02 – Fiber Laser MOPA Matrix Test Leads to Understanding (20:59)

0:00welcome to another five ELISA learning

0:03laughs today we’re going to carry on a

0:05little bit more with the mobile Acer

0:07designed because I had a problem last

0:12time if you remember and I couldn’t find

0:14out very much information about the

0:17actual pumped laces that were used with

0:20this system and to be honest I wasn’t

0:22particularly worried because I didn’t

0:24think it was relevant

0:25now as I mentioned to you before I was

0:28presented with this list of numbers here

0:30and these charts for all the 16 variable

0:34pulses that this machine can work with

0:37now this is a 20 watt machine and some

0:41of these pulses can deliver up to

0:44something like about 12 kilowatts or

0:46down here around about 7 or 8 kilowatts

0:49peak power sounds amazing

0:53I’ve got no concept of what damage these

0:57shapes these pulses can make to material

1:00because look we’ve got a pulse here

1:02which is only 2 nanoseconds wide that’s

1:05to billions of a second what damage can

1:0912 kilowatts for 2 billionths of a

1:11second to do this machine does not

1:14produce 12 kilowatts continuously it can

1:17only deal with it in dot dot dots spots

1:20and those spots are only going to be as

1:24good as the lens can produce and the

1:27spot size on here is 0.065 of a

1:30millimeter 65 microns I thought I would

1:33do a very quick test to get some sort of

1:38concept of what these numbers actually

1:40mean so I jumped onto this locust mark

1:42software otherwise known as easy cared

1:45and produce this matrix of the 16 pulses

1:50across here and down here I’ve chosen 14

1:52different frequencies from a thousand

1:55two kilohertz down to 5 kilohertz so

1:59first of all I ran it a couple of times

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

2:01on a piece of anodized black anodized

2:04aluminium and I think from where you’re

2:07looking

2:07you’ll quite clearly see that there is a

2:10lot more damage down here

2:13it’s difficult to see on aluminium the

2:16problem with aluminium is is a what I

2:19call a binary material it’s either black

2:22or white there’s nothing in between I’ve

2:27generated these shades of grey on here

2:30by virtue of some patterning that I’ve

2:33done and I will talk about that in a

2:35future session but instead of aluminium

2:37I chose to use a piece of stainless

2:40steel

2:40it might look from your point of view at

2:44the moment as though it’s it’s pretty

2:47brand we’ve got some stuff down here and

2:48then nothing else here’s an enlarged

2:50picture of the stainless steel obviously

2:52the background has now removed a

2:53reflectivity of it so it’s much easier

2:55to see now the thing about stainless

2:57steel is it is not a binary material as

3:01we heat it up it produces various oxide

3:04colors on the surface so there are

3:06shades that we can pick up now what you

3:09can’t see with some of these colors here

3:11is the variation in depth that you can

3:14probably see much clearer under a

3:16microscope but trust me basically what

3:18we’ve got is very little power here

3:20decreasing to almost nothing over here

3:23and no power at all down in this corner

3:27hang about I’ve got nothing down here

3:30what’s happened down here now this was

3:33my very simplified model of this machine

3:35the Mopar and in essence what we have

3:38we’ve got two fibers we’ve got one thin

3:42one and one thick one this one here is

3:44I’d call a signal conditioning fiber

3:46that produces the signal the pulses that

3:49we’re looking for and then they drop

3:51through to another fiber which amplifies

3:53the shape of that pulse to get an output

3:57now we went through clearly last time

4:00the way in which the laser actists this

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

4:02type of laser the fiber laser works so

4:04I’ve drawn these pump lasers in blue

4:07these are laser diodes which are

4:09injecting photons into the cladding here

4:13and the photons are whizzing around in

4:16here and they’re colliding with these

4:18special ytterbium electrons which are

4:21sitting in this very thin core and so

4:24here we’ve got the seed

4:26leisa which is a laser diode which

4:29injects coherent beam and runs through

4:32here collecting as it goes more and more

4:34photons so the beam then passes through

4:37to cross to this amplification stage

4:39where these laser diodes have already

4:41injected photons into here and excited

4:44this core up to its maximum energy

4:47potential now that was the way that

4:50which I was expecting it to work I did

4:53have a problem as you can remember

4:54saying I couldn’t quite work out how

4:57these laser diodes were working if you

5:00assume that these laser diodes are on

5:02all the time which is what I did so

5:06let’s take this example here of eight

5:07nanoseconds and 252 kilohertz so every

