Bouncing Back

This article first appeared in a slightly different form around Christmas. It was wihdrawn due to a 3rd party misunderstanding now amicably resolved for us by author Allan Staniforth. We would like to thank Allan for allowing us to use the article and also Speedscene magazine, where it first appeared. For those unfamiliar with it, the excellent Speedscene is the journal of the Hillclimb and Sprint Association and is an invaluable source of information on the innovative thinking that goes into speed event vehicles.

Allan was a former Monoposto Club member and has given practical talks to the club on suspension theory and practice. He designed the Terrapin and through his Terrapin Services consultancy has advised on vehicles as diverse as Minis, Locaterfields*, a Lola 598, Radicals, Van Diemens and an F1 Ensign. For a background about the Terrapin and a few words about Allan, who crossed the finishing line in May 2009, click here.

Allan Staniforth -"an ability to go on working with little or no deterioration with nil maintenance", though lubrication is beneficial

Asked to pick the most widely used, cheapest, simplest and most reliable part used in virtually every competition car for the past 60 years you would be hard pressed to find an alternative to the steel coil spring.

With international controls on its technical composition, competition among hundreds of manufacturers, extraordinary simplicity, an ability to go on working with little or no deterioration with nil maintenance - the list of virtues goes on and on. Good reason therefore for understanding how it might be used to best advantage in any specific vehicle - including yours.

Even should a professional manufacturer, large or small, rather than you personally, have constructed your current transport, rally car, single seater, sports trackday or circuit racer, it does not mean the coils in its suspension are either original or ideal. Designers often face difficulties in seeking any ideal choice, whether personal or

A Van Diemen Coil Spring "an ability to go on working with little or no deterioration with nil maintenance"

commercially forced. One simple example: the massive sprung weight increase in both a sportsracer at 200mph at the end of the Mulsanne straight and a mundane saloon driver being joined by a wife, three fat friends and generous holiday luggage. In both cases the penalties can be very similar - reduced running and ground clearances for tyres and suspension links, poorer handling and road grip, altered wheel angles, a required pressure increase, even bodywork distortion. In either case there may appear to be a similar answer - fit stiffer (ie harder or higher rate) springs - but in both cases the penalties are serious and unwanted.

Despite the normal disclaimer that ' a little knowledge can be a dangerous thing' I prefer the risk to total ignorance. Consequently what follows is offered with some diffidence, as I am well aware that there are highly likely to be readers with their own views ranging from polite scorn to amused contempt, but the approach I will now outline has served me well with a widely differing range of drivers and vehicles.

So let us start, before ever looking at a coil in detail, by considering suspension frequency as a fundamental starting point, an aspect with well over a century of thought, application of brain-power, experiment and maths behind it. Although we are often not aware of it, virtually everything has a frequency at which it will vibrate and complex structures such as a steel or alloy wheel, a moulded rubber tyre, a hydraulic damper and a spring will interact into a further frequency all of their own. Ignoring the earliest simplistic versions of a slice off a tree-trunk, the pace of a marching man of 30 a minute (or 40 if you 're light infantry) was felt to be a good starting point for human comfort in the first sprung people carriers.

This proved a little sick-making in early 1930s America, riding on the earliest coil springs and IFS designs and by post WW2 and a slow invasion of Europe (still clinging to the inherently stiffer leaf against the coil), a figure nearer 60CPM (cycles per minute) was beginning to emerge as the comfort region for an average human being. One has to admit that the appearance of the hydraulic 'shock absorber' or more accurately, damper, complicated the whole picture and is still doing so, but they will not be considered here as they do not affect the basic approach and calculations to be outlined.

It is now reasonably accepted that an approximation of target CPM figures appropriate to their different jobs in life areshown in the box below left.

These figures are given as the roughest of guidelines, less as advice than a flavour of an approach.

The first step is normally the simple decision - what have I got or what do I need? The answer will provide two possible starting points: A - existing vehicle with unknown coils, B - as the designer with a chosen target to integrate with planned wheels, linkages, leverages and mountings.

A', identify the coil fitted by removing it from the vehicle with its damper. Check the free length and search for any marks of a rate, usually etched (good) on one of the flat, ground end faces, painted (not so good) or absent (not good at all). Happily this last case is not insurmountable. Ideally, find a man with a pukka spring tester who will be able to check it in minutes. But in the absence of such help all is still not lost. It is still practical to measure and calculate the rate to a quite surprising degree or accuracy on the bench.

Notc that the better quality coil springs are almost always marked with free length in rounded inches and a rate in lbs. They may alternatively be in metric or printed on one of the individual coils, but let us keep it simple in Imperial for the time being. The rate is the how much weight is needed to squash it one inch. The equipment needed to do your own detective work is simple, but essential.- a digital caliper, a tape measure and two 12 inch steel rules, paper and pencil and a calculator that can handle square and square root functions.

