Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Design, Fabrication and Characterization of an Adaptive Retroreflector (AR)

Design, Fabrication and Characterization of an Adaptive Retroreflector (AR) hv photonics Communication Design, Fabrication and Characterization of an Adaptive Retroreflector (AR) 1 , 1 1 1 2 Freddie Santiago * , Carlos O. Font , Sergio R. Restaino , Syed N. Qadri and Brett E. Bagwell Naval Research Laboratory, Washington, DC 20375, USA; carlos.font@nrl.navy.mil (C.O.F.); sergio.restaino@nrl.navy.mil (S.R.R.); noor.qadri@nrl.navy.mil (S.N.Q.) Sandia National Laboratories, Albuquerque, NM 94551, USA; bbagwel@sandia.gov * Correspondence: freddie.santiago@nrl.navy.mil Abstract: Recent work at the U.S. Naval Research Laboratory studied atmospheric turbulence on dynamic links with the goal of developing an optical anemometer and turbulence characterization system for unmanned aerial vehicle (UAV) applications. Providing information on the degree of atmospheric turbulence, as well as wind information and scintillation, in a low size, weight and power (SWaP) system is key for the design of a system that is also capable of adapting quickly to changes in atmospheric conditions. The envisioned system consists of a bi-static dynamic link between a transmitter (Tx) and a receiver (Rx), relying on a small UAV. In a dynamic link, the propagation distance between the Tx/Rx changes rapidly. Due to SWaP constraints, a monostatic system is challenging for such configurations, so we explored a system in which the Tx/Rx is co-located on a mobile platform (UAV), which has a mounted retroreflector. Beam divergence control is key in such a system, both for finding the UAV (increased beam divergence at the Tx) and for signal optimization at the Rx. This led us to the concept of using adaptive/active elements to control the divergence at the Tx but also to the implementation of an adaptive/active retroreflector in which the return beam divergence can be controlled in order to optimize the signal at the Rx. This paper presents the design, fabrication and characterization of a low SWaP adaptive retroreflector. Keywords: adaptive retroreflector; tunable lens; adaptive lens; polymer optics; divergence control; Citation: Santiago, F.; Font, C.O.; fluidic lens; tunable optics Restaino, S.R.; Qadri, S.N.; Bagwell, B.E. Design, Fabrication and Characterization of an Adaptive Retroreflector (AR). Photonics 2022, 9, 124. https://doi.org/10.3390/ 1. Introduction photonics9030124 Retroreflectors are passive devices that return the incident signal through the same propagation path. For our intended application on UAVs, a retroreflector is ideal, due to its Received: 25 January 2022 Accepted: 15 February 2022 size and zero power consumption. The fact that this is a dynamic link (with changing dis- Published: 22 February 2022 tance between the transmitter and the point where the signal is reflected occurring quickly or discretely) means that signal degradation is expected due to atmospheric turbulence Publisher’s Note: MDPI stays neutral induced effects, but also due to the general nature of a propagating beam. In order to with regard to jurisdictional claims in ameliorate these effects, we relied on low order adaptive optics correction, in this case, published maps and institutional affil- focus control. Due to the constraints in SWaP, we have designed and fabricated an adaptive iations. retroreflector which allows us to change the divergence of the beam in order to optimize the link, achieving higher link performance or longer distances than can normally be obtained with a passive system. This device enables the control of the divergence, which can be Copyright: © 2022 by the authors. used to optimize the return signal in a monostatic configuration or to increase the return Licensee MDPI, Basel, Switzerland. footprint of the beam in a bi-static, dynamic, or reconfigurable link (moving link), this latter This article is an open access article case being the motivation for the following types of devices [1,2]. distributed under the terms and Adaptive optical devices (also known as active or tunable devices) are devices that can conditions of the Creative Commons adjust their surface/curvature (such as deformable mirrors, fluidic lenses, elastic/elastomeric Attribution (CC BY) license (https:// solids) or modify their index of refractions (such as liquid crystals) in order to change the creativecommons.org/licenses/by/ optical properties of the element, such as its focal length. This leads to the design and 4.0/). Photonics 2022, 9, 124. https://doi.org/10.3390/photonics9030124 https://www.mdpi.com/journal/photonics Photonics 2022, 8, x FOR PEER REVIEW 2 of 10 Photonics 2022, 9, 124 2 of 10 design and fabrication of an adaptive retroreflector—the one described in this paper is based on fluidic optical elements, for which NRL has extensive expertise [3–5]. fabrication of an adaptive retroreflector—the one described in this paper is based on fluidic optical elements, for which NRL has extensive expertise [3–5]. 2. AR Design and Configurations Our adaptive retroreflector (AR) can be used with a corner cube retroreflector, solid 2. AR Design and Configurations or hollow, and consists of an optical fluid, encapsulated by an elastomeric membrane that Our adaptive retroreflector (AR) can be used with a corner cube retroreflector, solid can be deformed via an actuator—in this case, the same actuator we use for our adaptive or hollow, and consists of an optical fluid, encapsulated by an elastomeric membrane that polymer lenses. This actuator and its electronics have been designed for tactical applica- can be deformed via an actuator—in this case, the same actuator we use for our adaptive tions in which SWaP is key, making this ideal for UAV applications. polymer lenses. This actuator and its electronics have been designed for tactical applications Depending on the application, both the membrane and fluid can be replaced with an in which SWaP is key, making this ideal for UAV applications. elastomeric optical polymer (which can be made from the same material as the optical Depending on the application, both the membrane and fluid can be replaced with membrane) that can be deformed mechanically to make the adaptive retroreflector. This an elastomeric optical polymer (which can be made from the same material as the optical device enables contro-l of the divergence. membrane) that can be deformed mechanically to make the adaptive retroreflector. This The device can be fabricated in two ways: (1) using an elastomeric optical polymer, device enables contro-l of the divergence. or (2) a fluidic adaptive/tunable device with a hollow or solid retroreflector. While manu- The device can be fabricated in two ways: (1) using an elastomeric optical polymer, or facturing errors could change the operation of a retroreflector with an elastic polymer in (2) a fluidic adaptive/tunable device with a hollow or solid retroreflector. While manufac- front, by slightly changing the direction of return light, such errors can be easily quantified turing errors could change the operation of a retroreflector with an elastic polymer in front, and corrected, for example, by monitoring the overall optical performance with an inter- by slightly changing the direction of return light, such errors can be easily quantified and ferometer. corrected, for example, by monitoring the overall optical performance with an interferometer. For the elastomeric optical polymer option, the elastic polymer is molded to a desired For the elastomeric optical polymer option, the elastic polymer is molded to a desired initial initialshape shapeand andthe the change change on on the the polymer polymesurface r surfaccan e can be baf e fected affecteby d bmeans y mean of s o applying f applyin ag pr a essur presse/compr ure/comp ession ression to to the the polymer polyme . rAn . An elastomeric elastomeric optical optical polymer polymeris isa apolymer polymerthat that has high transmission at the user-desired operational wavelength and has elastic properties has high transmission at the user-desired operational wavelength and has elastic proper- which allow the solid substrate to be deformed. A second alternative to deform the polymer ties which allow the solid substrate to be deformed. A second alternative to deform the can be achieved by the use of dielectric elastomer actuation in which a voltage is applied to polymer can be achieved by the use of dielectric elastomer actuation in which a voltage is a pliable electrode and the polymer is deformed, creating the change on its surface. Figure 1 applied to a pliable electrode and the polymer is deformed, creating the change on its shows the configuration of the elastomer optical polymer, as well as three operational states surface. Figure 1 shows the configuration of the elastomer optical polymer, as well as three of the adaptive retroreflector. operational states of the adaptive retroreflector. Figure Figure1. 1.Schematic Schematico o f f an an adaptive adaptive retr reor tro eflector reflecto u rsing usina gsolid a soli elastomeric d elastomer polymer: ic polyme (a r) : ( flat a) f(lb at ) ( convex b) con- vex and (c) concave. and (c) concave. For a fluidic adaptive/tunable option, an elastomeric membrane encapsulates an optical For a fluidic adaptive/tunable option, an elastomeric membrane encapsulates an op- fluid which is mounted on the front of a hollow or solid retroreflector. The elastomeric tical fluid which is mounted on the front of a hollow or solid retroreflector. The elasto- membrane needs to have similar optical and mechanical properties to those described above meric membrane needs to have similar optical and mechanical properties to those de- for the solid option. The optical fluid, needs to be optically and chemically compatible scribed above for the solid option. The optical fluid, needs to be optically and chemically with the membrane and needs to have high transmission at the operational wavelength. compatible with the membrane and needs to have high transmission at the operational Polydimethylsiloxane is a common polymer that can be used for the membrane as well as wavelength. Polydimethylsiloxane is a common polymer that can be used for the mem- for the elastomeric solid option. For the optical fluids, there are numerous oils, polymers brane as well as for the elastomeric solid option. For the optical fluids, there are numerous and resins that have been studied (for example, water, glycerol, etc.) [3]. The actuation of oils, polymers and resins that have been studied (for example, water, glycerol, etc.) [3]. this system can be achieved by compressing/decompressing the flexible membrane, which The actuation of this system can be achieved by compressing/decompressing the flexible creates a change on its surface. This occurs by moving a cylinder along the optical axis of the membrane, which creates a change on its surface. This occurs by moving a cylinder along system, thus compressing the circumference of the flexible membrane. Besides the mechanical the optical axis of the system, thus compressing the circumference of the flexible mem- action, magnetic actuation or use of a compliant electrode (dielectric elastomer) can achieve brane. Besides the mechanical action, magnetic actuation or use of a compliant electrode actuation of the membrane. There are other actuation techniques that are situatable and could be implemented as well, such as those used by commercially available fluidic lenses, Photonics 2022, 8, x FOR PEER REVIEW 3 of 10 Photonics 2022, 9, 124 3 of 10 (dielectric elastomer) can achieve actuation of the membrane. There are other actuation techniques that are situatable and could be implemented as well, such as those used by commercially available fluidic lenses, for example, Varioptics, Optotune and Holochip [6– for example, Varioptics, Optotune and Holochip [6–12]. Figure 2, shows a conceptual sketch 12]. Figure 2, shows a conceptual sketch of the adaptive retroreflector based on the flexible of the adaptive retroreflector based on the flexible membrane/fluidic concept. membrane/fluidic concept. Figure Figure2. 