Index Home About Blog
Newsgroups: rec.crafts.metalworking
Subject: Re: A story: was: Re: Cleaning with compressed air
From: John De Armond
Date: Tue, 09 Mar 93 20:11:21 GMT (David Kassover) writes:
>In article <3494@tymix.Tymnet.COM> kopfj@bert.Tymnet.COM writes:
>...So, the put a 1" diameter ball in the hub, and turned on
>>the air pressure.  At 100 lbs pressure it was spinning about 7000 revolutions per
>I dunno, unless I got the geometry wrong (possible), or flubbed
>the calculation (entirely possible), I make the tangential
>velocity at the surface of the ball about 1250 miles per hour, or
>about mach 1.67  (assuming 1100 ft/sec = mach 1 at room temp and
>pressure (which might not be a valid assumption in the close
>neighborhood of the surface of the ball))
Thoroughly flubbed, as a matter of fact.  Here's the calc.
Given:	1" diameter
		7000 rpm
Circumference == pi * D
	3.14159 * 1 = 3.14159 inches
Multiplying by 7000 revolutions per minute gives 21,991 inches/minute.
Divide by 12 to get feet/minute == 1,832.6 feet/minute
Referring to "Engineering Units Coversions", the conversion from
feet per minute to miles per hour is 0.011364.
1,832.6 * 0.011364 == 20.8 miles per hour
Back to the original post, I suspect the bearing was turning MUCH faster
than 7000 rpm.  As much as an order of magnitude faster.
A similar story.
One of the motorcycle race engines I used to race required main bearing
changes after every race.  As a result, I had a large surplus of slightly
worn, approx 1" diameter bearings.  A fairly long road used to dead
end onto another road that went in front of our house.  Our driveway
was on the other side of the road.  I used to spin these bearings
up with 175 psi air and release them to run down the road.
The bearings would EASILY outrun cars driving at ordinary residental
street speeds, spraying sparks all the while.  When it hit a gravel or
other debris, it would jump high into the air.  Which resulted in
some funny instances.  Like when it chose to leap while under a car.
Or when it hops up in front of a car.  Driver reaction was interesting.
I learned several things from this experience:
*	This is a lot of fun.
*	It is fairly hard to turn the bearing while is is spinning due to
	gyroscopic effect.  Turning the bearing can make it seize from the
	high side loading.
*	When the bearing seizes up while you're holding it, the friction heat
	makes a nasty burn on your fingers.
*	If you happen to have your finger securely inserted into the hole
	when the bearing seizes, the torque will likely dislocate your finger.
*	When the outer race explodes from over-speed, the particles that fly off
	are whizzing and can cause injury.  The incident - which ended my
	career as a "bearing whizzer" - did not injure me but it DID dig
	significant sized divots out of the concrete.
If anyone really wants to try this, stand perpendicular to the bearing
so esplosive debris won't impact your body and have it aimed in the direction
you want to release it.  Wear a leather glove and hold the inner race by
only the fingertips.

