Arco gas?

Gordon McGrew wrote:

You guessed wrong. I've never EVER been a GM fan.

And there's nothing wrong with timing belts at all. What's insane is
using them on interference engines. Yes, lots of carmakers, including my
own manufacturer of choice, are doing it these days. But common use
still doesn't make it good engineering practice, it just means that
everyone is compromising design in order to maximize profit. Cut some
valve reliefs, for pete's sake! Or if that compromises combustion
chamber design too much, then bite the bullet and use a tensioned chain
drive. Heck, even GM got that right on the Northstar engines!

 

A) they DO stretch. B) they break.
Specific output, or "horsepower per liter," is a meaningless
measurement. Who gives a rat's rump how much an engine displaces, so
long as it doesn't result in a huge increase in packaging size or a drop
in efficiency? Adding displacement is *just* as valid a way to increase
performance as adding a turbo or variable valve timing. And, up to
reasonable limits, it is just as efficient. For example, did you ever
notice that the "archaic" Buick 3800 is one of the most efficient
engines in its power class?

Everything I've read seems to indicate that belt stretch isn't that far
different than chain stretch over time, given a slack-side tensioned
chain system with a good roller chain and a good dynamic tensioning
system. If slack chains were a real problem, you wouldn't find them on
so many high-end engines worldwide. Belts are a far cheaper option with
equal or SLIGHTLY better performance.... UNTIL the catastrophic failure.

 

Heh... That's one of the reasons that all my "old" cars are
Studebakers. Timing failure is almost a "never" event. And the majority
used phenolic camshaft gears...

JT

 

Of course you don't HAVE to rev it to 8K, but you can and I do
regularly.

I doubt that I will be driving it when it is 40 years old (I would be
in my late 70s by then). Will I still be driving it when it has 280 K
miles? I'll let you know. Others on this board have made it. Let me
know when you get yours to 8000 rpm.

Anything can be over-engineered to make it durable. Making it
light-weight, powerful, clean, efficient, small *and* durable is where
the real engineering is.

An optimist says the glass is half full.
A pessimist says the glass is half empty.
An engineer says the glass is twice as big as it needs to be.

 

Except that it does. You will have to increase your Buick V6 to 4.7L
to get the horsepower of a Honda 2.2L. How do you think that will
work in an S2000?

I don't know what "valid" means in this context, but there are some
real disadvantages to adding displacement such as size, weight and
efficiency.

Buick Lucerne Acura RL Accord Accord
3.8L 197hp 3.5L 290hp 2.4L 166hp 3.0L 244hp
3869 lbs. 4012 lbs. 3197 lbs. 3435
19/28 18/26 24/34 20/29

Just comparing the Buick to the RL tells the tale. The Acura weighs
140 pounds more, has 50% more horsepower driving all four wheels and
its mileage is within 1-2 mpg.

The V6 Accord weighs a little less than the Buick (how much of that is
due to the lighter engine?) has about 25% more horsepower and it gets
1 mpg more.

Of all these Hondas, the closest in horsepower to the Buick is the 4
cyl Accord. It has 16% less hp which is offset by the fact that the
engine probably weighs 300 pounds less than the 3.8L in the Buick. It
gets 5-6 mpg more than the Buick and will probably beat it in a drag
race. Pick the available manual transmission and it will go even
faster and burn less fuel.

If you want to compare with Toyota, check out the Avalon. 36% more
horsepower and 3 more mpg.

So what is so great about that Buick engine?

 

eh? you're still not getting it. high specific output means
interference. retaining high specific output means retaining timing -
and ability to cope with high rpms helps too. belts are a /great/ solution.

not if you want decent emissions or high compression with low octane.
the "secret" of the honda engine was getting rid of all the sharp edges,
recesses, angles and convolutions in the combustion chamber. these all
foul up the flame path requiring either lower compression or higher
octane to fix.

tension doesn't compensate for stretch. only if you have kind of
re-timing mechanism can you compensate.

 

wow, you don't get out much do you. go check out a japanese sport bike
some time.

honda motors exceed that longevity all the time. detroit rarely makes
it to half that.

 

Yes, I get it.

A) it only means its HARDER to avoid interference, and B) if you have to
have interference, time it with a chain and do the engineering to manage
the chain correctly.

