Probing the exhaust manifold flange bore positions using CHIPS and Renishaw touch probe

I haven't checked the setup of the Renishaw probe for some time now. Any eccentricity (runout?) of the probe tip will lead to errors in the measured bore coordinates. The eccentricity is measured using a mechanical DTI with very low tip force - obvs you don't want the probe tip to move before the DTI moves. We are trying to measure the runout of the probe as you rotate it 360 degrees. For this, you need to remove the electrical connecting cable.

This Baty DTI resolves 10um, which is fine for my machine, as I will be doing well to achieve much better accuracy than 10um. Or even to get close to 10um to begin with, after careful setup.

The eccentricity is corrected by tightening and loosening the 4 bolts holding the body to the arbor. There's a steel ball bearing that acts as a pivot. I managed to get the runout well below 10um, like 2-3um ie barely registering.

Also measured the diameter of the ruby tip. This has a nominal diameter of 6mm and measures at 5.985 using a 1um micrometer.

Using an old HST bearing as a precision ring gauge, I can check the calibration of the probe as it stands. This is almost convincing - an error of around 12um.

Then I ran the "tip map" function, which plots the measured radius at different angular positions around the ring gauge:

This was before adjusting the probe runout

I managed to get this output at one stage. No idea how this happened, as it's clearly bollocks. If only!!

And this was the final result.

Better check the flange is reasonably horizontal:

Then I started probing the bores, setting my origin at the centre of the first one.

2nd hole:

3rd hole:

4th hole:

Then on to the bolt holes. These are M8 threaded (6.8mm tapping drill) with a 6mm probe, so very little room around the 6mm probe tip.

1st bolt hole:

2nd bolt hole:

3rd hole:

4th bolt hole:

All done. Now what angle is the flange plate at?

This is the angle to the Y axis. If the measured angle was zero, the probed face would be dead parallel to the Y axis.

Here are the various bores and holes drawn up in a sketch:

Extruded, looks like this. The bolt holes for this side of the flange connection are 9mm (M8 clear).

And after some CAM and toolpath setup, I have some gcode ready to go.

The 2D drawing shows the extents of the model that need to be accommodated withing the 100 x 10mm stock.