Open Access
Add to BibSonomyAbstract
A direct-liquid-cooled Nd:YLF thin disk laser resonator is presented, which features the use of refractive index matching liquid (RIML) as coolant. Highly uniform pump intensity distribution with rectangular shape is realized by using metallic planar waveguides. Much attention has been paid on the design of the gain module, including how to achieve excellent cooling ability with multi-channel coolers and how to choose the doping levels of the crystals for realizing well-distributed pump absorption. The flow velocity of the coolant is found to be a key parameter for laser performance and optimized to keep it in laminar flow status for dissipating unwanted heat load. A single channel device is used to measure the convective heat transfer coefficient (CHTC) at different flow velocities. Accordingly, the thermal stress in the disk is analyzed numerically and the maximum permissible thermal load is estimated. Experimentally, with ten pieces of a-cut Nd:YLF thin disks of different doping levels, a linear polarized laser with an average output power of 1120 W is achieved at the pump power of 5202 W, corresponding to an optical-optical efficiency of 21.5%, and a slope efficiency of 30.8%. Furthermore, the wavefront aberration of the gain module is measured to be quite weak, with a peak to valley (PV) value of 4.0 μm when it is pumped at 5202 W, which enables the feasibility of its application in an unstable resonator. To the best of our knowledge, this is the first demonstration of kilowatt-level direct-‘refractive index matching liquid’-cooled Nd:YLF thin disk laser resonator.
© 2016 Optical Society of America
1. Introduction
In recent years, direct-liquid-cooled thin disk lasers (DLCTDLs) have attracted growing interests because of their excellent heat dissipation ability, especially when they are intensively pumped by high power laser diodes [1–4 ]. As is implied by the name, the disks in this configuration are cooled by liquid flowing over the largest surfaces of them, which can carry the heat away easily and rapidly. Compared with the conventional conduction-cooled solid-state laser, this novel convection-cooled laser presents new characteristics. Firstly, the possible surface bending introduced by the disk welding procedure, which is the most critical one for disk module, can be circumvented. Secondly, thermal lensing effects and external constraint on the disks are relaxed dramatically, due to the colinearity of the heat flux’s direction and the laser beam’s axis, and the elastic support of the gain media. Thirdly, it is compact and suitable for table-top applications, because of the integration of many pieces of disks into one gain module.
So far, the reported DLCTDLs were mostly based on Nd:YAG. In 2010, 100 kW average power was extracted from six gain modules, reported by Textron Defense Inc., in which the Nd:YAG ceramic thin disk were used as gain media and immersed in coolant [5]. One more promising case is the “liquid laser” scheme that was expected to deliver an average power of 150 kW [6]. However, the details have not been revealed yet. In 2013, X. Fu et al. reported a DLCTDL resonator, where a Nd:YAG disk was used as gain medium and deionized water as the coolant [7]. Unfortunately, the oscillator’s single pass loss was as high as 5.6%, considering the absorption coefficient (0.143 cm−1 at 1064 nm) of deionized water and the coolant thickness of 4 mm [7]. If multiple coolant layers are contained in the multi-disks module, loss will increase exponentially and limit the final performance severely. Furthermore, the disk should be specially coated to prevent the coating (usually antireflection coated at 1064 nm and 808 nm) being detached from the surface of the disk because of the rushing coolant. Instead, the heavy water used as a coolant can give a much lower absorption coefficient (less than 0.03 cm−1 at 1064 nm and 808 nm). However, the gain medium is recommended to be positioned at Brewster’s angle in order to reduce the interface losses (Fresnel reflection) resulting from the refractive index difference between the gain medium and the coolant [2], which makes the configuration complicated and the output power sensitive to the misalignment of the gain medium from Brewster’s angle. Alternatively, adopting a coolant, whose refractive index equals that of the gain medium and absorption coefficient is as low as possible at pump wavelength and laser wavelength, can relax these problems. Although Nd:YAG is the most popular candidate as the gain medium due to its high thermal conductivity, its high refractive index of 1.82 at 1.06 μm makes it difficult to offer a suitable refractive index matching liquid (RIML). On the contrary, Nd:YLF has a much lower refractive index of 1.47 at 1047 nm, for which it is much easier to find out a suitable RIML. Moreover, in the literature above, the cooling capacity of the channels inside the gain module, which limits the maximum permissible thermal loads, was not investigated experimentally. The uniformity of the coolant along the length and width of the channels was not mentioned neither.
