321 SS to 9cr

[rmillercwi]

I have a client asking me to research a 321ss to 9cr repair procedure. After doing my own research, I will also consult with a metalurgist as well as a selected weld engineer and compare notes.

A crack was found in a fairly heavy 1.5" thk. x 12"-16" long 321 stainless lug welded to a 9cr pad welded to a 9cr pipe. Existing filler material or weld procedure unkonwn. Cracks are small but can be seen with the naked eye. My guess is this lug is in an approximate 1200f environment or possibly higher continuously. I do not know how often it cycles.

The 321ss Lug to the 9cr pad weld is a full penetration weld. Not sure why they chose this combination (I would have attached a 9cr lug/s directly to the header with a hole in it for pinning the stainless support hanger)

Initial research has me leaning towards grinding out the existing weld cracks to penetrant test acceptable, Apply 400 f preheat with ER preheating (if possible) GTAW process using,ERNiCr 3 filler, slow cool to ambient prior to heat treatment followed by normal 9cr heat treatment (1350-1425f, couple hour soak, slow cool under insulation).... No wet type NDE until after heat treatment. The rub with this is perhaps the 321 cant handle the heat treatment.

I see some have recommended a butter weld (presumably a 1/4" thk layer of enco 82 butter weld to the 9cr pad and heat treat) in new construction. Would this then eliminate the need for heat treatment of the lug to the butter weld?

This is not possible without complete removal (which whether they realize it or not is going to happen sooner or later). Their intent for now is to buy some time for a permanent fix later.

 

[metengr]

I would strongly recommend the following to avoid working with inadequate information;

1. Have a small sample of the full penetration weld removed and pad weld removed on the 9Cr base material for chemical analysis by a met lab to confirm filler metal compositions.
2. Once the filler metal compositions are known, you can decide on a specific repair option and filler metal selection regarding the 321 ss lug to the 9Cr pad.
3. If the cracks penetrate into the 9Cr pad, you need to know the filler metal of the pad, as deposited.
4. I can see several options once the filler metals are identified.

Note: In lieu of lab chemical analysis a portable alloy analyzer may work.

 

[rmillercwi]

PMI was done on the lug, pad and weld. The lug and pad checked out @ 321/9cr but "no match" on the filler metal. Not unusual. Do the results below tell you anything?

Ti - 6.47
FE - 8.81
Mn - 6.89
Ni - 61.75
Cr - 14.85
Zn - 0.106

 

[metengr]

Ok. Keep in mind the PMI is qualitative. This seems close to an SFA 5.11 ENiCrFe-3. This is a nickel-base filler metal so from that standpoint, your selection of Inco 82 would be acceptable.

Regarding weld repair, I would remove the crack(s), ensure removal by LPT. Use Inco 82 to weld a butter layer over the 9Cr pad. PWHT. Complete the remainder of the weld using Inco 82. Perform a LPT after completion.

 

[rmillercwi]

Thanks Met. Presumably the 321 lug can handle the standard 9cr heat treat temps with no ill effects or are you insinuating removing the lug entirely, doing the butter weld...heat treat and then re-install the lug welding it with inco 82??

 

[metengr]

No need to take the lug off unless the cracks are deep into the full pen weld and/or into the 9Cr pad material. If necessary, the 321 lug material can take the PWHT for this application. The butter is in case you decide to remove the lug. I am more concerned about the 9Cr pad material for PWHT.

You may get lucky and only need to repair the crack excavation within the full pen weld using Inco 82 and not touch the 9Cr pad material. Thus is why you have options.

 

[rmillercwi]

Thank you. One more question and I let this go. So what would happen if the plant decided to do nothing more than grind out the cracks to LPT acceptable, heat it up to to 400f and weld it out with Inco 182 (doing all else by the book...good electrodes and controls, ER heating, slow cool under insulation) and let it go). No heat treatment.

Doomed to fail immediately and perhaps create a worse condition than exists today? Just asking your opinion. Keep in mind the cracks are small and all on the 9cr side of the weld at the toe. Perhaps due to nothing more than the final cover weld was just a little bit bigger than the butter weld (which I suspect is what happend here.)

 

[metengr]

rmillercwi;
See my last sentence below from my last post to you...

You may get lucky and only need to repair the crack excavation within the full pen weld using Inco 82 and not touch the 9Cr pad material.

The above would be your best scenario, preheat and weld repair without touching the 9Cr pad material.

IF the crack enters the 9Cr pad material, you should PWHT because of the material composition. This is an air hardening grade of alloy steel. Perhaps you can use a temper bead technique with the Inco 82 to attempt to temper the 9Cr heat affected zone. This is marginal at best and would require a WPS qualified.

IF no PWHT is considered, you would most likely crack again and possibly into the 9Cr base metal, which would not be good.

 

[NBrink]

Not to say it's the way you'd want to do it, but it's not that uncommon to skip PWHT for tube skin thermocouple welds on 9 cr heater tubes, just give them a high preheat and a slow cool and hope it tempers in the furnace.

This seems a similar situation. 1200 F is right up there...

Nathan Brink

 

[metengr]

Apples and oranges comparison. Depending on the extent of weld repair to the 9Cr pad material, the heat affected zone could be large in comparison to a very, very small HAZ from capacitor discharge welding to attach TCs.

 

[NBrink]

I mean, the weld for a vee pad thermocouple is tigged out, good inch and a half long, maybe half an inch across. Not a huge HAZ, but a heck of a lot more than a capacitor discharge weld. Usually done with 1/16 wire...

I don't see it as hugely different if the extent of the weld repair isn't too significant...

Nathan Brink

 

[weldstan]

While the repair can be successfully made, the lug material selection is woeful. Thermal fatigue will still be an issue as to service life. There will always be a temperature differential during both startup and shutdown situations and I suspect it will be considerable.

 

[davefitz]

You need to re-design the lug /pad to avoid the thermal stress due to the difference in the thermal expansion coefficients and very high temperature differentials, and to avoid the flux of elements across the DMW HAZ.

One possible change is to weld the 321 lug to a 321 sleeve , and the sleeve held in place( mechanically) by small 9Cr lugs.

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick