The optics industry is experiencing trends of both increasing laser power and the advancement of coating technology to accommodate this demand. However, an optic does not always need to utilize leading-edge coating technology to implement high-power lasers into a system. A second solution is to increase the size of the beam, and therefore the size of the optics, which will lower the overall power or energy density per unit area on the optic. This requires large beam expansion optics, as well as focusing optics further along the optical path.
A second catalyst for increasing optic size would be any light collection system that is collecting collimated light. The larger the diameter of an optic, the more surface area to collect. In both of these cases, and countless others, there are performance increases that can be realized by designing in aspheric lenses opposed to spherical lenses. Previously, designers may have been hesitant to design in aspheres with diameters larger than 100 mm, having concerns around manufacturability and metrology equipment available to verify an asphere so large. However, with advancements in both manufacturing and metrology, aspheres as large as 200 mm are now commercially available.
Large Versus Very Large
When discussing large aspheres, it is important to make a distinction between large aspheres and very large aspheres, which cannot be hand-carried by a single person and require mechanical support to move them around. Those present even more challenges and require planning out the manufacturing process up front in great detail.
Although it is definitely fun to align a 1.02 meter optic on a polishing machine using a sledge hammer, the focus of this article is on the manufacturing limits of mass-produced large aspheres. These lenses have design considerations and limitations in addition to the general manufacturability considerations for normal-sized aspheres.
An obvious manufacturing limit that comes to mind is the size of the asphere grinding and polishing machines. Many machine manufacturers conveniently label their models (e.g. CNC100, CNC200 or CNC300), which tends to be related to range of motion of the machine. Unfortunately, this does not mean that a “CNC200” machine can be used to machine a 200 mm diameter large asphere.
For starters, during the manufacturing process a larger diameter blank is used, which is then edged-down to the final diameter in one of the last processing steps. But more importantly, the size limitation of a machine is given by a combination of the machine’s kinematics and the shape of the optic. For instance, let us assume that the optic is placed facing up on a spindle in the center of the machine, and a spinning disk tool is moved radially, starting at one edge and finishing at the other, and the vertical position of the tool is adjusted by the machine as required by the shape of the optic. From there, it follows that for a concave optic the tool has to travel much less horizontally to machine the same sized part than for a convex optic.
The optical fabrication engineer might be able to squeeze out an extra few millimeters of range by tweaking the process parameters, but these most likely will negatively impact cost and/or quality and/or lead time. In the example above, one might reduce the diameter of the wheel, but this will limit the cutting speed and increase the process time and increase the tool wear. As such, these labels do not indicate a hard limit, but a transition from economical to expensive to not feasible.
In addition to the dimensions of an optic, an asphere grinding and polishing machine also has a limit on the maximum weight it is able to machine. Depending on the kinematics of the machine, the optic might be spun and/or translated and the motors that affect these motions need to have enough torque to generate the required acceleration. In some cases, this means that the machine has to be configured specifically for heavy workpieces, which might lead to longer cycle times and thus higher cost.
In general, machine manufacturers select motors strong enough to machine the weights typical of workpieces in the size class, so this should be less of an issue. However, keep in mind that during manufacturing the optic is typically bonded to a carrier for easy transfer and alignment between machines and measuring equipment, which also adds weight.
Speaking about measurements, the limitations of the metrology equipment should not be overlooked. And, of course, the metrology platform needs to have enough travel to reach the full diameter of the optic.
During manufacturing, an asphere is typically measured using a tactile profilometer. With increased size of the optic, it is also likely that the sagittal height is increased (but this very much depends on the actual design of the asphere). Another limiting factor of a tactile profilometer, in addition to the travel range, is the height of the stylus used. This limits how much it is able to reach over the vertex of a convex asphere to measure the profile of the surface on the opposite side (Figure 2).
A concave optic has an analogous limitation for reaching into the optic to measure the vertex. There are some tricks that the optical fabrication engineer can apply to squeeze some more mileage out of the metrology platform that he/she has at their disposal, but these will again affect cost and/or quality and/or lead time.
In addition, having to use a larger stylus may negatively affect the accuracy of the metrology due to the increased weight, flex, and instability, and thus also negatively affect cost and/or quality and/or lead time.
Typically, the non-aspheric backside of an aspheric lens has a limited influence on the manufacturability analysis and cost. For large aspheres this is no longer true. Obviously, the equipment used needs to be able to accommodate the size of the optics. More problematic is the metrology solution, typically a large aperture interferometer. If an optical shop also produces components such as prisms, beamsplitters, and windows, it can most likely leverage the existing equipment. Even so, not many asphere manufacturers have a standard solution to measure planar surfaces beyond 10 inches (254 mm).
For convex spherical backsides, the metrology solutions are even more limited, as investing in the large aperture interferometer and associated large aperture transmission spheres is often cost-prohibitive or unavailable. For both convex and concave spherical backsides, a larger diameter goes hand-in-hand with a larger radius of curvature (RoC). Typically, the RoC is controlled by moving a stage with the optic mounted along a rail between the cat’s eye position (where the interferometer’s beam contacts a single point on the spherical surface) and confocal position (where the point focus of the interferometer beam is at the radius of curvature). Thus, the range of RoC that can be measured is limited by the length of the rail.
In addition, the use of test plates for in-process control is risky and cumbersome for large diameter optics. Not to mention the same difficulties as mentioned above apply to the manufacture of the test plates themselves.
Of course, to measure the backside of an aspheric lens, one could make use of the available asphere metrology. However, this makes the manufacturing process costly and inefficient, as the spherical surface would be competing with the aspheric side for measurement time on an expensive platform and asphere metrology tends to be more time consuming and/or require additional skills not typically found in spherical optics craftsmen. As such, it is typically impractical to take a quick peek at the spherical surface using asphere metrology during manufacturing to monitor the process and adjust process parameters if necessary.
As mentioned earlier, as one of the last processing steps, the diameter of the part is edged down to the final diameter. If the optical shop does not have one or more dedicated edging machines, or if they are not large enough to handle the large diameter, the parts will have to be edged on the asphere grinding machine. This is both inefficient and expensive.
Surface Quality and Inspection
Arguably, the number of surface imperfections created correlates with the area processed. As such, it is more difficult to maintain a tight surface quality tolerance specification on a larger diameter optic, whether specified using the ISO or MIL standard. In addition, a larger diameter optic is more difficult to handle and, as such, at a higher risk of surface defects due to mishandling. In addition, surface inspection is especially cumbersome for large diameter optics, as they require a lot of handling.
The blank can come either as a cut-disk (a disc cut from a rod of adequate diameter) or a pressing (annealed in custom-made molds). For regular-sized aspheres, it can be a factor of 3- or 4-times more economical to use pressings for high-volume production, depending on the exact material. For a large asphere blank, the material cost becomes the driving factor over the labor cost, as the volume increases. As such, pressings become less advantageous to use for large asphere blanks, especially considering that pressings have a longer lead time and are limited to a center thickness of approximately 40 mm.
As mentioned previously, with increased size of the optic, it is also likely that the sagittal height is increased. This will negatively influence the coating uniformity, so keep in mind that specifying the same coating uniformity as typical for regular-sized aspheres on a large asphere will most likely incur a premium.
Keeping these manufacturing and metrology considerations in mind allows optical designs to incorporate large diameter aspheres into their optical systems. The resulting systems pave the way for high-power laser applications and high-throughput light collection systems. Sometimes bigger truly is better.