[ananthkamath]

Guys, please dont fixate on the math, counting the number of active coils in a spring is pretty spotty anyway. I was only throwing a worst case number out there to show how dramatic the effect can potentially be.

And also, just because this works on a 2006 Civic does not mean it will work the same way on some other car. There are non-linearities involved in any suspension (I don't want to get into that here) so the effect on handling is pretty hard to predict just by looking. All I've said in my post is, be cautious and get a feel of how the car drives on different surfaces, loading conditions and speed, before you go out and take the car to its limit.

If you think you're going to run out of talent before the chassis does (which is guaranteed for 99% of the population), then don't do anything stupid. That's all.

[SS-Traveller]

Originally Posted by GTO

Report after 180 kms on the highway with 5 onboard + a trunk full of luggage...

Without having checked the alignment?

Originally Posted by ananthkamath

Adding this as a separate post for emphasis so that it doesn't get lost in the text of the previous one.
...the effect on handling will be much less dramatic if one were to install this at both the front and the rear. It may however make the front stiffer, but at least the car will be safe to drive.

@Ananth: A very educative previous post, and an interesting observation here - but in that case (if installed at both front and rear), wouldn't making the front end stiffer / raising ride height alter steering geometry & characteristics much more than if the front end remained OE (and hence keeping steering and handling more akin to original settings)?

Also, the reason that the manufacturer recommends an alignment is because the ride height of the car may be affected slightly when this part is installed, so the alignment also changes along with that and it has to be basically corrected to where it was.

IMO, the primary change in a car with independent suspension at all 4 corners would be a change in camber - which, in most production cars, is not an easy parameter to adjust or alter anyway. What's your opinion?

[ananthkamath]

Originally Posted by SS-Traveller

@Ananth: A very educative previous post, and an interesting observation here - but in that case (if installed at both front and rear), wouldn't making the front end stiffer / raising ride height alter steering geometry & characteristics much more than if the front end remained OE (and hence keeping steering and handling more akin to original settings)?

IMO, the primary change in a car with independent suspension at all 4 corners would be a change in camber - which, in most production cars, is not an easy parameter to adjust or alter anyway. What's your opinion?

Hi,

You may be right about the steering geometry being affected, but it is true only for drastic changes in ride height, half or three-quarters of an inch or so does not really make a huge difference. Even so this is likely to have a minor effect on the safety aspect of handling which is what I'm worried about. The biggest change handling-wise will be that the overall roll stiffness will get higher so the car will definitely roll less especially when unloaded. This has secondary effects but lets not get into that.

About your second query,both camber and toe will change when ride height changes. From experience I would say that camber has a much less effect on tire wear than does toe, for the same % change. Nearly every car I've worked on has had toe adjustment (except for torsion beam rear axles), so I don't see that as much of an issue.

[ananthkamath]

Originally Posted by Sutripta

Hi,
What would be the load - compression graph with the polymer wedge?

Regards
Sutripta

Hard to say what it is just from looking at a picture. The donut undoubtedly has a rising rate, and the coil spring is more or less linear, so the combination of the two could be anything from a falling rate to a rising rate. Maybe someone needs to go test these on a spring tester or UTS if they can find it.

[Addy]

Originally Posted by kaushik_s

Just want to know if this can be used in Baleno? For rear suspension. I'm satisfied with the handling of the Baleno but it bottoms out with passangers in the rear and this might be a good way to avoid that. Is it advisable on a Baleno?

Can some one please answer Kaushik's question. Considering I have a Baleno too, I am eager to know. Thanks.

[sanagg1]

Originally Posted by ananthkamath

Hard to say what it is just from looking at a picture. The donut undoubtedly has a rising rate, and the coil spring is more or less linear, so the combination of the two could be anything from a falling rate to a rising rate. Maybe someone needs to go test these on a spring tester or UTS if they can find it.

How come the combination of a rising rate and linear rate could be falling rate ?

