17-7 STAINLESS? 304 CRACKING AT WELDS

[JAB64]

I'M LOOKING FOR A STAINLESS ALLOY THAT WILL HOLD UP BETTER
IN A PART MADE OF 16 GUAGE. IT IS SUBJECT TO CYCLICAL VIBRATION AND AN AMMONIA SOLUTION FOR CLEANING (NOT SURE OF CONCENTRATION). DOES ANYONE HAVE EXPERIENCE WITH 17-7 OR IS THERE SOMETHING BETTER (MORE DUCTILE). THANKS IN ADVANCE

 

[stanweld]

Based on what you've stated, 304L should prove reliable. However, you have not provided much information to go on. What material is currently being used? What is the temperature of operation? Is cracking initiating from the ammonia (process) side of the weld or base material heat affected zone? What are the values of the cyclic stresses encountered? Are there chlorides in the solution?

 

[JAB64]

THANKS FOR RESPONDING,
THE PART IS MADE OF 304 CURRENTLY. IT WEIGHS 1.6 LBS. IT IS ATTACHED TO A PNEUMATIC CYLINDER THAT CYCLES AT A RATE OF 60
CYCLES PER MINUTE. MY CUSTOMER INDICATED THAT SOME CRACKS ARE FORMING AT BENDS IN THE PART BUT THE WELDED JOINTS SEEM TO BREAK FIRST. THE MACHINE OPERATES AT 70 DEGREES F. THEY
WASH THE PARTS WITH AN AMMONIA SOLUTION AND THE ROOM IS FOGGED WITH A CHLORIDE MIST. HARD TO SAY WHICH SIDE THE CRACKS ARE STARTING ON BUT I THINK IT IS A COMBINATION OF STRESS AND CORROSION.

 

[stanweld]

It appears that you have a case of chloride induced corrosion fatigue cracking. Since cracking is occuring at both the weld and at cold worked locations, stress releiving heat treatments or shot peening to place the affected surfaces under compression are recommended if you wish to use the same type of stainless steel (304 or 304L).
Alternate austenitic stainless steels such as 316L and 310 or duplex stainless steels such as alloy 2205 might also be used. All have more resistance to SCC than 304 and may well prove adequate for the service.

I recommend performing some screening SCC testing to determine the most economical alloy and or material processing methods for your service.

 

[davefitz]

I am not sure the shotpeening will help at the weld. The welding process leaves residual microcracks, and the properties of the HAZ also may have less tougness than the base metal. Also, there are residual stresses in the weld zone , of the order of magnitude of the yield stress at the PWHT temperature. All things considered, the prediction model for the crack porpagation sould be via facture mechanics with some factor to account for the corrosion acceleration at the crack tip.

The shotpeenign will leave a residual compressive stress at the surface, but it will not close off the microcracks.

 

[stanweld]

JAB64,
As davefitz indicates, failure could also be primarily fatigue related. If you have not performed an appropriate design for fatigue, you should do so - noting that the residual strains from welding are additive to the applied strains in operation. Such an analysis will indicate whether higher strength materials, thicker materials or stress relieving heat treatments are warranted.

Welds should always be considered the weak link in systems subject to cyclic loading.

Additional questions; How is the weld made? Which process? With filler metal or without? Fillet weld, butt weld, corner groove?