Engraving on a curved surface

[Richard DiMaggio]

A couple years back I began cutting out scale model buildings for HO trains and the like, which led to marking on motorcycle windshields. (don;t ask... there is a thread there). I've never been able to afford to buy a boxed system, so I ended up building my own, one piece at a time. This latest system i built is made up of independent components. CO2 tube and power supply from China, linear motors, servo motors and controls are from Baldor. Alot of the physical hardware is 80-20 style indrustrial erector set aluminum. Very easy to use and modify. As far as software goes, there are two programs involved, one resides on the motion side, taking path data from the VB program running on the laptop. The laptop does all the path generation, typically created in Corel, then exported as a DXF file. Then my VB program modifies that with a "grab the next closest line" routine or forms the raster pattern depending on if the result needs to be vector or raster.
One problem that I was faced with was the fact that a windshield is curved. The arc is more than the allowable "sweet spot" focus on my 3" focal length lens. The windshield can be removed and flattened in some cases, but some windshields have metal frames and cannot be removed. Trying to come up with a mounting technique that would allow for rotation to keep the beam perpindicular to the marking surace was the challenge. Just thought I would share my solution.

http://www.youtube.com/watch?v=wBcACJDsrK8
http://www.youtube.com/watch?v=wDaAHA5ujPM

The first link shows the parts of the system. The second shows it moving...
Anyone else need help with marking on a curved surface, maybe I can help.
rj

[Mike Null]

WOW!!

I'm impressed. Can we see the result of the engraving? I have an award made of glass in a similar arc that I have to sand blast because of focusing issues.

You have quite an invention.

There are a lot of people looking for a reliable rotary device and you appear to have the talent for that as well.

[Rodne Gold]

Amazing!!!!!!!!!!!!!!! - can you make me one?

[Dan Hintz]

At work, so I had to skip forward a lot and forego audio, but it seemed like there was a lot of side-to-side movement of the piece itself... is this taken into account when moving the beam axis and making your mark?

[Richard DiMaggio]

Mike, here is a picture of the first one I did using this configuration. As you can see (Dan) the aspect ratio of the desired image is accounted for in the (aparent) motion of the windshield. You don't really think that I have the ability to do rotary motion with dual vertical axes without taking aspect ratio into account, do you?

Unless you have a solid block of glass you are trying to do Mike, I don't see why it could not be done. I was going for a system that could rotate larger radius objects, and a standard theta (rotary) axis just was not an option here. I had to have some way to load and unload it easily, and since the radius of the marking surface takes up about 23 inches wide (which equates to about 100 degrees on the surface) I didn't need to rotate 360. For smaller parts I have a rotary axis that works fine for that.


Rodney, how many do you want?

204726

[Dee Gallo]

This is way beyond my understanding, but impressive nevertheless! You have an inventive mind, that's for sure!

[Martin Boekers]

WOW! You have been lurking in the background too long.
Can't wait to see what else you can achieve!

[Rodne Gold]

I'll take a dozen to start :)
How does it work on compound curves , like a dome? Can it work on stuff like that?

[Richard DiMaggio]

My first attempt of lasing on a curved surface was indeed a dome. So, to answer your question, my software solution will indeed adjust a single Z axis to keep a domed part in focus as long as the radius is consistent along the surface of the part. Unfortunately, it does not keep it perpindicular. Depending on the actual radius of the part, this may not be a bad thing. You can have the beam quite a few degrees out of perpindicular without much degredation of the result.
In order to maintain beam perpindicularity, it would require a little different setup than what I have here. I would have to be able to move the "desired" tool path surface in both X and Y. I mean I can probably build the machine that could do this using a "Stuart platform", but that would require 6 axes of control. If you wiki "stuart platform" you will see what I mean. They describe it as being hydraulic, and the first ones were, however this has been duplicated with electric actuators with success by Moog and other companies. Now we're talking major bucks (to me anyway). Hmmm... I could probably do a lightweight version of it using steppers...

[Greg Bednar]

My first thought was you are sitting on several huge lumps of gold! I hope you have contacted some laser manufacturers or done something to bring this to market. - Excellent product.