Patent Application
20100142574
The present invention relates to a laser source with coherent beam recombination, in particular to a high-power laser source.
Coherent recombination of laser beams is a technique used to solve the problem of flux stability limitation of gain materials for the purpose of obtaining a high-power laser source. In particular, high-power fiber laser sources are thus produced. This technique makes it possible to obtain a laser beam of high luminance, but also of high coherence and high optical quality (diffraction limited). Laser sources based on coherent beam recombination thus make it possible to envisage power levels that it would not be possible to obtain from a single fiber, owing to the flux stability limitation. To give an example, an ytterbium-doped polarization-maintaining fiber laser makes it possible to extract 500 W of monomode continuous power. A laser source comprising a bouquet of around one hundred of these fibers, with a coherent recombination system, would allow 50 kW of monomode power to be extracted, which power is impossible to obtain from a single fiber. This shows the great benefit of coherent beam recombination.
The general principle of this coherent recombination technique, in which N elementary laser beams are recombined, is to distribute the necessary amplification over N gain media undergoing spatial monomode propagation. The summation of N coherent beams is carried out, as output from the N gain media. N is chosen to be as large as necessary depending on the intended application. The beams to be recombined may in practice be in the form of a one-dimensional (1D) array or a two-dimensional (2D) array as a P×Q. matrix. In the rest of the description, the case of a 1D array is considered, but everything that is described can apply just as well to a 2D array.
Coherent recombination thus consists in summing N coherent beams of the same polarization in parallel, each amplified by propagation in a gain medium. If the N laser beams that emerge from the N gain media are in phase, they interfere constructively and thus constitute a source having a luminance of N2 times greater than that of an elementary amplifier (i.e. 1 beam and 1 gain medium), while maintaining its beam quality (diffraction limited in the case of monomode fibers for example).
However, the N beams follow different propagation paths, and thus undergo different phase variations. These phase variations are due to index variations of many origins: environmental conditions (temperature, vibrations, mechanical stresses, etc.) passages through pumped gain media, etc. These various phase perturbations vary with time.
Thus, these laser sources require a dynamic control device for controlling the phase of each beam, which allows the phase differences between the various beams to be corrected and cancelled out in real time. This ensures that the laser beam resulting from the recombination is very stable under severe environmental conditions. Furthermore, such a device makes it easier to take into account any missing element.
Dynamically controlling the phase of each beam has other known advantages, such as that of providing a beam scanning function. This is particularly beneficial in optronic applications, such as for example designation, tracking or pointing, or communications in free space.
A diagram showing the principle of a laser source based on coherent beam recombination according to the prior art is illustrated in
This source comprises N incident laser beams Lij, where j=1 to N, N phase shifters Dj—one per incident laser beam—and N spatial monomode propagation channels gj—one per incident laser beam. In the case of a high-power laser source, these propagation channels are preferably gain media, advantageously fiber amplifiers. The N incident laser beams are spatial monomode beams of the same polarization.
A coherent recombination system 1 receives a beam made up of N laser sub-beams Laj, obtained as output from the N channels gj. It delivers a recombined laser beam fR as output.
The coherent recombination system comprises a device 2 for taking a part of the output beam to a phase-lock device 3. This phase-lock device delivers the feedback signal to each of the phase shifters Dj. The device 2 may for example comprise an array of microlenses in an example of coherent recombination of free-space beams.
The phase-lock device 3 generates feedback signals according to an appropriate feedback control algorithm applied to a tiny portion of the beam taken by the device 2. The device 3 measures, on this tiny portion of the beam taken, the phase differences between the sub-beams Laj. It generates the feedback signals to be applied to the phase shifters Dj based on these measures. Each of the N phase shifters Dj is thus under closed-loop feedback control via a corresponding feedback signal generated by the phase-lock device 3. This signal determines the effectiveness of the coherent recombination system 1 in bringing the N beams Laj at the output of the channels gj into phase.
The invention relates more particularly to the phase-lock device 3. This phase-lock device must meet various constraints, which determine the efficiency of the coherent recombination system 1 in bringing the beams into phase.
