Edited by Kenneth J. Korane
Mechanical equipment and related hardware routinely include springs in their designs. And in a perfect world, knowing the load and travel, an engineer can adjust the mating parts so that the design can use a stock spring. However, that is usually not the case, and springs are frequently an afterthought.
Often this is because springs are well-engineered and proven components. Springs operating within their design parameters will last a long time. And they come in thousands of different sizes and versions. Most common are compression and extension springs, made from various materials, with or without a finish.
Stock springs are often used for applications requiring less than 500 springs per year. Generally, however, it is not only better but more cost effective to contact spring manufacturers with specific requirements. From there, the manufacturer may recommend a new design and make a customized spring for little or no cost premium.
One cannot emphasize enough the importance of discussing design requirements with a spring manufacturer. Stock springs are great for prototyping, but their use in production often compromises other aspects of a design.
This article is meant to help size springs for prototyping, prior to seeking the expert advice of a spring manufacturer. Here are some important decisions engineers must make when choosing a spring.
Stock versus special
One way to compare stock and custom springs is to look at load tolerance. If a catalog spring has a theoretical spring rate with a tolerance, but is not designed for a particular load, the actual load at a given distance can vary three times as much as a spring specifically made to meet that load.
This means a stock item with specific dimensions and spring rate may not work in the field even though some samples perform well. For example, a spring specially made to produce a 10-lb load at 1.0 in. will hold the load ±1 lb. A stock spring, on the other hand will vary ±3 lb. So prototype testing might succeed with springs actually running at 7 to 8 lb or 12 to 13 lb, leading one to assume the 10-lb load was correct. In reality, the springs were only in the ball park. And if stock springs didn’t work, perhaps the 10-lb load was correct but the springs were too far off. Remember, catalogs are sales tools, not design tools, and are controlled by the sales department, not engineering.
If springs are used in applications where loads are not critical, then stock springs are acceptable. And remember, most stock springs are based on the original World War ll design standards, MIL-STD-29.
Knowing the spring’s operating environment helps determine material requirements. The most-common material is carbon-steel music wire, ASTM A228. Specifying the correct material keeps costs in line. For example, opting for a slightly higher modulus could increase material costs many times over. And as a spring’s maximum operating temperature increases, so does the material price. (The accompanying table compares the cost of various raw materials.)
Also note that compression and extension springs use shear modulus (G), whereas torsion, flat, and spiral springs use Young’s modulus (E). And tensile strength for all spring wires differs with wire diameter. The smaller the diameter, the greater the strength and unit load capability. Thicker wire makes a spring stiffer, but more coils and a larger outer diameter (OD) make it weaker.
Most stock items receive some type of finish. If you do not specify plating, the price is for a plain finish. And ordering a “plated” spring is not sufficient, as there are several different standard types of plating. Preplated wire is also available, but this always provides less protection than plating afterwards.
It’s important to know if the spring will experience temperature variations. High temperatures weaken all materials (except some expensive, specialty materials). High stresses make the problem worse. (Note that the maximum operating temperatures listed in the table are based on extremely low stress levels.)
Almost all stock items are made from music wire or Type 302 stainless steel. Music wire and most carbon steels are considered “room-temperature” materials, generally –40 to 212°F. Even then, the metal will still suffer some slight load loss as temperature rises, dependant on stress. At 250°F, music wire theoretically loses approximately 4.5 to 10% of its load capability at 80,000-psi stress. For a 0.062-in.-diameter wire at 110,000 psi, expect an 8 to 14% loss, assuming approximately 100hr at this stress level. Chrome silicon wire (alloy steel, generally not used for stock items) theoretically loses only 1 to 2% at 80,000 psi, and 1.5 to 5% at 110,000psi.
Type 302 stainless is not even tested at 250°F. However, it can be used at somewhat higher temperatures, provided one reduces stress levels. At 350°F, stainless steel gives approximately 1% loss at 80,000 psi, and 4% loss at 110,000psi. But the design should use only approximately 50% of the available stroke at 350°F. This can continue to 450°F with even shorter strokes.
This recommendation makes some assumptions about stock-spring design stresses that are not necessarily true in every case. Stock items are not designed for higher temperatures, so always consult a spring manufacturer to get the best possible result.
17-7 PH stainless steel performs better at higher temperatures but is more expensive and generally not used for stock items. At 600°F, 100 hr under load, and 80,000 psi, it experiences approximately 3.5% loss, and a 5.5% loss at 110,000 psi.
Nonferrous materials are generally not used to make stock items and temperature ratings vary widely. So-called superalloys are never used to make stock items. They can be heat treated for specific high-temperature behavior, depending on requirements.
Note that approximately 95% of the loss a spring will suffer occurs in the first hour. The Spring Manufacturers Institute (SMI) defines spring failure as a loss of 10% or more of its load.
Simple sizing methods
Here are a few guidelines when sizing a spring. In the best case, load requirements and mounting locations will be set beforehand. Knowing the preload height and required travel also helps.
