Uneven power

[Glenn Corser]

Lee, I agree with what you said above. As I was assembling some lamps today I was thinking about it and came to the same conclusion. But, something else occurred to me. I think the beam basically illuminates the mirror with a two dimensional distribution that has a long tail in the direction of the cutting assembly and may look like a gaussian in the direction normal to that. As the corner mirror moves farther from the upper left, the illuminated area becomes bigger due to beam dispersion. So if the spot where the middle of the beam hits were to become less reflective, then as you approached the upper left hand corner and the beam became more focused on the "bad" spot, you would reflect less power. As you moved away from the corner, less of the beam's energy would be reflected by the "bad" spot and you would, paradoxically, get more of the beam reflected to the cutting assembly. That and a buck will buy you a cup of coffee.

[Lee DeRaud]

I think the beam basically illuminates the mirror with a two dimensional distribution that has a long tail in the direction of the cutting assembly and may look like a gaussian in the direction normal to that. As the corner mirror moves farther from the upper left, the illuminated area becomes bigger due to beam dispersion. So if the spot where the middle of the beam hits were to become less reflective, then as you approached the upper left hand corner and the beam became more focused on the "bad" spot, you would reflect less power. As you moved away from the corner, less of the beam's energy would be reflected by the "bad" spot and you would, paradoxically, get more of the beam reflected to the cutting assembly. That and a buck will buy you a cup of coffee.

This is the point, in conversations at work, when people look at each other, shrug, and say, "Well, it sounds good."

It brings up yet another gap in my knowledge about this machine: just how fast does the beam diverge? Again, we're only talking about (in my case) a maximum of 28" of beam length...I wouldn't expect significant (or even measurable) divergence over that distance.

As an aside, the whole "bad spot on the mirror" thing strikes me as one of those positive-feedback deals where, once one develops, decreased reflectance equals increased heating, which makes the bad spot bigger, lather-rinse-repeat, very quickly accelerating until the mirror is effectively destroyed.

[Rob Bosworth]

Lee, as hard as it is to believe, the larger the beam diameter on a laser, the smaller the focus spot size. So you might be seeing the effect of beam divergence. As the beam gets further from the source, the larger the diameter of the raw beam. The large the diameter of the beam, the smaller the spot size will be at focus. The smaller the focus spot size, the more energy density of the spot. The greater the energy density at the focus point, the more effective your laser processing should be. It is very hard to wrap your mind around, but it is true.

[Richard Rumancik]

I agree with everything you just said. Problem is, it's not applicable to the situation at hand.

Reread my first post: what I'm seeing is that the apparent power at the front-right corner is 10-15% higher than at the left-rear, contrary to what any reasonable intuition and/or analysis would predict.

Ok, I'm a bit embarrassed as I read that first post wrong. I try to read the questions carefully but I guess it was a case of "knowing" the answer but assuming I understood the question. Your laser is operating exactly opposite to what I see. So now I am really puzzled . . . and I don't have an answer either.

[Dan Hintz]

The large the diameter of the beam, the smaller the spot size will be at focus. The smaller the focus spot size, the more energy density of the spot.

Within limits... as you approach the edges of the lens, the focal point isn't kept as tight, so you can actually lose power density.




Lee,

Beam divergence is in the single-digit millirad range. Over a large table, the beam may only grow by a couple of millimeters in diameter.

[Lee DeRaud]

Lee, as hard as it is to believe, the larger the beam diameter on a laser, the smaller the focus spot size. So you might be seeing the effect of beam divergence. As the beam gets further from the source, the larger the diameter of the raw beam. The large the diameter of the beam, the smaller the spot size will be at focus. The smaller the focus spot size, the more energy density of the spot. The greater the energy density at the focus point, the more effective your laser processing should be. It is very hard to wrap your mind around, but it is true.

Actually, that's the first suggested explanation for the phenomenon that I can wrap my mind around.
(I still find it hard to believe that there is that much beam divergence over the distances involved, but there ya go.)

[Lee DeRaud]

Beam divergence is in the single-digit millirad range. Over a large table, the beam may only grow by a couple of millimeters in diameter.

(looks for envelope to scribble on the back of...)
Call it 5 millirad, or a tick over 0.25 degrees...sin(0.25) is a little under 0.5% times max path length of 700mm = 0.35mm growth. A bit larger than I expected, but not eyeball-detectable. If I remember correctly, the nominal beam size on my 25W system is around 5mm, so that's about 7.5% diameter growth or about 15% area growth...
Hmmm...assuming spot area is the driver here, we're at least in the right order-of-magnitude for the kinds of power variation I'm seeing.

Given that it goes up as the square of beam length, I'm a bit surprised that the guys with bigger machines don't complain about this more. Do the higher-power tubes have bigger beams and/or lower divergence values? (Or perhaps it's partially offset by beam absorbtion/dispersion effects?)

[Rob Patterson]

What is your bed size?

