Going a little in depth with the front suspension

sdibbers

Well-Known Member
As some of you know I'm playing with making adjustable spring seats for Beryl. I've already made the ones for the front, and will need to make the rear ones soon so I can fit them all. I've been chatting with @cobraboy about his findings when he built his Group 2 replica. This got me thinking about the front suspension and how loads are distributed.

I did a bit of work on Beryl this week. Also took some time to scan the front suspension setup (I have the toys, so I thought I might as well!).

I'd borrowed a spring rate tool and measured the rates of some original springs recently. They came out at:

Front: 140Lbs/inch with a free length of 16.25"
Rear: 240lbs/Inch with a free length of 13.75"

When I measured the static installed length (I had the car's weight on the bottom ball joint to preload it, it cam out at 255mm (10.08").

I guess it means its under a preload of 875lbs. 16.25"-10.0"= 6.25"

6.25*140lbs= 875lbs. So @cobraboy front springs at 850lbs makes sense.

I did some maths on the corner weights (in an ideal world)

Stock car weight is listed at 2810lbs. It's stated that weight balance on the P6 is 55/45 favoring the front. So that should work out as:

Front corner weight: 772.75lbs per side
Rear corner weight: 632.25lbs per side

I'm going to recreate the front suspension in CAD and see about setting up simulations. It'll be interesting to graph the wheel spring rate vs the actual spring weight.

Here's a quick screen shot from the scan data.
1669054838661.png
1669054892257.png
 
All good stuff !
I wish we had played with this simultaneously, it would have been fun. I await with interest your findings on rear spring rate, I feel my car was still under sprung at the rear, and possibly front as well, but for a P6 it sure knew how to party.
Springtime for Beryl looks promising - see what I did there.............
 
Now you've got my interest, I had done some rough static calculation for this too, but hadn't modeled it.
I'll see if I can find my calcs.
Must get up to speed again on FreeCAD.

Found my Calcs.
1669060330854.png

For your Corner weight calcs, subtract the unsprung weights from your load on the spring. I can;t remember if I measured my weights for were just best guesses. I think by that point we had bought the springs so were going to use them.
My spring displacement ratios where not accurately measured either.
 
Last edited:
Now you've got my interest, I had done some rough static calculation for this too, but hadn't modeled it.
I'll see if I can find my calcs.
Must get up to speed again on FreeCAD.

Found my Calcs.
View attachment 22630

For your Corner weight calcs, subtract the unsprung weights from your load on the spring. I can;t remember if I measured my weights for were just best guesses. I think by that point we had bought the springs so were going to use them.
My spring displacement ratios where not accurately measured either.
Good point on the unsprung weights. I did calculate those before with estimated weights for the unsprung components.

Here's the estimated weights I put together a while ago and forgot until you mentioned this :)

1669063004442.png

To your notes on ARB's etc. Beryl already has a 1in Hex 4043 flame hardened front ARB. Koni adjustable shocks all around. (They are at 90% stiffness so probably will be sent off to be revalved - there's a Koni approved company here on the US East coast).
 
The book gives front/rear rates as 170/260 lb/in, with 2 different free lengths the same rates.
 
The book gives front/rear rates as 170/260 lb/in, with 2 different free lengths the same rates.
That makes sense. The springs measured were used 50 year old ones. So the 10-20lbs lower rate would make sense.
 
I have a fantastic spring collection, at first I thought they were breeding in the shed, but no, I bought them all :rolleyes:

The problem with the rears is going to be getting clearance in the shock tower on the base unit. Jim has realised this and re worked the area with more room and extra bracing.

I was also set on following the soft spring, firm damping approach. I may have taken this too literally and struggled finessing it, I was nearly there but never quite nailed it.
 
The prefix to the Milliken and Milliken - Race Car Vehicle Dynamics, Chapter 16 Ride and Roll Rates,
"Figuring the suspension of a car is almost a matter of making useful approximations. It is not an exact science. But neither is it a blind application of rate-of-thumb principles"

I work to Ride Rates, Wheel Rate in spreadsheet. For an non-aero road road car, aim for between 1.5 and 2.0 Hz. As I said for the front springs we only had what we could get, hence the slightly soft ~1.0Hz at the front. The rear should be a higher rate, so that the rear end deflection will catchup with the front, when the car travels over a bump to control pitch.

Then look at the roll rates (deg/g), aim for a non-aero road between 5.0 (firm) and 4.0 (very firm). With the 1" anti-roll bar I got 3.37 deg/g, but with the soft front springs and lack of camber recovery on the front suspension geom, I wanted an Extremely firm roll rate in an attempt to keep the front tyres working (upright by minimising roll).
This ended up with a ~58/42% biased slightly to the front, which will result in a predictable (understeer) car as the lateral forces increase and decrease in a corner. At the time we had an open diff, so was happy to have a front transfer bias I.e. not lift inside rear wheel.
1669106922807.png

Once this is all done, you start on the dampers. Required travel is now known (length of dampers and springs). The damping should be as soft as you can deal with. Start with the dampers at min and increase until you can't fill any rebound, then turn back one click.
At this point you should have a good handling car. Then the fun part starts, and you can turn the dampers up and down by one or two click to alter the dynamic behaviour of the car, I.e. Turn-in, power-on, power-off etc. You are altering the front/rear weight transfer as the car rolls, but not affecting the steady state handling in the corner (once the car is at max roll).
I campained a single-seater for many years sprinting and hillclimbing. I loved the science and art of tuning the dampers for each track and hill.

