Speeding up engraving

[Brett Winston]

Good morning everyone! I've got a few questions that I can't seem to find answers for. I'll post each under a separate topic.

I have the following mock plaque which is made up entirely of vectors. Sending this to the laser from corel x7 via rdworks v8 and using "scan" (engrave) for both layers (or just one layer - I understand using two takes significantly more time when both are being scanned) produces the results I'm looking for but drawing the border is slowing things down. Even if everything is on the same layer. It could go faster if I use "cut" for the border layer but that apparently limits me to thin "cut" lines and not the wider, filled border that I want. I'm aware I could defocus the lens by dropping the table a bit or perhaps redesign and use several thin lines for the border, but I was hoping for a solution that would ultimately cut down on the time to produce each plaque. Ideas?

CorelMockPlaque.jpg

[Ian Stewart-Koster]

apply an internal contour to the border-in 20 steps, 0.15 mm each - and treat that as 20 concentric vectors... or similar scenario.

[Gary Hair]

Rotate the design 90 degrees and it will run in about 1/2 the time.

[Brett Winston]

Rotate the design 90 degrees and it will run in about 1/2 the time.

Brilliant! (smacking myself on forehead)

[Ross Moshinsky]

I engrave plaques/awards essentially all day long. The issue is your border. Get rid of it and change it to a vector border. There are all sorts of different ones out there that look pretty decent. If that doesn't work, go with a double plate. If that doesn't work, get a designer/silk screened plate. On my machines, I'd expect the run time to be about 5-7 minutes for something like that will an appropriate border. With that solid border, I'd guess it will go up to 12-15 minutes.

As for Gary's advise, my experience is the opposite. Taking a portrait plate and running it landscape results in run times increasing dramatically. I'll be honest, I can't say what the difference is on a big border like that, because I won't run borders like that unless I absolutely have to.

[Brett Winston]

Ok, I tried rotating 90 degrees and there was no significant time savings. Went from ~17 mins to ~16 mins. I'm going to try several vector lines as a border next.

[Gary Hair]

As for Gary's advise, my experience is the opposite. Taking a portrait plate and running it landscape results in run times increasing dramatically.

If you are rastering then that's almost impossible. The distance traveled on the Y axis determines how much time the job will take to run and it's even more noticeable on a faster machine. The reason for this is that the LPI or scan gap determine how many times the laser must move back and forth to cover the y axis. If, for example, you have it set for 500 LPI then it must move back and forth 500 times per inch of Y, there is no way around that. If you have a graphic that is 2" x 4" it will take twice as many passes if you have it with 4" in the Y vs 2" in the Y.

[Bert Kemp]

So if the OP lowers his scan gap to say 250 lpi it should be twice as fast, hopefully its still acceptable

If you are rastering then that's almost impossible. The distance traveled on the Y axis determines how much time the job will take to run and it's even more noticeable on a faster machine. The reason for this is that the LPI or scan gap determine how many times the laser must move back and forth to cover the y axis. If, for example, you have it set for 500 LPI then it must move back and forth 500 times per inch of Y, there is no way around that. If you have a graphic that is 2" x 4" it will take twice as many passes if you have it with 4" in the Y vs 2" in the Y.

[Rich Harman]

If you are rastering then that's almost impossible. The distance traveled on the Y axis determines how much time the job will take to run and it's even more noticeable on a faster machine. ...

Well that's the thing, the speed of the machine is going to play a huge part in whether or not rotating saves time. The OP is using a Chinese machine and does not say what speed he is engraving at. The amount of time spent traveling in the Y direction is most likely insignificant compared to the time spent rastering. On a Trotec, where the rastering is so fast that it is just a blur, well, the time spent moving in the Y direction becomes significant so rotating can save some time.

The only way that I can think of to reduce the time on this plaque, and retain the rastered border, is to split the border into four segments. Rotate the piece 90 degrees (landscape) and raster the left and right parts (horizontal parts when rotated) of the border, then rotate back to portrait orientation and raster the rest. This would require a jig in order to make the border segments line up properly.

