Precision mounting a square peg in a (probably) round hole.

[zondar]

So the square peg or knife edge will be made of brass as well ? I would wonder how long the knife edge will last

No, the knives must be out of a hard metal or else they will wear too quickly. Wear on the knives would change the center of suspension (the point at which the pendulum is hung) and increase friction, both reducing accuracy. The plates on which the knives rest must be even harder, or the knives will quickly cut grooves in it, changing the above too.

Back in the early 1800's, when scientists were first having success trying to measure local gravity, "wootz" steel was used for the knives, and agate plates for the rests.

I bought two 1/8" tool steel blanks for the knives. I haven't decided what the plates should be made out of yet.

However, I'm starting to reconsider the form of the pendulum. I may go back to an earlier design in which the main elements (weights, knives, etc.) are all adjustable. In this form, the knives would be mounted on a piece of metal that can slide along the main shaft and be tightened down along it at arbitrary locations. That means splitting the knives in two so each half can fit around the shaft. This raises additional difficulties such as getting the two knives to be perfectly in-line and parallel. But at least the parts to be worked on can fit on a tiny lathe.

 

[compact8]

No, the knives must be out of a hard metal or else they will wear too quickly. Wear on the knives would change the center of suspension (the point at which the pendulum is hung) and increase friction, both reducing accuracy. The plates on which the knives rest must be even harder, or the knives will quickly cut grooves in it, changing the above too.

Back in the early 1800's, when scientists were first having success trying to measure local gravity, "wootz" steel was used for the knives, and agate plates for the rests.

I bought two 1/8" tool steel blanks for the knives. I haven't decided what the plates should be made out of yet.

My point is if the knife edge is made of steel. It will not work according to your previous post because it will change the uniformity of the material so I don't quite undertand what can be accepted and what's not.

 

[zondar]

My point is if the knife edge is made of steel. It will not work according to your previous post because it will change the uniformity of the material so I don't quite undertand what can be accepted and what's not.

Yes, but practical realities must be faced as well. It's impossible to physically make a truly ideal pendulum. You just do the best you can and hope the little errors and non-idealities don't kill your measurements (actually, you would calculate the likely contributions to error from each factor and account for it all in an error budget).

The measurement of local gravity requires about 5-6 significant digits of precision. There are many sources of error, from the measurement of the knife blade positions (which has to be somewhere near the micrometer level) to finding the center of gravity of the pendulum, to the uniformity of your materials, to the temperature and local air pressure (which buoys the pendulum), to vibrations in the support structure, to error in the pendulum's period measurement, and many more. Luckily, at least some things are easy to do much more accurately these days than in the 1800's, such as the period (time of each swing) measurement.