If the car is handling well, via the springs (coil and roll), you should find a minor change to the dampers will have a noticeable affect.

All that being said, for the P6 it is all a compromise. Lets read the first paragraph of this post again:
"Figuring the suspension of a car is almost a matter of making useful approximations. It is not an exact science. But neither is it a blind application of rate-of-thumb principles"
 
You can grab the spreadsheet at :
https://criticalsystemdesign.com/P6/RoverP6_DeDion_Roll and Weight Analysis.ods

It is not an XLS, but if you use MSExcel it should convert OK. Also, it was recovered from a crashed harddrive so may not be complete etc.
Also..... it was designed for an other car, so some of the other tabs etc may not relevant.
Thanks for this @Gargo I was familiar with the concept of having a slightly higher frequency spring for the rears to deal with follow through on bumps. But this adds some very helpful details. When I get a chance to use my laptop I’ll grab your data and take an in-depth look.
 
@Gargo Really interesting reading on your spreadsheet. what material were you referencing for the ARB? I'm running 4043 steel 1" Hex bar flame hardened to 65 brinnell (IIRC, made it a good few years ago.

The spring vs wheel displacement is very helpful. How did you calculate the front spring displacement? The arrangement makes for a non linear displacement over arc doesn't it?
 
@GargoHow did you calculate the front spring displacement?
I found this sketch, maybe you could check that against your scan?
1669155047008.png
If the sketch is correct, the geometry would give a ration of 7.25:13 I.e. 1" of travel is 0.55in at the spring. (Spreadsheet entry is 0.5)
Yes, it will not be linear, over the travel, but as per the book it is "a matter of making useful approximations."
Looking at it I think it is slightly rising rate, but you need to choose something to make the calcs. And there are so many other compromises and shortcomings to over come, don't get too hung up on it. All this theory is for nothing, when hit a pothole, while controlling a 4 wheel drift.

As for the roll bar ratings: I've got the 'magic number' of 19700 representing the modulus of elasticity, but it is not the Modulus of Elasticity in ksi, which seems to be typically around 30,000ksi. I'd need to look at the equations for Torsional stiffness of a bar again to work out how you get the 19700. Maybe tomorrow.
As for the difference in stiffness between the Alloy Steel, I'd not be too worried, it should be simple enough to find a figure for the material you know you have.
 
Hi Sdibbers,

I hope that you don't mind, but I will just provide a few notes, so as to avoid any confusion. The rate of a spring has either imperial units of lbsf/in or N/mm. Note that lbsf is used as it is a vector quantity describing a force. The relationship between force f, spring rate (constant) k, and displacement x is f = kx.
Pounds lbs and Pounds Force lbsf should not be interchanged as it can lead to spurious answers. Ideally, using SI units is much better as it is clear that we are dealing with force and not mass.

Mass and weight mean different things. The mass of a body is always the same regardless of where we are in the universe, whereas weight can change depending upon location. Mass is a scaler, and weight is a vector, meaning it has both magnitude and direction. A person may have a mass of 80kg, but their weight is 785N. In imperial units, the nominal mass of the Rover is 2810lbs, but its weight is 90482lbsf.

Hi Gargo,

I am afraid I am not aware of the magic number 19700 in relation to a roll bar, but Modulus of Elasticity E is a measure of the stiffness of a linearly elastic material. E is directly proportional to stress and inversely proportional to strain. Torsional stiffness within the context of a prismatic bar, let's say circular in cross-section, is a measure of geometry. It can be found by using the polar second moment of area.

Ron
 
I 'magic number' of 19700 representing the modulus of elasticity

Torsional Stiffness = GJ/L lb.in/radian
To convert to lb.in/deg, multiply by pi/180

G=Shear modulus
J=Polar moment of inertia (for a solid bar = pi x d^4 / 32
L=Length of shaft

My magic number is the combination of the Shear Modulus * J calc and the converting radians to degrees. pi/32 * pi/180.
So, using a G=11.5x10^6psi gives a magic number = 11.5x10^6 * pi^2 / (32*180) = 19704 psi

Hi Gargo,
I am afraid I am not aware of the magic number 19700 in relation to a roll bar,

Ron makes it feel like I'm being peer reviewed for a Masters....:cool:
A magic number in programming refers to a number that just appears without reference or explanation, this a good example. Please excuse the lack of documentation. I was young and never expected to get this old.

Regarding AISI 4340 alloy steel, see:

AISI 4340 Alloy Steel (UNS G43400)
Which lists "Shear modulus (typical for steel) 11600 ksi"
So, replace my 11.5x10^6 with 11.6x10^6 gives the now not so magic number of 19876psi
 
Its probably worth while actually measuring the vehicle weight as there are several differing values in the literature and all of them are based on the standard car with no options. air Conditioningand Power steering both add quite a bit as does Power windows. In the first two all the extra weight is at the front.

Leveling out your roll centre is also a good move as it will make your steering behave predictably. Also fixing the front bottom links so they don't fight one another will make things follow the numbers better.
 
Back
Top