welding of martensitic SS steam turbine nozzles 2

[GordonF]

Dear all

We have a steam turbine nozzle block that is made from a martensitic stainless steel (similar to 422). The nozzle assembly is fabricated from individual nozzles which are welded together. The weld material is a nickel based alloy believed to be Inconel 182.

We are currently investigating a repair of the nozzle assembly. My question arises to the reasons why a nickel based alloy was used in the fabrication, rather than a 410 martensitic stainless bearing in mind it is possible to PWHT.

Any comments would be appreciated.

Regards

 

[GRoberts]

I don't see why they would choose to repair with nickel based filler. The main reason to PWHT the nozzle after welding is to temper the HAZ anyway, so using a different filler does not aliviate that requirement. We always use E410NiMo to weld martensitic stainless (CA6NM). It is tougher than regular 410 filler, so it wouldn't hurt to use the slightly overmatching filler even if the base metal does not have the Ni or Mo. The good news is that the previos fabrication with nickel based filler does not restrict you from using a PWHT yourself.

 

[GordonF]

GRoberts, thanks for the reply. Following your response, I look up the specifcations for the E410NiMo filler. The maximum PWHT temperature specified is 620C (1150F)whereas the tempering temperature of the base material is in the range of 690C. I have also looked at that other weld procedures that have used E410NiMo and they have been PWHT above 620C, but it is not clear if this has been done intentionally. Any thoughts?

 

[GRoberts]

Gordon,

I think that all the procedures that you see call for a 1150F PWHT because that is what the base material specification calls for, and is the most common reccomended temperature. However, we do double tempers with this filler metal when required to be 23HRC or less, with the first temper at 1225-1275F, and the second temper at 1050-1150F. Of course there is nothing wrong with the standard 410 filler metal except it is more crack sensitive. What base metal specification are you working with?

 

[GordonF]

GRoberts

The material is a Parsons propriety alloy with composition of C 0.07-0.15, Mn 0.3-1, Cr 11.5-13, Mo 0.40-0.80, V 0.10-0.25.

We have tried the 410 filler with a minimum preheat of 250C which we maintained for 2 hours after completion. There was no cracking on cooling and a hardness traverse indicated a maximum of 430 HV before PWHT.

Thanks for you assistance

 

[mahadhatu]

GRoberts,

I am working with CA6NM alloy with low carbon content (<.03%) I am trying to determine the PWHT required after welding with ER410NiMo. The maximum hardness required is 250 BHN. As you rightly mentioned in one of your replies, the ER410NiMo calls for a PWHT with temperatures not exceeding 1150F. Would you still recommend a double tempering HT? Also would the lower carbon content help gain the 250 BHN max requirement?

The welding to be done on these castings is mainly for repair purposes.

Regards

 

[stanweld]

Beware of using ER410NiMo. This filler metal does not have the creep rupture properties equivalent to the high temperature martensitic stainless steel that you propose to repair.

 

[corvin]

The 410NiMo has a similar composition with the 410 filler wire, except for the much higher Ni content (>4%) which provides a less crack sensitive weld deposit. The addition of four percent Ni has several metallurgical effects:
- Decreases the possibility of delta ferrite formation
- Lowers the martensite start (Ms) temperature by about 125F. This is partially offset by the lower carbon specified for the 12Cr-4Ni composition
- The lower critical temperature Ac1 is lowered approximately 345F, below 1200F
The first of these effects, elimination of delta ferrite is beneficial as it results in a homogeneous martensitic structure, free of islands of weaker, less tough delta ferrite hence better fatigue properties. However, the delta ferrite as grain boundary films has the advantage of reducing creep crack propagation rate.
The extent of lowering the Ms temperature is important because it controls the final structure obtained. It is desirable to have an Ms temperature above 400F as the transformation temperature range (Ms to Mf) is a least 300F. From this point of view the 410NiMo is anyway less of a problem than the 422SS base material, anyway.
Lowering the Ac1 temperature can be a problem as it limits the use of high stress-relief temperature.
If the Ac1 temperature is exceeded during the stress relief, austenite is reformed which transforms to untempered martensite upon cooling. The result is an increase in strength and hardness with a consequential loss in ductility.
A solution to the problem is a double PWHT, a higher one followed by a second one that doesn't exceed 1150F (which takes care of the re-transformed austenite). In this particular case is important to determine if the higher PWHT temperature employed for the 422ss is not over-tempering the 410NiMo.
In conclusion, Inco 182 may be a good choice as long as undermatching of the properties of the base material is not a problem. This may be a solution for small repairs.
For a matching filler material, Thermanit MTS4 from Bohler may be the right option. The potential for hydrogen cracking is higher than with Inco 82, obviously.
The hardness of 410NiMo PWHT-ed to 1150F is around 260-320HB.

