GDT Question

[[Ga...]]

I am trying to figure out if GDT characteristics are based on infinite planes, or if the data is only connected to the points taken during a scan (inside a contained space of a feature). Specifically, I want to make sure that my parallelism is check correctly. I only want it to check from end-to-end of a plane, and not infinite space. See the picture provide. I want parallelism of the selected feature (orifice plate face) compared to bottom face of the workpiece. 

[[Ch...]]

It is not infinite and is based on the size of the feature you are measuring.  If you're programming without a model, then you can adjust the Length 1 and length 2 to simulate the true size of the plane.  Just as a thought experiment;  If it was infinite, then all your parallelism values would also be infinite.  If you projected the parallelism values out to infinity, then it would keep getting larger and larger, and eventually reach infinity.                                                     

[[Ja...]]

In addition to what's already been said, I am pretty sure the results from Calypso will only include the points you have taken regardless of what you set as length 1 and 2. For example, if your feature is 10"x10" and you only probe 3"x3" area then you will get a much better result than if you probe the entire 10"x10" area. You want to cover the entire 10x10 surface and include a point density great enough to capture the true form to measure per standard.

[[Ch...]]

You are somewhat correct.  If using normal features this is exactly what it does.  However, if you recall that plane into a theoretical plane, then it will project to the Length 1 and Length 2.  For example, I checked parallelism between two planes, the measured point results in a value of .0135mm.  However, when I recalled that feature into a theoretical plane and made it absurdly big, then my value changed to .1998mm.  This is something to watch out for when using theoretical features. 

 

I have another good example of cylinder measurements.  Let's look at 3 different measurement strategies of the exact same bore.  One with a 2.5mm cylinder,  one with a 3.175mm cylinder, one with the 3.175mm cylinder projected to 6mm using a theoretical cylinder.  200+ parts are in this graph.  As you would expect, the length of the cylinder has a large impact on measured values.  The length of your measured cylinders is something to look at very closely, especially if you are fighting microns and very concerned about fit and function.  Often you want to project the feature to simulate real life conditions.                               

 

[[Ja...]]

Please sign in to view this quote.

That's really good to know. Thank you.

Edited September 23

[[To...]]

I use the following technique when I recall feature points. 

1. Extract the feature from the model to get correct nominal values. 

2. Open feature, and select "Recall Feature Points"

3. IMMEDIATELY un-tick "Re-calculate Nominal Geometry"

4. Then, select the features to recall points from.

[[To...]]

If I need to create a new plane from multiple planes

1. Extract all the planes.

2. Do not populate with data.

3. Create new plane and open.

4. Select "Recall Feature Points"

5.. Leave "Re-calulate Nominal Geometry" checked and select planes to recall points from.

6. This method will create new plane from nominal sizes of data-less planes.  If there are actuals in the individual planes, the new plane will be re-sized based on where points where taken.