Newsgroups: rec.crafts.metalworking
Subject: Re: A story
From: John De Armond
Date: Wed, 10 Mar 93 07:45:28 GMT (David Kassover) writes:
>In article <> () writes:
>>What do you mean "playful people" ?  This is science damn it!  You should
>>have been around when we experimented with propulsion through a tube and an
>>alumina ball.  We used the pressure in a Nitrogen bottle.  Every now and
>>then that little sucker is seen, still bouncing around the lab.
>>You university types...   ;-)
>I've got to jump in here. (again).
>Not too long ago, I had an office across the hall from a
>laboratory.  One day I heard:
>Ping!!!  Clang, Clang, rattlle rattle rattle.
><giggle, guffaw>
>Scrape, shuffle...
>PING!!!  CLANG, Clang, rattlle rattle rattle.
>Oooh, that was a good one! <giggle...>
Well since we're posting our Adventures with Science, I thought I'd dig
into my archives for an article I originally wrote several years ago.
This is absolutely true to the best of my (intentionally fading) memory.
From: jgd@rsiatl.UUCP (John G. De Armond)
Newsgroups: rsi.postings
Subject: jgd; rec.models.rockets; Re: A tale of [nuclear] model rocketry
Date: 13 Dec 89 09:36:45 GMT
In article <> (Mike Smithwick) writes:
>Ok guys, you want great model rocket stories? I got great model
>rocket stories. . .
[a GREAT story deleted ]
>Beat that guys! :-)
Hey, with a challenge like that, who could resist :-)  Betcha can't
beat this one.
This story involves TRULY internal combustion propulsion - as in internal
to the atom.
The setting is the Sequoyah Nuclear Plant in Chattanooga, TN.  The date is
about 1979 or 80 and we're starting that sucker up for the first time.
I was the shift test engineer, running the startup test program on the evening
shift.  Most testing is conducted from the control room.  In order
to get to the control room, a stroll across the turbine deck is necessary.
This deck contains the huge (1200 MWe) turbines and a variety of support
equipment.  The floor is tiled and the 10 story tall walls are slightly
tinted glass.  A glorious sight when bathed in the afternoon sun.
As I was strolling toward the control room, I noticed an Aux. Operator
standing near a device called a moisture separator/reheater.  This device is a
large heat exchanger, about 40 feet long and 20 feet in diameter.  Its
purpose is to reheat the steam exhausting from the high pressure turbine
in order to dry it before being introducted into the low pressure turbines.
On top of this device is a large safety relief valve with a tailpipe
that extended almost 10 floors through the roof.  When this valve opens,
steam at about 900 psi exhausts to atmosphere through this ~36" tailpipe.
The tailpipe is hung from spring hangers and simply floats on the exhaust
flange of the safety valve which allows the pipe to move under
thermal expansion.
Anyway, this operator was standing along side the reheater.  In one hand
was a walkie-talkie and in the other hand was a lanyard that ran to the
manual trip lever on one of the safeties.  This was not unusual, as the
functionality of these critical valves is tested fairly often.  Normally
when the valve trips, there is some steam escaping around the valve, a
loud shriek and a large steam cloud on the roof.
As I was almost to the control room, the operator got some activity on
the handi-talkie and pulled the manual trip lanyard.  The noise
from the H-T had gotten my attention and I looked around just in time
for the valve opening.  FOOOMsssss!!!! The whole damn tailpipe jumped
up about 6 inches in the air before settling back down.
Since this behavior was quite abnormal, I asked the operator what was
going on.  He pointed to the elevator and suggested I go to the roof
to find out.  I rode the elevator 5 floors and hoofed it up 5 flights
of stairs and onto the roof.  I noticed about 10 guys standing around
near the tailpipe.
As I stepped out, I saw about 6 guys hoisting A 55 GALLON DRUM up and over
the tailpipe.  Whoosh.  It hit bottom 10 floors below.
A message on the handi-talkie and BOOOMssss!!!!!!!!  That damn 55 gallon
drum full of 600 pounds of water had been launched literally out of sight
by 900 psi of steam.
It stayed out of sight a good 30 seconds before it came into view again,
hurtling down over the Chickamauga lake.  When it hit the lake, it looked
like a depth charge going off.
I did a 180 degree twist and headed back down the stairs as fast as my
little feet would carry me.  As Shultz on Hogan's heros used to say,
"I saw notsing.. I hear notsing.. I know notsing..".  I did keep
a piece of strip chart recording that showed the dip in steam pressure
that documents the launch :-)
I heard a few days later that one of their ICBD (Inter County Ballistic
Drums) had been caught by a gust of wind and had come down on a car in
the parking lot, thus ending the era of the nuclear powered missile.
The funny thing is, no one would ever admit to knowing how that drum
ended up on the car, which ended up about 6 inches tall :-)
So if anyone asks if America has ever launched a nuclear powered
missile, you can answer truthful YES!
[BTW, I've waited 10 years to tell this story to ensure that my
memory of the names of those involved has thoroughly faded just in
case the nuke police were to get interested.]

Newsgroups: rec.crafts.metalworking
Subject: Re: A story: was: Re: Cleaning with compressed air
From: John De Armond
Date: Wed, 10 Mar 93 22:57:53 GMT (Jim Kirkpatrick) writes:
>>>>the air pressure.  At 100 lbs pressure it was spinning about 7000 revolutions per
>Hmm, as your own post quoted, the original poster said 7000 rev per second,
>capitalized, not 7000 rpm as you proceed to calculate with.  Looks more
>like 1200 miles per hour, multiplying your result by 60.
No, I didn't miss it.  I presumed he had made a typo or mistake.  Ain't
no way you're gonna spin a 1" ball bearing to 420,000 rpm using an
air hose.  50,000 rpm (maybe doable) works out to a surface velocity
of 149 mph.  I always figured that the bearings I spun up hit maybe 25,000
rpm which would give a surface speed of about 75 mph.

