A Novel (or Naive?) Method of Coil Spring Rate Selection 3

[ErikPSS]

TLDR: Can I find the softest spring by: height remaining after preload, at full shock compression = spring solid height? 16" extended shock with 12" 150lb/in spring under 715lbm static load compresses 4.75" to 11"; add 3.6" preload to get to ride height of 14.7". At full compression, there's about 7.5" available spring height on the shock. 7.5" available minus 3.6" preload = 3.9" for spring at full compression, very close to 3.5" solid height = good utilization?


I'm coming from off-road trucks, where maximum travel and soft springs are required.

I'm now working on a reverse trike, and solved for optimal shock travel by unladen ground clearance to ensure easy servicing -- my trolley jack minimum is 5". I'm now moving on to spring selection.

To re-cap, I ended up with a bit too much bump:droop ratio, which could be improved when laden with soft springs (more compression = less bump/more droop). In addition, the trike is obviously roll-happy, and so maximizing the decrease in CG with passenger mass via soft spring deflection is desirable. Finally, greater travel with soft springs gives increased grip.

It seems there are a number of common methods to select spring rate:
Copy from aftermarket kits or forum posts: unreliable, especially with a new & (technically) unpopular platform
Solve for spring rate with zero preload at ride height (the Hyperco calculator)
Solve for suspension frequency: given the wide range of acceptable results, I'm unclear if/why specific suspension frequency would be prioritized in suspension design

Why not solve for minimum spring rate? Probably because it would maximize unladen to laden ride height change, which is usually harmful, as it would negatively impact ground clearance and aero. With a heavy load and/or short travel, the car would be riding on the bump stops.

But in this specific application, since the trike isn't aero to begin with, I want to maximize that laden ride height reduction to lower CG to reduce load transfer. My passenger load is fixed; I don't have to worry about varying cargo load. And I have 3.7" of available compression travel in the shock to avoid the stops.

To best use a minimum spring rate, I would have to use the entire load available from the spring, just over solid height (coil bind). That is, the spring would almost go solid, or coil bind, exactly at full shock compression. So if I know total available spring height on the shock at full compression, could I subtract the actual preload height required (3.5") and the solid spring height (3.5") to give the free height remaining at full shock compression? It seems this clearance should be as close to zero as practical. Alternately, optimal spring rate is where height remaining after preload, at full shock compression = spring solid height?

Is this reasonable?
What is an appropriate shock compression or wheel bump load capacity?

 

[GregLocock]

Coil springs actually soften slightly as you compress them, in theory. But the detail of the loading at the ends is more important, typically this reduces the number of coils somewhat as the ends engage in the spring cap, hence increases the rate.

You would be unwise to try and take the full load seen in the wheel by the coilbound spring alone, a jounce bumper is mandatory. Otherwise you'll break something.

Typically non aero race cars are designed for 3g vertically and it would seem that many on-road cars have been successfully designed for that. That is rather low in my opinion, based on measured road loads.

Running coil springs to full compression can be OK, it depends on the exact geometry. Production springs are often designed to yield slightly at coilbound so that the length at the check load can be set (this is known as scragging).

Another option is a rising rate spring, in which the end coils have reduced pitch, or I suppose, increased diameter.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BrianPetersen]

I come at this from motorcycle roadracing.

The suspension needs to be kept off the jounce bumpers, with some reasonable margin left to account for irregular surfaces, under the highest loadings normally seen. You can exclude the once-in-a-blue-moon imaginary scenarios, but you can not exclude the loadings from the highest foreseeable braking load on level pavement and the highest foreseeable acceleration load on level pavement and the highest foreseeable cornering load. If you are on a track that leads to higher "g" loading as a matter of normal riding (e.g. the banking at Daytona) then that's going to be part of those highest foreseeable loads ... on the normal tracks that I ride on, which have "normal" uphills and downhills, it can be ignored.

The solid height of the spring doesn't factor into this other than "don't hit it". It doesn't matter if there's open space between the coils when the damper is fully compressed. The spring's free length matters and the spring rate matters (and it has to physically fit the installation) - nothing else. At full damper extension, the spring needs to have non-zero preload - otherwise it will come off its seats, and any physically possible suspension travel beyond that in extension won't be doing you any good anyhow. My observation has been that this is only a factor when someone wants high spring rates combined with low ride height (i.e. the kids trying to make their street cars look cool in their own mind).

Our cornering loads are different from yours. Our cornering loads compress the suspension at both ends. Yours load up the outer front wheel and unload the inner one. If there is a front antiroll bar, account for it. If the force-based roll center height is above ground level, account for it.

We don't pay attention to ride frequency. It's connected to the spring rate relative to the mass, and the amount of travel available. It seems that if you select reasonable spring rates with reasonable suspension travel available, the frequency comes out in an okay range. Stupidly high spring rates lead to stupidly limited suspension travel which leads to stupidly high ride frequency, and stupidly low spring rates lead to stupidly excessive suspension travel - it will corner like a '58 Buick, outside suspension completely compressed against the jounce bumpers, inside suspension completely topped out, tire sidewalls scrubbing against the ground.

The high-dollar teams figure this out with fancy data-logging connected to suspension travel and logged together with GPS data so that they can see what the suspension is actually doing everywhere on the track - obviously with brake pressure and engine torque output also logged ... I'm not that sophisticated ...

 

[BrianPetersen]

I am gathering from your spreadsheet that your sprung mass is 718 lbs per front corner. We don't know what's on the rear wheel, but presumably for vehicle-stability considerations, it's heavily front-biased ... let's take a guess at 718 lbs per corner.

In heavy braking, a bunch of that is going to transfer forward. How much, depends on how hard you are braking (let's guess 1 "g") and the center of gravity height to wheelbase ratio, which we don't know. In motorcycle roadracing, it all transfers forward ... on pavement with nicely warmed-up slicks, I can lift the rear wheel of my bike with the front brake. For the heck of it, and in the absence of proper information, let's make that assumption with yours. So that's 1077 lbs per front corner that we want to not hit the jounce bumpers with. Antiroll bar won't help, roll center height won't help, anti-dive geometry would help but probably you don't have much ... too much anti-dive geometry causes other problems so let's assume it doesn't have any. Bottom line, the springs have to take up 1077 lbs per front corner while keeping some travel in reserve for bump compression ... let's keep it 1" off the bump stops for argument's sake.

At 150 lb/in the "extra" load from braking eats up 2.4 inches (sounds like a lot) and you need that 1 extra inch in reserve, so you need 3.4 inches of compression available to do this. At 200 lb/in this eats up 1.8 inches plus the 1 inch reserve, you need 2.8 inches of compression available.

In cornering, the limiting case is all weight transfer to the outside wheel. That's right on the borderline of flipping over. How close it actually gets to that ... depends on actual cornering load and the center of gravity height and the track width, which we don't know. But again, it's the same idea except now you have to figure out what the antiroll bar is going to do.