Before we begin, there is one important factor to remember. When a FEATURE is created, the default evaluation method is set to LSQ (Least Squares). This will come into play later on.

Setting up a true position measurement is not difficult. Let’s start with the basics. First thing is to setup our program and establish the base alignment.

What is the base alignment? The base alignment is a three dimensional coordinate system that constrains all six degrees of freedom and is the basis of travel commands of the CMM. The base alignment defines the work piece on the CMM and is the reference frame for the construction of nominal geometry1.  Depending on the application, one may not need to control planer rotation, but for the sake of this example, and the fact the boss doesn’t like replacing broken probes, we’ll constrain all degrees of freedom we can. As a user, you can select whatever features you wish to establish the base alignment. After all, this is only intended for probe navigation. However, a poorly chosen base alignment can lead to hours of head banging on the desk. Probe vectoring is determined by your base alignments and any subsequent secondary alignments. A difference in probe vectoring and feature orientation will affect the location, form, and size. Although only intended for navigation, a  sound base alignment will aid in the overall part evaluation.

For this example, I used the primary and secondary datum to control spatial rotation (wobble / orientation) and my translations (XYZ origins); I used another random feature to control planer rotation (constraining the part from rotating around the Z axis). Once we have the base alignment established, we can create our features. For now, let us look at what happens with true position and one feature. We have one circle here with established nominal locations and actual data,

In our characteristics list, select a true position evaluation. Open up the true position window and lets begin……

Once you select your feature to measure (circle 9), Calypso defaults to using the Base Alignment as the datum.

Piece of cake! Fill in the tolerance and nominal XYZ dimensions. If you do not set the nominals in your feature, Calypso will put the feature ACTUALS into the true position nominals, so be sure to adjust the nominals in your features. CAD models should have the correct nominals built in, but those who program at the machine will need to adjust. Click OK then re-open the evaluation to update the results with your new nominals. (*NOTE* Any time you play around in a true position characteristic and change things around, your nominals will change, be sure to always check your position nominals after any change.) We have a result, so we are done, correct? Not quite. As I mentioned earlier, your base alignment features were using the LSQ evaluation, so you are not getting a “functionally” established coordinate system. Also, there are other factors that make using a base alignment in place of the datum reference frame potentially incorrect.

In addition, when using an alignment as your datum reference frame, any allowed coordinate system mobility is not allowed. An alignment is a completely locked down coordinate system with no capability of datum feature maximum material conditions and no rotational or translational mobility. The coordinate system is based on your BASE ALIGNMENT features. Depending on your selection, this may or may not be what the print is asking for. Also, a BASE ALIGNMENT determines the translation zero location by the intersection point of your features. However, the highest contact points that would touch a Datum Feature Simulator, in order of primary, secondary, and tertiary features, establish a real coordinate system. Calypso will establish the proper location through the form datum reference calculation per ISO 5459. This feature needs to be turned on in your CMM settings found under EXTRAS / WORKROOM / MEASURMENT. Early versions of Calypso do not have this feature.

What the Form Datum does is establish a FUNCTIONAL Datum Reference Frame coordinate system, similar to if you were setting up the primary, secondary, and tertiary datums on a functional checking gage. This is ONLY for a datum reference frame and does not modify how the base and secondary alignments evaluate their respective coordinate system. In the screen shot above, you can notice the different origin points are for a base alignment and datum reference frame.

With this in mind, you can choose whichever method you want to use. There may be complex situations where an alignment is the best option, but I highly recommend filling out the Primary, Secondary, and Tertiary fields for an accurate evaluation.

Our blue print only calls for a true position of the hole to datum B. Datum B is our center diameter, this should be easy. I even chose datum B as my X Y location so the nominals should be the same for the true position. Plug in the feature Circle 9 as the feature and Datum B (circle1) in the primary datum. As a mentioned earlier, a created feature will default to the LSQ evaluation, but in a true position characteristic,  Calypso will default the proper, functional, evaluation method (maximum inscribed element for IDs, minimum circumscribed element for ODs, and outer tangential for surfaces.) Hey great, our result shows in tolerance.

Just to be sure, let us do the calculations and figure out if the results are right. So where do I go? I guess we could look at the feature nominals. But didn’t we say earlier that calypso feature evaluation defaults LSQ? So we can just change to maximum inscribed element, that is the functional evaluation, right?

Break out the calculator and do some math (I know you’re thinking we bought this expensive software to do the math for us, but humor me, it’ll be fun) So the difference from actual to nominal is the numbers we plug into our formula.

Wait, that is not our true position result, what is going on? The feature actuals are based on the base alignment, the same alignment that has controlled all degrees of freedom, orientation, rotation, and translation. In our datum reference frame, we choose only datum B (circle1), which only controls the X and Y translation but nothing is constraining rotation and orientation. Calypso knows datum reference frames may be allowed to shift and rotate, fitting features into their tolerance zones. Since we did not control rotation, Calypso rotated the coordinate system until the feature was at the best result. So where can we see position actuals that reflect location based on the coordinate system? Under the RESOURCES button, choose CHARACTERISTIC SETTINGS EDITOR / ADDITIONAL POSITION RESULTS and set to ON. View your custom printout and the locations show up in the position characteristic.

If we do the math on these actuals, we now know where Calypso is determining the result.

Hey, that is still a tenth of a thousandth difference, what gives? Depending on how the result round is set in your parameters, this will influence the calculation. At this point, we are confident we have the accurate result.

I am going to get on a bit of a soapbox here. The blueprint may have only called positional tolerance to one datum, and there are applications where the orientation of a hole pattern does not functionally need to be oriented to a specific location. In this situation however, is it ok to allow the coordinate system to rotate freely? A circle rotated 90 degrees, but maintaining the proper calculated distance from the datum will still show a good result. For this particular part, will this really work? Even though the print allows this coordinate system mobility, it us up to us as programmers and metrologists to ensure the design intent has been met. Sure our result is legal to the drawing, but if the parts do not work, what benefit is that? As many of us have come across in our careers, blueprints may not truly reflect design intent, requiring a more thorough evaluation to ensure functioning product.

Understanding how this part works, a rotational constraint is necessary for the true position evaluation. Which feature to choose is up for debate and could lead to extensive discussion with engineering, that however is a different matter in itself. So I’ll choose the same plane I used to control planer rotation in my base alignment. Since the print datum reference frame was vague, I’ll choose which features I believe should be used for Primary, Secondary, and Tertiary. You can see, just by changing the order of the features in the datum reference frame, your result will vary dramatically. Which order should we follow? Again, we programmers must decide how to evaluate the part, preferably based on the functional requirements.

In order for a manufacturing company to take full advantage of the tools the CMM provides, the users must understand the “how and why” a result is generated and not take the numbers for granted. As CMM metrologists, we must provide quality feedback to production in order to make the required adjustments. If not we run the risk of chasing number and  scrapping parts. A good CMM report does not necessarily equate to a good product.

Next issue we will review material modifiers, true position of three-dimensional features, and what happens when your datum and features do not share the same space axis.

Nathan Corliss

Assistant Quality Manager at Seiler Instruments and Manufacturing

Vice-Chairman of the Zeiss North American User’s Group

ncorliss@seilerinst.com This e-mail address is being protected from spam bots, you need JavaScript enabled to view it

All screen shots are from Zeiss Calypso software.