8/12/2019 1991108

1/4

N EXPL N TIONOF THE MECH NISM NDBENEFITS OF SHOT PEENING

It is possible to enh ance certain properties of some m aterials oenable them to b etter meet the needs of critical designs. Mostof u s are familiar with the us e of h eat treatments to impr ove theproperties of metal alloys. Such treatments can be used toincreas e he ha rdnes s of a knife blade so that it will remain sharpfor a longer period of time than will an untreatedblade. S imilarly,the strength of a bolt can be increased through heat treatmentso that a smaller bolt can be used in an application that wo uldotherwise require a m uch larger bolt.

Another process that is used to improve properties is shotpeening which is the process in which hard shot is hurledagainst a work surface. In this case the properties that areenhanced are resistance to fatigue fracture and resistance tostress corrosion rather than strength or hardness.

Traditionally, shot peening s done by spraying hard shot againsta worksurface in afashion similarto sand blasting. Shot peeningcan also be done with captive shot where the shot is ntegratedinto a rotating brush or flap. In this case , the spinning brus h orflap is placed in close p roximity to the work surface so that thecaptive shot strikes the work s urface with each revolution.

In order to unde rstand he bene ficial effects of shot peen ing onemus t first understan d the m echan ics of surface stresses, andthe effect that peening has on these stresses. To this end, thefollowing discussion of these concep ts s presented. The detailsof both fatigue failure, and stress corrosion will also be review edbut not un til aftera preliminary discussion on surface stresses.

Surface Stresses and eening

As a means of unde rstanding he concep t of surface stress, andthe ben eficial effects of peening, one can begin by individua llytaking a close look at the concepts of peening , compressivestresses , and tensile stresses .

Peening Peening means nothing more than to ham mer. Shotblasting or rotopeening effectively hammers asu rface, althoughwith very small hamm ers, and with a large number of strokes .

Compressive Stresses Most people are probably amiliar withthe concept of compression but probably with differentconsiderations. Compression means nothing more than tosqueeze toge ther. If a part is squeez ed in a vice, it is subjectedto com pressive stresses.

Tensile Stresses Most people a re probab ly also familiar withthe concept of tension, which results from pulling forces. Atightly-stretche d wire is subje cted to tensile stress es.As an aid to be tter understanding the concepts of compressionand tension, consider a very fine bar of metal. A bar so fine, infact, that it is compos ed of asin gles tring of atoms. To understan dhow such a bar will function, consider that the atom s behave ikevery sm all, sticky and flexible spheres, much like rubber ballscoated with adhesive.

0 t

1:tress

rofile

STRESS-FREE BAR ONE ATOM THICK

Here is such a bar in a stress-free condition. It is stressbecause there are no com press ive or tensile forces acting oA stress profile is used to indicate the level of stress oveentire cross section of the bar. In this cas e, the small grapthe left of the ba r is the co rrespo nding stress profile. Convecalls for c ompressive stresses to be given negative values,for tensile stresses to be given positive values. In thisexample, there are n o stresses as is indicated by the neustress profile. It will be foun d that the s tress p rofile will source of some very valuable nformation. B ~ g i ny considwhat will happen if the bar is p laced in tension, as it would

one end were secured and the other end were pulled on brope.

BAR UNDER TENSILE LOAD

The bar was pulled on. The individual atoms elongated.\the whole bar got longer. The atom s are under tensile strThis is shown by the stress profile diagram. Notice thastress profile diagram indicate s that the entire cross-se ctiunder tensile stress. Continue by further stretching the ba

BAR BREAKING UNDER EXCESSIVE TENSILE LO AD

Here the applied tensile force exc eede d the forces that bothe atoms together and the bar broke. This is how fracoccurs. When tensile forces between atoms a regreaterthan forces that bind them together, the atoms separate and piece breaks.Now, let us examine com pressiv e forces. Beg in by considewhat will happen if one end of the bar is secured againblock, and the other end is pushed on until the bar has bshortened by one atom diameter.