Procedure (check and note in all cases):
1 Free length (flat to flat will indicate any collapse)
2 Wire diameter, preferably with any paint scraped off (or subtract 0.004" for spray or 0.007" for high gloss colour)
3 Thickness of the two combined steel rules
4 Lightly tape a rule to each side of the coil and set up gently clamped in vice jaws
5 Measure total diameter across rules (see diagram)
6 Use this data to calculate the accompanying equation (right) which, needless to say, is the product of a mega-brighter brain than mine.

Hopefully the example set of figures shown right for a very typical 8in x 2.25 10 x 175 lb/in coil will clarify the above procedure. Note that the free length of the spring has no effect on its rate.

Aim at the very highest degree of accuracy you can achieve as multiplications to the 3rd and 4th power are capable of introducing major errors. Having a proper hydraulic compressor available capable of measuring compressed length to 0.001", I can assure you that you can achieve an accuracy of +/- 10% at the very worst, more often 6% - 2% as the required tolerances for manufacture fall well inside these figures anyway.

You can now set about finding the frequency for what you have or what you would like to have. The great value of the frequency approach in my mind is that rather than bald quotation of a coil or wheel rate, it takes fully into consideration not only the coil, but also inclined mounting, the weight on it, as well as leverages that can alter its effects. All are combined to give an honest comparison between front and rear, car A or car B, or any experiment you want to make in search of better handling. They also allow you to work backwards in search of the new coil rate needed to provide you with specific frequency by reversing the maths sequence.

Again we will employ a set of typical figures for an average 'Locaterfield'* with wishbone front suspension and inclined coil/damper units operating from a point part way along the bottom arms.

First and foremost you have to bear in mind that the car will rarely, if ever, see your 175 lb spring as 175. but some lower figure. Which takes us to:

Step 1 The reducing effect of it being mounted at an angle to the link that is operating it, ie wishbone, rocker, push/pullrod etc to provide an Effective Rate (ER)
Step 2 The effect (often much larger) of any geometric leverage which must be measured and then squared.
Step 3 Divide the ER by Leverage squared to obtain Wheel Rate
Step 4 Obtain the four sprung corner weights (each scale corner weight minus wheel, tyre, upright. hub and bearings, brakes and half of any links and coil/damper)

Step 5 Insert data into the equation that will utilize it to provide the suspension frequency we have been seeking.

For an example of this final, vital step we again use a set of typical dimensions as sketched left.

A little study and reflection on this calculation will give you a good idea of the magnitude and effects of variations in the data fed in. It should also be clear how you can compare front to back, car A with car B, or quantify the probable results of, for instance, weight variations whether from a big fuel tank, a big wing, a big mother-in-law, or a set of magnesium wheels.

So what are the targets for which you should be aiming? We have already considered some figures for various areas of the sport. Long usage in road cars means they are going to be using some 10% lower figure at the front to reduce or eliminate fore/aft rocking motion (which. in extreme circumstances, can cause a rally car to somersault) but this may well be reversed in competition. It is a free world and coils are, relatively speaking, not expensive items with which to experiment. You can of course start with a chosen CPM and find what coil rate will be needed to provide it by reversing the sequence of the calculation.

What it will not give you, unfortunately, is exactly what you and your vehicle need, or precisely how either of you will react to such changes. Not to mention the effects of antiroll bars, damper valving, bump and rebound adjustments, aero downforce, tyres and the pressures therein. But you will have at least part of the picture, and maybe the chance of a profitable bargain the day you see a box of mystery coils going for a song.

Allan Staniforth (inc drawings)

* The term Locaterfield may not be familar to all. It was coined by hillclimbing legend Reg Phillips to describe Lotus Sevens, Caterhams and Westfields.

Javelin outboard pullrod - on your own for the leverages

Appendix 1

It might not be immediately obvious by what amount the rate of an inclined spring should be reduced. Looking at the diagram left, where the angle at the top of the triangle is 32degrees, the black line represents the spring which, pushing at an angle of 32degrees has a horizontal and a vertical component. With the dimensions shown, the downward push of the spring will be 17/20 of the total spring force and the sideways push 10.6/20. A recollection of school trigonometry tells us that if we know the angle of the spring, the downward push is the total push (ie its nominal rate) multiplied by the Cosine of its angle to the vertical.

So for the example where the spring is at 32 degrees to the vertical Cos(32) is 0.848. hence we reduce the spring's rate by 15%.

Although the example is for an outboard suspension, the same principles apply to any use of a coil spring be it rocker, pushrod or pullrod. Though in working out the leverages on a pushrod, you're on your own!

Further analysis can be found in "Race and Rally Car Source Book" by Allan, published by Haynes at £19.99.

Appendix by Tony Cotton