# Actualités

### Tutorial - Roundness Measurement Errors and Effects

**In this tutorial we discusses the method for cresting styli, the importance of centre & levelling and setting styli force.**

##### Asperity Effects on MC, MZ & MI Reference Circles

It is often necessary to use the centre of the measured component as a datum. When using the MC MZ and MI reference circles care must be taken when setting a datum, if there is an asperity on the part then this may have a large effect on the centre of the reference circle.

The drawing below shows a part with no asperity. The MC reference circle is constructed to totally enclose the data, the centre of which is shown.

The drawing above shows what happens to the centre of the reference circle when an asperity is present on the measured component. A new centre (datum) is calculated which in turn will lead to bad repeatability and instability in any subsequent concentricity, eccentricity or run-out results. Therefore, care must be taken when constructing datums using reference circles that are calculated from peaks and valleys i.e. zonal type reference circles.

##### Asperity Effects on LS Reference Circle

The Least Squares reference circle relies on area for calculation. The asperity shown on the drawing above although large in amplitude is very small in terms of area and subsequently has little effect on the centre of the reference circle.

This makes the LS circle a more stable reference when constructing datums. It should still be noted that good practice would be to ensure no asperities were present on any measurements.

##### Cosine Errors

When measuring roundness the best possible means of measurement is when the stylus centre and the component centre coincide with the measuring direction. If the stylus centre does not coincide with the component centre then this is a cresting error.

If we look at the situation in the drawing below, here we can see that the stylus tip centre is not in line with the component centre along the measuring axis. This means we have a cosine error in the measurement result.

Cosine errors cause a number of problems, as can be seen above the stylus is presumed to be measuring at the 0 degree position on the table whereas in actual fact the actual angular position is a few degrees off centre.

When trying to centre a component the eccentricity of the component needs to be measured and removed by adjustment. To do this it is often necessary to know the eccentricity position in terms of angle especially if the roundness instrument uses automatic means to calculate and remove eccentricity.

If there is a cresting error then it will become more difficult to centre the component. Although it is good practice to remove all of the eccentricity by mechanical means this is not always possible, residual eccentricity is usually removed by mathematical means. If there is a large cresting error these calculations will also be wrong.