Recall Curves

[[ZE...]]

Does anyone else see a problem when recalling curves into a standard geometric shape that there becomes an issue with the way Calypso compensates for the probe radius? I have a circle on a Torus I need to get a roundness on, but the required circle section is too close to the poles, so Calypso wont let me use "Circle on a Torus". Im trying to use 3D Curve, then recall it into a circle, but the recalled points end up way off. It appears to have something to do with the probe radius that the curves were measured with.

Brett Krienke | 04-04-2018 09:22 PM |

[[ZE...]]

I have done this several times with 2d lines without any issues.

Clarke Gilbert | 04-04-2018 09:28 PM |

[[ZE...]]

If it's real close to the poles then try Curve , best fit . Constrain Z only. You won't get a "roundness" per say, but you can find max and min deviation. The difference between the 2 would be the closest to a "roundness" reading that you can get with a cmm.

Remember you can't really align to the torus, and any run out will look like out of roundness. This problem gets greater as you get closer to the poles. At the equator it won't be as bad.

We get roundness callouts like this all the time, I don't think it's the right callout. Profile would be better, as you get closer to the poles it looks closer to a plane, so why ask for "roundness" ? That's my opinion. Feeble as it may be LOL.

Dave Scott! | 04-05-2018 04:54 PM |

[[ZE...]]

Try using 2D curve instead of 3D curve. A circle is a 2D element, so you'll want to match that up with a 2D curve.

This has been my experience at least.

John LevinerESC | 04-05-2018 05:21 PM |

[[ZE...]]

The 2D Curve gets the Vectors wrong in this particular application, and the Recalled points from the 3D curve are for some reason thrown off a full diameter of the the Probe size once they get brought into a circle. This may simply be a glitch on the simulation side. I haven't tried to actually run this on the CMM. I ended up using circle on a Cone to find the Circle paths on the Torus by calculating what a tangent cone would look like to the torus circle path in order to get the correct vectors.

Brett Krienke | 04-05-2018 07:06 PM |

[[Lo...]]

I have had this issue.. I was scanning an arc using curve. I recalled feature points in to a circle and my diameter was off by 0.2. I never did figure it out. ❓

[[Ch...]]

I recall curves into circles and play with the evaluations setting all of the time. I use this for the OD and ID of complex shapes.

[[ZE...]]

check your projection of circle

Dilipkumar Shah | 04-09-2018 03:50 PM |

[[De...]]

Did you set up the probe radius in the geometry of the stylus or does it still say 0?

[[Lo...]]

Please sign in to view this quote.

Playing with the evaluation setting does nothing when extracting off of a CAD model. All the nominal surfaces are perfect. When I extract it from the model it's correct. When I slice the model and do a curve on the same surface by recalling feature points, its 0.2 off. Inside / outside is set properly.

[[Lo...]]

Please see attachments.

Recall points.jpgextract.jpg

[[ZE...]]

The best way I have found to create a radius you require is to use Create feature in the CAD pull down menu. I will try my best to guide you through this.
When you open Create Feature A menu opens up as seen on screen shot "section". Highlight the Section pull down and define the plane where it is needed, shows up green, click on Section at bottom and a series of curves and straight lines in yellow appear.
At his point turn off the rendering Open the Creation Pull down highlight the radius you need to create, mine turned magenta. Screen Shot Creation
Now uncheck full circle then click circle Icon and it will create the circle you need vectored appropriately to part. See Screen Shot Feature.
Now all you need to do is edit strategy and Evaluation.
Dang need to send three replies to get you all the screen shots

Gregory Kozera | 04-12-2018 08:50 PM |

[[Lo...]]

Gregory, I measure this feature as a curve for form evaluation. Instead of re-measuring again for size, I would just recall the curve points into a circle and evaluate size. When you recall the points from the curve into a circle, the nominal size is off by 0.2mm. I agree that creating a circle in the manner you explained works for creating a circle however that would require A re-scan for size when I have already scanned this feature once.

[[Ch...]]

Here is a screen shot of what I do on supercharger rotors. Playing with the evaluation settings does have an impact on recalled curve but the order you do it in is important. I measure the profile of the rotor one time. Then I recall feature points those into multiple circles with different evaluation settings. I can report OD, Root Dia, Curve form, Output a Graph, etc. all with one scan.

male coated profile OD and Root.PNG

[[ZE...]]

Picture 2 Creation

Gregory Kozera | 04-17-2018 07:56 PM |

[[ZE...]]

Picture 3 Feature

Gregory Kozera | 04-17-2018 07:57 PM |

[[Me...]]

I am measuring a .012 OD and a .044 ID using a .3 dia stylus, recalling the points from a 2D curve. Run it on an off-line seat, the CAD model runs great, on the actual part it runs about the stylus radius off. Clicked all the boxes and radio buttons I could find with no changes. Went to Circles and came pretty close, but I don't feel comfortable using only 35 degrees of a circle. Tried the constraints for X,Y in Evaluations and other tricks. In training, they said 35 degrees is not a valid circle for consistent measurements. We usually use a contour system to trace these features but the customer wants a CMM value. Any other suggestions?

[[CA...]]

Please sign in to view this quote.

Hello Watton,
I am trying to do the same, that is my first time with Curve License, but I did't find how to associate a group or segment of points to some features, in my case OD and ID. I have scanned all the profile and now the problem, how to separate the geometries?
Best regards

[[An...]]

Please sign in to view this quote.

Would do the same.

[[CA...]]

Hi CMM Community
Sometime when I try to construct a feature, ex. a line, circle or curve using the scanned data from other 2d curve, the new feature appears with a big offset from the real points or surface.
So when we need some projection or interceptions the feature is useless.

[[Da...]]

Please sign in to view this quote.

You really should be using the Zeiss gears software for that. It would do all of that for you.

[[Ch...]]

We have Gearpro rotor and it is okay for what it is. If you are a company that is buying rotors from a supplier then GearPro Rotor will do the trick. If you are a rotor design and manufacturing shop then GearPro Rotor is too limited. You really need to be able to program on them in Calypso and GearPro rotor.

That and you can't do the type of reverse engineering with GearPro that you can with Calypso with Curve/Freeform

[[Ke...]]

Was there ever a fix to this issue?
I'm trying to recall a torus section into a circle, for a "Circle in contour best fit"... but, the curve feature is offset by the size of the probe diameter.

[[Ri...]]

Please sign in to view this quote.

That's not how the Circle in contour best fit works. There is no need to take your Curve into a Circle. You drop your Curve into the Circle in contour best fit, and then you choose what you are looking for.

[[Ke...]]

Please sign in to view this quote.

I don't have a cad model. The feature isn't a circle, but it is very close to being circular. So, I was trying to: measure as a circle (90degree segment), recall that into a curve, and then do "Circle in contour" for a set theoretical diameter size.