Can Larson-Miller Be Applied To Stress Relief Situation? 4

[tc7]

We were doing a routine local stress relief on a weld repair of a 4130 casting last night and our induction heater controls malfunctioned. Our normal stress relief for the repair in question would have been 1150 deg F for 2-hours. We could only get upto 1000 degrees out of the heaters as measured by thermocouples placed on the item near the repair area.

I ordered the operation to go ahead at 1000 degrees but increased time to 2 1/2 hours. This was a desperate "seat of the pants" call on my part. Later on I looked at the Larson-Miller equation to see how right or wrong my call may have been and discovered I was getting some absurd numbers, or else I just don't have a clue on what I am doing. Please comment on the following:

Larson-Miller: Thermal Effect = T(log t + 20)/1000

equation used:
T1[(log t1) + 20]/1000 = T2[(log t2) + 20]/1000

where T1 = 1150 deg F = 1609.7 deg R
t1 = 2 hours
T2 = 1000 deg F = 1459.7 deg R
t2 = to be determined

after plugging in above values, I get a t2 of ~245 hours.
So that means my stress relief at 1000 degrees for 245 hours provides the same effect as 1150 deg F at 2 hours. This seems absurd, so I conclude I am misapplying this equation. PLEASE ADVISE and thanks if you will help.

 

[SMF1964]

It has always been my understanding that the Larson-Miller Parameter was developd by Larson & Miller from a heat treatment parameter, so you would then be applying the equation as it's original application.

With regard to your actual application, the 'ridiculous' times you were calculating are actually closer to reality. The ASME welding code does provide extended times for lower temperatures in certain circumstances and they are substantially longer than a boost of a half hour that you applied (even though I recognize your application may not be a code weld). I suspect your 2½ hour heat treatment did nothing significant.

 

[unclesyd]

I have to disagree with the above post in respect to your thermal treatment not doing anything. As you state this is a stress relief and as allowed by code the reduction of metal temperature requires an increase in holding time at temperature. A reduction of 150F would require an increase in holding times to five hours per inch of thickness. In a pure stress relief situation there is reduction of approximately 90% of the stresses in the first few minutes at temperature. IMO your thermal treatment has achieve a reduction of over 90% of the stresses imposed by the weld repair.

If there is any doubt you can redo the thermal treatment at the recommended temperature to satisfy any question.

In your case stress relief and tempering are two distinct operations.

 

[metengr]

Stress relief and tempering are two very different mechanisms for steels. The LM parameter is used for determining hardness reduction via carbide growth using thermal kinetics effects during tempering.

As correctly mentioned by unclesyd, the relief of residual tensile stresses can be very effective at small times and does not follow the tempering correlation for steels. One added benefit is that you will reduce stresses during tempering. So, the extended times you see in the ASME B&PV Code are correlations developed for tempering not only stress relief.

 

[tc7]

I applied the LM equation after reading the small section on stress relief in the ASM Handbook #6 (pgs 33/34) and in the Heat Treaters Guide which has the same basic information. Interestingly, the ASM Handbook #6 shows an example very much like mine but they arrive at an entirely different result.

So Unclesyd, when you say, "A reduction of 150F would require an increase in holding times to five hours per inch of thickness. In a pure stress relief situation there is reduction of approximately 90% of the stresses in the first few minutes..." THIS IS KNOWLEDGE THAT I NEED MORE OF - CAN YOU ADVISE ON REFERENCES AND OTHER READING MATERIAL; web searches are not getting me any good results.
Thanks all.

 

[arunmrao]

As per my impression unclesyd,metengr corypad, edstainless,tvp,kenvlach are the ultimate references and there is no need to go beyond their collective wisdom.

" All that is necessary for triumph of evil is that good men do nothing".
Edmund Burke

 

[metengr]

tc7;
I did come across some information recently that I was able to re-confirm its location at the web site below. I was involved the NBIC in revising the local PWHT discussion;

http://www.forengineers.org/pvrc/PVRC BATTELLE PhaseII-PROGRAMl .pdf

Battelle has been doing research work on local PWHT especially related to weld repairs in the field. If you can download this document, there was a parameter that was apparently developed by MPC/Battelle for evaluating stress relief, not tempering.

If you need more specific information, I would contact the forengineers web site.

 

[SJones]

You should be using the Holloman-Jaffe parameter for stress relief!

http://en.wikipedia.org/wiki/Hollomon-Jaffe_parameter

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[ulyssess]

As said by metengr for LM, also Hollomon-Jaffe parameter HP describes kinetic effects during tempering. With HP, the effect of tempering parameters (temperature, time) on the hardness in HAZ, but also on TS, YS and toughness of base material can be described within some limits for many steel grades.
But the remaining residual stresses are, as said before by some authors, practically only depending on the applied temperature. Time has only a small influence (at least for usual pressure vessel materials).

 

[TVP]

There is some good information available from Pyromaitre on this subject. Click through the links on their website to read more about Larson-Miller and Holloman-Jaffe. The "research" section has a number of papers that can be downloaded as .pdfs.

http://www.pyromaitre.com/

 

[EdStainless]

The basic relationship holds, but when you go to different processes the constant changes. I am not at work, but I have a crib sheet with L-M constants for tempering tool steel, annealing stainless, and other basic practices. In general the constants range from about 16 to 40.

= = = = = = = = = = = = = = = = = = = =
Still trying to help you stop corrosion.
formerly Trent Tube, now Plymouth Tube
eblessman@plymouth.com
or edstainless@earthlink.net

 

[SMF1964]

Proof that a little knowledge is a dangerous thing and check facts before asserting an opinion: After reading the above discussion, I dug up a copy of ASME section I, Table PW-39.1 (the table I was thinking about when I wrote my earlier post) to find that carbon steel and low alloy (150°F below specified temperature) recommends 10 hours/inch of thickness. Although it probably is not applicable for your alloy, it provides a guideline to start with. Should have checked before spouting off. I sit corrected.

 

[tc7]

Don't worry about a thing SMF1964! You were the first to reply to the original question and your reply inspired a great chain of followup information. Sometimes these posted topics only get wide spread interest when something controversial is stated.

Thanks SJones and Ulyssess for mentioning the Holloman-Jaffe equation, this gave me some new directions to explore. There is another thread here on Eng-Tips that discusses this and has a good outline of the H-J equation:thread 794-92824. Although I am cautious of some of the information as no reference was cited and a Kelvin temperature conversion is applied to the holding temperature but not applied to heating or cooling rates.

I have also found Metengr reference by Sinha interesting, it uses a material coefficient of "18" instead of "20" as used by many other texts ,and seems to make a large difference in the computed times.

 

[SJones]

Here is a factored equation to account for heating and cooling rates:

Apologies if it goes pear shaped through my lack of formatting skills:

H = T*(20 + log torr)/1000

torr = t + T/(2.31*K1*(20 - log K1)) + T/(2.31*K2*(20 - log K2))

t = holding time in hours
K1 = heating rate K/kour
K2 = cooling rate K/hour

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[SJones]

Apologies! For 'torr' read "tau". I was under pressure at the time!

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[tc7]

Steve-
Do I presume correctly the your K1, K2 and T are all in degrees Kelvin?

How about a reference for that form of the equation?
Thanks

 

[SJones]

Holloman J H, Jaffe L D, 'Ferrous Metallurgical Design', John Wiley and Sons Inc., 1947, pp.245.

and

B Lee, TWI UK, "What Is The Holloman-Jaffe Parameter"

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/