Offset Plane Alignment

[[Ca...]]

Firstly, I apologize, I cannot post images of my part, or drawing. I know that makes this more difficult.

My part has a FCF that uses 3 datum target points on a cylinder, and some other surfaces. The datum points form an isosceles triangle, with one point being on center to the OD, and the other two being equally off center to the left and right. My original method, was to recall these three points into a plane. Because of the two off center points, the Recalled plane is at a slight angle, a bit less than half a degree. After some correlation with the customer, they've requested we use an Offset Plane, which I've never used, but my Advanced Training manual says is used for "raw cast parts", which this is not. The Offset Plane is not at the same angle as the Recalled plane. The Offset Plane is parallel to the cylinder. I've had much better luck calling the points and other surfaces into a Geometry Best-Fit Alignment.

I've used Datum Targets before, so I understand their purpose and application, so we don't need to go over that.

1. Offset Plane looks to "rotate" the plane in line with the CAD model. How does it do this? What is that rotation based on?

2. Which of these methods is a better choice, or is there a better option still?

Calypso 7.0
No Curve
No FFS

[[Jo...]]

The Offset Plane I am familiar with allows you to edit the nominals for each point. The targets should have a Basic dimension for the target, which looping the alignment will refine the vector and position. If the print is not a simple 3,2,1 alignment you might have to resort to best fit ? Can you sketch a print ?

[[Ca...]]

The basic dimensions for the target points are correct and the alignment is looped. The datum points are the primary datum, the other datum is two threaded holes. That's it. Attached is my sketchy sketch.

Untitled.png

[[Jo...]]

Please sign in to view this quote.

That's interesting. I don't think the offset plane feature is going to work for that. Are the two of primary targets defined at an angle so that you probe perpendicular to the cylinder ? I see the problem with skewed primary datum. The offset plane feature offsets the points in one vector, where you appear to have 3 vectors.

Thinking aloud here: The primary controls 2 degrees of rotation, and poorly 1 translation. The other 2 control 1 rotation and (between them ?) 1 translation ?

option to simulate a tooling ball at tangents ?

[[Ma...]]

That's really bad drawing.

[[Ca...]]

John,

The symmetrical datum targets do not have a basic angle attached to them. They are defined by basic dimensions from centerline and a distance from the third point. They do not have a listed "drop" from the peak of the radius, so it's whatever the CAD model says.



Martin,

That's an ongoing discussion.

[[Ma...]]

I have to correct John - primary control only one degree of rotation ( in Z i presume ) and translation in Z. 2 threaded holes can restrict rest of rotation and rest of translation - but very poorly.

If it's needed to measure it, I would start as basic - cylinders, symetry to get good start. After you have right axis you can start to experiment with RPS ( RPS321 is not suitable )

[[Jo...]]

Please sign in to view this quote.

I suspect the customer doesn't understand the implications of using Plane as the primary. As a plane it would constrain 2 axis, although skewed. The customer requested Plane.. I think the intention is 1 axis along the cylinder. 3 tooling spheres would be my guess is the intent.

[[Ma...]]

Plane will restrict only one axis and one rotation - even if it is stepped plane.
Full drawing or some critical parts would be helpfull.

Surely there is some degree of misunderstood of GD&T.
If they try to create fixture with only primary 3 points, then they never find right axis ( minimum 4 points ), but calypso won't help me with cylinder with fixed radius and 4 points - minimum 5 points.

I have no expierence with RPS - i was ok with plane, point and circle/cylinder.

We have bad drawing too - for towing hook - we have 5 points to establish alignment.

[[Ro...]]

I've seen this alignment on some older drawings. The intent, IMO, is to simulate a cylinder resting in a V-block... poorly. You can't create a cylinder from 3 points so you either have to use a plane, which only restricts 3 degrees of freedom and will not not use the centerline of the cylinder as an axis, or use circle sections and explain to the customer the intent. I would check this part as 2 circle sections at the locations of the points to create a cylinder (4 DOF) and use the symmetry of the 2 holes as the origin and rotation (2 DOF).

If they insist on a plane then your hands are tied and the results are going to be highly questionable. The 3 points don't have the same vector, so I think there is a valid argument against using a plane.