Have to be careful with this one. The US have a good term that VSR fits into. Snake Oil. It is very hard to find objective scientific and repeatable data about the technique. I put it into the same category as cryogenic treatment when applied to non ferrous materials. If the magnitudes of the vibrations are sufficient to induce micro-plastic events then perhaps it might work. That means shaking very vigorously in my opinion. The only way to be sure is to measure the RS before and after. But so little of the "literature" does this.  

Ultrasound Impact Treatment is often used for relieving the tensile RS of weld joints. The effect of UIT is twofold. First effect if just changing the stress sign from tensile to compressive due to subsurface work-hardening. The secongd is that ultrasound oscillations are excited and then penetrate inside the body producing the same effect as tempering, i.e. reducing the residual stress. The RS are measured using different techniques such as  X-ray diffraction, ultrasonics, by drilling holes and strain gauge transducers.

VSR to me is an interesting issue which, to the best of my knowledge, has not been addressed satisfactorily in the literature in terms of solid materials science evidence of the changes - or in terms of measured residual stresses.

(Sergei - Please correct me if I am wrong, and send references ! )

I'd be very interested to know if you manage to get some solid results - we can discuss how my group could contribute in terms of modelling and residual stress measurement.

It is a question very interesting and i get only theoretical studies about. There are some welded structures that where subjected to that treatment but the results was not so good represented and until now we don't know if that treatment can be computerate with the tempering. Personally i was the designer and nobody can assure me that the bulk of the material change his RS level or the microstructure. 

1. Vibratory Stress Relief is a controversial subject, and I agree with the previous responses in that regard. You can find many papers in the literature (try Google Scholar). Most of the evidence suggests that purely elastic vibration does not relieve the residual stress. Rather, the literature mostly supports the conclusion that the vibration must bring the total stress level (residual stress + vibration stress) above yield in order for the residual stresses to reduce. That level of vibration can be hard to achieve. VSR might work, but you are wise to be wary.

2. I assume you need to measure those stresses nondestructively if you do not have an extra part you can sacrifice. X-ray diffraction, when done with sufficient expertise, can give you a good value for the surface stress only. Measuring subsurface stresses nondestructively in that large part would be less practical.

Am I correct that your parts are new and the contact surfaces are already deformed, or did they deform after use? Was the “heat treatment” you mention done before or after the contact surfaces were machined?

Dear Colleagues,

I agree that ultrasonic impact peening mat have minor effect of relieving the internal RS. However, I found some publications relating to this topic when doing literature research on UIT.

Article The effect of tensile deformation by in situ ultrasonic trea...

Article Acoustic softening in metals during ultrasonic assisted defo...

Article Physics and mechanism of ultrasonic impact

It is given as RS. what is it ?  Is it  compressive or tensile stress?  As suggested by Omar, you can go for shot peening which can reduce tensile to zero or to compressive.  Laser shock peening(LSP) is also an option.  shot peening can bring some surface roughness and LSP can be a better choice without drastic modification in the surface roughness(in fact many a times no change in surface roughness). Peening process induces compressive stress, so it is your choice to check if compressive stress is ok.  you can find lot of publications. Xray diffraction technique can give the magnitude of residual stress, but for this the job seems to be too big.  If destructive method is ok you can try with hole drilling technique.

There are three types of residual stresses in metal body depending on the scale. In this case we are talking about macrostress that exists between different parts of the cast item. Therefore there could be both tensile and compressive stresses on opposite sides. I believe that no surface treatment is capable of relieving the macrostress.  The bulk of the item could be treated using retempering or acoustic softening by introducing ultrasonic oscillations or through passing high-current pulses since electroplastic effect . 

I agree with previous comments (Michael B. Prime) , any technique will only be useful if the combination of residual stresses and the vibration stresses is larger than the local/ global yield strength. I would suggest that you perform X ray at the outer surfaces ( after removing the color of course) and measure the surface residual stresses. if they are high enough ( relative to the local yield strength), then there might be a chance to affect them by application of some elastic stress. Though, I must agree with other colleagues on the fact that we lack some solid research results on this matter. 

That is the result of such treatment will be somewhat difficult to predict!

Hi Karel,

VSR is also for me an interesting issue.  According to the practical experience from my previous activities, the procedure is  applied more widely in the United States.In Europe there are currently some reluctance. 

Attached I send you also two articles on the subject.