2.Schematic Schematicof ofan anadaptive adaptiver etr retor roeflector reflector using using an an optical opticaliquid/fluidic: l liquid/fluidic:( ( aa ))flat flat( b (b ) )convex convex and (c) concave. and (c) concave. The latter configuration was selected for this paper. The device changes the divergence of The latter configuration was selected for this paper. The device changes the diver- the returned beam but can also work as a regular passive retroreflector if the system requires. gence of the returned beam but can also work as a regular passive retroreflector if the system requires. 3. Fabrication, Characterization and Results 3. Fa Her bric e, awe tion pr , C esent harathe cteroptical ization design and Rof esu the ltsAR as fabricated and in the configuration we used for testing. We describe the optical setups used for testing and comparing the device Here, we present the optical design of the AR as fabricated and in the configuration to a passive retroreflector. Data for the repeatability and arbitrary radius of curvature we used for testing. We describe the optical setups used for testing and comparing the measurements was acquired using a Zygo Verifire HD optical interferometer. Furthermore, device to a passive retroreflector. Data for the repeatability and arbitrary radius of curva- we show images taken comparing a passive retroreflector in comparison with AR, both ture measurements was acquired using a Zygo Verifire HD optical interferometer. Fur- actuated and flat. thermore, we show images taken comparing a passive retroreflector in comparison with AR, both actuated and flat. 3.1. AR Optical Design We used OpticStudio nonsequential tools to model the corner cube retroreflector and 3.1. AR Optical Design adaptive components and, for visualization purposes, a beam splitter cube was added, as We used OpticStudio nonsequential tools to model the corner cube retroreflector and shown in Figure 3. This was also the configuration chosen for the test. Note that we did adaptive components and, for visualization purposes, a beam splitter cube was added, as not model the thickness of the membrane, since the effects of the membrane are negligible shown in Figure 3. This was also the configuration chosen for the test. Note that we did in OpticStudio for this type of application. The model was performed using the volume not model the thickness of the membrane, since the effects of the membrane are negligible of the fluid, the fluid acting as a lens which changes its radius of curvature and center in OpticStudio for this type of application. The model was performed using the volume thickness. The figure shows a collimated beam, incident on a beam splitter cube which of the fluid, the fluid acting as a lens which changes its radius of curvature and center reflects part of the incident light and transmits a portion, which then impinges on the thickness. The figure shows a collimated beam, incident on a beam splitter cube which adaptive retroreflector. Light is reflected back from the retroreflector and reflected again reflects part of the incident light and transmits a portion, which then impinges on the from the beam splitter cube and incident on the detector. In field operations, the beam adaptive retroreflector. Light is reflected back from the retroreflector and reflected again splitter can be used to monitor the incoming beam and direct the adaptive retroreflector from the beam splitter cube and incident on the detector. In field operations, the beam in order to control the divergence. It can also be used without the beam splitter cube, splitter can be used to monitor the incoming beam and direct the adaptive retroreflector such that the beam can be monitored at the receiver side and the system optimized in in order to control the divergence. It can also be used without the beam splitter cube, such a power-in-the-bucket (PIB) configuration using well-known algorithms (e.g., stochastic that the beam can be monitored at the receiver side and the system optimized in a power- parallel-gradient-descent) [13,14]; the AR is then instructed by this information. in-the-bucket (PIB) configuration using well-known algorithms (e.g., stochastic parallel- 3.2. Fabrication of the AR gradient-descent) [13,14]; the AR is then instructed by this information. For this particular design we used a 12.7 mm corner cube retroreflector from Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 (at = 589 nm). The first step consisted in making the PDMS membrane which was then bonded to the glass support structure. The fluid was added to the membrane/glass structure and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication Photonics 2022, 8, x FOR PEER REVIEW 4 of 10 Photonics 2022, 9, 124 4 of 10 Figure 3. Optical design setup used for testing including the collimated source incident on the beam procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent as pictures of the assembled AR (in its actuator). The actuator was custom-made for our beam. adaptive lenses, and we were able to modify one to accommodate the AR. The actuator consists of a modified motor in a custom housing, with a maximum clear aperture of 3.2. Fabrication of the AR 19.5 mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power For this particular design we used a 12.7 mm corner cube retroreflector from consumption ~15 W, idle power consumption of ~1–500 W and temperature monitoring Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the of 0.01 C. The electronics can control two actuators at the same time and can run off three PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 Photonics 2022, 8, x FOR PEER REVIECR-123 W batteries (two batteries for a single actuator) with an average number of actuations 4 of 10 (at λ = 589 nm). The first step consisted in making the PDMS membrane which was then of about 6000 per set of batteries. bonded to the glass support structure. The fluid was added to the membrane/glass struc- ture and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well as pictures of the assembled AR (in its actuator). The actuator was custom-made for our adaptive lenses, and we were able to modify one to accommodate the AR. The actuator consists of a modified motor in a custom housing, with a maximum clear aperture of 19.5 mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power con- sumption ~15 W, idle power consumption of ~1–500 µW and temperature monitoring of 0.01 °C. The electronics can control two actuators at the same time and can run off three Figure Figure 3. 3. O Op pt ti ic ca al l d de es si ig gn n s se et tu up p u us se ed d f fo or r t te es st ti in ng g i in nc cl lu ud di in ng g t th he e c co ol ll li im ma at te ed d s so ou ur rc ce e iin nc ciid de en ntt o on n tth he e b be ea am m CR-123 batteries (two batteries for a single actuator) with an average number of actuations splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent beam. of about 6000 per set of batteries. beam. 3.2. Fabrication of the AR For this particular design we used a 12.7 mm corner cube retroreflector from Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 (at λ = 589 nm). The first step consisted in making the PDMS membrane which was then bonded to the glass support structure. The fluid was added to the membrane/glass struc- ture and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well as Figure 4. Schematic representation of the fabrication and assembly process of the AR. Figure 4. Schematic representation of the fabrication and assembly process of the AR. pictures of the assembled AR (in its actuator). The actuator was custom-made for our adaptive lenses, and we were able to modify one to accommodate the AR. The actuator 3.3. Laboratory Optical Setup 3.3. Laboratory Optical Setup consists of a modified motor in a custom housing, with a maximum clear aperture of 19.5 The AR was tested in various ways. Firstly, in the same way that we measured The AR was tested in various ways. Firstly, in the same way that we measured the mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power con- the radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we meas- sumption ~15 W, idle power consumption of ~1–500 µW and temperature monitoring of measured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a ured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a negative 0.01 °C. The electronics can control two actuators at the same time and can run off three negative (concave surface) of 185 mm. The second setup was to compare the performance CR-123 batteries (two batteries for a single actuator) with an average number of actuations and proof of concept of the AR in comparison with a passive retroreflector, as shown in of about 6000 per set of batteries. Figure 5. We used the HeNe 633 nm source of the Zygo interferometer and a 1550 nm was co- aligned for further testing. Data from the 1550 nm was not included but performance of the active surface component at this wavelength has been demonstrated in a previous report [5]. We were able to use the beam collimated or with the addition of a known divergence that could be removed with the AR and compared with the passive retroreflector. The setup with the beam splitter cube allowed us to look at the return beam with the interferometer and, on the other arm, to look at the output with a camera, photodetector or power meter, while we were able to use beam blocks to look at each retroreflector individually or additionally, enabling viewing of the interference fringes formed by the two. This facilitated alignment, but also monitoring of the difference when the AR is actuated. This same setup was also Figure 4. Schematic representation of the fabrication and assembly process of the AR. 3.3. Laboratory Optical Setup The AR was tested in various ways. Firstly, in the same way that we measured the radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we meas- ured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a negative Photonics 2022, 8, x FOR PEER REVIEW 5 of 10 (concave surface) of −185 mm. The second setup was to compare the performance and proof of concept of the AR in comparison with a passive retroreflector, as shown in Figure 5. We used the HeNe 633 nm source of the Zygo interferometer and a 1550 nm was co- aligned for further testing. Data from the 1550 nm was not included but performance of the active surface component at this wavelength has been demonstrated in a previous report [5]. We were able to use the beam collimated or with the addition of a known di- vergence that could be removed with the AR and compared with the passive retroreflec- tor. The setup with the beam splitter cube allowed us to look at the return beam with the interferometer and, on the other arm, to look at the output with a camera, photodetector Photonics 2022, 9, 124 5 of 10 or power meter, while we were able to use beam blocks to look at each retroreflector in- dividually or additionally, enabling viewing of the interference fringes formed by the two. This facilitated alignment, but also monitoring of the difference when the AR is actuated. uT se hd ist s oap m ee rf s oe rt m up fi r w st as o r a d ls eo r m usee ad s u to re p m ee rn fo ts rm of ftih rs et ro er p d ee ar t a m bie la its yuo re f m the en A ts R ob f y th ae c tru ea pte in at gat b h il eitA yR of fr to hm e A aR fl b at y sa tc attu ea tt o in agc t o h n ev A ex Ro fr ro cm on a c a fv laet s st ta at te e a to n d a c bo an cv ketx o o arf c la otns cta av te e, sw tah te il e an m de b aa su ck ri t n og at h fle at su stra ft ae c,e w fo hrim le w mie th astu hr ein in gt e th rfe e r so um rfe atcee r .fo Arn m i m wp ito h r tta h n et in no te te r:fe te ro m m pe et re ar t.u A renw im asp m or otn ain to t r n eo dte in : tte h m e - room, but temperature compensation of the AR was not used—the room environment was perature was monitored in the room, but temperature compensation of the AR was not st u asb eld e— an td het h ro uo s m co e m np ve ir no sn at m io en nt w w as as n o sttarb elq eu ain re d d t .hus compensation was not required. Figure 5. (Left) Layout of the optical testing setup. (Right) Picture of optical devices used Figure 5. (Left) Layout of the optical testing setup. (Right) Picture of optical devices used for meas- for measurements. urements. 