[P2164]

Uhhhhhh..... Ok guys, back to the garage with a cup of coffee and put boston on the player, this is to much for me.

[mk e]

Originally Posted by P2164

Uhhhhhh..... Ok guys, back to the garage with a cup of coffee and put boston on the player, this is to much for me.

Timing belt threads never go well and always end up with a "yes you should.No you shouldn't" pissing match

[tazandjan]

Mark- Incidentally, that is exactly the process Ferrari used to get to the three year number, developed using accelerated testing on the 360. On the V12s, they did not do that except for the 575M when they introduced the new belts, so still recommend 3 years for the 456 and 550, which use exactly the same parts. Reason for that, no accelerated testing like they did on the 575M and no failure analysis performed. By example, though, the earlier V12 cars with the same parts should be good for five years.

[mk e]

Originally Posted by tazandjan

Mark- Incidentally, that is exactly the process Ferrari used to get to the three year number, developed using accelerated testing on the 360. On the V12s, they did not do that except for the 575M when they introduced the new belts, so still recommend 3 years for the 456 and 550, which use exactly the same parts. Reason for that, no accelerated testing like they did on the 575M and no failure analysis performed. By example, though, the earlier V12 cars with the same parts should be good for five years.

I work in medical devices so EVERYTHING gets time zero, artificial aging followed by real time age testing or it doesn't get sold and expiration/end of service dates mean it's expired means its no longer safe and selling or using it will get you sent to jail

Funny enough though reliability analysis has only been making it's way into the industry for the last 10 years or so and wasn't accepted by the FDA prior to that as far as I know......but has been used in the aircraft industry since the 1930 when it came out of Prat & Whitney (I think) so I assume was accepted by the FAA. Another 1 hand doesn't know what the other is doing. We always had to use capability instead of reliability which requires a lot more samples be tested to get a meaningful answer.

[Saint Bastage]

Originally Posted by mk e

Timing belt threads never go well and always end up with a "yes you should.No you shouldn't" pissing match

Exactly why I stayed out of it and so should everyone else.

[tazandjan]

Mark- We used similar methods for nukes to make sure they went off when they were supposed and never, never, went off when they were not supposed to. Same thing for the spacecraft I worked on.

In the olden days they used engineering rules of thumb and everybody made sure it was not going to be their piece of the aircraft that failed. That is why the B-52H is still flying 50 years after the last one was built. Overdesigned to engineering rules of thumb and tough as nails.

[Markp]

So forecasting survival is a little out of my norm (mostly I do regression analyses and a lot of correlation work) however it seems a little disingenuous to predict failure using only the failures and not the ones that survive. It would seem that you would want to have the MTTF in order to have an accurate idea of when an item may fail. After all, if you could find no example of a failure below 10k miles then it would be assumed that no likelihood of failure exists at that point. Also by not characterizing the failure mode, we have allowed foreign objects to potentially bias the result and not accurately reflect true failure due to a structural life of the component.

For instance, if I slashed 10 timing belts with a knife and they suddenly failed at less than 10k miles, don't you think that might bias the results of your analysis unfairly?

Belt failures are rare, if they were not, we'd see more of them.
Belt failures are due to a variety of reasons, some having nothing to do with the belt itself.
Belt failures will increase with structural demand on the belt.

Statistics isn't my area of expertise, but I am curious regarding the points above.

Mark

[wetpet]

Still waiting for someone to post that they have personally experienced a timing belt snap because of wear or age............neither brett nor mk e, nor brian have done so. still waiting to hear from artvonne. Again, not saying it doesn't happen. I know it does. Not even brian, who i'm sure has seen belts break from wear or age will post his experience. How many have you seen? 3,5,10,20? My point isn't that they don't break, it's that there seems to be little in the way of actual sightings of this unicorn. and 3yrs is paranoia.

[Brian]

I guess reading retention is not one of your skills.

[wetpet]

no retention needed, it is all just above this post. all i have to do is scroll, which i am quite good at.

