## Lathe – turning parts so they are parallel

#### Turning a shaft so that it is parallel

It may seem slightly odd that this should be a separate page since if the lathe is set up correctly it will automatically turn parallel parts. Setting up the lathe so it will make parallel parts is covered elsewhere.

see

Doing this will, for most purposes, produce parallel parts. But if, of course, there are always errors.

If the best possible parallelism is required then the only method is not achieved by hopefully setting the lathe up as well as possible. The test is whether the part produced is as parallel as possible.

Firstly the part is held between centers. The workpiece is fitted with a dog and the dog is turned by means of a catchplate or a face plate.

At first,the part is turned simply to get a smooth surface. Because of the presence of the dog it is not possible to turn the whole length in one go. It is either necessary to make the point where the dog is smaller than the rest. This can then be turned off later. Or to turn the job round as required.

The diameter is measured at either end. Suppose there is an error then this is corrected by moving the tailstock across in the appropriate direction by half of this error. This is repeated till the error is as good as zero.

In principle this is very simple. But to get it right is not so easy.

First problem

If a large amount of material is removed the workpiece will heat up. As it heats up it will expand. If the centers are a tight fit the workpiece might be forced into being very slightly buckled.

Of course if the centers are loose to allow for this the diameter  will be different and it will not be parallel. One way round this is to do most of the turning with the tailstock center adjustes as required. The workpiece is then allowed to cool down. The tailstock is the adjusted to be a good fit and the finishing cuts are made.

Second problem

If the lathe uses a V to align the tailstock on the bed the tailstock will move in a straight line even if the V is slightly worn. The fact that it might move very slightly down at any point will make negligible difference to the diameter of the workpiece.

If the lathe does not use a V and there is any play at all in the movement of the tailstock from side to side then this can be copied across onto the workpiece. And, of course, it might vary as the tailstock moves along the bed.

The traditional solution to this was to set the tailstock up using another dummy workpiece at a fixed point on the bed. This would then fit the real workpiece of the same length.

#### Accuracy of the centers

The axis of the center at the headstock must be coaxial with the axis of the spindle. In the highest class of work this center is ground when in position using a toolpost grinder. Most amateurs will find that if they test it it will be as good as they are likely to get it. But it is possible to cut a center using a carbide bit if necessary.

If there is an error here, it will not show up until the angle of the workpiece relative to the center changes. Since taking the workpiece out and putting back in is often an essential step this becomes important.

It will be noticed that this does not apply to the center in the tailstock if it is a solid center  because the center here never rotates. If a live center is used then any error in it will limit the accurately possible. Accuracy here means that the axis of the cone on the center is coaxial with the axis of rotation of the center.

Sharpness of the cutter

If the cutter is not sharp it will produce unnecessary heat but, worse than this, it will be impossible to make very fine cut without forcing the workpiece piece away from the cutting edge.