[Gary Hair]

So if the OP lowers his scan gap to say 250 lpi it should be twice as fast, hopefully its still acceptable

For the most part it's simple math at that point - 500 vs 100 would be a factor of 5, 500 vs 250 would be a factor of 2.

[Gary Hair]

Well that's the thing, the speed of the machine is going to play a huge part in whether or not rotating saves time. The OP is using a Chinese machine and does not say what speed he is engraving at. The amount of time spent traveling in the Y direction is most likely insignificant compared to the time spent rastering. On a Trotec, where the rastering is so fast that it is just a blur, well, the time spent moving in the Y direction becomes significant so rotating can save some time.

The only way that I can think of to reduce the time on this plaque, and retain the rastered border, is to split the border into four segments. Rotate the piece 90 degrees (landscape) and raster the left and right parts (horizontal parts when rotated) of the border, then rotate back to portrait orientation and raster the rest. This would require a jig in order to make the border segments line up properly.

It would have to be a very slow machine to make it longer to run landscape than portrait, snail slow actually. If I knew the speed he was running and the LPI he was using I could run it in Job Control on my Trotec using the same settings and give you a pretty accurate number. For now, however, I'm sticking to my assertion that it would be almost impossible to have it run longer in landscape than portrait.

[Rich Harman]

... I'm sticking to my assertion that it would be almost impossible to have it run longer in landscape than portrait.

I agree, I think I misread your argument. Whether portrait or landscape the time spent rastering will be equal. It is the acceleration/deceleration (over-run) and the movement between lines that will differ. In landscape you have fewer lines so naturally fewer over-runs so regardless of rastering speed, landscape will always take less time. My point was that the difference in time won't be significant unless you are rastering at high speed - where the time spent in over-run and movement between lines makes up enough time of the whole process to be worth considering.

[Ross Moshinsky]

Sorry, but portrait plaques run faster when run in portrait. Especially when it's straight text or has a "normal" sized logo. The reason is simple. The space between lines. The laser can skip over that space. All that negative space adds up to crazy time savers vs when run in landscape and the laser is covering more or less the full and width of the plate.

I've done the time studies MANY times. Working with a 12x18 machine (that is actually quite fast (~110ips) I had to figure out the best way to run jobs. I have/had one job I needed to get done as quickly as possible. I do it every year and it's during a busy time for us. I figured filling up the table and letting the machine run would be most efficient. Not even close. Production times went up a huge amount. Why? Because the plaque has a decent amount of negative space between lines. The laser was covering a lot more area when the job was rotated and run landscape vs keeping it portrait. I can't recall the exact numbers but it was something like 8 minutes per plate vs 11 minutes. It was a significant difference.

I just did a time study using ULS's estimator for a 7x10 plate. It has a large "square" logo on it. The run time in portrait is 14:53. Rotating it and putting it in landscape it is 15:25. Ran another test, same text, tweaked to fit a bit better with a landscape logo that is what I'd consider more standard. Portrait = 12:00. Landscape = 13:53.

If you're talking about the most efficient way to engrave a 4x4 box, that's one thing. Talking about engraving plaque plates, well that's different. I have the data and the years of testing behind me to know what works and doesn't.

[Tony Lenkic]

Ross,

If you are engraving only lines of text than yes you are correct but when rastering a border surrounding text than Gary's option speeds up timing.
If you are to draw a border 6" wide and 12" high test how long it takes to engrave it in portrait and landscape.

[Gary Hair]

Exactly my point! I didn't specify a graphic vs text because the OP was talking about a graphic surrounding the text and I figured anyone replying would have read that.

Ross,

If you are engraving only lines of text than yes you are correct but when rastering a border surrounding text than Gary's option speeds up timing.
If you are to draw a border 6" wide and 12" high test how long it takes to engrave it in portrait and landscape.