5:13cycle takes four microseconds and four

5:16microseconds is 4000 nanoseconds so we

5:23want an 8 nanosecond pulse but I’ve got

5:264000 nanoseconds ahead of that pulse in

5:30which to charge up these in a cause I

5:34was expected to see pulses all the way

5:36down to 0 almost where the 8 nanoseconds

5:40is that line and the 250 I haven’t got

5:46250 because between 200 and 300 so it’s

5:49some way across there all right so it’s

5:52between these two so basically it’s

5:55there is my maximum power and if we take

6:00a look at the colors we can clearly see

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

6:03that we’ve got maximum power there

6:05because it’s fainter there and it drops

6:07off there but hang about it drops off

6:09there and then there’s world is just a

6:11hint of something there and then there’s

6:13nothing why is there nothing because

6:15this this model that I’d produced

6:18predicted that I would have this density

6:22all the way down to the bottom that was

6:24my problem all these are missing down

6:27here for a very good reason and the

6:30missing part of the puzzle was all to do

6:32with these blue pomp diodes how they

6:35work

6:36and I had to go back to Lotus later

6:39where their expertise was able to

6:43furnish me with the details

6:45the secret is these do not run at

6:51constant power they run at variable

6:55powers wind and just before the seed

6:59laser is injected into the system

7:02obviously the less power that we feed to

7:04these diodes and the less photons and

7:07the slower the excitation rate of the

7:09electrons in that core because it turns

7:12out that to put too much power in here

7:14for too long is not good for the

7:18lifetime of the core material okay this

7:21diagram will help rather than hinder

7:22your understanding of the process here

7:25we’re showing the frequency scale from

7:26zero to a thousand kilohertz the period

7:28between the pulses is getting longer and

7:30longer and longer as the frequency gets

7:33lower and lower and that’s quite

7:35important because here we can see the

7:37blue pomp diode intensity with this line

7:40and along the top here and the intensity

7:43is increasing with frequency and then it

7:47stays level

7:48my assumption with my model was that it

7:51was level from the start and that was

7:53where my failure occurred it takes but a

7:57few nanoseconds to promote enough

7:59electrons in here up to their excited

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

8:01level so we certainly don’t need 4000

8:06nanoseconds here to build up energy for

8:11an 8 nanosecond pulse hence the reason

8:13why they start off at 0 and gradually

8:16build you up towards the point where we

8:19are going to discharge a 8 nanosecond

8:21pulse and at that point there we’ve got

8:2320 watts of energy stored in the fiber

8:26it’s been arranged such that we get our

8:29peak power at 8 nanoseconds and after 8

8:32nanoseconds the power starts to drop off

8:35now why does it drop off well the power

8:37doesn’t drop off at all there’s still 20

8:39watts all the way down this line is just

8:43that the pulse repetition rate is

8:45increasing and what that basically means

8:48is let’s assume we had a cake

8:49one cake equals 20 watts and we’re going

8:53to cut the cake into 250 slices when we

8:57get to this end we’re going to cut the

8:59cake into a thousand slices so the

9:02slices at this end are much smaller than

9:04they are at this end we’ve still got a

9:06cake but what basically that means we’ve

9:10got less power

9:11per pulse as we get to the higher

9:14frequencies when we look at our eight

9:17known a second pulse here at 250

9:19kilohertz we can see that we’ve got peak

9:23power but they’re both black just either

9:25side of the ideal which is in the middle

9:29there so there we are at a hundred

9:31kilohertz and at 100 kilohertz we come

9:35up here and we find that we’ve only got

9:37a very small amount of power power

9:40roughly let’s just guess the same as we

9:43would have if we were at 700 and 700 is

9:48about here and yeah there’s not a lot of

9:53difference between those two the point

9:55being that we’ve got the decreasing

9:58power down here and the further we come

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

10:02down of the frequency range the less

10:04power we shall have to recharge the

10:07fiber and less power means we shall have

10:09no pulses now this graph showing its

10:12decrease in recharging ability as we get

10:15to the lower frequencies clearly

10:17explains why we haven’t got anything

10:21down here and conversely when we start

10:25increasing the frequency up here the

10:27power starts dropping off so as we go up

10:30the frequency chart the power is clearly

10:33dropping off so all of a sudden

10:36everything is explained it’s all nice

10:39and neat and tidy and we understand how

