems with apparently ripped materials by swtiching to a different braze alloy.

Tom


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessor.com

 

[metman]

tomwalz:

you said;

"Why use flux in a Nitrogen atmosphere? Can you just clean the parts and not use flux?"

I am not sure you can expect good wetting in N2 atmosphere. Even though N2 exlcudes O2 there is no reducing (cleaning action) of braze substrate. H2 on the other hand creates a reducing action and for some silver base braze alloys even in H2 atmosphere we had to use flux to achieve good wetting.

All of the other comments sound on target and if you get Zinc out of the chemistry you probably can braze in H2 without flux and elimiante the LME to boot but concur with suggestion to confrim problem by analysis.

Just my 2 cents.

 

[hacksaw]

you are putting a 0.0625" tube in a 0.062" hole at some temperature, what grade stainless are you using for the tube, and what allowance are you making for differential expansion on cool down?

does the block retain all dimensinal tolenences on cool down without brazing?