In this contribution, a kilowatt-level direct-liquid-cooled Nd:YLF thin disk laser resonator is presented. The RIML is used as coolant, so the gain media can be positioned at normal to the laser direction instead of at the Brewster’s angle. High uniformity are achieved for distributions of both pump intensity and flow intensity in the gain module by the use of rectangular planar waveguide and the elaborative design of cooling channel structure, respectively. The convective heat transfer coefficient (CHTC) is measured experimentally by using a single channel device at different flow velocities. The maximum permissible thermal loads in the disks are obtained numerically. Lasing performance is investigated at different output couplers and coolant flowing velocities. When the uniformity of pump intensity distribution and the flow velocity are 8% and 5 m/s respectively, 1120 W multimode laser output is achieved at the pump power of 5202 W with an optical-optical efficiency of 21.5%. To the best of our knowledge, this is the first demonstration of direct-RIML-cooled Nd:YLF thin disk laser resonator.
2. Design and experimental setup
The schematic diagram of the direct-RIML-cooled Nd:YLF thin disk resonator is shown in Fig. 1 . The overall dimensions of the laser are 300 mm (x) × 90 mm (y) × 1700 mm (z). Dual-end-pumped plano-concave stable cavity has been adopted, which consists of an output coupler (OC), a high reflector (HR, Rs > 99.9% @ 1047 nm) and two dichroic mirrors (M1, M2). The two dichroic mirrors are high transmittance (HT, Ts, p > 99.7% @ 805 nm) coated at 805 nm and high reflectance (Rs > 99.8% @ 1047 nm) coated at 1047 nm for light at the incidence angle of 45°. The cavity mirror is placed closely to the two dichroic mirrors to make the cavity length as short as possible. The length is chosen as 33 cm finally. The gain module contains ten pieces of Nd:YLF thin disks, eleven cooling channels and two fused silica windows. The circulating RIML passes through the gain module from the inlet to the outlet. Inside the gain module, it flows over the two largest surfaces of the disks, and takes away the deposited heat in the way of convection cooling. All the disks are uncoated. Both windows of the gain module are coated on one side (on the air side), which is antireflection coated at 805 nm and 1047 nm. On each side, a metallic planar waveguide is used to mix and homogenize the high power 805 nm pump beam from four laser diode (LD) stacks, after which the reshaped beam is projected onto the gain module.
Fig. 1 Experimental setup of the direct-RIML-cooled Nd:YLF thin disk resonator. LD, laser diode; CL, cylindrical lenses; W, planar waveguide; IS, imaging system; M1, M2, dichroic mirrors; HR, high reflector; OC, output coupler. GM, gain module.
2.1 Pump system
The pump system includes two parts, as shown in Fig. 1. One is the LD stacks, and the other is the coupling system involving two cylindrical lenses, a planar waveguide, and an imaging system.
In total, there are four LD stacks, two on each side. Every stack is comprised of 30 LD bars, having an emission area of 54 mm × 10 mm and providing a maximum output power of 1800 W at 805 nm. Each LD bar is collimated individually by a microlens. On each side, the pump beam with p-polarized from two LD stacks is combined spatially with a stripe mirror and reshaped by the coupling system. Finally, the pump beam is projected onto the disk, with a pumping area of 45 mm × 18 mm.
