CALYPSO Symmetry Plane / Datum C Showing Y Offset on Slot (Understanding Feature Location)

[[Sh...]]

I’m measuring a slot in CALYPSO and creating a symmetry plane from the two opposing slot walls. This symmetry plane will be used as Datum C.

The slot runs north/south, with walls on the X− and X+ sides. So the two walls define the slot center in X.

The behavior I’m seeing:

• Plane1 and Plane2 are stable and repeatable
• Plane1 Y ≈ Plane2 Y
• Symmetry plane is stable
• But the symmetry plane reports about 0.3–0.4 mm offset in Y from nominal

I do not believe this is actual part error.

I also:

• Matched scan strategies
• Switched to outer tangential evaluation
• Tried constructing from both planes and points = same result

Base alignment:

• Datum A → Z
• Datum B → X/Y origin
• Planar rotation → line feature

What I’m trying to understand:

Since the symmetry plane is constructed from two opposing walls that mainly define X,

 How does CALYPSO determine and report the Y location of that symmetry plane?

Is the reported Y value based on:

• Feature centroid
• Nominal CAD patch location
• Projection into the active alignment
• Something else?

And in this situation, what is the correct way to validate that Datum C (symmetry plane) is being reported correctly?

Looking to better understand how CALYPSO handles constructed feature location in directions not directly defined by the input geometry.

Thanks

[[Ma...]]

As you stated for your Base alignment: your Y origin is in datum B. Rotation is given by direction of that line feature - not it's origin.

If you need to obtain rotation where Y of that symmetry plane/line feature will be 0 then you have to use only point for rotation.

[[Sh...]]

Thank you, that helps.

So just to confirm my understanding — in my current setup the Y origin is coming from Datum B, but my planar rotation is defined only by the direction of the line, not its position. Because of that, the coordinate system can still shift in Y, even though the rotation is correct.

So the symmetry plane itself isn’t really “wrong” — it’s being reported in a coordinate system that isn’t fully constrained in position.

Is that the correct way to think about it?

[[Ma...]]

Your Y origin is given by datum B - correct.

I don't understand what you mean by coordinate system shift in Y.

Imagine having a slot where nominal in Y is -10.00 - you will get just direction of that slot. But if you use symmetry point in that slot, then that point will also by Y=0, but A1 or A2 of that slot will not be 0.00 in actuals

Only shift is when you have material condition in datum callout (M) or (L), but this is not available in alignments - you have to se this in datum section of geometric callout

Edited 23 hours ago

[[Sh...]]

Here is what I’m seeing:

• Plane1 Y ≈ -76.624
• Plane2 Y ≈ -76.626
• Symmetry plane Y ≈ -76.625
• Nominal Y = -77.000

So both planes and the symmetry are consistent and repeatable, but the entire result is about +0.375 mm off in Y.

I also verified the slot center on a vision system and it matches nominal, so I don’t believe this is actual part location.

Since the slot walls themselves do not define Y location, I’m trying to understand:

 What exactly is CALYPSO using to calculate the Y position of the symmetry plane?

In this case, the measured geometry is consistent, but the reported Y location does not match the actual part.

Setup:

• ZEISS DuraMax (VAST XXT TL3)
• CALYPSO 2023

[[Ma...]]

Wait - those planes are in X axis? Then this would mean that your touch point is at 76.624 in Y.

[[Sh...]]

 

Yes.  the planes are along X, so the measured Y is around -76.624.

But what I’m seeing is: If I adjust the nominal to match ~-76.624 and re-evaluate, the result still shows a consistent offset of approx 0.375 mm from nominal.

So it doesn’t appear to just be reporting the touch location there is still a discrepancy between measured geometry and nominal.

I also verified the slot center on a vision system and it matches nominal (77.000), so I don’t believe the part is actually off.

I’ve also recreated the features multiple times with the same result.

 

   S

 

 

 

[[Ma...]]

So in this case - please ignore those values. For planes you have to watch mainly axis in vector direction. By measuring strategy you will have actuals different to nominal.
It's because of calculation of an element from measured points.
If you will have a plane 100x100 and in your strategy you will have path 30x30, then your actual size will be roughly 30x30 and position also elsewhere.

For your sketch:

|TP 0.05| A | B | - here apply only one axis ( X )
|TP 0.15 | A | B | C | - here apply only axis ( Y ) - or if you have symmetry point from centers of circles, then you can have XY

[[Ri...]]

If you are using a polyline for the Planes, check the nominals in the polyline.

If the third column is not zeroes, then this can cause what you are seeing. You have to check every point in the polyline.

I have found many times that CAD models will sometimes allow points to be imbedded into a surface, causing this type of affect.

Edited 6 hours ago

[[Ri...]]

How far away from the origin is the center of the slot? Is that the 77mm distance?

Your planes are extremely skinny, so you will struggle to use a Symmetry Plane to get a proper clocking for this. Calypso has to take the feature axis and project it to the origin, so imagine you’ve taken a plane that is only at max 1mm wide, and projected it say 10x or more away, any calculation error in the rotation will get amplified. 
 

In cases like this, I’ve had great success with using Free Form Surface to create the rotation axis for me. 
 

Create an alignment that is your A B alignment, create a FFS that is your Datum C planes, make sure the FFS lives in that A B alignment, evaluation - best fit - only allow rotation around the main axis, create alignment. This alignment you create will become your A B C alignment. 
 

I can post some screenshots once I get into the office later if needed.