Back
3. The bottom end.

The bottom end of the motor (crankcase, crankshaft & bearings and induction system) forms the basis of the engine, and a look at most commercial engines shows that this part is largely underestimated.
In team race however the necessity of improved constructions is clear, because motor are simply falling apart there.

3.1. Crankcase and bearings

One of the most serious problems in conventional motors with ball bearings in an aluminium case, is that because of the different expansion coefficients of thermal expansion between aluminium and steel, the fit of the outer bearing ring in the case gets looser at rising temperature. The order of magnitude of this effect is 0.015 mm every 75 °C of temperature rise. (see fig. 1a and 1b)

Fig. 1 a.

Thermal expansion aluminium housing vs. steel outer bearing ring.
Aluminium housing made 0.01 mm smaller than bearing ).D. at 20 °C.
(Wall thickness steel ring = 1.3 mm, wall thickness aluminium housing = 4 mm)
At 20 °C, the ballrace outer ring is compressed 0.005 mm.

--- free expansion alumium housing
-.- free expansion steel bearing ring
___ resulting curve bearing outside diameter and diameter aluminium housing (bearing in housing)

Temperature below 65 °C, then bearing has an interference fit.
Temperature over 65 °C, bearing comes loose.

Fig. 1b

As fig. 1a, but aluminium house made 0.02 mm smaller than bearing O.D. at 20 °C.
At 20 °C, the ballrace outerring is applied, the relative stiffness of the outer ball race ring is increased and the stiffness of the aluminium housing is decreased.
At 20 °C, the ball race outer ring is compressed 0,005 mm, like in fig. 1a, but temperature range is extended up to 120 °C

--- free expansion alumium housing
-.- free expansion steel bearing ring
___ resulting curve bearing outside diameter and diameter aluminium housing (bearing in housing)
~~~resulting curve if outer bearing ring is stiffened with an extra ring, 1.5 mm wall thickness. (curves applies to O.D. extra steel ring)

The constant "hammering" and a turning moment will thus make the bearing outer ring creep around, ruining the housing. Before this happens, the strong influence of mechanical load on engine temperature will produce the very well known popping run followed by a sudden seizure.

Many TR-flyers recognized this problem and quite a few solutions were seen.
Jim Plaunt, for instance, used three set screws to prevent the rear race from moving in his 1976 Rossi RV.
A same type of solution was seen in one of Maslov's engines. He used a 0,6 mm diam. key between bearing and house.
Both solutions affect the roundness and/or stiffness of the outer ring and must be considered as not ideal.( See fig. 2a and 2b).
A far better solution is to Loctite the bearing at 100°C into the crankcase (use Loctite bearing fit with activator and do it quickly!). This method was quite widely used by us in MK I and MK II Bugls, and usually lengthened the life of an engine for a while. The only problem with this method comes if you must change bearings.

Some Russian engines showed steel rings shrunk and maybe also glued into the crankcase. This system at least solves the above mentioned problem of replacement of bearings, but the combination of aluminium case and a shrunk-in steel ring has a resulting thermal expansion coefficient somewhere in between aluminium and steel, so the problem is only half solved ( see fig 2c and 1b).

FIG.2 SYSTEMS FOR PREVENTING BEARING ROTATION


fig 2a
Maslov's solution


fig 2b
Plaunt's solution


fig 2c
steel in crankcase

One of the best solutions to the problem is from Henry Nelson. He uses simply a very strong interference fit (~0.025 mm), which will stand temperatures up to ~150 °C before the bearing comes loose. Normally crankcase temperatures will be under 100°C.
A few conditions have to be fulfilled.
Firstly the material of the crankcase has to be able to stand a relative high stress at room temperature.
Secondly, bearings of a special high clearance type are needed to compensate for the compression of the outer bearing ring at lower temperatures. At higher temperatures, the bearings will run with quite some play, both radial an axial, which won't necessarily be a big disadvantage, up to a point.
However, replacing bearings at room temperature will give slightly decreasing interference fits, and asks for special tools. Replacing them at temperatures above 110°C - 150°C, which is necessary to do it without applying force, will ruin bearings. (advice: let Henry do it: he knows how it's got to be done).

The solution that we think is best was firstly shown by Krasnorutsky: the all steel front housing.
We tried them first in MK I Bugls and the effects were quite amazing compared with the (worn out) original aluminium front housing: no sudden and unexpected seizures at healthy sounding settings anymore! Fig. 3 shows the principles of the FMV front housing, a screwed-in type, developed after a good look at Krasnorutsky's 1977 engine.
The front housing is Loctited (Loc-tite hot retaining compound + activator) at 50°C into the crankcase, which means that above 50°C no change in interference fit of the main ball-race exists.
The interference fit of the front race is not affected by rising temperature of the engine. The fit we use is about 0.001 - 0.003 mm of interference for the rear race and 0.000 - 0.001 mm for the front race. More interference means loss of play in the bearing. This means they just can be put in with a (strong) human thumb or soft hammering with a piece of wood.

fig 3


The FMV steel front housing

Getting the ballrace housings to size is an extremely accurate job, not helped by our deformable, normal clearance, bearings. With high clearance bearings more interference could be used and the relative accuracy of the house becomes less critical, but we still can't get these bearings in our sizes (8x19x6 and 7x14x3,5) without ordering large numbers.

Next