|08-27-2006 10:40 PM|
Or, replace the measuring laser with a CO2 laser and etch your name a few ten-thousandths of an inch deep into the whole top of your car.
That's just off the top of my head. Take a look at some of the CNC machines at http://www.cnczone.com/gallery/misc....ion=showalbums
Having the laser travel along the same axis as the rotisserie gives you just two lines to make parallel. Even then, any inaccuracies would be measured at the same time the measurement from laser head to car is taken, so they're compensated for in the software.
The reason I said worm gear is because of the fact they don't coast; they stop when you stop the input. The worm could be driven by a step motor and the backlash could be measured with each step. Since it would be the computer doing the measuring, it doesn't matter that it would be a tedious job. This avoids using expensive, high torque step motors and puts it back in reach of a garage engineer.
The computer could do the calibrating for a full rotation, producing a +/- deviation for each step of the motor. That compensates for any slop in the gearing. As long as it can measure the error, the data will accurately reflect the real surface.
Essentially, it's taking the same rotisserie that a lot of people custom make for body work and turning it over to your PC to rotate in tiny increments. At each stop, the laser measures the distance to the car surface at however many points along that axis that you choose and measures how far off from the reference target it is at each of those points. Same priniciple used to align industrial machines or to level dropped ceilings.
|08-27-2006 09:33 PM|
I think it would be better to bolt the car down to a solid frame and then have the laser move back and forth and all around the car. Step motors are used on computer controlled machines. No worm drives because they are not accurate enough.
|08-27-2006 08:45 PM|
What is the whole purpose of this?
What is the end result?
|08-27-2006 05:29 AM|
computer controlled car scanner
Ok, this is off the wall, but that's what happens at this hour of the morning. This is something I've thought about off and on for a couple of years, but don't have the skills or tools to implement. Maybe it's just too much pentrating oil and too little ventilation.
Put a car, without fluids of course, on a motorized, computer controlled rotisserie. (Just assume for the thought experiment that space in the building is no problem, so the car can be rotated 360 degrees about its long axis). On a frame of pipes or tubing, the plane of which is parallel to the long axis of the car, put a kind of trolley system that can travel the length of that frame on the horizontal pipes. On that trolley, a head unit carries a laser measuring device. (It could also carry an arm that moves in and out, perpendicular to the trolley, toward or away from the car, to duplicate a paint job or mill a full-scale model of a previously scanned car).
That head can carry lasers for checking deviation of the whole rig from the X and Y axes. As long as there are fixed points of reference, the frame that the trolley rides on does not have to be very precise. Let the computer read the deviation from those fixed points with each pass it makes.
Similar frames front and back, without the rotisserie, deal with those views.
I'm assuming something like worm gears in the rotisserie and fine motor control to drive them so that it can be moved in small enough increments to be useful.
It just looks to me like all the necessary components are available off-the-shelf for such a thing to be assembled in a lot of garages. Quite a few hotrodders already build rotisseries. It's not that big a step to put a bearing and shaft at one end and a worm gear transmission (for example) and a stepper motor at the other. The hobby robotics folks already do some strange stuff with open source programs controlling free roaming robots; it should be a cinch to run a laser back and forth along a pipe while collecting data (distance from laser head to car surface, distance from laser head to reference point, angle that the car rotated since last measurement). All that data gives you a 3D representation of the surface of the car, or other large object, in the computer.
All sorts of wild things could be done with the data after that. Replace the laser head with a milling head on a rigid arm that can travel the same path the laser did and mill a replica out of a big block of foam or wood or aluminum. Or, modify the shape in the computer and mill that. Use the result to build fiberglass molds.
Where are the engineers and technicians to blow holes in this contraption?