Many of the electromechanical assembly features of the IMU are repeated in the tracker assembly. Thus, each end of each axis fits into a pair of preloaded ball bearings which, in consort with the support-structure members, feature an overall structural stiffness capable of maintaining relative positions within microinch tolerances. To enhance this capability, the main support housing or case for the whole IMU and tracker assembly (fig. 19) is to be made of 6061 aluminum, which, for its low weight (density), offers good thermal conductivity, high structural and mechanical stability, and high resistance to corrosion. The enclosing jacket or housing for just the tracker unit itself (fig. 19) is to be made of a single piece of heat-treated Ti-6A1-4x titanium alloy, which, in addition to the above characteristics, has the added advantage of a thermal coefficient of expansion closely matched to that of the optical glass. Mounted as appropriate in gimbal or case at one end of each of the abovementioned two axes is a torque motor to drive the tracker around that axis (fig. 18). Each of these two motors has a torque capability of 60 oz-in. (0.42 N·m). Mounted as appropriate in gimbal or case at the opposite end (from the torque motor) of each of the foregoing two axes is a 1:36-speed resolver (fig. 18). Each measures the angular displacement of the gimbal, and the readings are relayed to the onboard computer. The net effect of these cited electromechanical features is to assure that the overall tracker pointing accuracy is not degraded from the design accuracy of the resolver, namely, about 4 arc-seconds. A few words are appropriate to interpret the color coding used in figure 19. Individual colors are listed and explained as follows, proceeding generally from outermost to innermost physical features: of the aircraft cabin. This specification is met by constructing a floor well of appropriate dimensions, mounting the tracker as deep therein as possible, and adopting a coaxial and folded optical lens system. Layout of the tracker optics is shown in figure 20. From this layout, the laser-beam (green band, fig. 20) path is traced readily from source through transmitter lens assembly, to deflecting mirror, and to deflecting prism centered and cemented on the back side of the objective lens. At this point, the laser beam is outbound toward a retroreflector target secured on a ground-control point. After reflection, the laser beam returns inbound (light-blue band, fig. 20) through the same objective lens, to the deflecting alinement mirror, and through the beam-splitter assembly (medium-blue band, fig. 20) for delivery to a quadrant-photodiode-receiver assembly and a range-receiver assembly; the former assembly receives 90 percent of the return signal, and the latter, 10 percent. The 4.5-in. (114-mm) diameter objective lens [11.902-in. (302.3-mm) focal length] was the largest that could be used because of the 24-in. (0.61-m) diameter clear limit originally placed on the aircraft cabin-floor well. To inhibit moisture condensation in the tracker optics, two silica-gel capsules occupy a cavity designed into the tracker unit housing. The capsules are packaged in a lint-free nondusting material, and, by choice, the gel does not feature any dye to indicate moisture content, thereby forestalling any opportunity for adversely affecting a lens coating. The variable density optical filter, shown in the path of the outgoing laser beam, is operated by an automatic gain control loop which monitors the intensity of the return laser beam as the sum of the four quadrant pulse magnitudes. Thus, as the return signal weakens or strengthens, owing, for example, to changes in attenuation as a function of range to the retroreflector, the servo-drive motor rotates the continuously variable filter just enough to strengthen or weaken the outgoing beam the needed amount. The intent is to maintain a constant sum for the four quadrant return-signal pulses. The range receiver assembly features a silicon avalanche photodiode detector and transimpedance preamplifier. It senses the return-signal pulse from which range to the retroreflector target is ultimately determined. The laser beam is generated by a pulsed gallium arsenide (GaAs) transmitter that is similar to the one designed for the laser profiler. Pulses are produced at a FIGURE 19.-The inertial measurement unit, laser tracker, gimbal-support structure, and housing. rate of 3.2 kHz in the near infrared at a wavelength of 904 nanometers (nm), with 25 watts (W) peak power, subnanosecond rise times, and widths of 40 nanoseconds. These large widths (four times greater than the profiler pulses) assure that more return-signal energy will be fed to the quadrant photodiode assembly (fig. 20) and to the tracking electronics. The laser beam meets Federal safety standards in the following man ner: 1. In the laboratory.-Eye safe for all distances greater than 18 in. (0.46 m) when using a suitable attenuation filter and limiting exposure time to a 2-hour maximum. 