Are large semi tractors relevant to this?
No, same as most drag cars, the driveshaft is open so the chassis has to react the engine torque and that results in twisting of the frame. With an engine bolted to the transaxle there is no twisting of the chassis along the longituidal axis. All of the torque is in the axis of the rear axle. That torque tries to lift the front of the engine up in the frame, but it does not cause the frame to twist.
The torque steer that is associated with FWD cars comes from having different length half shafts that are relatively flexible and have some windup, as well as the differing angles that the half shafts will have because of steering effects. This results in forces that try to turn the wheels to one side or the other and this results in “torque steer”. Or a feedback in the steering wheel that tries to pull the car in one direction or the other. But what we are talking about here is a RWD car with the engine bolted to the transaxle. Since the axle shafts on these cars are short, and the same or very close to the same length, and very stiff, there are no appreciable axle windup effects present. Since the drive is to the rear wheels and they don’t have any change in steering angle (like we see in FWD cars), we aren’t going to see any effects due to the driveshaft angles. In short, what’s happening here isn’t an effect of the driveshaft lengths or angles so there isn’t any torque steer that is in any way similar to a typical FWD car. And finally, since the transaxle and engine are bolted together there isn't any twist in the frame or suspension. Also real torque steer is present if the car accelerates hard, even if going in a straight line on smooth pavement. Don't confuse torque steer with roll steer or power on oversteer or understeer. Adding power in mid corner can in many cars cause the car to turn in or push out, but that's not torque steer, that's power on oversteer or understeer, and that's not the same thing.
The chassis in P Ferraris are very flexible and easily flexed by solidly mounted engines and transaxles. Over time they become seriously torque twisted.
It is relatable to vehicles that the engine is mounted to a transaxle because the engine is mounted to the chassis. The engine transmits torque through the transaxle through the differential, through the axles, through the tires and thus to the ground. The ground provides resistance that feeds backwards though everything back into the chassis which propels the vehicles forward. This resistance is torque all the same as engine output torque, no chassis is absolutely rigid there for every chassis flexes and it must react some how. Due to the engine and driveline spinning in a certain direction one wheel will always receive more torque from the chassis flexing and this torque coming back through the driveline, this force continues to transmit both ways as long as torque from the engine is applied, energy always continues until something absorbs it or until energy is no longer applied. Is it less noticeable in a mid rear engined vehicle with a transaxle and a unibody/mono chassis than say a ladder framed solid axle vehicle? Yes but it still exists.
This statement is correct for a car with an open driveline, but it is not correct for a drive system where the engine and transaxle are rigidly coupled. While the engine is mounted longitudnally, the torque from the engine is reacted by the transaxle and converted to a torque about the rear axle. With this type of engine/gearbox layout there aren't any significant chassis twisting forces present, that's the beauty of this layout. As a result of change in engine speed, there is a very small amount of torque that is generated that acts to twist the frame. As we have all seen when we jazz the throttle when the engine is in neutral, the engine tries to rock some. But remember, jazzing the throttle causes the engine to accelerate much faster than it can when the car is in gear, so the amount of torque reaction that can be generated is really small. When you jazz the throttle you can accelerate from 1,000 to say 5000 rpm in a second or so. When you are in second or third gear it takes more than a few seconds to accelerate from 5000 rpm to 7000 rpm. Since the forces due to acceleration are a rate squared relationship, if the acceleration rate is one half, the forces are one fourth, so as you can see, there is some force here but it is a very small a second order effect, and it isn't enough to even consider. The only measurable "torque" that is present here is a moment that is about the drive axle. There is no moment that is trying to "twist" the car about the longitudnal axis, and no more weight or force is applied to one rear wheel or the other as a result of the reacted torque. What I'm getting at here is that it is unlikely that what you are seeing is genuine torque steer. Ok, so it isnt torque steer from unequal forces or halfshaft effects, why did they do it? I spoke with a good friend who is one of the gurus of Vehicle Dynamics of one of the big three in Detroit and is one of the most respected guys in the business. He is also a part time racer and did very well with his sports racer when he was younger, so he understands race cars as well as vehicle dynamics and got his expert opinion last night. He also agrees that that the only significant forces are about the rear axle and don't try to twist the car, and that classical torque steer isn't going to be present, so long as the rear suspension is symmetrical it isn't going to be there. His comment relative to why you would make the wheelbase different was that he wasnt aware of the practice and that a difference in wheelbase of 7 mm isnt going to be very noticeable in any setup of a car. The only thing that it will do is VERY SLIGHTLY change the weight transfer (the understeer/oversteer) relationship when turning in one direction as compared to the other. What that means is that it will cause the car to push a wee bit less on corner exit when turning one direction or it could plant the outside rear tire just a bit better on corner exit under power. A 7mm difference over a 2400mm wheelbase is a very subtle setup effect and unless the driver is very perceptive he really isn't going to notice it. Whether it was there to compensate for the weight of the driver or was used on specific courses to help the car get onto the longest straight a bit better, that's something that only the guys setting up the car will ever know.