It has been shown that the uniformity of pump beam greatly influences the output power of oscillating laser [2]. With a better pump uniformity, the temperature gradient inside the disk could be greatly relaxed [8]. Therefore, a high quality coupling system is required. In our case, the coupling system consists of two cylindrical lenses, a rectangular planar waveguide and an imaging system. The two cylindrical lenses have focal lengths of 170 mm in x-axis direction (fast axis) and 80 mm in y-axis direction (slow axis), respectively, which focus the collimated pumping beam into a rectangular planar waveguide. The planar waveguide is 100-mm-long along the z-axis, 3 mm along the x-axis, 7.5 mm along the y-axis. An imaging system with a magnification of 6 is used to prepare a homogeneous pumping area with dimensions of 45 mm × 18 mm in the middle of the gain module. Figure 2 shows the pump profile acquired by a CCD camera (SP620U, Ophir-Spiricon, USA). The root mean square (RMS) of the pump power density, which is used to evaluate the pump uniformity [8], is 8% (inside the dotted rectangle in Fig. 2) typically in our experiment, showing a good pump uniformity. Furthermore, the RMS is always less than 10% even the pump beam propagates from the first disk to the tenth. The coupling efficiency of the pump system is 85% approximately.
2.2 Cooling system
As the heart of the laser system, the gain module is the most important part to enable an excellent performance. Figure 3(a) shows the schematic view of the gain module used in the experiment. Figure 3(b) is the top view of the gain module. The circulating RIML flows from the inlet to the outlet with a temperature of 22 °C. A flow homogenizer consisting of one honeycomb and three screens with different mesh sizes is used for breaking up larger eddies and suppressing the non-uniformity and fluctuations. Ten pieces of Nd:YLF disks and two windows are mounted inside the twelve steel frames. It should be noted that the disks in our experimental setup are flexibly supported, which significantly reduces the external constraint on the disk, thus the thermal stress caused by thermal gradient is dramatically alleviated accordingly [2]. A spacing of 0.5 mm between adjacent disks is chosen for the coolant flow. The size of the gain module is 260 mm (x) × 90 mm (y) × 55 mm (z) approximately.
Fig. 3 (a) Schematic view of the gain module, (b) the top view of the gain module, and (c) the details of a flow channel.
Figure 3(c) shows the design details of an individual flow channel based on the existing wind tunnel designs [9–12 ], which is divided into nozzle part, test section, and diffuser zone. The nozzle, whose contour is described by a fifth-order polynomial, has a contraction ratio of 3:1. Just after the contraction, the RIML enters the 0.5-mm-width (z) × 48-mm-height (y) × 38-mm-long (x) test section which consists of a developing zone, a laser disk zone and a relaxation zone. The developing zone is used to provide fully developed laminar flow before the RIML goes into the laser disk zone, while the relaxation zone is used to avoid the downstream effects [12–14 ]. At the exit of test section is a 25-mm-long diffuser zone, with a circular curve having a total angle of 3°, which provides some pressure recovery [12, 15 ].
The physical performance of the RIML (Cargille-Sacher Laboratories Inc., USA, the code of liquid is 5610) is shown in Table 1 . The main component of the liquid is siloxane which has low toxicity and corrosivity compared with some inorganic liquid. The absorption coefficient of the RIML at 1047 nm is less than 0.01 cm−1, corresponding to an absorption loss of 1% over a length of 1 cm, which is 92% lower than that when deionized water (absorption loss of 13% [7]) is used. Moreover, the refractive index value of 1.47 of the liquid matches that of the Nd:YLF disks and quartz windows. Unfortunately, the RIML has lower specific heat and thermal conductivity and higher viscosity compared with heavy water, which is not favorable for heat dissipation. However, if flow velocity is faster than 4 m/s and circulating pump is strong enough, things will be different, which will be the main subject of the following paragraphs.
