4th axis indexing

[Rocketmachinist]

I have a guy at my work who is trying to make some parts on the rotary. He has his wcs set to the center of rotation. But when he creates his 2nd planes he decided to move his t & c  plane origin off of his original location and selected the top of the part. Everything backplots fine but when he posts it puts out the numbers like .1 instead of 2.1. I told him to just work completely off of center of rotation but he didn't want to hear it. He said there was a way to shift the origin to wherever he wants but then inside the toolpath you just need to select the original origin and the t & c planes can be where ever he wants. Is his way possible?

[Werktuigbouwer]

Something like:

DWO macro 4th axis

Corresponding Youtube:

 

To post out a custom gcode, you need to use:

*tox$, *toy$, *toz$ or: *tox4$, *toy4$, *toz4$

For the macro you do need to know the rotation point.

I am working on a similar project but i first have to make the fixtures for that machine.

 

[neurosis]

7 hours ago, Rocketmachinist said:

Is his way possible?

Maybe with the right post?

Why does he have an issue with leaving the origin at the center of rotation?

[Rocketmachinist]

19 minutes ago, neurosis said:

Maybe with the right post?

Why does he have an issue with leaving the origin at the rotation of center?

Personal preference, it seems. Personally I like to program from the middle of rotation, then when I backplot the numbers that spit out on the backplot are the same as the part I am going to run, no conversion in the background.

[byte]

1 minute ago, Rocketmachinist said:

Personal preference, it seems. Personally I like to program from the middle of rotation, then when I backplot the numbers that spit out on the backplot are the same as the part I am going to run, no conversion in the background.

Does it post out correctly if the nci is moved using transform? Just curious,

AFAIK there is no reason this wouldn't work in theory, in nesting we use multiple TPLANE and CPLANE origins, in 5 axis we like to use a new WCS if supported by the post,

Is the result different in Machine Simulation?

[Rocketmachinist]

 

6 minutes ago, byte said:

Does it post out correctly if the nci is moved using transform? Just curious,

AFAIK there is no reason this wouldn't work in theory, in nesting we use multiple TPLANE and CPLANE origins, in 5 axis we like to use a new WCS if supported by the post,

Is the result different in Machine Simulation?

It's a symmetric part but no repeated patterns around the part. So when A is rotated to 180 degrees that is the only time those features show up. I don't know what it look like in machine sim or ever verify.

[byte]

21 minutes ago, Rocketmachinist said:

 

It's a symmetric part but no repeated patterns around the part. So when A is rotated to 180 degrees that is the only time those features show up. I don't know what it look like in machine sim or ever verify.

I meant a linear transformation of the entire toolpath to origin, then the origin could be ghosted(posting-off) and you could view the nci in both the job location and world locations by alternating back and forth 

It might help to download some other post processor to test out the results, that might give you a bit more context, there are some free five axis posts like the pocketnc one that are well developed and if the results are different, youj might be able to isolate the issue more quickly,

 

[neurosis]

1 hour ago, Rocketmachinist said:

Personal preference, it seems. Personally I like to program from the middle of rotation, then when I backplot the numbers that spit out on the backplot are the same as the part I am going to run, no conversion in the background.

If you look at the generic 5 axis post, it has a 'work shift' baked in to the post.  You can set it to post either way using a miscellaneous integer. 

Do you guys have any posts that are set up this way? 

[crazy^millman]

8 hours ago, Rocketmachinist said:

I have a guy at my work who is trying to make some parts on the rotary. He has his wcs set to the center of rotation. But when he creates his 2nd planes he decided to move his t & c  plane origin off of his original location and selected the top of the part. Everything backplots fine but when he posts it puts out the numbers like .1 instead of 2.1. I told him to just work completely off of center of rotation but he didn't want to hear it. He said there was a way to shift the origin to wherever he wants but then inside the toolpath you just need to select the original origin and the t & c planes can be where ever he wants. Is his way possible?

He is correct he can do it that way, but on the machine he will need a new workoffset for each index angle that is the same distance from that center of rotation to the shift point he is using. He has 100 different angles he will need 100 different work offsets. In Mastercam every unique C-T Plane will also need that same work offset. Can't help the ignorant and proud ones that will not stop long enough to see the error in their ways.

Now with G68.2 on the machine and a properly configured post that output the shifted distance from COR to each T-C Plane then he coudl get away with one workoffset, but is far and few between that I have seen a 4 Axis machine configured on a Fanuc Control to support it. Siemens will be CYCLE800 and on Heidnehain it will be TPLANE then OKUMA use CALLOO88 again if that has been setup correctly and configured for the machine and the post to correctly support that output.

[IbanezTim]

On 2/8/2023 at 9:35 AM, crazy^millman said:

He has 100 different angles he will need 100 different work offsets.

 

On 2/8/2023 at 9:35 AM, crazy^millman said:

Heidnehain it will be TPLANE

I can confirm this... Just recently did a VERY large gear with a 28' pitch diameter in 45° arcs on a Horizontal Boring Machine, which was much too large to put on center of our table. It would have been nice to (also easier for the guys running the part to double check things as they went along) move the tool plane to the face of the part. However I simply suck at editing posts so every tooth profile would plot right but shift my work coordinates the right distance but would still offset the actual program as if the shift wasn't there. In the end we ended up having to work everything off center of rotation for our coordinates, which ended up with us having to shift this massive gear to be on perfect center for "X". I was able to tell Mastercam where the part was sitting on "Z"  as the center of the gear on that axis was off in space. Also would have been nice to Transform my tool paths to the other planes, but couldn't get that to even plot correctly. Only been programing with Mastercam for a little over a year so no clue to what I was doing wrong there.  Overall still did fairly decent as no one in my shop ever attempted this before and the post had to be modified to even index to begin with. When I am all caught up I have to go back and continue to teach myself and edit the post so we can actually machine with the table, not only for indexing. 

[Tim Johnson]

On 2/8/2023 at 1:30 AM, Rocketmachinist said:

I have a guy at my work who is trying to make some parts on the rotary. He has his wcs set to the center of rotation. But when he creates his 2nd planes he decided to move his t & c  plane origin off of his original location and selected the top of the part. Everything backplots fine but when he posts it puts out the numbers like .1 instead of 2.1. I told him to just work completely off of center of rotation but he didn't want to hear it. He said there was a way to shift the origin to wherever he wants but then inside the toolpath you just need to select the original origin and the t & c planes can be where ever he wants. Is his way possible?

We do this every day on our fourth axis horizontals only the 4th axis home location is a variable to allow the same program to run on several machines. Our program zero points are coming off of each piece of  material that's running.