Using vibration to control stress during martempering 6

[Greenleader]

We know you can use sub to low sonic waves to relive stresses in carbon, alloy steel cased by machining, welding etc.

When martempering, (marquenching), carbon alloy steels, is it possible to use the same vibrational frequencies that reduce stress, to reduce the built in stresses caused by the transformation of austenite to martensite? Could vibration be used to reduce retained austenite during martempering?
Metastable austenite is very plastic. It seems you could take advantage of this state and manipulate crystal formation with sound waves.
Any studies out there? I can not seem to find any myself. Thanks

 

[swall]

Vibrational stress relief, if it happens at all, is of low intensity, i.e. the degree of stress relief would be low. With marquenching, the whole idea is to relieve thermal stresses caused during cooling to the marquench temperature. But, you are still going to have the austenite to martensite transformation stresses. Since the transformation time is so short, there is not time for any kind of stress relief to help during transformation (which I think is the gist of your question). Anything that is going to be done to alleviate the transformation stresses has to wait until after transformation.

 

[rorschach]

I agree with Swall, every metallurgist I've discussed this with says it is either an "old wives tale"/voodoo or if it happens at all it is so minimal as to be useless.

 

[Greenleader]

What is exactly the "wives tale" you refer to?

 

[Greenleader]

If vibrational stress relief is a wives tale, than my company believes it, as I am starting to write a BAC spec for vibrational stress relive with the controlling engineer in the near future. I've done a lot of work with this the last 12 years or so.

I just wanted to know if it could work on metastable austenite to improve grain.

 

[rorschach]

Old Wives tale: aka urban legend, Snake oil, false statement, junk pseudo-science. In other words, no metallurgist I have ever discussed this purported phenomeona with believes it actually happens, or if it does happen, it is of such low intensity as to be not worth trying. I have discussed this with metallurgists who have been part of internal company studies trying to substantiate this phenomeona and they have been unable to prove it does anything. The claims of vibrating a weldment or casting to relieve internal stresses have been circulating for years but the only claims I've seen are by people who sell shaker equipment....

 

[Greenleader]

A little off topic but yes, if you judge claims made by manufactures, just hype.
We actually discovered this by accident several years ago when we received a bad batch of castings from a vendor. The flaws would show up at a certain process and would consistently manifest themselves. We did many controlled experiments, and found out that a certain freq. and process, does indeed relive internal stresses caused by heavy machining etc. MRD was impressed enough to design a similar device, (currently one is 1 of a kind ) to be used for large, single machined components.
They spent the time and the money and decided it works. By the way, we don’t sell shakers.

 

[bklauba]

Gentlemen,

I wrote a paper, soon to be published by the ASM, on Vibratory Stress Relieving. My coauthors in this effort were:

- Dr. C. Mel Adams, PhD, Metallurgy, MIT. Dr. Adams cofounded the Welding Research Department at MIT, has been the Dean of Engineering at the U. of Wisc. and the U. of Cincinnati, and Chief Staff Metallurgist at Forensic Technologies Int'l. Amongst his discoveries is the hydrogen embrittlement properties of T1, a problem that is common amongst low-carbon, high-strength steels (such as ASTM 514, B, H, and Q), HY80 and HY100, ASTM 710, and similiar steels.

- Dr. John T. Berry, Coleman Chair of Engineering at Mississippi State University. Dr. Berry has taught at Georgia Tech and the U. of Alabama, and is an expert at FEM simulations of complex material behavior, such as dendrite formation and fluid motion.

The paper was recently presented at ASM Int'l's Trends in Welding Research, held every 3 years, this time in Pine Mountain, GA.

Agreed, far too many salesmen are out there pushing vibratory stress relief upon applications that are inappropriate. AT THE SAME TIME, the VSR Process can be used to assure the dimensional stability of a wide array of components, especially if they are large, and works extremely well on many stainless steel grades. This type of experience was undoubtedly what has caused Greenleader to ask the question he has. Indeed, for some applications, the only way to do it effectively is with the VSR Process.

In my opinion it should be approached experimentally.

 

[Greenleader]

Thank you bklauba. I look forward to your paper. The process we use is an internal, tailor made system. We find it works quite well on complex parts, with very thick, and very thin cross-sections.

