1000 HP dragbike with Ultracapacitors

15327 Views 35 Replies 8 Participants Last post by Karter2, Nov 8, 2016

kennybobby

Discussion Starter · Oct 18, 2016 (Edited)

2

i want to split out the discussion from the other thread to focus on the technical details of the fast-charge.org dragbike project.

Sifting thru their blog pages seems to be the only way to find little bits of info about the details.

Their goal is to do 1/4 mile in 6 seconds at 200 mph, and they call out 1000HP. The electric power supply will come from ultracapacitors, looks like the Ioxus Titan 60 x 108mm axial cells, 2.7/2.85V-2000F, weighing 390 gms each. The drawing indicates 3px210s, so it's ~600V pack (somebody double check my count on that). 630 cells x 390 gms = 540 lbs per dcb.

No word on the motor, but some about the gearbox:
a single speed reduction gearbox, using two sets of 31:59 gears to give an overall ratio of 3.81:1. This reduces the motor's peak speed from 22,000 rpm down to 5,750 rpm, after which a standard motorcycle chain/sprockets set gives us 2.5:1 reduction to the rear-wheel rpm of 2,300.

See less See more

1 - 20 of 36 Posts

kennybobby

Discussion Starter · #2 · Oct 18, 2016 (Edited)

Calculatus eliminatus:
[edited]
[The acceleration for a 6 second 1/4 mile is 73.33 ft/sec^2, or about 2.27 g., from x=.5A t^2]

Take [their advertised] 1000 Hp over 6 seconds to determine the energy (or work over 1/4 mile) to determine that the [force on the accelerated mass] is 2500 lbs.

[1000 hp x 550 x 6s = 3,300,000 lb-ft, or 1,243 Whr.]

[Divide the 2500 lb force by the acceleration to get the mass = 34.1 and the bike weight is ~1100 lbs.]

Equate the kinetic energy, [1243 Whr] to the capacitor energy, 1/2 CV^2, using 3p = 6000F and 6000/210s = 28.6 F for C, it looks like they are assuming a voltage drop of only 9.4 V in the pack during the 6 second burn.

dcb

· #3 · Oct 18, 2016 (Edited)

also they have stated they want a sub 7 second bike (1/4 mile), comparable to the rocket. The rocket has a 14.kwh pack that makes about 1 megawatt for 7 seconds or thereabouts, and weighs 250 pounds. Basic physics would mandate that replacing the batteries with capacitors (and any necessary converters) and expecting the same performance would require that the energy output at that power level divided by the weight of the energy source (batteries vs caps and converters), needs to be equal.

I recon the caps alone weigh 541 pounds, and 1000hp is %25 less than 1MW. So already the energy source is over twice as heavy and less powerful than the current NEDRA leaders packs, and 28F really sounds like it is gonna come up short, but I haven't dug into the farads required much (seemed pointless given all the other red flags).

F=M*A

edit, I'll see if I can do an ltspice simulation of 28F @600v putting out 1000hp (even though it needs to put out more than 2MW for 7 seconds to be competitive on a density basis)

See less See more

kennybobby

Discussion Starter · #4 · Oct 18, 2016 (Edited)

just trying to run back-of-the-envelope numbers to see where/if any assumptions run off the tracks...

so they need a motor that turns 22,000 RPM and makes 1000 HP, so 238 ft-lbs of torque at 22k using 1255 Amps from a 600V pack.

1255 Amps/3p is about 420 Amps per cell over 6 seconds is 2520 A-s or 2520 Coulombs. Each 2000F cell holds 5400 Coulombs at 2.7V, so theoretically there is enough charge. Another way to look is the entire pack, 28F x 600V, holds 17,160 Coul and a 6-second burn of 1255 Amps is about 7,600 Coul.

Now here is where the problem surfaces, when the cells give up their charge the voltage drops. From a best case 2.85V x 2100F = 5985 C full less the 2520C burn leaves 3465 C. Now what is the capacitance of the cells? If 2000F, then the voltage will be 1.73 per cell, or a sag/drop of 1.11 V per cell. Over 210 cells, the pack will sag 234 Volts, which is a big difference to the 9.4 V sag used above to determine the capacitance required for the pack. This is a big deal breaker, plus can an inverter be made to tolerate this much sag and still drive the motor with the required power to make the run?

i think this should be tested with just one cell to prove whether or not it sags under 420 Amps for 6 seconds (and how hot it gets).

