Base alignment help

[[Ry...]]

Hello all, I am running into a issue with my base alignment not going to all of my features. I have a large round part with an outer diameter of approx. 72 inches. I have this set up as my XY zero in my base alignment, as I need to check the coaxially of a center shaft. When I scan the outer diameter the actuals for x and y come in at zero. But when I move the probe to roughly were zero should be on the part it shows the x value as negative 20" and the Y as 2. I have set base alignment to zero selected and have re run the manual several times. It almost seems like the base alignment isn't translating in the mcs correctly. 

I'm at a loss of what this could be. I am running calypso 7.8.06

Thank you for the help

[[Ma...]]

In CAD menu you can enable showing probe tip so you can see if it moves like it should and orientation.

Without any more specific data i can not help more.

[[Ry...]]

Please sign in to view this quote.

attached is a photo of the probe shown in the cad the probe is physically located over the center of the part but shows it way off in X

[[DW...]]

Please sign in to view this quote.

Please sign in to view this username.

 Can you post a screenshot of your base alignment setup window?

[[Ca...]]

I saw you mention coaxially, which makes me wonder if a cylinder is involved. Are you using a cylinder for your XY0? If so, I would recommend switching that to a circle for the base alignment to see if that better controls your XY0.

[[Ry...]]

Here is a screen shot of my Base alignment Dat FL is the large circle

 

[[Ry...]]

[[DW...]]

Please sign in to view this username.

 Do not use a circle to try to control rotation. It does not and can not control rotation. A 3d line recalled from two circles and projected to a plane? Sure. But not a single circle. I don't know what your part looks like exactly, is it a symmetrically round part? No actual clocking features? Either way, get rid of the circle for planar rotation, and then do "Execute manual run now." Let me know if this corrects the behavior. 

[[Za...]]

DWC makes a good point about circles being used in planar rotation, but if you feel you must use a circle as your planar rotation, it helps to define the planar rotation first. any point reducible feature used as planar rotation will have the direction defined from the origin to the chosen point-reducible feature

Edited Wednesday at 06:31 PM

[[na...]]

Please sign in to view this quote.

Please sign in to view this username.

 IIRC, that was the case back before 6.0 version.

I'm doing it like OP for years and it's working all the time.

 

 

 

Please sign in to view this quote.

Please sign in to view this username.

 But circle derives a point, doesn't?

[[ni...]]

A circle in planar rotation isn't a problem, it will do the exact same thing as a 3d line construct by RT BA & Dat FL. For Your problem you should verify if all your circle and plan are defined and taken by the same stylus. If you are on a stylus star and run the manual alignement with the wrong stylus it will confused calypso and do this sort of problem.

and don't think it's your problem but verify there is not rotation / translation your base alignment

[[Ry...]]

Please sign in to view this quote.

Thank you I checked and they are all in the -Z direction we have an XTR so we do a lot of L probes with multiple positions. 

 

The part is symmetrically round with only one clocking feature that is the circle I used as the rotation. Normally I use a 3 d line through the centers as of the two and I did on an other version of this program for trouble shooting and I got the same results. Although I don't see how the rotation would cause this since it seems to only being having the issue in a single axis its almost like the base system is being translated in x but I don't see how that can happen. Especially when the base features run just fine

[[ni...]]

have you check there is no translation or rotation in your Base alignment ?

[[DW...]]

Please sign in to view this quote.

Please sign in to view this username.

 I am not going to start a war here, but something "working" and being "correct" are not the same. I have ran/used programs where someone else decided to use a circle to clock rotation. "Hey this tapped hole is always machined at the 12 o' clock, I'll use that to control rotation in Z!" And then the report comes out, and the poor dudes on the machine shop floor are running in circles trying to adjust their machining centers/fixtures to get the numbers back in the green when it is really the CMM program spitting out constantly changing numbers based on a single point with no direction "clocking" the entire program's rotation.

 

Please sign in to view this quote.

Please sign in to view this username.