There is a well-known effect of fatigue (vibrational) stressing on samples containing residual stress, either through work-hardening or other processes.  This is known as cyclic softening, and research on this topic goes back to Bauschinger.  There is a huge literature concerning it.  Coffin showed how it could be used in processing to increase the ductility of a metal being cold formed so that it could be formed further.  Cyclic softening usually can be accomplished in relatively few cycles, depending on the strain amplitude of the cycle, before the onset of fatigue damage.  Fear of creating damage may be one cause of reluctance to use the method.  The questioner is encouraged to search the literature under the term: "cyclic softening".   Residual stress can be determined non-destructively by X-ray methods provided the metal is not severely cold-worked, and the geometry of the subject product allows access to the region of the residual stress.

Given the geometry of the component I would see re heat treatment as the best option for distortion control. I have come across this company that offers the service commercially. I wouldn't expect too much from a theoretical understanding perspective, but they certainly are very experienced in the practical application of the technique

http://www.v-s-r.co.uk/

We have been investigated efficiency residual stress relieving by VSR method and have capability for it. We also have capability for residual stress measurements. We discover that sometimes effect is good sometimes not sufficient what mostly depend on the shape of construction. More rigid construction has not reduced residual stress enough, but if construction is not so rigid, the effect is better. In some cases we manage to reduce residual stresses especially in weld joints for approximately 30%. We have been measured residual stresses before and after using VSR method. VSR method has also other advantages like dimensional stability of construction during machining. The constructions which are VSR treated are more dimensional stable. But I advised you when you use VSR treatment, please measure residual stress before and after treatment, this will for sure show how efficient the VSR treatment is. Sometimes you will be surplices how VSR treatment efficient is sometimes not. A lot depend on the shape of construction so measurements of residual stresses are required too to be sure, the drop of current in engine is not enough.

Here is one of my article and example of VSR treatment and RS measurement on part of vehicle transporter for liquid slag.

Have to be careful with this one. The US have a good term that VSR fits into. Snake Oil. It is very hard to find objective scientific and repeatable data about the technique. I put it into the same category as cryogenic treatment when applied to non ferrous materials. If the magnitudes of the vibrations are sufficient to induce micro-plastic events then perhaps it might work. That means shaking very vigorously in my opinion. The only way to be sure is to measure the RS before and after. But so little of the "literature" does this.  

I am on the Board of a company founded by a scientist (and FASM) who has something like 33 U.S. and international patents for an outstanding process using vibrational energy to reduce stresses, which actually sometimes increases strength, as well.  You might want to contact Donna Walker, at [email protected], and let her know that Bill Tobin referred you to her. .  

Jeremy Robinson has it spot on. Any mechanical technique can only reduce residual stress by plasticity. Plasticity + lots of vibratory cycles = fatigue. Attached is an old paper where we did some before-and-after measurements and saw no effect. Essentially, look at the origin of the residual stress and look for mitigation techniques that can be validated and don't potentially cause hidden damage.

I was sceptic about it, but I can prove that it works in welded structures made from stainless steel. Small experiment should not work, mass and inertia is important to introduce the required forces

We are in process of experimenting with US on aluminum alloys to find out the effect of US waves on the bulk material structure as compared to that of well-known US impact treatment when  the most part of the US energy is spent for subsurface plastic deformation. It was revealed already that the US wave amplitudes on the workpiece surface are at least  one of magnitude  higher in the first case. 

Cyclic softening under repeated stressing was first discovered by Bauschinger and rediscovered decades later.  The science of cyclic softening has been thoroughly studied in the last few decades but as frequently happens, not thoroughly enough.  Coffin patented a cyclic softening process of  intermediate "annealing" during cold rolling of metals.  Cyclic softening is most effective in reducing the hardness (and associated residual stress) if the stress is high enough to excite dislocation motion.   In theory, it is probably possible to reduce residual stress by vibrations of sufficient amplitude.  As Jeremy Robinson points out, this is going to be difficult in practice.  I cannot remember the report (at least in the open literature) of the limiting conditions needed to reduce stress or hardness of a hardened metal or alloy, by cyclic softening, and for different  methods of hardening.  I think such a project could be very interesting and useful.

I believe cyclic softening has almost nothing to do with US-assisted softening or acousto-plastic effect first discovered by Blaha (1955). MD simulaitons show that vibrations imposed on static indentation greatly increase the kinetic energy of the atomi sytem and thus generate dislocations.

Article Vibration and Acoustic Emission Monitoring the Stability of ...

Article Ultrasonic-assisted aging in friction stir welding on Al-Cu-...

Sergei and I have to disagree.  US vibrations are just high frequency vibrations superimposed on a static stress (in his limited application) so as to produce cyclically induced creep.  It is just a matter of mean stress and how many cycles are applied.  If the material is already hardened by cold working, then cyclic softening will occur.  It is interesting that people with different backgrounds differ on the basis of the assumptions introduced from their experience.

please review the attached

Article Using DIC on VSR in welded structures

Karel,

If you are dealing with a power plant generator rotor you are working with serious people and proper engineering.  I find it difficult to imagine that such people would make a facile recommendation in the absence of underlying experience and confidence. 