See less See more

dcb

· #5 · Oct 18, 2016 (Edited)

ok, got it sorted, 28F @ 600v is *just* enough to make 1000hp for about 7 seconds on paper, but it looks like they are making some terribly naive assumptions about being able to convert a 3 volt source at 300,000 amps at the end of the race (plus extra weight and conversions and that they are 254,000 watts short of the rocket).

edit: so it starts at 600v @ 1250a, and at 6 seconds it is at 3750 amps at 200v, and then it falls off a cliff, surely before the race is over.

See less See more

dcb

· #6 · Oct 18, 2016 (Edited)

there is also an increased risk of reversing a cap with no margins and over a 3+x voltage drop over a string of 210 caps @ thousands of amps.

edit: apparently that isn't catastrophic, possibly some capacitance loss. http://www.cooperindustries.com/con...pacitors/BUS_Elx_App_Notes_App_Guidelines.pdf

Karter2

· #7 · Oct 18, 2016

Just a couple of points..
Did they clearly statr the pack configuration of 3p, 210s ? Or is that a guess ?
On their facebook page they stated 640 caps in the pack ?..just a estimate ?
How sure are you they are not using the 3000F caps ?

Im not being critical here, just trying to understand why yesterday you were suggesting it would need 22,000 of the 3000F caps to do this, but now it looks like they are a lot closer with <1000 caps ??

dcb

· #8 · Oct 18, 2016 (Edited)

yah kenny threw out the 3600F figure in the other thread, but the simulation (rough approximation) and his research (digging through blogs instead of a scientific or engineering style organization of testing and data by the folks making the claims) have indicated that it is off, but 28F at 600V is still quite a bit short, if you were to take the existing NEDRA leader and replace the battery with 28F ultracaps and a boost converter made of unobtanium.

edit: I don't think it matters if they are 3000F, assuming they are proportionally heavier. ultras have better power density, but it takes energy to do a 7 second 1/4 mile.

Karter2

· #9 · Oct 18, 2016

Even avoiding the calculations, i dont understand how they dont see the simple fact of the pack weight is a big hint that something is wrong with the basic concept.

dcb

· #10 · Oct 18, 2016 (Edited)

more swags on the rocket bike:
http://www.nitrousexpress.com/oldwebdocs/hpcalculators.htm

if you put in 2000 lbs, and 1340 flywheel hp(battery hp?), you get a 7 second 1/4 mile. So the rocket bike is probably less than 2000 pounds w/rider, that 13 inch GE motor though... So the target is somewhere around the 1MW:2000lbs ratio (sustained for 7 seconds), which doesn't *quite* jibe with the Calculatus in post 2 (746kw @ 2600 lbs = 6 second 1/4 mile). There's a lot of drag at 200mph though, so I dunno how the calculator works.

edit: here is all that page does, I don't see drag factored in, ah well, expect more nasty guesses

Code:

document.input.et.value=Math.pow((document.input.weight.value
    /(document.input.hp.value*.854)), (1/3))*5.825;

kennybobby

Discussion Starter · #11 · Oct 19, 2016 (Edited)

Karter2 said:
Just a couple of points..
Did they clearly statr the pack configuration of 3p, 210s ? Or is that a guess ?
On their facebook page they stated 640 caps in the pack ?..just a estimate ?
How sure are you they are not using the 3000F caps ?

Im not being critical here, just trying to understand why yesterday you were suggesting it would need 22,000 of the 3000F caps to do this, but now it looks like they are a lot closer with <1000 caps ??

# of cap cells was from reading the blogs, and then:
i just counted the 3p modules as shown in their diagram above, and
i counted the number of boxes in their mock-up picture, 8 top row, 16 middle and 18 bottom. Each box holds 15 cells. Both methods gave 630.

i didn't look at the facebook, so unaware of 640.

Cap size 2000 vs 3000:
This was determined both by weight and length.
A ruler was shown in the mockup photo on the boxes, ~3 boxes per 12" ruler, so about 4" length per cell (the 2000F cell is 108.1mm),
The caption under the shipping pallet indicated 250 kg, the 2000F cell weighs .39 kg, the 3000F cell weighs .51 kg. 640 cells x .39 = 249 kg. Maybe they got 10 extra to use for prototype and spares.

i think yesterday we were using standard EE math for series and parallel strings of capacitors to get to the 3600F, which was calculated by equating the kinetic energy of a 2500lb bike at 200 mph with the Capacitor energy with a 50V sag--this was before parsing thru all the blog pages trying to find tech details.

