Considerations when testing springs with force systems

Spring testing is one of the most difficult testing processes. A common misconception is because springs are viewed as relatively simplistic components in terms of design and function, they must be just as easy to test and verify. There are also many deleterious factors and improper techniques affecting the accuracy and methodology used when testing springs. This article covers critical considerations and basic force testing precautions when testing extension and compression springs.

Testing system load capacity

Using a load sensor suitable for the spring being tested is a major consideration in accurately testing springs for spring rate, spring constant, initial tension, and free length. Don’t attempt to test a spring with a load rating less than 20% of the load system’s load cell sensor. For example, if the force system has a 500N (110 lbf, 50kgf) load cell, that sensor is acceptable for testing between 100N (22 lbf, 10kgf) and 500N.

Load cell deflection

A load cell sensor must deflect in compressive or tensile directions for the sensor to provide a measured output. Additionally, the entire load string used in the test will have some deflection ‒ for example, the testing fixture, the test frame, and the crosshead. Deflection may need to be compensated for when testing springs. Some force testers have a deflection compensation feature within the corrections setting, which is very helpful.

Grip string alignment

The grip string is the combined load cell sensor and testing fixture (top and bottom) used to test the spring. Hooks are often used for testing extension springs. Platens and specialized testing fixtures ensuring spring containment during compression are used for compression springs.

It’s absolutely critical the spring being tested is perfectly aligned to the grip string. The spring should be in the center of the platens to ensure even loading.

Correct spring preparation

Accurate measurement of spring rate, spring constant, free length, and initial tension requires the spring is designed and manufactured correctly. Compression springs should have a perfectly flat surface at both ends (top and bottom). If not, length measurements are compromised, leading to inaccurate and inconsistent spring rates.

Extension springs, with initial tension, should have loops perfectly symmetrical to each other. Extension spring loops should have equal distance between the inside of the loop and the first coil. Extension spring loops should have a common seating location within the loop ‒ for example, when the spring is under load, the position where the loops align should be parallel with each other.

If there are inconsistencies in the spring’s physical makeup, measurements will be affected. If a manual system had previously been used, such as a force gage with manual test stand, it’ll be difficult to correlate the data. Force gages can’t compensate for physical inconsistencies of the spring, and don’t have any deflection compensation.

Correct testing fixtures

Compression springs can be tested using platens if the platens aren’t over-sized to the spring’s outside diameter. Accurate compression measurement suggests the platen shouldn’t be greater than 25% of the spring’s outside diameter. Compression springs should use a testing fixture that’ll contain the spring should the spring walk ‒ for example, move to the point where the spring can shoot out of the platen.

When testing extension springs, there’s always some movement of the spring as it’s pulled apart. The end loops will typically position themselves to a perpendicular location along the grip string. However, if the hooks used to secure the loops have a rough finish, or are too large for the spring, it’s common for the loops to exhibit frictional errors affecting accurate measurements of free length and initial tension. Hooks used for extension spring testing should have a low-resistance surface to minimize any frictional effects. Hooks shouldn’t be oversized compared to the spring loop opening and type.

Friction effects

Belleville washers are very difficult to test accurately. Test fixturing is critical since frictional forces caused by the spring flattening during compression and against the platen surface is very likely.

To eliminate frictional effects, use the proper type of platen, typically a smooth surface with a hardness of HRC60 or better. Consider using some type of lubricant such as fine machine oil on the surface of the platen to counter frictional loading.

A compression spring will unwind slightly during compression, so a small degree of spring movement can be expected. Use hooks with a smooth surface. Applying a small amount of grease on the hooks will help to minimize frictional effects.

Correlating data

When changing or upgrading from one force testing system to another, it’s important to correlate the data. Examine existing test methods and system components including the type of load sensor and testing fixture. Make sure the load resolution is identical to the previous system. If the load resolution is different, different results can be expected for spring rates, free length, and initial tension.

If the load resolution is different, different results can be expected for spring rates, free length, and initial tension.

Testing speed

Testing speed is generally subjective and dependent on the application. In a production environment, testing speed will be greater for better throughput. In a lab environment, testing speed will tend to be lower. It’s always best to test at the lowest speed possible. Manual testing of a spring using a force gage and manual tester isn’t recommended since velocity is unknown and inconsistent.

Hysteresis effects

There’s some hysteresis when checking a compression spring not fixtured over a pin. The coils wind up when compressed, and the end coils twist and move on the platen. Platens can be put on bearings to eliminate this effect for extremely precise springs.

A solution is measuring the load while approaching the load target and then measuring it again after going past the load target (10° or 15°) and approaching the load target while coming back down in load. Take the average of these two readings. This averages out the hysteresis. Make sure to check the amount of over travel doesn’t cause a set of the spring.

The L.S. Starrett Co.
https://www.starrettmetrology.com