10:43the motor laser works so now we found

10:47another parameter which we can play with

10:49ie

10:50frequency which will allow us to control

10:52the power in the pulse beyond the peak

10:56power values that are set

10:58I think we could also play with this but

11:01this is a much more uncontrollable

11:03section down below that peak power I

11:07think we stand much more chance of using

11:10this part of the curve it does add to

11:13the pool of variables yet something else

11:15that we can use to change the energy

11:18density at the surface of a piece of

11:21material this machine only produces dots

11:24and we only damage the surface by

11:27converting the light energy into heat

11:30energy within the atomic structure of

11:32the material we’ve got all these

11:33variables to play with to caused that

11:36heating effect so within the program the

11:39software program we have the opportunity

11:41to change the percentage power well I

11:44don’t know quite what that’s doing I’ve

11:46got an idea that and that might actually

11:48be changing the power of the pump basis

11:50so that we can regulate the if you like

11:54the amount of energy it might be doing

11:56exactly the same as this when we

11:58increase the frequency if it is then

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

12:01we’ve got two ways of changing the power

12:04per pulse or the energy per pulse one of

12:07them is with a percent power button and

12:09the other one is with thee if you like

12:11the pulse repetition rate which is in

12:13fact the frequency and although we’re

12:15changing this frequency we’ve also

12:17changing the frequency of a specific

12:20pulse that we’ve decided to use now

12:24after percent power we’ve got 16 pulse

12:27durations there the question I don’t

12:28know is whether or not if I choose a 15

12:32nanosecond pull will it either drop back

12:35to a 13 nanosecond pulse or will it go

12:37out to a 20 nanosecond pulse or will

12:40actually respond to something in between

12:42maybe I’ll find out directly from Lotus

12:45later so for each one of these specified

12:48pulses there is a maximum power or a

12:50maximum amount of energy but in addition

12:53to that we’ve also got other

12:55opportunities as we’re finding out by

12:57playing with the power to change the

12:58energy per pulse so we’ve got the

13:01opportunity of playing with the pulse

13:03energy now the other thing that can

13:05dramatically change things is speed now

13:09unlike the continuous

13:12beam laser we are pulsing on this

13:14machine and that gives us a completely

13:17different set of circumstances I’ve got

13:20a line which is not 0.1 millimeters long

13:23and I’m going to run a thousand

13:26millimeters a second

13:28along that line so that’s going to take

13:32roughly a hundred microseconds to scan

13:36that at a thousand millimeters a second

13:39if I’m using the two nanosecond pulse at

13:43eight hundred and fifty kilohertz which

13:47is the top one on this chart then

13:51basically here’s what I’m going to get

13:54I’m going to get a huge number of pulses

13:57along that point one of a millimetre

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

14:00line I’m going to get a huge energy

14:03density in the area because I’m going

14:06over the same spot again and again and

14:09again and again now this is where

14:12something else comes into play there

14:16haven’t got my decent piece of paper

14:17here at the moment but if we take a

14:19quick look at the relationship between

14:21the pulse length and the repetition rate

14:24which is the frequency if I just round

14:28that up to a thousand for the sake of

14:31argument I’ve got one microsecond per

14:34cycle and I’ve got two nanoseconds of

14:39power within that cycle so that’s two

14:43nanoseconds per thousand nanoseconds

14:47which is a ratio of 500 to one or

14:51basically 498 of the power being off and

14:55one of it being on so there’s a heating

14:58to cooling ratio of almost 500 to one so

15:02there’s a 500 delay time to allow the

15:06heat to cool down and if we take a look

15:09down this set of numbers here we shall

15:11find that it drops down to as little as

15:14114 to 1 at the bottom here for these

15:17longer pulses so there’s another factor

15:20there which is could be a very important

15:22factor the fact that we haven’t got as

15:24much cooling to

15:25down here in other words we likely to

15:27build up the heat in the surface quicker

15:29the whole way in which his machine works

15:31is by creating heat in the surface so

15:34this I believe is a very important

15:36factor when it comes to assessing which

15:38of these we should choose to do certain

15:40types of job now I think probably people

15:43have been round and they’ve trolled all

15:44these numbers and they’ve put them in

15:46thousands of hours of testing and

15:48they’ve found things that work I haven’t