Use Hooke’s Law:
F = -kx
to determine spring rate or spring constant. Here, F = force exerted on the spring, x = displacement, and k = spring rate, typically in lb/in. or N/m.
Theoretically, spring rate is linear, provided operation does not get too close to the free length or solid height. Compression springs are weakest in the first 15 to 20% of available stroke. This portion of the stroke provides preload tension and prevents shock loads. A spring reaches the strongest point as it approaches the solid height — when all coils are fully compressed and in contact. Not crowding the solid height is critical, and it is important not to travel beyond this point. Most stock compression springs take a set (in other words, get shorter and, therefore, weaker) even before reaching the solid height.
Accounting for free length and compression height leaves a 60 to 70% maximum operating range. In general, the shorter the actual stroke within this range, the longer the life. Finally, never let a spring act as its own stop. Limiting travel with a stop also increases spring life.
Hole or rod clearance
Another design consideration is hole and/or rod clearance. A common error when designing with springs is crowding the OD and ID too tightly with surrounding parts and expecting tight tolerances from a stock spring to suffice. Generally, use shaft and counterbore clearances of 0 .015 in. for spring ODs to 0.375 in. with larger clearances as the OD increases, to perhaps 0.031-in. clearance at 2.00-in. OD.
These are gross approximations that assume stock springs tend to be centered, and they do not account for design or manufacturing variations. For example, considering tolerances, the nominal diameter of a 0.240-in. spring approaches the limit for a 0.250-in. hole. More clearance is encouraged. Some companies include counterbore or rod size in their catalogs. There again, discuss requirements with the manufacturer. The space needed for the spring and its mating parts is predicated on the spring’s actual dimensions.
The bottom line: Gather as much information as possible about the intended application. Contact a manufacturer and explain your requirements and learn their capabilities. Remember, stock springs are intended for testing and prototyping, not for production. Always use custom springs unless you’re certain the requirements aren’t warranted.
Good spring companies will do the design for free, but remember them when it is time to buy parts. There are over 10,000 spring companies in North America, about 250 of them are capable of aiding you with a design. If you contact one who cannot, keep looking.
Here is an example spring-design project with a 1-lb load and 0.054-in. required travel. Spring rate is 1/0.054 = 19 lb/in. From one manufacturer’s catalog, a stock precision-compression spring with 18-lb/in. spring rate serves as a good reference. This gives free height, OD, wire diameter, and solid height as datums, letting one calculate the prototype spring size as follows:
From the catalog, free height = 0.500 in. and solid height = 0.145 in. The available stroke is therefore 0.355 in. Calculating 15% of available stroke as the minimum deflection for preload,
0.355 × 0.15 = 0.053 in.
This establishes the minimum deflection for first load (preload). It’s the weakest load point.
Subtracting this from the free height,
0.500 – 0.053 = 0.447 in.
gives the maximum installed height.
Next calculate the maximum usable stroke, at 85% of available stroke:
0.355 × 0.85 = 0.302 in.
Beyond this stroke the spring will not have repeatable loading.
Finally, subtract the maximum usable stroke from the free height to find the minimum operating height = 0.198 in. This exceeds the 0.145 in. solid height by about 27%.
Spring load at the preload distance (0.053 in.) is about 0.95 lb; 0.054 in. required travel has 0.97-lb load. This gives a total load of 1.92 lb at 0.107-in. stroke. Given the size, shape, and loading, one can approach a manufacturer about making a spring. The key is to avoid extremes near the solid height and full extension, and to use as little of the available stroke as possible.
Determining the spring size also establishes an approximate working area. From this, one can design internal shafts, counterbores, and other required features.
The results can be fine-tuned for more-precise loading by using a stronger spring or limiting movements with a stopping device. If the spring doesn’t fit into the allotted space, one option is to divide the load among several springs. Multiple springs can also provide more-even load distribution or manage higher loads. Once a spring has been selected, create a prototype to verify fit and function.
Good intentions gone bad
A major spring manufacturer relates this anecdote regarding custom versus stock springs. One of their customers, in a cost-saving move, decided to redesign all of their products to use only stock springs. They purchased some 600 different springs, all in small quantities of 500 per year or less, and felt 90% could be converted to stock items.
After 10 or so redesigns, spring costs for each fell from about $600 per year to $375. But costs increased approximately $3,000 to $5,000 per device in new tooling to change the mating parts to accommodate new springs. It took only a year for them to scrap the project.
It will almost never make sense to change mating parts to accommodate a stock spring. Use stock items purely for testing to see if they are in the ball park, not with the expectation of actually using them in real applications.
What’s the spring index?
Manufacturers sometimes refer to the “spring index,” a term not commonly used by design engineers. Spring index indicates the ease of manufacturing as a function of cost. Springs with smaller indexes, under approximately 3.5, will increase costs, and it is impossible to go below approximately a 2.5 index. Larger indexes are only constrained by the spring manufacturer’s equipment. If a specific size spring uses standard diameter wire and has a spring index of 6 to 10, there will not be a manufacturing or cost issue. Just remember, tightening tolerances on the diameter also increases costs.