Induction Brazing

History

There is greatly increased breakage coincident with the switch in brazing alloys from BAg-3 (50% with Cadmium) to BAg-24 (50% Cadmium free).      

Overview

Successful brazing is the joining of parts to form a composite.  The stronger and more uniform the join is the stronger the composite material will be.  For more information on How to Braze see our Special Brazing Project articles.

Essential Points

• Cleanliness of the parts being joined- See Cleaning Steel before Brazing for more information on this topic 
• Selection of the proper braze alloy 
• Preservation of the cleanliness during heating - fluxing 
• Heating cycle 
• Handling of the parts during and after heating

Analysis

1. Typically the BAg-24 (50% Cadmium free) alloy makes a join 30% – 40% more susceptible to tip loss and breakage than the BAg-3 (50% with Cadmium) alloy. 
2. Very little flux was being used.   Due to thinning and sparse application the amount of flux used was perhaps 10% of what should be used.          
3. The heating cycle was much too long.  The parts got much too hot and were held at the too high temperature for much too long a time.   
4. The parts were worked around much too much.   Parts at temperature should be moved once to break the surface tension and release any trapped flux and gases then they should be positioned in their desired final location and held there until cool.   
5. The carbide does not wet well.  It has been sandblasted but not surface treated.     
6. The heating coil is too far away from the parts and is the wrong shape. 
7. Too much braze alloy is being used and some of it is placed improperly. 
8. The steel fixturing draws a great deal of energy thus slowing the cycle time. 
9. We didn’t test for steel cleanliness because we didn’t have a can of Easy Off oven cleaner.   However the steel wet well once we used enough of the proper flux.     
10. The operators do not know when to stop brazing.  They put a piece of braze alloy on top to tell them when it gets hot but even after that melts they keep working the parts. 

If nothing else the parts are being severely burnt.  It is possible to braze these parts more successfully just by following the immediate steps below.  

You can dramatically improve cycle time with proper coil design and placement.

You can get a better part by using the “as cast” steel holders than the machined holders.

I think the braze alloy is being oxidized in the joint.  Furthermore I think the lack of fluxing and the overheating are extreme enough that the copper is being burnt away in the corner giving problems.  The carbide is not very receptive to wetting so there is very little joining taking place.  

Wetting of the corner

The braze alloy flows to the heat.  The coil is placed below and behind the braze joint.  The heat moves through the steel into the braze alloy.  The way the part is heated now the braze alloy will be drawn to the back of the part, towards the hot steel.  With the length of the current heating cycle and the movement of the parts the braze alloy would have a great deal of opportunity to flow back and away from the front, especially the front corner where the damage occurs.    

Flux

The parts are horribly under-fluxed.  The brazing flux is much too heavily diluted so it is impossible to use enough of it.

Flux traps oxygen.  It loses its effectiveness with time, temperature and exposure to oxygen.   The problem corner is last heated so it is exposed to oxygen at high temperature for a long time.  I think the flux loses its effectiveness well before the braze alloy in this area comes up to temperature.  

I believe that much of the braze alloy in the effected corner may be drawn back into the part leaving the copper exposed with only a very thin layer of braze alloy on it.  As the braze alloy comes to full heat and is held at full heat for much too long time I think it may actually burn away. 

Below are pictures from an experiment.   I put piece of trimetal Shim under the coil with flux.  I overheated it and you can see the picture at right.  This doesn’t prove that this is happening but it establishes that it could be. 

Additional Steps

1. Give the operators clear indicators for when to stop brazing
2. Use a different joint design
3. Use purified flux
4. Switch to a stronger cadmium free alloy.
5. Use different coil placement.
6. Use a different coil design
7. Use treated carbide
8. Use ceramic fixturing
9. Use “as cast” steel parts
10. Investigate other carbide grades. 

Explanations

-Give the Operators Clear Indicators for When to Stop Brazing 

-Use a Different Joint Design

Typically this sort of joint would have braze alloy all the way around.    Currently trimetal braze alloy is put under the carbide parts and between them but nothing is put behind them.  

The extra wire put on top of the carbide serves very little purpose.  The reason given is to tell the operators when the parts are hot enough.   However the location where heat is essential is where the braze alloy, steel and carbide join.  Furthermore, the operators do not quit heating the parts when this piece of alloy melts. 

-Use Purified Flux

We saw the difference it makes in the test.  Do not dilute the flux.  If, for some reason, they must dilute it then use deionized (not distilled or purified or spring) water only.  We find Culligan to be the best source here. 

-Use Different Coil Placement

I think you could cut he cycle time by 75% with proper coil design and ceramic fixturing. 

I think that coil the current placement could be contributing to the failure.    The way the coil is placed now the corner where the failure occurs comes to temperature late in the heating cycle.  There is a possibility that the braze alloy is drawn off the trimetal shim and flows towards the heat at the back of the part.   This is what we saw on the part Bob broke.   There was a puddle of alloy on the inside middle.  

Move the coils as close to the work as possible.  Switch the big, round, single coil for smaller, square-sided coils with multiple winds. 

-Coil Design

The whole coil generates energy that could be used to heat the part.   Now maybe one-quarter to one-third of the coils is being used.  The fixture is too far from the machine but the big problem is that too much of the coil sticks past the part and the coil is too far away from the part.  The reason given for using such a long coil is to avoid overheating the end.  Here are two coils we use that do not overheat the end because they are open.  These coils also double and the tubing is square.  

This is the coil I used to test your parts.  In the left picture you can see how close the part is to the coil.  The inverse square law comes into play here so that every time you double the distance from the coil you cut the effective energy by 25% (4 times the distance means you only have 1/16 the available energy.)    

This not the coil shape I would use for production for this shaped part.  Too much energy is wasted on the high hoops.  However here it served to heat the parts very evenly by heating the full length of both sides. 

 -Use Treated Carbide

The carbide wets very poorly.   The top part in the pictures below has a piece of cadmium free braze alloy wire on it.  It melts into a lozenge shape.  Next to it is a piece of High impact braze alloy.  These parts were severely overheated.   You can see the uneven flow and the bunt appearance.  

Most of the surface coverage now comes from a wiping motion, what Bob refers to accurately as a “butter knife” effect.  This is important for two reasons.  First, if the braze alloy doesn’t stick to the carbide then the parts are more likely to come apart.  Second, the brazed structure is strongest in every way when it is uniformly bonded.  Having parts that do not bond fully is like having a car with three good shocks and one bad shock.   

High Impact alloy on left  - 50% Cad Free on right (now used)

-Ceramic Fixturing

The steel fixture is sucking up a lot of energy.  Ceramics would work very well.  The only drawback might be that ceramic fixturing can break while metal will not.  The type of ceramic we are discussing here is similar to kitchen bowls.  They will break but they are not fragile.  The fixturing is pretty inexpensive and we keep extras.  Lead time is usually a week for replacements if necessary.  

You can easily grind ceramic fixturing within 0.001” and it will retain its shape and dimensions during heating. 

The easiest way to do R&D is with soft firebrick from any ceramics supply store.    It is very soft an easy to work.  You can cut it with any saw.   We typically do mock ups and test with the soft firebrick then go to a ceramics shop and have them build the fixture for us.