Laser Applications Newsletter
25 November 1999
Laser Kinetics Inc.
Mtn. View, CA 94041
Issue 3 , Vol 3


ICALEO '99, the 18th International Congress on Applications of Laser & Electro-Optics, just finished up in San Diego. While the fabled San Diego weather didn't meet its usual standards (I heard a lot of comments such as "It was warmer back in Indianapolis"), ICALEO continued to solidify its status as the world's most important laser applications conference. The attendance was well over 400, and more than 200 papers were presented over the 4-day congress.

Diodes not Ready to Take Over Yet
The topic of this year's plenary session was Diode Laser Technology. Richard Craig of diode manufacturer SDL claimed that diode lasers will replace all other kinds of lasers. This may well be, but it didn't look like this would happen any time soon. The current areas of material processing where diodes are useful are soldering, plastic welding and heat treating. These are low-irradiance applications; Nd:YAG and CO2 lasers do welding, drilling and cutting because they can deliver more than 1,000,000 watts per square centimeter to the work. The problem with semiconductor lasers is that it's easy to make little ones, but you can't make big ones. Material processing applications require high power, and all known methods for getting high power out of semiconductor lasers involve ganging them up. Unfortunately, a collection of lasers doesn't work as well as a single large laser unless all the emission is in phase. The brightness, or intensity per unit angle, is not high enough for most material processing work. This is the current stumbling block, and I didn't hear any good ways around it (that doesn't mean it's impossible; it's just that no one was willing to describe anything promising).

So, while it's easy to get kilowatts out of diode arrays, it's not as simple to focus this power into a small enough spot to do keyhole welding, percussion drilling or high-quality metal cutting. As Reinhart Poprawe of the Fraunhofer Institute noted, "we are just not bright enough". Until we do get bright enough, Nd:YAG and CO2 lasers will continue to be the workhorses of material processing. Right now, the best way to get high brightness from a diode laser is to pump a laser rod with it.

Little Stuff
This year, ICALEO had a separate microfabrication conference. This area is growing rapidly, partly because many microfabrication applications are used in the electronics industry and partly because lasers offer excellent solutions to manufacturing problems in other industries.

Lasers are great at drilling holes, and a lot of holes need to be drilled in this world. Circuit board vias (little holes that let different layers in circuit boards connect to each other) have been mechanically drilled in the past, but the holes are getting smaller so they don't take up as much room on the board and tiny little drills break very easily. Lasers from CO2 to excimer are now taking over the via-drilling market. A CO2 laser can drill 1000 holes per second, while UV lasers do a cleaner job at a lower process rate.

What lasers can take away, they can also give. Laser deposition is used in electronics and medical devices to "write" materials in precisely defined patterns.

Outer Space in the Lab, or Take me to Your Ladder
Since we're going to have a Space Station, we'll have to do welding on it. Welds in space, however, are liable to be much different than welds performed on Earth because things like convection will be drastically altered by the lack of a strong gravitational field. This is the reason Osaka University is examining the effects of microgravity on laser welds.

Now you might expect this to be a costly project, involving flights in zero-G aircraft if not Space Shuttle experiments. The fellows at Osaka are smarter than that. They do their experiments right in their lab by dropping the test fixture off a ladder! With a fiber-delivered YAG beam and some timing electronics, welds are made during the short drop time before the test rig hits the padded floor.

Oh, by the way, they are able to get good weld quality in all common structural materials at atmospheric pressure by control of pulse shape. They haven't gotten good aluminum welds in vacuum/ zero-G yet, but at least it's easy for them to do process development.

Weld Monitoring Still not as Good as Weld Process
One serious obstacle to the adoption of laser welding for critical components is that it's impossible to tell if a given weld is good without destroying the part. A lot of work has been done over the years but the signals being read don't correlate 100% to weld quality. Dave Farson at Ohio State University has examined this problem and explored sensor fusion and fuzzy logic as methods to improve the reliability of weld monitoring.

For example, you can look at a weld's temperature with an infrared detector while simultaneously reading the plasma light emission, plasma charge and acoustic emission. Any one of these signals can give a false reading, but if you do a statistical analysis of all the signals, it's possible to get a more reliable answer.

The bad news is that more reliable is still not reliable. In Farson's tests, the equipment was 87% accurate in detecting weld defects. Since any acceptable industrial welding process produces better than 99% good welds, the use of such a monitoring system will not enhance product quality; it will only increase costs due to rejection of good parts.

HRL Fixes Fiber Delivery
As part of the Precision Laser Machining consortium, HRL (which seems to be the new name for Hughes Research Laboratories) was given the task of preserving the high beam quality of the TRW-developed high-power diode-pumped Nd:YAG laser when the beam was delivered through an optical fiber. It appears that they have accomplished this and more.

HRL took a low-power, high-beam quality Nd:YAG laser and delivered it through a single-mode fiber (which preserves beam quality but can't handle high power) to an end effector for a high-power fiber. They then sent the beam down the high-power fiber and into 3 Nd:YAG amplifiers. The beam coming out of the fiber, of course, was of very poor beam quality. The amplifiers added their own distortion, but that didn't matter because the final optical component was a phase-conjugate mirror. This is a magic device that works by 4-wave mixing (If you really want to know about this, read Yariv, A. and Pepper, D. 'Amplified reflection, phase conjugation and oscillation in degenerate four-wave mixing' Opt. Lett. 1, 16-18, 1977. Then explain it to me.) and returns a beam that has the exact opposite wavefront distortion as the incident beam. Therefore, after this beam goes back through the amplifiers and optical fiber, it reappears with the same beam quality as the low-power laser that we started with, only amplified a lot.

The kicker here is that the only laser that has to be any good is the little seed laser; you don't need diode-pumped slab amplifiers. HRL used old Lumonics lasers with the optics removed. It's not trivial to make a phase-conjugate mirror, but HRL did it and got a diffraction-limited beam out of a multi-mode fiber. I don't know the practical limitations of this technique, but it will be important if it can work in industry.



The next ICALEO (and the first without an apostrophe) will be held October 2-5, 2000 at the Hyatt Regency in Dearborn, MI. It will be a big one, with conferences on automotive applications (of course), microfabrication, aerospace, electronics, medical and rapid prototyping. There will be short courses on laser application topics, too. The call for papers is out; look at

The first big symposium of the year will be in San Jose. Photonics West covers biomedical optics, high-power lasers and applications, integrated devices and applications and electronic imaging in 85 different conferences.

Call 650 575-4919 or e-mail us for more information.