This was copied from another forum with a theory on the engine failures talked about here recently. I figured it would be good for much banter...
______________________________
Hi. This Mr. CrankyPants come to talk to you about engine failure and
our tendency to monkey see, monkey do.
I notice that on this group, and others, there is a typical American
idea that any guy with a screwdriver is a better designer than a bunch
of highly paid Japanese engineers, despite the fact that these
engineers can design a 190 mph 12,000 rpm crotch rocket and race it
internationally.
Actually, the LS650 isn't that conservative a design. It's pushing the
limits of large single rpm and horsepower output for a production
bike. Little cylinders can put out inherently more horsepower, because
you can spin them faster, because their parts are little and easy to
reverse.
The flamefront ignited by the spark plug has to travel all the way
across the monster piston bore to burn instead of explode. To minimize
this, the spark plug should be in the middle. This flame front speed
limits the rpm also, because the time to do the job at high rpm gets
smaller than the time needed to propagate the flame. In a big single,
you can get shock wave explosions from hot spots, also, especially if
there are carbon deposits from running too rich. More later on rich.
The Savage is carbureted, not fuel injected. This gives rich and lean
spots in the compression space because the carburetion stops and
starts, and there isn't enough length in the manifold to thoroughly
mix the droplets and let them vaporize. More uneven burning.
In any lean burn area, the temperature rises. The engine is already
running lean due to smog regulations, and backfiring. So the would-be
motorcycle engineer decides to solve this lean burn problem
with a 152.5 main jet, which brute force floods the cylinder with raw
unburned gasoline, which is an excellent oil solvent, which removes
the oil film on the top piston rings.
This causes scuffing and perhaps even deposition of piston aluminum on
the cylinder walls which generates a lot of friction, which overloads
the connecting rod big end bearing, which fails, overheating the
conrod, which breaks and goes through the case.
(Big end bearings tend to flash out the oil due to high linear speeds
between the big piston pin and the bearing surface on the conrod. But
you need a big bearing because that big piston and rod generate
relatively large riciprocating forces that require large bearing
surface. More limitation on rpm and thus horsepower. All of this just
to get that lovely thump.)
This ain't rocket science, but if you don't know this, and many other
things, you are asking for trouble dropping a low-back pressure (loud)
muffler on the bike, which makes it run lean at some rpm and rich at
others.
I have no doubt those engineers sweated blood designing this
sweet-running big single, getting the spark advance curve right, the
carburetion smooth and powerful, and the clearances just right.
Then we go at it with our limited engineering knowledge, and convince
each other that anything that anyone else has done must be OK because
the bike hasn't blown up yet.
I'm sure this sounds awful conceited and arrogant. My apologies, but
I've run hundreds of thousands of miles on stock BMW's including the
250cc single, British, Japanese and Italian bikes. I've only driven a
Harley 250 cc Sprint, which is Italian, and an FLH brand new, for
about fifty miles, which was long enough.
I know in my bones, from years of reading technical design manuals and
talking to dozens of shop managers that if an engine fails, the first
question is "Was it modified by the owner?"
People just don't seem to understand how exquisite a balance of design
features must be maintained to get performance AND reliability in any
aircooled engine.
Tinker at your own risk. You might not blow it up. Ten for fifteen
thousand miles is just about the right failure point for a
miscarbureted big single. BSA Gold Star, Matchless GS80 and 600
Typhoon, all easy to blow up by carb and muffler tinkering. Things
that make you go fast and make noise decrease reliability.
Have a nice ride. The Savage is already a VERY nice bike with a hard
seat. Change the seat. Change the handlebars. Add saddlebags and GPS
devices. Get an iPod. Goop up the squeaky speedometer. Don't overload
the alternator. Leave the carburetion and the exhaust stock. Don't be
greedy. The fairy tales are right.
Ormond Otvos Richmond California FSU Physics, 1961. Do the math.