Table 1. The parameters of the RIML
(a) The flowing state
In fluid mechanics, the Reynolds number (Re), a dimensionless number, is usually used to characterize the fluid’s condition. Laminar flow, transition flow, and turbulent flow correspond to the conditions with Reynolds number of less than 2300, 2300~4000, and larger than 4000, respectively. Reynolds number can be expressed as
where L is the hydraulic diameter of the channel, υ is the kinematic fluid viscosity coefficient and u is the flow velocity of the RIML. The maximum available value of the flow velocity is 5 m/s. For a cuboid channelwhere S = 24 mm2 is the open area and χ = 97 mm is the wetted perimeter of the channel. According to Eq. (1) and (2) , the Reynolds number is calculated to be 177. So we conclude that it is in the state of laminar flow (The experiment below will also verify this). By the way, although the CHTC of turbulent flow state is higher than that of laminar flow state, it is not easy to compensate for the wavefront aberration induced by turbulent flow field using deformable mirror or some other adaptive optics systems due to its randomness and high speed perturbation.A simple setup with only one channel is prepared and tested to analyze the flowing dynamics of the fluid inside the gain medium and its cooling capacity. Figure 4 illustrates the structure of the test device which is similar to one of the individual flow channels (in Fig. 3(c)) in the gain module except for the coolant’s flowing direction (along the y-axis direction in the test device). To leave out the complicated installing and fixing of gain module, the flowing direction is along the x-axis in the gain module as shown in Fig. 3. Two core issues are investigated in the following two sections, including the distribution of flowing velocity and cooling capacity of the channel.
As mentioned above, the wavefront aberration caused by the turbulence is not easy to be corrected, so it is quite necessary to make the coolant in laminar flow state in the laser disk zone. Particle image velocimetry (PIV) is a reliable and accurate technique for measuring velocity field of a liquid flow [16, 17 ]. More detailed and completed description about this technique can be found in the references [16–18 ]. Figure 5 shows the ensemble-averaged 2D velocity vector maps obtained with PIV technique in the laser disk zone. The arrows characterize the motion vector of the tracer particles, representing the flowing direction of the coolant at some certain time. The coolant used here is deionized water with a viscosity coefficient of 1 × 10−6 m2/s. When the flow velocity is adjusted from 2 m/s to 5 m/s, according to Eq. (1) and (2) , the Reynolds number ranges from 1979 to 4949. As shown in Fig. 5, it is obvious that the velocity distribution is very uniform when the Reynolds number is small (Fig. 5(a)). With increasing flow velocity, the uniformity becomes worse and worse (Fig. 5(b)-(d)). Meanwhile, the fluid state enters into transition state firstly and the turbulent state finally. In this experiment, when the fluid state changes from laminar flow to turbulent flow, the verified critical flow velocity is between 3 m/s and 4 m/s approximately, instead of 2.3 m/s, which corresponds to the critical Reynolds number of 2300. This discrepancy is attributed to system error in the experiment and the fact that the critical Reynolds number corresponds to a numerical range instead of a fixed value.
Fig. 5 The velocity distributions in the laser disk zone with the Reynolds number values of 1979 (u = 2 m/s), 2969 (u = 3 m/s), 3959 (u = 4m/s) and 4949 (u = 5m/s).
The viscosity of RIML is 28 times as much as that of water, so a much lower (28 times lower in theory) Reynolds number can be expected when RIML is used. This means that the RIML should still be in a laminar state even when the flow velocity is tens of meters per second. Actually, the flow velocity is only 5 m/s in the gain module, which guarantees a uniform distribution of flow velocity. Therefore, the designed cooling channel inside the gain module fulfills the requirement of laminar flow with quite large margin.
(b) Cooling capacity of the channel
Except for achieving a uniform distribution of flow velocity, the cooling ability is also of great importance. A higher CHTC can lend the gain media higher thermal load capacity. Another test facility has been constructed to measure the CHTC with different flow velocities. The structure is similar to that shown in Fig. 4, except for that one of the fused silica windows is replaced by a copper-block with same size. A thin insulating layer is coated between the front surface of the copper (We define the two big surfaces as front surface and back surface shown in Fig. 6(a) ) and the resistor for isolating. The thin film resistor (made up of platinum) with a thickness of less than 1 μm and resistance value of 7.2 ohms is used as a simulative heat resource. The effective aperture of the thin film resistor is 45-mm-height (y) × 15-mm-width (x) as shown in dashed box in Fig. 6(b). The coolant is the RIML. A thermal camera (SC655, FLIR, Sweden) is used to acquire the temperature distribution on the thin film resistor depicted in Fig. 6(b). With the heat flux density of 5.7 × 104 W/m2 which is achieved by controlling the current on the simulated heat resource and liquid temperature at inlet of 22 °C, the results of the temperature distributions on the surface of thin film resistor and in the flow direction (y direction) down the centreline (x = 9 mm) with different flow velocities from 1 m/s to 4 m/s are shown in Fig. 7(a) and (b) , respectively. It is easy to find that the temperature profiles on the copper are rather uniform due to the copper’s high thermal conductivity and the heat resource’s high uniformity. As the flow velocity increases, the measured temperature decreases. The temperature difference between the coolant and the front surface of the copper is 39.1 °C when the flow velocity is 4 m/s, which is 33% lower than that in the case of 1 m/s. Therefore, a higher flow velocity could enhance the cooling capacity.
