the simulation to more closely match
dyno tests, but there will still be
questions. Even so, pipes that did well in
simulations have done well in dyno tests
and in actual SAW hydros. I believe the
ranking of pipes as shown by the
simulations won’t change, just ignore the
power shown.
Another question is why use power
rather than torque to compare pipes.
Torque is a measure of how much work
is being done. The dimensions are (in
British engineering units) foot-pounds.
Power is a measure of how fast that
work is being done. Its units are foot-
pounds per second. We’re racing, so the
speed is important. The plot of rpm
against time for a rapidly accelerating
boat is very close to the engine’s power
graph. In cars, where the resistance is
easier to estimate than in boats, this
principle is used in “dashboard dynos”.
All the examples shown are for a
particular 26 cc engine. Larger
engines and different port timing will
need different pipe designs.
I conclude that for the fixed diameter
and length pipes shown, the G3s pipe
has the best peak power and should be
good for all around racing. A shallower
baffle cone angle gives more over rev
power but sacrifices low end power. In
the real pipes we have tested, it also may
sacrifice peak power, but is still good for
SAW engines. The broad power curve
makes prop matching easier. Steeper
baffle cone angles have the opposite
effect.
Diffuser design is more complicated.
The simple two stage diffuser in this
example is close to optimum for the
fixed pipe diameter used. However,
more diffuser stages should allow
expansion to a larger pipe diameter,
generating a stronger low pressure pulse.
In any case, some angle from the start of
the exhaust outlet is superior to fixed
diameter headers. Some designs may
work better with 1” headers, but this
design does not.
Finally, the stinger diameter was held
constant in the above examples. Often, a
bigger stinger will give more power,
especially as the rpm is increased,
increasing exhaust flow. Testing with
removable stingers in a 35 cc engine
showed that we needed a larger stinger
as we modified the engine for higher
rpm power.
PROPWASH
7
Next let’s look at standard 7/8 and 1” header pipes compared to the G3s pipe
whose area increases steadily from the exhaust outlet area. The 10 mm first section
represents the area change in the adapter from the exhaust outlet to the round pipe
header.
All the header style pipes and the G7 pipe with a steady increase from the flange to
22 mm diameter like the 7/8” (22 mm) header have broader power curves than the
G3’s pipe. They have lower peak power, though. The 1” (25 mm) header has lower
power than the 7/8” header pipe.
Before drawing conclusions from the above simulations, there are several
reservations. The friction and heat loss approximations in the simulation I use are
designed for larger engines. I notice that the peak power is overstated and that the
power doesn’t drop off as fast as it does in our dyno tests on the same engine and pipe.
All of the power curves will drop off more quickly after 16,000 rpm in a real engine.
For reasons I don’t fully understand, the low end power is understated in the
simulation. See the graph below of real dyno results with a different pipe compared to
the G3s pipe’s simulated power graph. It is possible to change some assumptions in
April 2012