The Mark Ortiz Automotive
Reproduction for sale subject to restrictions. Please inquire for details.
Mark Ortiz Automotive is a chassis consulting service primarily serving oval track and road racers. This newsletter is a free service intended to benefit racers and enthusiasts by offering useful insights into chassis engineering and answers to questions. Readers may mail questions to: 155 Wankel Dr., Kannapolis, NC 28083-8200; submit questions by phone at 704-933-8876; or submit questions by e-mail to: email@example.com. Readers are invited to subscribe to this newsletter by e-mail. Just e-mail me and request to be added to the list.
FRONT-DRIVE HILLCLIMB CAR WITH TWIST BEAM REAR SUSPENSION
Rather than a single question and answer, this month we have an exchange between me and a questioner. As usual, the questioner’s words are in italics and mine are in standard font.
My car is a VW Polo GTI that I run in a lightly modified road registered hill climb class. Mine is a 9N3 MY2006 Gti - same floorpan as the 1.4 super/turbocharged TSI model getting around now. I run OEM GTI springs that are 15mm lower than normal Polo springs on the front. On the rear I have either the OEM 142lb springs or some even lower 130lb springs that I've been using. The car has front (22mm) and rear (20mm) adjustable anti-roll bars both set to their hardest settings. My nonadjustable front camber is -3/4 degree and on the rear is -1 degree.
I have seen pictures of my car coming out of tight 2nd gear corners and the inside rear has lifted. The car feels very good in those corners and gets the power down well considering that it has an open diff. The same pictures also show that the front outside wheel is vertical. However, in faster 2nd and 3rd gear corners when I'm trying to get the power down the car is washing wide. I had initially thought that the lower 130lb rears were too soft and switching to the taller 142lb rears would get my elastic and geometric roll resistance up at the back and be the answer.
This was until I took a closer look at the pictures of my car when it was single wheeling. The inside rear which was well clear of the road surface appeared to have had to droop a long way before it could have detached. Now I'm thinking that on these more open corners maybe the inside rear on steady state cornering or when getting the power down is unable to leave the ground as the roll angle is less than when the car gets chucked up an uphill hairpin. The extra rear grip and more equal rear load sharing is exacerbating power on understeer.
Looking at images of IRS FWD hatches showed that their inside rear only had small amounts of droop. To my mind they would be able to single wheel much earlier than my car. So what is it with twist beam rear axles that governs how much the inside rear will droop? Is it twist beam torsional rigidity alone, rear spring rate/height alone or a combination? The twist beam couldn't ever droop to the point of topping out the inside rear damper could it? Or are my fronts too soft and inside rear
lift is happening in tighter corners only because the car is pitching, which it wouldn't do so much in faster corners?
I feel a bit stuck. In my case I've fitted the stiffest aftermarket rear anti roll bar that I can (which piggy backs the torsion beam) so I've reached the practical limit of adding more torsional rigidity into the twist beam. Stiffer rear springs will increase my rear roll resistance. If that forces me to soften off the anti roll bar then isn't that just going to let the inside rear droop even more. I remember reading in your Feb 2010 that twist beam and IRS rear ends behave very differently in terms of building forces on the outside wheel when single wheeling starts, and you also wrote about droop limiting cables. Does any of that apply here?
What exactly are your rules regarding springs and anti-roll bars? What do you have for shocks, and what are the rules on those?
To stay 'type 1 or type 2' I can only have modifications located in factory positions. So I can uprate springs, dampers and anti-roll bars and make geometry changes. They allow my rear anti-roll bar when no OEM rear bar was fitted because it fits with the 'lightly modified' principle of the class and for parity with other types of cars. They baulked more at my rear shock tower cross brace but allowed it.
I can only run road legal tyres e.g. Toyo R888, Yoko AD08R. The hillclimbs I do last between 45-90 seconds which is why a lot of the street registered guys use 'street semi's' rather than proper R spec semis in the belief that they will get up to temperature quicker and rightly or wrongly I have followed suit. I'm still unsure if I should be running warm pressures (34-36psi) to begin with from cold for a stiffer sidewall or if I should be using low 30-32psi pressures for maximum grip (but poorer turn in and steering feel).