Cheers

[ananthkamath]

Because the coil spring rate reduces as the number of active coils increases due to the donut compressing under load. However the act of compressing the donut itself is a rising rate so all of this is conjecture unless someone goes and tests the actual assembly. It is hard to do this analytically.

[AbhiJ]

Very nice and simple product...

Can be copied fairly easily and duplicated MUCH cheaper (about Rs. 500/pc) w/o compromise in quality

Small tweaks in the Hardness of the Rubber will really change the way this product works..

Harder the Rubber = Harder the suspension = Less deformation under stress.

[Sutripta]

Originally Posted by ananthkamath

Hard to say what it is just from looking at a picture. The donut undoubtedly has a rising rate, and the coil spring is more or less linear, so the combination of the two could be anything from a falling rate to a rising rate. Maybe someone needs to go test these on a spring tester or UTS if they can find it.

I think one can make a fair guess as to its nature. Of course actual values are another matter altogether.

The UTM which I have (easy) access to is unfortunately a 100t one, with lowest scale at 10t.

Couldn't agree more that one must have hard data, numbers before doing any serious modifications. People here are learning this now about modifying engines and dyno runs. Hopefully, the same philosophy will spread to all fields.

Originally Posted by sanagg1

How come the combination of a rising rate and linear rate could be falling rate ?

Cheers

Originally Posted by ananthkamath

Because the coil spring rate reduces as the number of active coils increases due to the donut compressing under load. However the act of compressing the donut itself is a rising rate so all of this is conjecture unless someone goes and tests the actual assembly. It is hard to do this analytically.

I don't think it will be falling rate anywhere in this example. But as you said, needs confirmatory testing.

Even the most commonly used falling rate springs (plate spring/ belleville spring/ dome etc) will not be falling rate until and unless a lot of design work has gone into it.

Aren't some bump/ jounce stops designed with falling rate characteristics?

Regards
Sutripta

[ananthkamath]

Originally Posted by Sutripta

Aren't some bump/ jounce stops designed with falling rate characteristics?

None that I know of. Why would you want a falling rate for a suspension springing medium? I meant to say in the previous post that falling rate is undesirable, and this spring system may in fact have a falling rate. I dont have the time right now to do the math as an example, but its easy enough to do given a few simple assumptions.

[vina]

Hi

For the academically inclined here are some notes (I'm sure GTO will post the actual experience after his long drive)

some notes on the spring (based simply on physics from 1st year engineering - there is nothing so complex in spring calculations that this will change substantially for most real springs) -