[P2164]

Originally Posted by wetpet

no retention needed, it is all just above this post. all i have to do is scroll, which i am quite good at.

Good night to all!

[ECSofVirginia]

About 10yrs I've been working with these cars, I personally haven't worked on a car with a broken timing belt.

As Brian says, most come in for the accepted intervals and have the service performed before they hit the failure point.

The few cars I've seen near the point of self destruct due to timing belt system failure showed most of the fatigue in the tensioner bearing. A couple came in due to "noise" from the engine which was causing by tensioner bearings with no grease. A few came in simply because several years had passed with no record of servicing and the current owner wanted to use caution.

On the other hand, I've seen numerous VW/Audi 1.8T engines (I'm an enthusiast of these models) not survive the factory 105k timing belt service interval. We often change these belts/tensioners at 85k.

The Ferraris seem to develop enough oil leaks and other maintenance issues between belt changes that it rarely makes sense to avoid the service. It overlaps with other regular maintenance issues.

I'm not sure that you would get the kind of data you're asking for. Unlike a lot of cars, not enough people risk the service intervals to pool numbers. Not to mention different engine with different stresses and belt designs will have differing thresholds.

[mk e]

Originally Posted by Markp

So forecasting survival is a little out of my norm (mostly I do regression analyses and a lot of correlation work) however it seems a little disingenuous to predict failure using only the failures and not the ones that survive.
.............

Mark

Any analysis that is slapped together from available data vs a controlled study is going to have problems, and the one I slapped together from just the data in this thread is not all that good. The main issue is I don't know the age of the failures.....and the factory spec has 2 parts mileage or age, I suspect age breaks a lot more belts than mileage.

We can't count the belts that didn't fail in a reliability analysis because those are basically a test that was not completed. To use that data it would need to be something more like you're used to seeing, lots of sample and we could talk about a limit that was above the minimum mileage/time in the study because even though we have variable data (time or mileage) we don't know ANYTHING about the failure point other than we didn't reach it so I think we'd have to treat the data as attribute and call them PASS to the minimum value in the group.

The general rule is if it's a something like a dimension its a capability analysis generally with at least 30 samples, but better with 3 lots of 30. But if its a performance spec of some kind it's a reliability analysis using 3-20 samples. We use 3-6 samples for development work and 15-20 for verification work.

Anyway my only point in throwing the analysis together was that of all the chat about what everyone thinks about timing belts I've never seen anything that makes me question the factory specs on belts, oil, octane, or anything else. They are the ones with all the good data and I see no reason not to trust their analysis.

[Markp]

Originally Posted by mk e

Anyway my only point in throwing the analysis together was that of all the chat about what everyone thinks about timing belts I've never seen anything that makes me question the factory specs on belts, oil, octane, or anything else. They are the ones with all the good data and I see no reason not to trust their analysis.

Well I think we can agree on that... After all we're shooting in the dark to some extent here. I don't disagree with Ferrari's 3 year/30,000 mile major service at all. It's early but that's why it's called preventive maintenance, you are well out in front of the problem.

[P2164]

Originally Posted by ECSofVirginia

About 10yrs I've been working with these cars, I personally haven't worked on a car with a broken timing belt.

As Brian says, most come in for the accepted intervals and have the service performed before they hit the failure point.

The few cars I've seen near the point of self destruct due to timing belt system failure showed most of the fatigue in the tensioner bearing. A couple came in due to "noise" from the engine which was causing by tensioner bearings with no grease. A few came in simply because several years had passed with no record of servicing and the current owner wanted to use caution.

On the other hand, I've seen numerous VW/Audi 1.8T engines (I'm an enthusiast of these models) not survive the factory 105k timing belt service interval. We often change these belts/tensioners at 85k.

The Ferraris seem to develop enough oil leaks and other maintenance issues between belt changes that it rarely makes sense to avoid the service. It overlaps with other regular maintenance issues.

