Previous in Forum: Managing Automation: Progressive Manufacturing Awards   Next in Forum: How to Minimize Conveyor Spillage?
Close
Close
Close
2 comments
Participant

Join Date: Jan 2006
Posts: 1

Image sensing problems with laser metal depostion

01/12/2006 6:48 PM

At UMR there is a lab experimenting with LMD, it is the Laser Aided Manufacturing Processes (LAMP) Lab. I am a graduate student there of Mfg Engineering with an undergrad in EE. I am working with a Fastcomm iMVS-155 image sensor, it senses 400-1000nm waveslengths. We use a 1kW diode laser that outputs at 808nm in wavelength. I am to monitor the melt pool created by the laser on the substrate, however, working with H13 tool steel and Ti64 have proven to be difficult, and the fact that the laser is within the sensing range has not proved to be helpful. Using an elliptical model, the camera views the melt pool from directly above it, not at an angle, the images are binarized, and then the melt pool's center of gravity is found and calculations are made on the image for major and minor axix length, ect. This setup works with a YAG laser, as it's wavelength is 1063nm, but we don't use that one anymore.

I am trying to figure out something that would save me from buying a new camera. A 0 degree incidence angle hot mirror was trialed, but did not allow me to see the melt pool as planned. I need as much contrast as possible in the scene, to the camera can threshold the images, convert them to binary mode and do calculations. There seems to be an almost constant gray level throughout the image and when the camers removes fixed pattern noise (FPN) I lose most of the melt pool. I have manual control over the cameras A/D max and min, gain, offset max and min and have tried turning down the brightness and turning up the contrast, like on a monitor, but that proved only to be mildly effective. I have also tried added extra light to the scene, but the laser is pretty bright. An optical notch filter would be ideal, but they only can be made custom for 808nm and that takes more $ than we have to spend.

Maybe if someone could confirm that when steel is heated on a surface is makes a hemisphere shape and all the light of the pool is being reflected elsewhere and not straight back up to the camera??? I wondered if anyone had any thoughts of how to improve the quality of image the camera is getting???

Register to Reply
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Anonymous Poster
#1

Camera Trick Shot

01/13/2006 1:03 AM

The marginal problem here is probably that the camera has no reference to focus. Speaking from a photographic standpoint, you need a blue lens on the camera. Then if using an auto focus feature, focus the camera before the weld begins. This is usually done by applying slight pressure to the shutter button without actually taking a picture. The blue lens should allow for easier focusing (at least between 600 and 1000nm lasers). Cheers, Cowardly Anonymous :-)

Register to Reply
Associate

Join Date: Dec 2005
Location: Wheaton IL, 30 mi W of Chicago
Posts: 49
#2

LAMP

01/13/2006 11:10 AM

Melted metal does form a curved surface, the point radius of a particular point on the surface of the melted metal is determined by the surface tension of the liquid metal, the adhesion of the base to the liquid and the base shape of the liquid footprint. It is also determined, to an extent, by gravitational force. If all the easy ways of changing the curve of the liquid have been tried and not knowing how to practically increase or decrease gravity, I would try using a magnetic field(s)to change the shape of the liquid metal surface. Of course this would only work with magnetic materials and there would have to be some attention paid to magnetically shielding sensitive devices. Good luck with your work!

Register to Reply
Register to Reply 2 comments
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.

Previous in Forum: Managing Automation: Progressive Manufacturing Awards   Next in Forum: How to Minimize Conveyor Spillage?

Advertisement