F cars with cam belts

You guys need to have a beer together

In fact, I'll join and I'll even buy

 

Sadly I owned a Cadillac Catera - which was really an Opel with an Engish sourced V6 interference engine with belt driven cam. GM had a secret warrantee to 100km on them. If youbrougth your car in for service near that point, the dealer was supposed to replace and not tell the customer. The Sadly part is I dropped off the car on Sept 11th 2001. About two weeks after I got it back the belt failed and engine ate itself on the freeway.

I asked about the belt when I found out about the warrantee and found out they skipped the replace at the dealer because they got backed up after 9/11.

So it had just about 105k km. It turned out that I was the third person that week who had blown their Catera engine and they had no more spares in North America. I'd have to wait six weeks for replacement, and no loaner - I should have sued.

 

I had an Eagle Talon in the early 90s that was recalled for a timing belt issue. The belt broke before I could get the car into the dealership. I think it cost me about $1200 to fix which was a lot of money to me back then. This car had around 30k miles on it.

 

I live in Connecticut, by the way. Fantasy Island in a little south of me. (Sorry you LI guys. )

Things in tension break when the tension exceeds the tensile strength of the thing in tension. That is why it breaks. When it breaks depends on when that happens, and the strength of the thing at the time. I'll go back to my Honda example. The car was running fine on that grungy, contaminated, weakened belt. All the way from NH to CT, about 100 miles. But when shut off and a restart was attempted, it failed. That's when it broke, contamination was why it broke. That's no fantasy. That's a reality. However, to break when it did the torque applied to the belt when starting had to be greater that the torque that was being applied to the belt before, and at the instant, the engine was shut off. Otherwise the belt would have already failed, or the starting torque would not have sheared the teeth off the belt. That's logic and science applied to failure analysis. The starting torque must have been greater than the running torque, or it not have failed, unless (caveat) the belt suddenly weakened further while the car engine was off.

You can sight a zillion example of when a thing doesn't happen. They say nothing about when it does. To even start to have an idea of why something failed you have to know something about it at the time of failure and how it got that way. Say someone brings you one of those 308 with sodium filled valves. It has a broken timing belt. You pull the heads and find the heads broken off a few valves, holes in the pistons, etc. What happen first, broken valve or broken T belt? You really can't tell. Even if there is a sign of a belt problem you can conclude the belt broke first. Maybe the belt was weak, but maybe the valve still broke first, jamming up the works, and then the belt, being weak, broke. That may be a fantasy, but it's no bigger a fantasy that assuming the belt broke first. All you experience and you gut may be screaming BELT, BELT, BELT. Or since sodium valves are know to fail, maybe VALVE, VALVE, VALVE. But reality is that you don't know for sure which. All you know is $$$$.

Oh, look. It's getting cloudy outside. Guess that snows on the way.

 

This is getting hilarious. Anyway I just want to say something that is more of a criticism towards Ferrari than it is towards anyone here.

If the OEM cannot unequivocally explain the Failure Cause, then they are not doing their customers any service. If they do not know, then design it out. Actually, it should have been part of the early Design stage (DFMEA) where all know Failure Modes are assessed, a Severity rating assigned (this is a clear 8 or 9/10 due to safety), Occurance (I'd say 5/10 or 1/10000 field failures) and Detection (5 years/30k miles with little failures suggests an 8).

That's an alarming RPN if anyone here knows anything about this process. That should have driven a design change immediately. There is no way any OEM or Tier 1 supplier should be able to get away with something like this. Not these days. I'm in the biz and I have millions a parts in the field. I'm expected to explain every single Failure Mode, its Cause and fix below 200k miles/15 years or I'm out. To burden customers with a major service ($$$$) every 5 years is not only bad quality, but bordering on bad business imo.

 

Over the last 4 years, I have seen more chain problems then belt problems. BMW, Land Rover, Mercedes, all have chain stretch problems. Which result in engine lights, and noisy engines. Somtimes complet engine failure. Its brutal.

 

Careful with first hand knowledge and facts, you will be debated by some here.

 

I have a 2.4 JTDm. I assume every 5 years, though I do 14k/annum. It's going next year so I'm going to leave it but by year it's due, not sure about the mileage and don't much care. Won't be worth a jot when I move it on anyway.

 

Perhaps we could collect six examples and make a chart.

 

Why six? I'm sure the experts can tell us everything you need to know about chain drives with just one.

 

I don't know what Statler meant either but 6 is about the minimum sample size for statistical significance. Of course studies can be run with sample sizes as little as 2 but the error is significantly increased.

PS. My offer is still on for any Statistical Analysis of know failures......so long as we can move beyond the debated Causes.