STRESS FREE BAR BEFORE COMPRE SSION

ontinued on Page

The Shot eener Volume Issue a

8/12/2019 1991108

2/4

- *.1 :Bm

4 .

->*&$. .

Stress Stresarollle rofile

BAR UNDER COM PRESSIVE LOAD BAR TWO ATOMS THICK UNDER TENSILE LOAD

Notice what happened. The bar was pushed on and as a resultthe atoms got a little narrower and a little taller as the bar gotshorter. These atoms are now under compressive stress. Butremember that atoms are springy, and that they would rather beround. They are pushing against the blocks and if the blocks

were removed would return to its original length. What if theexperiment iscon tinued and the bar is comp ressed an additionaldistance?

roflle

BAR UNDER H IGH COMPRESSIVE LOAD

Notice that aga in the atom s got narrower and taller as the bar gotstill shorter. Notice also how the compressive stress alsoincreased (as is indicated by the stress profile diagram), but thebar did not break, as is did when excessive tensile forces wereapplied A crack can never form in a bar where there are onlycompressive forces. The bar may eve ntually "buckle", but failureby crack propagation will not occur. Is there another way inwhich a compressive stress can b e created in the b ar? What ifeach end of the bar were restrained and another atom placed ontop. If this extra atom is then knocked into place, the new atoman d all the original atoms will be com pressed.

Whe n a tensile force 1s applied to the bar it elongates , just asdid before. Notice that the tensile stresses are distributeuniformly across the cross section of the bar . This isshown in tstress profile.What about placing the bar in compression? This can be don

just as before by block ing one end, and then pushing the othend in the distance of one atom diam eter.

rollle

BAR TWO ATOMS THICK UNDER COMPRESSIVE LOAD

As in the previous case involving compression, the individatoms got narrower and taller. Notice that the compressivstresses are distributed uniformly across the bar's cross sectionas is shown in the stress profile.

What would happen if an additional atom were squeezed inthe top layer? This could be done without securing the bbetween blocks, or pulling on it with a rope. In other words,external forces (and therefore no external stresses) need beapplied to the bar.

+

- o

I: . tressrollletress

rofile

UNRESTRAINED BAR OF TWO-AT OM LAYER THICKNESSSTRESS FREE BAR RESTRAINED AT EACH END

- o

4tress

rofile lI

tressrotile

RESTRAINED BAR AFTER INSERTION OF ONE AT OM

Sure enough, the bar is now in compression. In fact, it looks ustlike it did after it was compressed by the movable block in theprevious example. Also, since it is in compression, the atomswant to return to their unstressed state and try to do so bypushing on the fixed blocks. This is essentially what hap penswhen a part is peened. Surface atoms are d riven into the part,and the surface is pu t in compression.So much for tensile and com pressive stresses on a small bar.Now consider a bigger bar, one that is two atoms thick, and seewhat happens when it is put in tension.

UNRESTRAINED BAR OF TWO-AT OM THICKNESS AFTER INSERTI

ONE-ATOM INTO TOP LAYER

This is very interesting. The top layer is in compression , just when the bar was a single layer between blocks. Apparently, thbottom layer is able to restrain the top layer as had the blocks ithe earlier example. But look at the bottom layer In restrainithe top layer, the bottom layer gets stretched. Look at the stresprofile diagram. It verifies our suspicions, showing that the tlayer is in com pression, and that the bottom layer is in tensio

ontinued on Page 9

The Shot Peener Volume 5 Issue Page 8

8/12/2019 1991108

3/4

Notice something else. The top layer elongated because of theadditional atom . The bottom layer also elongated , but not quiteas mu ch as did the top laye r. Since both layers are tightly boundtogether, this differen ce in elongations caused the bar to bend.Cons ider what would happe n if the bar were too thick to ben d.

% eop layer would still be in compression, and the layerme mediately below would still be in tension. The core would b eunaffected, and without external forces, would be under nostress.

tressrofile

LARGE BAR PEENED ON TOP SURFACEatoms a re now too sma ll to b e individually seen)

To further extend this experiment, consider a very large bar thathas been peened over its entire surface. Such a bar is shownbelow, along with its stress profile diagram. Notice hat the outerlayer is in compression, the layer just beneath the surface is in

tension, and the core is under no stress.

