An extension spring is the only elastic element in a typical mechanism that carries no built-in end stop. A compression spring runs out of travel at solid length and protects itself. An extension spring, pulled further than the designer intended, simply moves into permanent deformation and never returns to its original dimension. Everything else in this article follows from that one difference.
Not a compression spring turned inside out
Visually the difference comes down to coils lying against each other and loops formed at both ends. Structurally there are more differences, and each of them changes the way an enquiry has to be written.
Overload protection has to come from the mechanism, not from the spring. A hard stop, a slot of limited length or a lever geometry that cannot travel further – something in the assembly has to define the maximum extension, because the part itself will not.
Calculation and design are covered by EN 13906-2, which applies to cold and hot coiled extension springs made of round wire and bar, loaded along the axis and working at normal ambient temperature.
Initial tension: force that exists at zero deflection
A cold coiled extension spring can be wound with its coils pressed against one another. That produces initial tension: before the coils separate at all, an initial force F0 has to be applied. The characteristic therefore does not start at zero, and only above F0 does it rise linearly.
Sometimes this helps and sometimes it causes trouble. It helps where an assembly has to stay closed without play, in a flap, a lever or a latch mechanism. It causes trouble when the engineer sizes the spring from the rate equation alone, forgets to add F0, and finds the measured force on the first prototype higher than the calculation promised.
How much initial tension is achievable depends on the spring index. Tight windings, where D/d is small, allow high values. Above an index of roughly 12 initial tension effectively disappears and the spring behaves as if coiled with a gap. Hot coiled springs have none at all.
The hooks decide how long the part lasts
The body of the spring works in torsion of the wire. The hook works in bending, over a small radius. Stress concentrates in that transition, so that is where the crack starts, no matter how conservatively the active coils were calculated.
The common solutions are a full loop on the spring axis, a side loop, an extended hook and a raised hook with an increased bend radius. Versions with a larger radius carry noticeably higher dynamic loads, at the price of a longer wire length and a higher unit cost.
A working rule for design: the inner bend radius of the hook should be at least about 1.5 to 2 wire diameters. A sharp bend is cheaper to produce and takes less space, but in a dynamic application it cuts service life by an order of magnitude.
The relative position of the two hooks is a separate matter. The angle between the loop planes needs to be stated explicitly, together with its tolerance. Without that information the parts will be made with both hooks in one plane, which is not always what the mechanism needs.
How to write the characteristic into an enquiry
Four quantities matter: the free length measured inside the hooks (LK), the initial force F0, the rate R and the force F1 at a defined working length L1. In most cases LK plus two forces at two lengths is enough for a manufacturer to work with.
State which of them governs. Tolerancing free length, initial tension and rate at the same time, within standard accuracy classes, is usually self-contradictory. The manufacturer will have to decide what to hold – better that the decision comes from the design side.
Materials: same grades, sharper consequences
The material range overlaps with compression springs, but the consequences of the choice are harsher, because the hook works in bending.
| Material | Standard | G [N/mm²] | Temperature range | Typical use |
|---|---|---|---|---|
| Patented wire SH / DH | EN 10270-1 | 81 500 | -200 to +80 °C | dry mechanisms, internal hardware |
| 1.4310 | EN 10270-3 | 70 000 | -200 to +160 °C | humidity, washing, outdoor use |
| 1.4401 | EN 10270-3 | 68 000 | -200 to +300 °C | chemical environments, higher temperatures |
| 1.4568 | EN 10270-3 | 73 000 | -200 to +200 °C | high loads combined with corrosion resistance |
For parts working outside the building envelope, in windows, gates and architectural hardware, a stainless grade is often the only sensible answer. A coating on an extension spring is harder to keep intact than on a compression spring, because the coils touch each other and abrade the layer in the same place on every cycle.
Tolerances: DIN 2097 where no EN equivalent exists
Quality requirements for cold coiled extension springs are set out in DIN 2097. The standard treats force as the governing parameter, which means the coil count may vary within a range and is not quoted as a fixed value.
That is the opposite logic to EN 15800 for compression springs, where the accuracy grades cover dimensions and forces in parallel. The difference matters when specifications are copied between documents – carrying „grade 2” across from a compression spring drawing to an extension spring drawing means nothing technically.
The measuring method belongs in the documentation as well. Free length can be measured inside the hooks, over the hooks, or across the body without the loops: three different numbers for the same part. Missing that line is one of the more frequent causes of dispute at first article inspection.
Fatigue: why shot peening does not rescue this part
In a compression spring, fatigue life is raised by shot peening, because the media reaches the wire surface along the whole length of the body. In an extension spring the coils lie against each other, so the treatment effectively reaches only the outer surface of the body and the hooks.
The result is a lower fatigue margin than a compression spring at comparable stress levels. For high cycle applications the part is therefore designed at visibly lower working stresses, or replaced by a different solution altogether – a compression spring acting through a lever.
Active coils, body length and the prototype surprise
In a compression spring the end coils are clearly separated from the active ones. In an extension spring that boundary is blurred: part of the wire is consumed by forming the hook, and that length works in bending rather than torsion. The more elaborate the hook, the less wire remains for the body within the same installation length.
This is where the classic gap between calculation and first article comes from. The engineer computes the rate for an assumed number of coils, the manufacturer has to fit hooks of a given shape into the same space, and the real active coil count comes out lower. The rate rises, and the force at the specified working length is higher than intended.
For a first project it is therefore worth ordering a short sample run with a measured characteristic instead of releasing a production batch straight away. Correcting the coil count after measurement is cheap. Correcting it after ten thousand units have been assembled is not.
How the spring is mounted changes the load distribution
An extension spring hung on an undersized pin concentrates the contact pressure on a narrow strip of the hook bend. A pin close to the inside dimension of the loop spreads the load over a longer arc and measurably extends service life.
Misalignment works the same way. If the spring axis does not coincide with the direction of the applied force, additional bending appears in the hook, in a plane no standard calculation accounts for. In lever mechanisms, where the angle changes through the cycle, this is worth checking in the extreme positions rather than at rest.
Where these parts actually work
In construction products and window hardware, extension springs handle self-closing, weight compensation and tension in locking mechanisms. The conditions are rarely gentle: cycles counted in hundreds of thousands, operation below zero, condensation on the surface.
We manufacture them as part of our technical spring manufacturing, including versions with non standard hooks bent to size. A large share of that production goes to the construction industry – hardware, gate systems, fasteners and mechanisms exposed to outdoor conditions.
Before the drawing is released
Check three things: whether the angle between the hooks is stated, whether initial tension has been included in the force balance, and whether the mechanism has a mechanical limit on maximum extension. Those three lines account for most of the failed parts that reach us from finished assemblies.
Have a mechanism where the extension spring does not reach the required cycle count? Send us your enquiry – we will review the hook design and match the material to the working conditions.