A first constraint is the rate of phase correction. This is because gain media, which are preferably fiber gain media (i.e. fiber lasers), are generally very long and very sensitive to environmental perturbations. This imposes a high phase correction rate, typically of the order of several kHz.
A second constraint lies in the measurement to be made for correcting the phase. It is not a question of measuring the aberrations of a single beam, as in other imaging applications, since the coherent recombination system 1 receives N beams. Therefore the phase shifts between each of these N beams have to be measured so as to recombine them efficiently. However, this does not mean analyzing all the aberrations either: since the N beams are all monomode, each is thus virtually diffraction-limited and free of aberration. The phase shifts to be measured and to be corrected thus correspond to zero-order aberrations, that is to say phase pistons, and possibly to 1st-order aberrations, that is to say “tilts”.
It may be considered that each of the beams received is seen by the coherent recombination system as a sub-pupil. Each of the N sub-pupils seen by the coherent recombination system is phase-shifted relative to the other sub-pupils by a constant phase shift, i.e. by a piston, that is to say a zero-order aberration, or by a tilt of the wave surfaces of each of the pupils, that is to say a 1st-order aberration.
A phase-lock device for a beam recombination system must allow these aberrations to be measured and corrected, at the necessary rate, in order to be very effective.
Various systems for measuring and correcting the phase are known.
In a system based on an interferometric analysis method, the phase shift between the sub-pupil of each beam and a reference is measured. This reference may be delivered by one of the N beams, or more simply by an additional, reference beam, which is phase-modulated. This reference beam acts as a local oscillator. If the propagation media are fiber gain media, the reference beam may be brought to the entry of the coherent recombination system via a non-amplifying fiber.
According to this method of analysis, the phase-lock device 3 goes back to the phase difference of each sub-pupil relative to the reference by demodulating the detected signals in phase quadrature.
This method of analysis has the drawback of being complicated to implement for a large number N of beams since, roughly, it consists in producing an interferometer for each sub-pupil. This therefore requires the provision of a detection system for each channel gj with the alignment problems that this entails, in addition to managing an additional reference beam. In practice, this method may be implemented for a small number of beams, typically up to N=4 for example. In a system in which there may be around one hundred beams, this method becomes very tricky to implement.
Another method is known, which uses the phase shifters Dj (the bandwidth of which must be adapted accordingly) for impressing on each of the N beams, in addition to the phase correction, an RF phase modulation (at several tens or hundreds of MHz), i.e. a much more rapid modulation than the frequency of the phase corrections (typically less than a few tens of kHz). This modulation must be different for each beam. Each sub-pupil detected is thus discriminated from the others by a different RF frequency. The detection uses a single detector, typically a photodiode. The associated signal processing makes it possible to go back to the phase difference of each of the detected sub-pupils by analyzing the actual phase shift at each RF frequency. This method, which requires only a single detector, has the advantage of being simpler to implement than the interferometric analysis method. In addition, it does not require a reference beam. On the downside, it requires more complex processing electronics, especially electronics capable of generating N different RF frequencies. However, this is not its main drawback. This is because with such a method, for applications of the high-power laser source of the data transmission type in free-space communication systems, each of the N beams is modulated at an RF frequency, which frequency could lie within the bandwidth of the signals to be transmitted. This is a disadvantage that limits the application options of this method.
Also known are wavefront analyzers, especially phase-shift interferometers, multiple-wave interferometers, and wavefront analyzers of the Shack-Hartmann sensor type. However, these analyzers are ill-suited for phase-shift measurement in a coherent beam recombination system. This is because they are very complicated and difficult to implement. However, above all their correction rate is low, generally less than 1 kHz, thereby precluding a real-time correction at the desired rate. Furthermore, a Shack-Hartmann wavefront sensor does not allow zero-order aberrations (phase pistons) to be measured.
In the invention, a phase analysis device was sought which does not have the various disadvantages of the devices and methods of the prior art and which meets the desired performance characteristics as regards the capability of measuring zero-order and 1st-order aberrations, simplicity of implementation, and correction rate.