Solofast, I hope you realize your spot on analysis is ruining this excellent comedy... Best wishes, Kare
Sorry, it's the engineer in us... We can't help it. One thing that would be neat though is for Jim to post the LeMans setup sheet data for his car. It would be really interesting to see how these cars were set up and how that compares with today's setups. In those days racing tires were strictly bias ply so they didn't need to run as much static negative camber as today's lower profile radial tires, but I'd be real curious as to what kind of toe settings they used (both front and rear) and they bias the caster in the front, and what camber settings that they actually ran. Also I wonder if they biased the corner weights or if they used a square setup. Most of the time this kind of stuff is lost to history and luckly Jim has been blessed to have this data. As he has realized, if you didn't know any better you would most likely set the car up the way we do today and they obviously didn't really do it that way back then. Similarly with 002C, what kind of setup would you use? Today we would routinely cornerweight the car, properly balance it and adjust the suspension to a much finer degree than they did in the early days. The net result might, in both cases be something that would drive a lot differently than it did back in the day.
Jim, the staggering may relate to the sense of rotation of the engine. Please confirm that all cars you mentioned have engines rotating in the same direction ( CCW when looking from behind the car ?).
The three Ferrari's do. Ferrari often changed wheelbases by a small amount. P3 and P4 wheelbases differ by 12mm.
Very interesting. The only thing I will add is that the rotation of the prop effects the direction the hull wants to go well before full power is employed and any form of hull roll happens. This is paticularly true with sail boats which are specificly designed to limit roll with hundreds of pounds of ballast in the keel. I am not an engineer but spent many years with boats. The interesting question I have about high powered competition cars, is this torque effect on the chassis primerlly from the engine block or the drive shaft[no torque tube] or both. just one man's question tongascrew
What happens with open driveshaft cars that have the gearbox bolted to the engine is that you get a boatload of torque going down the driveshaft. That same torque has to come out the other way, in that it's reacted by the engine mounts and eventually the chassis, but the torque isn't just the torque of the engine, it's the engine torque multiplied by the gear ratio of the transmission. If you had 400 ft lbs of torque at the flywheel and the car had a 2.5:1 first gear you are looking at 1,000 ft lbs going down the driveshaft, and that is what the chassis has to react, and it can be a lot higher if you add the effect of a torque converer, or the shock load of the flywheel. That's why you can feel older open driveshaft cars twist up when you launch them. That same 1,000 pounds of torque goes into the rear axle, reacting (over a 5 foot track) 200 pounds less pounds of vertical force on one rear wheel and adds 200 to the other wheel.
In the day they also used tires to stagger setup. They'd put tires with different number of plies on different sides in some cases. They also had special shim cutting machines to make shims that could make changes. They also set up chassis differently after a chassis had crash damage with these shims.
