Let's Discuss Engine Simulation Software
 

I'm taking an High Performance Engine building and tuning class currently and am looks at option for engine simulation software. We use Dynosim5 currently but I was wondering what else is out there and how the software packages compare to each other.

Dynomation5 ($500) - Dynomation-5 Engine Simulation
Dynosim5 ($100)- DynoSim5 Main Page
Ricardo Wave - WAVE - 1D Engine Gas Dynamics
Lotus Engine Simulation -
GT Power - Applications
Virtual 4 stroke ($?) - Virtual 4-Stroke Design and Simulation Software


Please chime in with your experience with any of these software. Their strengths and weaknesses and what not.

Also if anyone has run miata engine simulation before and can share any specs you might have to reduce my research time that would also be greatly appreciated. :party:
I'll be posting results here for discussion and feedback as well as I go through this project so check back often if your interested in this sort of thing.

.

A quick overview of Dynomation5 for those of you who aren't familiar with engine simulation software.

Lonely thread we have here. But that's okay. I'll just leave these here.

The stock torque curve is pretty off early on in the rpm range but the hp curve seems to track okay.

The 14.7 psi turbo curve seems to be pretty close to this other turbo dyno linked below.

https://www.miataturbo.net/attachmen...ine=1422599102

https://www.miataturbo.net/attachmen...ine=1422598958

99 stock
https://www.miataturbo.net/attachmen...ine=1422598958

99 gt2560
https://www.miataturbo.net/attachmen...ine=1422598958

Hey there, didn't see this thread until now. I'll chime in.

I have experience with Lotus Engine. I had to teach myself the software. I will say, it has it's limitations and quirks, but it's still useful to understand what's going on in an engine.

I have done parametric optimizations of intake and exhaust runner length with Lotus. That was pretty fun. I maximized Volumetric Efficiency over a 4000 RPM window, it was pretty awesome, although it took my computer a weekend to do this optimization. It ran simulations of runner lengths down to the nearest 0.5" to come up with the results.

From what I have read, GT Power is the most accurate software, but to really obtain accuracy, you have to know flow coefficients of every surface, materials, geometry, etc. Pretty much every parameter of an engine, including ones that can only be found through expensive and time consuming testing. Supposedly, GT Power has let certain formula teams to get within 2% of reality, and to improve upon this dyno testing was the only way.

I wouldn't believe it unless I saw it in person, but that's what I have heard.

I think Lotus is more accurate than Dynomation, but that's just a hunch as I haven't used Dynomation before. Dynomation seems simpler, which will often lead to a less accurate solution, but would be good for quick comparisons.

Anyone else have any experience with Engine Simulation?

I've fiddled around with dynosim2, nothing serious. Ran into the same thing you did with it overestimating the power and torque down low but looking about right once in the power band. Its interesting to see in your plots that +1 valves lowered power. Is there any parameters that would let it know how shrouded the valves are, or even the placement of the valves relative to the diameter (IE angle and valve location).

I think the problem with the low end being off is related to the cam profile. You might have the lift and duration right but I bet the flanks are set more aggressive than real life. IIRC my dynosim came with some actual v8 cams already programmed and and the ones that were actual cams vs entered looked more reasonable, on v8s of course.

Leafy, I don't see any Hp losses with the bigger valves. just very minor gains. Less than I expected and was hoping to see tho.

There are no inputs for things like that. It might put some shrouding factor in based on bore size and exhaust and intake size and how much area they take up but I'm not sure.

Thanks for Chiming in. I agree with you that dynomation is more basic then the others but is simple to set up and should give you a good idea of changing a single variable and looking at how that affects everything rather than outright dead on dyno plots.

I've heard GT power is very accurate but really requires an OEM engine development level of understanding. Just too much work for an aftermarket modification tool. Ricardo wave is also supposed to be accurate and I'm hoping to give that a try to compare.

I used Engine Analyzer Pro several years ago while working on a project. It was obviously written for Windows 3.1 and never ported to new libraries, which annoyed the hell out of me and only supported a single-core (no multithreading to get quicker results on the various runs). But it worked.

okay....

:ugh:

:sadwavey:

Cyotani,

How does Dynomation input boost? Do you just input estimated boost levels at various RPMs?

Also, does the software calculate Volumetric Efficiency, or is it something you input? If it's an input, your low RPM power and torque values would probably be off because the VE input is too high.

On a side note, I think Lotus Engine is about as advanced as any normal person would want to go. GT Power is the kind of simulation tool that requires a small team of engineers to do extensive testing to get all of the information to put into the model.

Lotus Engine has a few really cool features, such as you can input the compressor and turbine maps from a given turbo and it will find the mass flow rates, boost levels, etc all on it's own. At least, in theory.

I have always wanted to set up a BP in Lotus, but my license is only for single cylinder models, which can still be a semi-decent representation, you just multiply power and torque by 4. The main limitations of a single cylinder model are that it does not take into consideration the intake plenum feeding 4 cylinders, and the exhaust pipe carrying the full amount of exhaust, so results will be higher than expected.

The overarching theme with Engine Simulation software is that, the simpler it is, the further from reality it will be. For every input you can think of that would affect the way the engine operates, if this is not a value the software takes as an input, it is just assigning some unknown value. For every assumption, it will be the incorrect assumption.

You input the turbo, trim, A/R etc. and the max boost level. No boost by RPM input but that would be nice if you had actual spool data.

VE is calculated by the software. I can post those and the Boost plots up but I hand entered them into excel to overlay them and was to lazy to do those plots as well.

did you also input you're running a 0.5" exhaust?

elaborate please.

At the risk of exhuming a 4.5-year-old thread...cyotani, you currently have a version of your built-engine for sale, do you not, and you've included its torque/hp curves (in naturally-aspirated form)? How do these curves compare with the software's prediction?