h a l f b a k e r yYeah, I wish it made more sense too.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
High-volume car body panels are made using massive
hydraulic presses, for which the tooling costs tens of
thousands.
For smaller volumes (restoration, or exotic vehicles),
metal panels can be made by traditional panel-beating,
but this is fearsomely expensive.
I would like to propose an
alternative - a method of
producing complex bespoke body panels under computer
control.
A sheet of steel is held firmly between two frames, leaving
it unsupported except at its edges. The business-end of
the machine consists of a gun, from which individual ball
bearings (say, a few millimetres in diameter) can be fired.
The speed of the ball bearings can be controlled precisely,
and of course the gun can be moved in the XY plane with a
precision of a few microns. Lasers on either side of the
gun can be used to measure the displacement of the target
(the steel panel) in the Z axis.
Under computer control, the machine plays back and forth
over the metal sheet, firing ball bearings with sufficient
force to gradually, millimetre by micron, coax the metal
into the right form. The used ball bearings are, of course,
collected and reused.
For tight or complex curves, it might well be necessary to
have guns on both sides of the sheet, independently
controlled.
The process would be very slow and very noisy, but useful
for producing precisely-shaped one-off panels, with no
need to retool between jobs.
Father Ted
https://www.youtube...watch?v=vcNiYQgsBUU
[spidermother, Jun 15 2014]
deformatation without penetration
http://www.bluerhin...ts_test_targets.htm
[not_morrison_rm, Jun 15 2014]
In particular....
http://www.bluerhin...tml_files/10169.jpg
[not_morrison_rm, Jun 15 2014]
[link]
|
| |
Pardon? Can't hear you... |
|
| |
"Machine Operated by a young red haired girl in a halter top with a CS degree named Dimples" Rated PG |
|
| |
If the pellets are small enough so that their dents can overlap
smoothly, then to make them go fast enough to actually dent the sheet
of metal, wouldn't you risk puncturing it? |
|
| |
I don't think so. Each pellet is only moving the
metal by maybe a few tens of microns at most. |
|
| |
If you wanted to, you could make the pellets the
same size as a panel-beating mallet head, but I think
smaller pellets (and many, many more small
impacts) will do a better job. |
|
| |
May I, as an American who has one, suggest to an
Englishman, an English wheel? |
|
| |
The English wheel is a great tool in the hands of a
skilled craftsman. However, it lacks the ability to
automatically produce a given shape from a CAD file
in the hands of an unskilled person. |
|
| |
This converges on a sandblaster. I think Vernon is
right that steel would be too abrasive. Maybe balls
of ice? |
|
| |
But a sheepish +; mostly for the redhead. |
|
| |
Well, it's all going to depend on the diameter and
speed of the pellets. At some point, it becomes the
same as panel-beating, which deforms the metal
without abrading it. |
|
| |
(+) Pretty sure this would work. |
|
| |
I'm guessing that the pellets will at least need to be many times larger than the thickness of the panel. But then it's just a question of optimising; even if the pellets need to be a centimetre or more in diameter for best results, the concept still flies. |
|
| |
A bit like shot peening, then, which does not abrade
but compresses the surface to reduce stress cracking. |
|
| |
//the pellets will at least need to be many times
larger than the thickness of the panel// |
|
| |
I'm not sure if that's the case. Imagine tapping a
bluntish nail on a 2mm steel sheet - it will make a
small indentation. Hence, a ball with a diameter
similar to that of the blunt nail ought to do the
same, if fired at a high enough speed, no? |
|
| |
It's about mass, as well as contact area. A projectile of small mass cannot transfer much of its energy into moving a large mass; to make a dent (rather than punch a hole) in 2mm steel, you need to move considerably more than 4/3*pi cubic millimetres of steel, so you need a projectile considerably more massive than a 2mm diameter steel pellet. |
|
| |
Sand blasting and shot peening have been mentioned; in each case, projectiles which are of relatively low mass mostly have a surface effect. |
|
| |
Presumably this could instead be done with a static-mounted hammer and anvil, and a robotically-controlled frame which holds the sheet metal. The hammer bangs away in one fixed place, while the robotism twirls and shiggles the sheet under the hammer. |
|
| |