You mentioned stainless, I would also like to add aluminum alloy as being very effected by this.

As to my original question, do you know of any studies concerning the application of sound waves on carbon steel to effect grain during thermal processing?

 

[mtnengr]

Is Vibratory Stress Relief a hoax? Some say yes others say no. I think that the misnomer is to include “Relief” as part of the title. I suggest that four publications gives one a proper perspective:

1. G. Gnirss, Vibration and vibratory stress relief.
Historical development, theory
and practical application. Welding in the World,
Vol. 26 No. 11/12, pp284-291, 1988.

2. R. Dawson and D.G. Moffat, Vibratory Stress
Relief: A Fundamental Study of Its
Effectiveness, J. Engr. Matl, and Tech., Vol 102
pp 169-176 Apr 1980.

3. T.E. Hebel, Sub-Resonant stress relief: what it
is; when it’s used. Heat Treating, pp 29-31,
Sept. 1989.

4. Y.P. Yang, G. Jung and R Yancey, Finite Element
Modeling of Vibration Stress Relief after
Welding, Proc. 7th Int. Conf. on Trends in
Welding Research, Pine Mountain, Georgia, May 19,
2005.

I’ll let the readers read and draw their own conclusions.

It is my opinion that Ref. 3 has no verifiable engineering basis while Ref. 4 gives us a complete insight as to the true effects of VSR. Clearly, VSR is a masking of additional stresses over an existing stress state. Plastic strains are introduced at specific points along the work piece and through the thickness. The amount of plastic strain is dictated by the load level and duration imposed. Further, the path of effective stress to effective strain is different for loading and unloading. This is called the Bauschinger effect. The instantaneous tangent modulus changes for different cycles along a portion of this path. Therefore, one has a different structure for each cycle to consider. With these changes in material behavior, the actual natural frequency of the structure changes—slightly lower. This is demonstrated in BKlauba’s paper. Although, there is no discussion as to the change in compliance of the supporting blocks.

If one uses the natural frequencies as a guideline for developing the plastic strains, the pattern of strains will be dictated by each normal mode pattern of the individual component which in turn is being affected by its surrounding structural elements. You are not assured that you can “treat” a specific point on the structure. Further, points that do not require “treatment” may develop unwanted plastic strains.

VSR is therefore a cold work process not a relief process. VSR can not change the basic material properties as can heat treat stress relieving. VSR does not change the phase, bond or crystalline structure of the material as do heat treat stress relieving. Luder’s lines, stretcher strains or worms will form in the grain structure during the cold working process. VSR does allow the dimensional stability of the structure to eventually take place. VSR does have an effect on the fatigue life of the structure. The degree is dependent upon the duration and number of cycles the structure experiences during treatment. In BKlauba’s case, only 24,000 cycles were needed.

 

[Greenleader]

Very interesting. I need to read all the papers before I can comment though. Probably a couple of times.
Vibration is only half of the equation for our purposes at work. We found you could even induce stress by "masking" the part in certain areas.

Thanks for the interest, but would heat and vibration TOGETHER be able to improve crystal structure in carbon steel?

 

[jackboot]

We use VSR for dimensional stability. While I can't say exactly how or why it works - there appear to be enough experts in the room to explain their opinions.

However, consider we are using this to hold dimensional tolerances that are 0.0015 included (or +/- 0.00075) on a 19.75 inch diameter. This is a complex welded part that is made from high tensile strength steel (A514).

The parts are used in a service subject to fatigue loading with no problems to date. We have no basis to say it does or doesn't work - except our parts don't move.

jackboot

 

[bklauba]

Greenleader:

The source of residual stresses should not be an issue. I see no reason why, providing the geometry of the part and its condition are appropriate, that an effective vibratory stress relief cannot be performed.

Geometry (together with setup) will determine what opportunities you have to resonate the part, along with the limitations of your equipment. Max frequency and unbalance, speed regulation, and instrumentation of your gear are the parameters that limit its applicability. The physical condition of the material is also a factor, in that ductility is required. Severely through hardened or tempered materials resist the VSR Process, often to the degree of no benefit. On the other hand, as cast, welded or forged conditions are highly suitable.

