ASM Online Member Community

Expand all | Collapse all

F91 heat treatment requirements - needlessly stringent?

  • 1.  F91 heat treatment requirements - needlessly stringent?

    Houston Admin
    Posted 11 days ago
    ​We forge and heat treat F91 valve bodies for an OEM that supplies the power generation industry. Usually these are standard ASTM A182 material with a few supplemental requirements for things like min temper and hardness which are easily attainable. However, they recently approached us asking if we could meet some additional requirements being flowed down from their customers (utility companies), including:

    • Every individual forging in the HT load needs to have a minimum of 4 contact thermocouples (i.e. ~3 dozen in the load).
    • The rate of cooling off the normalize cycle must be, at minimum, 108 F/min (i.e. it must be quenched rather than air cooled)
    • Between normalizing and tempering, the forging cannot be exposed to atmospheric humidity above 80% (i.e. it must be quenched in oil, not water or polymer, and transferred immediately into the temper furnace after leaving the oil tank)
    • Restricted chemistry on residual elements and interstitial ratios

    I understand the motivation for the tight chemistry because it is associated with better creep strength, although I cannot say exactly how big of a difference it makes compared to the standard chemistry (i.e. how much increase in component life is associated with the residuals being this low and whether it justifies the increased cost). However, these heat treat requirements are, in my opinion, totally excessive, for several reasons:

    • Our furnaces are regularly surveyed to +/- 25F per AMS 2750, and these parts are a fairly uniform, thick section. We ramp them to temp to mitigate the risk of cracking and the soak time is more than long enough to ensure they are uniform in temperature.
    • The rate of cooling actually necessary to obtain complete martensitic transformation is about 9F/min, which is attainable by air/fan cooling. Cooling 12x as fast offers no benefit that I am aware of, except it forces me to oil quench which increases my risk of cracking. Furthermore, it would require that the spiderweb of 30-40 contact TC's be lowered into the oil with the load, which is violently agitated and, almost inevitably, some would get blown off (this would invalidate the quench since the cooling rate needs to be recorded to ensure it met the 108F/min minimum).
    • We have never had an issue with these parts cracking from sitting at ambient (which, in Houston, is above 80% humidity more often than not) for a short amount of time between cooling off the normalize cycle and going into the tempering furnace.

    Does anyone have any experience heat treating this alloy to suggest that any of these requirements are actually necessary? I am feeding these concerns back to my customer so that he can in turn relay them to his customers, but I am wondering where they might have come from in the first place.

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


  • 2.  RE: F91 heat treatment requirements - needlessly stringent?

    Rocky Mountain Admin
    Posted 10 days ago
    Sean,

    I believe the 108°F/min cooling rate requirement must be a typo in the requirement document you were provided.  Per the EPRI Grade 91 recommendations document (EPRI report 3002006390, available for free at the EPRI website), the required minimum cooling rate in normalization for Grade 91 steel is 9°F/min, which is generally achievable by free air cooling or by forced air cooling for components with section thickness greater than 3 inches.  I'm not sure why anybody would specify 108°F/min, except by accident.

    The tighter chemical composition requirements derive from EPRI work showing that heats with higher impurity content are more susceptible to creep cavity formation and therefore lower creep ductility and poor cross-weld and multiaxial creep performance. See EPRI reports 3002003472 and 3002004370 - I think also available to free. 

    The requirement for control of component temperature and/or humidity between normalizing and temperature stems from industry experience of stress-corrosion cracking of Grade 91 components that were left in an untempered condition in uncontrolled environments for a prolonged period of time.  We have seen a few of these at Structural Integrity - see the article below for an example.

    Hope this helps.  Feel free to get in touch if you have further questions.

    Terry

    Metallurgical Lab Case Study: Grade 91 Elbows Cracked Before Installation | Structural Integrity Associates
    Structural Integrity Associates remove preview
    Metallurgical Lab Case Study: Grade 91 Elbows Cracked Before Installation | Structural Integrity Associates
    Structural Integrity (SI) personnel visited a power plant construction site to examine four Grade 91 elbows (ASTM A234-WP91 20-inch OD Sch. 60) that were found to contain axially oriented surface indications. The elbows had not yet been installed.
    View this on Structural Integrity Associates >



    ------------------------------
    Terry Totemeier
    Structural Integrity Assoc.
    Boulder CO
    ------------------------------



  • 3.  RE: F91 heat treatment requirements - needlessly stringent?

