> I am in the market for a new crankshaft for my 1950 MG-TD. However, I
> am confused about what to buy. What is the difference between a normal
> crankshaft and a "steel billeted (sp?)" crankshaft? Are there other
> types of crankshafts available? If so, what are the differences?
Billet steel is the usual term. It means the crankshaft started out
as a single piece (billet) of metal the size of the cumulative diameter of
all the throws. This is then successively machined down to the size
of the crankshaft, with journals (the round parts where the bearings sit)
and throws (the straight parts between the journals) machined in place.
This is a very expensive way to make a crankshaft, but it results in a
very strong unit. (Visitors to the Monterey Historics had a chance to
see photos of the stages involved in machining a Miller V-16 crank, used
in unlmited powerboat racing in the Thirties, from a single billet of
steel. It was awe-inspiring, as was the engine, which was bigger than
a Sprite.)
Forged steel is the next strongest. In this, a piece of steel of the
approximate shape of the crankshaft is forged into a much less approx-
imate shape, usually by a multi-ton drop-forging machine. Then the bits
that aren't crankshaft are removed by a milling and grinding machine.
The forging process aligns the crystals of the steel, if my old knife-
making theory hasn't oxidized; the shock and pressure applied to hot
steel makes it stronger.
(Sidebar: How important is crystal alignment? The new Ford "Modular"
V8s use connecting rods forged in one piece. To get them to split
so they can put them on the crankshaft, they are scribed and carefully
broken, so that the metal faces line up like the coin around the necks
of the father and daughter separated by a shipwreck 15 years before
in romance novels. :-) The purpose, though, is to keep as close an
alignment as possible within the crystal lattice in the top part of
the rod and the cap, so that loads are transferred evenly without
stress-inducing gaps in the lattice. Weird stuff, metal.--sf)
Nodular iron comes third, though with the right heat treating and
alloys some people consider that this is equal to forged steel. In this
technique, nodules (little balls) of iron are heated to just below melting
point, then (I think) pressed in something very much like a drop-forge.
I think the main advantage to this is in making multidimensional cranks
such as those for inline sixes and other engines whose crankshafts have
throws in more than just one plane. (Four-cylinder engines have
basically "flat" cranks, with all the up-and-down bits in the same
plane.)
Cast iron is cheapest, strong enough for most ordinary purposes. Few
high-performance cars come with cast-iron cranks, as the stress limits
are lower than for the other techniques mentioned here. But if you're
making crankshafts in the millions, it's cheaper to pour them out of a
huge bucket of iron than to assemble them in one of the other methods.
Once manufactured, there are ways to treat the crank that can lengthen
its service life or extend its operating range. You used to be able to
get crankshafts nitrided, a thermochemical treatment that doped the
surface layer of the crank with a nitride (and I've always wondered
just *what* nitride it was) that toughened the outer layer of ferrous
material. This made it both more resistant to wear and also to cracking.
(The shape of the throw-journal interface, BTW, is crucial to preventing
cracks; you don't want sharp cylinder-plane interfaces, you want a
radiused edge that allows for even distribution of stress at the edge.)
Most crank failures occur when stress risers at the surface cause cracks
from flexing, and then these cracks propagate through the center of the
crank. This is why aligning the subsurface crystalline structure of
the crank is so important for high-performance operation, and why people
will pay extra to have a crank machined from a solid block of steel in
which the crystal lattice is uniform and strong by nature of the
manufacturing process. (Actually, *most* failures are probably caused
by wear on the surface that finally overcomes the ability to remachine;
I suppose it's most CATASTROPHIC failures that result from surface cracks
propagating through the metal.)
The problem with nitriding is that the chemical is toxic, carcinogenic,
and generally a Bad Thing, even worse than freon. (Half-smiley.) Let's
just say it was banned by OSHA even before there was such a thing as the
Green Police, as it's not good for the people who use/do it, and there's
no way to safely discard the used-up goop. Because of this, nitriding
has been replaced in the last 25 years or so by a technique called
Tuftriding, which produces similar results but uses more friendly
chemicals. (Again, I don't know what chemicals; I'll have to ask the
friendly folks at Kaeding, the shop that does all the machining on my
race motors.) Same basic idea: make the top surface more resilient so
that it doesn't wear out or get cracks. There's also shot-peening,
which uses metal shot to put thousands of tiny dimples in the top
layer of the crank's metal. The effect here is to micro-forge the
outer surface of the crystal structure, reducing its tendency to
crack in one further way. Shot-peening is more commonly done to
rods; I'm also not clear on whether you would shot-peen the journals
(it seems like a Bad Idea, but if you did it before micropolishing
it might be a Good Idea. I'll have to ask Kaeding the next time I'm
in there for something.)
But surface treatments of crankshafts, while more than a band-aid fix,
isn't the most important feature for crank longevity. Metal compound,
manufacturing technique, and mechanical finishing are still the most
important ways to protect an engine from the inside out. Be sure when
you have a crank reground that it's micropolished to assure the smoothest
possible surface; any parts of the metal that stick up from the journal
are going to hang on the bearings and cause wear. While you've got the
crank out of the car, it's relatively inexpensive to have it balanced,
but be sure you have it balanced with the rest of the reciprocating
mass -- this means the pistons, rods, balancer, flywheel, and clutch
pressure plate. Racing engines are typically balanced to tolerances
of half a gram or less; Kaeding matched my rods to within half a
gram. And just for fun, find out how far out the originals were --
my brand-new, still in wax paper, British Phlegmsucking Leyland rods
were more than 25 grams (almost an ounce) out of balance from lightest
to heaviest.
> Finally, do you know of any place that has new crankshafts available at
> this time? No one had any available last time I contacted the British
> car parts catalog shops.
Moss used to carry new, reproduction XP?G cranks. Al Moss used to
have a picture of one taped to the tachometer of his vintage-racing
MG TC, just before the redline indicator. Barring that, you can try
used parts places. If you end up getting a used crank, be sure to
have it crack-tested and inspected by a reputable machine shop before
you throw it into your TD. The best way to identify a good machine
shop is to find out how much racing work they do. While having a
race car in the product line is a good marketing tool for a car
manufacturer, for a machine shop it's a visible sign that they have
the skills, equipment and -- I can only call it craftsmanship and a
work ethic to do things to very precise tolerances under conditions
in which failure is more than just expensive.
--Scott "Lookin' for fun and feelin' cranky" Fisher
|