LARGE BAR PEENED ON ALL SURFACES

Now, considerw hat w ill happen to such a bar if it is put in tension.This would happe n if something useful were done with the barlike using it to suspend a bridge deck from a truss.

LARGE BAR PEENED ON ALL SURFACES SUBJECTED TO A TENSILE LOAD

By putting the bar in tension, tensile forces were added to theentire cross section. The core, formerly unstressed, now sees auniform tensile stress. The layer ust below the surface, formerlyin tension, now sees even grea tertensileforces . The outer layer,formerly in compression, is now neutral -th at is, under no stressat allThat is remarkable By peening the surface of a bar, it is possibleto create conditions that result in zero surface stresses, evenwhen the bar is under a total tensile load. More than remarkable,'%iss highly practical. Rem emb erthatcrac ksform , and structures

only where tensile stresses exist. Tensile stresses areespe cially detrimental if they occ ur where sma ll cracks (or evenscratches) already exist. These cracks elongate when they areexposed to tensile forces, especially if the tensile forces are

cyclical in nature. There will always be scratches on the s urof a bar. There shou ld, however, be no scratches inside of a bIf conditions exist such that tensile forces are not on the surface, then any pre-existing surface imperfections will nocaused to grow. By keeping the outer layer in compression,possible to g reatly increase the performance characteristica given structure.

It should be pointed out that if the surface cr ackgoe s throughcompressive layer, that it will enter the area of high tenstress, and catastrophic failure will likely occur. It is for

reason that defects must be removed if peening is to hadesirable effects.

Metal Fatigue and Stress orrosion

Metal fatigue was alluded to in the previous discussion wcrack p ropaga tion, as a result of cyclical tensile stresses, wmentioned. It is actually the growth of s mall cracks that cauastructure to fail in afatigue mode. Since such a structure wohave been able to p reviously withstand much greater loads, tmod e of failure is just as strong as it ever was. F ailure occurbecause unobserved and growing cracks effectively reducthe cross section of the member until it no longer was ablesupport the previously acceptable load.

If you watched ten thousand logging trucks cross a highwbridge, and then the same bridge failed with just you and yVolkswagen on it, you could reasonably assume that you wethe victim of a fatigue failure. What prob ably happen ed was some initial flaw in a critical member was caus ed to grow etime a truck passed and put the part in tension. The craincreas ed in size with each a dditional truck until the truckbefore you caused the crack to g row to such an extent that tcritical mem ber was not able to even sup port the load of you ayour ill-fated vehicle. If the part had been treated so thatsurface never saw tensile stresses great enough to cause cracpropagation, then fatigue failure would not have occurred.Stress corrosion is a higher class of corrosion than the type thmost peop le are familiar with. We know that iron will corroit is expos ed to salt water. In this situation, only two thingsrequired: asubs trate (in his case iron), and acorros ive agent this case salt water). F or stress corrosion to occ ur, three thinare necessary. As before, both a substrate and a corrosivagent are required, but there is the additional requirement ththe substrate be under a tensile stress. If only the corrosiagent, or only the tensile load is present, stress corrosion will occur. In other words, if either the tensile stress or corrosagent is p revented, then stress corrosion also will be prevente

Some stainless-steel alloys are subject to stress corrosiwhere the corrosive agent can be what would otherwise harmless environmentalsalts. Corrosion s prevented by assurthat parts that are e xposed to the environment never experientensile stress es. In the case of aircraft landing gear, this is doby shot pee ning the expose d surfaces. This is why it is vitimportant to re-peen any areas in which corrosion or otdefects were ground out. Not only did the grinding operaremove the surface defect, but it ikely removed he compressstresses that h ad b een imparted with the original peening. If tarea is not properly re-peened, then corrosion can be expecteto reoccur, and will probably occur much more rapidly second time around.

Continued o Page 1

The Shot Peener Volume Issue Page

8/12/2019 1991108

4/4