What's new
What's new

O/T - Safe flywheel speed?

Terry Keeley

Titanium
Joined
Oct 18, 2005
Location
Toronto, Canada eh!
Trying to figure out a safe rpm limit for a steel flywheel 6" diameter x 1" thick. I'll turn it from solid so the shaft and wheel are one piece.

Machinery's Handbook (27th Edition) has a formula on pg. 191 in the "Mechanics & Strengths of Materials" section called "Safe Speed Formulas for Flywheels and Pulleys". This says a max of about 17K but it only shows steel with 60 ksi tensile strength and doesn't give the safety factor. I'm planning to use 4340 which is around 160 ksi. I'm hoping to turn it to 30K briefly for use in a small inertial dyno.

Any material engineers here? Anyone point me to some more detailed information?
 
One thing to think of as well is you spindle and its design and manufacture. You'll want every tolerance on your shaft to be a tenth or so. And perfect fits on your bearings. For balancing i would recommend a shop that rebuilds turbo chargers.
One surprising thing about bar stock is that even parts within a tenth are not balanced.
 
Rather than going for more speed, put more material in the rim.

Tom


Good idea from the safety point of view..

Problem is, "V^2" goes up a lot faster than "m". The payoff for more speed is just way better, especially if weight is a consideration, and it usually is.
 
Good idea from the safety point of view..

Problem is, "V^2" goes up a lot faster than "m". The payoff for more speed is just way better, especially if weight is a consideration, and it usually is.

..until it blows up from over speeding.

If this is more than just a simple flywheel, then its time to pull down the design books. Years ago there was "push" for flywheel powered vehicles. A lot a time was spend designing the optimum shape and construction.

Tom
 
How well it's supported and how well balanced it is will influence the maximum safe speed. How are you dealing with those two aspects?

Will be supported solidly with pillow blocks. Going to turn it between centers so it should come out close but will check the static balance carefully before running it.


Glad that guy changed his mind. 1000 lb flywheel at 10K, ya right.

Anyone here run a Solidworks simulation like in that second link?

Basic formulas and a reference here- Solid Disk Flywheel Design | Engineers Edge | www.engineersedge.com

I'd also thing balance will have to be dead perfect and you also get into critical speed of the shaft and such. In any case you want a suitable safety containment device. That will probably cost way more than the flywheel.

That's what I was looking for, thanks! Anyone care to check my calculations? Using that formula at the bottom of the page I get like 184.4 N/mm2, that's only about 26.7 ksi, seems way too low to be right.

Plan to build a solid steel (1/4" plate?) box around it, don't want to be hit by 8 lbs of steel bouncing around the room!

I could gear it to slow it down but direct drive is so much simpler, and accurate. I could go to 5" diameter and 2" thick for the same inertia effect or 4" and 5" long but the weight goes up a bunch.
 
Rotor diameter 11.2 cm

These folks tested their prototype to 60K rpm but there is no mention of if they hardened the 300M steel rotor.

http://power.eecs.berkeley.edu/publications/theses/PerryPhDThesis.pdf

Ultimately, 300M alloy steel was selected for the rotor. The low alloy steel 300M is
very similar to AISI 4340 in composition and can be hardened to an ultimate strength of 2.0GP a
(290ksi) and has a fatigue strength of 1.0GP a (145ksi) for over 1 million cycles [1]. It was
chosen because of its high strength, availability, and relatively low cost.
 
You need professional assistance. Pillow block bearings? Shaft material and diameter. Unidirectional bending fatigue calculations at your projected cycles. 1/4 inch shatter shield. But its your dream so go for it, but with caution.
 
You need professional assistance. Pillow block bearings? Shaft material and diameter. Unidirectional bending fatigue calculations at your projected cycles. 1/4 inch shatter shield. But its your dream so go for it, but with caution.

Ya, I hear you. But professional engineering will put the project way over budget, lol.

