h a l f b a k e r y
Is it soup yet?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
Banish driveline oscillations with inerter technology
The inerter <link> is an interesting component in
automotive engineering. It was introduced to motorsport
by McLaren where it was used to counteract the bouncy-
oscillations in the suspension that are a big problem in F1
cars with big bouncy tires and aerodynamics that amplify
When an inerter is used within the
suspension system, upward movement of the wheels is
used to generate inertia - for example by spinning up a
flywheel. If the suspension is to oscillate it must first slow,
stop and then re-accelerate the flywheel in the opposite
direction. In this way, it inhibits rate of change not
velocity, like a conventional oil damper.
Oscillations also exist in the driveline, an educational
example of this is to take a reasonably powerful front
wheel drive car, stomp on the loud pedal and dump the
clutch. In the absence of fancy electronics etc, what often
happens is that the steering wheel will wobble
accompanied by some telltale chirruping noises from the
tires as you safely pull away**. This is because torque can
oscillate between the two wheels dependent on the
varying traction and type of differential fitted.
So, fit the differential with an inerter. To do this, each
drive shaft drives an idler in the differential rotating cage
assembly which in turn drives a flywheel that is essentially
coaxial with the drive shaft/cage. With significant up-
gearing, this would lead to the flywheel spinning up
whenever there was relative movement between the drive
shaft and rotating cage, i.e. one wheel is slipping.
This will be particularly good at damping driveline
oscillations with the added benefit that it isn't just
throwing energy away like viscous diffs, it's all there still in
the flywheels. It would be interesting to know how it
drives, I imagine it would cause initial turn-in understeer
and very stable mid-corner handling on constant-radius
corners***. Probably very suitable to a center differential,
or something limousiney.
*If the car rises up slightly, the undertray develops less
downforce. Conversely if the car squats down slightly, the
aero package generates more force pushing it down
further. Add in springy tires, energy and any form of initial
disturbance and you have an expensive space hopper.
**That was confident, hesitation-free acceleration to the
speed of traffic flow officer. I can assure you, any
momentary loss of traction was well controlled and almost
certainly attributable to the damp road surface at that
particular junction, and the 6 previous junctions. The
council really should do something.
*** Possibly some strange straightening-out induced
oversteer behavior - send lots of budget, and a specced-
out S-class, I'll find out.
[bs0u0155, Sep 22 2020]
[bs0u0155, Oct 01 2020]
||Will the increase in weight and the component on component wear and servicing requirements be worth the amount of bounced torque recovered?* Not to mention the space needed for the design.
||*The fundumental knock/question of any new mechanical add on.
||Aren't there any existing components of a car that could be
made to do dual duty as flywheel masses? For instance, what if
a disk were carved out of, say, a spoiler and made to spin in
place, in the existing plane of the spoiler?
||This will add some significant wear on the clutch.
||//This will add some significant wear on the clutch//
||Thinking about this, I think I need to clarify. Take a look
at the first image of the spur-gear differential in <link>,
it's a lot easier to imagine than with the complications of
bevel gears. We have the gear attached to the drive shaft
driving the first smaller gear. Those connect over as a 1:1
gearbox to the other drive shaft.
||Imagine that first idler shaft penetrates the casing. On
the outside it has a larger spur gear connected to the
shaft. Mounted on the driveshaft, co-axially with a
bearing is a small gear driven by the larger external spur.
This is connected to the flywheel. A ballpark mass for this
would be 5kg or so.
||So accelerating straight, you would have to spin that 5kg
up to wheel RPM, since it's connected to the differential
carrier and in a straight line is geared to it 1:1. I disagree
that this will add any significant load to the clutch,
there's a lot more than 5kg rotating mass already in the
system, and the clutch already has to cope with the
forces associated with accelerating the whole car.
||Where significant forces do come in, is when the wheels
have mismatched speeds. In a one wheel slipping
situation, assuming the flywheel drive gear ratios are
100:1 (2x10:1) then one wheel being stationary and the
other accelerating up to the equivalent of 20mph
(340rpm/5.6rps) in 0.5s then the flywheel will be
accelerated up to 34,000 rpm delivering about 360 Nm
(265 ft/lb) torque to the system/other wheel. That's ball
park stuff, but actually better than I thought.
||*you only need to have an effect on one side of a
differential since they're mechanically linked, but having
identical sides might be worth it for packaging &