They are great from the manufacturers point of view because it gets rid
of some hard design problems. And you can usually make them last beyond
the warranty period, so when the valves go through the piston tops the
manufacturer can blame the owner for "poor maintenance."

Its no "secret" at all. Hell, my 1993 Chrysler 3.5L v6 has 10.1
compression, is NOT an interference engine (though it uses a belt) and
runs on regular gas (though mid-grade is recommended.) The second
generation of that engine did turn into a slight interference design
(low probability of breaking a valve since the duration of the
interference event is very short, though non-zero). All it takes is a
more rigorous piston and chamber design process to minimize detonation
trigger points AND keep the required clearances. Is it "without
compromise?" No, of course not. But its perfectly rational for a street
engine where you know a certain percentage are going to break their
timing belts. At least if you want to have satisfied customers.

Dynamic tensioning makes sure that the timing error is only in one
direction, and also stabilizes the timing. IOW, wear in the links of the
chain causes the cam timing to retard and ONLY to retard if you have a
tensioner on the slack-side of the chain. But without dynamic tensioning
the interations between the masses and loads on the spinning
cam/crank/chain combination can cause the cam timing to wander between
an advance and a retard. Cam-in-block v8s use chains all the time
without any tensioning, but the chains are so short that the total error
(slop in each link times the number of links) is negligible no matter
what. Racers go ahead and either use gears or tensioners even for
cam-in-block engines, but its certainly shooting a squirrel with an
elephant gun for a street engine. Overhead cam V-type engines have a lot
more links so they can accumulate many times the cam timing error of a
cam-in-block engine, but with a good dynamic tensioning system its STILL
NEGLIGIBLE and stable enough for any street engine. Certainly negligible
enough to prevent valve-bashing, even if there's some slight performance
degradation near the end of life of the chain- but that's on the high
side of 300k miles compared to 100k for a belt!

 

jim beam wrote:

I don't mind engineering-based discussion, but this is the kind of
brand-o-philic bunk that makes me cringe just about every time a tech
thread wanders into a rec.autos.makers.BRAND group.

The idea that detroit "rarely makes it to half that" is no less
nonsensical than my hyperbole about rubber-band timng, which was a
response to more brand-o-philia. I mean hell, I've got 3 Detroiters over
250,000 miles and one over 450k. And a whole hell of a lot of Honda
engines don't make it past the low 100k mile range, too. Only a small
percentage of people care enough to come online and discuss their cars,
you know. And they tend to be the one that care for and get the most out
of their cars.

Yeah, I have a preferred brand. But I can tell you exactly what the
engineering weaknesses are as well as the strenghths. Its not "just
better than everything else."

 

The overall statistics disagree with you.

 

no. see below.

why? a belt does it cheaper, better, and doesn't require "management".

they're not "design" problems, they're deterioration problems. why
band-aid a fundamentally dumb design?

that's bullshit. detroit spends more money on "designed to fail" than
japan spends on "designed to work". plastic brake pistons? auto
transmissions with only 2 planet gears? cast cranks? dude, you really
have that ass-backwards.

detroit didn't know squat about how to run low octane at high
compression until the japanese showed them how. and 1993 is /how/ many
years after the event?

it's not "rigorous", it's simply taking the time to bother to identify
the problems. something detroit never did. all they did was run lower
compression or higher octane with rich mixtures and retarded ignition.
duh. once you know the solution, production costs exactly /zero/ more
than the older designs.

what's with this "street engine" excuse you keep using? why drive
de-tuned crap when you can drive a honda?

that depends on which side the adjuster is.

yes, to some extent.

right, but you don't have to /have/ the tensioner on that side.

which is precisely why belts are so superior - doesn't happen with them!

Cam-in-block v8s use chains all the time

it's reduced, but definitely not "negligible". i've seen short chain
stuff like that about 10 degrees out, and that's not negligible.

by that rationale, we'd still be using side valves and manually adjusted
ignition timing.

which is why chains on ohc's are such a bad idea.

why bother? a belt doesn't stretch, doesn't need "management" and costs
a good deal less.

ok, that's a fundamental misattribution. "valve bashing" only occurs if
the belt breaks. which it doesn't do if maintained on schedule. broken
chains "bash" valves just as effectively.

