is first inserted and locked in the headstock 3. Since quartz tubing is rarely perfectly straight, it may he desirable at this stage to heat up the tubing close to chuck 3 sufficiently to soften it and true it up by pressing on the free end to the extent necessary. The free end is next locked in the tailstock chuck and heat is then applied successively to both ends just enough to soften the quartz slightly and relieve any residual strain.
The burner is now positioned towards the left side of the tubing and the oxyhydrogen flame turned full on. The burner is held stationary for an interval of time corresponding to the initial vertical rise in curve 26 in FIG. 2 and the tubing gradually necks down as indicated at 32 in FIG. 3, forming the left throat. The burner is then initially displaced fairly rapidly to the right whereupon displacement of the tailstock to the left at a relatively constant rate begins, as shown by curve 25. As the tailstock moves to the left, it causes the plastic quartz to be upset and to increase in diameter as indicated at 33 in FIG. 4. Meanwhile surface tension in the quartz is tending to neck it down and reduce the internal diameter. The result is a gradual thickening of the walls which takes place smoothly and evenly so long as the motions of the tailstock and burner are smooth and regular. The heatsoftened zone in which the quartz thickenes is always behind the flame, that is it is the region over which the flame has just previously passed. Thus in this example the movement of the burner is to the right and the build-up of quartz occurs to the left of the burner, as may be seen in FIGS. 4 and 5. At the stage of FIG. 5, the burner is held substantially stationary on the right flank of the quartz bulge for an interval of time after movement of the tailstock had ceased. This is in order to collapse the right flank slightly and make it correspond in curvature to the left flank of the bulge. Thereafter the burner is moved rapidly over and again held stationary for an interval of time to form the constriction or right throat as indicated at 35 in FIG. 6. The burner is now shut down to stand-by, the completed bulb is disengaged from the chucks, and any excess tubing cut off to leave the necks 31a, 31b at the desired length. It may be preferable to use tubing precut to size so that the bulb necks come out the right length.
As an example of a bulb formed according to the invention as described above, a piece of quartz tubing 4" long and having a 4 millimeter bore and 6 millimeter outer diameter was used. In processing this tubing into a bulb as illustrated in FIGS. 6 or 7, the shrinkage in length of the tubing was P/s inch. The finished bulb had a chamber length of approximately %6 inch and a neck-to-neck distance of approximately % inch.
A characteristic feature of the one-piece quartz bulbs or envelopes according to the invention is the improvement effected in the quality of the quartz of the bulb proper. Quartz formed by drawing normally contains fine lengthwise extending cavities caused by trapped air or gases and known as "streamers." These are undesirable because they may hasten devitrification and development of a leak. In the bulbs according to the invention which are the result of build-up of quartz by simultaneous necking and upset, the cavities are converted into bubbles. This reduces the possibility of development of leaks and also means improved optical properties.
In envelopes formed by the union of parts by melting and fusing, the joints are sources of weakness where devitrification tends to originate and develop more rapidly. This is avoided in the one-piece envelopes according to the invention where the necks merge through constricted throats into the central bulb portion without joints or discontinuities.
Instead of starting the flame at the left side and displacing it towards the right as illustrated in FIGS. 3 to 6, the flame may be started on the right side and displaced towards the left. In such case, the flame must move the extent of tube shrinkage plus the finished bulb
length. This means in effect that the burner movement represented by curve 26 in FIG. 2 must be reversed and added to or superimposed on the tailstock movement. With the burner moving towards the left, the quartz 5 again builds up in the zone just previously passed over by the flame but it is now located to the right of the burner.
In forming bulbs by the method of the invention, it is important to have only a short section of quartz tub
10 ing heat-softened at one time. If heat is applied at two points simultaneously, the column in-between may go off axis; similarly, if a long section is heated, it may bend or go off axis. By heat-softening only a short section at one time, the quartz is caused to bulge out symmetrically
15 as the tailstock is moved in and the desired bulb formation with simultaneous wall thickening takes place.
The preferable practice is to neck the tubing at one end of the proposed bulb, form the bulb, and then neck the tubing at the other end. However, this sequence is
20 not essential. One may neck both ends and then run the burner oevr the length of quartz tubing in-between in order to thicken the walls and form the bulb. Alternatively, the bulb may be formed and then the burner moved first to one end and then to the other in order
25 to neck down the tubing and form the throats.
Bulb 37 in FIG. 7 -which is relatively thick walled and has an elongated discharge chamber illustrates generally the shape or cross section most readily formed by the process of the invention. I have found this shape par
30 ticularly suitable for use in metal iodide arc lamps. Where a more spherical bulb is desired and a lesser wall thickness is sufficient, a bulb configuration such as illustrated at 38 in FIG. 8 may be obtained by blowing out the bulb of FIG. 7. This is done by heat softening the
35 bulb in the glass lathe with an oxyhydrogen flame and applying a puff of air under pressure to one neck while blocking the other. When the bulb is blown out in this fashion, the wall thickness is reduced. Such bulbs are suitable for use in short arc gap high pressure mercury
40 or xenon discharge lamps.
The bulb -shape illustrated in FIG. 7, which may be referred to as the normal shape, tends to radiate equally in all directions except for the obstructions presented by the necks in the polar zones along the bulb axis. For
45 some applications, an increase in radiant flux density at right angles to the axis or in the equatorial zone is desirable. In accordance with my invention, this is readily achieved by thickening the bulb walls in a region intermediate the neck-bulb junctures. A lens effect is thereby
50 obtained which increases the flux density in a direction other than along the axis of the lamp. Thickening of the bulb wall in a selected region is obtained by increasing the degree of upset and this in general requires a greater rate of displacement of the tailstock for a given
55 rate of displacement of the burner. FIGS. 9 to 11 illustrate various bulb configurations according to this aspect of the invention.
In FIG. 9, the wall thickness- of bulb 39 is increased in the equatorial zone in a symmetrical pattern. This configuration is particularly useful where a circular distribution of light is desired, as where the lamp is to be mounted with its axis vertical to give a light beam concentrated about the horizon all around.
C5 The bulb configuration illustrated at 40 in FIG. 10 causes the flux density to increase towards the right and to diminish towards the left, -that is the flux is concentrated in a region between the equator and the right pole. This shape is well adapted to use in parabolic reflectors where
70 the lamp is mounted with its axis on the polar axis of the reflector. This bulb cross section may be achieved by increasing the rate of upset in the first stage of bulb formation to give the steep left edge of the bulb while reducing it in the later stage to give the gradually slop
75 ing right side of the bulb.