Disc versus fiber

Have solid-state lasers finally matured enough to offer competition?

Th 166732

Have solid-state lasers finally matured enough to offer competition?

One of the highlights of last year’s ICALEO conference (organized by the Laser Institute of America-LIA) was an invited paper given by Dr. W. O’Neill, from the University of Cambridge UK, on new developments in solid-state lasers. He explained the principals of the diode disc laser and two versions of the fiber laser. To get an idea of the impact on these technologies one may compare the flashlamp-pumped Nd:YAG laser with a steam engine, the disc laser with the auto motor, and cladding-pumped fiber lasers with the electric motor. Where a skilled technician is needed to operate the steam engine, a trained operator can just switch the electric motor on and off as needed.

Th 166732
Click here to enlarge image

There are still a large number of flashlamp-pumped Nd:YAG lasers used in industrial production. Over time, operators have learned how to run this equipment-similar to mechanical engineers who learned how to run a steam engine. Flashlamp-pumped Nd:YAG lasers have a limited beam quality that is due to the thermal lensing of the Nd:YAG rod. This lensing effect is due to the temperature gradient over the cross-section of the rod. The lensing effect will be reduced if the temperature over the volume of the active material is more homogeneous. The first step in this direction was made by replacing the flashlamp with diode bars that emit light in a much smaller spectrum so they pump more efficiently and, therefore, less heat is generated in the rod. The next step is to change the geometry of the rod. Because heat has to be withdrawn from the surface of the active material one may choose a geometry where the surface-to-volume ratio is increased compared to the traditional rod (typical dimensions: diameter 9 mm, length 150 mm). There are two ways to go. One is increasing the diameter and reducing the length, which leads to the disc laser. The other is reducing the diameter and increasing the length, which leads to the fiber laser. Table 1 gives typical data for dimensions of the active components of the rod, disc, and fiber lasers.

By changing the geometry one is forced to adapt the way pump light is guided into the active material, how the resonator is created, and how the output coupling is realized.

Th 166728
Figure 1. Basic setup of a disc laser with four discs.
Click here to enlarge image

The disc laser-Figure 1 shows the basic setup of a disc laser. The pump source is a stack of diodes. Curved mirrors are used to couple the pump light into the disc. The resonator is formed by a rear mirror and an output-coupling mirror. The thermal lensing is low compared to the rod laser because the surface-to-volume ratio is higher. The power per unit of volume is high. Water cooling is needed because the cooled surface is relatively small compared to the power per unit volume. Water-cooling is also needed for the diode stacks. Several discs may be combined to achieve multi-kilowatt output power.

Fiber lasers-There is one fundamental difference between the rod and disc laser on one side and the fiber lasers on the other side. Fiber lasers don’t use Nd:YAG as active material. They use glass fiber doped with rare elements. Beside Neodymium (Nd) one may use Erbium, Ytterbium, or Thulium. The wavelength of the laser radiation depends on the type of doping, but in all cases the wavelength is in the 1 μm region. In the early days of industrial lasers Nd-doped glass rods were used to build lasers. Their drawback was the limited thermal loading that they could withstand. This drawback is overcome by using fibers that have a very high surface-to-volume ratio. The laser power per volume is low compared to the disc and rod laser but may be compensated by increasing the volume.

Th 166729
Figure 2. Schematic drawing of double-clad fiber.
Click here to enlarge image

Fibers (see Figure 2) have a doped core that is surrounded by two layers of cladding, an inner cladding and a low-refraction index outer cladding. The core diameter is typically 9-40 μm. The outer diameter is in the range of 100-750 μm.

Th 166730
Figure 3. Principal of a fiber laser with end-face pumping.
Click here to enlarge image

Fiber laser with end-face pumping-The typical setup of fiber laser with end-face pumping is shown in Figure 3. The pump source is again a stack of diode lasers. Lenses and mirrors are used to focus the light of the diodes into the end faces of the fiber. Pump power is mainly going through the inner cladding and from this cladding to the doped core. The resonator is formed by the rear and output coupling mirrors. Water cooling is only needed for the diode stacks. The fiber may be cooled by air due to the high surface-to-volume ratio and the low power per unit volume.

Th 166731
Figure 4. Principal of a fiber laser with cladding pumping.
Click here to enlarge image

Fiber laser with cladding pumping-Figure 4 gives the basic setup of this type of laser. The heart of the laser is a fiber with a length of 5 to 20 meters. This fiber has a doped core with an undoped cladding. The pump light is generated by a set of single diodes. Fibers are used to couple the pump light into the inner cladding. Subsequently the pump light couples into the doped core. The resonator is formed by Bragg gratings at the ends of the doped fiber. Compared to the disc and end-face pumped fiber laser this laser is less complex in its construction. However, new and more advanced production technologies are needed to produce this laser. The output power per 5- to 10-meter length is low compared to the power generated per disc. But modules can be combined by fusing fibers using the same technology that is used to combine the elements of the basic module. Recently IPG Photonics (Oxford, MA) announced the delivery of a 10kW cladding-pumped fiber laser. With a wall plug efficiency of 30% or better, water cooling is only needed for high-power units (>1 kW).


Th 166733
Click here to enlarge image

Most important for industrial applications are the costs per unit of production, e.g. cut length, weld length, etc. Every end user will have to do homework. However, an indication of the cost per weld volume is given in Table 2. From this data one may concluded that fiber lasers are an attractive alternative to the conventional CO2 and rod Nd:YAG laser. Keep in mind that the data in Table 2 are based on bead-on-plate welding trials and do not include system components and labor costs.


Compared to the flashlamp-pumped Nd:YAG laser, one advantage of the diode-pumped disc and fiber lasers is that no downtime is needed for lamp replacement. Lamp replacement also has the risk of introducing contamination to the optical resonator components. The disc laser and the end-pumped fiber laser still use discreet optical components (lenses and mirrors) to form the resonator. So skilled engineers are needed for alignment. Normally this will be done at the laser manufacturer’s facility. In-field realignment may only be needed when lenses or mirrors are replaced-a situation that may never occur over the lifetime of the laser. A fiber laser with cladding pumping will only have discreet optical components in the beam-delivery optics. So the needed maintenance will be very low.

Final remarks

It is interesting to observe the recent developments in solid-state lasers. The approach chosen by the companies that build disc lasers is improving the concept of the rod lasers. They use the same material and lenses and a similar mirror. They are also forced to use water cooling. The real invention in laser technology is the fiber lasers with cladding pumping. The technology used to build these lasers is mainly developed for telecommunications application. The downturn in this industry has forced manufactures to look for other applications-which they found in the industrial application of material processing.

More in Surface Treatment