To mtnengr:

I have taken a poll of my metallurgist friends, and they all agree that interpretting the modus operandi of vibratory stress relief as cold working is "wierd". (Some were less polite.) Few, if any, of the overall effects of cold-working can be detected on a real-life, large precision workpiece that has been VSR Processed. Rather, the workpiece appears to have undergone the benefits of long term storage (if it were a casting it would be called "curing".)

At the ASM Conference, one of the first questions asked after the paper was presented was, "Would not a better description of the VSR Process be vibratory stress redistribution?" My coauthors agreed with this point, as did I.

But here the dichomoty 'tween those operating out of a lab, and those "in the trenches" dealing with large, complex parts bearing exacting dimensional tolerances, starts becoming apparent. In response to the Q at the ASM, I added, "If I showed up at a machine shop manager's office, and stated that I was there to talk about vibratory stress redistribution, for the purposes of dimensional stability, I think that the manager was scratch his head, give me a confused look for a moment, but then say, 'Oh, you mean stress relieving. Why didn't you say so? Sit down, and have some coffee. Can you help us with this?'

Some describe this problem as one of semantics. I think it is a bit more than that, and the differences in benefits and limitations of both PWHT and the VSR Process should be both kept in mind and continue to be explored. In the paper we presented two of the many case histories that we have (at the conference, we presented five) where PWHT-ing was either unable or highly inappropriate to solve problems that were resolved with the VSR Process.

Picking up on the semantic arguement approach, some of the guys in the office have conjured up some alternative names for vibratory stress relief, which have been crafted so as to both minimize offending those who are so quick to engage in this arguement, plus add a comic perspective. My favorite (this week, at least) is: Kumquats on parade.

Shifting back to a serious discusion: In our private email coorespondance, I posed the following situation and question: If a company employing you as a consultant asked if the VSR Process would be a viable choice to enhance the dimensional stability of a production run of semi-trailer sized stainless steel or lo-C, hi-strength steel (say ASTM 514, Grade Q), what would you say? Would you answer that the, in your opinion, vibratory stress relief is cold working? . . or that it would be better to label it as vibratory stress redistribution? (Your answer to me directly was that this is a sales problem, not a technical one. You stated that you would tell it like it is.) But you didn't answer the question: WHAT WOULD YOU SAY to the company as to their selection/choices/options of processes to consider when faced with the challenge of machining such parts to respectable dimensional tolerances?

YES or NO (within the context described) to vibratory stress relief (or whatever you wish to call it)??

The work of Dawson & Moffat (Doug Moffat sent me a fresh copy of the paper, with his compliments), is sited in the ASM Paper, along with Yang's. Yang's work bears some resemblence to Bill Hahn's work, which can be seen at:

ceer.alfred.edu/research/vibatory.html

When the ASM paper is posted on the net, I will add/post the website address as a followup to this corresopondance. It should be out within 60 days. (The lawyers are still haggeling over security/proprietary/patent issues.)

Bruce Klauba
Airmatic Inc.

 

[mtnengr]

To B Klauba:

You have a product. That product was hopefully developed by some engineer. In that process, the engineer tested the product to determine just what could be said about the functionality, its limitations and purpose of the product. ABC company buys your product (Kumquats on parade) and uses it for relieving stresses on a cast steel gondola they manufacture. Due to your product, stresses are introduced that change the state so that the critical transition temperature is above 40F. Your engineer did not investigate this case. The environmental service temperature drops, the part fails and many people are killed. If you did not warn ABC company not to use your product for that environment or did not disclose the fact that its properties could be dramatically altered to a near failure position, your legal staff would be active for many years trying to defend your company. A similar scenario can be envisioned for your semi-trailer customer.

The point is that you are making a claim that VSR is better than sliced bread. Prove it—under all of the circumstances that you envision. Define the intent and purpose of the product. Use accepted engineering terminology to describe the process. Some of your competitors just say “I don’t know how it works, but it does.” In my opinion, this is worse than saying nothing at all.

The problem I have with your present discussion is that you are not using nor referring to terms I am familiar with. For over fifty years, I have heard most of the ubiquitous statements provided to me by my customers during my long term consultancy. “Trust me” is one of them, but not accepted. It appears that we are at that point in re-evaluating VSR.

