Coil Springs

I have for most of my life, managed to avoid knowing much about springs and spring theory. I takes them out and I puts them in. 

My current quest to understand helical compression springs comes from trying to understand what would be accomplished by "cut 1/3 of the active coils off the spring and stretch the remainder out to the original length". The purpose of the spring butchery is ostensibly to reduce the amount of force required to activate the mechanism, however the tiny bit of spring theory I believe to have a passing understanding of, would stand in complete opposition to that course of action. Cutting the spring increases the required force to obtain a given deflection, no?

Isn't applying compression load to a helical spring analogous to applying downward force to the end of a bar with the opposite end fixed in space, such that shortening the bar results in an increase in the applied force required to obtain a given deflection?  So wouldn't cutting off a length of the spring increase the amount of force required? Precisely opposite of the intended result.

Reducing the number of coils on a given spring will increase the spring rate.

"stretching" the trimmed spring sufficiently so that it has the same free length as the untrimmed spring would require yielding the spring wire material.

Edited by Kelpiecross, 05 April 2017 - 04:55.

Hey Joe! Good to hear from you again. Thanks for chiming in!  I look forward to the email.

Reducing the number of coils on a given spring will increase the spring rate.

Because you're reducing the total mass of material in the spring?  And so we're on the same page: changing the number of coils in a given spring - IE reducing the total length of wire used in the spring will increase the rate. On the other hand, if you change the number of coils while maintaining the length of wire, say by altering the median diameter, the spring rate will stay the same. So in a most pedantic sense, it is not the number of coils, it is the length, yes? And with no packaging constraint, there are many combinations of wire diameter, median diameter and length to obtain the same spring rate.

In practical terms I think that would translate to your original statement, as beyond the theoretical, I'm not sure how you would change the number of coils in a given packaging scenario without changing the length as well. 

"stretching" the trimmed spring sufficiently so that it has the same free length as the untrimmed spring would require yielding the spring wire material.

Yes, however that's how a spring is made in the first place (yielding), so in this specific scenario where you're intentionally altering the length, you'd want it to maintain the new shape. As near as I have been able to discover, this would not be detrimental to the spring

I am not quite sure what you mean - but the action of a coil spring is one of twisting (not bending) - like a torsion bar but in helical form. So the shorter the total length of the material in the coils opened out - the stiffer it becomes.

Right. That is in essence what I was alluding to - that just like trying to bend a bar, the shorter the length (not height), the lower the deflection for a given force.  You'll have to elaborate on the twisting action in the coil though - that one is escaping me.

 

Because you're reducing the total mass of material in the spring?  And so we're on the same page: changing the number of coils in a given spring - IE reducing the total length of wire used in the spring will increase the rate. On the other hand, if you change the number of coils while maintaining the length of wire, say by altering the median diameter, the spring rate will stay the same. So in a most pedantic sense, it is not the number of coils, it is the length, yes? And with no packaging constraint, there are many combinations of wire diameter, median diameter and length to obtain the same spring rate.

 

In practical terms I think that would translate to your original statement, as beyond the theoretical, I'm not sure how you would change the number of coils in a given packaging scenario without changing the length as well. 

 

Yes, however that's how a spring is made in the first place (yielding), so in this specific scenario where you're intentionally altering the length, you'd want it to maintain the new shape. As near as I have been able to discover, this would not be detrimental to the spring

 

 

Right. That is in essence what I was alluding to - that just like trying to bend a bar, the shorter the length (not height), the lower the deflection for a given force.  You'll have to elaborate on the twisting action in the coil though - that one is escaping me.

It's not about reducing spring mass. If you reduced spring mass by reducing wire diameter, spring rate would drop. Cutting a coil off increases spring rate. Same mass reduction, different effect.

Here's a calculator to play with and help figure out what's going on. http://www.acxesspri...calculator.html

The wire of a coil spring deforms torsionally when loaded just like a torsion bar. The spring rate is (essentially) independent of free length in this case.

As Fatboy says, the stiffness of a coil spring will go up if you remove coils and don't change anything else. 

To be exact, the stiffness is inversely proportional to the number of coils.  So, if you start out with 12 coils, let's say, and you take away a third of them, so you then have 8 coils, the new stiffness will be 12/8 = 1.5 times the original stiffness.  There are some minor corrections for the end coils that can make the actual results slightly different but they aren't too important.

Stiffness is also proportional the fourth power of the wire diameter (ie, a BIG effect) and inversely proportional to the cube of the coil diameter (ie, a bigger coil is much softer than a smaller diameter coil).

Kelpiecross is correct--coil springs work by the compression causing the wire forming the coil to twist it about its axis.  It might be non-obvious because the spring is coiled up but that is what is happening.  Just like a long torsion bar that is coiled around itself.

Okay, it's all starting to gel, at least a little bit. The torsion thing is interesting - it is neither obvious nor intuitive. Length vs. stiffness is pretty clear. 

With regards to wire diameter - two springs of equal coil diameters and length, one of .5" diameter wire and one of .5625", the latter is in fact substantially stiffer, not marginally stiffer.

This learning is of course clouded by the "internet solution" (remove x # of coils and stretch the remaining portion) offered to the problem (reduce the amount of force required). Rather, it underpins the notion that I don't yet understand how the mechanism really works.