1. Any coil spring with N turns is roughly equivalent to N springs with 1 turn each (or N/2 springs with 2 coils each).
2. Assuming spring constant (or spring rate) of a spring is k, the spring rate of N such springs is series will be k/N and N such spring in parallel will be k*N. Since a coil spring acts like multiple coil springs of one turn in series, reducing the number of turns will increase stiffness (or effective spring constant) and increasing the number of turns will decrease stiffness.
3. In GTO's setups assuming there are 5 turns to begin with what the change does is :
   • As a first step (effectively) separates the 5 turn spring into a 4+1 turn spring at first, before insertion of the rubber (this doesn't mean anything in practice - both are same. this is just for the maths and modeling)
   • As a second step it puts the rubber as another spring in parallel to the "1" turn in the 4+1 turn (Ananth is wrong in stating that it is equivalent to 2 turns taken out. Two explain it -> k = -dF/dx where x is the displacement of spring and F is the force generated. The rubber is preventing displacement of exacatly one turn).
     • To be fair, 5 coil turns doesn't mean it'll act as a 5-turn spring, you really need to look at the effective play of the spring originally possible and final gap possible. Most likely GTO's car has 4.5 turns effective if he can see 5 turns.
     • All calculations here assume constant k throughout spring's operation.
   • If each original spring turn had a spring constant of k1 and the rubber has effective spring constant of k2 then for five turns the original spring constant was k1/5 and final will be 1/[(4/k1) + {1/(k1+k2)}]
     • assuming k2>>K1 this is about k1/4
     • For all values of k2 this is more than k1/5 which means the overall spring will always be stiffer - no exceptions
   • So if the rubber acts much stiffer than 1 turn (say 10 time or more) then the spring becomes stiffer by roughly 25%.
   • If the rubber is as stiff as roughly one turn of coil spring then the spring becomes roughly 10% stiffer.
   • By the way - while steel's stiffness remains constant under a wide range of practical load values, this may not be so with polymers. It is possible that the k2 increases with increasing load thus causing the overall spring to get stiffer as the load increases. However if k2>>k1 to begin with, then this wouldn't matter.
4. If all we want to know is how much stiffer the spring becomes (in percentage - not actual spring constant in N/m units) a simple ruler can measure that - GTO (or anyone else with this installation) can take the ruler (vernier calipers will be better - but a ruler from your kids pencil box will give good enough results) to the spring and measure
   • With empty car (no load or car over jack) the measure gap between the turns with rubber between them and the turns with air between them.
   • With loaded car (say 5-people, + 3 sitting on the open boot) measure the above values.
   • The displacements (i.e. decrease in the gaps) for the rubber turn and for the non-rubber turn.
   • The ratio of the two displacements (smaller to larger; it will be R=k1/(k1+k2) where k1 and k2 have meanings described above. For rubber much stiffer than steel (i.e. K2>>k1) this will be very small (practically 0). For K1=k2 (i.e. rubber about as stiff as steel, R=0.5 should be observed.
   • Once the above are performed, you can tell the overall spring would have become stiffer by a factor of N/(N-1+R) where N is the effective number of turns.
   • For 5 turns, and assuming rubber is much stiffer than steel, this is about 1.25 or 25% stiffer than the original. For rubber as stiff as one turn of the coil it'll be 1.11 (roughly 10%)
5. Longer rubber than one full turn will make spring stiffer by two effects:
   1. It'll reduce number of "free turns" (i.e. air-gap turns) from 4 in the above example (or N-1 overall).
   2. All of the "effective rubber spring" is in parallel to itself (unlike the coil spring) - this means k2 keeps increasing when you use more rubber.
      1. If k2>>k1 to begin with this doesn't matter much
6. when hitting a bump etc. (i.e. when you expect to hear a thud) the spring constant just at the time of thud will be exactly k2 - all of the coil is fully pushed and only rubber is acting. I think this is the 5mm the manufacturer is claiming. The prevention of thud probably comes from the fact that the thud sound most likely doesn't come from the coil itself but from other locations (shock absorber, jounce bumpers) and those other locations can not have contacting surfaces any ore because even fully squeezed rubber will add 5mm spacing (basically this rubber is acting as a jounce bumper too)

As far as increase in loading over the spring is concerned (people are worried about cracks in the spring) keep in mind that even in the original spring, the 4 turns that now have air gap had exactly the same load - there is no change or circumstances as far as those 4 turns are concerned (more on that below) and the one turn that has rubber in between now has extra support i.e. less loading.

The problem can arise at the ends of the rubber insertion, there the shear forces on the steel in the coil will most likely be more than the spring was designed for (look at how the coil spring end are made - almost parallel to the base on which the spring sits. This is not how the sheer forces act at the ends where the rubber is stiff). Thus steel at both ends of the rubber insertion is susceptible. This can be reduced to one point susceptibility by keeping the rubber either one top end of the spring or the bottom.


This is as far as simple maths goes. I don't know about the driving and handling characteristics (best found by driving anyway) and I think I need to sleep now

[dot]

Very nice analysis.

Originally Posted by vina

By the way - while steel's stiffness remains constant under a wide range of practical load values, this may not be so with polymers. It is possible that the k2 increases with increasing load thus causing the overall spring to get stiffer as the load increases. However if k2>>k1 to begin with, then this wouldn't matter.

Rubber is viscoelastic. So you are right, the stress strain response will have a lag. Thats why two modulus, elastic/storage modulus and loss modulus, are used when interpreting mechanical properties of a rubber or thermoplastic. Specially when vibratory forces are acting on the part.

But as you said if k2 >>k1, this may not be important.