A contraption like a daisy-wheel printer rotates different shapes of anvil under the work point. Perhaps a number of different shaped hammers could be used, typewriter-style as well. That gives the robot lots of options for shaping convex, concave and other types of surface. |
|
| |
//so you need a projectile considerably more
massive than a 2mm diameter steel pellet.// |
|
| |
Hmmm. I'm not sure. WIth a high enough
velocity, the pellet will punch a hole through the
sheet. With a low enough velocity, it will just
rebound without permanently deforming the
sheet. |
|
| |
Are you saying that there is no velocity between
those extremes, whereby the pellet will dent the
steel, even by a few microns? |
|
| |
//Are you saying that there is no velocity between those extremes, whereby the pellet will dent the steel, even by a few microns? |
|
| |
Yes. But only to be annoying. |
|
| |
Falling back on empirical evidence (the last resort of the scoundrel, I know) it seems our transatlantic cousins, in the drive to use something other than lead in shooting, have done quite a few ad hoc tests of steel shot/plate steel interactions, see linky, Not to mention road signs et al. |
|
| |
Something like a rapidly-repeating captive-bolt gun might work better - or even the 'hammer' action from a hammer drill |
|
| |
It might well. On the other hand, it lacks the
machine-gun-esque charm of the original. |
|
| |
Very true. Another approach might be to use a
high-pressure jet of water. High-pressure water jets can be used to cut all sorts of materials, so they should be able to do a bit of gentle panel-beating. |
|
| |
A constant jet might have difficulty to control the
plastic yield. But a pulsing jet might be better. |
|
| |
Edit: like a high powered ink jet... |
|
| |
OK, I confess I like the water-jet method, even
though it lacks explosions. |
|
| |
So, a water jet that can be moved in the X, Y and
(probably) Z directions, with control over force
and jet diameter. We might as well add pitch and
yaw (useful if you're making a steep curve, as the
jet can be tilted to remain at right-angles to the
metal as the curve forms). And jets on both sides
of the sheet will allow reflex curves etc. |
|
| |
The whole thing is looking frighteningly plausible.
I don't see it working for mass production, but
more like a 3D printer for low-volume work. The
advantage over a 3D printer, however, is the
ability to work in cheap sheet steel or aluminium.
If the thing were scaled up, it would also make
aircraft skin panels. |
|
| |
Not only would a water jet be useful for shaping the material but it could also cleverly control the thickness. If you are making a car body panel for example, you only need thickness (and strength) in certain areas. By having the water jet mixed (under conputer control) with a fine sand as it shapes certain parts of the panel it would be able to abrade away the surface to a pre-defined pattern.
Explosions will be just outside the manufacturing plant, in the form of massive fireworks celebrating the succesful IPO of
Water-Jet Shaping Technologies Inc. |
|
| |
Another idea for a 3D dimpler would be to use multiple
focussed laser beams and immerse the sheet in
water. This would provide the gratification of an
explosion of steam at every pulse. |
|
| |
Also, I wonder if the intertia of the water would
work rather like the shot-bag that panel-beaters
use. As I understand it, when shaping a panel with
a mallet, the beater will use either a solid dolly or
a shot-bag behind the work piece, allowing local
deformation whilst supporting the metal overall. |
|
| |
The water might work rather like that, if the laser-
generated steam pockets were formed
instantaneously. |
|
| |
Another approach would be to use actual explosions. Something like a fuel injector from a car engine would be scanned across the metal surface at a distance of a few milimetres, spraying jets of vapourised fuel into a minature spark generator hundreds of times a second. The shockwave of the explosions, perhaps focussed by a suitable shielding, would be used to deform the metal. This approach has the advantage that fuel-injector technology is cheap and reliably produces explosions. |
|
| |
(hippo) Sorry but gasoline burns in an engine. |
|
| |
What about a shaped charge? It shouldn't take unfeasibly too much computing power to calculate from the desired final shape, how the explosive should be arranged. The robot sculpts the explosive charge, you lay the flat sheet of metal on top, and then press the big red button. There is a deafening flash and your perfectly shaped panel tumbles back down beside you. |
|
| |
I suspect that the use of high explosives might put
this out of the reach of the average hobbyist. |
|
| |
...and suicide bombers would have rather a large
order backlog after the first production trial. |
|
| |