Thanks all for your input and sharing your knowledge.

Something that might help understand the theory. If you look at the original spring with a compressive force applied, it is easy to see that the same force is being applied to any segment of the spring eg the bottom half. The deflection however reduces in proportion to the length of the segment being considered eg the bottom half only deflects half as much as the full spring. Same force, half the deflection = twice the rate.

You can have one of two situations: 

1 - if the actuation force remains the same (eg, the dead weight of a car) then the deflection of the spring will go down by the ratio of the new higher spring rate to the original spring rate.

2 - if the deflection remains the same, then the actuation force will increase by the ratio of the new spring rate to the original spring rate.

Thicker Wire diameter will increase spring rate, cut a coil off of it and it will be stiffer again but more likely to break as well. There is several computations and you can get the same rate from two different looking springs but the makers know the 'safe' way to wind them so they do not break.

The old furphy that a sagged coil is softer is not true, it is just shorter and changes the ride height. And ofcourse corner weights.

They actually do it as part of the production process sometimes. It's called scragging the spring. The reason they need to do it is that the spring length at a specific load is defined and has a tolerance, if they've stuffed up the manufacturing they'll be outside tolerance. Whether a one-off event past the plastic limit is enough to affect the fatigue life I don't know. It's not always possible in compression, quite often we seem to end up running springs that go coilbound before they yield (a waste of steel, usually because of some packaging limits).

Edited by Greg Locock, 08 April 2017 - 21:34.

"Stretching" a spring would surely take it past its elastic limit, making it functionally useless.

Not exactly. Placing an axial extension force on a helical compression spring sufficient to cause it to assume a free length 1/3 greater than it had would obviously require yielding of the material. But it would not render the spring functionally useless. Helical compression springs are manufactured by plastic deformation of the spring wire. Spring winding is done using both hot and cold wire. Hot wound coil springs are usually heat treated after winding. High performance coil springs are also shot peened to improve fatigue performance, but stretching a coil spring beyond its yield limit would negate the fatigue benefits of residual compressive surface stress from shot peening. Very high performance compression springs that require consistent performance in service, like engine valve springs, are compressed to coil bind (pre-set) as the final step in manufacturing. This is to eliminate any compression set that might alter the spring's properties (such as installed force) in service.

Okay, it's all starting to gel, at least a little bit. The torsion thing is interesting - it is neither obvious nor intuitive. Length vs. stiffness is pretty clear. 

 

With regards to wire diameter - two springs of equal coil diameters and length, one of .5" diameter wire and one of .5625", the latter is in fact substantially stiffer, not marginally stiffer.

 

This learning is of course clouded by the "internet solution" (remove x # of coils and stretch the remaining portion) offered to the problem (reduce the amount of force required). Rather, it underpins the notion that I don't yet understand how the mechanism really works.

 

Thanks all for your input and sharing your knowledge.

The main geometric property of torsion bars as related to stiffness is the polar moment of inertia, and it's unit is meter^4. So any change in cross section area or shape will lead to major stiffness change:

http://www.tribology...ators/t14_7.htm

O how to 'see' torsion in a coil spring:

Edited by saudoso, 10 April 2017 - 10:11.

Thanks all for your inputs here. I know more now than when I started this thread which is always the goal.

Not exactly. Placing an axial extension force on a helical compression spring sufficient to cause it to assume a free length 1/3 greater than it had would obviously require yielding of the material. But it would not render the spring functionally useless. Helical compression springs are manufactured by plastic deformation of the spring wire. Spring winding is done using both hot and cold wire. Hot wound coil springs are usually heat treated after winding. High performance coil springs are also shot peened to improve fatigue performance, but stretching a coil spring beyond its yield limit would negate the fatigue benefits of residual compressive surface stress from shot peening. Very high performance compression springs that require consistent performance in service, like engine valve springs, are compressed to coil bind (pre-set) as the final step in manufacturing. This is to eliminate any compression set that might alter the spring's properties (such as installed force) in service.

I have never met a valve spring that does not 'flatten off' and 'go soft'

I have used premium brands that will lose height very early then stay the same.

I also wonder at some manufacturers cams, they recomend X rate spring and to get the expected performance you always have to use 10%stiffer springs.. Though they generallydo DROP to recomendations. For instance a Crane flat tappet cam will generally say  130lb on the seat and around 300 open. Go 10% harder with new springs and sometimes you will find the engine is stronger up high by a 10% increase. eg 6500-7200 hundred useabkle power and most manufacturers are the same. And ideally bed the cam in on SOFT springs be3fore you go to bigger ones. Many simply remove the inners. Beehive springs is a little harder.

And I am very critical on springs, if a spring coil binds at X height my max open height is always higher.

I have found this with coilover springs as well, they all sag a bit and you simply adjust the corner weights to suit. BUT from that decreased height the rate is still the same.

For most racers with fairly basic equipment you can check valve springs and softer coil over springs with a workshop press and bathroom scales. that with an accurate measuring instrument such. Then comprss 1 " and measure rate.

Most bathroom scales will require a piece of steelacross the bed to stop distortion,, especially glass one! Then use the retainer to find your answers. Quite simple generally for springs up to 300lb.. That means you can check open pressures on most valve springs as well.

It would give you a longer spring travel.

...and a higher rate. And it may take a set at high loads.