Brushless Drag Bike!

[liveforphysics]

If it was for something that just drops out of the sky and people die, like a real helicopter or something, then I wouldn't say they were under-rated, and would try to stick to the 45-90c thing.

For a dragbike looking to break all the electric records though, and will only be loading the cells for 6-7seconds at a time, 100C will be no issue at all, and I wouldn't hesitate to burst them to 150-200C at the launch.

After all, it's a dragbike, so range and capacity doesn't matter, and you should be looking to push the limits of technology.

Each 6s5Ah 45-90c Nano-Tech pack can be counted on to provide 11kw continuous, and >15kw at the launch.

If you want a pack that does 700kw constant, 1000kw launch, you just need 64 packs. With no special volume discount, that's $90/pack, or under $6,000usd to get the battery side of things handled.
132lbs of batteries for the pack. That's way better than we originally estimated, in cost, volume, and weight.

 

[Arlo1]

Very awesome luke!
I worked out the math and the total KW is still the same.
The weight of the bike is ~740 lbs without rider! And 950lbs with rider!
These are all calculations at this point we will not know what we need for total KW till we make a couple runs because the nitro bikes are so hard to get real numbers from!
I do belive that a brushless drag bike is the ticket and will be the Real eye opener!

 

[Doctorbass]

64 packs is 8 x 8.... so just imagine how SMALL could be a 1MW pack.. like 18" x 18" x 7"


Doc

 

[liveforphysics]

64 packs is 8 x 8.... so just imagine how SMALL could be a 1MW pack.. like 18" x 18" x 7"


Doc

Yep. It's pretty freaking amazing.

Poor Kilacycle and that new faster bike that are stuck with A123 sponsors.

Just spending the 6 grand to buy there own battery would save them so many 10's of thousands in chassis/tire/motor/controller costs in the long run from having a 200+lbs lighter and way more compact bike.

 

[Kingfish]

Hi Arlo

Thanks for inviting me on over to your project

I’d like to work of a fun little example of using 14-inch OD x 8-inch ID x 1.0-inch thick N45 arc magnets in a Non-Halbach Array (NHA) and determine how many rotors we can fit onto your bike.

Could you provide these figures when you can please:

• Rear Tire size/OD
• Rear sprocket: Width & teeth.
• Measure the width of the engine (estimate if the motor is gone); this will be the maximum width that we’ll work with.
• What is the maximum speed you expect to reach.
• Motor Heat will be a problem and we may have to get creative.
• Do you normally have fire-suppression loaded on these bikes?

Best, KF

 

[fizzit]

Well if you're only going for 6 seconds, you could just superchill the motor beforehand with a bunch of dry ice, and then when it's time to race, run the cooling system to spread the cold

 

[Arlo1]

The red bike we have for sure is proly this one http://www.weekendframe.com/
The nitro frame I might be able to get is like this http://www.weekendframe.com/

The sprockets dont mater we will build the motor and gear it how we need.
The red fame is a 29x9x15 tire
And the nitro bike runs a 31x14x15 tire so the rim size on both is 15 inches.
The area in the frame is 30" along the top, 17 along the bottom, 16" high in the back and 13" wide thats on the red one the nitro bike is bigger.
A 18" wide motor would not be a problem!
The motor will likely not be the bulk in the frame I feel the controler will be the huge part!
This is the chain they run http://www.pingelonline.com/ek_630shbz_chain.htm its 3.70 a pin

 

[x88x]

This sounds like an awesome project, I'll definitely be watching! I'm pretty new to the whole battery situation, but the Nano-tech's sound great for a dragster application, especially for the price. I wouldn't be surprised if you can land a battery sponsor though. I know when White Zombie got their Dom Kokam sponsorship, they said Dow was quite excited to have a PR outlet to prove the durability and capability of their batteries.

 

[Kingfish]

Arlo, thanks for the details – it really helps!

Wheel RPM Calcs:
We need someplace to start, so let’s do the math for a 29-inch wheel at 100 mph.

