17-4 Forging Metallurgy 3

[USMechE6]

Hi All,

I'm tasked with specifying a hollow cylinder to be made out of 17-4 PH (H900 heat treatment) and to be used in a high pressure and wear application. The O.D. will be 12" and it will be in excess of 100" long. I can't go into too much more detail but I'm hoping for some insight into the metallurgy I should call out. I do not have much experience with metallurgy other than what I've tried to learn on my own so any general info would be great.

If I call out a forging, will the recystallization during the solution anneal at 1900 F 'cancel out' the effect on grain structure of the forging, or is there some HWRR that would maintain a finer grain structure post treatment? I imagine this depends on the duration of the soak at 1900 and I.D. of the cylinder.

I've read some literature to indicate that finer grain structure will lead to better wear characteristics (I know it depends on what's wearing on what). This is something I would like to achieve. Would a minimum of HWRR of 4:1 enable finer grains than if it were say, spun-cast? Also I'm thinking that calling out 'open-die forging' will be specific enough?

Thank you for any help.

 

[EdStainless]

With the PH grades you can play with grain size.
If you anneal and then age, then re-anneal and re-age you will reduce the grain size.
This will increase the toughness and only very slightly reduce the strength.

You could make this part from cnetrif cast hollow that was then forged.
Are you looking at any surface treatments?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[USMechE6]

Thanks for your reply. Interesting, that makes sense.
I'm not looking at any surface treatments currently. We actually have had these made in the past by another vendor which used their proprietary coating on the I.D., but which cracked due to pressure cycling and differences in modulus, strength, etc. They've made all of our subsequent cylinders out of just 17-4 PH, H900.

Some additional background. We've recently begun using a cylinder that has had heavy particulation due to wear of the I.D. I managed to track down that it was a spin-casting (quenched), and displays a mottled appearance that returns with use despite repeated attempts at polishing.

The other cylinders we've used w/out issue were 17-4 forgings (no surface treatment), but were made so long ago that the specific work history of them is lost (one annealing/aging as far as I know)...so I'm trying to understand what conditions might enable me to recover their performance characteristics in terms of wear.

 

[metengr]

Your best bet would be to go to a reputable forging house and work directly with them. I have used the vendor below and they are excellent

http://www.jorgensenforge.com/pdfs/JFCbrochure.pdf

 

[USMechE6]

Thank you both very much for your thoughts and info. One additional question - considering the cycling of my cylinder, what is a good chemical composition standard to specify in order to minimize fatigue failure? I know there's several ASTM Standards but I didn't see anything for 17-4. I've been told that too high of silicone content can lead to fatigue failure. Also, should I call out that the casting process be done in vacuum to avoid impurities?

 

[EdStainless]

If fatigue and toughness are an issue then move to 13-8PH that has been double melted, one mill sell it as SuperTough. This grade, melted this way has superior microstructural cleanliness and the property benefits that come along with it.
A second option is to move to 15-5PH, which is a more stable version of 17-4. It is easier to get good uniform properties in heavy sections. I believe that 15-5 is also available as re-melted material (AOD-ESR).
And don't let someone use the casting chemistry, require the wrought product chem that is listed in the AMS specifications.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[MagBen]

Besides, 15-5 offers superior transverse toughness and ductility plus a high degree of forgeability.
As ED noted, both 15-5 and 17-4 are available as arc + ESR melted.

 

[USMechE6]

Thank you gents. Those sound great from an engineering perspective. Unfortunately I will probably be confined to 17-4 (by non-engineers...) due to revalidation that would be needed for products running through this equipment if I went with a different material. So I'm thinking I'll go with vacuum re-melted 17-4 using the AMS wrought chem. Any thoughts on silicon content?

 

[MagBen]

High Si is not good for toughness. Normally Si needs to be lower than .5%. However, you do not want it too low, since you need it to de-O2.
b.t.w. ESR is not under vacuum. vacuum melted (say VIM) 17-4 is not common, while VIM 15-5PH is popular.

 

[EdStainless]

In these alloys I don't care what the method for remelting is, I have found that AOD-ESR (no vacuum at all) is nearly as clean as VIM-VAR and it is actually available.
Perhaps 15-5 can be a sell for you since it is really just a modified 17-4, the documentation even says that.
In 15-5 I believed that we used Si 0.20-0.50, but we also had a bunch of other restrictions.
Such as S 0.002% max, and S+P+N+O+0.1Nb <0.060% and Nb was 5xC, 0.15min, 0.30max.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[USMechE6]

Awesome guys, thank you so much for your insight. Big help!

 

[WKTaylor]

USMechE6

Ed Stainless has provided a strong argument for making Your part from 15-5PH VAR ILO 17-4PH... especially since there is the high purity version is readily available [AMS5659].

CAUTION. The H900 or 925 Conditions are prohibited for use in aerospace structural parts MF 15-5PH or 17-4PH, due to low toughness and being prone too SCC failure. See MIL-STD-1587 Table I Restricted materials.

NOTE. I have used 15-5PH [AMS5659] at H900 Condition for bushes and small press-fit parts; and also used 15-5PH [AMS5659] at H925 Condition for small'ish structural and mechanical parts, that are then stress-relieved after all machining... and then shot peened [all surfaces]... and then coated with plating's [must be embrittlement relief baked] or bake-on SFL.

NOTE. IF these parts were used in moderate/low service temperature [LT 500F] and relatively benign corrosion environment, then I'd tend to lean towards PH13-8Mo AMS5934 [extra high toughness], H1025 or H1000 Condition, for best possible long-term service at the stress level You imply [~=H900 for 17-4 & 15-5].

Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[tbuelna]

The OP mentioned the cylinder was 12" OD x 100" L, and was forged. What forging process was used to produce a cylinder of this diameter and length? Given the large 12" diameter and 100" length, was it pierced and hot drawn from billet? As noted above, 17-4PH is not used at H900 condition in aerospace due to SCC concerns. 15-5PH or PH13-8Mo would be OK at this hardness, but it will be more difficult to find these materials in large bar diameters. Also as noted above, you might want to make your forging from a vacuum-melt quality material. A forging made from vacuum-melt quality material will give much better fatigue performance than any cast material. But these vacuum-melt quality materials are not usually available in large diameter bar sizes. So you might have to forge a preform shape from smaller bar to use for the final tube forging.

It is important to make sure the hot working is performed with the material at a high enough temperature to prevent cold laps, seams, etc in the forged material. It is also necessary to machine off at least 1/8" from any hot worked surface of the forging to remove contaminated material. It would also help to take a look at the recommendations in a heat treat process spec like AMS 2759/3.

 

[EdStainless]

If you use centri cast material remember that the ID is the 'top' of the casting and has all of the slag and junk. I would take at least 1/2" out of the ID before honing to final size.
We have extruded cast parts (turning them into forgings) and only had success when we took 3/8" off of the OD and >0.500" out of the ID.
There are people out there that will centri cast from clean melt billets, and do it with Ar shield gas.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[USMechE6]

Unfortunately the forging process that I am trying to recover occurred several years before I began at my company. The documentation has been lost by my company and the vendor. So I'm not sure what forging process to call out - one forger that I spoke to said mandrel-forged. Would this assure better microstructure at the I.D. from a wear perspective than say a billet open-die-forged or rolled and then bored out, or an initially centri-cast tube that's then forged? The problematic cylinder that was cast had an I.D. of 5.74", so it definitely had a lot bored out to reach the final I.D. I'm also unsure of what HWRR to go with. I know it depends on a whole lot, but I'm thinking a minimum of 4:1 - thoughts on this?

I'm also unsure as to what specifications to go by, have seen many. ASTM A564, ASTM 705, AMS5643, AMS 5604, UNS S17400, ASTM A693, ASTM 484, ASTM 479, ASTM A788, AMS5659. Also do you guys have any recommendations for void testing?

This is not for an aerospace application, but will be subject to pressure fluctuation, relatively low operating temperature. Calcs have been done by engineers before me to show that the nominal dimensions with 17-4 at H900 would be sufficient from a strength/fatigue perspective. I just don't know what specifics in the forging process would be to call out for optimizing the microstructure and other mechanical properties would be. If I can stay with 17-4 or 15-5 that's probably best.

 

[WKTaylor]

USMechE6... Aerospace has learned hard lessons from materials.

In aerospace, the problem with 17-4PH/15-5PH H900 temper is primarily SCC [stress corrosion cracking] related. SCC is aggravated by multiple factors that cannot always be easily predicted.

OK, OK, OK... Avoiding a forging parting plane, using cylinder hand-forging techniques, should eliminate worst source of SCC initiation and make for a beneficial circumferential/axial grain-flow.

Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[USMechE6]

Thanks for the warning, I do appreciate it. Fortunately we're not operating in a high-temp or highly corrosive environment. We've had success with our forged 17-4 (and 15-5 I recently found out) cylinders at the H900 condition - it's just now we need more. If I go with a different heat treat, I need to redesign the O.D. features of the cylinder due to the lower nominal yield and tensile strengths to maintain the same F.O.S. Given the accompanying tooling and equipment that would have to change as well, this is something I am looking to avoid.

 

[EdStainless]

USMechE6,
You could look at having these hot extruded.
When this is done you have the same structure as a hot forged part.
There are a few press around that could push this part.
I would do this with remelted 15-5PH or 13-8PH, and I would prefer the 13-8.
An extrusion would require modest stock removal from the ID, perhaps 0.150" (0.075" off of the wall).

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[tbuelna]

The 100" length of your cylinder is a bit beyond the capability of ring rolling machines. So I imagine your previous cylinders were produced using the mandrel forging process suggested by your vendor. Mandrel forging starts by piercing a hole thru the heated material preform, and then placing a cylindrical mandrel of the appropriate diameter/length inside the hole. The heated preform OD is hammered between a set of dies while being continuously rotated/translated thru the dies. This draws down the OD and lengthens the material preform. This hammer forging process usually results in a circumferential/axial grain orientation, which would seem to be acceptable for your application.

Specifying a minimum HWRR helps ensure the finished forging material is sound and has the grain characteristics desired. You did not mention what the forged cylinder's wall thickness is. But based on the 12" OD, 100" length, and limited availability of large diameter wrought bar in the alloy required, I think you will need to start with a forged preform made from smaller diameter bar stock. Remember that this mandrel hammer forging process only reduces the material OD as it increases length, with the ID surface being constrained by the mandrel. Unless your cylinder forging has a very thick wall, it should not be too difficult to achieve a 3:1 or 4:1 HWRR from hammer forging over a mandrel.

 

[USMechE6]

Thanks for the additional info guys. Mandrel-forging seemed to be a reasonable route. I know the other alternative would be just open-die forging a square and then boring it out, which another vendor suggested, but I imagine this is a less economical use of material. I also came across a paper that showed mandrel forging would be better for the I.D., since the alternative leads to more of a gradient of properties and structure through the radius. The wall would be pretty thick. There would also be a 16" O.D. flange, ~2" long at one end.