Thermal Shock changing material hardness

[goomba]

I have a finned cooling coil constructed from 304L stainless steel. It is installed in an oven that is operated at a temperature of 175°C. During the cooling cycle, chilled water is pumped through the coil. At the beginning of the cooling cycle, water is fed into the coil at the full rate of about 12 gpm. This imposes a thermal shock on the coil and occurs approximately 8 times per day. After operation of approximately 2000 hours, some failures have occured which have been verified as chloride stress corrosion cracking. My question is this: If the coil was properly annealed, is it possible that stress could be developed in the material from the thermal shock? The material analysis showed rockwell hardness in the area of C34. If the coil was properly annealed it should have been in the area of B78.

 

[Metalguy]

Was the high hardness found on the ID and OD? If so, I'd suspect it wasn't annealed. But Cl can cause SCC even on annealed 304. Was the SCC transgranular?

 

[mcguire]

With a hardness of Rc34 you have a severely cold-worked piece of 304. That suggests the possibilty of plenty of residual stress. That combination and a tiny bit of corrosion is sufficient for SCC. Annealing the structure will alleviate the stress and the SCC if not the corrosion.
If you have significant thermal shock it will show up as distortion of the component. Absent this I wouldn't be too concerned about thermal stresses. The unannealed condition of your 304 is sufficient to acccount for the problems.

 

[flames]

The only way to go from a relatively low hardness of annealed 304 to the high hardness that you see is if you go way beyond the yeild strength of the material. You can recheck the design calculations of the exchanger, but hopefully the exchanger was designed so stay well below the yield strength of the material. If you had any doubts I would look for someone with lots of knowledge in stress analysis. Normally your thermal cycling will then flucuate in the linear elastic region and with no real dislocation movement so you shouldn't see much change in your hardness.

Your condition of thermal cycling would if anything led to either fatigue or corrosion fatigue (appears you had 600 to 700 cycles), but this would have been determined in your failure analysis.

 

[mcguire]

Hey Goomba
Here's some quantification of my earlier response. 304 with a hardness of 34Rc has to have been cold-worked around 50% ( at room temperature ). That material should have a yield stress of about 120,000psi. It qualifies as 1/2 hard temper material. This cannot have happened from thermal gradients. It sounds like mixed steel or improper processing.

 

[kenvlach]

A high residual stress as evidenced by the hardness accelerated a problem that would have occurred eventually. This stress was from the original tubing manufacture and HXer fabrication; flames is correct – thermal stresses high enough to have increased the hardness would have resulted in plastic yielding/deformation. However, the elastic expansion and contraction certainly played havoc with the normally passive surface layer, so would create transient SCC even with proper initial annealing.

For some background on this particular HXer problem, see thread338-42610

goomba,
1) were you able to obtain the water analyses or more info on the third party additive package? A number of water treatment chemicals become pitting initiators when boiled down (as during your furnace heating cycle).
2) Am curious, did your management shoot down earlier suggestions from mcguire and myself?

 

[goomba]

Thanks for the responses. I needed to know if the stress in the material could have been caused by the thermal shock because the 180° return bends were supposed to be annealed. It is obvious they were not and I should be able to get some concessions from my coil vendor. I am pretty much required to use the facility chilled water and can't use some glycol-based fluid. The problem is that the supply and return come from overhead. As the oven heats up the water has to be boiled out of the coil which increases the concentration of the solids in what is left in the coil. I am considering replacing the 304L coil with AL6XN.

 

[Metalguy]

Good idea, but I'd use an alloy with even more Mo.