ASM Online Member Community

Expand all | Collapse all

Unusually high YS in H900-aged 15-5?

  • 1.  Unusually high YS in H900-aged 15-5?

    Posted 04-23-2021 15:51
    We buy 15-5 VAR reforging stock (billet) to AMS 5659 from a reputable American mill for a particular aerospace product we make. The mill performs H900 capability testing on the MTR, which specifies a min YS/UTS of 170/190 ksi, respectively. Historically, they have averaged in the mid 170's for YS and upper 190's for UTS. The YS/UTS ratio is consistently in the 88-90% range.

    Recently, however, we received two heats which reported YS/UTS of 196/199 ksi (ratio = 98.5%) and 200/201 ksi (ratio = 99.5%), respectively. Although these UTS values are certainly on the high side, they are not unprecedented - the YS values, however, are significantly higher than I have ever seen, either from this mill or elsewhere. The elongation and RoA were 14/43%, respectively, which is a little lower than we usually see but still comfortably passing for H900 (the specimen was transverse so the requirement is 6/20%, respectively).​
    I contacted the mill about this and we went through the standard due diligence of overchecking the stress strain curves, confirming that they extracted the specimens from the standard locations in the bar, checking the tensile specimen fracture surfaces, etc. and everything was done correctly; the material truly was that strong (it was tested at a 3rd party lab to boot). The mill commented that they do occasionally see YS/UTS ratios >98% and, while we could retest, there'd be no real point since the test didn't fail and also since another ingot from the same parent heat recorded average values (a third ingot reported 191/202 ksi, a 94.5% ratio).

    I am reaching out to this forum hoping to gain insight into what might be happening mechanistically here. I agree with the mill that there is limited value in retesting and there is nothing wrong with the material, but when you see statistical outliers like this it makes you wonder what's going on. My assumption is that it has to do with lot-to-lot variation in response to the H900 aging cycle, which is only 1 - 1.25 hours instead of 4 - 4.5 hours for other aging conditions (including H925, which is only slightly hotter) so it's a relatively quick precipitation reaction where the difference between truly peak aging and slightly under/overaging is probably minutes at temp, if that. I am assuming that the stars just aligned with these particular heats/tests and they were pulled at exactly the right time to achieve a truly peak age, but that's mainly because it's the only explanation I can come up with.

    Very interested to hear what you have to say!

    ------------------------------
    Sean Piper
    Product / Process Metallurgist
    Ellwood Texas Forge Houston
    Houston TX
    ------------------------------


  • 2.  RE: Unusually high YS in H900-aged 15-5?

    Boston Chapter Admin
    Posted 04-24-2021 10:28
    Since it's capability testing, I assume only the specimens were put through treatment, correct?  Do you have the furnace charts (with history of time, temperature, and rates) of those treatments?

    I'm curious regarding the chemistry, especially carbon, nickel, niobium, and copper levels.

    Would there be any opportunity to test a specimen to the as-quenched condition or the H925 condition for the longer heat treatment time?  This seems to be a "lack of temper" issue.

    I'm not a huge fan of H900 or CH900 for PH SST materials because of ductility concerns, but I understand that they are necessary for certain applications.

    Sincerely,

    Dan D.

    ------------------------------
    Daniel Denis
    ASM Chapter Council Vice Chair
    Hartford Chapter
    Senior Consultant - Metallurgy & Materials
    Structural Integrity Associates, Inc.
    ------------------------------



  • 3.  RE: Unusually high YS in H900-aged 15-5?

    Posted 04-26-2021 14:05
    Dan,

    It turns out that our end customer for this product puts it into service in a condition other than H900, but for byzantine historical reasons this is what we certify capability to (both at the mill and on a test forging per heat). However, we have occasionally experienced H900 failures in the past where we were able to demonstrate capability to a different age condition and get the parts accepted.​

    Regarding chemistry, I don't want to give away too much about the mill's aims but I wouldn't say that these high-YS heats were significantly different from what they've provided historically in the primary strengthening elements. That said, we have seen significant differences in the past between this mill and other mills which tend to provide different aims (notably lower copper and higher nickel); in this case the other mill records consistently lower YS and UTS, I presume mainly due to the lesser degree of copper precipitation and secondarily due to incomplete martensitic transformation upon cooling to room temperature due to the increased austenite stabilization.

    Regarding the samples, you are correct that only the specimens were heat treated (in a lab furnace). I might be able to get those charts if I ask nicely.