However when the coil is fully compressed k2 comes in picture, again just as you said. In that scenario, the rubber will act as a dampner, I feel.

Originally Posted by vina

I think I need to sleep now

Writing a technical post at 4:40 a.m. Amazing.

@Sutripta, I have access to a UTM with a 50kN load cell, will this do?

[vina]

thanks dot for the past few days I have had insomnia, did something useful last night with it.

From the pictures GTO has posted (especially the one in which sone guy is cutting the rubber) it seems the rubber can be cut relatively easily with a simple knife. The cross section of the cut doesn't show anything special - the material should be fairly isotropic.

For an isotropic material that can rival one turn of a car's suspension spring in stiffness to be able to cut like that is something hard to swallow.

GTO: can you let us know what is the thichness of the rubber after you took the car off the jack? Somehow I feel it should be way less than 1 inch (or it couldn't be cut in a manner seen here).


By the way, I remember someone had used Air Lift balloons in the rear suspension earlier. I think GTO had replied on that too (I'm not sure and can't find the thread now)

If someone still remembers can you comment on which method can potentially give better results?

Apparently the ballon in question is also a very easy installation and completely reversible.


Amazon.com: AIR LIFT 60769 1000 Series Rear Air Spring Kit: Automotive is the link of (a similar) product.

at airliftcompany.com you can find full catalog and technical information on how air-spring works ( basically PV=nRT equation with heat transfer gives pretty good progressive rate spring and damping).

Can someone

[dhanushs]

Nice post.

Originally Posted by vina

If the rubber is as stiff as roughly one turn of coil spring then the spring becomes roughly 10% stiffer.

Umm.. could you please clarify?. If the rubber is as stiff as one turn of the spring, there is no change in stiffness theoretically. (?)

Originally Posted by vina

If all we want to know is how much stiffer the spring becomes

Again, I'm a bit confused by your post.

Assuming Rubber is much.. much stiffer than the spring.

For measuring the approx change in stiffness. K is proportional to M/X . M is mass and 'X' is deflection upon mass. ie, (pitch upon no load - pitch upon load).

So we have to find 'X' with no rubber and 'X' with rubber. The ratio will give you the change in stiffness, right?.

Well, this simple test might really help us predict the handling or body roll characteristics of the car after installation.

Originally Posted by vina

when hitting a bump etc.

The thud is the sound of suspension bottoming out, or belly hitting the ground.

The Rubber reduces the suspension travel, and hence increases the GC on load, and hence there is no bottoming out, or scraping the belly.


P.S - Please feel free to correct me as and when required!.

[vina]

Originally Posted by dhanushs

Nice post.



Umm.. could you please clarify?. If the rubber is as stiff as one turn of the spring, there is no change in stiffness theoretically. (?)



Again, I'm a bit confused by your post.

Assuming Rubber is much.. much stiffer than the spring.

For measuring the approx change in stiffness. K is proportional to M/X . M is mass and 'X' is deflection upon mass. ie, (pitch upon no load - pitch upon load).

So we have to find 'X' with no rubber and 'X' with rubber. The ratio will give you the change in stiffness, right?.

Well, this simple test might really help us predict the handling or body roll characteristics of the car after installation.



The thud is the sound of suspension bottoming out, or belly hitting the ground.

The Rubber reduces the suspension travel, and hence increases the GC on load, and hence there is no bottoming out, or scraping the belly.


P.S - Please feel free to correct me as and when required!.

The comment I wrote about rubber being as stiff as one turn of the coil meant that in that case the overall spring will be stiffer by 10%.


You are right about M/X, however to find the ratio of stiffness with and without the rubber you'll need to find X for the overall spring with and without the rubber. My method saves the hassle of removing and reinstalling the rubber by estimating both from the same measurement setup.


On your comment on the thud sound, I fully agree - suspension can't bottom out fully now (at least 5mm extra spacing) - that's what I wrote; Underbelly hitting something on the road is unlikely by (probably) 0.5 inch to 1inch (depending on how much the rubber is squeezed under the conditions).