The cylinder head is the most important part of the engine to get
right if high power output is the aim. Unless the head flows well, the camshaft,
induction and exhaust systems are limited in what they can do. Modifying a head
properly is part science, part art and relies on flowbench development and
experience. There are many different head designs and each type will require a
particular approach to obtain the best results. There are, however, a number of
"ground rules" which will at least give a good starting point to the
process of head modification - this article is designed to set out some of
those rules and will hopefully help you make a more informed choice when you
buy a road or race head. We'll look at the different parts of the port in order
of importance to flow - this order might surprise some of you because the
straight part of the port which is easy to reach and comes in for much attention
with polishing wheels is not the most important in determining the final flow
potential.
This is the most important part of the head to get right and often
overlooked. The seat has to perform a number of functions:
1) - Physically seal properly so that there is no leakage of combustion gases.
2) - Cooling - most of the heat picked up by the valve head is conducted
through the valve seat into the head and thence to the cooling system. If the
valves run hot then power drops and detonation is more likely. Thin seats can't
conduct enough heat away.
3) - Flow well - the profile of the valve seat is the controlling factor in low
and mid lift valve flow. The port itself only really becomes a restricting
factor at high valve lifts.
Nearly all production cylinder heads have a 3 angle valve seat as
standard although it is often thought that cutting a 3 angle seat is just a
performance modification. I see too many reconditioned heads where just a
single angle cut has been used, resulting in a much wider seat than is standard
or desirable and losing a great part of the low and mid lift flow. Production
heads tend to have valve seats that are smaller than the head of the valve
itself - for best flow these need recutting to the same size as the valve head.
Seat widths are critical and on race engines there is a common tendency to use
too narrow a seat. Narrow seats neither cool the valve head properly nor do
they flow well - especially at low lift. They also burn out faster.
A good width for inlet seats is between 4% and 4.5% of the valve
size. The choice within this range would depend on cam lift and other factors.
Seat widths towards the narrow end of the range can improve high lift flow at
the expense of low lift and therefore might be chosen for race engines or for
use with high lift cams. For road engines it is better to use seat widths at
the high end of the range to get longer seat life and to better suit lower lift
road cams. For example a CVH has 42mm inlet valves - 4.5% of 42mm is 1.9mm and
that's a good choice of width for most heads.
For exhaust seats a rule of thumb is just to use the same width as
for the inlet. Because exhaust valves are usually smaller than inlets this
results in a proportionally wider seat than on the inlet side. In percentage
terms 5% of the exhaust valve size is a good starting point for the seat width.
Using the CVH as an example again, it has 37mm exhaust valves and 5% of this is
1.85mm.
Most heads use a 45 degree angle for the seat itself. A top cut of
30 degrees and a bottom cut of 70 degrees are good choices for the other two
angles. These 3 angles together enable the flow to make a nice smooth
transition from the port into the chamber or vice versa on the exhaust side.
The width of these other two cuts will depend on the head design but as long as
each of them is at least as long as the seat width, then the seat geometry will
work well. Many tuners use a 60 degree bottom cut, but I have always found that
70 degrees outflows 60 by a significant margin.
On the valve there should be a back cut at 30 degrees behind the
45 degree seat to smooth the flow round past the valve head and stem. Most
production valves do not have this back cut and it is worth quite a few
horsepower for very little work. The 45 degree seat on the valve should be the
same width as the 45 degree seat in the head.
A final point - you can't cut the seats properly if the guides are
worn oval or oversize. So part of any head modification should be to replace
worn guides before any seat work is started.
This is the second most important part of the port - the area from
say 1/2 inch above the seat inside the chamber to 1/2 inch below the seat in
the throat. Any parts of the chamber that shroud the seat need to be opened out
to clear the valve head by 25% of its diameter. i.e. a 40mm valve ideally needs
10mm clear space all round to give sufficient room for the flow to pass through
without restriction. Many heads have chamber walls closer than this and if so
then a certain amount of flow loss will take place. Hemi heads and multi valve
heads where the valve stem is angled away from the vertical do not tend to
suffer this type of shrouding. Bathtub heads like the mini and MGB are the
worst offenders and using a big valve can show a flow loss in some
circumstances because the shrouding gets worse.
