Valve Train Geometry

It must be winter... somewhere :-)

I've noticed that I get more questions about engine-building during
the cold months, along with comments such as '...not driving the bug
now that WINTER is here.'

But that's another subject.

The subject here has to do with valve train geometry and comments such
as:

"I understand that the rocker arm is supposed to be at the mid-point
of the valve stem when the cam is at the mid-point of it's lift... but
what I don't understand... (fifty to a hundred things folks don't
understand)"

First off, what the fellow(s) don't understand to begin with is the
thing they THINK they understand has to do with adjusting the valve-
train geometry for a WATER COOLED engine, in that they cite only the
OUTPUT side of the equation... that bit about the valve needed to fall
at the mid-point of whatever and so & so.

What is being overlooked here -- and why VW valve-train geometry tends
to be a bit tricky -- is the fact that on a horizontally-opposed
engine the geometry equation has and INPUT and an OUTPUT.

Put your thinking cap on for a minute and THINK about it. The push
that actuates your valve originates at the camshaft. When the lobe of
the camshaft PUSHES the cam-follower, it PUSHES the push-rod, which
PUSHES against the rocker arm.

That's the input side of the equation.

The OUTPUT side of the equation is when the rocker-arm PUSHES against
the valve stem, causing the valve to be lifted off the seat. That is
the OUTPUT side of the equation and as you can see, it's not nearly as
complicated as the input side of the equation.

The tricky bit here is what happens when you try to transmit a PUSH
around a corner using a push-rod and a lever (ie, the rocker arm).
The push-rod reflects some amount of LINEAR motion. Push an inch on
the bottom and you'll see an inch on the top. But levers aren't like
that. Push on a lever and you are dealing with NON-LINEAR motion --
the lever is prescribing a segment of an arc. Push an inch on a lever
and you will ALWAYS see something LESS-THAN an inch at the output.
That is the essence of valve-train geometry. The fact you must solve
for TWO equations rather than just one, as is common with water-cooled
engines, is why there's so much confusion out there.

When trying to 'push around a corner' you will always have some
losses. The whole idea behind setting up your valve train geometry is
to align the components so as to give the LEAST AMOUNT OF LOSS.

When someone addresses only the output side of the equation, it isn't
uncommon for them to give away up to 25% of their cam's potential
lift.

The other biggie is that when you're dealing with a non-stock engine,
you must determine your cam's INSTALLED LIFT for all EIGHT valves.
This is because a re-worked or after-market crankcase is liable to
have the center-lines of the crank and cam shafts OFFSET by some small
amount. That is, the deck height for the barrels will be different on
one side of the crankcase than the other.

Once you know the cam's lift, you can determine the 'half-height.'
This is what you need to determine the 'half-arc' point for the INPUT
side of the equation. The tricky bit is that when setting the half-
arc point for the input, you must make provision for the same thing on
the output.

Actually, it's not as difficult as it sounds. What's difficult is the
need to UNDERSTAND what you're doing. Once you understand why the
valve-train of a horizontally-opposed engine is different from a Ford
or a Chevy, it's pretty easy to make up a simple jig that allows you
to do most of the work on the bench rather than on the engine. Of
course, if you're only building one engine, it doesn't make a lot of
sense to build a jig... you simply use the ENGINE as the jig.

The jig allows you to set-up each head on the bench, a real time-saver
when you're doing more than one engine at a time... but I don't think
it's justified when doing just one engine.

-Bob Hoover