BEAM COMBINING: High-power fiber-laser beams are combined incoherently
Combining laser beams can boost power at the target far above that produced by a single laser. Incoherent beam combining achieves propagation efficiencies of greater than 90%, while avoiding the complexities of coherent or spectral beam combining.
PHILLIP SPRANGLE, ANTONIO TING, JOSEPH PEÑANO, RICHARD FISCHER, AND BAHMAN HAFIZI
Incoherently combining the beams from multiple high-power fiber lasers has a number of advantages over other beam-combining methods, and can result in compact, robust, low-maintenance and long-lifetime high-energy laser systems. In initial experiments, we have combined the beams from four lasers with a beam director consisting of individually controlled steering mirrors. We achieved propagation efficiencies greater than 90% at a kilometer in range, with a total power of 2.8 kW on a target with a 10 cm radius.
Incoherent beam combining is fundamentally simpler than other beam-combining techniques--for example, spectral or coherent beam combining. Incoherent combining of laser beams is achieved by overlapping the individual laser beams on a target with a beam director consisting of independently controlled steering mirrors with optional adaptive-optics capabilities.1 This approach does not require phase locking or polarization locking of the individual lasers, and can be readily scaled up to a compact and reliable directed-energy system.
The high beam quality and efficiency of fiber lasers make them ideal candidates for directed-energy applications. Although a number of companies manufacture high-power fiber lasers, IPG Photonics (Oxford, MA) currently holds the record, producing more than 3 kW per fiber of single-mode (M2 approximately 1) laser radiation.2 Another company, Nufern (East Granby, CT), expects to have a 1 kW single-mode fiber laser available in 2008.3 These multikilowatt single-mode fiber lasers are robust, compact, nearly diffraction-limited, have high wall-plug efficiency, random polarization, and large bandwidth. A 1 kW single-mode IPG fiber-laser module, emitting at 1.07 µm, has a dimension of approximately 60 × 33 × 5 cm (excluding power supply), weighs about 20 lb, has a wall-plug efficiency of about 30%, and has an operating lifetime in excess of 10,000 hours.
To operate in a single mode, the core of the fiber must be sufficiently small. For example, the IPG single-mode 1 kW fiber lasers have a core radius of about 15 microns. Multimode IPG fibers, on the other hand, operating at 10 kW and 20 kW per fiber, have core radii of about 100 and 200 µm and a beam quality M2 of about 13 and 38. These higher-power fiber lasers with larger values of M2 have a more limited propagation range. In 2008, IPG is expected to have a single-mode fiber laser operating at 5 kW.
In an example that illustrates the essence of incoherent beam combining, the beams from a hexagonal array of seven fiber lasers are combined with a beam director of individually controlled steering mirrors (see Fig. 1). The individual fiber lasers have an initial spot size large enough so that diffractive spreading is not significant over the propagation range. For example, a straightforward calculation involving only diffraction shows that a Gaussian beam with a 4 cm spot size that is focused onto a target at a range of 5 km will have a spot size of only 4 cm on the target. Typically, atmospheric turbulence will cause more beam spreading than diffraction.
FIGURE 1. Incoherent beam combining is achieved when individual beams, each with its own steering mirror, are overlapped on a target.
Incoherent beam combining of fiber lasers is readily scalable to higher total power levels. For multiple incoherently combined fiber lasers, the total transmitted power scales as the number of lasers, while the radius of the beam director scales as the square root of the number of lasers. A 500 kW laser system, for example, could consist of 100 fiber lasers (5 kW/fiber) and have a beam director radius of about 40 cm. Excluding the power supply, the fibers and pump diodes would occupy a volume of about 8 m3.
Atmospheric propagation of incoherently combined beams
The total angular spread of a beam propagating through the atmosphere can, in a simplified analysis, be written as
where θdiff is due to diffraction, θquality is due to the finite beam quality, θturb is due to turbulence, θjitter results from mechanical jitter, and θbloom is due to thermal blooming. For beams less than 100 kW in power, the contribution from thermal blooming can be neglected, and for the purposes of this discussion we will also neglect the small contribution of mechanical jitter. θdiff is the spread of a diffraction-limited beam, and, θquality = (M2 - 1) θdiff is the spread due to higher-order modes. For single-mode lasers (θquality of approximately 0) propagating over long distances, turbulence dominates diffractive beam spreading. For multimode lasers, on the other hand, turbulence contributes significantly less to beam spreading than does beam quality (that is, θquality is much greater than θturb). Adaptive optics can compensate for turbulence but not diffraction and beam quality, so while adaptive-optics techniques can enhance propagation efficiency for single-mode lasers, they have little effect with multimode lasers.
This conclusion is emphasized in an example that shows the propagation efficiency with and without adaptive optics for the four different laser systems described in the table (see Fig. 2 and table). The target is assumed to be a circular disc with a surface area of 100 cm2. Each of these systems has a 50-cm-radius beam director, and delivers a total power of 100 kW. The theoretically ideal case of a diffraction-limited 100 kW laser is included for comparison; the M2>/sup> values in the other three cases are realistic values for the respective laser powers.