The valve throat should be bored out to about 85% to 88% of the
valve diameter. This will leave enough material for a 70 degree bottom cut of a
decent width beneath the seat. Throats smaller than this restrict high lift
flow and throats bigger than this create too abrupt a bend through the valve
seat area. Many heads I see have no bottom cut at all and a throat bored out
right to the inside diameter of the valve seat. It might look impressively big
but it hurts flow. In early 2000 I did more development work on the seat and
throat area of the 8 valve Golf GTi Slick 50 cylinder heads. Despite the
already high power levels achieved in previous years, I found another 7 bhp
taking the engines up to 165 bhp at the wheels just from detail work in this
part of the cylinder head. Morever these gains were found by making the throats
smaller rather than larger!
This part of the port affects mid and high lift flow. There are no
hard and fast rules to shape here but the aim is to get as large a bend radius
as possible between the throat and the manifold end of the port. Waterways can
restrict the amount of metal that can be removed from the inside of the port
bend; the Golf GTi 8 valve head being a prime example. Quoting dimensions and
bend radii is pointless because lets face it, who has the measuring equipment
to take into a shop to check a head out before buying it? An eye experienced by
years of flow bench testing is the best tool to have when examining this part
of the port. At the least though it can pay to run a finger round the inside of
the port bend to make sure there are no obvious lumps and bumps there and that
the shape is smooth and fluid.
There is one absolutely vital thing to bear in mind though - it is
not possible to shape this part of the port properly without removing the valve
guide. If you can see cutter marks on the guide then you know it never came out
and there is a good chance that other corners were cut as well while the job
was being done.
This is the easiest part of the port to reach with cutting and
polishing tools and so it's not surprising that cheap and cheerful heads have
most time spent on them in this area. Mirror finishes and huge amounts of metal
removal are a sign of a cylinder head done more for show than for go. This part
of the port only affects high lift flow and rarely needs to be bigger than 80%
of the valve diameter for inlet ports. For road heads, 75% of the valve
diameter is usually ample. For exhaust ports it is less critical if the port is
bigger but it shouldn't be smaller than 80% of the exhaust valve diameter. For
square or rectangular ports you can't work from just a diameter obviously but a
similar port area to what a correctly sized round port would have is a starting
point.
When sizing this part of the port, consideration needs to be given
to the valve lift and the inlet manifold design. A low lift cam will not
require such a big port because the valve flow will also be lower than with a
high lift cam. Similarly, if the inlet manifold is poor then there is no point
in making the port bigger than it need be. The aim is too keep gas speed as
high as possible but without restricting valve flow.
Mirror finishes are a complete waste of time on either inlet or
exhaust ports. A smooth ground finish is fine and a final linish with an 80
grit flapwheel or emery wrapped round a stick is all that should be needed to
take off any remaining high spots. Within a given budget you can be fairly sure
that the more time there was spent on polishing, the less time was spent on
shaping and valve seats to really improve flow. However it is equally true that
if the port looks like it was done with a hammer and chisel then you ought to
have reasonable doubts about the machinist's craftsmanship and flowbench
experience.
It is not a good idea to shop for cylinder heads based on price -
the same amount of money won't necessarily buy you the same amount of flow
increase or the same amount or quality of workmanship. I've seen so called
"fully ported" road heads where the old badly worn guides weren't
changed, the seats and valves weren't recut at all and the work consisted of a
bead blast to make things look pretty and a quick polish inside the ports with
a flapwheel. The price was no lower than a properly done head with new guides,
3 angle seats and nice porting from another firm. A con yes, but not an
uncommon one. So try and find out in detail what you are about to get for your
money - does the work include any of the following:
Chemical clean or beadblast.
Guides replaced as necessary and on the basis of how much wear before
replacement is deemed necessary.
Are the seats recut with 3 angles and on what type of seat cutting machine (the
main carbide tooling systems are Serdi and Sunnen)
Are the valves refaced and back cut?
Are the guides removed before port work is done and replaced afterwards?
Will the gasket face be skimmed as part of the job?
Does the price include strip down and reassembly of valves or valve train?
None of the above will guarantee you any flow increase but at
least it might help prove that the important elements of the work are done
properly. In the next article we'll start to look at actual flow figures and
the sort of increases that are possible from different modifications.