Re: [alfa] Re: single-row timing chains

[bearbvd@xxxxxxxxxxxxxx (Greg Hermann)]

At 5:09 PM 11/8/04, Joe Elliott wrote:
>At 7:33 PM +0000 11/8/04, alfa-digest wrote:
>
>I wouldn't be so sure.  For sure, flinging the mass of that second
>row of chain around at engine speed requires a not-insignificant
>amount of power (although I'd have guessed maybe 2-3hp, I'm too lazy
>to do the math).  Then there's the friction between the various metal
>parts, as well as friction with the oil.  If you go to smaller
>pulleys, then you're saving rotating mass as well, which doesn't
>necessarily take any power to drive at constant speed, but does
>require power to accelerate from one speed to another, right?

With a typical timing chain, it is not operating anywhere NEAR the
'strength' limit of the chain. (Defining 'strength' as the fatigue life of
the roller link plates (which are the weaker ones, as the hole through them
is bigger).

Roller chain power capacity follows a multi-faceted curve. In a straight ,
non-moving test of tensile strength, a pin will usually break. Power at low
speeds is limited by fatigue strength of the roller links. Power at higher
speeds is limited by 'roller impact'---the chain rises and falls a bit as
it engages successive teeth on the sprockets, and as speed increases, the
impact from this action will BREAK rollers. Power at still higher speeds is
limited by galling between the pins and bushings. Centrifugal force becomes
a pretty significant factor in chain loading at higher speeds.

Going to smaller sprockets can indeed save rotating mass, but it will ALSO
increase roller impact, since the chain rises and falls more as it engages
a proocket with fewer teeth. Thus smaller sprockets could have a disastrous
effect on chain life and reliability.

The big factors in chain wear on an OHC chain drive are pulsations in
loading due to the uneven torque demands of driving the camshaft(s) and,
most of all, 'whip' in the long straight runs of chain. Good chain guide
and tensioner design is ESSENTIAL to good chain life and reliability.

The biggest single improvement in chain reliability/durability/power
capacity over the past few decades has come from the use of 'cold- headed'
( cold forged, if you will) rollers--as opposed to the older practice of
curling the rollers out of a piece of strip stock (which leaves a crossways
seam in the roller where the two edges of the curled strip stock butt
together). Wear tests on some chain with headed rollers demonstrated that
the curve for roller impact failure was raised so high that a test could
never reach it--because the transition from fatigue limited power to
galling limited power happened BELOW the roller impact curve when headed
rollers were used.

There are other improvements that have happened with chains that live in
nasty environments--such as motorcycle final drive chains. Reynold was ,
IIRC, the first to introduce the 'stepped' bushing. Basically this involved
putting a small step on the OD of the bushings at both sides---so that the
roller link plates were pressed onto the bushings against a positive stop.
Fatigue testing PROVED that step bushing chain was NOT as strong under
cyclical loading as conventional chain--BUT--in a motorcycle final drive
environment, with things like suspension deflection (NOT meaning suspension
travel) entering into the picture, it seemed that step-bushing chain kept
its assembly alignment/clearances. This change was proven CONCLUSIVELY to
be of no benefit, in fact a detriment, to the strength/life of chain in
typical industrial applications --where good lubrication and reliable
alignment are ever-present (barring interference by Bubbas) !!

More recently, things such as O-ring sealed chain have hit the market. (an
O-ring is inserted between the pin and roller link plates on each side so
as to keep factory applied lube IN and DIRT out of the pin--bushing bearing
area. This stuff is strictly for use in NASTY--such as motorcycle final
drive type-- operating environments. Would only be a detriment in a good,
well oiled, constantly aligned industrial environment.

 There would also be a point to it in 'clean' environments--say food
processing machinery. This sort of use gives away strength by using
stainless steel link plates, and no lube except inside the O-ring seals.
Can be steam cleaned without destroying the chain.

As far as improved strength due to metalurgy or different manufacturing
processes, NOT much change at all, except that perhaps the metalurgy of,
say for instance, Mexican made chain has likely improved VASTLY. Back when,
Diamond and Morse were WAY ahead of the rest of the American made pack in
these areas. Jwis was the absolute class of the European stuff.

The only other manufacturing tweak--which was well into use back when-- was
(and likely still is) 'ball-burnishing' ( pushing a hard, polished steel
ball of the appropriate size through the holes) of the holes in the pin and
roller link plates --a process performed after making the plates (a punch
press operation) and after heat treat. This greatly improves fatigue life
by burnishing out any lines or scratches left by the punching die in the
holes--and also by leaving a residual compressive stress in the surface of
the steel at the surface of the holes. The latter is helpful because
fatigue cracks consistently originate at and propogate from the holes.

The link plates are and virtually always have been made (punched from)
medium carbon (no alloy) steel. (1036 or 1040, IIRC) The section thickness
of the links is such that none of the increase in hardenability available
with alloys is NEEDED (or necessary, nor would it make any difference if
alloy material were used, other than to increased $$$$$ ).

Pins are, and virtually always have been made from drawn wire (and, again,
they are of medium carbon steel, carburized to attain a higher 'case'
hardness on the wear surface). One of the bugaboos back then was pin
brittleness problems due to hydrogen embrittlement (a gremlin that would
work its way into the carburizing heat treat process from time to time. I
expect that modern , computerized process controls in the heat treat
operations have likely licked this problem.

Other than the above, not really many changes in chain tech over the past
35 years. Other than the fact that the Japanese likely no longer make THEIR
chain out of material that, back then, appeared to be related to used
Tootsie Rolls !!

I'm not aware of ANYONE _ever_ having made any 3DR or 3SR chain on the
American side of the pond. If 'Diamond' still existed today (?????) and
made 3DR, I would buy and use it in a heartbeat !!

I used to run the Jwis 3DR variant on a seriously hot-rodded 1600, as well
as on a very hotted up Jag 4.2--with nary a problem. Once 'acquired' a box
containing 50 feet of it 'for testing purposes'--and never did use it all
up. This usage was based on DIRECT personal knowlege of the results of
fatigue and wear tests run on 3DR chains from a numbER of mfgrs.  IIRC, a
good friend of mine also used (may still use) Jwis 3DR on his GTZ vintage
racer--first on a 1750, and later (I guess he felt the need to get legal),
on a 1600. (Although, the change may have simply have been because he
didn't think it would make any different---I recall a statement from him to
the effect of , "The only thing keeping the TZ from going faster is the
problem between my legs !!" :-)

Greg (among other things, actually worked on development of the cam drive
for the Ford FE 427 'Cammer' engine while at Morse)

  Keep
>in mind that however silly you think timing belts are, lower mass and
>lower friction compared to chains was a major motivation for their
>widespread use in engine designs of the '70s and '80s.  Certainly
>halving those losses in a traditional chain must be worth something.
>(Perhaps that's why your lawnmower and snowblower likely have
>sidevalve engines--giving up that much power to drive the valvetrain
>isn't worth it in a small, crude engine that might not gain a whole
>lot from an OHC design.)
>
>However, I wouldn't say that there's *no* loss in reliability.  Even
>if a modern single-row chain is just as strong in various tests as a
>double-row chain from 30 years ago, two rows is going to offer some
>redundancy in the event of certain failures.
>
>Joe Elliott
>'82 GTV-6 (toothed rubber belt, and no complaints)
>--
>to be removed from alfa, see /bin/digest-subs.cgi
>or email "unsubscribe alfa" to majordomo@domain.elided
--
to be removed from alfa, see /bin/digest-subs.cgi
or email "unsubscribe alfa" to majordomo@domain.elided