High Temperature Oxidation

[SMF1964]

Does anyone have any data on high temperature oxidation (900 F) of Cr-Mo steel as a function of microstructure (martensite/tempered martensite/pearlite)? I have a dissimilar metal weld that has failed (classic DMW creep along the fusion line) but also has a slight flare in the OD profile leading up to the fusion line on the ferritic side (which also shows significant oxidation/exfoliation from sitting in the penthouse environment for 30 years). The flare starts where the heat affected zone ends.

I'm thinking the flare is merely due to more oxidation of the tube away from the HAZ and the HAZ has lost less material, but I can find no published data to support this. Any thoughts?

 

[mcguire]

Microstructure could affect oxidation resistance to the extent that the micro-chemistry changes between these regions. This could happen because the amount of chromium in solution, which does affect oxidation rate, would be less in pearlite than in martensite, making the more quenched areas more oxidation resistant.

 

[metengr]

From my experience with boiler tube failures, the oxidation resistance of Cr-Mo tube materials is directly related to alloy content (typically chrome content), rather than microstructure.

I have not see any published data on high temperature oxidation characteristics related to microstructure for a given alloy content.

How close is the DMW failure to the high crown seal or penthouse floor?

 

[metengr]

There is one other possibility. The oxidation of Cr-Mo tubes that have been exposed to many hundreds of thousands of operating hours at elevated temperature normally starts as intergranular penetration. Once the oxidation occurs preferentially along the grain boundaries, over time they are consumed and the matrix oxidizes resulting in tube wastage

As you go thru boiler start-up and shut-down cycles, exfoliation of the oxide exposes fresh tube material, and the process begins again. It may well be that the microstructure of the HAZ is not as prone to intergranular oxidation. This is only speculation on my part.

 

[SMF1964]

metengr (first comment): The DMW is a foot or less above the high crown seal. The stainless comes up through the floor to the DMW, and low alloy steel is above that.

metengr (second comment): You raise a possibility. It occurs to me also that perhaps the martensitic HAZ, having a higher strength, might be less susceptible to the flexing needed to exfoliate the oxide. On the other hand, wouldn't the modulus of elasticity would remain the same regardless of it being pearlite or martensite? so it shouldn't matter.

Puzzles like this keep reminding me why I got into failure analysis.

 

[metengr]

SMF1964;
At the operating temperature of the boiler tube you will not be in the elastic regime. The tempered martensite will have a different creep rate and strength level in comparison to the unaffected tube material (away from the DMW).

I've got one for you along the same lines. DMW failure from a supercritical boiler in the high crown seal (furnace side). Confirmed the filler metal is Inconel. The crack was along the toe of the weld - except the tube only lasted 8 years. Found three more DMW's during a recent tube repair. Looks to me like fatigue versus classic creep failure along the weld fusion zone on the T22 tube side.

 

[mcguire]

If the carbon steel is flared out near the stainless and assuming the stainless is austenitic, then the greater thermal expansion of the stainless is sufficient to cause the ferritic steel to expand in diameter by creep close to the weld.
This will cause a flaring even if oxidation of the carbon steel is uniform, as would normally be expected.

 

[SMF1964]

metengr: this failure has fatigue associated with it as well - see my comment at the end of the last note.

mcguire: Would your mechanism apply to both the ID and the OD? In my case, this is an actual increase in wall thickness.

 

[davefitz]

One of the main issues with DMW's is the different coefficient of thermal expansion between SS and Cr-Mo: as you heat the tube from ambient to 1050 F +, the SS expands 50% more than the ferritic. This tends to form a high shear stress, and failure usually originates in the mid wall ( consistent of the location of max shear stress).

The flaring you discussed may be related to the Cr-Mo creep relieving to reach the same hot OD as the SS. Although the SS might have a lower yield stress, the Cr-Mo has a much lower creep strength.

Goodlong term availability of the DMW's has ben found by using a Nickel based weld filler ( incolnel) , but using a very wide weld widht and applying the weld using an automatic orbiatl welder. It had been found that if the weld was applied manually, the welder would decrease the amps to limit weld splatter, but would then cause insufficient weld penetration.

 

[mcguire]

SMF1964
Yes, the differential thermal expansion would cause the ferritic steel, which has much lower yield strength at high temperature, to grow in the thickness direction also.