Mass Effect in Normalized 4340 1

[PaulDoherty]

Does anybody have a feel for what the expected hardness would be for 4340 in a 9-1/2" diameter after normalizing at 1600 F? All of the mass effect data I have come across so far just goes up to 4" diaemeter. Thanks.

 

[redpicker]

I'd expect around 248-321 BHN. I have to give it a wide range; it's kind of difficult to predict, since the cooling rate isn't controlled.

rp

 

[PaulDoherty]

Thanks RP.

This part is a large pintle - a cylinder with a tapered nose. Our customer will be welding the base of the pintle to a large plate and has specified the normalizing instead of quench & temper for weldability, but wants the wear surfaces of the pintle as hard as possible.

Do you feel that it would be wise to pre-machine the nose area before normalizing in order to reduce the mass in that area and minimize tha amount of (presumably harder) stock that we are removing after the heat treatment?

Paul

 

[TVP]

Paul,

You may find that the normalized microstructure is preferable to a hot-rolled or annealed microstructure even though the hardness may be a little higher. Annealed microstructures with some spheroidized carbides will have a hardness in the low 200's HB, say 210-220 HB. Tool wear may be lower with this treatment, but the resultant surface can be less than desired, due to the "gummy" cutting behavior of high fracture toughness steels like 4340.

 

[PaulDoherty]

Thanks TVP. My thought on pre-machining was more related to getting the nose wear surface as hard as possible, assuming that the normalized microstructure transitions to more of an annealed microstructure toward the center of the part, because of it's large mass. By premachining, we would be removing less of the harder exterior material in our finishing operation and presumably have a better local cooling rate in the nose area due to reduced mass.

I agree with you that machinability is actually 'better' in these alloys when they are harder because we stand a better chance of breaking the chip.

Paul

 

[redpicker]

I don't see why a normalized structure will improve weldability. The weldability of 4340 is poor and will require a relatively high preheat and postheat to prevent cracking of the HAZ, regardless of the prior microstructure.

With the normalized bar, you aren't going to see a big difference in hardness through the diameter, so machining prior to heat treatment isn't going to do much.

rp

 

[unclesyd]

i agreed with TVP in that the center hardness will be around 220-225. This would be essentially an annealed condition.

Are you going to heat treat the bar after the welding?

If so what is the heat treatment and desired hardness of the finished part?

If you are not going to heat treat after welding there might be a way to overcome the low center hardness by either rough machining the bar prior to normalizing or having the point of the pintel a close to the end of bar as possible. If you rough machine the bar you will get the same effect in a cross section that is equivalent to that size bar. If you machine the point as close to end of the original bar as possible you will get the hardness corresponding to the dia of a bar corresponding to the depth form the end.

 

[PaulDoherty]

Thanks to all for the input. This is a great group.

The fabrication drawings call for the pin to be shrink fitted into 3" deep bore in a 4" thick A572 Gr 50 plate that is square shaped (the plate, not the bore), roughly 5' x 5'. They show a 3/4" fillet weld (all around) between the pin & plate. I am not responsible for this aspect of the job. My thinking is that this pretty much rules out any possiblity of a Q&T process after welding.

As for weldability, here the the rational (not mine):

"The problem with Q&T is that the crystal structure is BCT (Body-Centered Tetragonal) martensite which tends to crack when welded to BCC (Body-Centered Cubic) ferritic steels (A36, A500, etc.). In the normalized condition, the structure on 4140 is BCC ferrite. That makes welding fairly trouble free. It is the crystal structure that causes the weld problems, not the hardness."

At the time, 4140 was being proposed and we have since determined, as most of you would have no doubt warned us, that in this diameter, the normalized hardness is not appreciably different that the annealed hardness.

My thought is that with 4340 and a rough machine > normalize > finish machine process, we could get the wear surfaces as hard as possible while keeping the allegedly weldable normalized structure.

I have attached a drawing if you are interested.

 

[TVP]

FYI, the rationale that you have been provided is not correct, at least not for 4340 that has been tempered at a temperature greater than ~ 200 C. It is true that quenched steels with sufficient carbon content and hardenability have a crystal structure change from FCC (austenite) to BCT (martensite). However, the tempering process converts this BCT (martensite) structure back to the BCC structure, meaning ferrite + cementite, it's just that the cementite particles are so fine that they are not revealed using light microscopy. Alloy carbides are also precipitated in grades like 4340 when tempered in the typical range of 300-700 C.