 

I'm not here to change your religion, but fact isn't on your side.

Were you even BORN in 1975? Or owned an American car from that era? Or
studied the engineering papers of the time? Rich mixtures, my ass.
Detroit spent more time mucking around with *LEAN* combustion and
advancing the timing to make a lean mixture burn than sanity would have
dictated, not rich/retarded!! It kinda worked, but my criticism of
lean-burn combustion is much the same as my criticism of belt timing: it
workswhen new, but it doesn't age well. At the same time, Honda was
dicking around with CVCC- and to their credit got it to work in small
low-output engines. But even Honda abandoned stratified-charge
combustion later on when the simple expedient of closed-loop digital
ignition and fuel management with fuel injection proved so much more
effective starting in about 1981.

Again, I'm not going to argue religion. The whole concept of a "street
engine" has nothing to do with build QUALITY (quite the opposite,
actually.) It refers to an engine that maintains its performance and
efficiency under the abuse of ownership by the masses who don't post to
rec.autos.tech, or even THINK about their engine more than once every 5
years.

This sentence makes no sense. YES, the tensioner IS always on the
slack-side. Even in (good, well-designed) belt-timed applications.
Better chain systems, like BMW, Mercedes, Cadillac Northstar, Chrysler
4.7, etc have tensioners on the slack side and sometimes vibration
snubbers on the tenisoned side.

Yes, it DOES. Belts stretch, or they wouldn't need tensioners.

Maybe on a 1975 small-block Chevy with a nylon cam gear, but I've dug
into mid 90s pushrod v8s with steel cam gears and chains with 200k
miles and less than a degree of cam timing error.

Straw man. An equal and opposite straw-man would be that since the
technology exists to build 800-horsepower engines for both NASCAR and
IRL, then all engines on the road, whether import or domestic, should
have 800 horespower available. As an engineer, one is obligated to think
about the needs of the particular application, not just blindly throw
technology at a device "because I can."

You keep saying that.... doesn't make it true.

Since when are there belt systems without tensioning and damping systems?

And fails catastrophically at 1/3 the life of a chain.

at's on the high side of 300k miles compared to 100k for a belt!

No, the argument was made earlier in THIS THREAD, that without the
"precision" of belt timing, the cam timing would drift enough to make
valves kiss pistons. Go back and look it up.

But at 300k miles instead of 90k.

 

which "facts" are these? so far, you've not cited one that is relevant.

eh? "lean" for detroit means lambda < 1.6.

that's not a lean burn problem - that's a detroit q.c. problem.

cvcc is not "stratified charge". modern honda motors, the one with two
plugs per cylinder - /those/ are stratified charge.

your context has been to mean a de-tuned motor - the detroit solution
with low specific output. honda run at /way/ higher specific output and
at least 3 times longer without blinking.

ok, you don't understand the concepts. you can have a tensioner on
either side - but it's cheaper on the slack side. if it's on the tight
side, it reduces retardation as the chain stretches.

that's an economic decision, not an engineering decision.

see above. you're unclear on the concepts.

no, belts do not stretch. "tensioners" on belts are to take up slack on
first fitting - then they're locked tight and simply act as rollers.
tensioners on chains /have/ to move as the chain slackens - and
continues to slacken with each passing mile.

absolutely no way - chain's not even that close when new.

eh? you said tensioners on short-chain are overkill for street engines.
they're not if you want to address valve timing drift. they're not
fitted because detroit is cheap-ass.

dude, have you ever worked on these things? the only time you'll ever
find a belt that's slack is if it wasn't fitted properly. the material
itself does not stretch.

eh? cite /one/ example of an active tensioner or damper on a belt.

no it doesn't. it's recommended change interval is 1/3, but that's for
99.99999% reliability. meanwhile the chain slacks off and your motor
burns excess gas as it loses timing. enough to pay for a new belt most
likely.

no dude, you're misreading - it's been about whether interference is
justified with a belt. you've never seen valve timing radical enough
for interference to be an issue with mere drift.

with a chain, you have 250k of crap valve timing!

 

Your Honda engine is FAR more more "de-tuned" from Honda's IRL engine
than a Chevy LS-3 or an SRT-8 Chrysler Hemi is "de-tuned" from a NASCAR
engine. You should write Honda and complain about your horribly de-tuned
engine! You DESERVE better! Honda is screwing you!

