Right, I've pulled the pin . . lets see what happens next
Commissioning a Re-Built ‘Airhead’ Gearbox
There has been some debate recently about the methods adopted when re-building an Airhead gearbox, in particular around whether one can rely on a shimming plate and shims or whether they have to be built by ‘feel’.
The argument against relying on a shimming plate concerns stiffness in the action when the built-up box is cool, this being seen as the box being over-shimmed and therefore too tight.
The concern was that this would pre-load the ball races leading to premature wear and failure.
As one who has had apparent success with a shimming plate and has done some research into the subject I thought it prudent to record my thoughts and findings from practical experience.
What is ‘Shimming’ ??:
Shimming is the reduction of free play in a shaft by the insertion of thin washers (shims) of known thickness such that the ‘float’ of that shaft in its mounting is reduced to the recommended tolerance.
In an Airhead gearbox, we aim to shim the three shafts to a clearance of 0.05mm.
To do this we install the three shafts in the main gearbox case and then check the clearance of the rear bearings in their housings in the rear cover.
This involves installing a Shimming Plate over the rear bearings on the rear face of the gearbox case. I fit a new gasket to the face, then bolt-down the plate leaving part of the three bearings standing proud above the plate surface.
One then has to use a Depth Micrometer to measure the height of each bearing above the plate in millimetres – taking a number of readings around each bearing.
Secondly, the depth of each bearing housing is measured in millimetres from the inner surface of the rear cover – again, taking multiple readings around the periphery of each housing.
Thirdly, measure the thickness of the Shimming Plate, in millimetres, using an Outside Micrometer.
Fourthly, measure the thickness of the oil baffle that sits at the rear of the Intermediate Shaft with a Micrometer – it is effectively one of the shims.
Now there is a decision to be made as the readings taken with the Depth Micrometer will not be consistent around each bearing or housing. I aim for the tight end of average – more height of bearing, less depth in housing - as this will produce thinner shims. The differences are usually very small.
Finally, calculate the shim thickness for each shaft in millimetres thus:
(Depth in rear cover – 0.05mm) – (Height of bearing above plate + thickness of plate) = Shim
It is a good idea to measure the selected shims – as a stack – with the Micrometer to ensure the combined thickness is correct – there is a surprising variation in the standard shim thickness.
Completing the build:
Having selected the appropriate shims for each shaft, they are placed above the three bearings and the heated rear cover is tapped down over them – the bearings are an ‘interference fit’ in the cover, hence the need for the cover to be at around 160 deg C.
With the hot cover in place the nine machine screws are quickly inserted and pinched-up and then the input shaft is given a tap to seat the rear bearing in the cover.
When the box is cool the cover screws are torqued down to 9 lbs/ft. before the output flange is bolted up to a torque of 161.5 lbs/ft.
When everything is finished I find that the box is stiff to turn by hand. The output flange usually turns but the input shaft needs an old clutch plate attached before it can be easily turned.
Warming the gearbox case with a hot air gun results in the box freeing-up.
Is this a problem ??? Some think it is, I don’t think it is and this is why . . . .
Loading the bearings:
The worry appears to be that having the box tight will put axial load on the ball races that are designed for radial load only. True, but it is a minimal load for a brief initial period and is far less than the on-going load that the bearings are subjected to in normal use.
Airhead boxes contain two types of gear pinion – straight cut and helical.
The front pinions of each of the three shafts are Helical gears that transfer the drive from the input to intermediate to output shafts when the box is in 5th gear (top).
Helical gears are chosen because of their smooth power transition from one to another. The straight cut gears are reserved for the intermediate gears that are not usually used for long distance cruising. The problem with Helical gears is that they produce Side Thrust.
This side thrust can be so severe in the more powerful Boxer engines that it can draw the output shaft back through the front bearing, putting extreme load on the rear bearing and the rear casing. This can cause premature failure of the output shaft bearings and to minimise this effect BMW fit circlips to the output shaft in front of the forward bearing.
