Race Engine Forum UK

Why?
Reliability, pure and simple.

I don't think that quoting Gregs car is a good example, it is solely a drag car and will be low mileage, it will have the cambelt regularly checked, maybe even changed after every event or at least regularly, far less than the Lancia recommended interval anyhow which only the foolhardy would adhere to.

Ok, I did some research, here is what I came up with:

Belts generally don't fail at high Rpm, they fail at low Rpm.

At above idle speeds things are nice and smooth and there is less of a load on it, on start up and idle the jerkyness of the cams will snap it or it can jump teeth.

8v integrale belt failures are rare, 16v ones are far more common, this could be down to the same width belt having to open twice as many valves. I think that the 16v belt must be set up very very carefully, fit high lift cams and you are sailing ever closer to the edge.
The recommended tension of a 16v belt is less than that of an 8v.


What contributes to failure?

1. Exceeding belt change interval.
2. Balancer belt breaking and taking out cambelt.
3. Badly fitted belt, this includes;

A. Not replacing the tensioner bearings at the same time.
B. Incorrect tension.
C. Swinging the tensioner anticlockwise instead of clockwise before applying tension.

I know that fitting a wider belt won't eradicate the effects of a sloppy mechanic, but it will lend a greater safety margin. The 16v integrale has a bad name for belt snapping, mostly undeservedly I would imagine,I would dearly love to do a PM on any such failed engines!

Martin.

I don't think that quoting Gregs car is a good example, it is solely a drag car and will be low mileage, it will have the cambelt regularly checked, maybe even changed after every event or at least regularly, far less than the Lancia recommended interval anyhow which only the foolhardy would adhere to. Ok, I did some research, here is what I came up with: Belts generally don't fail at high Rpm, they fail at low Rpm. At above idle speeds things are nice and smooth and there is less of a load on it, on start up and idle the jerkyness of the cams will snap it or it can jump teeth.

Martin, hi

I know you are posting in good faith and trying to be helpful but you are wrong. The life cycle of a belt is determined much in the same way as Sn predictions are made in metals and that prediction comes from an analysis of the effective tension in the belt at various belt loads (turning torques) and engine speeds. There is no one rule. Every engine has different characteristics and the belt peak load might be 2000N at 3000rpm on one engine and 1000N at 5000 on another. Superimpose a peak load on resonance and that belt's going to fail. The stiffer the belt system (tighter) the higher up the rev band the peak loading will be. The loading depends on the whole reciprocating and rotating architecture: inertia of cams and camwheels, cam profile, cam phasing, spring rate, crank vibrations etc etc.

The Integrale belt is a very good design and is one of those based on current cambelt design philosophy that the land and teeth should share equal load. Doing that helps to inhibit the most common form of belt failure which is a fatique fracture at the root of one or more teeth under cyclical load on the nylon case of the tooth form.

Greg's engine is not the only example I could cite but it is a good example because that engine has extremely high effective tension on standing starts and gear changes, an aggressive input from the engine architecture and the highest possible cyclic rate of belt loading. If that belt was at risk from a load/resonance input it would have failed & fast. Albeit that the loading is applied for a relatively short time on each run, these high stress runs are made over and over on the same day and practice too and it is likely that the load the belt sees is equivalent to a very high mileage on any other car.

GC

Hello Guy,

Thanks for your explanation, it fills in another chapter in my understanding of cambelts. If you had the time could you possibly tell me how you would set up a belt on a 16v integrale to the correct tension?

Thanks,
Martin.

P.S. Greg changes his cambelt every 6 miles.