• Tire Diameter (d) = 29-inch = 2.4167 feet
  Circumference (c) of d = π d -> π * 2.4167 = 7.5922 feet (per revolution)
  If 1 mile = 5280 feet, 100 miles = 100 * 5280 = 528000 feet.
  With 1 hour = 60 minutes, rpm = (528000 feet / 7.5922 feet) / 60 = 1159.1 rpm or 19.32 rps at constant velocity.

Avoiding the games we can play with gearing, let’s pretend this is a mega-hub drive with 18-inches of width to play with; the wheel and motor will turn at 1:1.

A large number of rotor-stator pairs will be added, and the motor will rotate at a ratio of 11.591/mph at constant velocity presuming no tire-to-ground slippage.

A Little Drag Race:
Let’s calculate for 200 mph in 6 or 7 seconds because that’s at least a noteworthy beginning. Using uniform acceleration…

• s = [(u + v) * t] / 2
  where
  s = displacement
  u = initial velocity
  v = final velocity
  a = uniform acceleration
  t = time.

We solve for s so that we can get to a in the next step.
Let’s set t to the worst case at 6 seconds. We know u will = 0, therefore…

• s = [(0 + 200) * 6] / 2 -> (200 * 6) / 2 -> 1200 / 2 = 600 feet.

If v = u + at, and we know u = 0, then v = at, or a = v/t -> a = 600/6 = 100 feet/s^2
Now we take a / c to get the wheel acceleration: 100 / 7.5922 = 13.171 rps

Presuming that at the start of the event, when t = 0, the tire is not spinning; we have zero rotation, is that correct?

• At the end of the first second, the tire is rotating at 13.171 rps.
• At t = 2, it’s 26.343 rps.
• By the time we reach 6 seconds 600 feet later at 200 mph that wheel will be spinning at 79.028 rps which is 4741.7 rpm

Can someone double-check my math please?

Arlo, what is the target speed you wish to reach in what time?

Fun, huh! KF 8)

 

[Arlo1]

Well as for targets. I use the nitro harley as an example the fastest one at Vegas was 222mph in 6.27 seconds so faster then that, because in all honesty I want to kill kilacycle! Is 300 not possible? I realize we need a place to start but when you ask a target I don't aim for something kinda cool I aim for world records!

 

[Arlo1]

I posted the gear ratios earlier in this thread I think. As for wheel rpm it's not 4700 rpm. It's a 31 inch OD tire and it grows a lot with rpm! So from my calculations the wheel rpm at 200 mph is 2168!

 

[Arlo1]

And for a 29 inch OD tire at 200mph it's 2318 rpm and remember both of these numbers are with zero growth!

 

[John in CR]

If it's a hubbie it will stop the frock motor BS once and for all. The front end will need weight anyway, so why not one in the front wheel too with traction control built into the controller?

 

[Kingfish]

Arlo,
I picked the 29-inch wheel because it will spin faster than the 31-inch given the same motor; it provides the worst case

Calcs for the 31-inch wheel…

• Tire Diameter (d) = 31-inch = 2.5833 feet
  Circumference (c) of d = π d -> π * 2.5833 = 8.1158 feet (per revolution)
  If 1 mile = 5280 feet, 100 miles = 100 * 5280 = 528000 feet.
  With 1 hour = 60 minutes, rpm = (528000 feet / 8.1158 feet) / 60 = 1084.3 rpm or 18.07 rps.
  At 200 mph at constant velocity it is 2168.61 rpm. My number matches yours.
  The 29-inch tire at 100 mph is spinning at 1159.1 rpm, and at 200 mph at constant velocity it is 2318.2 rpm. Again my number matches yours.
  We’re in agreement on this part. :wink:

Accleration:

• This time I used a different formula for calculating acceleration…
  a = (vfinal – vinitial) / t
  We know vinital = 0, therefore
  a = vfinal / t -> (10560 ft/s) / 6 seconds = 48.89 ft / s^2
  29-inch wheel = 7.5922 ft/rev -> 48.89 / 7.5922 = 6.439 rev/s^2; after 6 seconds = 38.636 rev/s^2 or 2318.2 rev/min^2 … which matches a 29-inch wheel with an angular velocity of 200 mph!