3.4. Results/Discussion 3.4. Results/Discussion 3.4.1. ROC and Surface Measurements 3.4.1. ROC and Surface Measurements The first test consisted of measurements of the ROC, positive and negative, in order to The first test consisted of measurements of the ROC, positive and negative, in order evaluate the performance of the device. Figure 6 shows a set of measurements, including a to evaluate the performance of the device. Figure 6 shows a set of measurements, includ- (left) measurement for a positive ROC of 255 mm and a (right) measurement for a negative ing a (left) measurement for a positive ROC of 255 mm and a (right) measurement for a ROC of 184 mm. The top row shows the 3D surface profile, and the bottom row indicates the negative ROC of 184 mm. The top row shows the 3D surface profile, and the bottom row 2D profile. The black circles in the figure are software masks used to remove unnecessary indicates the 2D profile. The black circles in the figure are software masks used to remove back reflections created by dust particles in the reference sphere. Within the respective unnecessary back reflections created by dust particles in the reference sphere. Within the figures, the left column (A or C) is the raw measurement and the right side (B or D) is with respective figures, the left column (A or C) is the raw measurement and the right side (B the dominant aberrations removed. As mentioned before, fabrication was not optimized or D) is with the dominant aberrations removed. As mentioned before, fabrication was for the surface figure, but what can be seen is the typical dominant aberration of coma not optimized for the surface figure, but what can be seen is the typical dominant aberra- and astigmatism, which are characteristic for this type of fluidic structures. Coma is due tion of coma and astigmatism, which are characteristic for this type of fluidic structures. to gravity and astigmatism is due to fabrication or assembly procedures. For the positive Coma is due to gravity and astigmatism is due to fabrication or assembly procedures. For ROC case demonstrated below, coma is the dominant aberration. On the negative ROC, the positive ROC case demonstrated below, coma is the dominant aberration. On the neg- there is a combination of coma and astigmatism, because measurements were taken close to ative ROC, there is a combination of coma and astigmatism, because measurements were the negative resting ROC (fabricated ROC) of the membrane for the fabricated AR device. taken close to the negative resting ROC (fabricated ROC) of the membrane for the fabri- The fabricated aberrations were more noticeable closer to the resting ROC because, for cated AR device. The fabricated aberrations were more noticeable closer to the resting this type of actuation mechanism, this is the point of contact where boundary conditions ROC because, for this type of actuation mechanism, this is the point of contact where are established between the membrane and actuation surface for the clear aperture. At boundary conditions are established between the membrane and actuation surface for the this point, the amplitude of any existing aberrations can be enhanced. Another aberration that cleacan r apbe ertu noticed re. At tis his tr p efoil ointon , thboth e amp ROCs—this litude of anwas y exipur stinely g ab due errato tiothe ns c assembly an be enhin anthe ced. actuator Anothe.rW ae be developed rration thapr t c ocedur an be es nofor ticefabrication d is trefoil and on b assembly oth ROCs that —th rieduce s was the purdominant ely due to aberrations which are implemented when building adaptive polymer lenses, with the caveat the assembly in the actuator. We developed procedures for fabrication and assembly that that redu we ce can the d minimize ominant a coma berrat based ions w on hicthe h ar application, e implement but ed w do henot n bu completely ilding adap eliminate tive poly- it. mThe er lepr nsocedur es, with e to the eliminate caveat thcoma at we during can min fabrication imize coma is b extr ased emel on y thcomplex, e applicatcostly ion, bu and t do time consuming if performed at the active surface. There are other ways to minimize it, including using a corrective element along the optical path of the system or close to the active surface, and this is a typical configuration used in commercial adaptive/tunable lenses [11]. Photonics 2022, 8, x FOR PEER REVIEW 6 of 10 not completely eliminate it. The procedure to eliminate coma during fabrication is ex- tremely complex, costly and time consuming if performed at the active surface. There are other ways to minimize it, including using a corrective element along the optical path of Photonics 2022, 9, 124 6 of 10 the system or close to the active surface, and this is a typical configuration used in com- mercial adaptive/tunable lenses [11]. Figure 6. (Left) Positive ROC measurement (A,B columns) 2D and 3D profiles with dominant aber- Figure 6. (Left) Positive ROC measurement (A,B columns) 2D and 3D profiles with dominant ration removed, in this instance being coma (B column). (Right) Negative ROC measurement (C,D aberration removed, in this instance being coma (B column). (Right) Negative ROC measurement columns) 2D and 3D profiles with dominant aberration removed, in this instance, coma and astig- (C,D columns) 2D and 3D profiles with dominant aberration removed, in this instance, coma and matism (D column). astigmatism (D column). Using the data from Figure 6, a Zernike fit was performed using the Mx software Using the data from Figure 6, a Zernike fit was performed using the Mx software tools tools from the interferometer and coefficients of the fit for both the positive and negative from the interferometer and coefficients of the fit for both the positive and negative ROCs ROCs are shown in Table 1. are shown in Table 1. Table 1. Results from the Zernike fit coefficients obtained from the data in Figure 6, for positive ROC Table 1. Results from the Zernike fit coefficients obtained from the data in Figure 6, for positive ROC (top) and negative ROC (bottom). (top) and negative ROC (bottom). ROC = 225 mm ROC = 225 mm Zernike Fit Zernike Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) ZFR 3 0.018 2 0 1 + 22 ZFR 3 0.018 2 0 −1 + 2ρ2 ZFR 4 0.085 2 2 2cos(2) ZFR 4 −0.085 2 2 ρ2cos(2θ) ZZFR FR 5 5 −0. 40 0.407 7 2 2 −2 2 ρ2sin (2sin(2 2θ) ) ZFR 6 −0.122 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.122 3 1 (2 + 33)cos() ZFR 7 2.191 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 2.191 3 1 (2 + 33)sin() ZFR 8 −0.047 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.047 4 0 1 62 + 64 ROC = −184 mm ROC = 184 mm Zernike ZernikeFit Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) Z ZFR FR 3 3 −0 .00.049 49 2 2 0 0 −1 + 2ρ 12 + 22 ZFR 4 −0.211 2 2 ρ2cos(2θ) ZFR 4 0.211 2 2 2cos(2) ZFR 5 −1.823 2 −2 ρ2sin(2θ) ZFR 5 1.823 2 2 2sin(2) ZFR 6 −0.108 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.108 3 1 (2 + 33)cos() ZFR 7 −2.115 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 2.115 3 1 (2 + 33)sin() ZFR 8 −0.372 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.372 4 0 1 62 + 64 Figure 7, shows data taken for the AR at the same ROCs mentioned above but in a perpendicular configuration in order to eliminate the effects of coma due to gravity. Note, that for the data no terms have been removed. Astigmatism and trefoil were noticeable but the large magnitude due to coma was absent. Photonics 2022, 8, x FOR PEER REVIEW 7 of 10 Figure 7, shows data taken for the AR at the same ROCs mentioned above but in a perpendicular configuration in order to eliminate the effects of coma due to gravity. Note, Photonics 2022, 9, 124 7 of 10 that for the data no terms have been removed. Astigmatism and trefoil were noticeable but the large magnitude due to coma was absent. Figure 7. (A) Positive and (B) negative ROC, 2D and 3D surface representation for the perpendicular Figure 7. (A) Positive and (B) negative ROC, 2D and 3D surface representation for the perpendicular setup. Right side shows a picture of the setup. setup. Right side shows a picture of the setup. The same procedure was performed on the results from Figure 7 and Table 2 shows The same procedure was performed on the results from Figure 7 and Table 2 shows the Zernike fit coefficients for the perpendicular measurements. the Zernike fit coefficients for the perpendicular measurements. Table 2. Results from the Zernike fit coefficients obtained from the data in Figure 7, for positive Table 2. Results from the Zernike fit coefficients obtained from the data in Figure 7, for positive ROC ROC (top) and negative ROC (bottom). (top) and negative ROC (bottom). ROC = 225 mm Perpendicular ROC = 225 mm Perpendicular Zernike ZernikeFit Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) Z ZFR FR 3 3 0.0.038 038 2 2 0 0 −1 + 2 ρ1 2+ 22 ZFR 4 0.137 2 2 ρ2cos(2θ) ZFR 4 0.137 2 2 2cos(2) ZFR 5 −0.390 2 −2 ρ2sin(2θ) ZFR 5 0.390 2 2 2sin(2) ZFR 6 0.065 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.065 3 1 (2 + 33)cos() ZFR 7 −0.071 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 0.071 3 1 (2 + 33)sin() ZFR 8 −0.099 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.099 4 0 1 62 + 64 ROC = −184 mm Perpendicular ROC = 184 mm Perpendicular Zernike Fit Zernike Fit Coeff Value (λ) n m Representation Coeff Value () n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 Z ZFR FR 1 1 0.0.000 000 1 1 1 1 ρcos(θ cos( ) ) ZFR 2 0.000 1 −1 ρsin(θ) ZFR 2 0.000 1 1 sin() ZFR 3 −0.202 2 0 −1 + 2ρ2 ZFR 3 0.202 2 0 1 + 22 ZFR 4 −0.052 2 2 ρ2cos(2θ) ZFR 4 0.052 2 2 2cos(2) ZFR 5 −1.248 2 −2 ρ2sin(2θ) ZFR 5 1.248 2 2 2sin(2) ZFR 6 0.126 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.126 3 1 (2 + 33)cos() ZFR 7 −0.112 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 0.112 3 1 (2 + 33)sin() ZFR 8 −0.224 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.224 4 0 1 62 + 64 3.4.2. Repeatability Measurements 3.4.2. Repeatability Measurements Repeatability measurements were taken using the setup in Figure 5. The data collec- tion consisted in changing the actuation state by a known encoder count from a flat state Repeatability measurements were taken using the setup in Figure 5. The data collection cto on a s ic st o en dvi en x/c ch oa nn cg aiv n eg stth ae tea , c w tu ha il te io r n ec so ta rt d ein bg y ta hk e n eo nw co n de en r c p oo d se it rio co nu a n st w fre olm l aa s t fh la et d sta atta e ftro om a convex/concave state, while recording the encoder position as well as the data from the inferferometer. The encoder data is in the form of a set of three numbers: the set position by user (state of the lens), the temperature compensate position (once thermal compensation is activated) and the measured position. This last position, or the difference from the set Photonics 2022, 8, x FOR PEER REVIEW 8 of 10 Photonics 2022, 8, x FOR PEER REVIEW 8 of 10 the inferferometer. The encoder data is in the form of a set of three numbers: the set posi- the inferferometer. The encoder data is in the form of a set of three numbers: the set posi- Photonics 2022, 9, 124 8 of 10 tion by user (state of the lens), the temperature compensate position (once thermal com- tion by user (state of the lens), the temperature compensate position (once thermal com- pensation is activated) and the measured position. This last position, or the difference pensation is activated) and the measured position. This last position, or the difference from the set position, was recorded. Readings from the interferometer PV(λ) (peak to val- from the set position, was recorded. Readings from the interferometer PV(λ) (peak to val- position, was recorded. Readings from the interferometer PV() (peak to valley) and power() ley) and power(λ) were recorded as well. Data was taken for a delta for encoder counts of ley) and power(λ) were recorded as well. Data was taken for a delta for encoder counts of were recorded as well. Data was taken for a delta for encoder counts of 150 and 300 from 150 and 300 from flat, on both positive (convex or higher encoder counts), and negative 150 and 300 from flat, on both positive (convex or higher encoder counts), and negative flat, on both positive (convex or higher encoder counts), and negative (concave or lower (concave or lower encoder counts) direction. Figure 8 shows a sequence of consecutive (concave or lower encoder counts) direction. Figure 8 shows a sequence of consecutive encoder counts) direction. Figure 8 shows a sequence of consecutive measurements from flat measurements from flat to positive and Figure 9 shows measurements from flat to nega- measurements from flat to positive and Figure 9 shows measurements from flat to nega- to positive and Figure 9 shows measurements from flat to negative for a delta of 150 encoder tive for a delta of 150 encoder counts. Note, the hexagonal pattern was a result of the facets tive for a delta of 150 encoder counts. Note, the hexagonal pattern was a result of the facets counts. Note, the hexagonal pattern was a result of the facets of the corner cube. This was of the corner cube. This was noticeable in this configuration based on the testing setup of the corner cube. This was noticeable in this configuration based on the testing setup noticeable in this configuration based on the testing setup with the interferometer using a with the interferometer using a transmission flat. For the ROCs the measurements dif- with the interferometer using a transmission flat. For the ROCs the measurements dif- transmission flat. For the ROCs the measurements differed, since we were using a reference fered, since we were using a reference sphere and the spherical wavefront matched the fered, since we were using a reference sphere and the spherical wavefront matched the sphere and the spherical wavefront matched the deformed membrane, not the retroreflector. deformed membrane, not the retroreflector. deformed membrane, not the retroreflector. Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. In Figure 10, data is presented in graphical (with error bars based on the standard In Figure 10, data is presented in graphical (with error bars based on the standard In Figure 10, data is presented in graphical (with error bars based on the standard devi- deviation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 deviation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 ation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 points points for delta 300, and all cases starting from the same initial flat position. The average points for delta 300, and all cases starting from the same initial flat position. The average for delta 300, and all cases starting from the same initial flat position. The average and and standard deviation for the encoder position and peak-to-valley for the cases are and standard deviation for the encoder position and peak-to-valley for the cases are standard deviation for the encoder position and peak-to-valley for the cases are shown in shown in the table. shown in the table. the table. Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was placed at a distance of approximately 1500 mm and the response from a collimated beam recorded and the AR was actuated in order to focus the beam on the screen. Photonics 2022, 8, x FOR PEER REVIEW 9 of 10 Photonics 2022, 8, x FOR PEER REVIEW 9 of 10 Photonics 2022, 9, 124 9 of 10 Flat PV(λ) Convex PV(λ) Convex PV(λ) Concave PV(λ) Concave PV(λ) (Δ=150) (Δ=300) (Δ=150) (Δ=300) Ave. encod. Pos 27,490 0.564 27,640 2.261 27,790 4.248 27,340 2.408 27,190 4.256 Stdev 0.16 0.02 0.00 0.05 0.00 0.02 0.00 0.02 0.00 0.04 Flat PV(λ) Convex PV(λ) Convex PV(λ) Concave PV(λ) Concave PV(λ) Figure 10. Graphical and tabular representation of the actuation sequence for the cases described (Δ=150) (Δ=300) (Δ=150) (Δ=300) above. Delta values refer to the change in encoder position from the flat state. Ave. encod. Pos 27,490 0.564 27,640 2.261 27,790 4.248 27,340 2.408 27,190 4.256 Stdev 0.16 0.02 0.00 0.05 0.00 0.02 0.00 0.02 0.00 0.04 Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was Figure 10. Graphical and tabular representation of the actuation sequence for the cases described Figure placed 10. atGraphical a distance and of tabular approxr iep ma resentation tely 1500 m ofm the an actuation d the ressequence ponse fro for m the a co cases llima described ted beam above. Delta values refer to the change in encoder position from the flat state. above. record Delta ed an values d the rA efer R w toathe s ac change tuatedin inencoder order tposition o focus fr th om e bthe eam flat on state. the screen. Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was placed at a distance of approximately 1500 mm and the response from a collimated beam recorded and the AR was actuated in order to focus the beam on the screen. Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown (left) Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors overlapping (left) conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors over- lapping in the screen with the AR actuated to focus the beam at that particular distance. in the screen with the AR actuated to focus the beam at that particular distance. 4. Conclusions Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown 4. Conclusions (left) conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors over- We have presented the concept of an adaptive retroreflector. This concept was de- We have presented the concept of an adaptive retroreflector. This concept was devel- lapping in the screen with the AR actuated to focus the beam at that particular distance. veloped during a data campaign to study the atmospheric turbulence in a dynamic link, oped during a data campaign to study the atmospheric turbulence in a dynamic link, with with the end goal of an optical anemometer for UAV applications in which the propagation the end goal of an optical anemometer for UAV applications in which the propagation 4. Conclusions distance is changing rapidly. The concept of the AR was then designed, fabricated and distance is changing rapidly. The concept of the AR was then designed, fabricated and We have presented the concept of an adaptive retroreflector. This concept was devel- tested in a laboratory environment as a proof of concept. This particular device can operate tested in a laboratory environment as a proof of concept. This particular device can oper- oped during a data campaign to study the atmospheric turbulence in a dynamic link, with from the VIS to the SWIR and preliminary parameters of its performance were studied. ate from the VIS to the SWIR and preliminary parameters of its performance were studied. the end goal of an optical anemometer for UAV applications in which the propagation The next step will consist of fabricating a device following the tighter tolerance procedures The next step will consist of fabricating a device following the tighter tolerance procedures developed distance ispr ceviously hanging for rapadaptive idly. The lenses. concepA t o follow-up f the AR w report as thwill en dconsist esignedof , fperforming abricated an ad developed previously for adaptive lenses. A follow-up report will consist of performing calibration tested in ain laa bo laboratory ratory env envir irononment, ment as a including proof of thermal conceptcompensation . This particula and r de quantification vice can oper- a calibration in a laboratory environment, including thermal compensation and quantifi- of atlosses e fromadded the VIS by toabsorption the SWIR aand/or nd preliscattering minary padue rame to tethe rs omembrane/fluid f its performance w combination ere studied. cation of losses added by absorption and/or scattering due to the membrane/fluid combi- in comparison with a conventional retroreflector. The latter case will be studied in more The next step will consist of fabricating a device following the tighter tolerance procedures nation in comparison with a conventional retroreflector. The latter case will be studied in detail develin opa ed field prev experiment iously for awher daptie ve we len can ses.compar A folloe wthe -up losses reportdue willto cothe nsis addition t of perfo of rm the ing more detail in a field experiment where we can compare the losses due to the addition of membrane/fluid combination with the losses of a conventional retroreflector (e.g., due to a calibration in a laboratory environment, including thermal compensation and quantifi- the membrane/fluid combination with the losses of a conventional retroreflector (e.g., due diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. cation of losses added by absorption and/or scattering due to the membrane/fluid combi- to diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. A power-in-the-bucket configuration will be used to compare the divergence control of the nation in comparison with a conventional retroreflector. The latter case will be studied in A power-in-the-bucket configuration will be used to compare the divergence control of adaptive retroreflector and a conventional one. While the overall losses depend on configu- more detail in a field experiment where we can compare the losses due to the addition of the adaptive retroreflector and a conventional one. While the overall losses depend on ration and materials, our experience with fluidic lenses has shown that the transmission the membrane/fluid combination with the losses of a conventional retroreflector (e.g., due losses are negligible compared to effects induced by turbulence. to diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. A power-in-the-bucket configuration will be used to compare the divergence control of the adaptive retroreflector and a conventional one. While the overall losses depend on Photonics 2022, 9, 124 10 of 10 5. Patents A provisional patent application has been submitted, U.S. Patent Application Serial No. 62/695,310. Author Contributions: Conceptualization, F.S. and C.O.F.; methodology, F.S., C.O.F., B.E.B. and S.R.R.; formal analysis, F.S., S.N.Q. and S.R.R.; writing—original draft preparation, F.S., C.O.F., S.R.R.; writing—review and editing, S.R.R., S.N.Q. and B.E.B.; funding acquisition, F.S. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data is contained within the article. Conflicts of Interest: The authors declare no conflict of interest. References 1. Font, C.; Apker, T.; Santiago, F. Laser Anemometer for Autonomous Systems Operations. In AIAA Infotech@ Aerospace; AIAA SciTech: San Diego, CA, USA, 2016; p. 1230. [CrossRef] 2. Font, C.; Santiago, F.; Apker, T. Atmospheric Turbulence Measurements in Dynamical Links. In Propagation Through and Characterization of Atmospheric and Oceanic Phenomena, OSA Technical Digest (Online), 3rd ed.; Paper M2A.3.; Optical Society of America: Washington, DC, USA, 2016; pp. 154–196. 3. Santiago, F.; Bagwell, B.; Martinez, T.; Restaino, S.; Krishna, S. Large aperture adaptive doublet polymer lens for imaging applications. J. Opt. Soc. Am. A 2014, 31, 1842–1846. [CrossRef] 4. Santiago, F.; Font, C.; Restaino, S. Adaptive Polymer Lenses at NRL. In Applied Industrial Optics 2019, OSA Technical Digest; Paper T2A.2.; Optical Society of America: Washington, DC, USA, 2019. 5. Santiago, F.; Bagwell, B.E.; Pinon, V., III; Krishna, S. Adaptive polymer lens for rapid zoom shortwave infrared imaging applications. Opt. Eng. 2014, 53, 125101. [CrossRef] 6. Berge, B.; Peseux, J. Variable focal lens controlled by an external voltage: An application of electrowetting. Eur. Phys. J. E 2000, 3, 159–163. [CrossRef] 7. Berge, B. Electrocapillarite et mouillage de films isolants par l’eau. C.R. Acad. Sci. Ser. II Mec. Phys. Chim. Sci. Terre Univ. 1993, 317, 157. 8. Mugele, F.; Baret, J. Electrowetting: From basics to applications. J. Phys. Condens. Matter IOP Publ. 2005, 17, R705–R774. [CrossRef] 9. Heikenfeld, J.; Smith, N.; Dhindsa, M.; Zhou, K.; Kilaru, M.; Hou, L.; Zhang, J.; Kreit, E.; Raj, B. Recent Progress in Arrayed Electrowetting Optics. Opt. Photonics News 2009, 20, 20–26. [CrossRef] 10. Yiu, J.; Batchko, R.; Robinson, S.; Szilagyi, A. A fluidic lens with reduced optical aberration. In Intelligent Robots and Computer Vision XXIX: Algorithms and Techniques; Proc. SPIE 8301; SPIE: Burlingame, CA, USA, 2012; p. 830117. 11. Blum, M.; Büeler, M.; Grätzel, C.; Giger, J.; Aschwanden, M. Optotune focus tunable lenses and laser speckle reduction based on electroactive polymers. In MOEMS and Miniaturized Systems XI; Proc. SPIE 8252; SPIE: San Francisco, CA, USA, 2012; p. 825207. 12. Vorontsov, M.A.; Sivokon, V.P. Stochastic parallel-gradient-descent technique for high-resolution wave-front phase-distortion correction. J. Opt. Soc. Am. A 1998, 15, 2745–2758. [CrossRef] 13. Font, C.O.; Gilbreath, G.C.; Bajramaj, B.; Kim, D.S.; Santiago, F.; Martinez, T.; Restaino, S.R. Characterization and training of a 19-element piezoelectric deformable mirror for lensing. J. Opt. Fiber. Commun. Res. 2010, 7, 1–9. [CrossRef] 14. Ma, S.; Yang, P.; Lai, B.; Su, C.; Zhao, W.; Yang, K.; Jin, R.; Cheng, T.; Xu, B. Adaptive Gradient Estimation Stochastic Parallel Gradient Descent Algorithm for Laser Beam Cleanup. Photonics 2021, 8, 165. [CrossRef] http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Photonics Multidisciplinary Digital Publishing Institute

Design, Fabrication and Characterization of an Adaptive Retroreflector (AR)

Loading next page...
