Engineering technology

Transcription

1 Philips tech. Rev. 39, ,1980, No. II 315 Engineering technology H. J. Akkerman When work started on the construction of the cyclotron in 1946 there was a need for engineering support on the site. The postwar shortages meant that mechanical engineers had to manage as best they could withvery few machines; the cyclotron project had one milling machine and a discarded pre-war lathe, and very Research Laboratories in Eindhoven, and the close cooperation that resulted from this situation was to have an important influence later on the development of the workshop at IKO. In the fifties nuclear physics was a flourishing and expanding field of research, and IKO found them- Fig. 1. General view of the workshop. In 35 years it grew from a little local workshop with four craftsmen to become a centre for engineering technology with a staff of 34. few tools - a single calliper gauge, for example. A few heavy timbers were put together to make a work bench. In 1948, thanks to the American Marshall aid plan for Europe, a new lathe was acquired but it was a large British machine not well suited to our work. The most important components of the cyclotron, however, were made in the workshops of Philips H. J. Akkerman is with Philips Research Laboratories, Amsterdam Department, and is head of the Mechanica! Technology Section at IKO. This article arose partly out of discussions with J. J. Arendse (Philips Research Laboratories, A msterdam Department) and G. C. Gerritsen (IKO). selves in increasing need of special engineering techniques so that they could construct better instruments for the work in hand. At that time, mechanicalengineering technology was also in a phase of rapid development and many new opportunities were being offered by the many new materials and processing methods becoming available. As a result the workshop grew into an engineering department with specialized sections for design and drawing, metrology, bonding techniques, vacuum and cryogenic engineering, electroplating and mechanical processes (fig. 1). The department could tackle other problem as well,

2 316 H. J. AKKERMAN Philips tech. Rev. 39, No. II because of the support it had in Eindhoven. This placed it in a position to carry out extensive projects such as the electron accelerators and 'BOL' for IKO, and the ion-implantation equipment for the Philips group at IKO. A few examples will be quoted to give an irnpression of the kind of work that was carried out in the engineering department during the last 35 years. Vacuum components The vacuum system formed an important part of the cyclotron system [11. The pressure in the vacuum chamber of the cyclotron (volume 800 I) had to be less than 10-4 Pa. When the cyclotron was in use there was a strong magnetic field, and the components also had to be able to stand up to the radioactive radiation released. In about 1946, when the cyclotron was being built, vacuum components for such applications were not an everyday item, and were not therefore cornmercially available. This resulted in all kinds of standard designs being developed for vacuum components, including a design of vacuum valve that was unique at the time [21. This valve had a toggle-lever construction enabling the pressure required to close a valve to be created by a relatively small force (fig. 2). A large number of these valves with ori fice diameters between 50 and 350 mm have since been manufactured (fig. 3). Another then novel feature of the valves was that they could be operated either pneumatically or by hand. In the middle fifties pneumatic control systems had come into use in the Netherlands and a system of this type was used for remote control and automation of moving parts of the cyclotron. Pneumatic automation was also used for the vacuum-lock system for the cyclotron target. In this system a target was introduced into the vacuum of the cyclotron and brought out again after bombardment, without altering the state of the vacuum in the chamber of the cyclotron in any way. The automatic system not only made it easier to change the target, it also made it safer. A short vacuum locking time was also important for specimens with a short half-life; the seven successive operations necessary were completed in eight seconds. Cobalt needles In 1969 about forty cobalt needles were manufactured for the former Philips-Roxane company. Each needle contained a small rod of radioactive cobalt. The needles were used in the treatment of cancer. The difficulty was that the needles had to be machined after Fig. 2. Diagram of the vacuum valve developed in Only a relatively small force is necessary to press the valve V vertically down on to the seat S. Fig. 3. Some examples of vacuum valves with orifice diameters between 8 and 3S cm. One of the features of these valves was that they could be operated either by hand or by compressed air. Co Fig. 4. Cobalt needle, diameter 1.1 mm, tilled with radioactive cobalt (Co).