"De-tuning" as you call it, is what we call "design margin" in
engineering. Deal with it, its good for you even if you don't think so.
Every automaker practices this method, otherwise there would be no such
things as hot-rodders and rice-boys.

Apparently its you who don't understand the concepts.

Yeah, it would cause cam ADVANCE as the chain stretches. Guess which one
reduces peak performance more? Advancing- because it reduces horsepower
an return for a little more torque right off-idle.

In the first place, why the hell would the cost be ANY different?

In the second place, your second argument is simply not true at all. You
cannot put the primary dynamic tensioner/damper on the tension side,
because the normal operating load of the cams and any accessories driven
off the cams would tend to gradually compress the tensioner during
normal operation, allowing the slack side of the chain or belt to go
VERY slack indeed. The only thing you could do is use a fixed tensioner
on the taut side, which went out with the VW Rabbit engine in the 80s.
But you can put a dynamic tensioner/damper on the slack side, because
during normal operation it will take all induced tension OUT of the
slack side and keep the belt or chain as snug as possible. On the brief
transients of reversed belt tension when decelerating, the damper
prevents the tensioner from slacking off significantly.

All you are doing is proving that you've never degreed a cam on a
cam-in-block engine. True roller timing chain sets are easily that close.

Why address something that doesn't NEED addressing? Why tilt at windmills?

"Detroit" DOES use them.... when necessary (Cadillac Northstar, Ford
Modular, Chrysler 2.7, 3.7, and 4.7). And it doesn't when they're
unnecessary (GM LT and LS smallblocks, Chrysler 5.7 and 6.1 Hemis, all
of which produce more power than 4 Hondas)

Unfortunately, I do own a belt-timed engine and have owned others
without dynamic tensioning in the past.

That's fundamentally wrong. While its true that belt makers go to great
lengths to reduce stretch, it is not zero.

No, "dude," I'm not misreading. You're right, I've certainly never seen
valve bashing with mere drift, but that is EXACTLY what one of your
non-engineering-based friends has been arguing is a reason for belts.

BMW doesn't think so. Mercedes doesn't think so. Cadillac doesn't think
so. Chrysler doesn't think so. Ford doesn't think so. Toyota doesn't
think so.

 

that's stupid.

eh? you don't advance the timing, you compensate for the stretch that
would otherwise retard it.

eh? a tensioner on the tight side has to be a lot more robust to last,
that's why!

jeepers - i've seen some dumb-ass non-comprehension of engineering
concepts in my time, but that's spectacularly clueless. ever heard of a
"ratchet"???

tensioners on the slack side allow timing to retard. fact.

yes i have and you don't know what you're talking about. you're lucky
if factory is that close new. within 500 miles, it's drifted at least
that much again.

why did you bring it up then?

so which song do you want to sing here? - you started with no
tensioners. now suddenly you want to reference them as commonly used.
do you want it both ways?

it's fundamentally right steve. there is no "stretch" mechanism in the
belt any more than there is "stretch" in the walls of your tires, or
"creep" in your windshield. and any wear is trivial and pretty much
zero too. if you've experienced a "slack" belt, you've experienced the
consequences of one that wasn't fitted properly.

no, that was the misinterpretation /you/ used to try arguing against them.

they're simply making an engineering decision that because /their/
engines are low rev plodders, and driven by people that rarely keep them
more than 5 years, there will be no problems during the life of the
original owner. and they're dead right - none of those drivers will
know nor care! but that doesn't mean their chains /don't/ stretch, that
they /don't/ suffer timing drift, and that those motors /don't/ lose
efficiency because of it - because they all do! or are you going to
argue that day is night just to take a contrary position again?

if you do care to respond steve, please try sticking to facts this time.
thanks so much.

 

Steve wrote:
u who don't understand the concepts.

Ouch, In my earlier haste I completely mis-stated this. A stretching
chain or belt can never advance the *average* timing of the cam, only
retard it. Without damping, there can be excursions of advance and
retard, but since the cam on AVERAGE is being dragged along by the
crank, it will lag behind the crank (retard) to whatever extent it is
allowed to do so by the cam drive.

