Lowering the Pendulum down onto the Ball

Working from the top of the pendulum, one must lift the entire thing by hand and lower it down onto the flat, either with the ball on the flat, or already fixed in the ball support somehow. By the way, I counsel against the second option above. There is a great temptation to use a bit of Blutack or a similar clay-like adhesive for securing the ball into the ball support, so that it can be reliably lifted up together with the ring. But, as I have found, the Blutack will accumulate in the recess below the ball, and will eventually start to squidge out at the most inappropriate moment, even falling on the flat and vitiating a whole string of observations. Moreover, bits of Blutack adhere to the ball, and must be cleaned off if, in order to use a new spot on the ball, you rotate it in its holder. Therefore I advise you to eschew the Blutack! Simply set the ball at the desired spot upon the flat, and then lower the ring down onto it.

The problem with doing this is that the pendulum complete with bob is quite heavy - more than 15 kg. And the ball and flat, although they are made of tool steel, have a very small contact surface - theoretically a point. If there is any clonking at all - even slight clonking - when lowering the ring down onto the ball, the ball and/or the flat may be damaged at this contact point. Even if this hasn't actually happened, you don't know that it hasn't; and therefore the resulting observations are basically suspect. Thus it is absolutely required to lower the ring very very gently indeed. This can scarcely be done by hand even with an assistant.

Therefore, using simple 'Ronstan' sailing dinghy hardware, we constructed a sling. You can see it in the upper left of this picture:

This is a full-on photo of the hoist. Basically it consists of two three-pulley blocks and a length of line run round them, thus giving a 6:1 mechanical advantage. More important, however, is the friction in the system, which means that, when lowering the ring, one can very easily feed out the line millimeter by millimeter with one hand, in complete control. This system works very well.

Here is the ball resting on the flat, ready for the ring to be lowered down onto it with the hoist. Since at this stage the flat has been leveled and is accurately horizontal, you will find that the problem of the ball rolling off does not arise - Van de Waals forces are quite enough to stop it moving.

You can select the position of the ball on the flat at this stage, so as to use a new and untouched portion of the flat. You will find that the flat does inevitably suffer some damage: an 8-hour experimental session, say, leaves a little zigzag mark on the flat as the point of contact of the ball moves to and fro, and as it slowly migrates to one side (typically this zigzag mark is about 2 mm square). No doubt the ball also suffers some wear. [With our Sarawak experiments, we used a ball coated with tungsten carbide, whereas the flats were not coated but were simple tool steel.] Therefore both the bottom point of the ball, and the position on the flat, should be periodically shifted. Allais used a new ball for every experiment, but basically that was in order to average out the errors for each ball, because they weren't accurately made in the first place. (He says that he changed the flat every week; that's all he says about it.) I think it is better: to use a very accurate ball and flat; to take extremely good care of them - no skidding or clonking; and to change them periodically rather than constantly.

There is a very important point here: A PARACONICAL PENDULUM HAS NO CENTER! That is, the bottom dead center point of the pendulum - the point at which the pointer points when the pendulum is absolutely stationary - of course shifts precisely according to the position of the ball upon the flat. Since the flat is 35 mm long and 25 mm wide, this shifting can be considerable. For reasons of cost and convenience, it is desirable to be able to use - use up - the entire surface of each flat, which means that we must be ready for the bottom dead center point to be positioned anywhere within about three centimeters along the line of swing of the pendulum and within about two centimeters across it. The sensing arrangements must allow for this, and so must the launch arrangements - see later.

The last photo shows the ring lowered down onto the ball. It was impracticable to take several shots during the lowering process, because of the lack of space for photography. It is quite inconvenient to be working up a ladder, near to the ceiling.

In the next version of the apparatus, for tying to the ring, rather than using a loop of line round the ring clipped to the lower hoisting block as shown, it would be better for a loop to be made on the brass plug at the top of the ring, so that one could simply clip the block to that loop, in one operation.

Back to the description of operating the English paraconical pendulums in Sarawak.

Back to the website main page.