Two simple rules to improve results.

[theTTshark]

Well done Freddy. I think you've hit the nail on the head right there.

[MCoach]

Freddy,
Care to post your OptimumLap files?
I'd be interested to add in a 2wd EV and aero IC car
If you don't want to give up the track files we can use this one:
Attachment 1186

If these are added, add 4wd IC car. You'll like what you see.

Downforce > all. The power is just there to get you from one corner to the next.

[Z]

Freddy,

Thank you very much for bringing numbers to the forum.

What I find most interesting is how "close" the Briggs-Kart is to the E-AWD. The Briggs has LESS THAN ~7% of the power, but is still ~70+% as fast as the E-AWD.

Also worth pointing out that a very simple kart frame can take much more power and still work well. MCoach's Outlaw-karts (specs given earlier) have similar power to your E-AWD, but significantly less mass. And I recall seeing somewhere that these karts sometimes run the Jawa speedway bike engines (= 500 cc air-cooled single with 80+ hp)...

So, could you run one more kart simulation, with power somewhere between Charles' 8 hp Briggs and MCoach's 80+ hp Outlaw?

What I have in mind for the engine is the JUNIOR JAWA. This is a 250 cc, sleeved and short-stroked, version of the "Senior" Jawa, intended for under-16 year old boys starting speedway racing (the t-shirt reads "No brakes, no gears, NO FEAR!").

The conservative specs on this engine are 30+ Nm up to ~10 krpm, giving 30+ kW (many quote ~45 hp), and 12 krpm redline. This engine has "bare" mass ~25 kg, so make the total car+driver mass = 200 kg (up 30 kg from Charles' car). This should cover the bigger wheels, mandatory safety stuff, etc., of FS cars.

If I were to build this car (which I am sure could go under 130 kg dry in "all-steel" version), then I would fit it with a two-speed gear-box, with low GR = ~15:1, high = ~9:1 (for the typical R = 225, 10" tyres). If you want to do it as a single-speed kart, then maybe GR = ~12:1.

My simulations of the 2-speed JJ-engined car, with ~65%R, suggest it can get just under 4 seconds in Acceleration, depending on set-up details such as ride-height, anti-squat, etc. A turboed version of the JJ (which has the same bottom-end as the 500 cc Jawa, so can take the same power), should comfortably beat the 3.4 seconds you have for the E-AWD (ie. ~80 hp JJ in a RWD-only car).

Lastly, since 2005 I have always pushed the "brown go-kart, WITH AERO-UNDERTRAY". As MCoach repeated above, the whole point of building a simple (= "no-bling" = "brown"), lightweight (= "go-kart" style) car, is that the lower the mass, the more cornering-Gs can be leveraged from any given level of aero-downforce.

So, to start with some conservative, round numbers, could you model this car with CD.A = 0.5 m^2 and CL.A = 2 m^2. I reckon a lot more DF can be had with a good undertray, and possibly less drag (for less Fuel usage). But that might entail a bigger "Cost" risk. Chasing higher numbers first time around means more time taken for design, and more fiddly aero parts to make, so much later build-finish, so less time for durability testing, so bits start falling-off at comp, so (maybe) LOSE MANY POINTS.

(I was going to have a long rant about "Cost" = "Risk". No time now, but deep thinking students might want to read Sun Tzu's "Art of War", Chapter 4.)

Anyway, the above "2-speed Junior-Jawa FS-car/kart with aero-undertray" would be relatively easy and cheap to build. And it should also have exceptionally good fuel economy.

But ... HOW FAST IS IT?

Z

[rory.gover]

Z's JJ does FSAE-A 2016 track at a 86.7 sec pace (with equiv. R25Bs 10" on 8" wide rims). Fuel consumption is estimated 0.144 L/lap. Concerns here are the 2.0 m^2 to 0.5 m^2 ratio (L/D = -4, okey dokey), as well as the added complexity of a non-integral gearbox and turbo doesn't seem to me as typically "brown".


Interestingly, the single speed 12:1 ratio does a 86.9 sec lap, even though limited to about 77 km/h, and fuel usage plummets to below 0.13 L/lap. With a more readily achieved L/D (~2), results don't get too much worse.


These are pretty well comparable with the numbers you get from a 190 kg (+ driver) with a big single or twin with more achievable aero targets. Fuel mileage may vary. N.b. you need a well-trained driver to get these results, the sim is spec driven rather than design driven etc. Didn't take into account that JAWA burn Methanol, so there's that.

[Tim.Wright]

Tim W,

I have looked a bit more at the VI-Grade comp, but it is taking me ages to download some of the stuff. So ... can you fill in some of the details?