A few of these have been built and I haven't heard of any coming apart but they've mostly used lower rpm and/or smaller wheels so it's the wheel I'm worried about. Pillow blocks will be custom made with drip oilers and Swiss bearings that have been proven to handle that rpm. The base will be 3/4" MIC 6, I'll use Thompson "60 Case" 15 mm for the shaft, a clutch and one-way bearing. I'm pretty confident all this is OK, it's just the wheel I'm worried about. It'll look similar to this:

zenoah-on-dyno-jpg.73686


Anyone care to check my results with this formula?

solid-flywheel-10.png


p = 7800

u = 0.3

v = peripheral velocity in m/s. My wheel is 152.4 mm in diameter at 30,000 rpm.
 
I think you have the wheel and mounts pretty well understood. You seem to be ignoring the stuff that drives it. High rpm flex coupling, gears, pulleys etc have not been mentioned. I hope it is not chain drive.
Bil lD.
 
Ya, I hear you. But professional engineering will put the project way over budget, lol.

A few of these have been built and I haven't heard of any coming apart but they've mostly used lower rpm and/or smaller wheels so it's the wheel I'm worried about. Pillow blocks will be custom made with drip oilers and Swiss bearings that have been proven to handle that rpm. The base will be 3/4" MIC 6, I'll use Thompson "60 Case" 15 mm for the shaft, a clutch and one-way bearing. I'm pretty confident all this is OK, it's just the wheel I'm worried about. It'll look similar to this:



Anyone care to check my results with this formula?

solid-flywheel-10.png


p = 7800

u = 0.3

v = peripheral velocity in m/s. My wheel is 152.4 mm in diameter at 30,000 rpm.

Question on your "v" term. How did you get the 0.35? I have:

Rotations * (pi * diameter) meters * minutes
aMinute aasdeterf rotation asdasdff second

(30e03) * (152.4e-03 * 3.1415) * (1/60) = 239.39 which is much higher than your 30*0.35 = 10.5
 
Do you follow two stroke stuffing on Youtube?

I'll have a look thanks but it looks like mostly gas stuff, I'm all nitro.

I think you have the wheel and mounts pretty well understood. You seem to be ignoring the stuff that drives it. High rpm flex coupling, gears, pulleys etc have not been mentioned. I hope it is not chain drive.
Bil lD.

Want to stay away from gears/chains/pulleys and go direct drive for simplicity and to avoid the losses from them.

Thought initially to use some rubber "lovejoy" couplings but don't think they'd be balanced and accurate enough at 30K so going to make my own solid 7075 set screw couplings. Everything will be accurately aligned then dowel pinned to maintain alignment.

Question on your "v" term. How did you get the 0.35? I have:

Rotations * (pi * diameter) meters * minutes
aMinute aasdeterf rotation asdasdff second

(30e03) * (152.4e-03 * 3.1415) * (1/60) = 239.39 which is much higher than your 30*0.35 = 10.5

Thanks for checking. The 0.35 is just the example given, their wheel is 350 mm in diameter and is turning at 0.3 rad/sec. You got the same O/D velocity I did tho at 239.4 m/s. Do you end up with 184.4 N/mm2 or 26,700 psi?

Again, it seems way too low considering Machinery's Handbook said I was way over-revving it with 60 ksi material.
 
I race and build two strokes. The 2 stroke dyno shop I been in on the high rpm stuff uses a timing type belt reduction, so dyno spins at 1/2 rpm of engine rpm.




Ya, I hear you. But professional engineering will put the project way over budget, lol.

A few of these have been built and I haven't heard of any coming apart but they've mostly used lower rpm and/or smaller wheels so it's the wheel I'm worried about. Pillow blocks will be custom made with drip oilers and Swiss bearings that have been proven to handle that rpm. The base will be 3/4" MIC 6, I'll use Thompson "60 Case" 15 mm for the shaft, a clutch and one-way bearing. I'm pretty confident all this is OK, it's just the wheel I'm worried about. It'll look similar to this:

zenoah-on-dyno-jpg.73686


Anyone care to check my results with this formula?

solid-flywheel-10.png


p = 7800

u = 0.3

v = peripheral velocity in m/s. My wheel is 152.4 mm in diameter at 30,000 rpm.
 








 
Back
Top