I hope that we can agree that VSR is a method that introduces vibratory stresses in a part and in turn suggests the following items need to be considered:

1. These stresses can only affect the final state of stress (strain) at any given point. At low amplitudes, the stresses do not alter the bond strength of the grain structure and do not change the basic crystalline structure of the material.

2. Because of the limitations of forces needed to excite the structure, VSR dwells at a natural frequency of the structure. Therefore, it can only be effective at specific nodal patterns associated with that natural frequency. Forces can be introduced that are proportional to the damping ratio, 1/2?. This natural frequency may become nonlinear due to the softening of the material resulting from plasticity. Otherwise, very large shaker systems would be required to introduce adequate stress levels.

3. The final (superposed vibratory stresses and the original residual stresses) stress state at these nodal patterns often approach the presumed elliptical failure surface for that material. Other failure surfaces are possible and its shape is dependent upon the class of material. The failure surface (that relates the three principle stresses) can be considered to be fixed in space and ever expanding and convex or the entire surface can shift in space.

4. There is no evidence that VSR alters the creep characteristics of the material. If VSR does, then a relaxation function needs to be defined as a function of VSR levels.

5. If VSR is the process as you have described and if you believe that the material does not experience plasticity, you need to describe the mechanism that alters the residual stresses within the body. Heretofore, the referenced finite element approaches are on the presumed thesis that inelastic behavior is that mechanism. It should be noted that with the recent demise of Dr. Hahn, his study can not be effectively examined.

You are only discussing the effects of VSR on dimensional stability and not the other possible side effects. If you have not established the true limits of the process, you had better clearly define what the limitations are. Otherwise there will be a new class of lawyers on your heels.

With respect to your question (YES or NO, within the context described), it would be NEITHER if I did not have laboratory proof of the limitations under those exact (or near) circumstances or it would be YES or NO if I did have proof.

Prior to 1980, VSR had not been “accepted” as a stress relief process. Twenty-five years later, there are more suppliers in the field, but I have not seen any advancement on the definition of its limitations. The fact that you tie down some eccentric mass to a part (no matter how massive) and vibrate that part near or at a resonant frequency really does not demonstrate the effectiveness of the process. Since dynamic stresses are introduced throughout the part and not just at one locale, the sparseness of the induced stresses needs to be addressed. If the resulting stresses are below the proportional limit of the material, no change in the material properties or dimensions can take place. It is though you had done nothing. The only possible change in the structure’s dimension would be due to creep or a change in load—like transporting and/or material handling. The magnitude and direction of these new principle residual stresses under these new conditions would have to be known and their effects on the dimensional stability needs to be determined.

There has been a review of studies by H. Mughrabi and H. Christ, “Cyclic Deformation and Fatigue of Selected Ferritic and Austenitic Steels: Specific Areas”, ISIJ Intl, Vol. 37 (1997), No. 12, pp. 1154-1169. I would hardly classify Dr. Christ as “weird.” They make the same claim (as I) about the presence of Luders bands ( a result of the cold work process). They go on further to describe the behavior of steels when they approach their fatigue limit. And Yes, plastic strains are observed and the grain structure is changed into what is called “dislocation arrangements.” Thermomechanical fatigue is also touched on.

It has been suggested that using VSR during the welding process, residual stresses are greatly reduced. (S. Aoki, T. Nishimura and T. Hiroi, “Reduction of Residual Stress of Welded Joint Using Vibrational Load,” SMiRT 16, Wash. D.C., Aug. 2001). This would suggest that the material properties are being accelerated during the solidification of the weld material over the “normal” VSR process. The mechanism that reduces these residual stresses has yet to be identified.

It would be interesting to have a blind test as to the effectiveness of VSR. If one had the resources, prepare six identical specimens for each material class. Produce three specimens that have residual stresses built in, three that are virgin. Using transmission electron micrographs or X-ray diffractometer, establish the residual stresses before and after treatment for each specimen. Pass these specimens to VSR suppliers, but do not tell them which specimen is treated. Let the results fall where they may. I tried this challenge, twenty-five years ago and got no takers. I wonder if now we have any advocates that would take the bait. NIST, are you interested?

