- Dynomation 2 is an engine simulator. This is a
program which displays a running engine that's been
assembled on the computer screen and is defined by its
input parameters. Designing engines this way has the
potential to save barrels, dynamometer time and much of
the tuned exhaust pipe making process since many of these
characteristics can be simulated on the computer.
Different tuned pipe dimensions can be entered into
Dynomation 2 to see what affect it produced upon the
engines power - though there are some limitations to the
parameters that can be entered. It can then be further
enhanced by making it more specifically targeted either
by selecting a more appropriate operating range or
dialing the dimensions into a more tightly focused rev
range. The changes can then be run through the simulation
process again to see if it works as planned. The
simulations are very time consuming but design time on a
computer is cheaper than buying cylinder barrels and
making pipes. To use this as an option a whole engine
(disassembled) is needed to measure the parameters from,
or at least the cases with crank installed, barrel, head,
reed cage and intake pipe. Dynomation requires so much
data from the upper and lower end of the engine it would
be impossible to do it without having these engine parts
next to the computer to measure everything as accurately
as possible. Doing it this way will allow the least
amount of error between the actual engine and the
simulated one. When the simulation process is over the
result is a massive amount of information - from it the
best engine/pipe/porting for the application can be
determined. The idea is for the output information to
steer the operator directly and precisely toward an
obvious choice. One of the features of this software is
that is can simulate up to 3 cylinders running at the
same time (a triple cylinder engine) and it can simulate
them as though they were each fitted with an individual
tuned exhaust pipe or a 3 into 1 pipe like a watercraft.
There is a note in the manual about the stability of the
program when running such complex configurations - it
could crash.
-
- It is impossible to enter an engines parameters into
this (or any) software without first knowing the correct
values - they have to be discovered. I spent a
considerable amount of time discovering my test engine -
as a computer "sees" it. The amount of time it takes to
enter the necessary data to properly simulate an engine
is quite lengthy. There is a tremendous amount of engine
parameter measuring to do if the "virtual" engine is to
be just like the real one. When thinking of it this way
it leads me to think that "more is better." Though it's
tough to accurately measure some of the parameters the
software calls for - it is required to make the
simulation as real as possible. When measuring the
transfer ports I found it helpful to cut some paper into
shapes until the shape matched the tunnel as accurately
as I could. I would later call these pieces of paper,
templates. In this case transfer port templates. When
made properly, the template could then be measured very
well and its average duct length and entrance area
recorded for the software program. Though there is no
mention in the manual of performing this extra task to
produce good results, I found it to be very helpful. The
manual describes the right place to take the readings
from - it just doesn't say how to come by the
information. Anything that can help to make that
difficult measurement more accurate, or easier to record
is a good thing. I considered this extra step necessary
because it would complement the degree of accuracy of the
overall simulation. Trying to guess at any of the
required information fields makes the purpose of the
simulation useless - one bad entry and the whole things
goes bad.
-
- I entered the exhaust pipe dimensions into the
required data fields. I had to carefully measure each
section of the Blaster's pipe and record the data. This
was not a problem since the pipe only had 7 sections to
it. The exhaust has such a tremendous effect on the
overall performance of a two stroke engine it is
absolutely crucial to get it right. Even small changes in
the diameter or length of a cones section has a
noticeable impact upon the engines performance. It is the
job of the software to reflect the small changes and
produce a result that is worthy enough that it can be
relied upon. As I mentioned above there is a limitation
on this data entry parameter. The pipe sections must be
divisible by the simulation mesh size. For more
information about this the manual point the user to
Blair's books once again and notes that using a size of
"10 mm" works best for most simulations. When using the
10 mm size it is necessary that all pipe section lengths
be evenly divisible by 10. That's somewhat of a snag.
Once the engine data is entered a simulation can be
run. The output data requires a specific targeted
operating range - such as 4000 to 9000 for example. The
simulator will start at 4000 RPM and run through the
process 6 times as it "warms up" to get the pipe and
internal engine temperature up to what it should be. Then
it proceeds to run between 20 and 50 simulations at that
RPM in order to establish a solid set of baseline data at
that RPM. The time it takes to run a single simulation is
about 15 seconds. The 6 warm up runs and 20 real runs at
15 seconds amounts to about 5 to 6 minutes at that RPM.
It will continue to run simulations at any specified RPM
increment. Typically this would be 500 RPM's or 1000
RPM's for roughing in the data. The process can be
stopped if the data is not coming out like it was
thought. This is an advantage that an engine simulation
program can offer and something that some other software
simply can not duplicate.
Some of the data output information includes pressure
(intake, transfer & exhaust) vs crank angle @
specific RPM's, reed lift @ RPM, exhaust temperature @
RPM, trapped cylinder pressure, peak cylinder pressure,
charging efficiency, trapping efficiency, delivery ratio,
PMEP, BSFC, piston speed, HP and torque. Obviously this
information has to be understood in order to be put to
good use. Dynomation 2 doesn't try to explain its output
data it simply displays it for the interpretation of the
operator. It is up to the operator to decipher the data,
and make use of the huge amount of relevant information
that it gathers. While the simulation is running it is
very interesting to watch the wave graphs and other on
screen animation. A visual representation of this
magnitude can let the mind fly with ideas. It provokes
design changes to watch the outcome. Simulation
definitely have a place in the two stroke software
arena.
-
- One of the most important adjustments to make when
running a simulation is to keep the BSFC (brake specific
fuel consumption) at .650 lb../HP-HR at the peak. This
also requires an increase of .150 lb/hp-hr for each 1000
RPM over peak power and an increase of .150 lb../HP-HR
for every 2000 RPM's below peak power. It has been my
experience that this requirement requires a bit of
attention to get it dialed in. Within the Dynomation 2
help section there is a useful chart for determining the
more extreme values of this parameter. The values are
listed according to the fuel octane requirement of the
engine. A BSFC value of .7 can be used for pump gas
engines, while .65 to .69 requires a solid 92+ octane.
Higher output engines may run between .55 and .64 when
using 100+ race gas when their engine is being built for
long circuit type races. Engines built for drag racing
running 100+ octane or engines built for long road
courses running 110 octane fuel may use a .5 to .54 BSFC.
The extreme of this specification is using a BSFC less
than .49 and is only for drag race engines running 110
octane fuel or higher.
-
- When running the simulations a couple of things are
noticed - the reeds don't close until well after the
exhaust port opens and as the RPM's rise the point when
they close also rises. At 5000 RPM they were closing at
about 100 degrees ATDC but at 6000 they are not closed
until well after the transfer ports open. This is due to
the inertia of the incoming charge and the latency
associated with the reed material. This pipe has the
reeds opening again just after BDC and almost closing
about 40 degrees after that - then opening again. At 7000
RPM the same situation is observed though when the reeds
open for the second time they do not close nearly as much
and are continuing to feed during all but about 45
degrees of crankshaft rotation. At 8000 the reeds do not
close until about 25 degrees before BDC.
-
- More to come - Rick
|
