F1technical.net

xpensive wrote:

tommylommykins wrote:do you also get slightly sinusoidal rotation speed from the geometry of CV joints?

You do indeed with a conventional joint, but that is directly neutralized with two joints on the same shaft interacting.

only neutralized if the wheel axis of rotation is parallel to the diff axis ?

BTW why would you want a CV joint on unsteered (ie rear) wheels ?

to account for suspension travel. On this type of suspension, the differential is fixed to the chassis, while the wheels can move up and down relative to the chassis. Thus, the driveshaft cannot remain at the same angle relative to the wheel or the differential. The CV joint allows this movement to occur while ensuring the same angular velocity across the joint.

Tommy Cookers wrote:

xpensive wrote:

tommylommykins wrote:do you also get slightly sinusoidal rotation speed from the geometry of CV joints?

You do indeed with a conventional joint, but that is directly neutralized with two joints on the same shaft interacting.

only neutralized if the wheel axis of rotation is parallel to the diff axis ?
...

Not necessarily, the pulsating angular speed (w2) of the driven shaft as a function of driving (w1) of a universal joint is;

w2= w1 * cos alfa / (1- sin^2 alfa * cos^2 w1t)

As long as the "claws" on the intermediate shaft are parallell and alfa is the same in the two joints, pulsations are neutralized.

In other words, two universal joints can link two shafts with a combined 60 (30+30) degree angle difference without pulsations.

Try it and you'll be sorely disappointed. There are other effects that do not cancel.

Greg Locock wrote:Try it and you'll be sorely disappointed. There are other effects that do not cancel.

O'boy, and what effects might that be, I for one am waiting in suspense for you to drop the big one now?!

Inertial. The input and output speed variation effects cancel, as you say, but the accelerations required to change the speed of the intermediate shaft over one revolution are not cancelled. The torque pulses created get worse with angle, rpm, and inertia.

Jersey Tom wrote:

rjsa wrote:Torque knows only direction, not point of application.

True but kind of deceiving. There are suspension topologies where the torque from the engine is resolved internally and/or there are no anti's and it's all nice and easy.

There are others where you can have significant longitudinal anti's, body roll with drive torque, etc... all which has to be accounted for. The torque ultimately has to become a force at the ground after all.

Anyway what I say still holds. No matter where you apply torque to a solid body, for the same amount and direction you will have always the same resulting forces.

Greg Locock wrote:Inertial. The input and output speed variation effects cancel, as you say, but the accelerations required to change the speed of the intermediate shaft over one revolution are not cancelled. The torque pulses created get worse with angle, rpm, and inertia.

Let's try and put some numbers to it then;

- On an F1 car, the shaft's Rpm is some 2300 at 288 km/h, while the torque can be as much as 2200 Nm.

- Up to an angle of 10 degrees, the speed pulsations of the intermediate shaft are less than 4%, see above.

- Conculsively, at 2300 +/- 100 Rpm at a frequency of 40 Hz, you might have a point there, but I have no
idea of the inertia of the intermediate shaft, is it 400 mm long, 20 mm diameter and made of steel?

Lycoming wrote:to account for suspension travel. On this type of suspension, the differential is fixed to the chassis, while the wheels can move up and down relative to the chassis. Thus, the driveshaft cannot remain at the same angle relative to the wheel or the differential. The CV joint allows this movement to occur while ensuring the same angular velocity across the joint.

CV joints were made for fwd, they are worse than pointless for non-steered wheels

when steering the wheel axis is out of parallel with the diff axis, the attempted variation in wheel angular velocity can be fed back to the driver (much of the typical variation cancels internally ?)

before the Mini fwd had (for 30 years) Hooke type joints, steering angles were minimised by design (and wise drivers)
certainly its transverse layout would not have been possible without Mr Rzeppa's CV joint design

BTW ......
I schemed a negative camber rear axle conversion (live axle/Hotchkiss axle), using internal flexures

has anyone ever made anything like this ??