 
/lp/multidisciplinary-digital-publishing-institute/design-fabrication-and-characterization-of-an-adaptive-retroreflector-1phzrjCPk4
Publisher
Multidisciplinary Digital Publishing Institute
Copyright
© 1996-2022 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer The statements, opinions and data contained in the journals are solely those of the individual authors and contributors and not of the publisher and the editor(s). MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Terms and Conditions Privacy Policy
ISSN
2304-6732
DOI
10.3390/photonics9030124
Publisher site
See Article on Publisher Site

Abstract

hv photonics Communication Design, Fabrication and Characterization of an Adaptive Retroreflector (AR) 1 , 1 1 1 2 Freddie Santiago * , Carlos O. Font , Sergio R. Restaino , Syed N. Qadri and Brett E. Bagwell Naval Research Laboratory, Washington, DC 20375, USA; carlos.font@nrl.navy.mil (C.O.F.); sergio.restaino@nrl.navy.mil (S.R.R.); noor.qadri@nrl.navy.mil (S.N.Q.) Sandia National Laboratories, Albuquerque, NM 94551, USA; bbagwel@sandia.gov * Correspondence: freddie.santiago@nrl.navy.mil Abstract: Recent work at the U.S. Naval Research Laboratory studied atmospheric turbulence on dynamic links with the goal of developing an optical anemometer and turbulence characterization system for unmanned aerial vehicle (UAV) applications. Providing information on the degree of atmospheric turbulence, as well as wind information and scintillation, in a low size, weight and power (SWaP) system is key for the design of a system that is also capable of adapting quickly to changes in atmospheric conditions. The envisioned system consists of a bi-static dynamic link between a transmitter (Tx) and a receiver (Rx), relying on a small UAV. In a dynamic link, the propagation distance between the Tx/Rx changes rapidly. Due to SWaP constraints, a monostatic system is challenging for such configurations, so we explored a system in which the Tx/Rx is co-located on a mobile platform (UAV), which has a mounted retroreflector. Beam divergence control is key in such a system, both for finding the UAV (increased beam divergence at the Tx) and for signal optimization at the Rx. This led us to the concept of using adaptive/active elements to control the divergence at the Tx but also to the implementation of an adaptive/active retroreflector in which the return beam divergence can be controlled in order to optimize the signal at the Rx. This paper presents the design, fabrication and characterization of a low SWaP adaptive retroreflector. Keywords: adaptive retroreflector; tunable lens; adaptive lens; polymer optics; divergence control; Citation: Santiago, F.; Font, C.O.; fluidic lens; tunable optics Restaino, S.R.; Qadri, S.N.; Bagwell, B.E. Design, Fabrication and Characterization of an Adaptive Retroreflector (AR). Photonics 2022, 9, 124. https://doi.org/10.3390/ 1. Introduction photonics9030124 Retroreflectors are passive devices that return the incident signal through the same propagation path. For our intended application on UAVs, a retroreflector is ideal, due to its Received: 25 January 2022 Accepted: 15 February 2022 size and zero power consumption. The fact that this is a dynamic link (with changing dis- Published: 22 February 2022 tance between the transmitter and the point where the signal is reflected occurring quickly or discretely) means that signal degradation is expected due to atmospheric turbulence Publisher’s Note: MDPI stays neutral induced effects, but also due to the general nature of a propagating beam. In order to with regard to jurisdictional claims in ameliorate these effects, we relied on low order adaptive optics correction, in this case, published maps and institutional affil- focus control. Due to the constraints in SWaP, we have designed and fabricated an adaptive iations. retroreflector which allows us to change the divergence of the beam in order to optimize the link, achieving higher link performance or longer distances than can normally be obtained with a passive system. This device enables the control of the divergence, which can be Copyright: © 2022 by the authors. used to optimize the return signal in a monostatic configuration or to increase the return Licensee MDPI, Basel, Switzerland. footprint of the beam in a bi-static, dynamic, or reconfigurable link (moving link), this latter This article is an open access article case being the motivation for the following types of devices [1,2]. distributed under the terms and Adaptive optical devices (also known as active or tunable devices) are devices that can conditions of the Creative Commons adjust their surface/curvature (such as deformable mirrors, fluidic lenses, elastic/elastomeric Attribution (CC BY) license (https:// solids) or modify their index of refractions (such as liquid crystals) in order to change the creativecommons.org/licenses/by/ optical properties of the element, such as its focal length. This leads to the design and 4.0/). Photonics 2022, 9, 124. https://doi.org/10.3390/photonics9030124 https://www.mdpi.com/journal/photonics Photonics 2022, 8, x FOR PEER REVIEW 2 of 10 Photonics 2022, 9, 124 2 of 10 design and fabrication of an adaptive retroreflector—the one described in this paper is based on fluidic optical elements, for which NRL has extensive expertise [3–5]. fabrication of an adaptive retroreflector—the one described in this paper is based on fluidic optical elements, for which NRL has extensive expertise [3–5]. 2. AR Design and Configurations Our adaptive retroreflector (AR) can be used with a corner cube retroreflector, solid 2. AR Design and Configurations or hollow, and consists of an optical fluid, encapsulated by an elastomeric membrane that Our adaptive retroreflector (AR) can be used with a corner cube retroreflector, solid can be deformed via an actuator—in this case, the same actuator we use for our adaptive or hollow, and consists of an optical fluid, encapsulated by an elastomeric membrane that polymer lenses. This actuator and its electronics have been designed for tactical applica- can be deformed via an actuator—in this case, the same actuator we use for our adaptive tions in which SWaP is key, making this ideal for UAV applications. polymer lenses. This actuator and its electronics have been designed for tactical applications Depending on the application, both the membrane and fluid can be replaced with an in which SWaP is key, making this ideal for UAV applications. elastomeric optical polymer (which can be made from the same material as the optical Depending on the application, both the membrane and fluid can be replaced with membrane) that can be deformed mechanically to make the adaptive retroreflector. This an elastomeric optical polymer (which can be made from the same material as the optical device enables contro-l of the divergence. membrane) that can be deformed mechanically to make the adaptive retroreflector. This The device can be fabricated in two ways: (1) using an elastomeric optical polymer, device enables contro-l of the divergence. or (2) a fluidic adaptive/tunable device with a hollow or solid retroreflector. While manu- The device can be fabricated in two ways: (1) using an elastomeric optical polymer, or facturing errors could change the operation of a retroreflector with an elastic polymer in (2) a fluidic adaptive/tunable device with a hollow or solid retroreflector. While manufac- front, by slightly changing the direction of return light, such errors can be easily quantified turing errors could change the operation of a retroreflector with an elastic polymer in front, and corrected, for example, by monitoring the overall optical performance with an inter- by slightly changing the direction of return light, such errors can be easily quantified and ferometer. corrected, for example, by monitoring the overall optical performance with an interferometer. For the elastomeric optical polymer option, the elastic polymer is molded to a desired For the elastomeric optical polymer option, the elastic polymer is molded to a desired initial initialshape shapeand andthe the change change on on the the polymer polymesurface r surfaccan e can be baf e fected affecteby d bmeans y mean of s o applying f applyin ag pr a essur presse/compr ure/comp ession ression to to the the polymer polyme . rAn . An elastomeric elastomeric optical optical polymer polymeris isa apolymer polymerthat that has high transmission at the user-desired operational wavelength and has elastic properties has high transmission at the user-desired operational wavelength and has elastic proper- which allow the solid substrate to be deformed. A second alternative to deform the polymer ties which allow the solid substrate to be deformed. A second alternative to deform the can be achieved by the use of dielectric elastomer actuation in which a voltage is applied to polymer can be achieved by the use of dielectric elastomer actuation in which a voltage is a pliable electrode and the polymer is deformed, creating the change on its surface. Figure 1 applied to a pliable electrode and the polymer is deformed, creating the change on its shows the configuration of the elastomer optical polymer, as well as three operational states surface. Figure 1 shows the configuration of the elastomer optical polymer, as well as three of the adaptive retroreflector. operational states of the adaptive retroreflector. Figure Figure1. 1.Schematic Schematico o f f an an adaptive adaptive retr reor tro eflector reflecto u rsing usina gsolid a soli elastomeric d elastomer polymer: ic polyme (a r) : ( flat a) f(lb at ) ( convex b) con- vex and (c) concave. and (c) concave. For a fluidic adaptive/tunable option, an elastomeric membrane encapsulates an optical For a fluidic adaptive/tunable option, an elastomeric membrane encapsulates an op- fluid which is mounted on the front of a hollow or solid retroreflector. The elastomeric tical fluid which is mounted on the front of a hollow or solid retroreflector. The elasto- membrane needs to have similar optical and mechanical properties to those described above meric membrane needs to have similar optical and mechanical properties to those de- for the solid option. The optical fluid, needs to be optically and chemically compatible scribed above for the solid option. The optical fluid, needs to be optically and chemically with the membrane and needs to have high transmission at the operational wavelength. compatible with the membrane and needs to have high transmission at the operational Polydimethylsiloxane is a common polymer that can be used for the membrane as well as wavelength. Polydimethylsiloxane is a common polymer that can be used for the mem- for the elastomeric solid option. For the optical fluids, there are numerous oils, polymers brane as well as for the elastomeric solid option. For the optical fluids, there are numerous and resins that have been studied (for example, water, glycerol, etc.) [3]. The actuation of oils, polymers and resins that have been studied (for example, water, glycerol, etc.) [3]. this system can be achieved by compressing/decompressing the flexible membrane, which The actuation of this system can be achieved by compressing/decompressing the flexible creates a change on its surface. This occurs by moving a cylinder along the optical axis of the membrane, which creates a change on its surface. This occurs by moving a cylinder along system, thus compressing the circumference of the flexible membrane. Besides the mechanical the optical axis of the system, thus compressing the circumference of the flexible mem- action, magnetic actuation or use of a compliant electrode (dielectric elastomer) can achieve brane. Besides the mechanical action, magnetic actuation or use of a compliant electrode actuation of the membrane. There are other actuation techniques that are situatable and could be implemented as well, such as those used by commercially available fluidic lenses, Photonics 2022, 8, x FOR PEER REVIEW 3 of 10 Photonics 2022, 9, 124 3 of 10 (dielectric elastomer) can achieve actuation of the membrane. There are other actuation techniques that are situatable and could be implemented as well, such as those used by commercially available fluidic lenses, for example, Varioptics, Optotune and Holochip [6– for example, Varioptics, Optotune and Holochip [6–12]. Figure 2, shows a conceptual sketch 12]. Figure 2, shows a conceptual sketch of the adaptive retroreflector based on the flexible of the adaptive retroreflector based on the flexible membrane/fluidic concept. membrane/fluidic concept. Figure Figure2. 2.Schematic Schematicof ofan anadaptive adaptiver etr retor roeflector reflector using using an an optical opticaliquid/fluidic: l liquid/fluidic:( ( aa ))flat flat( b (b ) )convex convex and (c) concave. and (c) concave. The latter configuration was selected for this paper. The device changes the divergence of The latter configuration was selected for this paper. The device changes the diver- the returned beam but can also work as a regular passive retroreflector if the system requires. gence of the returned beam but can also work as a regular passive retroreflector if the system requires. 