Since it’s a grass roots class and I'm not exactly flushed with cash my front dampers are still OEM gas Sachs struts (MacPherson strut front end). I have gotten my hands on (but have not fitted yet) some eccentric bolts that will allow me to shift each of the front wishbone subframes/cradles and eke out a little more than the factory camber and castor. Apart from the rear bar I have fitted Koni 'yellow' sport dampers w/rebound adjustment so I assume I have more compression damping in the rear which I would have thought would assist turn in and midcorner but not be enough to hurt corner exit. I have tried going back to stock dampers but it was worse. Come to think of it, the rear Konis are oil damped.
To try to combat my understeering I even tried running Yokohama AD08R's on the front only and street tyres on the rear. It actually worked very well on tight circuits but you could really feel the extra slip angle in the rear when speeds got up.
You mean the rear Konis are non-pressurized? (Almost all shocks are oil damped.) What setting are you using on the rebound adjustment?
Can you get a larger diameter rear bar, or make one?
RE the dampers, when compressed, they do not push back out. I'd assumed from that that they are oil damped not gas but probably have that wrong. I guess they are non-pressurized then. The rebound damping is set to minimum which was Koni's recommendation (assuming that their fronts were set to 1 turn out from minimum). The adjustments do work and have a marked impact on the rate at which you can pull the shaft back out, so that side of things is working.
The rear bar is the stiffest retail one I could find. It’s 20mm and has 3 settings. It’s on hard (the shortest arm setting) now. The twist beam is a U-channel. Could welding bridging strips across the open side of the beam torsionally stiffen the bar? Or could it be clamped in a few places across the beam somehow?
While I have tried both the OEM and Koni dampers with the lowered softer rear springs, what I haven't done is try the car with the taller and stiffer OEM springs with standard dampers.
At the time that I upgraded the suspension with front and rear anti roll bars I was already using the rear Konis. I only briefly swapped back to the OEM Sachs dampers as a test whilst using the lower rear springs and didn't find an improvement.
“Gas shocks” don’t damp using gas. They use oil for that. The difference is that they use a trapped volume of pressurized gas to allow for the volume change inside the unit as the shaft moves in and out (called shaft displaced volume). This permits the full diameter of the unit to be used for the piston, and allows the unit to damp in both directions without needing a foot valve to prevent cavitation. The Konis and all non-pressurized shocks have an inner tube that has the piston in it, and a space between that and the tube you can see. The outer compartment has some oil and some non-pressurized air in it, and oil flows into it through a valve at the bottom of the inner cylinder called the foot valve.
The reason I was asking about the dampers is that if you had stiffer damping in the rear, that could account for the car three-wheeling only in tight (hence short-duration) turns. But it doesn’t sound like that’s the case.
So I’m not sure exactly why the car doesn’t tricycle in faster turns. I’ve certainly seen Golfs/Rabbits/Polos that do. You could just reduce front anti-roll bar stiffness and/or spring stiffness, but if the car rides bumps well enough now it would be better to increase rear roll resistance instead. You definitely could do that by boxing the twist beam. You’d want to check whether that would be legal.
Making anti-roll bars isn’t real hard. A little diameter change does a lot. Stiffness varies with the fourth power of diameter. Going to 22mm or 7/8” would give you about half again as much stiffness as your 20mm bar. The most common material for solid ones is 4140 steel. You buy it in the normalized or annealed condition, heat it and bend it with torch, and if you want flattened ends, beat
those to shape on an anvil after heating cherry red or a bit hotter. Drill holes as needed, then heat treat by hardening to full hard by heating and quenching, then tempering to similar hardness to your existing bar. In some cases you can get away with not heat treating. If the material is not sufficiently hard, it generally won’t break but it will take a set when you corner hard or hit big one-wheel bumps. The car will handle differently in a right turn immediately following a left one or vice versa, and the readings on your wheel scales will be subject to mysterious large variations. If you see that, you should heat treat the bar. Rear ride height has very little effect on geometric roll resistance in your car, although front ride height has a big effect on front geometric roll resistance. Rear ride height does have significant effect on anti-lift in braking, but normally this is not a concern.