[Richard Rumancik]

I'd like to know if other members tend to see the same phenomenon as Lee is seeing on their tables.

Rob, I don't disagree with what you are saying, but do you see what Lee is seeing on every laser that you have used? Or just some? I had it in my head that laser "effectiveness" would generally drop off as to get farther from the source.

Rob, your explanation would suggest that cutting should be better the farther you go from the source. But working against this is (a) the spherical aberration of the lens, and (b) that if you have a larger incoming beam, your depth of field reduces. The spherical aberration causes loss which increase as the cube of the beam diameter, the depth of field is reduced by the square of the beam diameter. So on one hand you get benefit (processing efficiency) from the increased incoming beam diameter but a couple other things work can against the improvement. Ideally the effects would all cancel out and you'd have a constant "processing efficiency" over the table but I'm sure this is not likely to happen.

My explanation assumed the losses outweighed the potential improvement. Maybe it is opposite in Lee's case? (My explanation describes what I see on my system, but was 100% wrong for what Lee is seeing.)

[Lee DeRaud]

What is your bed size?

12"x16"...max path length around 28" or 700mm.

[Lee DeRaud]

Rob, your explanation would suggest that cutting should be better the farther you go from the source. But working against this is (a) the spherical aberration of the lens, and (b) that if you have a larger incoming beam, your depth of field reduces. The spherical aberration causes loss which increase as the cube of the beam diameter, the depth of field is reduced by the square of the beam diameter. So on one hand you get benefit (processing efficiency) from the increased incoming beam diameter but a couple other things work can against the improvement. Ideally the effects would all cancel out and you'd have a constant "processing efficiency" over the table but I'm sure this is not likely to happen.

I agree that the various effect tend to cancel each other out, with the caveat that they are all slightly variable machine-to-machine and also variable with respect to environment (dust/humidity/etc). That cancellation, to whatever degree it occurs, is a good thing.

[Dan Hintz]

(looks for envelope to scribble on the back of...)
Call it 5 millirad, or a tick over 0.25 degrees...sin(0.25) is a little under 0.5% times max path length of 700mm = 0.35mm growth. A bit larger than I expected, but not eyeball-detectable. If I remember correctly, the nominal beam size on my 25W system is around 5mm, so that's about 7.5% diameter growth or about 15% area growth...

Your math appears to be a bit bodged... 5 mrad (which, incidentally, is the spec given to the average-powered ULS tubes, +/1 mrad... good guess) = 0.9 degrees, which is roughly 1.6% beam divergence. Over a 700mm path, that's a divergence of 11mm.

Wow, that value surprises even me! Did I do my math right?

EDIT: Math was good, I missed pi. 5 mrad --> 0.0286 degrees, leading to a roughly 0.5% divergence. Over 700mm, that's 3.5mm... more in line with what I initially expected.

[Richard Rumancik]

. . . Do the higher-power tubes have bigger beams and/or lower divergence values? . . .

Well, if you go by the Synrad series, I would say you are right on both counts. Take a look at this pdf for the Synrad product line. The higher power tubes have larger beam diameters and lower divergence. I don't quite understand the 1/e**2 part, but the "beam waist diameter" is roughly the minimum diameter of the beam which will occur near the exit of the laser tube assembly. So for our discussion we can call it the initial beam diameter.

Regarding the math - shouldn't the growth be more like 3 or 3.5mm in your calculation? Dan, there seems to be something wrong with the .9 degree estimate . . .


Synrad says that a full-angle beam divergence of 4 milliradians means that the diameter will increase 4mm/meter. So if you started off with 3.5mm initial (waist) diameter (Synrad 48 series tube) and it increased 700/1000 x 4mm you would end up with about a 6.5mm beam at the end of travel. I know your tube is not a Synrad but the numbers are probably in the same order of magnitude. So the beam diameter could roughly double over a 700mm path. That reduces the spot size by 1/2 and reduces the depth of field by 1/4. Smaller spot size can be good due to increased power density, but smaller depth of field is bad for cutting effectiveness especially for thicker materials. How it all plays out with all the other variables seems to be uncertain.

[Lee DeRaud]

5 mR (which, incidentally, is the spec given to the average-powered ULS tubes, +/1 mR... good guess) = 0.9 degrees, which is roughly 1.6% beam divergence.

180/pi * 0.005 = 0.28 degrees, right? The divergence would be the sine of that: sin(0.28 degrees) = 0.005 (or 0.5%). But yeah, that's more like 3.5mm beam growth, which seems high.

Again, that's using a 5mm initial beam diameter figure, which is what I dimly recall from a discussion with the vendor. (The mirrors and lens are about 15mm, which makes 11mm beam growth, um, problematic.)

[Dan Hintz]

Yep, missed the pi in my initial calculation (must have fat-fingered the calculator keyboard)... I knew that sounded huge. 0.286 degrees.

5 mrad --> 0.0286 degrees, leading to a roughly 0.5% divergence. Over 700mm, that's 3.5mm... more in line with what I initially expected.