Thanks. this is a very clear explaination of what's going on. I had trouble with the theory that all this chassis twist came from the engine.In road cars the engine mounts are some what flexible and absorb some of the vibration and twist effect from the engine. With all out race cars the engine mounts are usually bolted direct to the chassis without any cushoning. This I would think would actually increase the twist effect to some, probably small degree, with the the vast majority of it still generated from behind the gear box into the drive train. Ferrari had a problem with the early large displacement Lampredi engines when mated with 212 type gearboxes which couldn't handle the additional power/torque. We also see FerrariGilco continuing to increse the cross bracing of the chassis as the power increased. In an earlier post I asked the question who actually discovered this chassis twist and the solution with the different lengths. Maybe someone out there can enlighten us. just one man's opinion tongascrew
Good idea! Don't be surprised if there is no stagger. At that time Gilco made it clear to S F the X bracing between the main chassis tubes was more important than the weight savings by eliminatin the X bracing. This was clear with the two early chassis Ferrari ordered from Gilco with no X bracing. Gilco warned Ferrari at the time that the weight saving would be more than offset by the chassis twist effect on the suspension and they were correct. And these cars had engines of very limited power and narrow tires with very limited cornering. just one man's opinion tongascrew
Good points. One of the advantages of the mid engine designe was the elimination of the open driveshaft and the ability of the engine block and transaxle to be mounted so as to even better stiffen the chassis and controle twist particularly in the backend where so much of the twist occures. just one man's opinion. tongascrew
Expanding a bit on the good points made by Solofast , we may consider that the only reason for having the engine in a racing car is to generate a robust torque ( rotating force ) which over crankshaft and gear train can spin the wheels thus allowing for vehicle acceleration. The engine torque induces reactions which have to be balanced (carried to ground ) over the 3 axes of the vehicle : longitudinal. transversal and vertical. The longitudinal ( pitch ) and the transversal (roll ) reactions are grounded via frame, suspensions and tires. The torque in the vertical (yaw) axis, , which would tend to induce a turn in the vehicle, is balanced via steering wheel, suspensions , front tires or by wheelbase staggering. As soon as the engine spins always in same direction, the staggering is needed at one side of the car. Of course staggering was an extreme "finesse" to smooth as much as possible the handling of racing cars to prevent unwanted vehicle reactions. MB might have had CW rotating engines, which would have required wheelbase staggering on the vehicle left side. Sorry for the over-simplification of this matter ( vehicle dynamics ).
Could I make a simple suggestion? Why not set it up even, run a few laps, get some times - And then retry it with the offset? If you want a scientific blind result, don't tell the driver.
Carroll Smith ("Prepare to Win") says: "The wheelbase should be adjusted until the left side is equal to the right...", and he should know what he's talking about, but if the car is faster with a staggered wheelbase, and/or easier to drive quickly, then I guess that's the point on a race car. Depending on what you used as a reference point when measuring the wheelbase, it could make a significant difference if the front wheels are not pointed EXACTLY straight ahead when the measurements are made.
This is what we do. We set the car up as per Targa Florio set up sheets. We then test on the road and on the track. I give my feedback and Sal tweeks. For Daytona where we ran 0846 last year before the 24 we modified the set up a bit. We use this tool to set up the wheelbase and stagger. Image Unavailable, Please Login Image Unavailable, Please Login
Very interesting - so you do think the stagger helps? Just as a side point, are you going to try this on the P4/5C?
On Vintage P cars which have very flexible chassis I do especially as modern tires tend to tramline on crowned roads. I like them set up so at speed you can take your hands off the wheel and they go straight without tweeks they won't. You can drop 70,000 lb. on P 4/5 C and the roll cage won't deflect more than 1/2 of an inch so that chassis is MUCH stiffer and likely won't flex appreciably. I will track test P 4/5 C a lot before deciding on final road set up.
Jim, I've just read through this entire and most interesting thread. My first thought was back to the TR-3 that I raced when in the Navy in the late 1950's. The first time I had it on a road racing circuit I noticed the car's very slight tendency to steer left. Took it to my racing shop and learned that the front wheels had castor deliberately set to counter the crown effect on all public roads (in the late 50's there was no interstate highway system and almost all roads, even U.S. highways, were only two-lane). I hadn't noticed this tendency to "climb" the road camber until racing on the perfectly flat airport runways on which we used to lay out our racing circuits. It occured to me that the slight difference in wheelbase might have a similar effect. From then forward I always set the castor for perfectly neutral steering.