2. In flight.-Eye safe for all distances greater than 892 ft (304 m). This type of laser beam permits range determinations well within the stipulated precision of ±0.5 ft (0.15 m) over distances as great as 6,000 ft (1.8 km), which is the maximum planned operating slant range to a retroreflector target. Divergence of the laser beam is 6 to 7 milliradians (mrad) and the fields of view of the quadrant-photodiode and range-receiver assemblies, respectively, are 8.3 and 2.6 mrad. The quadrant photodiode receiver assembly senses the return signal as four independent pulse magnitudes related to the sectors above, below, left, and right of the quadrant dividers. If the tracker boresight catches a retroreflector target off center, the target image will be unevenly distributed in the four quadrants, and the four pulse magnitudes proportional to the sighting error in each quadrant, and the corresponding control currents drive the two torque motors (see fig. 18) to keep the target image centered in the tracker field of view. The discussion given in the section, "Laser Profiler," refers to a key mathematical expression termed the "range equation" which is equally important in this section in arriving at suitable design choices for the laser transceiver. Although the following discussion of the range equation closely parallels that given in the "Laser Profiler" section, there are significant differences that relate primarily to the laser beam ranging on a specially designed cooperative retroreflector target instead of the uncooperative land-surface target. Thus, for the laser tracker, the range equation has the form: A2 = K2πR2, where K2 is simply the appropriate trigonometric function of a used with R to compute the footprint radius. To develop a similar expression for A4, the area of the circular footprint of the returning laser beam at the receiver, consider first that the retroreflector is designed to have a high reflection coefficient and a shaped surface such that virtually all the laser beam that is intercepted is reflected straight back along the line-of-sight path to the source. In effect, therefore, the retroreflector becomes a new point source for the laser beam, with its own constant divergence angle, a'. (Ideally, a' should be nearly equal to, but no less than, a.) Thus, the radius of the circular footprint at the receiver is a function of a' and R, and the area, A4, is determined from the relation, σ = attenuation coefficient, atmospheric path, product of K2 and K4. and R = range. The interrelations among the foregoing parameters are ilustrated schematically in the three parts of figure 21. These parts, studied in sequence from left to right, show (A) the significant geometric features of the laser light path from transmitter (source) to retroreflector target and back to receiver, (B) the critical points in the flow of light energy from the lasing source to the target and return, and (C) the mathematical description of the principal losses in energy flow which allows a rigorous statement to be written (eq 3) that shows the net change in energy level between laser source and receiver. Obviously figure 21A is schematic in the sense that, in the real tracker instrument, the field of view is coextensive (not side by side) with the field irradiated. To arrive at the particular form of the range equation given in figure 21, note from the geometry shown in part (A) that the radius of the circular footprint of the laser beam at the retroreflector is a function of the constant beam divergence angle, a, and the range, R. Thus, the footprint area, A2, is determined from the relation, The tracker may be operated in either a search mode or a track mode. In the search mode, when the tracker must search until it acquires a retroreflector target, the tracker gimbals are positioned automatically by computer program in much the same manner as described in the last paragraph in the section titled "Electronics." Now, however, the appropriately processed gimbal angle data are sent to the tracker gimbal servoamplifiers to drive the tracker torque motors. Included in the computer program for this pointing mode is continuing iterative calculation of the commanded tracker gimbal angles to maintain the line of sight toward the expected position of the retroreflector target. This takes into account the known target location, as well as the present aircraft position, attitude, and velocity, and then superimposes a tracker spiral scan motion that slowly opens up, in a step-by-step programmed way, around the initial, most probable target direction. At the maximum intended flight operating altitude of 3,000 ft (920 m), the spiral scan will sweep the tracker footprint over a terrain area that totals about 4.3 × 104 ft2 (4.0 × 103 m2) in 20 seconds. Normally, this is more than sufficient to find the retroreflector target. However, if the target is not found and is determined to be |