    Regarding retesting specimens in other conditions, I could request this but, given that there was no actual failure, in a way I'm asking for help on an R&D project more so than I'm asking for a retest. I would not hold my breath on this, and I also don't know how much it would get me if the root cause of the high YS is indeed a nuance in the H900 cycle (and whether these samples happened to get pulled at true peak age). I might be able to sell it if there were a specific end goal in mind though - what results would you expect in these conditions?

    ------------------------------
    Sean Piper
    Product / Process Metallurgist
    Ellwood Texas Forge Houston
    Houston TX
    773-524-8985
    ------------------------------



  • 4.  RE: Unusually high YS in H900-aged 15-5?

    Posted 04-27-2021 07:48
      |   view attached
    Greetings Sean,

        At Solar we age tons of PH grades which are solution annealed directly from the mill. Over 90% of the time the material is conforming after the age treatment. We know this because we machine and pull our own tensile samples in house. The majority of the time our tensile results agree with the Mill's capability tests. When we do get those outliers, we typically re solution treat and re age - not changing a thing in our re age treatment. The material always responds correctly to my second re age. This is true for different heats of material all certified the same mill heat!

        So why this variability? I wrote about this phenomena several years ago (see attached). I personally blame the differing MIL specs for raw material and the AMS spec for finished parts. Pyrometrically they are written differently. In our commercial heat treating world we always process with WORK THERMOCOUPLES imbedded into the material no matter what the job is. The mills do not because they are not required to per the old raw material MIL spec. Therefore, running by a control TC, they really do not know the true SA temperatures of their loads nor the true cooling times. They go by the old 1 hour per inch of material adage which only tells you the temperature of the furnace not the material. Their load sizes are also not comparable to ours.

         My purpose for writing this article then was to point out the inequities in the specification which ultimately affect the final heat treatment. Our metallurgist Greg Scheuring has several case studies that proves this theory and how it affects tensile strengths and impact qualities. Feel free to contact Greg (Ext 2253) or myself (Ext 2224) at the number below.

    I hope this helps you.

    Bob Hill
    President
    Solar Atmospheres of Western PA 
    724-982-00660

    ------------------------------
    Robert Hill
    President
    Solar Atmospheres of Western Pennsylvania - Hermitage
    Hermitage PA
    (724) 982-0660
    ------------------------------

    Attachment(s)



  • 5.  RE: Unusually high YS in H900-aged 15-5?

    Boston Chapter Admin
    Posted 04-27-2021 08:52
    @Sean Piper

    That's very interesting regarding chemistries, as I've found that there is a trend in reduction of expensive alloying additions if specs can be satisfied by re-balancing.

    @Robert Hill brings up some great points regarding heat treat and control of temperatures in work pieces versus control of furnace temperature only.

    In my estimation, the combination of a more greatly martensitic structure with reduced temper relief will provide a PH SST with elevated properties and low ductility.  I ask about As-Quenched and higher temper testing because I suspect that you're looking at a material with copper precipitates but with a highly stressed martensitic structure, providing the very high strength values.

    In a 4-hour PHT for H925, you know that even with furnace variability (per @Robert Hill's comment) that you'll have sufficient heat to temper and knock down the strength and increase ductility.  I believe that the H900 treatment is more sensitive to those inconsistencies, including under-heating the work piece.​​​​

    Thank you for the great insight and for allowing me to continue to discuss!

    Sincerely,

    Dan D.

    ------------------------------
    Daniel Denis
    ASM Chapter Council Vice Chair
    Hartford Chapter
    Senior Consultant - Metallurgy & Materials
    Structural Integrity Associates, Inc.
    ------------------------------



  • 6.  RE: Unusually high YS in H900-aged 15-5?

    Posted 04-27-2021 14:36

    Dan,

    It turns out that our end customer for this product puts it into service in a condition other than H900, but for byzantine historical reasons this is what we certify capability to (both at the mill and on a test forging per heat). However, we have occasionally experienced H900 failures in the past where we were able to demonstrate capability to a different age condition and get the parts accepted.​

    Regarding chemistry, I don't want to give away too much about the mill's aims but I wouldn't say that these high-YS heats were significantly different from what they've provided historically in the primary strengthening elements. That said, we have seen significant differences in the past between this mill and other mills which tend to provide different aims (notably lower copper and higher nickel); in this case the other mill records consistently lower YS and UTS, I presume mainly due to the lesser degree of copper precipitation and secondarily due to incomplete martensitic transformation upon cooling to room temperature due to the increased austenite stabilization.