3. Fabrication, Characterization and Results 3. Fa Her bric e, awe tion pr , C esent harathe cteroptical ization design and Rof esu the ltsAR as fabricated and in the configuration we used for testing. We describe the optical setups used for testing and comparing the device Here, we present the optical design of the AR as fabricated and in the configuration to a passive retroreflector. Data for the repeatability and arbitrary radius of curvature we used for testing. We describe the optical setups used for testing and comparing the measurements was acquired using a Zygo Verifire HD optical interferometer. Furthermore, device to a passive retroreflector. Data for the repeatability and arbitrary radius of curva- we show images taken comparing a passive retroreflector in comparison with AR, both ture measurements was acquired using a Zygo Verifire HD optical interferometer. Fur- actuated and flat. thermore, we show images taken comparing a passive retroreflector in comparison with AR, both actuated and flat. 3.1. AR Optical Design We used OpticStudio nonsequential tools to model the corner cube retroreflector and 3.1. AR Optical Design adaptive components and, for visualization purposes, a beam splitter cube was added, as We used OpticStudio nonsequential tools to model the corner cube retroreflector and shown in Figure 3. This was also the configuration chosen for the test. Note that we did adaptive components and, for visualization purposes, a beam splitter cube was added, as not model the thickness of the membrane, since the effects of the membrane are negligible shown in Figure 3. This was also the configuration chosen for the test. Note that we did in OpticStudio for this type of application. The model was performed using the volume not model the thickness of the membrane, since the effects of the membrane are negligible of the fluid, the fluid acting as a lens which changes its radius of curvature and center in OpticStudio for this type of application. The model was performed using the volume thickness. The figure shows a collimated beam, incident on a beam splitter cube which of the fluid, the fluid acting as a lens which changes its radius of curvature and center reflects part of the incident light and transmits a portion, which then impinges on the thickness. The figure shows a collimated beam, incident on a beam splitter cube which adaptive retroreflector. Light is reflected back from the retroreflector and reflected again reflects part of the incident light and transmits a portion, which then impinges on the from the beam splitter cube and incident on the detector. In field operations, the beam adaptive retroreflector. Light is reflected back from the retroreflector and reflected again splitter can be used to monitor the incoming beam and direct the adaptive retroreflector from the beam splitter cube and incident on the detector. In field operations, the beam in order to control the divergence. It can also be used without the beam splitter cube, splitter can be used to monitor the incoming beam and direct the adaptive retroreflector such that the beam can be monitored at the receiver side and the system optimized in in order to control the divergence. It can also be used without the beam splitter cube, such a power-in-the-bucket (PIB) configuration using well-known algorithms (e.g., stochastic that the beam can be monitored at the receiver side and the system optimized in a power- parallel-gradient-descent) [13,14]; the AR is then instructed by this information. in-the-bucket (PIB) configuration using well-known algorithms (e.g., stochastic parallel- 3.2. Fabrication of the AR gradient-descent) [13,14]; the AR is then instructed by this information. For this particular design we used a 12.7 mm corner cube retroreflector from Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 (at = 589 nm). The first step consisted in making the PDMS membrane which was then bonded to the glass support structure. The fluid was added to the membrane/glass structure and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication Photonics 2022, 8, x FOR PEER REVIEW 4 of 10 Photonics 2022, 9, 124 4 of 10 Figure 3. Optical design setup used for testing including the collimated source incident on the beam procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent as pictures of the assembled AR (in its actuator). The actuator was custom-made for our beam. adaptive lenses, and we were able to modify one to accommodate the AR. The actuator consists of a modified motor in a custom housing, with a maximum clear aperture of 3.2. Fabrication of the AR 19.5 mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power For this particular design we used a 12.7 mm corner cube retroreflector from consumption ~15 W, idle power consumption of ~1–500 W and temperature monitoring Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the of 0.01 C. The electronics can control two actuators at the same time and can run off three PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 Photonics 2022, 8, x FOR PEER REVIECR-123 W batteries (two batteries for a single actuator) with an average number of actuations 4 of 10 (at λ = 589 nm). The first step consisted in making the PDMS membrane which was then of about 6000 per set of batteries. bonded to the glass support structure. The fluid was added to the membrane/glass struc- ture and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well as pictures of the assembled AR (in its actuator). The actuator was custom-made for our adaptive lenses, and we were able to modify one to accommodate the AR. The actuator consists of a modified motor in a custom housing, with a maximum clear aperture of 19.5 mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power con- sumption ~15 W, idle power consumption of ~1–500 µW and temperature monitoring of 0.01 °C. The electronics can control two actuators at the same time and can run off three Figure Figure 3. 3. O Op pt ti ic ca al l d de es si ig gn n s se et tu up p u us se ed d f fo or r t te es st ti in ng g i in nc cl lu ud di in ng g t th he e c co ol ll li im ma at te ed d s so ou ur rc ce e iin nc ciid de en ntt o on n tth he e b be ea am m CR-123 batteries (two batteries for a single actuator) with an average number of actuations splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent splitter cube with the outputs: (a) a collimated beam, (b) a converging beam, and (c) a divergent beam. of about 6000 per set of batteries. beam. 3.2. Fabrication of the AR For this particular design we used a 12.7 mm corner cube retroreflector from Thorlabs, polydimethysiloxane (PDMS) as the membrane, glass support structures for the PDMS, and an optical fluid with an index of refraction of 1.45 and an Abbe number of 45.0 (at λ = 589 nm). The first step consisted in making the PDMS membrane which was then bonded to the glass support structure. The fluid was added to the membrane/glass struc- ture and the corner cube was bonded to it. The last step was to mount the AR into the actuator and start the testing—the assembly steps are shown in Figure 4. An important note: for this proof of concept, we did not follow the special fabrication procedures that we normally utilize to reduce the surface wavefront error which involve the reduction of coma induced by gravity and astigmatism resulting from the materials and fabrication procedures. Figure 4 shows the top-level schematic of the assembly procedure, as well as Figure 4. Schematic representation of the fabrication and assembly process of the AR. Figure 4. Schematic representation of the fabrication and assembly process of the AR. pictures of the assembled AR (in its actuator). The actuator was custom-made for our adaptive lenses, and we were able to modify one to accommodate the AR. The actuator 3.3. Laboratory Optical Setup 3.3. Laboratory Optical Setup consists of a modified motor in a custom housing, with a maximum clear aperture of 19.5 The AR was tested in various ways. Firstly, in the same way that we measured The AR was tested in various ways. Firstly, in the same way that we measured the mm, optical encoder with a resolution of ~50 nm, speed of ~2.5 mm/s, peak power con- the radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we meas- sumption ~15 W, idle power consumption of ~1–500 µW and temperature monitoring of measured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a ured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a negative 0.01 °C. The electronics can control two actuators at the same time and can run off three negative (concave surface) of 185 mm. The second setup was to compare the performance CR-123 batteries (two batteries for a single actuator) with an average number of actuations and proof of concept of the AR in comparison with a passive retroreflector, as shown in of about 6000 per set of batteries. Figure 5. We used the HeNe 633 nm source of the Zygo interferometer and a 1550 nm was co- aligned for further testing. Data from the 1550 nm was not included but performance of the active surface component at this wavelength has been demonstrated in a previous report [5]. We were able to use the beam collimated or with the addition of a known divergence that could be removed with the AR and compared with the passive retroreflector. The setup with the beam splitter cube allowed us to look at the return beam with the interferometer and, on the other arm, to look at the output with a camera, photodetector or power meter, while we were able to use beam blocks to look at each retroreflector individually or additionally, enabling viewing of the interference fringes formed by the two. This facilitated alignment, but also monitoring of the difference when the AR is actuated. This same setup was also Figure 4. Schematic representation of the fabrication and assembly process of the AR. 3.3. Laboratory Optical Setup The AR was tested in various ways. Firstly, in the same way that we measured the radius of curvature (ROC) of lenses using a Zygo interferometer—for this device we meas- ured two ROCs as a test, a positive (convex surface) with a ROC of 234 mm and a negative Photonics 2022, 8, x FOR PEER REVIEW 5 of 10 (concave surface) of −185 mm. The second setup was to compare the performance and proof of concept of the AR in comparison with a passive retroreflector, as shown in Figure 5. We used the HeNe 633 nm source of the Zygo interferometer and a 1550 nm was co- aligned for further testing. Data from the 1550 nm was not included but performance of the active surface component at this wavelength has been demonstrated in a previous report [5]. We were able to use the beam collimated or with the addition of a known di- vergence that could be removed with the AR and compared with the passive retroreflec- tor. The setup with the beam splitter cube allowed us to look at the return beam with the interferometer and, on the other arm, to look at the output with a camera, photodetector Photonics 2022, 9, 124 5 of 10 or power meter, while we were able to use beam blocks to look at each retroreflector in- dividually or additionally, enabling viewing of the interference fringes formed by the two. This facilitated alignment, but also monitoring of the difference when the AR is actuated. uT se hd ist s oap m ee rf s oe rt m up fi r w st as o r a d ls eo r m usee ad s u to re p m ee rn fo ts rm of ftih rs et ro er p d ee ar t a m bie la its yuo re f m the en A ts R ob f y th ae c tru ea pte in at gat b h il eitA yR of fr to hm e A aR fl b at y sa tc attu ea tt o in agc t o h n ev A ex Ro fr ro cm on a c a fv laet s st ta at te e a to n d a c bo an cv ketx o o arf c la otns cta av te e, sw tah te il e an m de b aa su ck ri t n og at h fle at su stra ft ae c,e w fo hrim le w mie th astu hr ein in gt e th rfe e r so um rfe atcee r .fo Arn m i m wp ito h r tta h n et in no te te r:fe te ro m m pe et re ar t.u A renw im asp m or otn ain to t r n eo dte in : tte h m e - room, but temperature compensation of the AR was not used—the room environment was perature was monitored in the room, but temperature compensation of the AR was not st u asb eld e— an td het h ro uo s m co e m np ve ir no sn at m io en nt w w as as n o sttarb elq eu ain re d d t .hus compensation was not required. Figure 5. (Left) Layout of the optical testing setup. (Right) Picture of optical devices used Figure 5. (Left) Layout of the optical testing setup. (Right) Picture of optical devices used for meas- for measurements. urements. 