Going on and off the power in top gear will therefore shunt all three shafts forward and back in normal use. This is very useful when we first come to use our rebuilt boxes.
What are we shimming ??
When one thinks of shimming we usually think of a shaft that has some movement between bearings – such as the ‘bottom end’ of a simple single cylinder motor. In this case it is a simple matter to bolt it together then use a Dial Gauge on the end of one shaft to measure the total movement.
A simple calculation determines the thickness of shims required on either side of the bottom-end to produce the desired running clearance.
Airhead boxes are not like that. The bearings are pressed onto the shafts (Input shaft roller bearing excluded) and the bearings are then shrunk into the cases by heating the cases. So what are we shimming as there isn’t any apparent running clearance???
The answer is that under load, when the motor is putting 50 BHP or thereabouts through the box, the bearings can move in the case and the shafts can move in the bearings.
But you won’t be able to do that on the bench by hand !
We limit the travel of each shaft to 0.05mm in order to preserve the gearchange clearances as the shafts move forward and back in the case as well as limiting side load on the bearings.
These clearances are specified by Getrag ZWN – who make these gearboxes for BMW.
What happens during the build ??
We heat the main case and drop the three shafts in – their front bearings hopefully seating fully in the front housings. A tap with a rubber mallet will settle them as a precaution.
However, when we come to fit the rear cover different forces come into play. The box is usually sitting with the rear facing up so that the oil baffle and shims can be positioned on top of the bearings. As a further precaution I use a small smear of grease to hold the shims to the bearings as they can be easily dislodged.
The rear cover is heated to around 160 deg C and then tapped down over the bearings. The case screws are then inserted and the input shaft tapped back to seat the rear bearing.
Whilst doing this, the hot rear cover is cooling, heating the main case as it does so. The case expands a bit in the process.
As the rear cover cools it grips the bearings, then, as the cover and case cool to room temperature they contract slightly – pulling the bearings towards the centre of the box and putting them under side load.
The shafts are not free to align themselves with their housings. In addition, the stack of shims will not necessarily be compressed together at this stage.
It is only when the box is put under load that the shafts settle down. It is, however practically impossible to check the running clearance of the shafts – hence the need for accurate measurement during the build.
It has been suggested that the rear cases can warp, meaning that individual measurements of the bearing housings are meaningless. All I can say is that I have never noticed a cover that wasn’t flat when placed on a good flat surface. If it was bowed upwards it wouldn’t have shown up, but the resulting clearance would have been greater than required and wouldn’t preload the bearings once the box had settled down.
What do the manufacturers do ??
Anton Largiader has published the transcript of a report on a visit to the Getrag-ZWN factory by German Motorcycle Magazine MO in February 1996. They were particularly interested in the ‘Oilhead’ boxes but the article says that the processes described are also applicable to the ‘Airhead’ boxes.
After discussions on design & testing they came to production and testing. MO spoke to Getrag development engineers Klaus Sommer and Eberhard Schaetzle.
SCHAETZLE: Every fully assembled transmission goes onto a test stand and is thoroughly tested with hot test-oil. Afterwards the hot oil is extracted and along with it the first wearing-in material. The transmission does not need to be broken in. It can be fully loaded from the very beginning.
So new Getrag boxes have been warmed-up and run under some load as part of the initial running-in process. This will have settled the shafts in the cases.
My own experience:
Having started rebuilding Airhead boxes by working on two of mine I found that they settled down quickly and were free after a short time on the road. Since then I have rebuilt many more and have not had any problems reported to me. Quite the reverse, people seem very pleased with the results.
I conclude that one needs to go one step further than the shimming process in order to settle the shafts into the box as it is not possible to seat the Intermediate and Output shafts in the way that the Input shaft can be seated and this requires the box to be put under load.
Bob.
.
. 
, no complaints from me on both of mine - 80 and 100 GS - still running really nice!