I'm not sure that you would get the kind of data you're asking for. Unlike a lot of cars, not enough people risk the service intervals to pool numbers. Not to mention different engine with different stresses and belt designs will have differing thresholds.

Very good information, thank you for your input.

[barcheta]

Originally Posted by Saint Bastage

Exactly why I stayed out of it and so should everyone else.

Amen...

[Artvonne]

Originally Posted by P2164

Artvonne you live 40 miles from me and I thought I knew everyone in NW Arkansas that owned a Ferrari or had one yet I don't think I know you. Are you new to the area?

So personally you have had many belts break due to age or mileage? Your the first so how many off hand?

No, I have not had any break, personally. I have however, seen many engine with blown belts. On the "other" forum, if you do a search and if you can get back into the old site, there are quite a number. Then there are the ones Butch was always running into, Dave Helms has seen more than he can probably recall, Brian Crall has seen a lot.

Me personally, I have seen engines, 308 engines, up close, with my own eyes,that blew a belt. Others, like Dave Helms for example, have sent me pics of stuff that they saw in their shop. I have some pics of a 348 that flung a belt at 23K miles, its not pretty. The most extreme was the engine I saw at T Rutlands back in 2001. It was FUBAR. Destroyed. Farther back, maybe 1983 or so, I recall seeing a 308 in Hemmings for cheap, needed repairs, so I called the guy. He said one of the belts let go and it ate a bunch of valves. Road & Track had a 308 article discussing the issue IIRC, its a known problem and is truly the cars Achilles Heel. I think one of the saddest examples was a guy who frequented the F List years ago. He let a friend borrow his 348. The guy said he started hearing a squealing noise and before he could pull over to shut it down, it quit. One of the idlers come apart and burned through the belt cover while it burned the belt. It bent all the valves and damaged both heads. He struggled for quite a while trying to scrounge up the parts to fix it, but in the end it just didnt make sense to dump $30K into a $30K car, and he dumped it for salvage cost.

Yeah I moved down here 3 years ago after losing half a mil and a house and am starting over, licking my wounds as it were.

[Artvonne]

Originally Posted by ECSofVirginia

About 10yrs I've been working with these cars, I personally haven't worked on a car with a broken timing belt.

As Brian says, most come in for the accepted intervals and have the service performed before they hit the failure point.

The few cars I've seen near the point of self destruct due to timing belt system failure showed most of the fatigue in the tensioner bearing. A couple came in due to "noise" from the engine which was causing by tensioner bearings with no grease. A few came in simply because several years had passed with no record of servicing and the current owner wanted to use caution.

On the other hand, I've seen numerous VW/Audi 1.8T engines (I'm an enthusiast of these models) not survive the factory 105k timing belt service interval. We often change these belts/tensioners at 85k.

The Ferraris seem to develop enough oil leaks and other maintenance issues between belt changes that it rarely makes sense to avoid the service. It overlaps with other regular maintenance issues.

I'm not sure that you would get the kind of data you're asking for. Unlike a lot of cars, not enough people risk the service intervals to pool numbers. Not to mention different engine with different stresses and belt designs will have differing thresholds.

This is a good post that makes a lot of the key points. Another thing to add, and Helm's, Crall, Butch, etc., often eluded to it or said it outright: many of the failures are not reported at all. Most buyers want cars with squeaky clean records with no major failures. A belt failure SCREAMS neglect.

[Brian]

Originally Posted by Artvonne

Yeah I moved down here 3 years ago after losing half a mil and a house and am starting over, licking my wounds as it were.

Sorry to hear.

In the words of some half American guy named Winston Churchill, "Never, Never, Never give up".

[mk e]

Originally Posted by Artvonne

Another thing to add, and Helm's, Crall, Butch, etc., often eluded to it or said it outright: many of the failures are not reported at all. Most buyers want cars with squeaky clean records with no major failures. A belt failure SCREAMS neglect.

I have no doubt that's the case.....