Newsgroups: rec.crafts.metalworking
Subject: Re: A story
From: John De Armond
Date: Fri, 12 Mar 93 20:57:11 GMT

kopfj@bert.Tymnet.COM (John Kopf, X6305) writes:
>>Yes, but remember that the tangential speed HAS TO BE LESS THAN THE
>>AIR SPEED comming out of the hose.  7000 rps and 1" dia is about
>>1830 ft/s.  Aint no way you are getting supersonic flow out of a
>>100 psi air hose.  (without a pretty trick nozzle!)
>>The original poster has to have meant 7000 rpm, or if they did mean
>>7000 rps, I'd like to buy their air hose!
>I posted the original story, and meant 7000 revolutions per SECOND!  (We
>estimated this from the fact that the spinning ball was putting out a
>pure tone of ~7000cps, and couldn't come up with any reason why this was
>not the RPS of the ball -- rather than some high harmonic of it).
Take whatever RPM  you estimated and divide it by the number of balls in
the race.  If there were 8 balls, a speed of 875 revolutions per second
(52,000 RPM ) is believable, though doubtful.
>Also, the tangential speed does NOT have to be the air speed -- a sailboat can
>travel much faster than the wind when moving at right angles to the wind.
But you're not dealing with an airfoil here.  This is a simple case of
impingement and the air velocity must be sufficiently higher than the
ball bearing speed to deliver sufficient impingement force to balance
the frictional and aerodynamic losses.
That said, a race speed much faster than the air velocity is possible simply
because the balls and cage travel at a fraction of the velocity of the
outer race.

Newsgroups: rec.crafts.metalworking
Subject: This Ball Bearing Thread
From: John De Armond
Date: 13 Mar 93 07:34:45 GMT

I got a little tired of the wild speculation in the thread on spinning up
a ball bearing with an air hose.  There was a claim of being able to
generate speeds over 400,000 RPM.  To settle things, I decided to
actually measure it.  The experimental setup is as follows:
A SKF #6302 (OD=1.65") 7 ball ball bearing, minus its grease shield, is
mounted on a dowel.  A blowgun with an outlet orfice diameter of 0.14"
(much larger than stock) is connected through a pressure regulator and
gauge to 175 psi shop air.  A mist lubricator introduces a very small
mist of 1% light machine oil in 1,1,1 trichloroethane.  This was
experimentally determined not to affect the speed of the bearing while
providing some lubrication.  For each data point, the angle of the
blowgun was adjusted for maximum speed as indicated by the sound.  It
was determined that the best speed always occured when the nozzle was as
close as possible to the bearing cage and angled roughly so the air
impacted perpendicular to the cage where it wraps around the balls.
This demonstrates that the propulsion is simple impingement and does not
involve any ambient air entrainment.
A mark was applied to the outer race of the bearing using Dykem.
A General Radio Strobotach was used to measure the bearing RPM.  For
RPM above 25,000, the strobotach was used on the second harmonic.
Ten points of data were recorded.  The testing was stopped at 110 psi
because the bearing had started to self-destruct.  At the end of this
test, the previously new bearing was worn, loose and pretty worthless.
The data is as follows:
PSI		Speed
40		11,000
45		12,500
50		14,500
55		16,000
60		18,000
70		21,000
80		24,000
90		25,000
100		27,400
110		28,500
This data was plotted and curve fit using MathCad.  The curve shows
a definite tendency toward flatening above 90 psi as would be
expected.  That data was least-squares fitted to a linear equation.
The coefficients are:
a = 257.05
b = 1796.67
For the equation y = ax + b.
Using this to compute the theoretical speed (neglecting the flattening)
gives a speed of 53,000 for 200 psi.  In practice, that speed would
be unobtainable.  The bearing life would be measured in seconds.
My original speed estimates were pretty close.

