Chapter 26 My thoughts on Opto Mechanical

This chapter takes a different approach than the previous ones. Rather than diving into a specific technical procedure, I want to share my overall view of opto-mechanical alignment.

This perspective was recently prompted by a request to review a journal paper on the subject. After a quick look at the paper, I honestly did not know whether to laugh or cry. The paper was indeed about alignment. It featured pictures of interferograms of tilted and defocused surfaces showing that the test object was not well aligned. The text, however, was full of mathematics covering Fourier transforms, Zernike polynomials, and the use of interferometers for alignment.

While I agree that all these mathematical and optical tools can be used in the process of alignment, I question why one needs them. I declined to review the paper because the math fell outside my field of competence.

To me, alignment is getting the optical axes of individual optical elements coaxial with a common reference axis. Because the optical axis of an individual element is defined by its centers of curvature, alignment simply requires a means of optically locating those centers and moving the elements to the locations specified by the system’s optomechanical design.

Centers of curvature are defined by discrete points in space that have three degrees of freedom. While you could use two tilt coefficients and one focus Zernike coefficient to specify this location, there are simpler, more direct ways to achieve the same result.

I should step back a moment to clarify that I certainly believe in interferometry and the use of Zernike polynomials during the fabrication of individual optical elements. The combination of interferometry, computer-controlled polishing, magnetorheological finishing, and ion milling allows us to manufacture optical elements to extreme precision.

However, once those near-perfect components are handed to the assembly technician, fabrication is over. There is nothing they can do with these elements other than properly orient them in six degrees of freedom according to the opto-mechanical design.

The assembly technician does not need an interferometer to find a center of curvature. As discussed in previous chapters, several simpler methods of finding them exist.

I suspect an interferometer is the default tool associated with alignment because optical engineers are familiar with them as optical testing tools, yet they know much less about the tools and methods used to locate centers of curvature, and assembly methods in general.

Sure, you can roughly assemble an optical system, use an interferometer to measure the wavefront error at numerous field angles, and use that information to guide your alignment adjustments. But a basic question remains: Why not just assemble the system according to the design in the first place?

The easiest way to achieve this is to leverage a coordinate measuring machine (CMM). Instead of using the machine strictly as an inspection tool, it can be thought of as a high-precision, long-travel X-Y-Z stage to position an optical center-of-curvature sensor in place of the usual touch probe.

The process is highly systematic. First, the optical sensor is zeroed on a master calibration ball, exactly like a mechanical touch probe is zeroed. Next, tooling ball features on the optical bench or mount are located so the machine establishes the coordinate axes of the mechanical mount.

From there, the optomechanical design specifications indicate the exact locations of each element’s centers of curvature. Elements are assembled one by one, precisely where the design dictates. When the final element is installed, the job is done.

While I have never personally implemented this next approach, an even better method may be to use the CMM in conjunction with a Bessel beam. By utilizing a Bessel beam, you start with a reference axis rather than relying on a single reference point, such as the master ball.

As each optical element is installed, you can directly measure the position and angle of the beam leaving the element with the optical sensor to verify if the beam’s location matches the nominal design.

The advantage is accessibility. You never have to locate a center of curvature that is located behind an element. You are always working upstream of the last element installed.

Using a CMM coupled with a center-of-curvature sensor changes optical alignment from an iterative process to a deterministic method of opto-mechanical assembly. This approach not only simplifies the build but inherently produces a precise record of the installation quality of each element, mapped directly to the optically critical datums that describe their five or six degrees of freedom.