 It very much is a problem, and will not do the same thing as a 3d line. A single point has a position, with no direction. A 3d line contains two or more points, has a position and direction.

 

For anyone learning the most ideal way to constrain all six degrees of freedom would be to use three mutually perpendicular planes. 

Since not every part is a cube, you will have to use other standard geometries correctly to constrain translation in X, Y, and Z and rotation in U, V, and W. 

It is of the utmost importance as a programmer to understand this concept in the physical realm (fixtures) and virtual realm (object oriented programming in 3d space). 

You directly influence the results (good or bad) with your measurement strategy. 

Best of luck out there.

[[na...]]

Please sign in to view this quote.

I get what you're saying, but I'll have to test this (test again, it's been a while since I clocked with 3D Line).

Anyone else with proofs?

[[Ri...]]

Please sign in to view this username.

 Please share some data to support your claim. 

 

Everyone should keep things civil, but if there is data to backup your claim, we'd all greatly appreciate it. 

[[To...]]

Please sign in to view this quote.

The problem is not using a circle for the planar rotation. It is using a tapped hole to clock the part.  Even a 3d Line created by recalling the center circle and the planar circle will give you the same problem. Use a better planar feature.   In the OP's example, a "single point" created from this circle doesn't need a direction/vector, it only needs the ability to stop rotation about the Z axis, which it can do quite well.  The center of this circle is projected to the top plane, so if the hole has an angular deviation from part to part, or you check the same part at different depths, you will not get the same results, even if it is not tapped.

Edited yesterday at 01:55 PM

[[To...]]

Please sign in to view this quote.

Please sign in to view this username.

 I can't see the RT BA circle in your screenshot showing the probe being off location.

Edited yesterday at 02:00 PM

[[Ze...]]

I've always used a 3d line for clocking, but it's because I've had alignments rarely flip 180 degrees around the axis of rotation when using a point reduced feature for some reason. This has happened in other metrology software I've used, as well. It's rare, but it's happened.  

[[To...]]

Please sign in to view this quote.

The direction is assigned with the dropdown menu on the right side of the planar option

[[To...]]

I often find that when someone's alignment gets flipped around, there are a few common issues.  1) The spatial feature was not stable, i.e. a long, narrow plane.  2) During the manual alignment, the points were not spread out "far and wide".  3) During the manual alignment, taking points on features done on wrong order, i.e. points were taken on the Origin circle, but Calypso was looking for point on the Planar circle.  This happens when you knew the features were in a particular order but you inadvertently started with "run from characteristics" selected.

[[Ry...]]

Please sign in to view this quote.

Please sign in to view this username.

 The RT BA circle is unfortunately rather small in the scale of the part put it is the small circle at the tip of the blue arrow 

Please sign in to view this quote.

I have re run the manual alignment multiple times in taking particular care to make sure that they are being takin the correct order. (I have made that mistake before.) Every time that happened though when the program started on the base it wouldn't run the base features correctly. But in this case it runs the base alignment features fine in then errors out when I try to take get values from the features in the center. (not shown in the screen shot)

[[Ke...]]

Have you tried re-creating the program from scratch? It's been a minute since I've had a random bug in a program, but 🤷‍♂️ (world's worst "shrug" emoji 😂)

[[Ry...]]

Please sign in to view this quote.

I Think the problem that I am having may have got lost after the discussion about how to define the rotation. The X and Y rotation seems to be accurate. Its the X origin that seems to be translating strangely. In my previous screen shot with the probing object shown the probe is roughly over were the actual center of the part is on the table after running a manual alignment, and the machine taking a scan of all all of the base alignment features. What is really throwing me off is that the actuals for the large 72" circle is showing 0,0 for x,y. Yet when I move the probe to 0,0 based off of what calypso is saying in the base alignment its off by around 20". I also checked in the specials settings for the Base system and there are no accidental translation or rotation modifiers in it.

[[Ry...]]

Please sign in to view this quote.

Please sign in to view this username.

 I have that's actually the program I have been showing is the second version. Which is why the center features are not showing in the screen shots