    Regarding the samples, you are correct that only the specimens were heat treated (in a lab furnace). I might be able to get those charts if I ask nicely.
    Sean Piper,  04-26-2021 14:04
    Hi Sean,
    This is an interesting situation.  Certainly the use of any spec is just an agreement between purchaser and supplier. The ranges of expected properties in the end product is what they're buying.  And if the end user intends to repurpose the material or retreat it, presumably they are working from someone's expectation that "if it will meet H900, then it will meet Hxxxx when we give it the final treatment". 
    However such a statement sounds suspiciously like what we called "dead guy's rules", where the process being followed was established without adequate documentation so that nobody now knows exactly what or why something is required. When a material such as the overachiever heat appears, it may just be "if more is better, then too much is just right" extra good, or it may have a combination of properties that takes it outside of the original specification and industrial experience. Aerospace and medicine are two areas where the known is always better than the new/unknown until there have been years of experience with the new.

    In any case, my limited experience with these grades was with 17-4, though the differences are mainly in that 15-5 is rebalanced to further minimize the chances of residual ferrite.  Those experiences were in situations where going to higher hardnesses were bad, since to a first approximation the SCC resistance of the parts goes inversely with the hardness.  A problem with these grades is that there is substantial overlap in the final property ranges between the different heat treats, so that just doing hardness (or probably tensile) _may_ not be able to differentiate two heat treats.  For instance, room temperature tests of Condition A overlap the acceptance ranges of the intermediate heat treats.  I think that the experience with SCC resistance is always based on whether the actual heat treat temperature/time was done.

    One of the failures I dealt with was a seacoast atmospheric exposure where the part, which was from hardness probably an H900 or H925-ish treatment, had SCC.  Another was a specialized application where one end of the part was subjected to an operating temperature ~625F, and over long exposure had increased hardness and lowered impacts/SCC resistance from increased precipitation / ordering, though the other end of the valve stem at lower temperatures still met the original spec (I think it was H1100).  For the latter situation, impact testing clearly showed a substantial embrittlement accompanied the increased hardness. 

    So my bottom line is that 1)if it were my situation, I would ask for those heat treat charts (WITH records of furnace calibration) to at least assure your own interest that the material did indeed have the specified heat treatment; 2)I would communicate the remarkableness of the results to the end purchaser on the off chance that they have someone (not a dead guy or "she's not here anymore to ask") who can evaluate whether an overachiever heat is truly suitable for their end use.

    ------------------------------
    Paul Tibbals, P.E.
    ------------------------------



  • 7.  RE: Unusually high YS in H900-aged 15-5?

    Posted 04-26-2021 08:41
    Edited by George Shannon 04-26-2021 08:41
    I'd agree with Dan that chemistry might be playing a role - it would be unusual, though not unheard of, for ingots in the same heat to have slightly different chemistries.  If your certificate of test shows minimal differences, I'd personally chalk it up to testing variability.

    George





  • 8.  RE: Unusually high YS in H900-aged 15-5?

    Boston Chapter Admin
    Posted 04-27-2021 08:53
    Thank you @George Shannon!  I agree with you that with the small sampling numbers there can be some significant variability.

    ~Dan D.​

    ------------------------------
    Daniel Denis
    ASM Chapter Council Vice Chair
    Hartford Chapter
    Senior Consultant - Metallurgy & Materials
    Structural Integrity Associates, Inc.
    ------------------------------



  • 9.  RE: Unusually high YS in H900-aged 15-5?

    Posted 04-27-2021 10:12

    Sean,

    Another area you may want to investigate is the actual ageing heat treat practice used.  I am assuming that AMS 2759/3 was the heat treat specification utilized.  When I was a Nadcap Heat Treat Auditor I saw quite a bit of variation in how AMS 2759/3 was applied.  Given the low ageing temperature of 900F I would focus on how soak times were actually determined.  Was Furnace Instrumentation Type C or Type B used?  Weight of the furnace load and ramp-up cycle are also significant factors to investigate.

    I have seen a lot of interesting heat treat practices in my career.  Just because a specification may allow you to do something does not necessarily mean it is a good idea to do so.

     Sincerely,

    Robert Werber, PE
    Planish Inc.



    ------------------------------
    Robert Werber
    Principle Engineer
    Planish Inc
    Oregon City OR
    9712218453
    ------------------------------