1
Introduction

Despite our best efforts and intentions, Mr. Murphy-a rather jovial fellow who happens to be a bit sensitive to these things, but who also has a fondness for Schadenfreude-will remind anyone developing optical hardware of his inescapable Law, whether they like it or not. For those who are not prepared, the reminder will be unexpected, and schedules, budgets, and careers will eventually be broken; for those who are prepared, his reminder will be much less painful-and soon forgotten, as the customer's happiness at receiving working hardware reminds us why we are engineers in the first place.

Fortunately, Pasteur's antidote to Murphy's near-deadly "snakebites"-that chance favors the prepared mind-is our opportunity to remove most of the fatalism from engineering. The type of preparation that this book provides goes under the name of optomechanical engineering-an area of optical systems engineering where "the rubber meets the road," and thus has the highest visibility to managers and customers alike [1].

It is sometimes said that optomechanical design is a relatively new field, but the truth is a bit more complicated. Not surprisingly, it is as old as optical engineering itself, a field that dates back to at least the early 1600s when the Dutch were assembling telescopes.1 Notable contributions by people who are otherwise known as great scientists-but should also be recognized as optomechanical engineers-include:

• Galileo-While he did not invent the telescope, a practical telescope architecture using refractive (lens) components is named after him.
• Isaac Newton-To develop his theories on the nature of light, he invented the first practical telescope using reflective (mirror) components.
• James Clerk Maxwell-In addition to his brilliant discovery of the electromagnetic nature of light, he developed a structural theory of trusses and experimented with photoelasticity and kinematic mounts.
• Joseph Fourier-He made many contributions in the areas of heat transfer and thermal design, including the discovery of Fourier's law of conduction and the Fourier transform for analyzing vibrations, heat-transfer problems, and more recently, electrical circuits.
• William Thomson (aka Lord Kelvin)-He is best known for his work in thermodynamics, developing the Kelvin temperature scale. In addition, he deserves to be recognized for his work in kinematics, inventing the Kelvin kinematic mount.

In short, these were people who were trying to solve difficult physics problems but were unable to do so until they first solved the instrumentation problems of how to make an apparatus stiff, stable, repeatable, and so on, that is, solve the state-of-the-art optomechanical problems.

More recently, we are still trying to solve difficult problems including the following applications and even quantum optomechanics [2]:

• Aerospace: infrared cameras, spectrometers, high-power laser systems, etc.
• Biomedical: fluorescence microscopy, flow cytometry, DNA sequencing, etc.
• Manufacturing: machine vision, laser cutting and drilling, etc.

In the following sections of this chapter, we first take a look at what a typical optomechanical system might consist of (Section 1.1), the skills needed to engineer such a system (Section 1.2), and the mindset needed to do this efficiently (Section 1.3).

1.1 Optomechanical Systems

If we buy an optomechanical system today, what would we expect to get for our money? Figure 1.1 illustrates a complex biomedical product known as a swept-field confocal microscope-a microscope with some unique features that allow it to image over a wide field-of-view with high resolution [3, 4].

Figure 1.1 A complex optical system such as a swept-field confocal microscope requires a large number of optomechanical components packaged into a small volume.

Credit: LOCI and Laser Focus World, Vol. 46, No. 3 (Mar. 2010) [3].

Given its complexity, the designers of this microscope had to struggle with many issues that are not obvious to the eye, including:

• Assembly and alignment-Can the optical components all be assembled in a small package and maintain critical alignments such as the "Focus to CCD" distance (for which an adjustment is provided)?
• Structural design-Are the overall structure and the optical submounts stiff enough to keep things in alignment due to self-weight or shock loading?
• Vibration design-Have scan mirror vibrations been isolated from the optics and prevented from causing the optics to move ever so slightly (but more than is acceptable)?
• Thermal design-With components such as the piezos and galvanometers dissipating heat in such a small volume, is there even enough surface area to transfer this heat without the external box temperature getting excessively hot?
• Kinematic design-If the microscope needs to be repaired, is there a way to remove critical optics that allows them to be replaced in the field, without a major realignment at the factory?
• System design-Have all the interactions between the elements been considered, for example, the effects of heat on the optics?

Before getting to these topics, it is important to first understand that common to all optomechanical systems is the use of electromagnetic waves known as "light" (Fig. 1.2). This refers to the wavelength of these waves-on the order of 1 µm, but extending down to 0.1 µm or so and up to ~30 µm-and distinct from "radio" waves, with much longer wavelengths. Controlling the curvature and direction of optical wavefronts with lenses and mirrors is what allows us to create optical images, or determine the wavelengths present, or measure the power transported by a wavefront. Keeping mechanical parts aligned and stable to <1 µm-an extremely small dimension equal to ~40 micro-inches (or 0.04 milli-inches, often pronounced in abbreviated form as "mils")-is one of the many challenges of optomechanical engineering.

Figure 1.2 Optical electromagnetic wavelengths ("light") can be divided into infrared, visible, and ultraviolet bands.

Credit: NASA (www.nasa.gov).

1.2 Optomechanical Engineering

So an optomechanical system has a few lenses and a detector-how difficult can this be to design? As we have just mentioned, the small size of an optical wavelength-and thus the mechanical accuracy required-is one of the difficulties. Paul Yoder and Dan Vukobratovich have published the majority of recent books showing us many of the implementation difficulties, and have many useful details and hints on building hardware [5-10]. Even "just" a packaging job is not straightforward, for example, as many laser jocks have discovered when trying to convert their laboratory hardware into a commercial product (Fig. 1.3).

Figure 1.3 The transition from laboratory to marketplace is critically dependent on the skills of the optomechanical engineers.

Photo credits: Permission to use granted by Newport Corporation; all rights reserved.

An example of one of the steps required for the transition from lab to marketplace-or even optical designer's desk to working prototype-is shown in Figure 1.4. Here, a lens designer has determined that a three-lens system is required to meet the customer's needs (or "requirements"). The lens designer's deliverable to the optomechanical engineer is an optical prescription from lens design software such as Zemax or Code V, consisting of lens geometries (surface radii and diameter), materials, and spacings between the lenses.

Figure 1.4 An example illustrating the steps required to move from an optical design prescription to a complete optomechanical design.

Credit: G. E. Jones, "High Performance Lens Mounting," Proc. SPIE, Vol. 73, pp. 9-17 (1975).

The lens designer must also provide a tolerance analysis showing how sensitive each lens is to misalignments such as tilt, centration, and spacing errors-and how much of each is allowed. In addition, the lens designer must deliver a fabrication analysis for each lens, specifying the tolerances on the lens surfaces, refractive indices of the lenses, and so on-a topic we will look at in detail in Chapter 3. Ideally, the optical prescription, alignment analysis, and fabrication analysis must be developed in coordination with the optomechanical engineer; in practice, this is not always done, with the consequences illustrated with an example in Section 1.3.

Given these inputs from the lens designer, what the optomechanical engineer must then determine is as follows: are the fabrication and alignment tolerances (i.e., allowable variations) feasible, given the manufacturing, technical, environmental, cost, and schedule constraints? That is, is it possible to take the lens designer's prescription and convert it into manufacturable hardware such as that shown in the bottom...