CeramicSpeed DrivEn - Faster than any chain

[TobinHatesYou]

Belts are less efficient than chains. 2 belts are less efficient than 1 belt.

[bremerradkurier]

Might be something for track racing, assuming it can handle Quad-zilla wattage.

[UpFromOne]

Might also be something for wishful thinking.

[kwakekeham]

Chain drives are already 98% efficient in the optimal chainline at 300W, so something like this is chasing 5-6W, and more realistically 2-3W.

Why isn't anyone talking about this point!!!! It's just shut up and take my money and "the industry needs innovation" happens to be equal to "better throw those standards out the window". Which is fine. However this isn't as big as everyone thinks. I've literally heard people trying to math this yesterday into hilarious proportions.

It's driving me nuts. They literally said they made a 49% improvement over a chain system to achieve about 1% loss... so that means they went from about 5 watts to 2.5 watts.

Here is the math people.

250 watts
chain = 5 watts loss
49% improvement on LOSS = 2.5 watts
1% their lilteral claim = 2.5 watts
Chain therefore = 2% loss.

So, on friction facts bench setup which had tested everything at 250 watts (which is what they used, however in this testing it was dialed up to 350 or 380 for some reason which would likely widen that margin). But obviously they aren't using Molten Speed Wax as their reference which they (Ceramic's speed now owned Friction Facts) tested to be 3.5watts, so the gap if you're willing to wax a chain and this thing with about 28 ceramic bearings, 24 or more which are open to the air (capping it will lose you watts likely putting this less efficient than a chain) is a whopping 1 watt.

To achieve similar results
Wax chain with Parafin or Molten Speed Wax or similar = 1.5 - 3 watts (2ish being normal against most lubes)
Tidy up those cables out front = 1 watts

Done. You now have an efficienty gain of about the same or more. Feel free to send me the thousands of dollars you've have spent on this.

The way the cassette is designed there is one or two exact positions which it could shift. Lets take a worse case scenario of 60km/hr will work out to (rounding here) 8 rotations per second. You have one single position to shift and lets say slop of 2 degrees (overlap teeth for 12 and 13... only one position). The time the actuator needs to move from one gear to another is.... drumroll please... 7 microseconds. The bearings are 5mm tall approx, say a gap between gears of similar, so total movement of 10mm's. Average velocity is 14 m/s. Assuming linear ramp and to achieve this speed, you really need to get to double at the peak so 28m/s to achieve the average. You have half that time needed for acceleration in old school v = a*t, with t being half the time so 3.5 micro seconds (0.00035 seconds) we need accel (and decel) of 80000 m/s^2.

So just eyeballing and comparing it to other bearings I'll just throw in half the size and weight of a 6805 which is 24 grams, but they have 12 on the "moving" element, so 144 grams + structure. I'll be nice, 200 grams total. I won't even include actuator mass or the transformer effect where the effect on inertia of a geared system is actually squared. F = m * a = 0.2 *80000 = 16000 N of force or in mass about 1600 kg. This force isn't impossible to achieve in a system but at the 80000 m/s^2 acceleration... well..... good luck guys.

Probably tear the teeth apart trying to do this any other way, but that's non of my business.

Edit
POWER
P = f * d / t = (16000* 0.01)/(0.0007) = 228 571 watts for a motor driver on my calcs.

[CarlosFerreiro]

Off topic, but it got me thinking, is an different slant on drivetrain development maybe to use single sprocket internal hub gears, with all their lower efficiency issues, and then just add in 20W or whatever of little electric drive compensation to cancel out those losses? Drivetrain renormalization?

[uraz]

Pure innovation

[stevey028]

Chain drives are already 98% efficient in the optimal chainline at 300W, so something like this is chasing 5-6W, and more realistically 2-3W.

Why isn't anyone talking about this point!!!! It's just shut up and take my money and "the industry needs innovation" happens to be equal to "better throw those standards out the window". Which is fine. However this isn't as big as everyone thinks. I've literally heard people trying to math this yesterday into hilarious proportions.

