Symmetry Plane

[[je...]]

I'm trying to create a symmetry plane using a cad model. The two planes are scanned. The individual planes are close to the correct location according to the cad model. I drop in the symmetry plane and recall the feature points for the two scanned planes. I want to evaluate the symmetry plane distance which should be 1.720". Calypso is giving a nominal of 0.188" and and actual of 0.108". What gives?

forum.png

[[No...]]

It's a common mistake to recall points in a symmetry element. Better don't do that. Calypso will calculate 'something', but probably not what you want. A symmetry feature is typically a geometric construction, not a measurement. So you should measure each of the planes individually and then recall the features (not the points!) to the symmetry plane.

[[Da...]]

I see in version 6.4, you can only put features in symmetry

symmetry1.PNG

[[An...]]

Please sign in to view this quote.

Under version 6.4 there is also the symmetry plane!
Measuring - special geometries!

[[Aa...]]

The OP definitely used the symmetry plane feature, not the construction.

The problem is with the nominals. Specifically the nominal normal is wrong. Looking at the image, the nominal normal should be in the X direction, whereas the nominals shown in that dialog have the normal in the Z direction (or really close to it). "i" should be 1.0. "j" and "k" should be 0.

It looks like the OP let it "Re-calculate Nominal Geometry". Unfortunately, when one of the lengths of a symmetry plane is significantly smaller than the width, this can often end badly, with Calypso guessing the wrong normal direction, as it has in this case. (It ignores the measured normal vectors of the points in doing so.)

[[Ch...]]

One thing to keep in mind when working with Symmetry Planes Vs a just play Symmetry, is that Symmetry has no form and is a perfect feature. Symmetry Plane has form. Thinking about that concept will usually help me know which one to use.

[[No...]]

When would I use a symmetry plane with form? That sounds a bit contradictory to me.

[[An...]]

_

[[Ch...]]

Norbert,
To be honest, I have never used a Symmetry Plane that I can recall. The form value in the Symmetry Plane always scared me. I suppose a print could possibly call-out and tolerance a Symmetry Plane. Does anyone else use Symmetry Plane? If so, what about the form value?

[[Aa...]]

Symmetry Plane is to Width as Cylinder is to Diameter.

So, to Norbert's question, when would you use cylinder with form?

Or, to Chad's point, does the form value in cylinder scare you?

The symmetry construction doesn't have a size associated with it. It's merely a mathematical average of the resolved locations of its associated features. (Kind of like recalling the circles measured on a cylinder into a 3D line.) That's the real reason to use Symmetry Plane.

[[je...]]

I was using the Symmetry Plane to establish a Datum which was called out as the width between two planes. Using a Symmetry construction with two planes will give you a single plane between the two planes used in the construction. The Symmetry Plane will result in a feature of size which in some cases in Calypso allow you to add the MMC option.

I tried to change the nominal vector from K to I with no luck. Constraining the evaluation to X and or the nominal vector didn't help either. Luckily the feature was close enough that I could calculate the position with confidence without using the Symmetry Plane.

symmetry construction.png

[[An...]]

_

[[No...]]

Please sign in to view this quote.

I understand, but I wonder what the exact algorithm for the calculation is. How does Calypso know which points "belong together"? Does it separate the points by the features selected (I doubt that)? To me it sounds more like a simple best fit, and that's why I mistrust it 😕
One of my colleagues used to create 3D lines by recalling the points of a number of circles on a cylinder instead of recalling the circle elements. Unfortunately he used the line to establish the base alignment, which worked for a long time, but one day it failed completely (collisions etc.). Symmetry plane sounds to me like it would do the same and the result is a bit unsteady. I may be completely wrong, but I'll stay away from it 🤣

[[An...]]

I work a lot with symmetries, but I have never needed the symmetry plane!
For me, this is another feature that Caypso does not need!

[[Aa...]]

Please sign in to view this quote.

That's not how Symmetry Plane works. To put it in terms of ISO 14660, Symmetry Plane is an Associated Derived Feature, and what you've illustrated is an Extracted Derived Feature

It's a very poorly extracted one, at that. For one, 14660-2 3.2 requires that "the connecting lines of sets of opposite points are perpendicular to the associated median plane".

[[Ch...]]

From Zeiss Form and Location Error Manual.

2018_07_17 6_09 AM Office Lens.jpg

[[To...]]

ouch, i have a crick in my neck now.... 🤣 🤣

[[Da...]]

Please sign in to view this quote.

OK, I see it now, my bad. I have never used it.

[[An...]]

_

[[Aa...]]

Please sign in to view this quote.

Another illustration of everything that drives me nuts about Calypso. The description on that cookbook page makes it sound like Symmetry Plane actually calculates the Extracted Derived Feature. But if that's the case, then why are Outer Tangential and Inner Tangential among the options for Evaluation Technique?

As stupid little experiment, I created a symmetry plane that recalled the measured points of four fictitious Symmetry Points I created. Three had an actual distance of 0.5", and the fourth had an actual distance of 0.7". All four were centered on the Symmetry Plane The resulting Form on the Symmetry Plane was... .1"! If it were calculating the Extracted Derived Feature, the Form should have been the same thing as Flatness of the Extracted Derived Feature, which would be Zero. If, on the other hand, I create a Plane and make it a Theoretical Feature recalling the four Symmetry Points, I get a Form of Zero. (Incidentally, this plane, which represents the Extracted Derived Feature, also gives me the option to apply Flatness.)

But the cookbook isn't going to tell you this, because setting Calypso up to calculate midpoints for each opposing element with a data set large enough to actually measure something would be excessively cumbersome, even for Calypso. At the end of the day, it's easier to give us a solution that's manageable but wrong than to give us one that's right but unwieldy.

(Taking the analog back the other way, you can't directly apply straightness to a cylinder in Calypso. If you want to measure the straightness of the cylinder axis, you have to measure a series of circles and recall the circles into a 3D line. Then you could apply straightness to the 3D line. If you simply try to use the form of the Cylinder, you get an incorrect result.)

[[Aa...]]

Please sign in to view this quote.

😃 Yup.

You could translate that to a cylinder, too.
Symmetry of 2 planes ==> axis of cylinder
All points are located on a perfect flat plane ==> all points are on a perfect cone
Distance ==> Diameter

Note: the direction constrained to LSQ isn't something that happens automatically in Calypso. You have to tell it to do that.