 

[mahadhatu]

A low carbon (0.03%) CA6NM after welding and PWHT gave me inconsistant hardness numbers. The hardness on base as well as HAZ was around 246BHN but in the weld it was 277BHN.

I used E410NiMo-16 covered welding electrode to weld 1" thick plates and a double tempering PWHT ( 1250F, AC and 1150F,hold 8h, slow cool @100F/h). Does any one have a clue as to why the hardness was not consistant?
Is the 410NiMo capable of attaining a hardness lower than 250BHN?

Thank you.

 

[Carburize]

mahadhatu- your results are typical of many others experience with this alloy. Some have been successful in reducing the hardness by multiple temper post weld heat treatments - 3 or 4.
Frequently suppliers will offer to supply purchasers with fabrications which are welded under the processes and controls you outline but with no guarantee on final hardness because of the problems.

 

[GRoberts]

We regularly obtain hardnesses below 23 HRC (NACE MRO175 requirement), which would be about 243HB according to ASTM A370. A couple questions on your heat treat. Did you cool to less than 100F between heat treat cycles? This is required if you are going to meet the hardness requirements. How long was the 1250F temperature held? Also, find a filler metal with as low a carbon as possible. Less than .02% should help, but it can be done with .03%C filler also. Carburize's suggestion is also good. Mutiple cycles bring down hardnes and improve toughness also. Longer times at temperatures help too, but don't improve toughenss as much as multiple cycles.

 

[mahadhatu]

Thank you GRoberts and Carburize,

The heat treatment was like this,

Heat to 1250F, hold 4 hours
Air Cool to room temperature
Heat to 1150F, hold 8 hours
Slow cool @100F/H to 600F, Allow to cool in furnace without openning the furnace door. Final temperature - 250F when removed from furnace.

The filler metal has .03%C

Also, this weld test sample was previously treated once with an exactly same double tempering cycle. I am unable to understand why the base metal and the filler metal react so differently to the heat treatment when the composition is very identical.

The NACE MR0175 says that the hardness permissible is 255BHN when measured using a Brinel Hardness tester which has been empirically determined to be equivalent to 23HRc.

Thank you

 

[Carburize]

I do not have a copy of the NACE document available but I believe that the Rockwell C number is the definitive hardness and is probably somewhat easier to meet than using other hardness tests and then converting.

 

[stanweld]

GordonF,
The Parsons alloy is an X20 derivitive. Bohelr Thyssen manufactures matching SMAW electrodes under trade name MTS 4 and bare wire fillers under MTS 4 Si. PWHT is 1400 F for 4 hrs. Recommended preheat is 450 F.

 

[GRoberts]

Some things you might try is increasing the time of the 1250F cycle to 8 hours. Also, how close is the temperature control on the furnace. If there is overshoot or too much deviation and the temperature gets much above 1150F, some of the martensite could transform into austenite and then be untempered martensite when you get finished. Depending on the actual heat to heat variations on the filler metal, the lower critical temp could be as low as 1165F. Try 1125F for the second temper and see if it comes out any softer.

Carburize is also correct in the NACE MRO-175 lists 23HRC as the hardness limit, any other hardness scale would be a conversion from HRC. I haven't used Brinell on a weld, but HRC is easier to meet than vickers or HRA/HRB.

 

[mahadhatu]

GRoberts/Carburize,

Here's what I did, after welding CA6NM plates with ER410NiMo welding rods, I PWHTed it at
1250F for 4 hours,
Air cooled to room temperature.
tempered at 1125F for 8 hours,
furnace cooled at ~150F/hr to about 300F and then in air to room temperature.

The result, hardness in base metal: 241-250BHN, weld: 255 - 270BHN. The hardness in the weld never came down. Do you have access to cooling curves of this alloy?

Thank you.

 

[GRoberts]

Sorry, mahadhatu, I don't have any cooling curves for you. I'm wondering though, were you not able to hold the 1250F for longer than 4 hours? Also, I haven't had a need to use it, but some people suggest cooling to below room temperature between heat treat cycles.

At this point, it looks like you might have at least 3 options left to try.
1. Keep fiddling with the current heat treatment. Trying a longer time at 1250F, or cooling colder between cycles. I have also heard a rumor that waiting 24 hours between cycles helps, but never tried it. I have recently (in the last few weeks) tried a double 1125F, and it did improve the toughness, so it may affect hardness too. I also tried 1150/1125, but it didn't make a tougness difference compared to 1125/1125. This would be after the 1250F of course.
2. Find a filler metal with lower carbon.
3. Try a different heat treatment, such as Nomalization, followed by a combo of 1250/1125.

If you have any of your previous test plates left though, can you experiment with the heat treatment on remnants of those tests instead of welding new ones? It could save a lot of time & $$.