3 Philips tech. Rev. 39, No. II ENGINEERING TECHNOLOGY 317 Fig. 5. Boring bar with six accurately positioned cutting tools. With this device the desired profile for a detector-telescope hole could be machined in a single operation. The profile of the hole can be seen in fig. 3 of the article on BOL, see page 304. they had been filled with the radioactive material. The needle was made from a stainless steel tube with a diameter of 1.1 mm (fig. 4). This was sealed off at one end by a clearly marked stopper with a nylon thread. The radioactive rod was then inserted and the other end was also sealed by a stopper, which was ground to form a three-cornered point. For safety reasons the joints had to be soldered with silver. Various pieces of ancillary equipment were developed so that the filling, soldering and grinding operations could be carried out remotely at a safe distance. Fig. 6. Combination of three rotary tables mounted at right angles la one another for machining a hemisphere on a horizontal boring and milling machine. BOL In lko's 'BOL' instrument for research in nuclear physics [3], 64 detector telescopes had to be accurately positioned. Each telescope had to be pointed at the centre of the sphere to an accuracy of 5 urn. This was accomplished by drilling centring holes to the required accuracy in two hemispherical castings. Each casting was machined on a horizontal boring and milling machine. The workpiece was placed on a structure of three rotating worktables that could move perpendicular to one another. The point of rotation of this structure was set to the centre-line of the machine, using ball and feeler gauges to achieve the required accuracy. For the drilling a boring bar with six cutting tools (fig. 5) was used so that the entire profile of a telescope could be made in the workpiece in a single operation (fig. 6). 'Wire chambers' Two 'wire chambers' have been constructed for the observation of particles in nuclear-physics experiments at [KO. A wire chamber operates in an analogous way to a Geiger-Müller counter, but contains a large number of anode wires, in this case as many as 1200 per chamber. The wires - of tungsten - have a diameter of 0.02 mm and are stretched across a frame Fig. 7. Machining the locating grooves with a numerically controlled machine at the Philips Engineering Workshops in Eindhoven. A large number of grooves 70 urn deep are made in the strip, which is attached to a supporting plate about 2 metres above the ground (on the left in the figure). [I] See for example G. Luijckx, The cyclotron, this issue, page 290. [2] H. J. Akkerman, Fijntechniek 2, 154, 1962 (in Dutch). [3] See for example R. van Dantzig, BOL, this issue, page 302.

4 318 ENGINEERING TECHNOLOGY Philips tech. Rev. 39, No. 11 of dimensions 1220 x 110 mm. Two sheets of polyester good soldered joint this method also ensured that the film with a conductive coating are placed on either wire was located at the bottom of the groove (fig. 8). side of the wires and act as cathodes. A special semi-automatic machine was built for posi- In the construction of the unit it was essential for tioning, tensioning and soldering the large numbers of the wires to be accurately located in relation to the wires (there were about 2400) [4l. cathode films and each other. For a spacing between the wires of 4 mm the tolerance was ± 10 urn. This accuracy was achieved by etching tracks in printedcircuit material, forming grooves in it to a depth of 70 urn and positioning the wires with the help of the grooves. The grooves were made using a numerically controlled milling machine at the Philips Engineering Workshops in Eindhoven. The machine was fitted with a slotting mechanism, with a diamond tool, designed for this application (fig. 7). The wires were attached in position by 'reflow' soldering. In this process a tin layer applied earlier is made fluid again with an electrode that can be set for time, temperature and pressure. Besides creating a w Fig. 8. Soldered joints s of tungsten wires won copper contacts c in a 'wire chamber'; the position of the wires is determined by the grooved strip g. [4] H. J. Akkerman, R. P. Arink and P. H. Thobe, Mikroniek 20, 230, 1980 (i 11 Dutch). /

5 Philips tech. Rev. 39, No A 200 kv ion-implantation system in the FOM Institute for Atomic and Molecular Physics. This Institute is located at Watergraafsmeer, like IKO; the Philips group at IKO is working closely with the Institute on research into ion beams. The 200 kv source is located in the cage that can be seen in the upper left-hand corner. The ion beam generated here can be guided by means of a magnetic switch to anyone of three experimental beam lines: - a bench for ion scattering - Rutherford Back-Scattering - that was installed by Philips (the left-hand beam line, as seen from the high-voltage side), - a bench for surface-structure investigations (the central beam line), - a bench for measuring ion-induced X-ray radiation (the right-hand beam line).