2004 sebring 2.4l 16 v engine 118000 kms on clock

Joe Pfeiffer wrote:
> Steve > writes:
>
>> KirkM wrote:
>>
>>> The change interval for the timing belt in the owners manual specifies
>>> 101,000 miles. I am planning to do it at 95,000.
>>> As you indicate, I will have them change the water pump at the same
>>> time. I had a 1990 Dodge Spirit with a 2.5L engine.
>>> The change interval for the timing belt was specified was for 50,000
>>> miles.
>> Which is patently bizarre. If the timing belt on a 1990 2.5 were to
>> break, the engine spins harmlessly to a stop with all the pistons
>> clearing all the valves. New belt and you're on your way. Conversely,
>> if the belt breaks on a 2.4, you're in for new valves (at least) and
>> maybe a complete overhaul. 2.5s weren't notorious belt-breakers,
>> either.
>
> Not sure why that makes it bizarre -- I look at those figures and see
> that either timing belt materials are improving or Chrysler is getting
> less conservative with its change interval.
>
> The change interval should be set on making sure it's changed before
> it breaks, not on how much damage happens if it does break.

You may be right, but it would not be unusual or necessarily wrong on
there being some influence on the chosen change interval by the severity
of the consequences of a timing belt breaking.

In fact, as an engineer and engineering manager, I used to be involved
in design FMEA's (Failure Modes and Effects Analysis) for parts being
designed for sale to Delphi (GM). The analysis was a joint or
coordinated effort by Delphi and the supplier (Ford/Visteon and Chrysler
have the same process as dictated by, at the time, QS9000 - I have no
idea what they use now).

When a potential failure mode for a given design and manufacturing
process of a part was identified, there were three categories that were
quantified and multiplied by each other to determine an RPN (Risk
Priority Number). One of the categories was (likelihood of "Occurence"
("O"), another was "Severity" (of the effects of the failure) ("S"), and
the third was "Detectability" ("D") (of a the failure when it occurred).

The number chosen for the two categories for the particular failure went
from (IIRC) 1 thru 10. For the likelihood of failure, the higher the
number, the greater the likelihood of that failure occurring. For
severity, the higher the number, the more severe the effect of such a
failure (1 being "not discernable", 9 and 10 being two different degrees
of safety and/or government regulation violation). (You can Google
"FMEA", and find tutorials and other info. on this kind of stuff - in
fact, because I've been away from it for about 7 years, I was a little
fuzzy on some of the info. and did just that to refresh my memory when
composing this post - yes - I cheated.)

The RPN (the product of the 3 numbers) established the priority of
tweaking the particular design or process failures until all RPN's were
below an acceptable threshold. Also - there were certain over-riding
rules. For example, a 9 or 10 in severity was an automatic "MUST FIX"
as long as, say, likelihood of occurrence was above 2 or 3 (I forget the
exact details, but you get the idea. For example, for any failure in a
brake or wiper motor application, as long as the there was some credible
likelihood of occurrence (a brake pad triggering a nuclear explosion
would not be a credible failure), that particular failure was put in the
"Must Fix" category.

It is a formalized and expensive process to be undertaken by a technical
committee (buzzword: Team) defining the type of thought process we all
go thru individually every day - for example - if you're on a family
vacation, compare what your actions (both immediately and delayed) might
or might not be if a knob fell off the radio on your car vs. if steam
started pouring out from under the hood. It goes back to severity of
consequences of taking or not taking action.

Anyway - I can see such a process resulting in two different change
intervals the manufacturer decides to specify on two different engines
with timing belts with identical statistical failure periods, the
difference being the severity of a timing belt failure depending on if
an engine is interference or not. I'm not saying a given manufacturer
would *necessarily* specify two different change intervals in the two
scenarios where the *only* difference was interference or not
interference, but it would not surprise me at all if that were typically
the case. I'd be just as surprised if it wasn't the case.