It's driving me nuts. They literally said they made a 49% improvement over a chain system to achieve about 1% loss... so that means they went from about 5 watts to 2.5 watts.

Here is the math people.

250 watts
chain = 5 watts loss
49% improvement on LOSS = 2.5 watts
1% their lilteral claim = 2.5 watts
Chain therefore = 2% loss.

So, on friction facts bench setup which had tested everything at 250 watts (which is what they used, however in this testing it was dialed up to 350 or 380 for some reason which would likely widen that margin). But obviously they aren't using Molten Speed Wax as their reference which they (Ceramic's speed now owned Friction Facts) tested to be 3.5watts, so the gap if you're willing to wax a chain and this thing with about 28 ceramic bearings, 24 or more which are open to the air (capping it will lose you watts likely putting this less efficient than a chain) is a whopping 1 watt.

To achieve similar results
Wax chain with Parafin or Molten Speed Wax or similar = 1.5 - 3 watts (2ish being normal against most lubes)
Tidy up those cables out front = 1 watts

Done. You now have an efficienty gain of about the same or more. Feel free to send me the thousands of dollars you've have spent on this.

The way the cassette is designed there is one or two exact positions which it could shift. Lets take a worse case scenario of 60km/hr will work out to (rounding here) 8 rotations per second. You have one single position to shift and lets say slop of 2 degrees (overlap teeth for 12 and 13... only one position). The time the actuator needs to move from one gear to another is.... drumroll please... 7 microseconds. The bearings are 5mm tall approx, say a gap between gears of similar, so total movement of 10mm's. Average velocity is 14 m/s. Assuming linear ramp and to achieve this speed, you really need to get to double at the peak so 28m/s to achieve the average. You have half that time needed for acceleration in old school v = a*t, with t being half the time so 3.5 micro seconds (0.00035 seconds) we need accel (and decel) of 80000 m/s^2.

So just eyeballing and comparing it to other bearings I'll just throw in half the size and weight of a 6805 which is 24 grams, but they have 12 on the "moving" element, so 144 grams + structure. I'll be nice, 200 grams total. I won't even include actuator mass or the transformer effect where the effect on inertia of a geared system is actually squared. F = m * a = 0.2 *80000 = 16000 N of force or in mass about 1600 kg. This force isn't impossible to achieve in a system but at the 80000 m/s^2 acceleration... well..... good luck guys.

Probably tear the teeth apart trying to do this any other way, but that's non of my business.

Edit
POWER
P = f * d / t = (16000* 0.01)/(0.0007) = 228 571 watts for a motor driver on my calcs.

Well put...could'nt agree more on this one. But still good to see innovation.

[euph]

I agree with all the "good to see innovation" posts. Here and on other sites. However, I agree with the maths and the feeling from the post above; this is no more efficient and in practise likely to be heavier, less efficent, slower and clunkier than what has been the standard forever. People have to try new things to spur development though so I dont want to seem to against this.

Credit where credit is due though, CeramicSpeed have incredible skill at reading the market of cycling and creating a kind of "veblen good". Their components create real gains, I wont deny that. But the average Joe will never be able to see those benefits.

Hopefully I dont start a war with this post. Just thought I would chime in with my opinion. Catn deny that the carbon driveshaft is pretty.

[spartan]

https://www.youtube.com/watch?v=-9gQ1KRhesM

20% lighter. can change gearing without major headaches. integrated powermeter. super $$$$$$$$$$$$$

[alcatraz]

How about some better ways to design a drive shaft. Could be the way to go.

[TobinHatesYou]

Has it been mentioned that this design introduces lateral forces on both the dinner plate/wheel and the shaft+chainstay? Without lots of reinforcement, the first time anyone tries to put a decent amount of torque into this drivetrain, it’s going to jump out of mesh or break entirely.

[ome rodriguez]

[amngwlvs]

You beat me to it; came back to this thread to post exactly that picture!

[TonyM]

Wow! That’s a good one!

[Rogntudju]

Ridiculous, mechanical non sense...