|
In 1971 I was promoted to the staff of Grubbs (in those days there were two classes of worker, the works employee and the staff employee) Staff members were mostly engineers, physicists and specialists, so promotion was a compliment. During the next nine years I worked in a variety of departments, The research labs where new analytical instruments were developed, the drawing office and the telescope engineering department.
|
|
The first car I owned was a Sunbeam Rapier. With twin Zenith carbs it could just reach 100 MPH |
|
The forward 16” triple turrets of the Battleship New Jersey (BB62) Guns like these can hurl shells weighing 2 tons over twenty miles! |
|
Moving to the staff side of the works was a great opportunity to work with some fine engineers. Grubbs was organised into two main divisions: The instrument section which produced spectrometers, gas analysers and other analytical instruments and the Telescope section which produced astronomical telescopes of all sizes for some of the great astronomy sites all over the world. A third section, the Optical works supplied optics for both sectors.
My first position was with the instrument section where I worked in the research and design department. This was a self contained section where new prototypes were designed, made and tested before passing to production. At this time Grubbs were trying to secure their place in the marketplace under stiff competition from rival companies like Perkin Elmer and Nicolet. Then, as today, the only way to stay in business was to develop new instruments which were superior to the competition. I recall three which caused many headaches. The first of these was a flameproof gas analyser which was intended for the mining industry. The instrument had to be completely sealed so that there was no chance of a spark causing an explosion in the mine. I think we succeeded but the mine safety board were never fully satisfied so the instrument did not pass to production.
The second instrument was a new spectrometer called the Wavelength Master. The idea was to have an instrument which covered so wide a wavelength range and with so high a resolution that the competition would be left floundering. Unfortunately, the design was so ambitious that we could never get the instrument to meet its specification. I recall working all day and through the night until the next morning because we were up against a deadline. This proved to be in vain and the Wavelength Master joined the other special projects on the scapheap! The third instrument was a Raman Spectrometer (for those technical types see here RAMAN ). This was a very ambitious program as Raman techniques were still in their infancy and the machine used the new-fangled stepper motors. The most memorable incident concerned with this machine occurred when the leading engineer, a rather strange person called Hugh decided that the case needed to be stiffer. He hit upon the idea of glueing the bottom plate to the shell to form a very stiff box. This was not a bad idea but, unfortunately, Hugh took it upon himself to carry out the work. When we arrived the next morning we were greeted with the amusing sight of Hugh trying to lever the composite off the bench where the cement had anchored it!
In 1978 I moved again to the telescope engineering department. This was quite a change of scene, instead of working with bench-top instruments I was working on structures which weighed several tons. The art of making a good telescope structure was to have a stiff and rigid structure which could support the mirror systems in a way which introduced no stress to the optics yet the whole structure had to be moved in two axes smoothly and with great precision. The two critical elements were the guidance system and the primary mirror support. I was fortunate enough to work on both of these aspects.
The Grubb guidance system was a small telescope which was attached to the main telescope and aligned to point in the same direction (sounds easy but when the target is a point of light some light-years distant it becomes a tricky task!). Inside the guider a bi-prism separated the star image into four sections. Each section of the image was detected in turn, sampled and compared to each other section. The telescope was moved constantly to keep the amount of light in each section equal. That’s all there was to it! The problem was that the electronics were fairly primitive (by today's standards) and not really up to the task. An electronics engineer called Dave Atherfold and myself spent many weeks in the bowels of the Edinburgh Royal Astronomy trying to get the guidance system to function correctly. The location was chosen because the basement was built into the granite bed rock so was stable, cool and very dark. The problem, as I recall, was that the photomultiplier tube was working near its limit and thermal noise was interfering with the signal. Eventually we got it to work without having to buy a house in Edinburgh!
Another interesting saga was the primary mirror support controller. The mirror which was 98 inches in diameter sat on a hydraulic support consisting of three concentric rings of support pads all interlinked to evenly support the mirror’s weight of several tons. The problem with this kind of support is that, as the telescope rotates from the vertical to near the horizontal, the force exerted by the supports has to decrease in proportion. As the telescope nears the horizontal the radial supports have to bear more load and the axial supports bear less. (Otherwise the mirror would be displaced longitudinally and the telescope would defocus. The solution proposed was, essentially a cylinder containing a weighted piston which was supported by air pressure acting on a diaphragm around the piston. The location of the piston in the cylinder was monitored with micro switches. When the cylinder was vertical the piston exerted maximum force on the diaphragm and the pressure supporting it had to be increased. As the cylinder rotated towards the horizontal (with the telescope) the piston exerted less pressure on the diaphragm, the air pressure was reduced to compensate. The pressure of the hydraulic support system was slaved to the air pressure Here again we had problems, friction being the main enemy. We experimented with tiny vibrators to keep the piston from sticking in its bearings but, eventually, it took a radical redesign to get the device to work.
This subject brings me neatly to another Grubb innovation, the mercury girdle. Mirrors, in a telescope, are supported underneath and around their periphery. The edge supports consist of pivoted weights allowing gravity to support the mirror in proportion to the angle it makes with the vertical. Someone had the bright idea of surrounding the mirror with a girdle (much like an inner tube in a car tyre) filled with mercury. In theory this would provide a near perfect support but, when a leak developed (as it often did) you had a couple of gallons of toxic mercury covering the floor! The obvious dangers (and some technical difficulties) led to the abandoning of the project.
|
