how many guys use analyzer software to...

[grumpyvette]

how many guys use analyzer software to...

 

 

how many guys use engine analyzer software , like DD-2000, engine analyzer, and similar programs to check out potential changes and mods, and have you found the real world results to be similar to prodicted results?

I have several versions in the shop to use for spotting potential problems and areas that need improvement, sure the experiance helps a great deal, but

these are very useful tools but the less expensive options are better at guessing trends in the engine mods than acctual results, but thats to be expected simply because they don,t take into accound , NEARLY the same amount of data and ASSUME things the more complex programs actually calculate

now thats NOT always a bad thing because the old (GARBAGE IN/GARBAGE OUT) factor applies to most software and if your not 100% sure what your entering is factual data, the simple programs might give closer results than a more precise simulation with some bad data entered

 

heres a few examples

 

http://www.proracingsim.com/desktopdyno.htm

 

http://www.virtualengine2000.com/

 

http://www.land-and-sea.com/dyno-max...e/dyno-max.htm

 

http://www.motionsoftware.com/

 

http://www.dynamometer.fsnet.co.uk/d...o-software.htm

 

http://www.engineprosoft.com/

 

http://www.auto-ware.com/software/eap/eap.htm

 

Ive played with several of these dyno software programs and they run from extremely simple and fairly accurate to extremely complex and difficult to use, but very accurate

 

now I tend to get the best results accuracy wise with software results matching real world results, with EA PRO

 

http://www.auto-ware.com/software/eap/eap.htm

 

and

http://www.motionsoftware.com/

seems about the fastest and reasonably close at predicting TRENDS

[Michael]

Garbage in --> garbage out; true enough. But sometimes truth in --> garbage out, nonetheless.

 

I’ve used DD2000 with very marginal success. DD2000 arrives at a higher torque curve at nearly every rpm, whenever a given component is replaced with a bigger one. Got a 650 cfm carb? Great – now replace it with a 850 cfm, and watch the torque curve rise magically. There is no penalty in throttle response and no discernable penalty in quasi steady-state torque. Nor is there a penalty for mismatch of components. The program is very sensitive to intake duration – for some reason, this one parameter has a completely magical effect.

 

Now to change the topic somewhat, and consider for a moment what a good research code (in mechanical engineering, not in “practical†automotive mechanics) should do...

 

A good computational model needs, first of all, the CORRECT GRID – that means full knowledge of the geometry of the air cleaner, carburetor, intake manifold, intake port, combustion chamber and cylinder, exhaust port, header and exhaust tract. It needs the right model for wall surface roughness, including the various bumps and protuberances along the way.

 

Next, a good computational model needs to be solving the “right†set of equations. This is the full 3-D compressible Navier-Stokes, with the right constitutive laws for 2-phase flow, and the right equations for energy and entropy.

 

Next, it needs the right boundary conditions, such as the piston and valve kinematics.

 

And finally – this perhaps the most difficult part – it needs the “right†turbulence model, to “close†the expression for the viscous stress tensor in the momentum equation.

 

This is the sort of computation that would be run in university graduate schools with connections to the OEMs. What I would like to see is a better connection between the research world and the practical applications world.