 

I did too when we were servicing a big percentage of chain driven motors. No overhead cam drive is without issues. It is one of the major challenges of the design. Chains wear out and tensioners break, just not as often but when they do it is much more expensive. They are all trade offs. Gears are harder to service, noisy, heavy and expensive. Belts are reliable, light, inexpensive and have far less timing change over their lifespan. They work so well they have been in use in American V8s in racing for some time. They got smart and ditched chains a long time ago. The one downside to belts is life span is measured on the calendar, not just the odometer.

 

Let's bring back the push rod

 

Chevy is teaching us just how good they can be.

OHC does predate it, by quite a bit. Push rod motor has many advantages. They were not designed by dummies. The only motor that ever beat them in Can Am racing was turbocharged and cost 10 times as much.

 

I'm personally not interested in why belts fail. My interest lies in how long a car can run without a failure under typical use, i.e. based on field data. The limited data provided to me shows that 21 belts made it 7 years or more (one car was a 348 with only 1 belt). 33, 6 years or more. How many samples would you need to determine that the odds of a belt breaking for any reason, are less that 1 in 1000, or 1 in 10,000 belts after X years and/or y miles? Ferrari apparently thinks 3 years gives a 1 in xxxxxx chance of failure, but Ferrari doesn't tell you what xxxxx is. Regardless, there is always the odd case where a belt fails well before that period because it defective and slipped through QC or the tech installed it incorrectly.

I'd like to get back to data collection, but I don't see anyone else offering up numbers.

 

I think the winter problem has more to do with snow getting in between the belt and the idlers. Happened to my X1/9 back in 76. My first experience with changing a timing belt. The 70s had some cold winters. "Scientists" then were claiming we were heading into the next ice age.

 

We can simplify this by using existing data we gathered during the TR Differential Failure issue. One of our members conducted a survey of mileage driven which was used as a decent baseline of all TR's on the road today (expanded to 7000 cars with the same distribution). This is also very important data to have for the Belt Reliability Analysis BUT, if we base the Belt failure on the TR only, then maybe we can make progress.

I can add my data points.

1987 TR
I have 2 know belt replacements on my car
First at ~ 7500 miles and around 2001 (14 years) without failure
Second at 8750 miles and around 2013 (13 years) without failure (bearings busted up though but no failure)

The more we gather, the more accurate our model can be. The "no failure" samples will be used to censor the population. Failed parts will determine the Reliability distribution and predict future failures.

We should also decide whether we base the analysis on Mileage or Age. Either way is fine but I'd suggest Mileage since we have an underlying population distribution that I believe is still valid.
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Just one other note. If we decide to Analyze by Time, then we would have to go back to when the Cam Belt Engines started getting produced on high volume suspect models. That would be year 1. I think they are still being produced today so we can stretch the boundary out to today or 2016 models for argument sake.

It would be interesting to see how the distribution will look like but I highly doubt it would be Normal (i.e. steady rise, a peek and taper off). We would still need to know the following:

Belt/Bearing changes without failure from the time the car was new
Time when Belt/Bearing failed after the last belt change
Time without failure and without a Belt/Bearing change is actually the general population

Again, Analysis by Mileage or Time is fine.

 

I know of three 348's that went just over 9 years

 

Changed a friends timing belt on a DD non interference. Very cold out, rodent crawled up into the motor in between belt. Attempt to start car, no start. Towed to heated garage, took me a while to finally figure why it wouldn't start. Animal made belt jump time.

 

Went as in FAILED or DID NOT FAIL in 9 years?

 

So, I was watching Motor Week do their review of the 17 BMW M4 Competition. Then they go to Pad Goss and wouldn't you know it, he is doing a segment on timing belts/chains. With regard to T-belts he explains that they go bad based on time or miles and you should not ignore the manufactures recommendations. Then he starts to talk about how the T-belts fail. He says, to paraphrase, "See these pulleys. Notice how the one on the crank is smaller? Because it is smaller it transmits more torque to the belt. And when you try to start a car with an old belt that higher torque shears the teeth off the belt and the crankshaft just turns. This can result in damage to or even destroy the engine."

 

I could start you off with 6 Landover timing chain replacements in 3 months time (different cars of course) but all with the 5.0l engine (charged or not) and all under 100k km. 2012 and 2013 model years. Dealer would do nothing for the customer.

Bmw/mini..I keep the chains guides and sprockets in Stock.

Benz..chains and balance shafts in the six cylinders including the diesel.

 

Did not fail

 

I know of one TR in right now , a 1990 going for its first engine out

I personally have gone once to 8 yrs on my TR , and know a few others that have gone in under the 10yr mark. Not be issue they were worried, all of them ( myself included ) had oil leaks develop, therefore time to go in for the service.