The idea at the basis of the invention is not to process the N beams separately but to consider the output from the N channels gj as one object, of which an image whose aberrations it is desired to correct is taken.
According to the invention, a “phase diversity” processing method is applied to this image in order to correct at least the zero-order aberrations (pistons) thereof.
Phase diversity is a known processing method used in imaging, and especially in astronomy, in particular to correct the images received on a telescope. More particularly, phase diversity is used in this field for correcting a wave surface or wavefront disturbed by the propagation in the atmosphere or by the misalignments of the imaging system (the telescope). It is also used in what are called “cophasing” sensors for producing large-diameter telescopes by optical aperture synthesis, interferometrically recombining several sub-pupils of smaller diameters, which may be segments of primary mirrors or of separate telescopes.
In a simplified manner, the phase diversity method as normally used to improve the resolution of an imaging system, such as a telescope, consists in analyzing a focal image and a defocused image, linked by a known nonlinear relationship. It has the advantage of being simple to implement from the set-up standpoint. On a downside, adaptive image processing means are required.
The method may be described as follows, in relation to
Consider an object o, an image i of which it is desired to form. The telescope 4 is used to acquire the image i of the object o in the focal plane of the telescope. It is possible to go back to the object o and to its aberrant phase from this detected image i. To do this, a second, defocused image i′ of the object o is acquired. This second image is produced by means of a deforming element 5, which adds a known deformation to the wave surface. Thus, additional information about the phase is obtained thanks to the acquisition of the first image i and the second image i′. The two images i and i′ obtained are linked by a relationship characteristic of the deformation induced by the deforming element 5. Adaptive signal processing then makes it possible to go back both to the aberrant phase and to the object o. The second image i′ has therefore provided the phase measurement with “diversity”, hence the name of this method. Typically, the diversity may be provided simply by defocusing the image plane by a known length ΔZ, for example by means of a glass plate as illustrated in
According to the invention, a phase-lock device of the phase-diversity type is used in a coherent beam recombination system to analyze the phase differences between the beams and to control the phase shifters adaptively so as to obtain, as output, an optimally recombined beam.
This thus involves a very different approach from the prior art, by analyzing the phase collectively, that is to say by not considering the incident beam as a juxtaposition of N beams, but rather as a single beam, the image of which it is desired to form. If all the N beams are in phase, the observed image will virtually be that of a single main lobe. The less the phase differences are controlled, the more this lobe will diminish and the more chaotic the image will be.
The invention makes it possible to deal with high-power laser sources based on a large number of beams arranged in a 1D or 2D array. To give an example, it allows a source comprising a 2D array of beams, for example 10×10 beams to be easily dealt with.
The invention therefore relates to a laser source with coherent recombination of N spatial monomode laser beams, comprising N phase shifters controlled by a phase-lock device, characterized in that said phase-lock device comprises:
in order to deliver an optimized recombined laser beam as output.
The processing means may furthermore measure phase tilts and determine corresponding phase corrections.
The optical deforming element is preferably an element for defocusing by a known distance.
Preferably, the N laser beams are amplified by gain media, so as to obtain a high-power laser beam as output, these preferably being fiber amplifiers.
In a variant, said optical device furthermore applies a shaping operation on the recombined beam as output, by impressing a wave surface profile approximated by phase pistons juxtaposed with said phase corrections.
Said approximated profile may be a phase ramp, a phase function for precorrecting subsequent perturbations or, more generally, a shaping function according to the use of the beam.
Still other objects and advantages of the present invention will become readily apparent to those skilled in the art from the following detailed description, wherein the preferred embodiments of the invention are shown and described, simply by way of illustration of the best mode contemplated of carrying out the invention. As will be realized, the invention is capable of other and different embodiments, and its several details are capable of modifications in various obvious aspects, all without departing from the invention.
Accordingly, the drawings and description thereof are to be regarded as illustrative in nature, and not as restrictive.