Fig. 6 The schematic diagram of (a) experimental system for measuring the temperature on the surface of thin film resistor; (b) the measured temperature distribution on the thermal camera, the detailed distribution in dashed box is shown in Fig. 7.
Fig. 7 The measured temperature distribution: (a) on the surface of the thin film resistor; (b) in the flow direction (y) down the centreline (x = 9 mm) with different flow velocities.
Due to the limited thickness of the resistor and insulation, the temperature difference between the front surface of the copper and the film resistor can be neglected. Considering the fact that the heat exchange between the copper and the ambient air is negligible, all the heat from the resistor is removed by convection heat transfer through the back surface of the copper. To evaluate the cooling capacity quantitatively, one-dimensional heat conduction equation and Newton’s law of cooling can be used to calculate the CHTC
where h is the CHTC, q is the heat flux density of 5.7 × 104 W/m2, d and λ 0 are the copper’s thickness of 10 mm and the thermal conductivity of 387 W/(m·K), respectively, T and Tf are the average temperature on the resistor and the temperature of RIML at inlet. Tf is maintained at 22 °C in the experiment.The CHTC versus the flow velocity is illustrated in Table 2 . The CHTC reaches 1437 W/(m2·K) when the flow velocity is 4 m/s, which is about 1.5 times higher than that in the case when the velocity is 1 m/s. With the measured CHTC, the convective boundary condition on the two largest surfaces of the disk can be determined and the temperature distribution in the disk can be calculated, which will be given in the next section.
Table 2. List of the maximum pump power the gain module can sustain with different flow velocities
2.3 Thin disk
Thin-disk laser has been proven to be an effective configuration for realizing high output power, high efficiency, and good beam quality [19, 20 ]. As mentioned above, Nd:YLF features a much lower refractive index of 1.47 at 1047 nm. Furthermore, as a uniaxial birefringent crystal, the thermal stress-induced depolarization loss can be neglected even at a high thermal loading [21]. So the a-cut Nd:YLF crystals are chosen as gain media in our experiments. The two largest surfaces of the crystals are polished but uncoated. The dimension of each disk is 48 mm × 18 mm (height × width) and the clear aperture is 45 mm × 18 mm. The thickness is 2 mm. The crystal’s remaining four surfaces are not polished, which is helpful for limiting the amplified spontaneous emission (ASE) effect. The crystals with different doping levels from 0.36 at.% to 0.76 at.% are used to make the thermal loading in each crystal nearly the same [2], as shown in Fig. 8 . The pump absorption efficiency varies within a small range between 9% and 10% indicating that the profile of pump beam distribution along the optical axis direction (z-axis) is fairly uniform. The total absorption efficiency reaches 94%.
