Like mechanical fasteners or types of plastic, springs all appear alike from a distance, but up close their differences become apparent. And the reason why they are all different is because, although their purpose might be the same – and the scope of this article covers only compression – every application differs in its detailed requirements.
Basics first. A spring is a formed piece of metal that resists compression (plastic, composite and even ceramic versions do exist but aren’t covered here). Push it, and it pushes back. The stiffness of the force with which it pushes back is called the spring rate.
The most familiar and common spring design is a coiled round wire. Here is how it works. Under load, the loops of wire that make up the spring twist, and the distance between the coils reduces. Releasing the load results in the loops of wire twisting in the opposite direction, and the spring lengthens. Spring rate decreases in proportion with the number of coils.
It turns out that the forces are not exactly perpendicular to the long spring axis, because the pitch of the coil throws the angle of the force slightly offline, points out Simon Ward, technical manager at TFC, a distributor of Smalley springs.
For nearly 50 years, metal processing techniques have been sufficiently advanced to coil wire with a flat section. TFC calls this process edgewinding, and emphasises the strengthening effect the process has on the material, compared to production by stamping (a selection of products is pictured above).
Helically coiling flat wire offers a slight height advantage, as the profile is flatter. But the benefits multiply with further processing, in particular adding kinks or waves to the curved wire. Doing that creates a spring based on bending functions; when curled, the waves resist compression in the same way that the coil does.
The first consequence of this design and construction is that wave springs produce the same force in about half the height of a coiled round wire spring. As Ward points out, this also means that designers can reduce the size of space that the spring fits in, without affecting performance. This is true both in terms of dimensions and also weight, which is of particular importance in aerospace applications. Reducing size and weight is an extremely useful feature when a driving trend of product design is miniaturisation. Indeed, there continues to be a demand for smaller and smaller springs, says Jimmy O’Shea, technical sales engineer at Rotor Clip. (Some of its range are shown below).
Catalogues list standard models’ spring rates along with other variables including wire thickness, diameter, estimated free height, number of coils, spring rate, wire material and design. It may be interesting to note that published wave spring rates are theoretical approximations based on the target operating range of a wave spring, according to O’Shea. During the first 20% of compression from its free height, a wave spring’s spring rate is lower than expected as it energises (although the exact amount can be unpredictable). At the other extreme, as it approaches total compression (solid), the spring rate exceeds the reported figure as it rises exponentially. He adds: “That is not something that you want to design to; you don’t want springs too close to solid. That shortens cycle life.”
Wave springs come in three main types, and each one operates in a different way. First are peak-to-trough springs, in which the undulations of successive coils are nested; they stack in sinusoidal phase with each other above and below. The second design is peak-to-peak springs (called Crest-to-Crest by Smalley) whose coils are out of phase with each other, creating a honeycomb appearance. Finally, single-turn wave springs can be made with a gap or overlap. Gap-type designs feature an undulating wire whose total length measures less than a complete rotation, while Overlap designs have additional length and without a gap.
Mechanically speaking, flat wire springs work differently to round wire springs. Because of their construction, there is no torsion as they compress; as a result, wave springs flatten axially in a direction perpendicular to the load. One design consequence of that is that wave springs should not be side-loaded, as this increases the risk that the spring will buckle.
Given that the turns of a round wire spring work independently, they may be more able to resist this. However, even so, O’Shea recommends that all wave springs, whatever the type, should be constrained either by the inner or outer diameter to prevent misalignment.
Peak-to-peak flat wire springs are particularly different to coiled wire springs. Increasing the number of coils reduces the spring rate. Adding a third turn to a two-coil spring reduces the rate by a third and adding a fourth to a three-coil spring reduces the rate by a quarter. This is because of the way the turns interact with each other, explains O’Shea.
He says: “When you have [wave spring] turns in parallel, when they are nested, all the turns deflect at the same time, so they are adding to the spring rate. They all deflect as if it is solid steel; it’s really bending a big bar. But when they are peak-to-peak, it’s almost as if each turn is acting on its own, and then there’s another spring acting on its own, and another. Each turn acts independently giving a load, and they all deflect at the same time.”
