Thanks for the responses from Will, Tim and Brian (hope I didn't miss anybody),
My goal, Brian, is both... improve both speed and accuracy. Using outline, like Will said, really improves surface finish. I also think control over beam offset is improved resulting in a more accurate part. Using an increased laser power coupled with the wider scan spacing reduces fill scan time significantly. It also allows the operator to adjust the strength of the inside of the part independently from the outside surface... excepting horizontal surfaces. 3D Systems' software does allow the user control over up-facing and down-facing surfaces. While this may not be possible with the DTM's, tilting the part so there aren't any horizontal surfaces will somewhat accomplish the same thing.
Laser-on and laser-off could probably be adjusted to better optimize the fill since we're not trying to control the surface finish of the part. I'll try the mylar and copy-file-to-scanner if I get a chance Tim.... you learn something every day. P.S. I knew those were the parameters but I couldn't remember the names at the time... really;) Unfortunately, because it is a raster scan, the best values probably vary as a function of part cross section. Brian and Mary, would you suggest any change for improved laser on/off numbers for this situation?
And, I didn't know there was an outline scan speed setting. Ideally, if I could increase the scan speed to the point of keeping the laser at 9-10 watts, that would eliminate some of the delay while the laser changes power, unless the delay is mostly caused by "loading scanner" when switching from fill to outline. There must be a maximum recommended scan speed for outline though. What might that be?
One more software improvement suggestion for the next release. Perhaps the non-linear scaling should be a function of distance from the top of the build rather than the bottom... Z height. I'm guessing but I think that the farther a part is buried, the more it tends to shrink thus requiring a slightly greater scale value. Rather than a function of Z, perhaps use (build-height minus Z).
Disclaimer: Engineer and out the other!
I've already told you more than I know.
Michael W Miller (email@example.com)
The Boeing Company M/C 17-PE B-2T65
Propulsion Experimental Hardware 206-655-3289
Rapid Prototyping 655-4366 Lab 655-4365
> From: Brian_Bauman/DTM@dtm-corp.com[SMTP:Brian_Bauman/DTM@dtm-corp.com]
> Sent: Friday, June 09, 2000 8:11 AM
> To: Miller, Michael W
> Cc: Mary_Michalewicz/DTM@dtm-corp.com; 'RPML www Mail List'
> Subject: Re: Applications Improvements
> Thanks for your questions. I hope to address some of the basic ones and will
> follow up with other people here on the more detailed ones.
> By opening up scan spacing and using outline, what is the goal of the excerise?
> Is it for speed or accuracy or both? Opening up scan spacing will create less
> dense parts unless you increase your laser power which in turn can create growth
> or bonus Z problems. Adjusting laser on laser off values with increased scan
> speed can be an option, but will hinder accuracy of small features. Then
> outline will have to provide the strength for those small features. Maybe two
> outlines will suffice this issue.
> With regards to delays in laser power changes is attributed to the scanner
> controller on the 2000. The 2500's and 2500 plus' do not have this problem.
> Should we have a salesman contact you? :-) The solution is to use the same
> laser power for both outline and fills in order to avoid the delay. You can
> adjust the scan speed for the outline to try to match the required affect using
> the same laser power. I don't believe changing laser power will decrease the
> life of your laser.
> With regards to 4 decimal places for feed distance, you can actually enter in
> this value and the machine will read it accordingly. The problem is that the
> display on your screen is rounding to 3 decimal places. Make sure you have 4
> decimal places selected for your global settings in build setup. This will be
> fixed on the 2.1 release.
> Good luck and keep us informed on your results.
> Best Regards,
> Brian Bauman
For more information about the rp-ml, see http://ltk.hut.fi/rp-ml/
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 23:03:39 EEST