User Guide

Dyno2000 Advanced Engine Simulation—93
Using Iterative Testing™
possible to begin a new Iterator test using any of the spawned engines as a Baseline
Engine to further “home in” on the desired results.
Halting And Restarting Testing
At any time during the Iteration process, you can stop calculation by click the Halt
button. Simply clicking Run will resume calculation with no data loss. If, instead, you
wish to reset the Iterator and start the simulation series over again, click the Restart
button, or you can click Setup, change any parameters or step values you wish,
then click Run to start the Iteration process from the beginning.
If you Halt the simulation, you can close the Iterator screen by clicking Close. If
you reopen the Iterator screen before you close the Dyno2000, you will be able to
resume calculations on the current Iterator series. However, when the Dyno2000 is
closed, all Iterator calculations and results that have not been spawned to engines
and saved to disk will be lost.
Tips For Running Iterative Testing
Setting up an Iterative series only takes a few seconds, however, if you include
too many parameters, ranges that are too wide, or step values that are too small,
you will create an Iterator series that contains too many tests. If you create a series
longer than 300 million tests (even fast computer systems will require one year or
more to complete 300 million tests) the Dyno2000 will request that you increase step
values for selected parameters.
The best way to find optimum components, especially cam timing, is to use large
step values (5 degrees or more) to “get in the ballpark” of the right values. Then run
a second Iteration series on the best engine, keeping the range of values narrow
(perhaps just a 5 or 10 degree range) and use smaller (perhaps 1 degree) step
Iteration Counter Limits
Setting up too many
parameters, ranges that
are too wide, or step
values that are too
small, will create a test
series that contains too
many runs. This experi-
mental setup requires
over 1.6 billion tests to
complete (would take
almost 6 years). The
Dyno2000 limits a test
series to no more than
300 million tests (about
one year of computation
time on a fast com-
puter!).