The present invention is illustrated by way of example, and not by limitation, in the figures of the accompanying drawings, wherein elements having the same reference numeral designations represent like elements throughout and wherein:
a to 4c illustrate an optical system for analyzing the transmittance of a pupil, in order to analyze the phase difference between the sub-pupils corresponding to the various sub-beams forming the image;
In the invention, the phase diversity method is applied to the coherent recombination of N beams in a laser source, as illustrated in
The phase-lock device 3 therefore comprises a first matrix M1 and a second matrix M2 of high-speed detectors, the first for detecting the image i of the wave surface emitted as output from the N monomode propagation channels gj and the second for detecting the deformed image id on an optical element 32 applying a phase deformation. The phase-lock device 3 further includes a signal processing device 31 capable of carrying out computations of the Fourier transform type, autocorrelation type, etc., in order to determine the aberrant phase of the image i and to deduce therefrom the phase corrections to be applied as input of each of the channels gj, to the N phase shifters Dj.
It should be noted that it is quite possible, and known, to detect both images i and id on a single matrix of detectors. This situation may be implemented by a person skilled in the art without any particular difficulty.
Such a phase-lock device according to the invention is simple to implement as it requires only one or two detector matrices. There is no reference beam to be provided, nor any modulation of the useful signal.
In a simple exemplary embodiment, the optical deforming element 32 is an element that defocuses by a known distance ΔZ, typically a plate of thickness ΔZ. The phase therefore undergoes a nonlinear deformation. In this example, the deformation is known, thereby simplifying the data processing, but it is quite possible to use a phase deforming element that applies any deformation. The sole limitation comes from the number of parameters needed to model the deformation, which makes the data processing for accessing these parameters relatively complex.
Hereafter, the simple case of an optical element 32 is considered, this being an element that defocuses by a known distance.
For the needs of the modeling, and to draw a parallel with the device shown in
The method thus consists in analyzing the pupil equivalent to the wave surface 43 emitted by an array of N laser beams output by the N channels gj.
It will be recalled that a laser beam has a spherical Gaussian field distribution. Thus, the transmittance of the virtual pupil 41 placed at the waist of the N beams is the juxtaposition of N sub-pupils spj having a Gaussian amplitude distribution (illustrated in
Assuming that the beams are perfectly parallel to the propagation axis, the tilts of the sub-pupils are zero.
As regards the modeling, the system is therefore equivalent to that of a known point object o and of a collimating lens 45 emitting a plane wave 44 perturbed by passage through an aberrant pupil 41, having the above-mentioned characteristics.
The following assumptions are made:
i(x)=o(x)*h(x)
where * is the convolution operator, i and o are the image and object fields and h is the impulse response of the optical device 40 (propagation, optical instrument and sensor); here the noise terms that would taint the image (i=o*h+b) are neglected.
The image of intensity Im, seen by a matrix detector M1, typically a CCD detector, placed at the focus of the analysis system is then Im=|i|2=i×i*.
When in the plane of the pupil, this equation becomes, in frequency space:
I=O×H,
where I, O and H are the Fourier transforms of i, o and h. H is the transfer function of the optical device 40.
Since the incident wave is assumed to be plane before being perturbed by the pupil 41, O(x) is in our case a Dirac function. Therefore O(u)=1 and i(x)=h(x).
In addition, we may write:
where P is the amplitude of the spherical Gaussian wave of each beam.
The phase of each of the beams is decomposed on the basis of the Zernike polynomials:
By assuming that the sub-pupils are phase-shifted only by a piston, it is possible to truncate the above sum to the 1 order polynomial, i.e. φn(u)=αn. It will be noted that for taking tilts into account, it would be necessary to truncate to the 3 order polynomial.
Therefore, in the focal plane, the following may be written:
The phase diversity is provided by defocusing the focal plane by the amount ΔZ. To denote the quantities corresponding to the defocused image imd, we will use the same notations, but with the suffix d, i.e.:
where φd is the phase diversity provided by the defocusing:
Thus, knowing o and φd, and by measuring the intensities Im and Imd of the observed images i and id, we are able by appropriate signal processing to go back to the desired phases φn.
This processing may for example consist of an optimization of a criterion J as a function of N pistons αn.
