Do away with manual alignments - use Readpoint

Well, that is fine if that is how it works, that is what ended up doing. I am just used to doing a manual alignment and PC-DMIS will prompt me to take the hits on my alignment features. I didn't expect it to blow past a manual readpoint. But if that is how it works, Ok. I will add a picture with a note showing how to position the probe.

Well, that is fine if that is how it works, that is what ended up doing. I am just used to doing a manual alignment and PC-DMIS will prompt me to take the hits on my alignment features. I didn't expect it to blow past a manual readpoint. But if that is how it works, Ok. I will add a picture with a note showing how to position the probe.

That's the purpose of the readpoint. Remove the manual alignment, which lessens the operator interaction.
That removes the possibility of an operator taking incorrect hits. It also shortens program run time by a few minutes, which saves money if you're running large quantities.

Thanks

"We don't use read points because we have every Tom, Dick and Harry in the shop running programs (over 150 people on 3 shifts). Learn to make the programs idiot proof, you set the part up on the CMM so the set up can be repeated, put good instructions in the program, you take the manual alignment hits then unmark them and anyone with a little knowledge of your parts and PC knowledge can run your CMM programs."

Odd?! I believe that using readpoints does exactly that, cuts down on operator error by making the program "idiot proof'! So easy a "Caveman could do it!" I now use readpoints for all program with CAD and without, and I love it! Anyone can run my programs by just following my Report Commands and placing the probe where told and shown in setup pictures....

I am not experienced in this industry, so excuse me if I am talking out of turn, however...

We skipped all these various methods of alignment and purchased a custom interface for PC-DMIS. So the only options the user has is to select a programme and click RUN. If anyone remembers the MM4, it is similar to that (I know because one of our CMMs run on it).

However, this works because we use fixtures to attach the components to to ensure the components are placed in the same positions every time by whoever uses it. Nothing in any of our software in this company involves any manual input from the shop floor staff. There is no access to the edit window, and our results get exported to the machine computer for viewing as a PDF. It is literally a case of load the component as shown in the picture on the interface, press run, walk off, come back and remove component when finished, then view results at your own computer. We have dramatically reduced run times, errors and probe damage by doing this. I may not know much, but what I do know is I sleep far easier knowing that any problems that occur are my fault.

I have read people talking about idiot proofing a programme. Best way to do this is to only let them press RUN, surely?

I do it almost exactly the same as MatthewHaines, though without a custom interface. No read points, everything fixtured. The operator views setup documentation, presses run, occasionally enters data used to identify what fixture it was machined on, then walks away and views the .pdf on his own computer. Occasionally I assist in interpreting results, but other than that I don't need to be involved.

I include coloured models of both custom fixtures, and TECO fixtures with my CAD info, so as soon as the program is opened, the operator has a pretty good idea as to how to load the part.

we as well have a fixturing alignment built in to our cmm but due to the fact we have two and they are not identical
i find that taking a manual point on the model or part first and creating an xyz origin works great.then i proceed with true datum alignment.after manual point is taken i turn it off and am good to go.also guys using cmm can take this point when method of holding part changes.

How is an external alignment made and what is it used for? I have never used them but I keep hearing about how useful they are. I use read points but only at the start of my programs for "Read Point Alignment".
I have heard that the external alignment is great when you have a lot of odd tip angles that while the program is running is at risk of knocking the probe off; so the external alignment well take the probe to have save place to rotate to the tip angle and then designate to the next feature to be measured. Is this an accurate statement and if so I do I make external alignments? I work in Aerospace and we have a need for a routine like that.

Thanks,

CMMGUY13

How is an external alignment made and what is it used for? I have never used them but I keep hearing about how useful they are. I use read points but only at the start of my programs for "Read Point Alignment".
I have heard that the external alignment is great when you have a lot of odd tip angles that while the program is running is at risk of knocking the probe off; so the external alignment well take the probe to have save place to rotate to the tip angle and then designate to the next feature to be measured. Is this an accurate statement and if so I do I make external alignments? I work in Aerospace and we have a need for a routine like that.

Thanks,

CMMGUY13

An external alignment is quite useful if you have a fixture on your table that does not move. For example we used to have a plate bolted to the granite, in the front left corner of this plate there were two bars about .5" thick in an "L" shape. We created a little program that would measure a plane on the plate, then a plane on the inner face of each bar. Intersect the bar planes to make a line and pierce the plateplane with that line for an origin point. Then we created what we called the "nest" alignment and after it was created we saved it as an external alignment.

Insert>Alignment>Save

Then when we had a part that we planed to "seat" in the nest, we would start the program by recalling the "nest" alignment entering DCC mode and program executes with no manual hits and no readpoint.

HTH

I have basically no fixturing (a machinist's vice, and some screwjacks, and that's it). We mostly don't have CAD files and I really don't know how to program with CAD.

What I typically do is to try to get the part on the table as level and square as I can. I pick two holes with the same centerline. These don't necessarily have to be datums, they are just whatever is convenient, and I can use them to "find" the datum features for the main alignment.

So, I park the probe 1/4 inch or so out from the hole (or corner, or whatever), and insert a readpoint. I set an origin on that readpoint. I usually have prehit and check jacked up a bit so I have a nice zone where the probe can look for a hit. Usually there is a machined plane of some sort (though I suppose this would work with a cast surface). I have the probe take three hits for a plane. I set a new alignment with that plane as an origin for whatever axis the probe is parallel to, and with the readpoint still the origin for the other two axes. Having the plane set up this way, I can control how deep in the hole the probe will go (since my idea of what 1/4 inch out from a hole is and somebody else's idea might be two different things). I use an autocircle to take four hits in the hole, and four hits in the other hole on the centerline. Now I have a plane and two holes for another alignment. From there, I either go looking for the datums (if the part is complex and the datums are hard to get to), or, if the features I am using are actual datums, I just repeat everything, but with more hits. Takes a few seconds longer, depending on the part, but it works, and I don't have to spend a buncha time driving around taking hits for a manual alignment.