Why a metal that springs back perfectly can still be permanently bent a moment later.
Push lightly on a steel ruler and it flexes, then snaps right back the instant you let go — as if nothing happened. Push harder, past some threshold, and it stays bent. Same material, same action, two completely different outcomes. The dividing line between those outcomes isn't vague or gradual — it's a specific stress value called the yield strength, and which side of it you're on determines whether the deformation is temporary or permanent. Most real forming operations, though, don't cleanly land on one side or the other — they involve both at once, and missing that is a genuinely common, genuinely costly mistake in metal-forming design.
Below the yield strength, an applied stress simply stretches the bonds between atoms slightly out of their equilibrium spacing — it doesn't break them or force any atom to a genuinely new position in the lattice. In this region, deformation is directly proportional to the applied stress: Hooke's law, stress = modulus × strain (σ = Eε). Because the crystal structure itself is untouched, removing the load lets every stretched bond snap back to its original spacing, and the part returns to its exact original shape. Nothing about the material has permanently changed — it's stored energy being released, not damage being undone.
Once applied stress exceeds the yield strength— the boundary between the elastic and plastic regions — the mechanism changes entirely. In metals, dislocations (line defects in the crystal lattice) begin gliding through the structure, permanently shifting planes of atoms relative to one another. That rearrangement doesn't undo itself when the load is removed, because unlike elastic stretching, it isn't stored energy — it's a structural change to where the atoms sit. Some elastic "spring-back" still happens even after plastic deformation: whatever elastic strain was present at the exact moment of unloading relaxes back, the same way it always does. But the plastic component underneath it — the part where the lattice actually rearranged — stays. That's the permanent shape change.
Bending a piece of sheet metal past its yield point — as almost every real forming operation does — doesn't produce pure plastic deformation. At the moment the forming load is at its maximum, the total deformation is a combination of an elastic component (which will spring back the instant the load is released) and a plastic component (which stays permanently). Release the load, and the elastic portion recovers immediately — a phenomenon the metal-forming trade calls springback — while the plastic portion remains as the actual final part geometry. Because of this, a bracket that needs to end up at a 90° bend has to be deliberately over-bent beyond 90° while under load, specifically to compensate for the elastic recovery that will occur the moment the forming tool releases it. Assuming the shape at maximum applied load is the final shape — and skipping that compensation — is a real, common, and costly metal-forming design mistake; the part comes off the press measurably under-bent relative to spec.
Once a material has been pushed past yield, releasing the load doesn't send it back down the same curved path it climbed — the plastic deformation already happened and isn't reversible. Instead, the unloading path is a straight line with the same slopeas the original elastic loading line (the same elastic modulus governs both), just starting from wherever the maximum-load point happened to land. That line runs down to zero stress at a strain value greater than zero: the permanent, plastic strain. The distance between the maximum strain reached under load and that final permanent strain is exactly the elastic strain that recovers — no more, no less. This is precisely why springback in metal forming is calculable rather than mysterious: it's governed by the same elastic modulus as everything else in the elastic region, which is why die designers can predict it and over-bend by a computed amount rather than guessing.
False, and it's the exact misunderstanding that produces out-of-spec parts. A real forming operation carried out beyond the yield point involves elastic and plastic deformation simultaneously — the total deformation at maximum load is a combination of the two, not one or the other. When the load is removed, the elastic portion genuinely does recover, following a path parallel to the original elastic loading slope rather than snapping instantly to zero strain. What remains afterward is only the plastic portion. "It didn't fully return to its starting shape" and "none of the deformation was elastic" are not the same statement— the elastic component is real, quantifiable, and springs back every single time; it's just riding on top of a plastic component that doesn't. Treating springback as negligible, or as evidence there was no elastic recovery at all, is exactly what causes formed parts to miss their target dimensions.
Explains the difference between elastic deformation (a temporary shape change that fully recovers once the load is removed, proportional to stress below the yield strength) and plastic deformation (a permanent shape change that begins once stress exceeds the yield strength, caused by dislocations permanently rearranging the crystal lattice) — and why real metal-forming operations produce both at once, requiring deliberate over-bending to compensate for elastic springback.
It's tempting to treat deformation as a binary: either a part springs all the way back (purely elastic) or it stays exactly where it was pushed to (purely plastic). Real forming operations beyond the yield point don't work that way — the total deformation at maximum load is a combination of an elastic component and a plastic component, and only the elastic component recovers when the load is removed. Assuming the shape under maximum load is the final shape ignores the elastic springback that is about to happen, which is precisely why formed parts can come out systematically under-bent relative to the target dimension if that recovery isn't compensated for.
Below the yield strength, stress and strain are related linearly by Hooke's law (σ = Eε): atomic bonds stretch slightly but the crystal lattice itself is untouched, so removing the load lets every bond return to its equilibrium spacing and the part returns to its original shape exactly. Above the yield strength, dislocations glide through the lattice and permanently shift atomic planes relative to each other — a structural change that doesn't reverse itself. When the load is removed after plastic deformation has occurred, the material still relaxes elastically by whatever elastic strain was present at the moment of unloading, following a path parallel to the original elastic loading slope (governed by the same elastic modulus) — but it lands at a nonzero permanent strain rather than returning to zero, and that permanent strain is the plastic deformation.
This is the physical basis of springback in sheet-metal forming, wire bending, and any cold-forming operation taken past yield. Because the elastic recovery is governed by the same elastic modulus throughout the process, it is calculable — die and tooling designers routinely compute the expected springback angle and deliberately over-bend the part while under load by that amount, so that once the tool releases and the elastic component recovers, the part lands at the actual target dimension. Skipping this compensation is one of the most common root causes of formed parts coming off a press or brake measurably out of spec.
The yield strength is the stress level that marks the boundary between the elastic and plastic regions of a material's stress-strain curve. Below it, deformation is proportional to stress and fully recovers when the load is removed. Above it, the material begins to deform plastically — permanently — because dislocations start gliding through the crystal lattice and rearranging it.
No. Even when a material is deformed well past its yield strength, it still retains whatever elastic strain was present the instant before unloading, and that portion still springs back when the load is released. The total deformation at maximum load is always a combination of an elastic part (which recovers) and a plastic part (which doesn't) — plastic deformation doesn't eliminate the elastic component, it just adds a permanent component on top of it.
Because the forming operation happens past the yield point, and the elastic portion of that deformation will spring back the moment the forming tool releases the part. Bending to exactly the target angle under load would result in a final part that springs back to an angle short of the target. Over-bending by the calculated springback amount compensates for that recovery so the final, unloaded part matches the target dimension.
No. Once a material has yielded, unloading does not retrace the curved plastic loading path. Instead it follows a straight line with the same slope as the original elastic loading line, starting from the maximum-load point and running down to zero stress at a nonzero permanent strain — because the same elastic modulus governs both loading and unloading, but the plastic deformation that already occurred isn't reversed.
The same elastic-recovery principle applies broadly to any material with a distinct elastic region and yield point, though the magnitude varies a great deal by material and stiffness. Metals are the most common context because so much manufacturing involves cold-forming metal past yield, but polymers and other engineering materials that are formed past their own yield points show comparable — often larger — springback.
Try our Materials Science & Metallurgy Studio
More calculators, simulators, and guides for this discipline.