--
Bill Putney
(To reply by e-mail, replace the last letter of the alphabet in my
address with the letter 'x')

Bill Putney > writes:
>
> Anyway - I can see such a process resulting in two different change
> intervals the manufacturer decides to specify on two different engines
> with timing belts with identical statistical failure periods, the
> difference being the severity of a timing belt failure depending on if
> an engine is interference or not. I'm not saying a given manufacturer
> would *necessarily* specify two different change intervals in the two
> scenarios where the *only* difference was interference or not
> interference, but it would not surprise me at all if that were
> typically the case. I'd be just as surprised if it wasn't the case.

Thanks, that's a lot of interesting information. Remember, though,
that what Steve was referring to was a case where the interference
engine had a much longer change interval than the non-interference.

Joe Pfeiffer wrote:
> Bill Putney > writes:
>> Anyway - I can see such a process resulting in two different change
>> intervals the manufacturer decides to specify on two different engines
>> with timing belts with identical statistical failure periods, the
>> difference being the severity of a timing belt failure depending on if
>> an engine is interference or not. I'm not saying a given manufacturer
>> would *necessarily* specify two different change intervals in the two
>> scenarios where the *only* difference was interference or not
>> interference, but it would not surprise me at all if that were
>> typically the case. I'd be just as surprised if it wasn't the case.
>
> Thanks, that's a lot of interesting information. Remember, though,
> that what Steve was referring to was a case where the interference
> engine had a much longer change interval than the non-interference.

True. I guess it was this that maybe I was responding to: "The change
interval should be set on making sure it's changed before it breaks, not
on how much damage happens if it does break." That gets back to the
severity thing (independent of the examples, unfortunately for the case
I was making, that had just been discussed where the situation was
reversed).

But - yeah - as you said, the change intervals got longer over the years
- probably due to the improvements in the belt technology. I don't
think it was because the manufacturers were getting more conservative
with their numbers, at least not in all cases. If anything, in those
days where 50k and 60k were the typical intervals, the manufacturers
were possibly stretching things trying to keep the intervals as long as 60k.

I remember Subaru had to put out some notices (I guess they were also
called TSB's back then) on some late 80's engines (E82 engine I think)
to shorten the change interval back to what it had been a couple of
years earlier. Don't remember the exact numbers, but it was something
like they had extended it to 70 or 80k, and backed them off to 50 or 60k
due to some customers belts breaking.

Sure wish they'd figure out a way to gear drive the overhead cams, or at
least quit driving the water pumps off the timing chain engines like on
my 2.7.

--
Bill Putney
(To reply by e-mail, replace the last letter of the alphabet in my
address with the letter 'x')

Bill Putney > writes:

> Joe Pfeiffer wrote:
>> Bill Putney > writes:
>>> Anyway - I can see such a process resulting in two different change
>>> intervals the manufacturer decides to specify on two different engines
>>> with timing belts with identical statistical failure periods, the
>>> difference being the severity of a timing belt failure depending on if
>>> an engine is interference or not. I'm not saying a given manufacturer
>>> would *necessarily* specify two different change intervals in the two
>>> scenarios where the *only* difference was interference or not
>>> interference, but it would not surprise me at all if that were
>>> typically the case. I'd be just as surprised if it wasn't the case.
>>
>> Thanks, that's a lot of interesting information. Remember, though,
>> that what Steve was referring to was a case where the interference
>> engine had a much longer change interval than the non-interference.
>
> True. I guess it was this that maybe I was responding to: "The change
> interval should be set on making sure it's changed before it breaks,
> not on how much damage happens if it does break." That gets back to
> the severity thing (independent of the examples, unfortunately for the
> case I was making, that had just been discussed where the situation
> was reversed).

And I can see why that would be correct -- though in this case, losing
a timing belt on even a non-interference engine would frustrate an
owner enough that it would be a Really Bad Thing.

> But - yeah - as you said, the change intervals got longer over the
> years - probably due to the improvements in the belt technology. I
> don't think it was because the manufacturers were getting more
> conservative with their numbers, at least not in all cases. If
> anything, in those days where 50k and 60k were the typical intervals,
> the manufacturers were possibly stretching things trying to keep the
> intervals as long as 60k.
>
> I remember Subaru had to put out some notices (I guess they were also
> called TSB's back then) on some late 80's engines (E82 engine I think)
> to shorten the change interval back to what it had been a couple of
> years earlier. Don't remember the exact numbers, but it was something
> like they had extended it to 70 or 80k, and backed them off to 50 or
> 60k due to some customers belts breaking.
>
> Sure wish they'd figure out a way to gear drive the overhead cams, or
> at least quit driving the water pumps off the timing chain engines
> like on my 2.7.