I can quite imagine that the reason for loss of tolerances has been residual stress effects and I can imagine that you may be correct that the underlying cause has been inadequate stress relief during manufacture.  However, I can equally well accept that cyclic vibration might be successful in relieving such stresses as the manufacturer has indicated. My question therefore would be, do they have practical successful experience of applying the concept?  If the answer is 'yes' and they can provide documentary evidence I would be inclined to allow them to give it a try in this instance as, presumably, they would be picking up the cost as the underlying cause of the current issue is their responsibility and I cannot imagine that applying the procedure would result in additional damage to the component.

You question how you could verify the effectiveness of the repair.  I would say that's simple enough - either the rotor returns to its original clearances or the repair hasn't been effective, in which case the manufacturer would have no option but to strip, heat treat and re-machine the rotor.  If the rotor goes back to original tolerances, I wouldn't expect that it would require further attention in the future because the deleterious residual stresses would no longer be present. 

I can sympathise with the reservations expressed by other contributors but if this were my problem my response would depend strongly on the established reputation of the rotor manufacturer and my confidence (or lack of confidence) in them, together with the relative ease or complexity of applying the vibrational stress relief procedure compared with the outage time for a complete strip/stress relief/re-machine/reassembly of the installation.   

If the major focus is upon dimensional stability, vibratory stress relief, if properly performed, is a viable alternative. For a number of applications, such as stainless steel components, bimetallic components, and HSLA steels, all of which run degrees of risk if thermally processed, it might well be the best choice.

Also, a significant number of machine tool builders, Ingersoll Machine Tools being one example, use the VSR Process both before and after rough machining, finding the ability to meet very tight target dimensions on large mild steel weldments, better than using PWHT. I invite those who have answered this inquiry to visit our website to see more examples and case studies (one such attached).

Let me reiterate. If you have a residual stress in a structure, that can only be relaxed by the movement of atoms or the creation of interfaces. In practice this usually means thermal treatment leading to diffusion; plastic deformation on a local or global scale; or micro-cracking. If you have a safety-critical component, you have to understand what effect a vibratory technique is having on the material, and not just the effect on residual stress or distortion, however beneficial.

The vast majority of our customers are using our process and equipment for purposes of dimensional stability. However, it also has been used and increasingly so, upon structures that undergo significant loading. Examples include press and shear frames, waterway gates, and (one of the larger set of structures we have done), the storm-surge gates for the City of New Orleans, which includes the 5 story tall, 177 ton pair. See http://www.nytimes.com/2012/06/15/us/vast-defenses-now-shielding-new-orleans.html?src=me&ref=general

You can see pictures of these setup for VSR Treatment on our website, both on the homepage and on our PWHT Risks page (since the construction was low-carbon / high-strength steel).

Addressing your answer, it is hard to imagine any form of stress relief occurring without plastic flow occurring. On what level and range remains an issue. The reduced effectiveness of our VSR Process on hardened or cold worked materials we believes give some clues as to the process's limitations and even the mechanism involved. Small and thick parts (low aspect ratio) are difficult to do, often having resonances outside the range of our equipment (150 Hz, but we are working on that.)

In practice, companies such as Ingersoll have pointed out that asymmetry of a structure can cause a problem during cool down in a furnace, establishing stresses that have little or nothing to do with the weld seams and HAZ's, but rather cooling pattern. They have found that they can solve the problem if they lengthen the cooling period, but find it impractical to leave, for example, a 40 foot long, 25 ton structure in their furnace for 4 - 5 days. Using their VSR gear before and after rough machining gives them better results, dimensionally, with the two treatments taking together less than 4 to 5 hours. They can then machine to accuracy to within several thousandths of an inch, full length.

These are but a few of hundreds if not thousands of parts that undergo such treatment, every year. What we have been working upon for the last few decades is identifying performance and workpiece criteria that allow our track record of such results to continue.

Is more research needed, of course, but it should recognize what the needs of industries are, that use this process on a production basis, with a raging river of anecdotes as their reference point.

How best to define standards or at least guidelines to those who would most benefit from this process ? ?

How best to minimize misapplication or misunderstanding of this technology ?

Can we work together on these questions ?

Following..

Many stator frames for hydro are built in segments, due to their size being prohibitive to ship. But they change shape during transport, which might be looked upon as an inadvertent type of stress relief. A bit of a problem during reassembly, and expensive.

Here is a photo of a 6 segment, 16 to 18 m diameter stator frame segment, all of which (36 pcs total) will be VSR Processed in Brazil, saving extensive fit-up problems in the power-house install.