15:51got thousands of hours left in my

15:52lifetime so I’m not certainly going to

15:54do that I want to try and choose some

15:56factors which will give me the

15:58information quickly so speed I mean if I

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

16:01run the same thing thousand millimetres

16:03a second but this time we’ll go to 350

16:07nanoseconds which is this one 350

16:11nanoseconds okay the heating and cooling

16:13ratio is different we’re not going to

16:16cool as quickly as the previous one and

16:18we’re going to probably have put more

16:19energy into it as well so the net effect

16:22should be greater but if we carry on

16:25running that at a thousand millimeters

16:27per second and we also run it at 25

16:30kilohertz because the pulse rate is so

16:33much slower what’s going to happen is

16:35we’re going to space these out and we’re

16:37not going to get as much of a heating

16:38effect so we’re certainly probably not

16:40going to get as much power into the

16:42surface this way as if we were to do it

16:45this way I don’t think but I don’t know

16:48that’s what we’re gonna have to try and

16:49find out so that’s how speed is going to

16:53affect us so as well as speed we’ve got

16:57scan patterns now there are several

17:00ideas for the scan pattern number one is

17:04we could just scan scan scan scan and

17:09put the lines beside each other and

17:12accept that is the heating effect for

17:14there and now we’ve got another heating

17:16effect here and another heating effect

17:18here or we can overlap them so for

17:21instance if this is a point O three spot

17:23on here if I shift it down by 0.01 five

17:27pitch then I will double up on that half

17:31of the heating and put a new set here

17:34and then when I put another line down

17:35that will get doubled up so they

17:38a different energy density if I choose

17:41to use a line spacing of 0.03 or 0.015

17:46there are other things in scaled pattern

17:49which the software allows us to do as

17:52well as this overlapping pattern we can

17:57scan in one direction only the other way

17:59is to scan and come back scan and you go

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

18:04back scan like that so we’ve got

18:07uni-directional

18:08or bi-directional scanning there’s

18:10another option as well and that is to

18:12send the dot in a circular motion in

18:16other words we start off with our dot

18:18there but we can do this with it and we

18:22can drive the dot round in a circular

18:24pattern to produce this overlapping

18:29high-density picture where we want high

18:32energy density into the surface to get

18:35the temperature up to do certain amount

18:37of damage then we’ve got the beam

18:44profile now this is an interesting one

18:46because they make a big thing about this

18:49machine having a Gaussian energy profile

18:56to the beam which means we’ve got a lot

18:58more energy in the center of the beam

18:59then we have at the outside edge of the

19:02beam which means that we should be

19:04cutting grooves in the material

19:08effectively which look like that so

19:10there’s the surface of material if we

19:12decide to scan across we’re likely to be

19:15putting grooves in which look like this

19:17now there are other beam patterns that

19:21could be chosen that would produce a

19:23much flatter more uniform shape rather

19:28than a Gaussian you could have something

19:30like that which which would produce a

19:32much shallower more even heating effect

19:35I don’t know whether that’s what we need

19:37or not but it is another variable

19:41so that’s beam profile and then of

19:44course finally we’ve got things like

19:46focus

19:48that’s a very emotive subject in some

19:52people’s eyes focus is focus and you

19:55must stay with the focus but I’m here to

19:59break the rules I’m very happy to run

Transcript for Fiber Laser MOPA Matrix Test Leads to Understanding (Cont…)

20:01this machine at whatever is required

20:03because as you change the focus you’re

20:05actually going to change the energy

20:07density in the in the little pool as we

20:10as we Inc as we change the focus and I

20:13going to change the size of the beam

20:17that’s a subject for a future session

20:19and then of course finally you’ve got

20:22material every material will have a

20:28different set of parameters is a year

20:31going to be long enough to try and find

20:32out how this machine works well on that

20:35note it sounds as anger to be spending

20:36more time with this machine than I am

20:38with my wife and I think that just about

20:40rounds off this second session and gives

20:43us a bit of a handle now a real handle

20:45on what this machine can do and the

20:48possible problems that we’re going to

20:49encounter in the future with all these

20:51variables so thank you very much for

20:54your time and I’ll catch up with you in

20:56the next session