Yeah -- the external belt drive water pumps are a lot easier to get
at...

Joe Pfeiffer wrote:
> Bill Putney > writes:
>
>> Joe Pfeiffer wrote:
>>> Bill Putney > writes:
>>>> Anyway - I can see such a process resulting in two different change
>>>> intervals the manufacturer decides to specify on two different engines
>>>> with timing belts with identical statistical failure periods, the
>>>> difference being the severity of a timing belt failure depending on if
>>>> an engine is interference or not. I'm not saying a given manufacturer
>>>> would *necessarily* specify two different change intervals in the two
>>>> scenarios where the *only* difference was interference or not
>>>> interference, but it would not surprise me at all if that were
>>>> typically the case. I'd be just as surprised if it wasn't the case.
>>> Thanks, that's a lot of interesting information. Remember, though,
>>> that what Steve was referring to was a case where the interference
>>> engine had a much longer change interval than the non-interference.
>> True. I guess it was this that maybe I was responding to: "The change
>> interval should be set on making sure it's changed before it breaks,
>> not on how much damage happens if it does break." That gets back to
>> the severity thing (independent of the examples, unfortunately for the
>> case I was making, that had just been discussed where the situation
>> was reversed).
>
> And I can see why that would be correct -- though in this case, losing
> a timing belt on even a non-interference engine would frustrate an
> owner enough that it would be a Really Bad Thing.

*BUT* - on a scale of 1 to 10 - where would you rank (1) "the vehicle is
temporarily disabled until you can pay $300-600 for a repair (including
the maintenance that was unwisely deferred)" vs. (2) "needs new
(possibly used or junk yard, or serious repair of existing engine)
engine or goes to scrap yard because the blue book is less than what the
engine repair/replacement costs"?

Now - you and I are resourceful and whichever the case, we get by
cheaper (and probably avoid the problem in the first place by doing the
belt pre-emptively or having a good idea of how far we can delay it
without too much risk. But for the typical consumer who knows nothing
and is at the mercy of whomever to put things back together (or sell
them a new car) for a huge pile of money - that's what you have to look at.

--
Bill Putney
(To reply by e-mail, replace the last letter of the alphabet in my
address with the letter 'x')

Bill Putney > writes:

> Joe Pfeiffer wrote:
>>
>> And I can see why that would be correct -- though in this case, losing
>> a timing belt on even a non-interference engine would frustrate an
>> owner enough that it would be a Really Bad Thing.
>
> *BUT* - on a scale of 1 to 10 - where would you rank (1) "the vehicle
> is temporarily disabled until you can pay $300-600 for a repair
> (including the maintenance that was unwisely deferred)" vs. (2) "needs
> new (possibly used or junk yard, or serious repair of existing engine)
> engine or goes to scrap yard because the blue book is less than what
> the engine repair/replacement costs"?
>
> Now - you and I are resourceful and whichever the case, we get by
> cheaper (and probably avoid the problem in the first place by doing
> the belt pre-emptively or having a good idea of how far we can delay
> it without too much risk. But for the typical consumer who knows
> nothing and is at the mercy of whomever to put things back together
> (or sell them a new car) for a huge pile of money - that's what you
> have to look at.

I think you're agreeing with me here...

Conversely, if
> the belt breaks on a 2.4, you're in for new valves (at least) and maybe a
> complete overhaul. 2.5s weren't notorious belt-breakers, either.
>

Not true. The 2.4l DOHC built for the 4dr. and convertibles will not bend
the valves. While the 2.4l SOHC built for the two door coupes will bend the
valves.

