All Activity

Yes, circle with size defined by the two points, and then a curve from that circle, which is used for a fitting circle. The linear size needs a selection like a section or curve. Cluster is to copy elements and inspections so that would help to create elements around A for every tooth for example.

@Nanno Muehl Thanks - that seems to work and it's much simpler instead of creating next elements just for makeing total of 4 points on circle. So i'll create circle by center and point → refferenced constr. → create curve from that circle ( refferenced ) → create circle from curve ( fitting ) Is that right? Next thing would be cluster i think - i hope i'll learn that too - it's so messy with naming

It should work if you use this function on the circle, then create a fitting circle from the curve: https://techguide.zeiss.com/en/zeiss-inspect-2026/article/cmd_curve_create_curve_from_2d_geometrical_element.html

Hello, I'm scanning a composite part with both plastic and metallic components on a Metrotom 1. My goal is to inspect the plastic part or metal and plastic at the same time, but I'm struggling to get the plastic part. What are the recommended parameters (kV, power, exposure time, and filter) for scanning composite parts with low-density materials (plastic) with metal components? Should I use Beam Hardening Correction? Any suggestions would be appreciated. Thanks!

Hello Daniel, you are right, the current situation within CALYPSO is, that as soon as the "Sigma" Value Limit is reached, the "old/good" Sigma value from the previous qualification is used in the Result Element. The Reason why the faulty qualification doesn't overwrite the good qualification is that it could cause a vector change that could lead to a collision in the following qualification. Development is working on a solution for a better visualization. Currently you can only see that the Stylus is not qualified in the Stylus System Management (marked as red) as well as the "default report". The default Report shows that the stylus is not qualified and additionally the correct/bad Sigma value. Best Regards Marcel

So using two points allows to create correct circle fails on GD&T size - so no MMR can be used

@Nanno Muehl So i've tested that circle. As i thought - it was using whole section to calculate. BUT - if i go to actual circle, constrain it's center point into symmetry point, deselect data and select only those two points, then projection into actual section plane, then it's calculating that circle from two points 🙂 But - is that correct? All hints are telling me it's using what it should be using 🙂

@Nanno Muehl Thanks, i'll look into it. Problem is i can not fit width as stated by base circle/cylinder. I can get GG or MIN and MAX distance. Tested with symmetry from two 2d lines and that worked. Tested with circle, but i am not sure if i can use fitting circle from section to use only those two points - i think it will fit into section measured curve. Any hint into pattern alignment? I don't think customer wants individual tooth to be evaluated separately - it will be machined, so i think he wants to evaluate symmetry ( or one axis TP ) on patterned alignment. Good is that customer is not returning parts even if we are making them a few years.

Good morning, as @Jochen Heller pointed out, a linear size is needed which cannot be calculated by two points in Inspect, Calypso seems to allow this. You could create a cylinder based section through the part, so the section results in two sections per gear tooth which can be used to create two fitting lines -> symmetry line. Linear size is then possible and thus MMR. Alternatively this: A pseudo fitting circle must be created by creating a circle with radius by point (one of the points) -> Curve from 2D element -> use this curve as selection for fitting circle.

Hi @RPSmetrologyTech. It looks like you've really been diligent on analyzing this problem. I like the form plots. Regarding your two last questions: 1. "Wouldn't the shape of the angle measurement be the same as angularity?" That's an understandable assumption. However, no, the shape of the scanned deviations for an angular measurement are fundamentally different from those in a GD&T angularity characteristic. The reason for this relates to the shape of the tolerance zones. Angularity establishes a linear tolerance zone that is controlled by a basic angle driven by the datums. In the case of your chamfer feature, the shape of the tolerance zone is two parallel planes. The .001064 deviation is the quantity of linear deviation from the center of that coplanar tolerance zone. Regarding colors: The magnitude (farther or closer from nominal) of deviations in the angularity characteristic will assign the color gradient based on departure from the coplanar tolerance zone. 2. "I'm also not sure why the cad eval doesn't change color when the characteristic is out of tolerance." The way that the color gradient in Calypso angular characteristic should be assigned is in an arc-shaped path relating to the angle. However, for angle measurement, all we get is a form plot that shows us the form deviation of the two features of interest. This is why there is no color gradient assigned, and it's all green. This CAD evaluation is still useful; however, because it shows you how the planes are calculated according to their form error, and significant form error can be a signal for out-of-tolerance results. I hope that answers those two questions. Chamfers are important features that are often downplayed as "simple." Also, the evaluation algorithm that is assigned to each of the features that construct your chamfer is very important. Outer tangential method will be prone to rocking the planes and introduce high variation in results. If you're using high density scanning, which appears to be the case, LSQ may give you more stable results. Either method may be valid, but it helps to be aware of the effects. Chamfers are also a rare case where I sometimes enable the "constraints" options in the evaluation tab of the features to prompt the characteristic to only consider deviation along vectors that matter functionally. I'm interested to hear how this chamfer measurement problem develops for you.