We estimated the permissible pump power in the crystal by assuming the maximum thermal stress to be 16 MPa (The typical stress fracture limit is 20 MPa to 50 MPa [21, 22 ]) and the maximum temperature on the surface of the crystal to be 100 °C (lower than the flash point of 121 °C of the RIML). The model in the calculation is similar to our previous work [23]. The finite element method is adopted to calculate the temperature and thermal stress in the disk. The pump power density is considered as a constant. Forced convective cooling boundary condition is imposed at the two largest surfaces of the disks with the measured CHTC and the coolant temperature of 22 °C. The four side surfaces of the disks are assumed to be adiabatic. When the thermal stress caused by the non-uniform temperature distribution is calculated, the external constraint around the disk has been ignored, since the crystal is flexibly supported. In the analysis, the fraction of absorbed pump light converted to heat is assumed as 30%. The simulation results show that the maximum permissible absorbed pump power is 5560 W when the flow velocity is 4 m/s, which is 33% higher than that in the case of 1 m/s. As shown in Table 2, it is worth noting that the maximum absorbed pump power is determined by the maximum principal stress at a high flow velocity, while it turns to the maximum temperature on the surface of the disk when the flow velocities are 1 m/s and 2 m/s. So the flash point is also worth considering when choosing the RIML. Obviously, the higher flow velocity can enhance the upper limit of pump power effectively. However, this should impose higher standard on the gain module’s sealing procedure. In our gain module, the maximum flow velocity should not exceed 6 m/s.
3. Experimental results
In this section, we will present the experimental results on direct-RIML-cooled Nd:YLF thin disk laser resonator. The flow velocity is around 5 m/s at the laser disk zone in the channels with the laminar flow pattern. The laser is operated in multimode due to the large-aperture stable cavity with large Fresnel number. Using the total 4-array-stacks with pump power (The pump power is measured after the light propagates through the coupling system) of 5202 W, the maximum CW output power of 1120 W is achieved, corresponding to an optical-optical efficiency of 21.5%, and a slope efficiency of 30.8%. The polarization extinction ratio is measured about 20 dB. The output power is recorded lasting 15 seconds, as shown in Fig. 9 . To avoid the potential risk, operation with more than 15 seconds has not been tested. The dashed line is mainly caused by the slow response of the power meter (Ophir-Spiricon, USA). The output instability is evaluated by the coefficient of variance (CV) [1] which is calculated less than 1.5% (small figure in Fig. 9). Figure 10 (rectangular point) shows the output power curves versus pump power.
The output power as a function of the pump power with different output couplers is shown in Fig. 10. By using the OC with transmittance of T = 10% and radius of curvature of the HR mirror of 1 m, the maximum output power is obtained. The lasing threshold for the 6% OC is around 1000 W, and increases to 1500 W with 10% transmittance. The slope efficiency of 35.1% for the 12% output coupler is slightly higher than that of output coupler with T = 10%. It can be expected that a higher optical-optical efficiency can be achieved with the OC of T = 12% when higher pump power is injected.
Figure 11 shows the near-field intensity distribution of the laser beam with the maximum output power. The size of the laser beam is close to that of the clear aperture both in the vertical and horizontal directions.
Four separate flow velocities are utilized for evaluating the laser performance of the direct-RIML-cooled Nd:YLF thin disk laser resonator, shown in Fig. 12 . When the flow velocity are 4m/s and 5m/s, the output powers are almost the same, and the curves show excellent linearity, indicating that the sufficient heat removal is achieved by the liquid cooling. When the flow velocity decreases to 3 m/s, the maximum output power drops to 1050 W, showing a reduction of flow velocity weakens the cooling capacity. Besides, the output power drops even further when the flow velocity reduces to 2 m/s and the output curve shows an obvious decreasing when the pump power ranges from 4387 W to 5202W. The maximum output power drops by 17.0% from 1120 W to 930 W. These can be explained by the weaker heat exchange between the disks and the coolant layers, therefore the oscillating laser suffers from serious wavefront aberration when it propagates through the gain module and thus the output power drops continuously. So the high-speed cooling which can provide excellent heat dissipation is preferred.
