Having fixed the differential clonk I could still hear a clonk sound, particularly on deceleration. Since the motor does a great job of providing regenerative braking, the noise occurred frequently and was becoming annoying. So it was time to fix that too.

The problem is that I have wire wheels and rather than have four sensible bolts holding the wheel to the hub, the MG designers went with the extraordinarily elaborate splined hub design. (What were they smoking? - and can I have some?).

The hubs and wheels have 75 tiny splines upon which the entire forces of acceleration and deceleration are placed - all at a diameter of 2". The torques are both ways, of course. Wear on the splines is inevitable and really worn splines can fail catastrophically causing the wheel to spin on the axle providing zero braking or acceleration ability. In the sub-catastrophic case, the splines merely cause the wheel to rotate by maybe 1 degree and clonk together as the car changes from acceleration to braking or vice versa.





The motion of the wheel in the above video is exaggerated to prove the point. To get this video I loosened off the wheel nut a bit. However, it illustrates the play that causes the spline clonk.

Here's some math...

Work out the torque and therefore, force on the splines and on the bolts of a standard wheel.

The specs and various test drives (by professionals) suggest that the maximum deceleration on the MG is 0.7G. I.E. 6.867 meters/sec/sec.

The wheel diameter is 0.6 meters

Decelerating at 0.7G means that the wheels are slowing down by a certain number of radians per second per second:

alpha = decel / (pi x wheel dia) x 2pi (in radians/sec/sec)

alpha = 6.867 / (3.141 x 0.6) x (2 x 3.141) = 22.89 radians/sec/sec

net torque = I x alpha where I is the moment of inertia of the wheel

To calculate the moment of inertia, which is 1/2 Mass x Radius2 for a uniform disk

we make the assumption that the entire mass of the car is essentially encapsulated in the wheels - I.E. if the car has a mass of 1000kg then that's the same as four wheels of 250kg each. (Obviously the weight distribution is not even and in particular the front wheels experience more force than the rear on braking.)

So we make a guess at the equivalent mass of the back wheels under braking at 250kg which makes the moment of inertia :

I = 250 x (0.6/2)2 = 11.25 kg m2

Hence the torque is:

net torque = 11.25 x 22.89 = 257.5 Nm (newton meters)

To counter this torque, the force at a distance from the center of the axis of rotation is:

force = torque/radius

For the case of the splines on a hub diameter of 2" (radius = 0.0254 meters)

force = 257.5 x 0.0254 = 10138 newtons (which is quite a lot!)

However, there are 75 splines so each one takes a maximum force of

10138 / 75 = 135 newtons each (= 30.4 pounds)

Let's say that we have two bolts on a 7.5" diameter (radius 0.095 meters), then the same force is only

force = 257.5 x 0.095 = 2703 newtons

and the force per bolt is

2703/2 = 1351 newtons (303 pounds)

From that SAE standards tables for bolt shear strength we can read that a 3/8" bolt is quite capable of handling 303 pounds.

Bolt DiameterShear strength
1/4"200 pounds
5/16"340 pounds
3/8"510 pounds
1/2"940 pounds
3/4"2260 pounds
1"4130 pounds


(if we used four 1/4" bolts then the force on each would be 151 pounds)

Here's the project :


There have been one or two negative comments about this project. Namely 'unsafe!' 'unbalanced wheel!' etc. The counter these arguments you have to note that the splines system is still in place and perfectly capable of stopping the car without my modification. All I am achieving is preventing the small amount of play on the splines. No safety issues at all. Indeed, the two bolts design is, but itself, capable of stopping the car: the calculations above are quite conclusive. Secondly the wheel is perfectly balanced - this was done by a pro at the local wheel shop. He actually had more problems with the two front whees where I have installed very small magnets to give my speed-o and CycleAnalyst computer a speed measurement. All the wheels are balanced and the car drives nicely.