simulink model of R6

[tom888]

hi

does anyone have a simulink model of a yamaha R6 engine?? or does anyone know how to create one??

cheers

[tom888]

hi

does anyone have a simulink model of a yamaha R6 engine?? or does anyone know how to create one??

cheers

[drivetrainUW-Platt]

simulink?????? wow thats a new request!

[Adambomb]

yeah...I hope you're not trying to make a numerical model of the engine on your own...if you value time and usable results that is

[tom888]

well im trying to make a simulink model of a car accelerating for 20seconds. the idea is to show the differences in performance between a normal geared car and a car with a CVT. seems to work, but the engine that i made is rather dodgy, though seems to give the right out puts.

[B Hise]

it sounds like what you need is a torque curve for an R6 engine in FSAE trim. Search the forums and you may be able to find the relevant info for your model.

-Bryan

[js10coastr]

http://www.calpoly.edu/~jjang/R6.JPG

[D J Yates]

Tom888, I do know how to model an engine in Simulink/Matlab, as i have done so myself. I can give you a rough outline as to what to do, but it'll take longer than the couple of minutes i have to write this post, so it'll have to wait until tomorrow. In the mean time, if you can tell me what data, or resources for gathering data (e.g. R6 engine on a dyno, ready to go), that would be most useful.

Asside, if you're researching CVT, google my employer and have a look at IVT also.

Adambomb, although developing a full engine model from scratch is not particularly useful if you want imediate reliable results to base your FSAE design on, it is certainly not a waist of time. From personal experience, i can assure you that the problems involved will teach you enough about engines, Matlab and Simulink to make the time spent on R&D more than worthwhile. Afterall, FSAE is still an educational exercise, not proffesional racing.

[Adambomb]

Agreed, given Tom's 2nd post it sounds much more feasible, that's some critical info left out of the first one! I believe js10coaster's post sums up what the problem looked like without the details. Moral: ask a vague question, get a vague answer.

With that out of the way...sounds like an interesting project though. I imagine it would be relatively simple, probably just need vehicle mass, drive ratios, torque curve, and maybe some simplified tire data if you wanted to go that deep. Make a quick and dirty algorithm for launch and shifting, and voila!

[tom888]

basically i need an engine that has throttle and load (from car) as inputs, and torque and rpm as outputs.
ive adapted a model from a far larger engine, but when i add the load from the car the engine goes crazy. without any load it gives the right torque/rpm curve.

the way ive designed the car (everything without the engine), is that it gets torque and rpm from the engine, goes through the transmission and down to the ground. the rpm is to control the CVT ratio. torque is used to find the car's motion. (ive added a function to work out rolling resistance air resistance etc). the car then sends the load back to the engine.

David, as of yet our R6 engines are still being rebuilt so i dont have the data for them but im using the data from our CBR 600F4 engines and they give a max torque of 62Nm at 10100rpm.
i was actually going to use an IVT but the uni prefered me to buy a CVT (ive also used CVTs in mini Baja so have that little bit of added knowledge there). IVT does seem to look more reliable though.
thanks for all your help so far. very much appreciated...

tom...

[tom888]

thanks for all your help guys.

ive just made the engine. it seems to work well, so alls good...

[SR-Mike]

I don't think you need to over complicate this. All you really need is a torque curve at WOT. Then you can use rpm as an input and torque as an output. If you feel that you must introduce a throttle input then I think it would be a valid assumption just scale the torque output at any given rpm linearly with throttle position.

If it is purely comparative performance that you are looking at and the absolute performance is of less importance, then as long as your models are semi realistic (engine, tyres etc) then they should be sufficient. Just make sure everything else remains constant between tests.

[tom888]

mike, thanks for your input. i was going to try do it that way, but for what i need a simple torque curve isnt really enough.(trying to acheive an almost perfect simulation).
the model i need and have now made, needs to have torqure and rpm as outputs, and throttle and load as inputs.

the only things that are considered constant are, ground surface and gradient,and air density. pritty much everything else is variable.

(the torque cant just be scaled to meet throttle as it not a linear change)