RE: CV joints on RWD vehicles

There are cars with U-joints on the rear axle, but nearly all modern cars have CV joints instead. U-joints are cheaper to manufacture, but CV joints will give a smoother power delivery (the "output" of a u-joint doesn't spin at a constant speed given a constant speed "input", it's more of a sine wave - hence why CV's are called "Constant Velocity" joints). Both serve the same function, just CV's are better in almost every way, not just increased range of motion (as required by FWD).

However, no F1 car has used CV joints in quite a while - they're just simply too heavy. Instead most small formula cars will use what's called "tripod joints". They're kinda like CV joints in how they function, just instead of using 6 ball bearings they use three sealed bearings.

xpensive wrote:

Greg Locock wrote:Inertial. The input and output speed variation effects cancel, as you say, but the accelerations required to change the speed of the intermediate shaft over one revolution are not cancelled. The torque pulses created get worse with angle, rpm, and inertia.

Let's try and put some numbers to it then;

- On an F1 car, the shaft's Rpm is some 2300 at 288 km/h, while the torque can be as much as 2200 Nm.

- Up to an angle of 10 degrees, the speed pulsations of the intermediate shaft are less than 4%, see above.

- Conculsively, at 2300 +/- 100 Rpm at a frequency of 40 Hz, you might have a point there, but I have no
idea of the inertia of the intermediate shaft, is it 400 mm long, 20 mm diameter and made of steel?

So, when Torque is polar moment of inertia times angular acceleration; T = J * rad/s^2 (1);

- A 400 mm long and 20 mm dia steel-shaft; J = m * r^2 / 2 (2), gives J = 5 * 10^-5, doubled to 10^-4 for good measure.

- From 2200 to 2400 rpm and back with 40 Hz, means a delta-w of 21 rad/s in 0.0125s, gives 1680 rad/s^2.

- Finally, Torque increase is according to (1) some 0.2 Nm in order to care for the inertia of the shaft, zip really?

xpensive wrote:

tommylommykins wrote:do you also get slightly sinusoidal rotation speed from the geometry of CV joints?

You do indeed with a conventional joint, but that is directly neutralized with two joints on the same shaft interacting.

To be clear, a conventional joint is a U-joint. A constant velocity joint provides essentially constant rotational speed.

Tommy Cookers wrote:CV joints were made for fwd, they are worse than pointless for non-steered wheels

Nonsense. Many powerful RWD cars with pretensions of civility use CVs (typically tripod inboard, rzeppa outboard). I have a sneaking feeling that other than off roaders CVs rather than Hookes for the rear axle are in the majority.

xpensive wrote:

xpensive wrote:Let's try and put some numbers to it then;

- On an F1 car, the shaft's Rpm is some 2300 at 288 km/h, while the torque can be as much as 2200 Nm.

- Up to an angle of 10 degrees, the speed pulsations of the intermediate shaft are less than 4%, see above.

- Conculsively, at 2300 +/- 100 Rpm at a frequency of 40 Hz, you might have a point there, but I have no
idea of the inertia of the intermediate shaft, is it 400 mm long, 20 mm diameter and made of steel?

So, when Torque is polar moment of inertia times angular acceleration; T = J * rad/s^2 (1);

- A 400 mm long and 20 mm dia steel-shaft; J = m * r^2 / 2 (2), gives J = 5 * 10^-5, doubled to 10^-4 for good measure.

- From 2200 to 2400 rpm and back with 40 Hz, means a delta-w of 21 rad/s in 0.0125s, gives 1680 rad/s^2.

- Finally, Torque increase is according to (1) some 0.2 Nm in order to care for the inertia of the shaft, zip really?

Slight quibble with numbers (20mm sounds a bit shy, and the yokes of the UJ are significant, but a lot will be carbon not steel) - it is an oscillating torque, not an absolute torque decrement. The reason i said you'd be sad if you tried a 30 degree articulation on the UJ is that these high amplitude high frequency torques are actually enough to break things.

Edit: I stand corrected, and figured it out.