Cylinder Head Seats and Guides

[[Da...]]

Hi all,

 

Does anyone of you check seats of the cylinder head on a daily basis ? I am wondering how do you 'find' and measure the seats very accurate ? We measure 2 lines on the seat and outer relief and then create intersection point and use this for a local alignment. Any other ideas ? I do not have the PCM license. All I am after is just to make sure that the probe measure the seat even if they are offset like too shallow etc. 

[[Pa...]]

My auto-mechanics never really merged well with my measurement geek side.

Can you share a drawing snippet?

How much off location are you likely to see?

There's probably a material detection or self-centering solution.

[[Cl...]]

Do you have a CAD model? What does the geometry extract as?

[[Ri...]]

By "Seats" are you referring to valve seats?

 

[[Ze...]]

When I did this in the past I aligned to the head face, valve guide, and clocked to the CB's that align the head to the block. Each seat had its own measurement. I used 3d curves placed at quadrants on the seats. We would take a baseline measurement of the new seats and then subsequent measurements after use to track their wear. The results were tracked by head serial numbers. 

[[Da...]]

This is not the same model, but the idea is the same. Basically we need to measure the 'seat' as a cone and generate the angle, runout back to the guides and depth. Usually there is no problem, but if I come off the fire face and the depth is not the same every time I might find the issue with 'finding' the seat. It will measure the cone in not exactly the same position every time. 

[[Ri...]]

When I programmed the heads for Cummins, I used two circle scans and 4 lines scans.

I had no problem reporting runout to the valve guides; it went thru a full MSA at Cummins.

One thing we did, was not to use the entire surface of the Head as a Datum, it had too much form variation. Their engineers went with three Datum targets to stabilize the head surface.

[[Pa...]]

You could certainly poke in there with a self-center point and a big stylus tip, evaluate as midpoint, and take the (x,y, and / or Z) of that point to execute a suitable probing.