Bryan

Joe Pfeiffer wrote:
> Steve > writes:
>
>> KirkM wrote:
>>
>>> The change interval for the timing belt in the owners manual specifies
>>> 101,000 miles. I am planning to do it at 95,000.
>>> As you indicate, I will have them change the water pump at the same
>>> time. I had a 1990 Dodge Spirit with a 2.5L engine.
>>> The change interval for the timing belt was specified was for 50,000
>>> miles.
>> Which is patently bizarre. If the timing belt on a 1990 2.5 were to
>> break, the engine spins harmlessly to a stop with all the pistons
>> clearing all the valves. New belt and you're on your way. Conversely,
>> if the belt breaks on a 2.4, you're in for new valves (at least) and
>> maybe a complete overhaul. 2.5s weren't notorious belt-breakers,
>> either.
>
> Not sure why that makes it bizarre -- I look at those figures and see
> that either timing belt materials are improving or Chrysler is getting
> less conservative with its change interval
>
> The change interval should be set on making sure it's changed before
> it breaks, not on how much damage happens if it does break.

Uh.... I only half agree. Yes, the change interval should be set to
insure that only a small percentage of belts ever break, BUT the amount
of damage that might happen should DEFINITELY skew the change interval
shorter for high-risk engines. If I were setting the recommendation and
putting the company's warranty at risk, I'd pad a high-risk engine by at
least 20,000 miles compared to a free-wheeling engine with the same
probability of breaking the belt.

Steve > writes:

> Joe Pfeiffer wrote:
>>
>> The change interval should be set on making sure it's changed before
>> it breaks, not on how much damage happens if it does break.
>
> Uh.... I only half agree. Yes, the change interval should be set to
> insure that only a small percentage of belts ever break, BUT the
> amount of damage that might happen should DEFINITELY skew the change
> interval shorter for high-risk engines. If I were setting the
> recommendation and putting the company's warranty at risk, I'd pad a
> high-risk engine by at least 20,000 miles compared to a free-wheeling
> engine with the same probability of breaking the belt.

You and Bill both make good points about tying the change interval to
the damage done. I guess where I'm coming from is that maintenance
intervals don't cost the company money, and anything that might render
a vehicle undriveable if it breaks (even if it doesn't cause any
actual damage) ought to have a change interval such that the
probability of failure is really, really low. Today we expect our
cars to be reliable enough that any story that includes "...so there I
was, stuck on the side of the road.." is likely to mean one less
customer for the company for the rest of that customer's life.

Joe Pfeiffer wrote:
> Steve > writes:
>
>> Joe Pfeiffer wrote:
>>> The change interval should be set on making sure it's changed before
>>> it breaks, not on how much damage happens if it does break.
>> Uh.... I only half agree. Yes, the change interval should be set to
>> insure that only a small percentage of belts ever break, BUT the
>> amount of damage that might happen should DEFINITELY skew the change
>> interval shorter for high-risk engines. If I were setting the
>> recommendation and putting the company's warranty at risk, I'd pad a
>> high-risk engine by at least 20,000 miles compared to a free-wheeling
>> engine with the same probability of breaking the belt.
>
> You and Bill both make good points about tying the change interval to
> the damage done. I guess where I'm coming from is that maintenance
> intervals don't cost the company money, and anything that might render
> a vehicle undriveable if it breaks (even if it doesn't cause any
> actual damage) ought to have a change interval such that the
> probability of failure is really, really low. Today we expect our
> cars to be reliable enough that any story that includes "...so there I
> was, stuck on the side of the road.." is likely to mean one less
> customer for the company for the rest of that customer's life.

We both know what bell curves look like. We both know that there is no
such thing as zero percent failures with something like a rubber belt
doing the job that timing belts do. When you talk in terms of parts per
million, while the bad p.r. from a broken belt takes its toll on
customer loyalty, there also is an impact on that loyalty on how often
maintenance that costs several hundred dollars and requires alternate
transportation is required. Then there's an additional cost (on
customer loyalty) of an engine that has to be replaced. All of these,
though talked about in fractions of a percent or parts per million, have
to be balanced out to compete with other auto makers who could be doing
an incrementally better or worse job in choosing those balances.

--
Bill Putney
(To reply by e-mail, replace the last letter of the alphabet in my
address with the letter 'x')