I set the tolerance in the result formula to a value like +-.000020" for the radius. For the sigma I put .000039" about 1 micron.

Reason I go with line to line is to avoid as much of form error as possible.... Unless features are large enough faces to be considered planes. If you need to continue scanning the planes, then I would make sure that your planes filter out the form errors properly as well as use LSQ eval instead of Outer Tangential. For CAD Eval: Once you set it click on a different characteristic and then back on the one you wanted to look at. Also make sure that you have "color gradient" turned on. Angularity itself is checking the form and orientation of the face compared to the nominal angle. It is no different than checking parallelism (form/orientation at 0°) and perpendicularity (form/orientation at 90°), everything in between those two angles falls under angularity. If the callout is a simple chamfer check, then doing grid scans and checking the form/profile is probably overdoing it.... I would try to picture it as if you set the part on the comparator how would you check/verify your result and then have CMM mimic that.

I will definitely try that when I have some time. I just want to thank all of you for your responses, you guys don't have to help but you do and that means alot!

@Ray Solomon Ok, nevermind. My question would be: Why can Calypso handle this with symmetry point and Inspect not? Why you can use pattern alignment in Calypso and same thing can not be done in Inspect? I hope I know Inspect enough to ask those questions. Only option to handle pattern of 6 points is using RPS. But then i have to make somehow alignment usable in callouts.

https://carl-zeiss-industrial-metrology-llc.helpjuice.com/zeiss_inspect/zeiss-inspect-managing-visibility-and-display-states/version/1?kb_language=en_US

https://carl-zeiss-industrial-metrology-llc.helpjuice.com/zeiss_inspect/zeiss-inspect-understanding-the-workflow-assistant/version/1?kb_language=en_US

Hello, As a new user to the software, I wanted to share a few articles with you that may just help navigate the software a little easier. See below:

I've tried line to line angle as well, and get nearly the same thing as the plane to plane angle. I'm having a hard time understanding the cad evaluation. Wouldn't the shape of the angle measurement be the same as angularity? I'm also not sure why the cad eval doesn't change color when the characteristic is out of tolerance. I'm doing grid scans on the planes.

My apologies, I copied the link to the wrong chat window I had up. for this, I think the issue is not the construction itself, but that MMC/MMR requires a feature of size with an actual size value. The tooth center derived from the two intersections on the basic diameter is only a symmetry point and has no actual size. ZEISS INSPECT documentation states that material requirements require a dimensional tolerance on the element, and symmetry itself is generally not considered an MMC-applicable characteristic. Therefore I don't think INSPECT can represent this callout exactly as drawn. It may be a drawing-definition issue rather than a software limitation.

No... but, whenever I needed to import a second model that needed to be realigned to match the first, I would: import the second model into a new program file, then realign that model to the correct position and save that new program. Then I would import the saved cad model from that newly saved program (model.sab) into the existing program.

I tried building on an existing inspection plan by adding a second model to the file. It would randomly delete the second model from the file and I had to reload it several times. We are using 2024 Calypso. I did most of the added programming with the first model hidden. When I went to test the program it started fine, but then the alignment was super twisted. I would do a manual start alignment and then it would do the basic alignment. The first iteration was fine, but then the second iteration was just wonky and it wanted to crash into the part. I don't think I fiddled with the basic alignment. It was like when I rotated the second model to make it match the orientation and location of the first model that first model (which defined the basic alignment features) rotated even though the first model was directly aligned with the second model.

We have programs made for running qualification checks on the probes that output the sigma value to a report. Once in a while a probe branch will not pass qualification and still output a "passing" sigma. The sigma for each branch is output with a Result Element formula getProbe("Down","3 MM X 50 MM").stdProbeDev. Obviously probes not qualifying is an issue in itself. Are there ways to check the pass/fail status within a program like this?

@Ray Solomon Thanks, but this is not helping in any way. I was trying at home in free version 26 - only option is using 2d lines, so perhaps making it shorter could help.

Unless I am misunderstanding what you're looking for, I recommend using Autorun. You can add a pallet of parts, input the XY offset, and run the entire pallet with a few clicks. This allows you to be able to walk away from the machine while measuring a number of parts. Here is a guide I helped develop in a past job. AutoRun_Guide_2019 (2).pdf

I am trying to update an existing program that requires a max fullness measurement of a helical gear. Originally, the requirement stated the point must lie with ±0.5º (roll angle) of the pitch diameter. When I look at the existing program in GearPro, the tolerance listed in the profile setup for "Profile - position of high point (dc)" shows a max/min value of 0.0661/-0.0661 under user-defined values. The angles are correct on the printout, but I am not sure how the tolerance is converted from the user-defined field to the expected angles. I need to change the expected max fullness point for a revision change, but am a bit stumped as to how I can define the requirement given the entry parameters.