The VF2017 Rules/Regs seem straightforward enough, only 4 pages!

The idea of competing in Accel, Skidpad, and Autocross events is also good, because it makes it harder to "optimise" a single vehicle design.

I also agree very much with the inclusion of a FUEL EFFICIENCY score (essentially a "Cost" on OTT designs), because it again makes it harder to find an "optimum". Just chasing maximum horsepower or biggest wings becomes counterproductive.

Question: I noticed that there are "cone penalties", so is this a "Driver-In-the-Loop" simulation? Something like the video games (GTA?)?

If so, then great, because that is the type of VD sim that I am now developing (slowly, in spare time). But my general proposal for a Virtual FS comp could also use the more conventional "computer drives the car" simulator.

I don't know so much about the competition but I use that same software for simulator/simulation work.

From what I understand the manoeuvres are all closed loop using the virtual driver included in the software. A part of the project is actually calculating the best trajectory and tuning the virtual driver's control parameters. The cone penalties I imagine are imposed when you stray more than X meters from a nominal centreline of the track.

[Charles Kaneb]

Freddie,

Thank you for the simulation. Looking at the output graphs I notice two things:

1) I do not think I'd bring a 15/59 gearset to a competition featuring courses like that. I'd aim to hit the rev limiter at the end of two or three of the straights. Would a 15/64 reduction, giving me about 10% more acceleration, change anything?
2) Even in <30 km/h corners, my kart is shown to have no cornering speed advantage. The point-mass simulator uses a single low-speed maximum lateral acceleration figure to represent all effects on the car (grip losses through roll and weight transfer, the need to use the grip to perform work to change the rotational kinetic energy on turn-in) etc. 1.5g is in concession-kart tire territory; on a "green" surface I get 1.9-2 g midcorner. Can you plot sensitivity to this?

Even if a FSAE car built along these lines can't use an MG Yellow tire, the tires used for a light car are very load-sensitive and on something that light would be well over 1.5 warm.

Sincerely,

Charles

[FrederikWe]

Charles, Z,

sorry for just making one statement and then stop replying to the thread.
I'm quite involved with the Setup of our 2017 car and right now pretty busy with fixing understeer Problems.

If i find time, i'll update the results with some other cars and also your finetuning of the Kart.

[BillCobb]

I'm quite involved with the Setup of our 2017 car and right now pretty busy with fixing understeer Problems.

This refers to a FSAE car or your Kart ? If it's the FSAE car, a discussion about this situation would be interesting to the Peanut Gallery as well as the players.

[Z]

... I'm quite involved with the Setup of our 2017 car and right now pretty busy...

Freddy,

Take your time.

I know exactly how you feel. (Sighhh ... now to repeat yet again details of rising/falling-rates on other thread...)

Z

[noah]

Tim W,

I have looked a bit more at the VI-Grade comp, but it is taking me ages to download some of the stuff. So ... can you fill in some of the details?

The VF2017 Rules/Regs seem straightforward enough, only 4 pages!

The idea of competing in Accel, Skidpad, and Autocross events is also good, because it makes it harder to "optimise" a single vehicle design.

I also agree very much with the inclusion of a FUEL EFFICIENCY score (essentially a "Cost" on OTT designs), because it again makes it harder to find an "optimum". Just chasing maximum horsepower or biggest wings becomes counterproductive.

Question: I noticed that there are "cone penalties", so is this a "Driver-In-the-Loop" simulation? Something like the video games (GTA?)?

If so, then great, because that is the type of VD sim that I am now developing (slowly, in spare time). But my general proposal for a Virtual FS comp could also use the more conventional "computer drives the car" simulator.

I would also be interested to hear any other interesting details of how this went.

Anyone from Adelaide care to comment?
~o0o~

Tim P,

I sense much confusion in you.

It would help if you listen to (or read) what others have to say, and then CONSIDER such, before blurting out comments like "...doubt it would be any use...".

But you did get this bit right.

Which is the theme of this thread.

Selecting "the right concept" is also the reason for doing good simulations, in that it is pointless aiming for an exotic, super-spec, car, when the sims show that a much simpler and quicker-to-build car, is very nearly as fast. And it uses less fuel, so it wins on points anyway!

Z

Sorry I'm late, but a little on the Vi-Grade Competition. My team participated last year and full disclosure is partnered with Vi-Grade for software access. It was a very enjoyable competition, although it was hard to get members engaged as most were focused on the physical car (I was the only one from the team who directly participated.). The program has DIL capability, my team and I are are working on fully implementing it. The competition itself does not use dil, the cone penalties are for the car driving off the designated path. If you (or anyone) has any specific questions about that competition feel free to ask.

Noah