Is the residual stress always developed and stayed in a specimen because of experiencing plastic deformation and then eliminating the external forces. Or there is other ways of inducing residual stress too?!
Residual stress in ceramics and glasses can not be relieved by plastic deformation like in metals.They can be caused by chemical treatments and ion implantation of surfaces as well as during bonding of materials with different cofficients of expansion. In glasses fast cooling will cause residual stresses in cooled glass that can cause cracking. That is why Hhglass neede to be anneled to minimize residual stresses.
When consider a crack, it can be appear à stress storage arround it after an open, a shearing of the lips and a closure of these last. There's no irreversible strain but residual stress appear. It is just a meta-stable state.
As far as I understand this term, stress is caused by strain and strain (theoretically) has nothing to do with plastic deformation. If a material is completely plastically deformed, there is no volume change (compared to the initial state). Only elastic deformation changes the volume.
Unfortunately, in practice we always have materials which are plastically deformed which still contain elastic components measureable as strain and therefore stress. On only plastically deformed material you cannot discover any stress (or strain).
Yes, for instance in composites where you have mismatch in coefficient of thermal expansion between the constituents. This mismatch results in residual stress after cooling down from processing temperature to room temperature.
Change of composition due to diffusion can also lead to residual stress. This is in particular the case upon introduction of interstitial elements into the surface of metals by carburising or nitriding. Of course the stresses can be that high, that a large portion of the residual stress is accommodated by yield at the process temperature. Nevertheless considerable stresses can be retained after the process, whereby thermal expansion mismatch will then also contribute, as mentioned above.
to Gert Nolze 1) "... strain (theoretically) has nothing to do with plastic deformation" - how can you characterise the plastic deformation in a simple tension test if not using plastic strain Epsplast=Lresidual/L0 - 1 ? + a billion more cases for which the plastic strain tensor Eijplast was invented.
2) "Only elastic deformation changes the volume." - this is an approximation which does work in many cases, but not always; in many other cases the pressure dependent plasticity has to be used. On the other hand, there are material classes which are elastically incompressible, i.e., elastic deformation doesn't change the volume there (Poisson coeft. = 1/2), which do manifest booth elastic and plastic behaviours. 3) "On only plastically deformed material you cannot discover any stress (or strain)." What would you discover then after a homogeneous tensile testing beyond the yield point and unloading? Elastic deformation is absent there (in homogeneous crystal, the lattice returns completely intact). If you see the specimen longer, what does remain there then? Only plastic deformation and corresponding strain as its measure in relative terms (return to comment 1 above).
Anyhow, superposition of elastic and plastic strains is behind this kind of residual stresses; better to say, the result of this superposition (additive or multiplicative in corresponding cases) = the total strain (as well as deformation) must satisfy the strain (deformation) compatibility condition at any rate, and this "must" can lead to residual stresses. In particular, inhomogeneous plastic strains are one of the ways to generate residual stresses.
Turning to the original question, I'd say that every phenomenon that per se causes its proper local deformations (represented by strains in relative change terms), which are incompatible (are in misfit) with surroundings -- plasticity, phase transformations, melt solidification, composition and thermal inhomogeneities, .... , all them must generate stresses without external loading merely to restore the strain compatibility.
Yes it is possible. The simplest example comes from thermal misift strains in parts with more than one material. That can happen with or without plastic strain. Another example is from chemically induced misfit strains such as from nitriding of steel. See Article Residual stress Part II – nature and origins
for a more detailed discussion.
Non-uniform plastic deformation is by far the most common cause of residual stress, and often accompanies thermal and chemical misfits, but those can sometimes cause residual stress without plastic deformation.
I can see where Xu Wang is coming from, but his reply assumes that residual is used in the phrase solely as a synonym for ( the stresses left) after plastic deformation. And that is often the case, but you can see from the variety of correct answers, that in a wider sense, you could interpret the term as residual stress, as the stresses remaining after some/any process. That process could be a thermal cycle with or without a phase change (maybe just a non-uniform thermal field has been imposed). It could be where you have a multi-phase microstructure which has been through some thermal cycle and their expansion coefficients (of the constituent phases) are dissimilar, (here you could look for/at papers by Gladman et al on tessallated stresses and so on).
So as others have stated the answer to your question is yes, within a material the yield point does not necessarily have had to have been exceeded for there to be residual, elastic stresses present.
Yes. It can be. There are several mechanical treatments available to induce compressive residual stress into the metallic materials. If suppose the induction of residual stress is not reached the Hugoniot Elastic Limit (Yield strength) of the material then probably there will be no plastic deformation taking place in the material. But the actuated internal residual stress is beneficiary one to retard the fatigue failures.
Relating to the answer by Biljana Mikijelj, an example of chemically introduced residual stresses (used to improve mechanical properties of a product) is the “gorilla glass”. Here, as Biljana mentioned, ion introduced into the surface results in compressive stresses there and tensile in the interior of the glass; making the whole product ductile.
Machining processes and shot peening both produce significant plastic deformation. Those are not good examples of residual stress without plastic deformation.
I totally agree with Dr.Prime : The basic necessity to have residual stresses is to have a "misfit". The misfit maybe because of differential inelastic strain or material properties. Differential material properties often leads also an inelastic strain but it is not compulsory to have it.
I should highlight that I intentionally used "inelastic" instead of "plastic", which is usually associated with classical plasticity (yielding) of the material. On the other hand, there are many other mechanism that a material can deform inelastically.
Usually we should talk about eigenstrains (Mura 1982) or stress free transformation strains (Eshelby 1957). One of the initial paper on the sources of residual stresses is from Reisner 1931. Each process, each load, each material may have its own nature of eigenstrain (plastic, thermal, phase change, misfit, ...). This eigenstrain is often incompatible and leads to accomodation by elastic strains. Sometimes the accomodation is more complex when the accomodation stress becomes higher that the yield stress. A counterexample of compatibility is a uniform plastic strain (standard tensile test) of a plastic strain that depends linearly of the position. See the expression of the compatibility equation. For more details, read the nice book from Korsunsky 2017. Hope it helps
Just two basic examples of non- plasticity induced residual stresses:
Roll a thick steel plate and joint the two opposite borders together, forming a pipe. Even if no plasticity occurs during this process, the inner surface undergoes compressive stresses whilst outer surface undergoes tensile stresses.
When an ingot of molten metal cools down, the outer surfaces start to solidify first. Because of thermal shrinkage during further cooling, the metal near the center of the ingot will undergo tensile stress during solidification. This possibly induce the so-called solidification cracks.
Similarly, when a polymer is cooled from a melt, residual stresses occur as a result of crystallization. This is especially true for volumetric bodies (not rods and plates). This is explained by different degrees of crystallinity achieved near the surface and inside the body volume.