After finding the hive drawing and photo with the 3p connecting plates, etc. it was possible to determine the actual capacitance and voltage of a 3p210s pack made using the 2000F cells. This configuration yields a total capacitance of 28F at 600V, and could only work if the pack voltage only sagged ~9 Volts. Feel free to double check and challenge my numbers, i don't want to make a mistake in the basic data that affects the calculations and predictions.

See less See more

Karter2

· #12 · Oct 19, 2016

No, thats ok Kenny, thanks for the clarification.
I am not too familiar with ultracap powerbanks so i just wondered why the big discrepancy between what we thought would be required conpared to what they are doing.
It still seems as if they are way off in their design thoery, which is even more surprising when you look at who they are working with.

kennybobby

Discussion Starter · #13 · Oct 19, 2016

i'm puzzled by it also, hence the investigation--do they know something that we don't know?

It's almost as if they are using the caps as battery cells with no consideration for the reduction due to stacking in series...

dcb

· #14 · Oct 19, 2016 (Edited)

more comparative data points for drag bikes (roughly speaking):
http://www.knfilters.com/news/news.aspx?id=2461

1030 lbs (rider?)
1000 hp
2 speed
< 5.8 second 1/4 mile @ 245mph
37 pounds of nitromethane per run (mix?)
which is maybe 3.5kwh equivalent, but we know they ICEs aren't real efficient, giant wall of flame from disassociated hydrogen from water in exhaust. (5.765 seconds at 227.10 MPH here https://www.youtube.com/watch?v=npNkwVw-V5I)

note, I learned there is some funny business too, some (all?) top fuel races are only 1000 feet instead of 1320 feet, so WTF.
https://en.wikipedia.org/wiki/Top_Fuel

<sigh> perhaps not top fuel dragbikes though. http://www.cycledrag.com/top-fuel-motorcycle-1000-drag-racing-when-is-now

See less See more

Frank

· #15 · Oct 19, 2016

I'm on a tablet so research is hard, but a number of years ago a group of students raced an EV1 (I think) in a 1/4 mile drag race. They pushed it to the line to converge energy.

Also, without following numbers here, don't caps decrease dramatically in voltage as they yield energy? Result would be massive sag, no?

Frank

· #16 · Oct 19, 2016

http://universe.byu.edu/2004/05/11/students-race-ultracapacitor-for-first-time/

http://electrifyingtimes.com/maxwell_EV1.html

I'm sure there's more out there on the byu project.

dcb

· #17 · Oct 19, 2016

Frank said:
Also, without following numbers here, don't caps decrease dramatically in voltage as they yield energy?

Yup, I put a constant power graph in post 5., granted races aren't really constant power, but they aren't constant efficiency either, so it is a first order approximation.

Frank said:
I'm sure there's more out there on the byu project.

they changed everything though, i.e. cut the weight in half by removing who know what and the EV1 has no lithium performance figures to compare to. I've been looking at other drag bikes because they should be similar in weight and aerodynamic drag and have some performance numbers to compare to.

zaxxon

· #18 · Oct 19, 2016 (Edited)

This simulation shows if one had a controller that could operate with an input range from 600 to 300 volts and supplied 2000 amps at 300 volts to the motor, then one would expect to get around 800 HP for 4 seconds using a 28F capacitor charged to 600 volts. (Note: This is for a linear controller. Result should be better with a switching controller)

See less See more

zaxxon

· #19 · Oct 20, 2016

With a switching type buck converter, this simulation shows one could expect to get around 800 HP for 6.3 seconds using a 28F capacitor charged to 600 volts.

See less See more

dcb

· #20 · Oct 20, 2016 (Edited)

ltspice is fun, but you might be over complicating it a bit

the controller design isn't relevant per-se, indeed changing everything around will invalidate the comparison between caps and batts (excepting for possible power conditioning of the cap output to suit motor/controller requirements).

600V 28F 800HP(596800W) discharge (ideal components)= 300V @ 6.3 seconds, check.

See less See more

1 - 20 of 36 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.