All Activity

Your strategy seems plausible. There's only a few things I would change if it were me. I don't know if its possible to establish the center line on the shaft instead of the threads, but that's what I'd be leaning towards. Also, I'd only establish the rotation point on the thread face, instead of your strategy of establishing both Y and rotation using the threads. This is just my theory. It all looks legit otherwise. The way Calypso see's it, if you leave blanks in the base alignment, then it pulls from the start alignment, if there are also blanks in the start alignment, then it pulls from the machine coordinates. Order of precedent is Base Alignment>Start Alignment>machine coordinates.

I've never seen this happen, but my measured position is flipped from the nominal. Has anyone seen this before? For context, I am measuring a cylindrical part on a rotary table. The primary Datums A and B are along the Z axis, measured at 3.200" and -3.200" respectively. I construct a 3d line from the circle paths and use that in my base alignment as my spatial and my X and Y. I was going to just take a fresh manual alignment and see if that addresses the issue, but wanted to see if there was anything I should look at first.

Thank you, i think i found what i shluld: Profile tolerance is set, TE_Profile is set and i still see weird values. Min / Max tolerances: on some sections its both having -0.15 on osome just max has -0.15 I used the update/recalculate, also accept buttons but it not helps. Is there some special order of doing this or what ?

You have to use segments and don't forget to select all segments and click on "measure in group" - that will avoid approach and retract from scan and it will scan whole curve on one go. Only problem is that all measured points will be in first segment. Without grouping it would act like paths in strategy.

@Herbert Spitzer Thanks - after trial i finally found working solution. I had a problem with my curve was scanned in division ( showing "2D curve(16)" ) when i created recalled curve without division i got finally correct measurement without problems. Thanks again for new approach

Good morning everyone, Do you know if there's a way to assign different speeds on the same curve? Thanks a lot.

@Richard Shomaker "tun on display of plot" thats what i was missing. Thanks for the tip!

When I have a program(s) giving me problems like this I will typically try disabling one result/file output at a time until I find the culprit. Have you already turned everything OFF for testing under "Results to File" and Measurement Tab > Multiple Printout?

It happens with every program, big or small. It started all of a sudden. I think it may be related to the fact that I turned on "Results to File" , but that was just for one single program and then I turned it off. I also turned on auto PDF save for some programs, but the lag still occurs for programs that have that turned off. It also happens when only one program is open. Lag is extremely frustrating, especially when you just re-run to pull up a report on results already measured. Thanks (Calypso 2020)

Richard, Thanks for this information! You are correct, the Gear world has its own language!

In addition, you will have to change the tolerance for each .act file you want to evaluate unless this has been changed in newer releases of Gear Pro.

A flatness plot would be very helpful.

Let me know if you resolve this. I have a similar issue.

For the start alignment, I'm using a -z probe and scanning two half circles on either side of the thread. recalling them into a 3d line. with a point in the y axis. I'm setting the base alignment plannar rotation to a point, but this just rotates my model along X axis. Am I missing a step?

The alignment isn't going to change the actuals of the measured plane flatness if you are just doing a recall feature points (they are still the exact same points collected). The alignment will absolutely have an impact on the measured plane flatness because each alignment is telling the hardware where to go to measure data. It seems like right now you are just throwing a lot of stuff at the wall, and trying to see what sticks, and unfortunately you have muddied the water by introducing so many variables, and not controlling any of them. Here's where I would start, and go from there: Create a new measurement plan. Use a stable base alignment: Use the face on the gear as it is the largest as the Spatial Rotation and Z Origin Only take (4) touch points equally spaced around the face (measure as a Plane) Use the ID of the gear as the X, and Y Origin Only take (4) touch points equally spaced around the diameter (measure as a Circle) Use the Keyway as the Planar Rotation Only take (2) touch points (measure as a symmetry point) Loop the Base Alignment 3x Place a break condition of 0.005mm Measure the Plane on the hub face with (2) circle paths (1) inside and (1) outside Scan at 5mm/sec 0.25mm point spacing Filter set to 2.5mm Outlier Elimination set to ±3 sigma (I wouldn't do adjacent points for a Flatness unless necessary) The speed/point spacing/filter might need to be adjusted to match your exact size, but I took a stab in the dark based on how it look.

@Don M. Usually you are limited to what the drawing stipulates. I've seen drawings have no specs on the involute form, and I've seen drawings have both the upper/lower limits of where the involute is to be valid. Sometimes it just isn't possible though due to the sizes required to achieve the spec limits, but I would always start by following what the drawing states. In most cases they just have the TIF, in those cases I just typically ensure that I am meeting that, and then let GearPro set the other limit. Obviously if there is a root/tip radius that I need to avoid I will adjust accordingly for that.

Now I'm just confused even more. So, I was getting frustrated with this whole situation not making sense, so I abandoned the program I had and started over. I was browsing the Zeiss Basic training instructions again, and had the idea to use the a cone to locate the part instead. I took a cone, plane and symmetry point in the keyway(cone/plane/sym as base align), fully defined alignment as it should be. I even swapped to a different probe to reach into the keyway, again, just to see what changes. I still added the 2 circles and created a 3d line, created 3 additional different alignments, (plane/circle), (Plane/circle/sym), and (3d line/plane) so I could see what changes without going back and forth and re-running the part. Made new planes from recalled points, but changed their alignments. As expected, nothing changes, because the alignment shouldn't matter. I've included a bunch of screenshots of the results and am at least now seeing just over .0001" of deviation between results, and are believable for what a ground surface should measure. What was happening before still doesn't make sense to me, especially since I recreated the same "bad alignment" program and got good results this time. Only thing I can think of is the fact that the 3.0mm long probe that is being used is a manual load, not from a probe rack, but I'm not getting the 'probe contacts closed' error, and the probes all re-qaul just fine. Is anyone able to make any sense out of what may have been going wrong previously? I just want to understand what's going on here.

I am logged in as master user. All functions are accessible, until the first time I save the inspection program. After saving, extraction from CAD model stops working. This includes defining points on the model. All other tools function normally.

I would think you could iterate your base alignment to get it there. I do something similar on twin screw compressor rotors. There's even a method where you don't have to iterate your entire base alignment and can iterate just the point. (see attached) Let me know if I'm mis understanding something cause it totally seems possible. Iterative Alignment.pdf

With the delta between the actual point and the desired nominal, and the thread pitch, you should be able to calculate the rotational angle to move to advance the thread the desired linear distance.

Curious how you guys determine the scan path for your profiles. The method we've used measures from the True Involute Form to the OD low limit (for external gears) or the ID high limit (internal gears).

Hi, you have already given the answer: for arbitrary mirror planes we can't ensure that the parametric is "mirrorable" in (most) cases with the intended behaviour of the customer. This leads to more questions than answers and therefore we prevented this option. Geometrically mirroring is supported with arbitrary planes! Regards

Hello, I'm having trouble aligning a part that's basically a cylinder with a thread in the center. Currently I have to take a manual space point and rotate the model manually until the point lines up with the model. Is there a better way of obtaining an accurate alignment for a part like this? Without translating the part in any direction, just rotate it until the point lines up with the model? I'm attempting to scan the threads with a duramax vast xxt.

I looked into the Auto Fetch Data and mine was checked. This didnt work for me. I unchecked and checked and it didnt work. I continue to see what is making this happen.

It's likely that you're not logged in as the Master user and therefore do not have the necessary privileges. Here's a couple articles on the topic of user privileges. How to set up - https://portal.zeiss.com/knowledge-base?id=523210 privileges defined - https://portal.zeiss.com/knowledge-base?id=443337