 

[metengr]

The welding of the 4340 pintle will require a local preheat of 450 deg F and post weld heat treatment of 1200 deg F to provide for adequate tempering of the 4340 base metal heat affected zone. I would limit the PWHT temperature at 1200 deg F to avoid hardening the A 572 Grade 50 base plate.

I would consider induction hardening the wear surfaces after PWHT in lieu of re-heat treating. 4340 alloy steel has the necessary carbon content for surface hardness and excellent hardenability that can work in your favor.

 

[redpicker]

I have a problem with the rational for weldability. First of all, in the Q&T condition, the crystal structure is BCC, same as with normalized. The only difference is the size and distribution of the carbides. Yes, martensite has a BCT structure, but on tempering, it forms ferrite and carbide. This occurs, literally, within seconds at temperatures above 500 F.

Second, welding on 4340 (or 4140, for that matter) is not "fairly trouble free" unless you consider preheating the pintle to 500 F and post-heating at 1000 F can be accomplished without any trouble.

It isn't the hardness or crystal structure that causes the welding problems, but the chemistry of the base material. The heat affected zone of the pintle will be reaustenitized during welding and when it cools, it will form a hard and brittle martensite. This will crack on cooling if the material does not have enough preheat to slow the cooling rate. Furthermore, if it isn't tempered with a post-heat, it will likely be too hard and brittle for service, so even if it doesn't crack on cooling, it is likely to crack in service.

IMO, you should use Q&T material at the hardness you need and use proper welding techinques. But, since that doesn't seem like an option, your second best is use normalized material as you have been told and when then that part fails, consult with a welding engineer for proper welding procedures. No need worrying about the hardness since the problem is going to be with the welding.

rp

 

[PaulDoherty]

Thanks for your replies, TVP, metengr & rp.

I had suggested A434 Cl BD material in liu of the original confusing & contradictory spec, but I was too intimidated by the title of the person that gave the above (excerpted) respose to the RFI to press the issue. Bolstered by your suggestions I will now reassert my position.

I hope to be able to contribute to this forum in the future.

Paul

 

[PaulDoherty]

TVP, metengr, rp:

Any comments on the following response to my suggestion to abandon the normalizing approach and use a quench & tempered material with proper welding procedures? Again, this was offered as a way to achieve hardness of 30-32 Hrc for 9-1/2" dia. 4140 after normalizing.

"The following is offered as a potential solution for consideration by the contractor: Accelerate cooling using a Molten Salt Bath (suggested bath temperature 1000 degrees F), followed by air cool. This process may allow the material to achieve the required hardness without forming martensite, which may pose a problem for welding"

Thanks in advance for your time.

Paul

 

[redpicker]

In my opinion, way too much trouble for too little gain.

It isn't the presence of tempered martensite in the microstructure prior to striking the arc, but the formation of untempered martensite in the welding heat affected zone after the welding has cooled that causes the problems with welding. The formation of the martensite on cooling from welding temperatures is not affected by the structure of the material, but rather on chemistry and the preheat given the material prior to welding.

If you can preheat the pintle to 400-500 F prior to welding, this should slow the cooling rate from welding temperatures so that it doesn't crack on cooling. It will also be necessary to post-heat at a temperature of at least 1000 F (metengr has suggested 1200F, which is even better) to temper the heat affected zones to safe hardness levels. This will be true if you use either 4140 or 4340.

This thread reminds me of a consulting job I did for a company operating rotating kilns. They used a 6.5" dia 4340 shaft to rotate the kilns. With poor maintenance, the pillow blocks that supported the shaft would eventually seize and have to be removed with a cutting torch. They would then end up weld repairing the gouges (no pre- or post-heat) and put them back into service. My client was experiencing the shafts breaking in two (transversely), but not all the shafts would break. He had not correlated the fractures with the weld repairs. When I questioned him about proper welding procedures for 4340, he said "yeah, I know you're supposed to pre-heat and post-heat, but we-ve been getting away without it since we're only using annealed material". Well, every crack started in a weld HAZ with a hardness greater than 55 HRC. They hadn't been getting away with anything.

rp

 

[TVP]

I agree completely with redpicker. The problem is martensite formation during cooling of the weld, not during cooling of the bar, machined part, etc. If you don't pre- and post-heat 4340, then the welds will be suspect/cracked/etc.

 

[stanislasdz]

Paul Doherty

here the CCT and TTT curves of the 4340 steel

I hope this can help you !