FIGURE 2. Propagation efficiency is plotted against range for the laser systems described in the table below. The efficiency enhancement due to adaptive optics can be seen by comparing (top), without adaptive optics, to (bottom), with adaptive optics. Adaptive optics enhancement is more significant with single-mode lasers. For all cases, the turbulence is moderate. In calculating the propagation efficiency, a 100 cm2 circular target is assumed.
A 100 kW incoherent beam combining system having a 50 cm radius beam director can have various configurations. The color blocks to the left of each row correspond to the curve color in Fig. 2. N is the number of fibers and R0 = 50 divided by the square root of N is the radius of the collimating lenses.
The propagation efficiency of each system is plotted as a function of range in a moderately turbulent atmosphere (see Fig. 2, top). The broken black curve represents the theoretically ideal 100 kW laser, and the degradation of each of the three lasers is indicated by the respective color-coded curves; the higher the M2, the greater the degradation. It is frequently stated that for incoherent beam combining, the effective beam quality associated with the combined beams is equal to the square root of N times M2, where N is the number of lasers. This is not relevant for ranges of interest in typical atmospheric conditions. The single-mode curve with N = 33 is always near the theoretical M2 = 1 limit.
Clearly, the greatest enhancement resulting from adaptive optics occurs for single-mode lasers (see Fig. 2, bottom). For the highly multimode laser (M2 = 38), adaptive optics does virtually nothing to improve the propagation efficiency.
NRL propagation experiments
We at NRL have recently completed a proof-of-concept field demonstration of long-range incoherent beam combining at the Naval Surface Warfare Center in Dahlgren, VA. These experiments used four IPG single-mode fiber lasers having a total output power of 6.2 kW. In the initial experiments, we transmitted a total of about 3 kW, and delivered about 2.8 kW to a 10-cm-radius target at a range of 1.2 km. The fiber lasers were operated at half power because of thermal issues in the beam director, which can be readily corrected in the next series of experiments.
A fiber-laser output coupler and the beam expander were used to adjust the focal length (see Fig. 3). Thermal effects caused an axial shift of the focus as the total laser power was increased to about 3 kW. This effect was compensated by changing the separation between the lenses in the beam expander. A beam director, output couplers, and steering mirrors were used in the experiments (see Fig. 4).
FIGURE 3. An output coupler and beam expander was used for each of the fiber lasers in the experimental demonstration.
FIGURE 4. Three of four fiber output couplers are in the foreground, and all four individually controlled steering mirrors can be seen in the background.
The power on the 10-cm-radius target was plotted as a function of time (see Fig. 5). After the output coupler reaches thermal equilibrium (after 200 s) the measured power is 2.8 kW, corresponding to a propagation efficiency of about 90%. Air turbulence causes the beams on the target to wander and change shape with time. At times, the four beams completely overlap, forming a single spot. At other times, four individual beams are separated by a few centimeters. The inset shows a picture of the intensity profile when the four beams are completely overlapped.
FIGURE 5. This plot shows the experimentally measured power at target versus time. The target was a power meter with 45 s response time and 20 cm diameter. The average wind speed during this experiment was about 2.5 m/s. The power on the target was as high as 2.8 kW, and the propagation efficiency was about 90%. Thermal equilibrium in the power meter is reached at about 200 s and the laser beams are turned off at about 320 s. The power buildup on the power meter is due to the temporally changing focal length caused by thermal effects. The inset shows the intensity on the target at 180 s.
Propagation experiments using the NRL fiber lasers on a 3.2 km range at full power are presently taking place at the Starfire Optical Range (Albuquerque, NM). These experiments will verify our computer model of incoherent beam combining and will help us devise closed-loop techniques to compensate for wandering of the beam centroid due to air turbulence. We will also investigate the effects of thermal blooming, which can be an important limitation under certain conditions. This latter investigation will use a stagnation tube to eliminate the cooling effects of transverse airflow.
1. P. Sprangle et al., J. Directed Energy 2, Spring 2007, 273; NRL Memorandum Report, NRL/MR/6790--06-8963.
2. V. Gapontsev, CLEO Europe 2005, paper CJ1-1-THU, Munich, Germany.
3. J. Edgecumbe et al., Conf. Proc. of Solid State and Diode Laser Technology Review (2007).
PHILLIP SPRANGLE is the chief scientist and head of the Beam Physics Branch at the Naval Research Laboratory (NRL), 4555 Overlook Ave. SW, Washington, DC, 20375 (firstname.lastname@example.org), JOSEPH PEÑANO is the head of the Radiation and Acceleration Section at NRL (email@example.com), ANTONIO TING is the head of the Laser Physics Section at NRL (firstname.lastname@example.org), RICHARD FISCHER is a research scientist and engineer in the Laser Physics Section at NRL (email@example.com), and BAHMAN HAFIZI is the head of Icarus Research, P.O. Box 30780, Bethesda, MD 20824-0780 and a research scientist at NRL (firstname.lastname@example.org).