Best Computer engine simulator?
Thread Starter
Intermediate
Joined: Oct 2002
Posts: 33
Likes: 0
From: Key West FL
Best Computer engine simulator?
I tried searching the archives but wasn't having much luck. I wanted to know everyone's opinions on the computer desktop dyno programs that are out there. Thanks for the help fellas! :cheers:
Race Director
Joined: May 2000
Posts: 12,712
Likes: 2,270
Re: Best Computer engine simulator? (tangel)
For the money Desktop Dyno is a good value. It's easy to use and appears to give fairly accurate results - top end power appears close to SAE gross, but peak torque is probably high.
Engine Analyser has more input, which makes it more flexbile, but a little harder to use. One thing I don't like about it is the detonation prediction feature, which retards timing with a corresponding reduction in output in that range. It appears too aggressive and I can't turn it of, but one thing I do like about EA is it allows you to predict SAE net or installed power, by attaching accessories, assigning exhaust system back pressure, and changing atmospheric conditions. DD2000 is basically SAE gross horsepower.
One thing to remember about these programs is that they are not exact predictors, but are very good at predicting trends. For example if config A appears supperior to config B then a real A will probably be better than a real B, and the percent difference will probably hold.
The biggest limitation of these programs is cam characterization, which is the heart of the engine. Published specs are often different that the actual timing in a running engine, which throws off results and comparisons.
The other mistake I have seen users make is to focus entirely on peak power. Unless you're running at Bonneville you will spend 99.99+ percent of your time driving BELOW peak power revs at LESS than WOT. Our driving conditions are not one-dimensional, so the engine should not be either.
Focus on torque bandwidth and keep the bottom end of the 80 percent torque bandwidth no higher than about 2000 revs for a street high performance engine.
Duke
Engine Analyser has more input, which makes it more flexbile, but a little harder to use. One thing I don't like about it is the detonation prediction feature, which retards timing with a corresponding reduction in output in that range. It appears too aggressive and I can't turn it of, but one thing I do like about EA is it allows you to predict SAE net or installed power, by attaching accessories, assigning exhaust system back pressure, and changing atmospheric conditions. DD2000 is basically SAE gross horsepower.
One thing to remember about these programs is that they are not exact predictors, but are very good at predicting trends. For example if config A appears supperior to config B then a real A will probably be better than a real B, and the percent difference will probably hold.
The biggest limitation of these programs is cam characterization, which is the heart of the engine. Published specs are often different that the actual timing in a running engine, which throws off results and comparisons.
The other mistake I have seen users make is to focus entirely on peak power. Unless you're running at Bonneville you will spend 99.99+ percent of your time driving BELOW peak power revs at LESS than WOT. Our driving conditions are not one-dimensional, so the engine should not be either.
Focus on torque bandwidth and keep the bottom end of the 80 percent torque bandwidth no higher than about 2000 revs for a street high performance engine.
Duke
