RE: Springs and all that (long)

[M.Battley@xxxxxxxxxx]

> > BUT, the middle eight will not be the same length as they used to be:
> 
> huh?  what law of physics causes them to get longer?
> 
> > some of the ends of the eight will be inactive, so the eight coils
> 
> inactive?  why?  the only section of a spring that would be 'inactive' would
> be one that is perpendicular to the center line of the spring.  unless i
> flatten an end of my cut eight coil spring, the entire length of the spring
> will contribute to the spring rate.

Not so. Think about how a coil spring works: the "wire" is in 
torsion, just as in a 116 torsion bar. However unlike a torsion bar 
the ends are not clamped to apply the torque. So there is a region at 
each end of the spring where the torque on the wire (and the 
resulting shear stress and strain) increases from zero to whatever 
level it is in the central uniform region. Part of the last coil or 
so, most of which is in contact with the spring seat, acts as a 
lever to generate the torque. The overall deformation (and hence the 
rate) depends on the sum of the shear strains along the length of the 
spring. 

When calculating the spring rate based on the spring dimensions and 
material properties the main unknown is the effective "active" length 
of the spring: this has to be estimated. Hence the, perhaps slightly 
misleading, term "inactive" for the ends: it's common terminology 
in the spring design texts that I've looked at. The end regions are 
not completely inactive, but just do not deform as much  
as the central region. In practice the length of this region will 
depend on the properties (wire diameter and material modulus) of the 
spring, the angle of the end helix, and the shape of the spring seat. 
In the case of the 105 front springs that I have tested it appears 
(based on the rate) to be only about half a coil. It would be 
possible to determine it more accurately by using a series of strain 
gauges to measure the shear stress field at the ends: or by some 
geometrically nonlinear finite element analysis (which would need to 
include contact modeling at the spring seats).

> > will be slightly longer than they were. Yes, the same applies to the
> > original 10 coil spring, however this had the last coil flattened
> > off at each end (which was the coil that was removed). The length
> > (height) of the inactive region will be greater for the 8 coil spring
> > than that of the 10.
> 
> again, why?  this seems counter-intuitive to me.
 
Because of the greater helix angle of the ends of the 8 coil spring. 
Whether this is true or not will depend on the length of the inactive 
region relative to how much of the spring is in contact with the 
seat.

> however, there's also the chance that the cut spring wasn't seated the same
> way the uncut one was seated.  you could change the ride height of a car by
> rotating the spring such that it's not seated properly.

That is true, and if suspension arm have been removed and replaced you 
could also end up with slightly different "preloads" on the 
suspension from bushes. And you need to drive the car for a while to 
let everything settle down to its final position.

Yes there are other possible explanations. All I'm saying is that I 
don't think that it is necessarily impossible that the spring could 
be the cause: for springs of particular geometry. It certainly is 
counter intuitive though!

Mark Battley
Auckland, New Zealand

1974 2000 GT Veloce

------------------------------

End of alfa-digest V7 #337
**************************