The Rails

1.     This is how we first started with our space and time-resolved spark emission spectroscopy experiments in 1969. The instrumentation was designed and assembled a piece at a time. Parts were fitted together after the fact. Sometimes, like here, delicate optics were up in the air more than a foot about a supporting bench, subjecting delicate optics to periodic surprise misalignment and vibrational drift. Once an experiment was set up, it would stay that way often for a couple of years, even though the work might suggest new approaches. Graduate students seldom could exchange equipment, which led to competitive isolation thinly disguised as "independence". Not that we were the only ones. Our instrument setups were more robust than many other spectroscopists used. But, a new approach was clearly needed. An entirely new premise was needed.
2.     The new premise preceded the equipment to make it possible. It was that graduate students should be able to interact on a spectroscopy experiment, either within a current generation or between generations, if their equipment was interchangeable, essentially regardless of its size or physical nature. To allow this to happen, all experiments would be built along a commonly dimensioned reference axis, called the "ERL". To this end, a set of rails and riders were made. These were sufficiently robust that an entire spectrometer could be carried on them, as well as small optical components. The key element was the vee-and-flat rider itself. This was the element that was mobile between experiment beds, each of which would have a commonly dimensioned set of round rails. All components of the experiment would be firmly supported using A-frame, kinematic methods at a common height above the surface of the flat rider.

3.      The rider (above) was designed to support optical devices of considerable weight at an experimental reference line ("optical axis") 10.5000 inches above its top (precision ground) surface. This dimension was taken from an existing Baird-Atomic 3.0 meter Eagle spectrograph, which at the time used the largest and heaviest arc/spark stand as a "mobile" external accessory.

4.      Any rider was positioned by a straight and true reference rail, and leveled by an auxiliary rail, both of which were along a common center line. The dimensions are shown above. A picture of a finished aluminum rider on two ground steel rods is shown at the left. The rail center lines were 15.000 " apart. Note the 1/4 x 20 hole pattern drilled into the top of the rider.

Eventually, a hundred or so riders were machined to high common tolerances. Some were machined from aluminum castings, while others were made from cast iron blanks. They were made the foundation of all instrumentation.

5.      The key to the system was that the mechanical reference and auxiliary rails would be duplicated in all of the labs. The reference rail would be made straight and true, while the top of the auxiliary rail would be leveled to it. This was to be done no matter the size of the apparatus to be mounted on the rails, or the size of the room in which they were installed. Then, any piece of equipment that was machined to fit on the top of one or two riders could be moved between labs.

If a student were in the middle of an experiment, and wanted to do another one on a different apparatus, she or he could dismantle their light source or other instrument part and move it physically to another lab in a matter of days, rather than months. It would make collaboration between students both possible and practical.

6.      Alignment of components along the 10.5" ERL ("optical axis") was done with HeNe lasers (above) and mechanical height gauges. Ground "bricks" were used to raise components above the riders (right, top), and underslung Vees were used to drop them below the rails (right, bottom).

7.      The rails and rider sets were mounted on "beds". These were of various designs, all depending on when they were made (simplest first) and what kind of experiments were initially envisioned. Above is the mechanical drawing that was used to make our first bed. The finished product is shown at the right. Note the quarter-wave spark source penetrating down between the rails, showing their superiority to a flat table or the like. The bed was made from a series of aluminum castings which were machined on their tops to hold adjustable vee blocks for straightening and leveling the rails. The whole assembled was lag bolted into the floor. It all worked marvelously well and was one of our better beds.

8.      An idea of the physical size of an "optical component" that can be carried by the rails and riders is shown above. In this case, the component is a complete two-meter Czerny-Turner monochromator manufactured by the Jarrel-Ash Company. The monochromator takes up the back section of the rails. Admittedly, this is not the kind of component that is going to be routinely moved between beds during a single doctoral thesis project. Still, it can be moved, and was in fact moved when the lab was turned over to Professor John Wright in 1982.

9.      Another bed design was based on bolted-together tetragonal pyramids. The base structure could be made of to essentially any length by bolting together more sections. The top of the bed was wide enough to hold three sets of rails, so that optical paths could be bent around the analyzing spectrometer. The bed shown at the left is a short pyramid section as it was first built. Only two rails are mounted on it using adjustable vee blocks to straighten and level the rails. The one shown below has several sections of pyramids bolted together, and eventually carried three sets of coplanar rails and one orthogonal set. It was so long that it had to be air suspended for torsional rigidity and vibration isolation.

10.       When an additional set of "outrider" rails are added to the above left bed, a complete experiment can be assembled on it in a few months. One such experiment is shown at the right. A 1.0 meter Czerny-Turner monochromator is at the back left (grey colored), along with a locally-built nitrogen laser next to it on the right in the black box.

The N2 laser radiation is spatially filtered and then wrapped around with three plane mirrors to focus into the spark gap, whose radiation then is time swept across the slit of the monochromator.

When this experiment was finished, the laser was taken elsewhere to be used in another experiment. In all, some four large experiments were conducted, one interactively, on this bed over a period of one Ph.D. thesis, proving the concept of a common optical axis to be valid.
11.      In addition to traditional optical components, entire arc and spark stands were carried on the rails with their discharges precisely located on the optical axis. The arc stand at the left was designed for rapid traversal of the bilateral, water-cooled electrode holders under computer control. The electrode jaws were run vertically on Thompson ball screws by high-speed, high-torque stepper motors. They were individually counter weighted to offset inertial drag.

Designed by James B. Peters and machined by Robert Lang and Robert Schmelzer, this arc stand was a true work of high-tech art. It worked like a dream. It was moved between many of the 8 beds carrying the optical rails. It now is in the lab of Dr. John Olesik at The Ohio State University department of geology.

12.      Since most of our work focussed on spark discharges, there were several kinds of spark sources and stands mounted on the rails. At the right is a traditional, bilateral-jaw spark stand, using non-computer-controlled electrode positioning. It was moved around many, many times between all of the beds. A particularly interesting spark source is shown mounted above and in the two pictures below. It is a quarter- wave transmission line source. Below right is shown the complete electrode rotation system of computer-controlled stepper motors, designed and built by Dr. Tsutomu Araki. Above and below left are views of the source mounted with its discharge arranged horizontally along the bed optical axis.

The beautiful Araki quarter-wave spark source on the blue rail.

13.      Possibly our most fascinating optical component was the 10x6 inch grating shown above. This huge optic was mounted on a rider-based assembly. It used precision ball-bearings and a large helical drive gear that allowed miniscule changes in its angular position relative to its feed mirrors. It was used in many experiments and was moved between two beds during its most active service.

14.      Large and small components that were not in active use at any particular time were stored on wooden wall mounts in a controlled environment room. They were covered with gauze so that they could breathe, but not be swept with dust from flowing air. They could be stored here for a year or two, untended and without deterioration.

15.      It worked. The rail designs made it to several other labs (that I know of) in various forms, with ones at the Ohio State University, Brigham Young University, University of Illinois, Urbana, Wayne State University and the University of Alberta. They also became part of a larger laser lab at the University of Wisconsin after I moved to St. Olaf in 1982. These pictures may encourage (or not!) others to try the concept of "interchangeability for interaction" in those graduate environments where interaction between members of a research group is desired.