P33 to P1 Repair weld?? 1

[bcase14]

We have a heat exchanger that has started to leak due to a crack/ separation of the tube header plate to the shell. This is causing a contamination of the lake water that is used to cool the Cooling water.
The plate is P33 (silicon bronze) and the shell is P1. It has been welded before (made in the 1950's) and the original weld has cracked in the HAZ on the carbon steel side.
We have tried to contact the manufacture of this but they have been sold out and moved so many times that the original WPS can’t be found.
We have tried several times to create our own PQR/WPS for the repair, however it always breaks in the HAZ on the bronze side and we have been unable to qualify this procedure.
Any help??

 

[metengr]

What welding process and filler metal are you using for this dissimilar metal weld joint?

 

[Vinch]

I think the correct term to use here is brazing. In this case, welding is impossible for such dissimilar materials.
HAZ cracking is usually not a problem with brazing.
More details please.

 

[metengr]

Welding is possible using GTAW or GMAW with ERCuSi or
ERCuAl-A2 filler metals.

 

[bcase14]

We are currently using a procedure given to us by the contractor who was trying to make the repair in the HEXs. The rod we used was first an GTAW aluminum bronze wire but now they are trying to qualify the procedure with silicon bronze wire. i am tring to get my hands on the WPS now.

 

[bcase14]

Ok I have new info on this
GMAW, ERCuAI-A2, .035 DCEN, AMPS145-195, VOLTS 23-25
They are testing it as a double bevel weld. I think the bend test will never work because of the dissimilar metals. Is it possible to test it as a fillet weld (as that is what the joint to be repaired is) and do a micro or macro etch test rather than bending?

 

[metengr]

You can use a less severe bend if you can demonstrate the minimum required ductility is met,(e.g. 10%). In other words, you can use a more generous radius fixture to reduce the bend strain from 22% to 10%, if this is all the joint can withstand because of the dissimilar material weld joint.

 

[bcase14]

Where can I find the exception that will allow me to the less sever bend?
Also found out that when they preheated the test to 325 it would break on the bronze side and then when welded with no preheat it would break on the P1 side either way it was in the HAZ.

 

[metengr]

Here it is

ASME Section IX

QW-161.5 Longitudinal-Bend Tests. Longitudinal bend
tests may be used in lieu of the transverse side-bend,
face-bend, and root-bend tests for testing weld metal or
base metal combinations, which differ markedly in bending
properties between
(a) the two base metals, or
(b) the weld metal and the base metal

 

[metengr]

Also, look at QW-466.1, TEST JIG DIMENSIONS

Scroll down to
Materials with 3% to less than 20% elongation

Dimensions are provide for the plunger radius and shoulder distance.

 

[muld0020]

The ERCuAl-A2 wire is the way to go. You must be careful when welding the joint since the thermal conductivity of the Al Bronze is roughly 5 times that of steel. You may want to look at preheat on the Al Bronze, or the actual direction of the welding consumable to point towards the Al Bronze is sometimes enough.

Preheating the Al Bronze causes more heat to flow to the lower-conductivity metal during welding (because of the larger temperature difference) and thus leads to a more even heating of the weld assembly. Heating the higher-conductivity base metal also reduces the cooling rate after welding, and this helps reduce post-weld thermal stresses.

The balance of heat is important in these joints.

I would also recommend GMAW or GTAW w/ GMAW fill.