Appraisal Of A Broken Camchain
24 Dec 2018

A new chain has the following dimensions:-

total plates pin length roller dia roller width pitch outside plate width inside plate width
118 11.45 4.55 ± 0.02 4.9 ± 0.02 7.75 ± 0.01 1.0 1.33

Figure 1 shows the remains of a cam chain from a 900 Tre, (top) that had supposedly only travelled 8,000 km. Compare that with a new chain (bottom).

A new chain has 118 pins. The distance over 35 link pins is 272 mm on the new one, or a total of 917 mm for the full length. The distance over 35 link pins of the old chain is 275 mm, or a total of 927 mm for the full length. This is a massive 10 mm longer than a new one, or 1.1% elongation.

The cam chain of a low revving engine should be changed when it reaches 1% elongation - MAX.

If you look at Figure 3, you'll see the measurements of a chain that has travelled 25,000 km. The chain is compressed in the top half, measuring 287 mm over 37 pins. When extended, as shown in the bottom half, the distance increases to 288 mm. Over 118 pins, these would be 915.3 and 918.5, a total elongation of 3.2 mm, or 0.35%

That pretty much sums it up! The broken chain is showing the age of one that has travelled 3x the distance of the 25,000 km chain. We can be sure that it hasn't travelled 78,000 km, so either, it has spent a most of it's life near red line, or, it has worn excessively due to poor adjustment.

So which is it?

Tax's Chain & A New One >
Fig.1: Tax's Chain & A New One (click to enlarge)
Fig.2: A Broken Link Pieced Together
Figure 2 shows some broken pieces placed back together. There is evidence of 2 kinds of damage. The first is a clean break at the weakest part of the link plate. The second is a gouge mark in a plate, made by a sprocket tooth, once the plate had elongated enough to become misaligned with the sprocket teeth.

Figure 4 shows one other tell-tale sign. An inner plate has broken cleanly on one side (the crystaline structure is revealed by the rough surface at the break), but is smooth on the other. The smooth surface of the break, can only be explained by a crack that had developed over time. As it progressed, it moved, and the two surfaces polished each other. So this failure didn't happen in a split second. It probably took many minutes, and maybe hours, before catastrophe. So the failure was not caused by any other engine part, but the chain itself, or it's adjustment.

25k Chain Measurements Break Up Close
Fig.3: Measurements of a 25,000km chain (click to enlarge). Top is compressed, bottom is extended.
Fig.4: Slow Break

Figure 5 is another view of the same link shown in Figure 4. Figure 6 shows another link that has failed very quickly - each broken surface is rough.

Slow Break
Fast Break
Fig.5: Evidence of Wear
Fig.6: Fast Break

So why did it break?

The full history of this engine is not known. It replaced a previous one that had suffered damage due to a failed clutch body, or alternator gear coming lose (not sure which). It supposedly came out of a demonstrator bike, so it may have spent time on the track, and probably suffered at the hands of the motorcycling media. Even so, the wear seems immense for an engine of only 8000 km.

One piece of evidence may well have been provided by a new mobile chain guide. This is the one that is moved by the chain tensioner at the rear of the engine barrel. When fitting this to an engine that had travelled 25,000 km, the author found that it required filing to reduce it's width and allow free rotation.
There is the possibility, that if such a guide were originally fitted to the 8,000 km engine, it would not have allowed any movement of the tensioner because it was not free enough to rotate, and so could not have taken up the slack of a wearing chain.

New Mobile Guide
Fig.7: New Mobile Guide

Skip forward to 2015.

The remains of the broken engine have recently been acquired by the author with a view to rebuilding it. As a matter of interest, the old moving guide was installed to see if it was tight. It wasn't.
However, it WAS offset. So badly in fact that it dragged on one side of the cam chain cavity. It could not move.

New Mobile Guide Good Mobile Guide Guide Comparison
Fig.8: Original Mobile Guide
Fig.9: A Good But Used Mobile Guide
Fig.10: Two Good Guides & the Crook One
Figure 8 shows that the original guide is bent towards cylinder 1. It can also be seen where the chain has worn a new track into the guide.

Figure 9 shows how a good, but used guide should sit in the chain cavity. The centre line is shown in yellow in both figures.

Figure 10 shows two good guides (on the left), both having seen 25,000 km and more likely 50,000 km in a Tornado engine, and the guide of Figure 8, all mounted on a shaft of similar diameter to that used to hold them in normal use. The two on the left sit straight and parallel, while that on the right curves to the right. All three guides were milled from a sheet of plastic. The current guides are cast plastic for reasons that now become plain. When cast, no stresses remain in the part, whereas when milled from a blank, the process of machining can release internal stresses, if they are present, causing the part to bend. When warmed in use, these stresses can bend the part still further and this is what has happened to the original mobile guide installed in this engine. It is doubtful that this part was installed in this condition by the factory.

Of course, if the tensioner operation had been checked at the first and second services, as the factory requires in it's service manual, the problem would have been picked up before the chain destroyed itself.

The Demise of a Chain

From the beginning, the engine warmed the guide, allowing the internal stresses to be released causing it to deform. When it had deformed sufficiently to strike the side of the chain well, the tensioner spring was no longer able to move the guide as the chain lengthened due to normal wear. The chain then slackened increasingly, allowing it to flop about freely. As it did so it induced even more wear, which accelerated it's demise. Eventually, the wear on the link plates would have been enough to weaken them, until one began to crack. As the crack progressed at an ever increasing rate, the links began to misalign, and eventually the inner plate was picked up by a tooth of one of the sprockets. The resulting stress broke both plates. The chain then proceeded to get caught up inside the engine, but before this, the camshafts had ceased to rotate and half the valves had been bent by the pistons. Within the blink of an eye all came to a shuddering halt.

Further Evidence

At least one 900 Tre engine has had it's cam chain skip a tooth on one of its sprockets, causing it to run poorly due to incorrect valve timing. A stuck or bent mobile chain guide, or tensioner, would explain this fault as well.

Conclusion

The evidence suggests poor engine build standards and a poor service regime. Benelli is now building the engines in-house and have changed the guide construction, so this kind of failure should be a thing of the past provided that maintenance is kept to a high standard.
In the author's memory, no engine failures of this type have been recorded on the Benelliforum for at least five years.

Bent Tornado Valves