What all of this means is that wave springs can be used for lots of applications. Ward at TFC recommends Crest-to-Crest wave springs for medium-duty, medium-deflection loads. O’Shea says that they offer a little more travel in small space than coil springs, and are good for precise loads.
Peak-to-trough multi-coil wave springs do also exist; TFC only launched its Nested Wave Spring range a few years ago; Rotor Clip offers these types as specials only. They are intended for much heavier loads; in fact, given how little they deflect, Ward compares their performance to cone-shaped disc springs when stacked in parallel (they resemble a deformed washer). For example, they are used on subsea oil and gas valves to overcome high pressures on a seal.
Finally, the last type is single-turn wave springs, which offer compression in a very small distance. States Ward: “We have even done 40mm-diameter springs that are 1.2mm high, but still allow 0.5mm of movement.”
This can be very useful in assembly, Ward continues: “If you are building a product, you might stack up a number of tolerances from components in an assembly, and this takes them up. Also, in aerospace applications might involve thermal expansion. In extremes of hot and cold, metals move at different rates, and this is a compensator.”
Preloading bearings is one of the biggest applications for both suppliers’ single-turn wave springs. For example, TFC works with a number of bearing manufacturers.
Static vs dynamic
Designers assume by default that springs (either coiled or flat wire) will be installed in a static application; in the case of unvarying load, springs can last indefinitely. Dynamic applications, in which the load increases and decreases, causing the spring to compress and relax, require additional analysis, says the Rotor Clip technical sales engineer. Not all spring designs are suitable for dynamic or cyclic loading. “There have been bridges that fail because of dynamic loading; it’s no different for this piece of steel,” he says. Important factors include the two heights that the spring is oscillating, the expected lifetime and frequency of oscillating: “Are they vibrating fast, seven days a week, or once per day for the next 50 years?”
The issue, he says, is metal fatigue, and the ability of the material to resist it. O’Shea states that designers use a standard formula using the change in stresses and material stresses to indicate theoretically how many cycles it should last, ranging from under 30,000 cycles to over 1 million.
Other important stresses acting on a spring involve the environmental conditions that it faces; the harsher the conditions, the more exotic the material, explains Ward. While carbon spring steel is the basic material, and is suitable for working in oily or greasy environments such as a gearbox, 17/7 stainless steel is better for working in conditions exposed to moisture, and for working at temperatures up to 343°C.
Springs that operate at a higher temperature than the material is really designed for can start to creep and lose load, and in extreme cases they might take a set, points out O’Shea. Rotor Clip also switches to stainless steel for smaller springs, as the superior material quality contains fewer impurities that might affect performance and lifetime.
While, mechanically speaking, every spring has a failure mode, and that is most likely either corrosion or mechanical fatigue, correct specification, material choice and design should ensure that the spring will last for its entire duty, stresses Ward.
On that subject, O’Shea emphasises the importance of early engagement with the customer on component design. He says: “The earlier we get into the process, the more helpful we can be, because once everything is designed and there’s a box for the spring, the constraints can be really difficult to work with.” Knowing the application, load, and travel required can lead to a simple specification of the space of the pocket required. “Otherwise we have to go backward to fix the problem, which is more challenging and sometimes more expensive,” he adds.
Although Rotor Clip and TFC both sell springs made in the USA, one difference between them is the ratio of standard to custom parts supplied. For the former, it’s about two-thirds specials, one-third standard product.
For TFC, specials make up three-quarters of sales. Reflects Ward: “A significant proportion of the work we do is bespoke. And that would apply to coil springs as well. There are so many varying factors in a spring. Diameters could be from 3-4mm up to 1m. There are different materials, loads, and different spring rates that customers need, depending on the application. It is impossible to catalogue every single permutation.” However, he adds that standard products can offer a place to start, at least.
O’Shea also stresses the importance of testing in the final application arrangement to verify that the specification is correct. He says: “Testing in-application is the be-all and end-all to ensure that the spring withstands the use you are expecting. The spring might be rubbing on the sides. It could be doing things in the application that the bench test can’t simulate, or a theoretical calculation can’t pick up.”