How does VSR affect these “dislocation arrangements”? Does VSR reduce or enhance the fatigue life of the structure? Does VSR reduce the impact strength? Worried? I would be.

 

[jackboot]

mtnengr:

I would say you know a great deal about metallurgy.

Would you explain the phenomenon whereby highly stressed steel parts will slowly "stress relieve" themselves over a period of time (with no thermal stress relief)? I recognize that the time span may well be years, decades, or possibly greater.

How would this mechanism be any different than that of VSR?

Why does thermal energy allow for "slippage" in the crystaline structure but kinetic energy does not?

The second law of thermodyanmics allows for everything to seek a "lower state".

Please explain. I would think the two mechanisms would be the same.

jackboot

 

[mtnengr]

To Jackboot:

For thirty years, I was V.P. Engineering of a lab that performed failure analysis on a variety of structures and conditions. My principal job was to determine or oversee the determination of mechanical stresses these structures would experience. Why did the part fail? The metallurgists would provide me with the mechanical properties, the modes of failure and alike. Maybe some of that information rubbed off.

Your second point is answered by my point (4) above. If VSR affects the creep properties of the part, provide me with the new creep relaxation function with respect to VSR vibration levels. But, others have suggested that the level of plastic strain affects the condition of the material. Can both be right?

The use of temperature on a part can directly change the phase of the material. Any elementary metallurgy textbook should give you a phase diagram for the presence of Ferrite and Austenite and what the grain structure would look like. Similarly, non-ferrous alloys have equivalent phase diagrams. The content of steels has a direct effect on the mechanical properties and hardness at these temperatures. There is a process called aging or age hardening. Aging may take place at room temperature or at elevated temperatures. Just how these changes are made can be best described in a “Bain-S Curve” relating phase change over time at specific temperatures.

Slippage of the grain structure generally has to do with the shear stress properties of the material. When a cold-worked material is subject to prolonged annealing temperature, the material recrystallizes. Grain structure improves and many of the mechanical properties also improve.

Kinetic energy can be transformed into work. The work done has to do with the determination of stress and strain over the volume less any thermal work being done either through internal or external losses. The work done is conditioned by the physical properties of the material. So in a round about way, kinetic energy can affect the slippage properties.

Dealing with the 2nd Law of Thermodynamics and material properties is well beyond by expertise and comprehension. I might suggest M. E. Gurtin, “Modern Continuum Thermodynamics”, Mechanics Today, Vol. 1, 1972, pp 168-213, Edited by S. Nemat-Nasser.

 

[Greenleader]

This has been a very interesting thread.
I'm dying to share what we found, and our solution, but they didn't use any of my money on the R and D, so I hope you all understand I'm not at liberty to say much.
Like I mentioned, vibration is only half the energy that we utilize. Metengr's point #2 is spot on. Our "apparatus" are very large, and can be adjusted both amplitude and frequency, among other adjustments. The shape and the size dictate nodes, or you can cheat by "masking" part of it. Keep in mind we use them strictly for aluminum forgings, and for "dimensional stability"

That being said, I would love to participate in metengr's experiment, I really have an opened mind, but alas I work in a production shop, that rarely has time for good R & D.
We have been doing fairly well recently, and have a new CEO as of today. Mayyyybe he is a good ole' engineer.

Now, how about vibes on metastable steel? Would it help austenite "slip" into martenste?

 

[unclesyd]

Greenleader

He probably brought some tape with him to stick the vibrating parts back togather when they break.

 

[Greenleader]

Your post is a classic example why letterman doesn't hire engineers to write jokes.
At least metengr addressed specific points.
KMA, Greenleader

 

[unclesyd]

b]Greenleader[/b],
My post was in relation to your comment on the new CEO and your apparent relation to the aircraft industry were a highly publicized CEO swap occurred.

I hope you understand that vibrations and aircraft are not normally considered good bed mates. One of the foremost airframe designers of my time, not yours, was once asked how to keep the wings of new design from cracking from vibration and falling off at a specific point. His answer was we need to perforate it like toilet paper, since it doesn't ever tear where it supposed to.

Now that we are the same level we can let the thread continue.