The drawback to putting a tensioner on the taut side is that as the cam
drive (be it chain or belt) stretches, the added slack will be
accumulated BETWEEN the crank and cam, maximizing the amount of retard
by the total stretch of the belt or chain. OTOH, if you tension on the
slack-side, the retard is minimized because all the extra length is
carried harmlessly on the slack side, and the only retard experienced is
the delta-length of JUST the length of belt or chain on the taut side at
any instant, which is necessarily less than 1/2 the total amount of stretch.

 

Is it? Quoting an interview on indycar.com....

"'For the 2007 season, a 3.5-liter engine (213.6 cubic inches of
displacement) will be used to provide longer engine life between
rebuilds and additional mid-range torque for the varied IndyCar Series
schedule – from street/road courses to short ovals to superspeedways.

“Less stress is good for reliability, and with the goal to push the
mileage targets higher still for 2007 this is the best solution for
Honda and the Indy Racing League,- Honda Performance Development
president Robert Clarke said. ”


So it would appear that Honda engineers recognize the benefit of what
you call "de-tuning," even in a race engine. But if you say its "stupid"
then I guess they're wrong. How silly of them for not consulting you....

True. But putting the tensioner on the taut side allows the timing to
retard MORE, because the delta-length for the WHOLE chain or belt
appears on the taut side, versus 1/2 or less the total delta-L if you
tension on the back-side. The change in length of the taut side ONLY is
what alters the cam timing, so it is CATEGORICALLY STUPID to put the
tensioner on that side. And no sensible manufacturer does it anyway, so
I don't know why you're bleating about it.

If you took time to notice that some of my engines listed above are OHC
and others are cam-in-block, you'd look a lot less foolish in claiming
I'm "singing a different song." I've said all along that some need
tensioners, some don't. Where needed, "Detroit" (and Germany, and Italy,
and Sweden, and Great Britain) use tensioners. Where not needed, they
don't.

Your whole argument seems to come down to "Honda uses belts, so chains
are stupid." Well, the high-end carmakers of the world don't agree with
that at all or BMWs and AMG Benz's and Lamborghinis would have belts.
They generally don't.

BMWs are "low-rev plodders." Intersting view of reality you've got
there, my friend. What do you think of Lamborghinis, then?

OK, here's a fact I'd like to know. What cam drive mechanism does the
Honda IRL V8 engine use? IF its a belt, I sure can't see it in any
pictures of the engine, but then maybe the engine is just shown without
the cam drive installed. I haven't been able to google up a definitive
answer...... but I'll bet you a donut its got a chain.

 

all "street engines" as you cal them are de-tuned. the point is,
detroit is much /more/ detuned. and there's a reason for it too.
cheapo crap manufacturing methods like cast cranks and 3 crank bearings
means you /have/ to de-tune.

bullshit. dude, you fucked up the first time. now you're /really/
talking through your ass. timing retards with the stretch between crank
and cam on the tight side. anything else is utterly totally irrelevant.

if you want to compensate for chain stretch, if you increase the length
of run by the same degree as the chain stretches. end of story.

stop bullshitting kiddo. you said first there's no need for tensioners,
then there is - you keep changing your story like you change your mind
on the theory you don't understand.

"generally"??? what ind of weak-ass excuse is that??? the "trend" is
away from belts because whiners like you bleat about having to change them.

fact: belts don't stretch, thus retain their timing.
fact: belts are reliable.
fact: belts don't need damping.
fact: belts induce much less rotational speed fluctuation and thus
induce less "noise" into sensor signals.

so why doesn't everybody use belts? because whiners like you bleat
about changing them!!!

bmw /are/ low rev plodders. honda s2000 is 9,500rpm red line from
factory. no bmw gets close. as for lamborghini, what's the life
expectancy then sport? you think the so-called "reliability" of a chain
is going to be a factor that's going to take /that/ motor up to 300k
miles? while you're checking your facts on that one, i also recommend
you look into their valve timing mechanisms and let us know whether they
bother to have a tight-side tensioner. report back when you find out.

see above. just like honda dropped wishbone suspension, they're
pandering to the whiners and using chains. nice little earner for the
bean counters if motors fail emissions because of timing drift and thus
need replacing sooner. but adoption doesn't change the facts - belts
are technically superior for the reasons stated above.