Rally’s a whole different breed of racing, there’s a ton of other factors contributing to a driftier style where necessary, for example, keeping a turbo spooled means sometimes you brake while mashing the gas pedal, and what better way to keep the revs high than to break traction. Similarly it’s better to follow whatever the guy in front was doing even if it’s slightly off line, as their tires already clear up some of the loose material on the path and leaves harder dirt which is grippier and faster.
I still want to see eMTB rally stages on single track though.
Venturi effect esk8 when? Sorry, I stole your picture Dallas, I couldn’t find one in my camera roll like this so I took the first good one I could find from the picture thread.
This has been implemented successfully in RC so maybe someone could give it a try, probably not a ton of performance given how small our decks but I bet a wider deck like the flux or even wider could generate a few pounds of downforce with a perfect implementation and a huge fan on the rear.
I was thinking about that on my aero build, but the fact that the skateboard deck doesn’t stay flat when turning makes it really difficult to seal the underfloor. An easier and more effective way to improve aero is to add a fairing to route air around your legs.
Also if you want to add a few pounds of downforce just eat a few burgers before a race.
I read that it is very dangerous when the skirts loose contact, because the downforce immedately reduces → loose traction. Very difficult to save for the driver.
But in RC is no driver in the car… so no problem with safety.
do you have an idea for the skirt design? they have to be very variable because of the lean. dont know how this can be done
Great minds think alike I guess . Back in '77, 1 year before the Bt46, lotus used a solid side sill and these kinda nylon brushes that extended to the ground to account for the fact that under-turning the car will lean. You could probably do something like this?
Yeah I’m looking mostly at improving aero. There’s some quick and easy gains like streamlining truck hangers to smooth out airflow or wheel fairings, but a paper mentions that the board only makes up ~12% of drag while 48% of total drag comes from the legs and feet, so that’s where my focus has been. Helmet is 25%, but there’s not a whole lot you can do there.
I’ve been thinking about this for like a year at this point . Im terrible with tech developments like that, my only dumb Idea was a computer controlling 4 individual escs and monitoring slip, Gs, and other stuff to send power to the outside in a turn. I have no idea how to do this though.
This is exactly what I was thinking about. Do it exactly the way a car would.
@lucid_layers vesc traction control is essentially useless. Think of it more as a digital differential that let you choose the max difference in RPM. Pretty much only useful for off-roading and even then not so much.
anyway… the vesc traction control would transfer the power to the inner wheel because its slower.
And as I understand, a “torque vectoring” would transfer more power to the outer wheel.
my car has the bmw “xdrive dpc” which is a kind of active vectoring. and I must say it feels very safe in any driving situation and any surface condition. but I guess that Audi does it even better with their long 4wd history.
Yes, you want power going to the outer wheel since it has more grip. Xdrive is great, the one it sounds like you have is the DPC which has an electronically controlled clutch pack that can change the distribution of power by slipping or locking clutches. I still think that emulation of Acura’s SH-AWD or I guess more accurately the AWD implemented in the Rivian or tesla cars since they will perform more similarly to how our boards work. Audis handle like pigs and all they do is understeer, I’ve driven plenty and you are not missing out on anything, they have 4(?) versions of Quattro currently in production and the only decent feeling one is in the R8, shocker its because it doesn’t have a boat anchor hanging over the front axle like the rest of their lineup.
hmmm really wonder why there is no implementation for that in esk8…
I could think of a “vesc app” doing this or of a ESP32 microcomputer with communication via CAN to the vescs.
These versions from Acura or Tesla, how do they calculate the power ratio for the wheels?
Do they just use the erpm of the motors or also use IMU?
Acura looks at steering angle (could be replaced by a horizontal yaw sensor for boards since that can be calibrated to steering) throttle position, lateral Gs, wheel speeds, braking input, and engine gear/torque, and it puts all these into an algorithm that tells the clutch packs, electric motors, and brakes what to do. Another interesting thing that the Acura system does in some models is overdriving the rear wheels by up to 1.7% faster than the front wheels and then an additional 1% to the outer wheel to shift power rearwards since that is where the weight will be under acceleration and the rear wheels will have more traction than the front meaning there is no point in sending 25% to each corner since that would not be maximizing available grip.
Rivian is slightly different because it has electric motors on all four wheels it can measure how much power it takes to induce slip and from that, it can determine a rough estimate of the frictional coefficient of the surface and then determine max possible power using all the same variables that I listed for the Acura system.
For the Rivian, electric motors have such fine control over power that the VSC/TC system will know exactly how much power is being put down, putting the force into this equation will give us the surface’s coefficient of friction.
F=Wheel force (known)
µ=Surface Fr
W=Vehicle weight (known)
Put that into the formula, F= µ * W, and we can find the Fr of the ground.
we can determine exactly how much force we can put down on this surface without slipping and use the benefits of the super fine control of electric motors to put down exactly that. Couple that with the other information listed in Acura’s system and you have a system that completely replaces the need for skill in cornering, just lean as hard as you can, mash the throttle and the board will figure it out for you, and keep you at that 100% optimization of grip level that no human could ever achieve. This obviously would take an incredible amount of skill and hours to write all these algorithms that honestly would be better spent somewhere else but if you could implement this it would undoubtedly be faster than any human not using this system.
this is a very interesting approach of the Rivian.
calculate the surface fr from slip.
If you compare the electric solution with gasoline driven cars. it sound so much simpler. no clutches and mechanical links are necessary. just sensors and software…
do you know any details how this is done?
No guarantee that I’m capable of programming this. but my background is software developer and I worked a lot with microcomputers in industrial application.
so this might be a topic where I can offer my part softwarewise.
hehe… it sounds like cheating.
but I guess there is a lot of personal finetuning you could do to the algorithms and also you have this weight shift of the rider, that have to be very precise to make the system work as expected.
. Back in '77, 1 year before the Bt46, lotus used a solid side sill and these kinda nylon brushes that extended to the ground to account for the fact that under-turning the car will lean. You could probably do something like this?
. Im terrible with tech developments like that, my only dumb Idea was a computer controlling 4 individual escs and monitoring slip, Gs, and other stuff to send power to the outside in a turn. I have no idea how to do this though.