Generally, when these cars are tricycling, the inside rear suspension is not at full droop. It is being held in partial compression by the twist beam and anti-roll bar.
Reviewing my February 2010 newsletter, what I said was that droop limiting devices work somewhat differently on beam axles than on independent suspension. I was referring there to ordinary beam axles, not twist beams as in the Golf/Rabbit/Polo. The VW twist beam is almost a pure trailing arm system with a built-in anti-roll bar, because the twist beam is almost at the front bushings. The twist beam provides about 25% camber recovery in roll (disregarding compliances) and a roll center at about 25% the height of the bushings. So changing the rear ride height an inch only changes the roll center height about a quarter of an inch. If the beam were in line with the bushings, you’d have a trailing arm suspension. That system would have no camber recovery in roll and the roll center would be at the ground regardless of bushing height. If the beam were on the axle line, you’d have a beam axle suspension. That would give 100% camber recovery in roll, and the roll center would be at bushing height. Changing rear ride height with that system would change roll center height by an amount identical to the ride height change.
Ok. Well if there is not a gross damping mismatch by having the Konis in the rear, then perhaps I'll retain them and get front Konis as soon as I can afford them.
I do have provision on my front anti roll bar to go a softer hole/setting. I had moved away from this to the stiffer setting because in steady state cornering the car had more understeer set up this way and was killing the sidewalls of the tyres, I assume because it was losing front camber. Going to the stiff front anti roll bar setting really gave the car a more flat planted front end but possibly at the expense of power on understeer? I can have a play there. But first I will explore a stiffer rear anti roll bar like you said.
So if I read rightly, the roll centre with my rear end will roughly be a quarter of the way up from the ground to the twist beam bush height? Ok, so I probably need to drop the idea that slight variations
in rear spring/beam height is going to meaningfully affect geometric roll resistance - I should just think of geometric roll resistance more or less as a constant considering what I use my car for, and instead just think in terms of the effects that changes to elastic roll resistance might have? And if I do add more rear elastic roll resistance with springs rather than more bar, that would have the side benefit of minimising squat when trying to power up a hill and help front end traction wouldn't it - or would that effect be marginal too?
My front end is at the OEM ride height that has the lower control arms and steering arms virtually horizontal to the ground so I have deliberately left the roll centre alone there. If I do then lower the rear, does it not then inadvertantly add small amounts of front castor relative to the ground plane as the car is in a slight bum down attitude whereas OEM had it at a bum up attitude? If so is there any problem with that - is it likely to be felt?
Reducing squat under power does not reduce rearward load transfer. If anything, we want the rear to squat, and prevent the front from rising, so that the c.g. will be as low as possible. However, the effect is very small, and simply lowering the car statically is just as good.
Lowering just the rear will increase caster. This may actually help the car. Lowering both ends a similar amount will have little effect on caster, in this car. Lowering the front does affect geometry adversely on a MacPherson strut suspension. In addition to the geometric roll resistance being reduced, camber recovery in roll gets poorer. However, the front will generally pick up a little static negative camber from lowering, which may help when the rules prohibit adding an adjustment or moving any points. The reduction in camber recovery and geometric roll resistance can be addressed with stiff springs and bars. A car thus lowered and stiffened will generally be faster on smooth surfaces, at the expense of its ability to ride bumps.
If there’s room, and if the rules permit, it’s worth considering using coil spring clamp/spreader adjusters. These can be used to either clamp adjacent coils closer together or spread them apart. This takes a coil out of action, which stiffens the spring. You also get a ride height adjustment that can be used to adjust diagonal percentage.