3.4. Results/Discussion 3.4. Results/Discussion 3.4.1. ROC and Surface Measurements 3.4.1. ROC and Surface Measurements The first test consisted of measurements of the ROC, positive and negative, in order to The first test consisted of measurements of the ROC, positive and negative, in order evaluate the performance of the device. Figure 6 shows a set of measurements, including a to evaluate the performance of the device. Figure 6 shows a set of measurements, includ- (left) measurement for a positive ROC of 255 mm and a (right) measurement for a negative ing a (left) measurement for a positive ROC of 255 mm and a (right) measurement for a ROC of 184 mm. The top row shows the 3D surface profile, and the bottom row indicates the negative ROC of 184 mm. The top row shows the 3D surface profile, and the bottom row 2D profile. The black circles in the figure are software masks used to remove unnecessary indicates the 2D profile. The black circles in the figure are software masks used to remove back reflections created by dust particles in the reference sphere. Within the respective unnecessary back reflections created by dust particles in the reference sphere. Within the figures, the left column (A or C) is the raw measurement and the right side (B or D) is with respective figures, the left column (A or C) is the raw measurement and the right side (B the dominant aberrations removed. As mentioned before, fabrication was not optimized or D) is with the dominant aberrations removed. As mentioned before, fabrication was for the surface figure, but what can be seen is the typical dominant aberration of coma not optimized for the surface figure, but what can be seen is the typical dominant aberra- and astigmatism, which are characteristic for this type of fluidic structures. Coma is due tion of coma and astigmatism, which are characteristic for this type of fluidic structures. to gravity and astigmatism is due to fabrication or assembly procedures. For the positive Coma is due to gravity and astigmatism is due to fabrication or assembly procedures. For ROC case demonstrated below, coma is the dominant aberration. On the negative ROC, the positive ROC case demonstrated below, coma is the dominant aberration. On the neg- there is a combination of coma and astigmatism, because measurements were taken close to ative ROC, there is a combination of coma and astigmatism, because measurements were the negative resting ROC (fabricated ROC) of the membrane for the fabricated AR device. taken close to the negative resting ROC (fabricated ROC) of the membrane for the fabri- The fabricated aberrations were more noticeable closer to the resting ROC because, for cated AR device. The fabricated aberrations were more noticeable closer to the resting this type of actuation mechanism, this is the point of contact where boundary conditions ROC because, for this type of actuation mechanism, this is the point of contact where are established between the membrane and actuation surface for the clear aperture. At boundary conditions are established between the membrane and actuation surface for the this point, the amplitude of any existing aberrations can be enhanced. Another aberration that cleacan r apbe ertu noticed re. At tis his tr p efoil ointon , thboth e amp ROCs—this litude of anwas y exipur stinely g ab due errato tiothe ns c assembly an be enhin anthe ced. actuator Anothe.rW ae be developed rration thapr t c ocedur an be es nofor ticefabrication d is trefoil and on b assembly oth ROCs that —th rieduce s was the purdominant ely due to aberrations which are implemented when building adaptive polymer lenses, with the caveat the assembly in the actuator. We developed procedures for fabrication and assembly that that redu we ce can the d minimize ominant a coma berrat based ions w on hicthe h ar application, e implement but ed w do henot n bu completely ilding adap eliminate tive poly- it. mThe er lepr nsocedur es, with e to the eliminate caveat thcoma at we during can min fabrication imize coma is b extr ased emel on y thcomplex, e applicatcostly ion, bu and t do time consuming if performed at the active surface. There are other ways to minimize it, including using a corrective element along the optical path of the system or close to the active surface, and this is a typical configuration used in commercial adaptive/tunable lenses [11]. Photonics 2022, 8, x FOR PEER REVIEW 6 of 10 not completely eliminate it. The procedure to eliminate coma during fabrication is ex- tremely complex, costly and time consuming if performed at the active surface. There are other ways to minimize it, including using a corrective element along the optical path of Photonics 2022, 9, 124 6 of 10 the system or close to the active surface, and this is a typical configuration used in com- mercial adaptive/tunable lenses [11]. Figure 6. (Left) Positive ROC measurement (A,B columns) 2D and 3D profiles with dominant aber- Figure 6. (Left) Positive ROC measurement (A,B columns) 2D and 3D profiles with dominant ration removed, in this instance being coma (B column). (Right) Negative ROC measurement (C,D aberration removed, in this instance being coma (B column). (Right) Negative ROC measurement columns) 2D and 3D profiles with dominant aberration removed, in this instance, coma and astig- (C,D columns) 2D and 3D profiles with dominant aberration removed, in this instance, coma and matism (D column). astigmatism (D column). Using the data from Figure 6, a Zernike fit was performed using the Mx software Using the data from Figure 6, a Zernike fit was performed using the Mx software tools tools from the interferometer and coefficients of the fit for both the positive and negative from the interferometer and coefficients of the fit for both the positive and negative ROCs ROCs are shown in Table 1. are shown in Table 1. Table 1. Results from the Zernike fit coefficients obtained from the data in Figure 6, for positive ROC Table 1. Results from the Zernike fit coefficients obtained from the data in Figure 6, for positive ROC (top) and negative ROC (bottom). (top) and negative ROC (bottom). ROC = 225 mm ROC = 225 mm Zernike Fit Zernike Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) ZFR 3 0.018 2 0 1 + 22 ZFR 3 0.018 2 0 −1 + 2ρ2 ZFR 4 0.085 2 2 2cos(2) ZFR 4 −0.085 2 2 ρ2cos(2θ) ZZFR FR 5 5 −0. 40 0.407 7 2 2 −2 2 ρ2sin (2sin(2 2θ) ) ZFR 6 −0.122 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.122 3 1 (2 + 33)cos() ZFR 7 2.191 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 2.191 3 1 (2 + 33)sin() ZFR 8 −0.047 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.047 4 0 1 62 + 64 ROC = −184 mm ROC = 184 mm Zernike ZernikeFit Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) Z ZFR FR 3 3 −0 .00.049 49 2 2 0 0 −1 + 2ρ 12 + 22 ZFR 4 −0.211 2 2 ρ2cos(2θ) ZFR 4 0.211 2 2 2cos(2) ZFR 5 −1.823 2 −2 ρ2sin(2θ) ZFR 5 1.823 2 2 2sin(2) ZFR 6 −0.108 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.108 3 1 (2 + 33)cos() ZFR 7 −2.115 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 2.115 3 1 (2 + 33)sin() ZFR 8 −0.372 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.372 4 0 1 62 + 64 Figure 7, shows data taken for the AR at the same ROCs mentioned above but in a perpendicular configuration in order to eliminate the effects of coma due to gravity. Note, that for the data no terms have been removed. Astigmatism and trefoil were noticeable but the large magnitude due to coma was absent. Photonics 2022, 8, x FOR PEER REVIEW 7 of 10 Figure 7, shows data taken for the AR at the same ROCs mentioned above but in a perpendicular configuration in order to eliminate the effects of coma due to gravity. Note, Photonics 2022, 9, 124 7 of 10 that for the data no terms have been removed. Astigmatism and trefoil were noticeable but the large magnitude due to coma was absent. Figure 7. (A) Positive and (B) negative ROC, 2D and 3D surface representation for the perpendicular Figure 7. (A) Positive and (B) negative ROC, 2D and 3D surface representation for the perpendicular setup. Right side shows a picture of the setup. setup. Right side shows a picture of the setup. The same procedure was performed on the results from Figure 7 and Table 2 shows The same procedure was performed on the results from Figure 7 and Table 2 shows the Zernike fit coefficients for the perpendicular measurements. the Zernike fit coefficients for the perpendicular measurements. Table 2. Results from the Zernike fit coefficients obtained from the data in Figure 7, for positive Table 2. Results from the Zernike fit coefficients obtained from the data in Figure 7, for positive ROC ROC (top) and negative ROC (bottom). (top) and negative ROC (bottom). ROC = 225 mm Perpendicular ROC = 225 mm Perpendicular Zernike ZernikeFit Fit Coeff Value () n m Representation Coeff Value (λ) n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 ZFR 1 0.000 1 1 cos() ZFR 1 0.000 1 1 ρcos(θ) ZFR 2 0.000 1 1 sin() ZFR 2 0.000 1 −1 ρsin(θ) Z ZFR FR 3 3 0.0.038 038 2 2 0 0 −1 + 2 ρ1 2+ 22 ZFR 4 0.137 2 2 ρ2cos(2θ) ZFR 4 0.137 2 2 2cos(2) ZFR 5 −0.390 2 −2 ρ2sin(2θ) ZFR 5 0.390 2 2 2sin(2) ZFR 6 0.065 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.065 3 1 (2 + 33)cos() ZFR 7 −0.071 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 0.071 3 1 (2 + 33)sin() ZFR 8 −0.099 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.099 4 0 1 62 + 64 ROC = −184 mm Perpendicular ROC = 184 mm Perpendicular Zernike Fit Zernike Fit Coeff Value (λ) n m Representation Coeff Value () n m Representation ZFR 0 0.000 0 0 1 ZFR 0 0.000 0 0 1 Z ZFR FR 1 1 0.0.000 000 1 1 1 1 ρcos(θ cos( ) ) ZFR 2 0.000 1 −1 ρsin(θ) ZFR 2 0.000 1 1 sin() ZFR 3 −0.202 2 0 −1 + 2ρ2 ZFR 3 0.202 2 0 1 + 22 ZFR 4 −0.052 2 2 ρ2cos(2θ) ZFR 4 0.052 2 2 2cos(2) ZFR 5 −1.248 2 −2 ρ2sin(2θ) ZFR 5 1.248 2 2 2sin(2) ZFR 6 0.126 3 1 (−2ρ + 3ρ3)cos(θ) ZFR 6 0.126 3 1 (2 + 33)cos() ZFR 7 −0.112 3 −1 (−2ρ + 3ρ3)sin(θ) ZFR 7 0.112 3 1 (2 + 33)sin() ZFR 8 −0.224 4 0 1 − 6ρ2 + 6ρ4 ZFR 8 0.224 4 0 1 62 + 64 3.4.2. Repeatability Measurements 3.4.2. Repeatability Measurements Repeatability measurements were taken using the setup in Figure 5. The data collec- tion consisted in changing the actuation state by a known encoder count from a flat state Repeatability measurements were taken using the setup in Figure 5. The data collection cto on a s ic st o en dvi en x/c ch oa nn cg aiv n eg stth ae tea , c w tu ha il te io r n ec so ta rt d ein bg y ta hk e n eo nw co n de en r c p oo d se it rio co nu a n st w fre olm l aa s t fh la et d sta atta e ftro om a convex/concave state, while recording the encoder position as well as the data from the inferferometer. The encoder data is in the form of a set of three numbers: the set position by user (state of the lens), the temperature compensate position (once thermal compensation is activated) and the measured position. This last position, or the difference from the set Photonics 2022, 8, x FOR PEER REVIEW 8 of 10 Photonics 2022, 8, x FOR PEER REVIEW 8 of 10 the inferferometer. The encoder data is in the form of a set of three numbers: the set posi- the inferferometer. The encoder data is in the form of a set of three numbers: the set posi- Photonics 2022, 9, 124 8 of 10 tion by user (state of the lens), the temperature compensate position (once thermal com- tion by user (state of the lens), the temperature compensate position (once thermal com- pensation is activated) and the measured position. This last position, or the difference pensation is activated) and the measured position. This last position, or the difference from the set position, was recorded. Readings from the interferometer PV(λ) (peak to val- from the set position, was recorded. Readings from the interferometer PV(λ) (peak to val- position, was recorded. Readings from the interferometer PV() (peak to valley) and power() ley) and power(λ) were recorded as well. Data was taken for a delta for encoder counts of ley) and power(λ) were recorded as well. Data was taken for a delta for encoder counts of were recorded as well. Data was taken for a delta for encoder counts of 150 and 300 from 150 and 300 from flat, on both positive (convex or higher encoder counts), and negative 150 and 300 from flat, on both positive (convex or higher encoder counts), and negative flat, on both positive (convex or higher encoder counts), and negative (concave or lower (concave or lower encoder counts) direction. Figure 8 shows a sequence of consecutive (concave or lower encoder counts) direction. Figure 8 shows a sequence of consecutive encoder counts) direction. Figure 8 shows a sequence of consecutive measurements from flat measurements from flat to positive and Figure 9 shows measurements from flat to nega- measurements from flat to positive and Figure 9 shows measurements from flat to nega- to positive and Figure 9 shows measurements from flat to negative for a delta of 150 encoder tive for a delta of 150 encoder counts. Note, the hexagonal pattern was a result of the facets tive for a delta of 150 encoder counts. Note, the hexagonal pattern was a result of the facets counts. Note, the hexagonal pattern was a result of the facets of the corner cube. This was of the corner cube. This was noticeable in this configuration based on the testing setup of the corner cube. This was noticeable in this configuration based on the testing setup noticeable in this configuration based on the testing setup with the interferometer using a with the interferometer using a transmission flat. For the ROCs the measurements dif- with the interferometer using a transmission flat. For the ROCs the measurements dif- transmission flat. For the ROCs the measurements differed, since we were using a reference fered, since we were using a reference sphere and the spherical wavefront matched the fered, since we were using a reference sphere and the spherical wavefront matched the sphere and the spherical wavefront matched the deformed membrane, not the retroreflector. deformed membrane, not the retroreflector. deformed membrane, not the retroreflector. Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back Figure 8. The AR was actuated from the flat state to a compressed state (or convex surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. Figure 9. The AR was actuated from the flat state to a less compressed state (or concave surface) and back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. back to flat. For each case the top row is the 3D surface and the bottom row the 2D surface. In Figure 10, data is presented in graphical (with error bars based on the standard In Figure 10, data is presented in graphical (with error bars based on the standard In Figure 10, data is presented in graphical (with error bars based on the standard devi- deviation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 deviation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 ation), and tabular, form for the sequences, with 20 data points for delta 150 and 10 points points for delta 300, and all cases starting from the same initial flat position. The average points for delta 300, and all cases starting from the same initial flat position. The average for delta 300, and all cases starting from the same initial flat position. The average and and standard deviation for the encoder position and peak-to-valley for the cases are and standard deviation for the encoder position and peak-to-valley for the cases are standard deviation for the encoder position and peak-to-valley for the cases are shown in shown in the table. shown in the table. the table. Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was placed at a distance of approximately 1500 mm and the response from a collimated beam recorded and the AR was actuated in order to focus the beam on the screen. Photonics 2022, 8, x FOR PEER REVIEW 9 of 10 Photonics 2022, 8, x FOR PEER REVIEW 9 of 10 Photonics 2022, 9, 124 9 of 10 Flat PV(λ) Convex PV(λ) Convex PV(λ) Concave PV(λ) Concave PV(λ) (Δ=150) (Δ=300) (Δ=150) (Δ=300) Ave. encod. Pos 27,490 0.564 27,640 2.261 27,790 4.248 27,340 2.408 27,190 4.256 Stdev 0.16 0.02 0.00 0.05 0.00 0.02 0.00 0.02 0.00 0.04 Flat PV(λ) Convex PV(λ) Convex PV(λ) Concave PV(λ) Concave PV(λ) Figure 10. Graphical and tabular representation of the actuation sequence for the cases described (Δ=150) (Δ=300) (Δ=150) (Δ=300) above. Delta values refer to the change in encoder position from the flat state. Ave. encod. Pos 27,490 0.564 27,640 2.261 27,790 4.248 27,340 2.408 27,190 4.256 Stdev 0.16 0.02 0.00 0.05 0.00 0.02 0.00 0.02 0.00 0.04 Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was Figure 10. Graphical and tabular representation of the actuation sequence for the cases described Figure placed 10. atGraphical a distance and of tabular approxr iep ma resentation tely 1500 m ofm the an actuation d the ressequence ponse fro for m the a co cases llima described ted beam above. Delta values refer to the change in encoder position from the flat state. above. record Delta ed an values d the rA efer R w toathe s ac change tuatedin inencoder order tposition o focus fr th om e bthe eam flat on state. the screen. Figure 11 shows a comparison of the AR with a passive retroreflector. A screen was placed at a distance of approximately 1500 mm and the response from a collimated beam recorded and the AR was actuated in order to focus the beam on the screen. Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown (left) Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors overlapping (left) conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors over- lapping in the screen with the AR actuated to focus the beam at that particular distance. in the screen with the AR actuated to focus the beam at that particular distance. 4. Conclusions Figure 11. Images taken in the laboratory at a known distance from the retroreflector are shown 4. Conclusions (left) conventional retroreflector, (middle) AR in the flat state, and (right) both retroreflectors over- We have presented the concept of an adaptive retroreflector. This concept was de- We have presented the concept of an adaptive retroreflector. This concept was devel- lapping in the screen with the AR actuated to focus the beam at that particular distance. veloped during a data campaign to study the atmospheric turbulence in a dynamic link, oped during a data campaign to study the atmospheric turbulence in a dynamic link, with with the end goal of an optical anemometer for UAV applications in which the propagation the end goal of an optical anemometer for UAV applications in which the propagation 4. Conclusions distance is changing rapidly. The concept of the AR was then designed, fabricated and distance is changing rapidly. The concept of the AR was then designed, fabricated and We have presented the concept of an adaptive retroreflector. This concept was devel- tested in a laboratory environment as a proof of concept. This particular device can operate tested in a laboratory environment as a proof of concept. This particular device can oper- oped during a data campaign to study the atmospheric turbulence in a dynamic link, with from the VIS to the SWIR and preliminary parameters of its performance were studied. ate from the VIS to the SWIR and preliminary parameters of its performance were studied. the end goal of an optical anemometer for UAV applications in which the propagation The next step will consist of fabricating a device following the tighter tolerance procedures The next step will consist of fabricating a device following the tighter tolerance procedures developed distance ispr ceviously hanging for rapadaptive idly. The lenses. concepA t o follow-up f the AR w report as thwill en dconsist esignedof , fperforming abricated an ad developed previously for adaptive lenses. A follow-up report will consist of performing calibration tested in ain laa bo laboratory ratory env envir irononment, ment as a including proof of thermal conceptcompensation . This particula and r de quantification vice can oper- a calibration in a laboratory environment, including thermal compensation and quantifi- of atlosses e fromadded the VIS by toabsorption the SWIR aand/or nd preliscattering minary padue rame to tethe rs omembrane/fluid f its performance w combination ere studied. cation of losses added by absorption and/or scattering due to the membrane/fluid combi- in comparison with a conventional retroreflector. The latter case will be studied in more The next step will consist of fabricating a device following the tighter tolerance procedures nation in comparison with a conventional retroreflector. The latter case will be studied in detail develin opa ed field prev experiment iously for awher daptie ve we len can ses.compar A folloe wthe -up losses reportdue willto cothe nsis addition t of perfo of rm the ing more detail in a field experiment where we can compare the losses due to the addition of membrane/fluid combination with the losses of a conventional retroreflector (e.g., due to a calibration in a laboratory environment, including thermal compensation and quantifi- the membrane/fluid combination with the losses of a conventional retroreflector (e.g., due diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. cation of losses added by absorption and/or scattering due to the membrane/fluid combi- to diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. A power-in-the-bucket configuration will be used to compare the divergence control of the nation in comparison with a conventional retroreflector. The latter case will be studied in A power-in-the-bucket configuration will be used to compare the divergence control of adaptive retroreflector and a conventional one. While the overall losses depend on configu- more detail in a field experiment where we can compare the losses due to the addition of the adaptive retroreflector and a conventional one. While the overall losses depend on ration and materials, our experience with fluidic lenses has shown that the transmission the membrane/fluid combination with the losses of a conventional retroreflector (e.g., due losses are negligible compared to effects induced by turbulence. to diffraction, or divergence introduced by atmospheric turbulence) at a propagation path. A power-in-the-bucket configuration will be used to compare the divergence control of the adaptive retroreflector and a conventional one. While the overall losses depend on Photonics 2022, 9, 124 10 of 10 5. Patents A provisional patent application has been submitted, U.S. Patent Application Serial No. 62/695,310. Author Contributions: Conceptualization, F.S. and C.O.F.; methodology, F.S., C.O.F., B.E.B. and S.R.R.; formal analysis, F.S., S.N.Q. and S.R.R.; writing—original draft preparation, F.S., C.O.F., S.R.R.; writing—review and editing, S.R.R., S.N.Q. and B.E.B.; funding acquisition, F.S. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data is contained within the article. Conflicts of Interest: The authors declare no conflict of interest. References 1. Font, C.; Apker, T.; Santiago, F. Laser Anemometer for Autonomous Systems Operations. In AIAA Infotech@ Aerospace; AIAA SciTech: San Diego, CA, USA, 2016; p. 1230. [CrossRef] 2. Font, C.; Santiago, F.; Apker, T. Atmospheric Turbulence Measurements in Dynamical Links. In Propagation Through and Characterization of Atmospheric and Oceanic Phenomena, OSA Technical Digest (Online), 3rd ed.; Paper M2A.3.; Optical Society of America: Washington, DC, USA, 2016; pp. 154–196. 3. Santiago, F.; Bagwell, B.; Martinez, T.; Restaino, S.; Krishna, S. Large aperture adaptive doublet polymer lens for imaging applications. J. Opt. Soc. Am. A 2014, 31, 1842–1846. [CrossRef] 4. Santiago, F.; Font, C.; Restaino, S. Adaptive Polymer Lenses at NRL. In Applied Industrial Optics 2019, OSA Technical Digest; Paper T2A.2.; Optical Society of America: Washington, DC, USA, 2019. 5. Santiago, F.; Bagwell, B.E.; Pinon, V., III; Krishna, S. Adaptive polymer lens for rapid zoom shortwave infrared imaging applications. Opt. Eng. 2014, 53, 125101. [CrossRef] 6. Berge, B.; Peseux, J. Variable focal lens controlled by an external voltage: An application of electrowetting. Eur. Phys. J. E 2000, 3, 159–163. [CrossRef] 7. Berge, B. Electrocapillarite et mouillage de films isolants par l’eau. C.R. Acad. Sci. Ser. II Mec. Phys. Chim. Sci. Terre Univ. 1993, 317, 157. 8. Mugele, F.; Baret, J. Electrowetting: From basics to applications. J. Phys. Condens. Matter IOP Publ. 2005, 17, R705–R774. [CrossRef] 9. Heikenfeld, J.; Smith, N.; Dhindsa, M.; Zhou, K.; Kilaru, M.; Hou, L.; Zhang, J.; Kreit, E.; Raj, B. Recent Progress in Arrayed Electrowetting Optics. Opt. Photonics News 2009, 20, 20–26. [CrossRef] 10. Yiu, J.; Batchko, R.; Robinson, S.; Szilagyi, A. A fluidic lens with reduced optical aberration. In Intelligent Robots and Computer Vision XXIX: Algorithms and Techniques; Proc. SPIE 8301; SPIE: Burlingame, CA, USA, 2012; p. 830117. 11. Blum, M.; Büeler, M.; Grätzel, C.; Giger, J.; Aschwanden, M. Optotune focus tunable lenses and laser speckle reduction based on electroactive polymers. In MOEMS and Miniaturized Systems XI; Proc. SPIE 8252; SPIE: San Francisco, CA, USA, 2012; p. 825207. 12. Vorontsov, M.A.; Sivokon, V.P. Stochastic parallel-gradient-descent technique for high-resolution wave-front phase-distortion correction. J. Opt. Soc. Am. A 1998, 15, 2745–2758. [CrossRef] 13. Font, C.O.; Gilbreath, G.C.; Bajramaj, B.; Kim, D.S.; Santiago, F.; Martinez, T.; Restaino, S.R. Characterization and training of a 19-element piezoelectric deformable mirror for lensing. J. Opt. Fiber. Commun. Res. 2010, 7, 1–9. [CrossRef] 14. Ma, S.; Yang, P.; Lai, B.; Su, C.; Zhao, W.; Yang, K.; Jin, R.; Cheng, T.; Xu, B. Adaptive Gradient Estimation Stochastic Parallel Gradient Descent Algorithm for Laser Beam Cleanup. Photonics 2021, 8, 165. [CrossRef]

Journal

PhotonicsMultidisciplinary Digital Publishing Institute

Published: Feb 22, 2022

Keywords: adaptive retroreflector; tunable lens; adaptive lens; polymer optics; divergence control; fluidic lens; tunable optics

There are no references for this article.