    Posted 9 days ago
    Could the 108F be a typo for 10.8F converted from 6.0 C?

    ________________________________________
    Larry D. Hanke, PE
    Materials Evaluation and Engineering, Inc.

    Sent from mobile device. Please excuse brevity and typos.





  • 4.  RE: F91 heat treatment requirements - needlessly stringent?

    Posted 9 days ago
      |   view attached
    Hi, Sean,

    Terry has given you some good guidance.  A Grade 91 CT diagram is attached; not sure who developed it but I got it from one of the guys on the ASME Creep Strength Enhanced Ferritic Steel committee.  You can cool Grade 91 a lot more slowly than 108°F per hour and still get martensite.  Way more slowly.  With sections over 4 inches or so, maybe a fan directed through an opening, but water quenching is silly.

    Also, from what I understand (and I'm not sure anyone really understands it), the cracking phenomenon is related to contamination of the surface by handling -- usually chloride salts + condensed moisture. RH is not the driver; condensation is.   Suggest you offer to MT the parts after tempering.  If it cracks, that's your problem.

    Walt


    ------------------------------
    Walter Sperko
    President
    Sperko Engineering Services Incorporated
    Greensboro NC
    (336) 674-0600
    ------------------------------



  • 5.  RE: F91 heat treatment requirements - needlessly stringent?

    Houston Admin
    Posted 8 days ago
    Larry,

    I myself was thinking the exact same thing, that they wanted the requirement to be 6C but they converted it to Fahrenheit incorrectly (10.8 vs 108). ​I will be sure to ask what the original Celsius requirement was.

    Terry/Walter,

    Thank you for the information - I read the referenced EPRI reports and they contained lots of good information.

    I'm wondering again about the 4 thermocouple requirement - it seems like that might have been formulated originally for something more like a pipe welding application where a localized PWHT is used and it's more difficult to ensure temperature uniformity. In this case, these forgings are being heat treated in a furnace significantly larger than themselves where the force of the air/gas blowing out of the burners circulates the atmosphere to encourage temperature uniformity (plus the normalizing cycle is hot enough to have a significant radiative heating component), so the odds that one section of a given forging would be significantly hotter/colder than another section after adequate soak is, in my opinion, very low. We heat treat aerospace parts all the time with a heat sink to represent the load (rather than a TC on every part) and it works fine. Do you think any value would really be added by using 4 TC's per part? Have you ever seen this required in other applications?

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



  • 6.  RE: F91 heat treatment requirements - needlessly stringent?

    Rocky Mountain Admin
    Posted 8 days ago
    Sean,

    Glad you found that information helpful.  Regarding T/Cs, in your case it sounds like a single T/C would probably be adequate.  The requirement for a minimum of four T/Cs stems from the discovery of many Grade 91 components (especially forged bends, tees, valve bodies) being found after installation in the power plant as being incorrectly heat treated (low hardness and non-martensitic microstructure).  Having more T/Cs gives the utility or OEM purchaser more assurance that the heat treatment has been done correctly and in a furnace with uniform temperature where the component is (e.g., flame not impinging on the component).  For local PWHT even more T/Cs are required!

    Terry

    ------------------------------
    Terry Totemeier
    Structural Integrity Assoc.
    Boulder CO
    ------------------------------



  • 7.  RE: F91 heat treatment requirements - needlessly stringent?

    Posted 7 days ago
    Sean, thank you for your questions and thank you Dr. Totemeier for the timely and accurate responses.  As a further point of background, EPRI's Grade 91 specification is based on over 15 years of research stemming from a wide variety of problems the industry has encountered with the performance of Gr. 91and multiple extensive test campaigns.  EPRI's research and guidance documents are all publicly available with some of the guidance and white paper documents available for free (such as the ones Terry referred to including: https://www.epri.com/research/products/000000003002006390).

    That being said, we welcome any feedback you have on the document as we are constantly evolving these and other guidance documents based on the best available research and feedback from industry.  We are currently working on a similar specification for Grade 92 and have other similar guidelines for high reliability power plants including dissimilar metal welds, hard facing application, etc. I would be happy to connect you with Dr. John Siefert who lead's EPRI's materials program which maintains these reports and Dr. Jonathan Parker who was the principle author.

    ------------------------------
    John Shingledecker FASM
    Electric Power Research Institute
    Charlotte NC
    (865) 201-1252
    ------------------------------