OK, I see the error of my ways. I will at this point yield and presume that the 31-inch wheel will resolve similarly.

If you don’t care about the distance we can move right along.

Rerun the calcs for 300 mph:

• Circumference of 29-inch wheel = 7.5922 feet (per revolution)
  300 mph = 5280 feet * 300 = 1584000 feet.
  rpm = (1584000 feet / 7.5922 feet) / 60 = 3477.3 rpm or 57.95 rps.
  Circumference of 31-inch wheel = 8.1158 feet (per revolution)
  rpm = (1584000 feet / 8.1158 feet) / 60 = 3252.9 rpm or 54.22 rps.

Regardless, if we keep the motor and tire geared at 1:1 we’re looking at a motor that can spin at 3500 rpm or faster, agreed?

~KF

 

[Arlo1]

If it's a hubbie it will stop the frock motor BS once and for all. The front end will need weight anyway, so why not one in the front wheel too with traction control built into the controller?

The front wheel is a waste of time and weight. The front wheel on a perfact nitro Harley is in the air for another 40 feet after the 1/4 mile mark!

 

[liveforphysics]

Fast motorcycles are unicycles.

 

[Kingfish]

Arlo, forgive me - I'm a noob with this sport. Are you planning on 300 mph in 1/4 mile?

I just ran the calcs and the acceleration will be 73.33 ft/s^2 and you will reach the end of the 1/4 mile in exactly 6 seconds, pulling 2.3 g's :wink:

What a ride! KF

 

[Arlo1]

Arlo, forgive me - I'm a noob with this sport. Are you planning on 300 mph in 1/4 mile?

I just ran the calcs and the acceleration will be 73.33 ft/s^2 and you will reach the end of the 1/4 mile in exactly 6 seconds, pulling 2.3 g's :wink:

What a ride! KF

Haha yeh 1/4 mile hence the drag bike build!
The calcultations will be more around 3-4 gs for the start I think they run a 180 mph 1/8 mile!

 

[Kingfish]

No wonder the seat is shaped to keep the pelvis in place! I was just thinking about that…

Tomorrow (Monday) I have to study & prep for a client, move some zeros and ones around, so I might not be as attentive: Work before play. Though I am thinking about this in the background ~ you can be sure of it!

Cheers, KF

 

[Arlo1]

No wonder the seat is shaped to keep the pelvis in place! I was just thinking about that…

Tomorrow (Monday) I have to study & prep for a client, move some zeros and ones around, so I might not be as attentive: Work before play. Though I am thinking about this in the background ~ you can be sure of it!

Cheers, KF

I think about this all the time......
I work way to many hours...... But I will be getting going on this!
So far the three big hurdles are.
#1 Money althogh sponcers will be easy to convince as this gets going.
#2 Controler that can push 750000 watts at 100 volts or what ever voltage I just chose 100 because its a number that seems good for good mosfets!
#3 Time I will make time as I need but it is very tricky although as I get going on this it will prove my drive and enthusiasim!
Engine design doesnt scare me I can build anything and its time to design and build an engine!

 

[Kingfish]

Arlo, selecting 100V is a good start. Are we sticking with the 700 kw or the 1000 hp? Allow me run with the hp for the moment.

1000 hp = 745.7 kw

Given P = τ * ω, solve for τ:

• ω = (2 Ï€ *rpm)/60 = rads/s
  Changing units to metric…
  300 mph = 483 km/h
  Circumference of 31-inch tire = 8.1158 feet / 2.4737 m (per rev)
  Linear Velocity = 483 km/h / 3.6 = 134.112 m/s
  rps = 134.112 m/s / 2.4737 m = 54.215 rps
  ω= 54.215 rps * 2π = 360.64 rads/s
  Resolving τ = P/ω -> τ = 745700/360.64 = 2189 Nm

Ke = back EMF constant; Solve for Ke:

• Ï„ = Ke * ω -> Ke = Ï„ / ω = 2189 / 360.64 = 6.07 V/rads/s

If Kt = Ke when expressed in the same units allows us to calculate current (I).
Solve for I:

• I = Ï„ / Kt -> 2189 / 6.07 = 360.64 A
  Given P = I * V, solve for V:
  V = P / I -> 745700 / 360.64 = 2068 V.
  I don’t see that as an option, so let’s try again using 100 Volts:
  I = P / V -> 745700 / 100 = 7457 Amps.