For example, a criterion J resulting from the least-squares method may be used, i.e.
J(α1, . . . , αn)=∥Im−ii*∥2+∥Imd−idid*∥.
The optimization of such a criterion J then consists in minimizing the difference between the intensities Im and Imd and the intensities of the images reconstructed from the estimation of the pistons φn in the expressions for H.
By applying Parseval's theorem and using I(u)=H(u), the following may be written:
The N desired pistons αn are then those which set this criterion J to zero.
The invention is not limited to just this criterion. Other criteria that are more sophisticated and astute may be employed without departing from the spirit of the invention.
In the foregoing, only zero-order phase shifts (phase pistons) were considered. It should be noted that to take the tilts into account requires only adding two 1st-order terms in the decomposition of the phase on the basis of Zernike polynomials.
Thus, a phase-lock device 3 according to the invention as illustrated in
It is possible to obtain the recombination in far field of a beam fR, the radiation pattern DR of which is monomode, as illustrated in
What is thus obtained is shaping of the beam fR and generation of an optical wave surface that are optimized for the application.
This in-phase shaping function illustrated in
It is possible to employ more elaborate shaping functions to generate more complex wave surface profiles for use in various applications.
This shaping of the beam fR with optical wave surface generation is most meaningful when the number N of beams (or sub-pupils) in question is very large. This is because only the piston of each sub-beam, i.e. a phase “plateau”, is controlled, and the resulting overall wave surface is therefore made up of “staircase steps”. The more numerous the steps, i.e. the larger N, the closer the emitted wave surface approaches the desired shape.
A more elaborate beam shaping example involves the deflection or scanning of the beam fR, which applies in particular to optronic applications such as for example designation, tracking or aiming, or free-space communications.
To produce such a beam deflection, in addition to annulling the relative phase pistons between the sub-pupils, a phase ramp profile r is impressed in such a way that each of the sub-pupils is shifted by a constant phase piston pj relative to its neighbors. This is what
The main lobe of the radiation pattern DR is deflected through the angle θ of the phase ramp. On the downside, its amplitude is somewhat reduced.
As explained earlier, the larger the number N of beams, the better the approximation of the ramp by the phase pistons. A very good resolution of the pointing error is then obtained.
Another implementation example is illustrated in
In a similar variant, it is possible to impress wave surface profiles of diverse forms for diverse applications. For example, starting from a two-dimensional matrix array 60 of channels gj, it is possible to impress a phase function f(φ) such that the beam is focused as a rectangular spot 61, as illustrated in
Thus, treating the N beams collectively as a single image benefits from a collective method that allows the phase on a very large number N of beams to be analyzed. To give an example, the phase diversity systems used in astronomy make it possible to process and optimize images consisting of 512×512 pixels with several adjustment parameters per pixel. In the invention, there is only a small number of adjustment parameters—three at most—namely the piston and optionally the tilt in two directions of each of the N monomode beams. The number N of beams that can be treated with such a coherent recombination system according to the invention may be correspondingly larger.
The invention applies in particular to high-power laser sources based on coherent recombination of beams using gain media as spatial monomode propagation channels gj, and applies especially to high-power fiber laser sources, that is to say those in which the gain media gj are fiber amplifiers.
It will be readily seen by one of ordinary skill in the art that the present invention fulfils all of the objects set forth above. After reading the foregoing specification, one of ordinary skill in the art will be able to affect various changes, substitutions of equivalents and various aspects of the invention as broadly disclosed herein. It is therefore intended that the protection granted hereon be limited only by the definition contained in the appended claims and equivalent thereof.
| Number | Date | Country | Kind |
|---|---|---|---|
| 05 08542 | Aug 2005 | FR | national |
The present Application is based on International Application No. PCT/EP2006/065203, filed on Aug. 10, 2006, which in turn corresponds to French Application No. 05 08542, filed on Aug. 12, 2005, and priority is hereby claimed under 35 USC §119 based on these applications. Each of these applications are hereby incorporated by reference in their entirety into the present application.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP06/65203 | 8/10/2006 | WO | 00 | 2/12/2008 |