In order to evaluate the wavefront aberration of the gain module, a HeNe laser beam with a diameter of 100 mm is used (without laser output) to pass through the module when it is pumped at 5202 W. Firstly, we measure the stationary wavefront as the reference wavefront which takes account of distortions present within the HeNe laser beam itself and those introduced by clamping the gain media. The root mean square (RMS) of the stationary wavefront is about 0.1 μm. After subtraction of reference wavefront distribution, the distortion due to thermal effects emerges. The wavefront of the output HeNe beam with subtracted defocus and tilt is exhibited in Fig. 13 . The peak to valley (PV) and the RMS of the phase distribution are 4.0 μm and 0.6 μm when the flow velocity is 5 m/s. It is believed that the wavefront aberration should be released when there is lasing, because the fraction of absorbed pump light converted to heat should decrease. Generally, the temperature gradient, thermal deformation of disks and stress-induced change of refractive index are the major factors that determine the wavefront aberration in solid-state lasers. The third term is always much weaker than the former two terms [24] and in our experiment, the refractive index of coolant matches that of the disks, thus the second term can also be neglected. Therefore the main cause of wavefront aberration is thermo-optical effect in our case. In addition, the thermo-optical coefficient of the coolant (in Table 1 at last line) is two orders greater than that of the disks. So reducing the temperature gradient of the coolant is the key factor to weaken the wavefront distortion in direct-RIML-cooled thin disk laser resonator. A detailed analysis of the wavefront distortion of the gain module will be carried out in the later work. With the high uniformity of pump intensity distribution, flow velocity distribution and especially the weak wavefront distortion of the gain module, the design of unstable resonator which would largely improve the beam quality is feasible in the future.
Fig. 13 Measured wavefront of HeNe beam passing through the gain module (with subtracted defocus and tilt).
4. Conclusion
We present a double end-pumped direct-RIML-cooled Nd:YLF thin disk laser resonator. The gain module is designed to fulfill the heat transfer capability requirement at kilowatt-level. It contains ten pieces of disks at different doping levels and eleven channels filled with RIML flowing over the largest surfaces of the disks directly. The laser aperture is 45 mm × 18 mm. A multimode laser output with the power of 1120 W is achieved at the pump power of 5202 W, corresponding to an optical-optical efficiency of 21.5%, and a slope efficiency of 30.8%. The experimental results indicate that the direct-RIML-cooled Nd:YLF thin disk laser resonator can be efficient in thermal management and has potential for compact and efficient sources.
With the merits of highly uniform distribution of both the pump and the flowing velocity, the proposed design has a great potential to further improve the power scaling by the increase in size, number of thin disks or gain modules. Future work will also be focused on the design of the unstable cavity to improve the beam quality.
Acknowledgments
The research was supported in part by the foundation of the Key Laboratory of Science and Technology on High Energy Laser, CAEP (No. 2014HEL04), in part by the National Natural Science Foundation of China (NSFC) (No. 61575172), in part by Zhejiang Provincial Natural Science Foundation of China (No. LZ15F050001), and in part by the Fundamental Research Funds for the Central Universities.
References and links
1. R. Nie, J. She, P. Zhao, F. Li, and B. Peng, “Fully immersed liquid cooling thin-disk oscillator,” Laser Phys. Lett. 11(11), 115808 (2014). [CrossRef]
2. X. Fu, P. Li, Q. Liu, and M. Gong, “3kW liquid-cooled elastically-supported Nd:YAG multi-slab CW laser resonator,” Opt. Express 22(15), 18421–18432 (2014). [CrossRef] [PubMed]
3. J. R. Wang, J. C. Min, and Y. Z. Song, “Forced convective cooling of a high-power solid-state laser slab,” Appl. Therm. Eng. 26(5-6), 549–558 (2006). [CrossRef]
4. P. Li, X. Fu, Q. Liu, and M. Gong, “Analysis of wavefront aberration induced by turbulent flow field in liquid convection-cooled disk laser,” J. Opt. Soc. Am. B 30(8), 2161–2167 (2013). [CrossRef]
5. A. Mandl and D. E. Klimek, “Textron’s J-HPSSL 100 kW ThinZag® laser program” in Conference on Lasers and Electro-Optics (Optical Society of America, 2010), paper JThH2. [CrossRef]
6. X. Fu, Q. Liu, P. Li, L. Huang, and M. Gong, “Numerical simulation of 30-kW class liquid-cooled Nd:YAG multi-slab resonator,” Opt. Express 23(14), 18458–18470 (2015). [CrossRef] [PubMed]
7. X. Fu, Q. Liu, P. Li, and M. Gong, “Direct-liquid-cooled Nd:YAG thin disk laser oscillator,” Appl. Phys. B 111(3), 517–521 (2013). [CrossRef]
8. P. Li, Q. Liu, X. Fu, and M. Gong, “Large-aperture end-pumped Nd:YAG thin-disk laser directly cooled by liquid,” Chin. Opt. Lett. 11(4), 041408 (2013). [CrossRef]
9. P. Moonen, B. Blocken, S. Roels, and J. Carmeliet, “Numerical modeling of the flow conditions in a closed-circuit low-speed wind tunnel,” J. Wind Eng. Ind. Aerodyn. 94(10), 699–723 (2006). [CrossRef]