Something tells me that 100 V will not work. I think we need Luke to chime in here and suggest some FETs that can take more voltage.

Also, a brief calculation suggests that if we use the 14-inch OD windmill magnets that the real OD of the magnets will be closer to 16-inches due to the air gap between the magnets. I haven’t done the shear on the rotor material and I think this will affect the air gap more than any other factor; we don’t want the magnets flying off. For the moment I will model in FEMM with an experimental gap just to see where we are in terms of Tesla – probably tomorrow night.

Check my math; keep me honest!
Cheers, KF

 

[Arlo1]

Luke and I have discussed fets I'n great depth and the fets rated for 100 volts are the best for low resistance there is fets rated for 150 volts but their resistance is higher! Above 150 volts we are no longer dealing with fets!

 

[Kingfish]

Arlo,
friend I must admit that I am struggling to wrap my head around pushing 7457 Amps at 100 V through a single-axle motor. A good engineering solution requires review of all possibilities, an unconstrained development of ideas.

Some quick math for visualization:

• Convert 7457 Amps into 3-phase current = 7457 Amps / √3 = 4305 A
  If we’re stuck in the mud using 100 V, then let’s apply this device:
  IRFP4468PBF 100 V, rated up to 195 A, TO-247
  Calculate the number of FETs/Phase at 4305 A:
  4305 A / 195 A = 22 FETs * 2 / Phase = 132 FETs total (an optimistic count)

That’s 132 points of potential failure!

What if we used: XFK360N15T2 rated at 150 V, up to 360 A, TO-264?

• I = P / V -> 745700 / 150 = 4971.3 Amps;
  4971.3 Amps / √3 = 2870 A;
  2870 / 360 = 7.9 or 8 FETS => 8 * 2 * 3 = 48 FETs

Better, but not great.

What’s the resistance problem again? I don’t get it and the reason is because there will be a lot more circuit resistance in a board supporting 132 FETs over one supporting 48 plain as day. And we haven’t really touched the surface yet. I understand completely that we can’t push these devices to the max rating, that we have to build them to be more robust and withstand tremendous loading.

The point I’m trying to make is we’re starting from scratch, we can do what we want on paper, let’s leave all options open because my goal is to reduce the amount of current running through the motor so we don’t fry it or anneal the magnets. The controller is second-banana in this regard.

Can you be amenable and allow me/us to provide all possible constructs?

Pragmatically, KF

 

[Arlo1]

I'm open to all suggestions but what we have is a couple great fets rated at 100 volt I was considering 300-400 4110 fets. Or 24 of these http://ixdev.ixys.com/DataSheet/L401.pdf rated at 680 phase amps per fet and the motor might need to be more then 3 phases! I think voltage spikes and inductance and fet load will decrease if more then one phase is pulling/pushing at once! So a 9 phase would have 3 pulling at once 1 just starting to pull one In the middle of it's pull and one just finishing!

 

[Arlo1]

I'm open to all suggestions but what we have is a couple great fets rated at 100 volt I was considering 300-400 4110 fets. Or 24 of these http://ixdev.ixys.com/DataSheet/L401.pdf rated at 680 phase amps per fet and the motor might need to be more then 3 phases! I think voltage spikes and inductance and fet load will decrease if more then one phase is pulling/pushing at once! So a 9 phase would have 3 pulling at once 1 just starting to pull one In the middle of it's pull and one just finishing!

Actualy I just did a quick calculation and 18 IXYS Hyperfets should be ok!
Now the nice thing is to help controler life and to possibly use less amps We can have all the windings from each phase switch from series to parrallel to make the rpm we need and keep the load down on the fets.
18/3=6 fets per phase Or 18 phase or 2 fets per phase and 9 phase!
Depending on the winding scheme I come up with we might need a fet for each end of the phase but I think I know what I can do!