Newsgroups: rec.crafts.metalworking
Subject: Re: This Ball Bearing Thread
From: John De Armond
Date: Sun, 14 Mar 93 07:47:06 GMT (Jack Mcbryde) writes:
> (Ole-Hjalmar Kristensen) writes:
>>Wasn't the original poster talking about a single ball, not a complete
>>ball bearing?
>That was the impression that I got. I have doubts about whether or not it got
>to 7000 rps, but I suspect that a single sphere is noticibly more durable than
>a bearing assembly.
Praytell how you would propose to:
a)	hold the ball bearing ball while you spin it with the air hose and
b)	How you propose for an exceedingly smooth and balanced object like
	a ball bearing ball to make a musical note - the basis for the
	first speed estimate.
A hint to resolving this dilema: the term "ball bearing" refers to the
assembled bearing and not an individual ball.
Hmm.  Someone suggested to me in Email that someone would want to argue
about my post.  Methinks he was a very astute Usenet observer.

Newsgroups: rec.crafts.metalworking
Subject: Re: This Ball Bearing Thread
From: kopfj@bert.Tymnet.COM (John Kopf, X6305)
Date: 15 Mar 93 18:08:32 GMT

Here's (an approximation) of the geometry:
|---------\                 /------------|
|          \   The Ball    /             |
|           \             /              |
|            \           /               |__
|             \         /                 __  Ain in
|              \       /                 |
|               \     /                  |
|             ==>\   /<== holes          |
|                 \-/                    |
|          Air Plenum (space)            |
The actual device was a "box" with an air connection on one side.  A
flange was bolted to this as a cover; this flange was machined from a
thick piece of plate and had a conical depression on one side, the
other side was machined away to produce a conical "pyramid", concentric
with the conical hole, rising out of a flat surface.  Several small
holes were drilled around the "cone", parallel with flat surface, and
located so thery were tangential to the inner conical hole.  This
flange was then inverted (point down) and bolted to the box.  The air
line was attached, and air blowing out of the small holes formed a
"vortex" within the conical hole.  When the SINGLE BALL (NOT a race!)
was dropped into the hole, it floated on the air and the vortex spun it
up to speed.  Note that the air was introduced UNDER the ball, and had
to force it's way between the ball and inner wall of the cone.
Maybe someone out there with access to a machine shop can BUILD one of
these suckers and report back their observations?

Newsgroups: rec.crafts.metalworking
Subject: Re: This Ball Bearing Thread
From: John De Armond
Date: Fri, 19 Mar 93 06:24:54 GMT (John Whitmore) writes:
>	_Journal of Physics E_ (GB) V.8 #4 p.291-295
>has an interesting article on an ultracentrifuge with magnetic
>	The authors used a variety of rotors, but a hardened steel
>element 12.7 mm in diameter (a half inch ball, near enough) was
>the most durable.  It got to 17.5 krps (that's over 1,000,000 rpm)
>before exploding.
This kinda puts into perspective the possibility of a 1" ball turning
400,000 rpm.  BTW, did they say how they propelled the balls?
Now I know this is getting silly but I couldn't resist building the
conical rig described by the original poster.  I used a plastic funnel
and 4 small copper capillary tubing jets and a 1" ball bearing ball.
The results were unexciting.  The air vortex suspended the ball and made
it spin but not at any interesting speeds.  The ball oscillated and
vibrated to the point that I could not get a good strobotach indication
but it appeared the RPM never exceeded 25K.  The limiting factor seems
to be the inherent imbalance in the ball and probably also turbulence.
The thing DID make a LOT of noise.  I can see where with a steeper angle
on the cone, the ball could vibrate at an audio frequency unrelated to
the actual speed.
In parallel, I wanted to analyze the problem so determine what the
theoretical speed might be.  Problem is, this problem has me way in over
my head in fluid dynamics so I complied with Pournelli's law (*) and
asked an expert.  My fluid dynamicist friend estimates the air exiting
the tip of my nozzles is at about 0.5 mach.  This would put the velocity
at somewhere between oh, 500 and 700 fps.  500 fps works out to about
114,000 rpm one a 1" ball if I did the math right.  Fits well with
I've asked him to model this problem on his supercomputer.  If he has
time, I'll report back the results.
(*) Pournelli's law: If you don't know, ask someone who does.

Index Home About Blog