10. R. D. Mehta and P. Bradshaw, “Design rules for small low-speed wind tunnels,” Aeronaut. J. 83(827), 443–449 (1979).
11. G. Diana, S. De Ponte, M. Falco, and A. Zasso, “A new large wind tunnel for civil-environmental and aeronautical applications,” J. Wind Eng. Ind. Aerodyn. 74–76, 553–565 (1998). [CrossRef]
12. P. M. Ligrani and R. D. Niver, “Flow visualization of Dean vortices in a curved channel with 40 to 1 aspect ratio,” Phys. Fluids 31(12), 3605–3617 (1988). [CrossRef]
13. L. S. Han, “Hydrodynamic entrance lengths for incompressible laminar flow in rectangular ducts,” J. Appl. Mech. 27(3), 403–409 (1960). [CrossRef]
14. P. Moonen, B. Blocken, and J. Carmeliet, “Indicators for the evaluation of wind tunnel test section flow quality and application to a numerical closed-circuit wind tunnel,” J. Wind Eng. Ind. Aerodyn. 95(9–11), 1289–1314 (2007). [CrossRef]
15. R. D. Mehta, “The aerodynamic design of blower tunnels with wide-angle diffusers,” Prog. Aerosp. Sci. 18, 59–120 (1979). [CrossRef]
16. C. D. Meinhart, S. T. Wereley, and J. G. Santiago, “PIV measurements of a microchannel flow,” Exp. Fluids 27(5), 414–419 (1999). [CrossRef]
17. R. J. Adrian, “Twenty years of particle image velocimetry,” Exp. Fluids 39(2), 159–169 (2005). [CrossRef]
18. R. D. Keane and R. J. Adrian, “Theory of cross-correlation analysis of PIV images,” Appl. Sci. Res. 49(3), 191–215 (1992). [CrossRef]
19. C. Stewen, K. Contag, M. Larionov, A. Giesen, and H. Hügel, “A 1-kW CW thin disc laser,” IEEE J. Sel. Top. Quantum Electron. 6(4), 650–657 (2000). [CrossRef]
20. Y. H. Peng, Y. X. Lim, J. Cheng, Y. Guo, Y. Y. Cheah, and K. S. Lai, “Near fundamental mode 1.1 kW Yb:YAG thin-disk laser,” Opt. Lett. 38(10), 1709–1711 (2013). [CrossRef] [PubMed]
21. W. Koechner, Solid-State Lasers Engineering (Springer, 2006).
22. X. Peng, L. Xu, and A. Asundi, “High-power efficient continuous-wave TEM00 intracavity frequency-doubled diode-pumped Nd:YLF laser,” Appl. Opt. 44(5), 800–807 (2005). [CrossRef] [PubMed]
23. Z. Ye, Z. Cai, B. Tu, X. Wang, J. Shang, Y. Yu, K. Wang, Q. Gao, C. Tang, and C. Liu, “Numerical approach to temperature and thermal stress in direct-liquid-cooled Nd: YAG thin disk laser medium,” Proc. SPIE 9255, 92550T (2014).
24. P. Ferrara, M. Ciofini, L. Esposito, J. Hostaša, L. Labate, A. Lapucci, A. Pirri, G. Toci, M. Vannini, and L. A. Gizzi, “3-D numerical simulation of Yb:YAG active slabs with longitudinal doping gradient for thermal load effects assessment,” Opt. Express 22(5), 5375–5386 (2014). [CrossRef] [PubMed]
