Constraints

[[Dy...]]

So, I've been programming now a little over (2) years. I haven't had much formal training other than (2) weeks at Zeiss for the Basic and Advanced classes. Most of the knowledge that I've gained has been through YouTube, or searching forums like Reddit and Zeiss. Though we covered constraints in the Calypso advanced class, I still have difficulties wrapping my head around constraints. I guess I don't see the use if you have restricted all (6) DOF properly?

Does anyone have any additional information on constraints that might be helpful?

Thanks in advance!

[[Cl...]]

Two things they come in handy for is when reporting size and location. Say you have a very small radii segment of <90° in the XY work plane and you are reporting size, you would constrain location (X & Y). Conversely if you want to report location constrain the features radius.

[[DW...]]

Clark is on the money. When you constrain something in the software, you will notice the actual will become the nominal value. If you have very small portions of a feature to measure, best practice dictates constraining the size or location. Constraining either size or location is going to increase your repeatability without sacrificing the result from an unstable/small segment of a feature. Also in this scenario, you may want to switch to single points instead of scanning.

[[Mi...]]

You have to be careful using constraints though as basically you're making calypso pretend something is nominal. Obviously it's not, or Calypso thinks it's not.

For small partial circles, Constraining Radius might help if you have to report the location. Constraining XYZ (Only 2 at a time) might help if you have to report size. To do both you would have to RFP so you have 2 circles to use... one for size one for location. I generally prefer to use other methods to help the CMM find the circle more reliably than relying on constraints.

The most common way I end up using constraints is on Cylinders/Cones with a high ratio of Diameter to length, on those I will constrain "Normal Vector" almost by default now so prevent Calypso struggling with calculation

[[Cl...]]

Please sign in to view this quote.

I like your use of constraints for cones/cylinders. I will use this going forward, thanks!

[[Mi...]]

Please sign in to view this quote.

You can't always use it... When I say a high ratio I'm talking 50mm+ diameter with only 2mm length. Things like that are quite common in our parts.

[[Cl...]]

What if that scenario was reversed? The diameter was 2mm and the length was 50mm+ length?

[[Jo...]]

Please sign in to view this quote.

I use a similar method when I want a radial distance from a "fixed" center. It helps identify the form in relation to origin, like a cam or eccentric surface.

[[Dy...]]

Please sign in to view this quote.

So, in other words, you are saying if i were measuring a small feature of size where the total feature isn't available to scan, I'm telling Calypso to ignore the Actuals and report the size only from the Nominal location?

Sorry if my interpretation is inaccurate, I just have to put into layman's terms for my simple mind to comprehend 😕

[[Ow...]]

See if the attached helps.
I put it together a few years ago for an internal class to help explain why constraining is necessary on small radius.
If the topic of measuring a small radius is brought up, this will turn into a lengthy discussion like it has in the past because it seems everybody has a different perspective on it.

I know it's not exactly what you're asking but, using curve is good way to measure small radius's as well and below is link to a Zeiss page that covers that.
https://portal.zeiss.com/knowledge-base?id=1665954

Small Radius limitiations on a CMM.pptx

[[Ch...]]

Thanks for sharing Owen,

Hmmm, if Zeiss knows the magnification error is multiplied on radii (circles) less than a given amount (<90°), they could default to constraining this automagically in Calypso (and allow user to uncheck constraints if needed).

Just a thought, I'll probably have to submit a request. 🤣

Instead of being "tribal knowledge" it would be a "standard" in Calypso.

btw, yes I sue this regularly to get accurate radii sizes when needed - works well.

[[Lo...]]

Please sign in to view this quote.

That is an awesome way to explain it. I've tried to break it down for years for people, simply. You have done that. Thank You.

[[DW...]]

Please sign in to view this quote.

Yes.

[[Th...]]

Gentlemen, what is everyone thoughts on the normal vector constraint for cylinder datums on runouts. Same for normal vector on planes for parallelism? I'm getting drastic differences once I activate the normal vector. Seems too good to be true... 

[[Ma...]]

I wouldn't restrict anything unless it's needed or reasonable.

Having small datum and controlled feature at long distance from datum would need consulting with engineer.
In that case others are using axis from both cylinders and then report that runout.

[[Ri...]]

Please sign in to view this quote.

I once had a white paper from a study at a university in California, I think it was Berkley.

I attended a seminar on this in California about 15 years ago, I paid $300 for a copy of the white paper.

This study was about utilizing CMM's and the math behind the CMM's (way over my head in the math department), the title of the study was:

Short Unstable Datums or Features

This paper explained the idiosyncrasies of the math behind the magic, where a feature can have unstable ratios of size vs length and they will or can produce axial anomalies that are factually, not true. The instability is not about the feature being able to move or a fixture that is not capable, in Calypso programming, it was about how the A1 and or A2 values can be skewed and inaccurate. The paper also touched on radii that may or may not have short arc lengths.

Thus, utilizing constraints in manners as discussed in this thread are valid methods to produce accurate results (IMO).

Typically, a secondary method of measurement will validate results.

[[Ky...]]

Please sign in to view this quote.

I'm not surprised you would be getting different results. Let us assume for a moment that the cylinder datum you are using is A and datum B is the feature in the base alignment (or whatever alignment you are measuring A in) that defines the same nominal orientation as A. If you constrain A as you describe, you are effectively doing the runout to B|A rather than to just A. In short, B is controlling the direction, A is controlling the position.

 

Like Martin said, I try to avoid constraints outside of very specific situations.