The Mark Ortiz Automotive
Presented free of charge as a service
to the Motorsports Community
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.
TEN YEARS IN RACECAR ENGINEERING
With the publication of the November 2011 issue of Racecar Engineering, I have now had a regular column in the magazine, based on this newsletter, for ten years. The first column was in the December 2001 issue.
FIRST-GENERATION MAZDA RX7 FOUR-LINK
Over here in Australia we have a race class called Improved Production. A popular car is the Gen 1 Mazda RX7 with a live rear axle.
The suspension must retain all the original links and all bushes must be elastomeric -- no spherical bearings. Additional longitudinal links are permitted and lateral location devises are free.
Conventional wisdom says that the original links are prone to binding. They consist of parallel lower trailing arms and shorter upper arms about 10in long and angled inwards towards the front. Most people have very soft bushes in the top arms and add another set of parallel top trailing arms.
My question is this---
The track is about 60in. So if the outside suspension is compressed say 3in in a corner and the inside extends 3in the total difference is 6in giving an axle angle of about 6deg compared to the chassis. I accept that this would twist the suspension arms and perhaps cause binding but would the degree of twist not be the same in the new parallel arms as well? ie if the original arms were to bind, would not the new arms also be at risk of binding to the same degree? Does the fact that the original top arms are angled make the situation any worse?
For readers unfamiliar with this suspension, it has four trailing links for longitudinal location and torque reaction, and a Watt linkage for lateral location. The upper trailing links do splay out a bit at their trailing ends, mainly just to clear the springs. This is not a triangulated four-link, where the
link angles in plan view are sufficient to allow the four links to provide lateral location without any other mechanism.
The upper links are much shorter than the lowers. This is partly for packaging, but also it makes the side view geometry (and accordingly, the anti-squat and anti-lift) roughly in accordance with Olley’s Rule: link lengths inversely to proportional to their height above ground. This makes the longitudinal anti’s roughly consistent regardless of ride height.
The problem is that the side view swing arm length fluctuates wildly with suspension displacement. Depending on where the suspension is in its travel range, the longitudinal links try to rotate the axle housing forward with compression, or rearward, or neither, at a varying rate. In ride, no problem results, but in combinations of ride and roll, the longitudinal links on each side of the car try to rotate the axle housing different amounts, and/or in different directions. Any time the links try to create differing rotational displacements of the housing, they effectively turn the housing into an anti-roll bar.
This might be lived with if the effect were consistent (the car has a rear anti-roll bar, so eliminating roll resistance is not the point), but the problem is that this component of roll resistance varies erratically in different combinations of ride and roll displacement. In a straight line, the system behaves well, even if the surface is undulating. In cornering, it behaves well as long as the surface is smooth. However, when presented with hard cornering and a bumpy or undulating surface at the same time, the system produces capricious variations in roll resistance distribution.
It is true that any cylindrical bushing only allows torsional movement by deforming, and that is inescapable if sphericals are prohibited. But that is not the main problem with the stock layout.
In the US, at least under SCCA rules, bushing material or construction is entirely free, but the links still have to be there. It is permitted to add “traction control devices”. The most common approach is to use either urethane bushings or sphericals in the lower links, replace the bushings in the upper links with foam rubber, and add a long central upper link, bent to clear the driveshaft tunnel. This effectively converts the suspension to a three-link design with a Watt linkage for lateral location.
That does make the system consistent. It doesn’t eliminate torque roll, which should be an objective in live axle suspension design, but it solves the most urgent problem.
Packaging and load paths permitting, my approach might be to react braking torque through a centered top link that acts only in compression, perhaps a shock absorber with a snubber on the shaft, and react drive torque with a link or chain or cable that acts only in tension, offset to the right and angled down at the front.
UNIFIED THEORY OF SUSPENSION SETUP FOR PARTICULAR TRACK?
I am an avid BMW CCA club racer. I have progressed quite far and have even won my championship a few times. However, there are still a couple of guys that I just can't catch. Other than money, I think one of the reasons they are beating me is setup. They are very good at nailing a setup for a given track.
I have a good baseline which was given to me. I soften it a bit for bumpy tracks like Mosport or Sebring. But I pretty much stick to it. Once in a while, all the planets align and I set a lap record. But it's always a surprise to me. I think this is my little setup lottery.
I am an engineer by schooling (electrical). I have come up through the BMW and Porsche clubs where I have been instructing for years. I understand all the underlying principles and have read all the books. But I have no idea how to unite them all into a unified theory of suspension setup. In this respect, I am quite typical of club racers and probably a good chunk of your readers.
Do you think you can give me some guidance toward developing these skills? Where should I look?
My car is a BMW E36 M3, 410hp, 2200lbs, Moton 3-ways, slicks, etc...
I don’t see any way to reduce something that complex to a brief set of principles or rules, especially if those have to apply to a wide variety of vehicles. For production-based cars with tightly regulated bodywork, it does get a little simpler than for winged cars, because at least we aren’t dealing dramatic variations in the aero properties, but even then it isn’t simple.
One universally applicable principle would be to know why you are trying a given setup change: what are the physics? What exactly are you trying to make the car do or not do?
The basic idea of having a baseline setup and softening it a bit if the track is rough makes sense. Depending on the car and the rules, it is often necessary to use higher static ride heights for the softer setup.
In addition to general roughness of the track, factors that may influence setup include:
· Dominant turn direction. Some tracks have most of their turns in the same direction.
· One or two turns being more important to lap time or passing opportunities than the rest. Most commonly, such turns are at the beginning of major straightaways.
· Predominance of tight turns or sweepers.
Issues such as these get particularly interesting if we have ballast we can move. Production-based cars may or may not have ballast, depending on the rules.
When one turn direction is dominant, or when one turn is unusually important, sometimes it pays to move ballast toward the inside of such turns. It may cost us elsewhere in the lap, but we may come out ahead overall.
When there are lots of short-duration tight turns, braking before those turns and forward acceleration after them become more important, and mid-turn speed becomes less important. Assuming rear wheel drive, more rear percentage will help the car put power down, and will also shorten braking distances.
When there are a lot of sweeping turns, and few long straights and tight turns, it may pay to move ballast forward a bit more, to work the tires more evenly in steady-state cornering.
With live-axle rear suspension, it often helps to use a bit less than 50% diagonal (meaning LR + RF), to compensate for torque wedge. If the turn before the most important straight is a tight right-hander, there is a case for being a bit more aggressive with the diagonal than we might otherwise.
But again, in all of this it is of paramount importance to know what one is trying to achieve, and have a clear understanding of how one expects a setup change to accomplish that.