Multiaxis Pocket

[Flycut]

I still have a lot to learn about the Multiaxis side of Mastercam.

Can anybody please suggest the best toolpath for this feature on a 4th axis? 

I've tried them all and none of them gave me what I was looking for.

Testing.ZIP

[Corey Hampshire]

Have you tried Multiaxis Pocketing?

 

***Edit, I did it in 2022 and now see that you were doing it in 2020. Pocketing isn't available in 2020.***

Edited June 23, 2021 by Corey Hampshire

[crazy^millman]

18 minutes ago, Flycut said:

I still have a lot to learn about the Multiaxis side of Mastercam.

Can anybody please suggest the best toolpath for this feature on a 4th axis? 

I've tried them all and none of them gave me what I was looking for.

Testing.ZIP

Victor check this and see if it helps. I posted with a generic 4 Axis HAAS post to test and it works.

5th Axis 4th Axis Pocket Testing

1 minute ago, Corey Hampshire said:

Have you tried Multiaxis Pocketing?

That is what I used, but if he doesn't have 5 Axis as part of his Mastercam then he will have to use Axis sub. I went ahead and added that process to the file. Have to ignore the bad backplot position.

[Flycut]

Thanks guys,

I do have 5 axis options now but only for the past 2 months. It's all still new to me.

I did have to resort to axis substitution to get the job done but knew there had to be a better way.

From my understanding A axis position relative to other features gets lost with axis substitution.

 

 Millman,

The multiaxis operations you sent me are perfect.

I'm certain I tried that along the way but did not get these results.

I'll dissect the parameters you have set and try to understand what I was doing wrong.

I owe you one.

Thanks again. 

[crazy^millman]

On 6/23/2021 at 11:55 AM, Flycut said:

Thanks guys,

I do have 5 axis options now but only for the past 2 months. It's all still new to me.

I did have to resort to axis substitution to get the job done but knew there had to be a better way.

From my understanding A axis position relative to other features gets lost with axis substitution.

 

 Millman,

The multiaxis operations you sent me are perfect.

I'm certain I tried that along the way but did not get these results.

I'll dissect the parameters you have set and try to understand what I was doing wrong.

I owe you one.

Thanks again. 

No one ever owes me anything. I am just honored to be able to help anyone I can help. I had to fight my way through the ranks to gain a lot of the knowledge I have. The better we all are the more we can help each other. All I ever ask is to pass it on to someone who is trying to learn.

[Flycut]

One thing I found odd is the position of your geometry for the axis substitution operations.

I normally just have my geometry in the middle of the part in the same plane.

Is this how you control where the pocket will be on the A axis?

I was only interested in the multiaxis operation at first but I'm going to have to take a closer look at everything.

 

 

[crazy^millman]

39 minutes ago, Flycut said:

One thing I found odd is the position of your geometry for the axis substitution operations.

I normally just have my geometry in the middle of the part in the same plane.

Is this how you control where the pocket will be on the A axis?

I was only interested in the multiaxis operation at first but I'm going to have to take a closer look at everything.

It is always a crap shoot with that placement. Since nothing had to be timed to that then I would just let it put it where it comes out to be, but if the timing is important then I do some tests with normal 3 Axis toolpaths at that plane and Axis sub to see what is correct and then adjust my geometry that is being rolled to align it. So yes the placement in relation to the center of the part control where the angle output is located from.

Dylan explained this not to long ago in a different thread. Maybe he will pop in and shed some insight on why unrolling geometry puts the back plot in the wrong place when rolled back. I just always dealt with it begin wrong in backplot since I had external CAV to check code. I am just not sure if a user has to unroll it a different way or move it. Like I said it was always not a problem as long as it was a single feature with no timing to something else. If it needs to time then I do what I am suggesting and do what it takes to dial it in.

I use 5 Axis toolpaths so much that has been a few years since I even bothered with Axis sub in a situation like this, but wanted to help teach you and others with the example so why I threw it in there. I was aware it was going to lead to a conversation about why the backplot is jacked up. Why I put my warning about it.

[Chally72]

To try to shed some flickering candlelight on the mysterious:

With unrolling wireframe geometry in order to path it for an axis sub, you must always rotate it by 90 degrees (or -90 degrees) in the Positioning field of the Unroll command. I forget if it's positive or negative, but like trying to insert a USB, I am guaranteed to get it wrong the first time. The reason for the unroll clocking having to be 90 degrees off has been lost to the annals of time and no one seems to be able to give me a concrete explanation, but suffice it to say we can't change how it functions now or everyone who uses a black background would pick up pitchforks. By doing this clocking on the unroll, the rerolled axis sub should be clocked correctly, and should have a correct toolpath display and a correct backplot display.

 

Now, what Ron is referring to, and what I have seen on two recent parts, is an issue where backplot does not match toolpath display on an axis sub operation in terms of clocking. I don't have a specific trigger for this issue, but from the instances I've run into, posted code is correct- it's merely one of the displays (I believe backplot) that is incorrect.

[crazy^millman]

24 minutes ago, Chally72 said:

Now, what Ron is referring to, and what I have seen on two recent parts, is an issue where backplot does not match toolpath display on an axis sub operation in terms of clocking. I don't have a specific trigger for this issue, but from the instances I've run into, posted code is correct- it's merely one of the displays (I believe backplot) that is incorrect.

Dylan thank you for chiming it I really appreciate the insight.

This has been around since the X4 days so like I said just ignore if the code is right and move on.

[Colin Gilchrist]

25 minutes ago, Chally72 said:

To try to shed some flickering candlelight on the mysterious:

With unrolling wireframe geometry in order to path it for an axis sub, you must always rotate it by 90 degrees (or -90 degrees) in the Positioning field of the Unroll command. I forget if it's positive or negative, but like trying to insert a USB, I am guaranteed to get it wrong the first time. The reason for the unroll clocking having to be 90 degrees off has been lost to the annals of time and no one seems to be able to give me a concrete explanation, but suffice it to say we can't change how it functions now or everyone who uses a black background would pick up pitchforks. By doing this clocking on the unroll, the rerolled axis sub should be clocked correctly, and should have a correct toolpath display and a correct backplot display.

 

Now, what Ron is referring to, and what I have seen on two recent parts, is an issue where backplot does not match toolpath display on an axis sub operation in terms of clocking. I don't have a specific trigger for this issue, but from the instances I've run into, posted code is correct- it's merely one of the displays (I believe backplot) that is incorrect.

I have no direct knowledge of the historical reason, but I do know that John Summers wrote a lot of the mathematical functions that underlie Mastercam.

In the Lathe Product, the "Back Plane" is used for the "Rotary C-Axis Zero Position". (Z-Axis Vector of 'Back'). If I had to guess, I would think that Roll/Unroll functions are making use of some "internal function" under-the-hood, and hence we've got a 90 degree positional difference between "Top C-Axis Zero" for VMC Machines, and "C-Axis Zero" for Lathe machines.

[Aaron Eberhard]

3 hours ago, Corey Hampshire said:

Have you tried Multiaxis Pocketing?

 

***Edit, I did it in 2022 and now see that you were doing it in 2020. Pocketing isn't available in 2020.*** 

I didn't download the file, but if he's roughing, in 2020 it was called "multiaxis roughing." 

[crazy^millman]

43 minutes ago, Aaron Eberhard - CNC Software said:

I didn't download the file, but if he's roughing, in 2020 it was called "multiaxis roughing." 

I used that and did dynamic at the roughing leaving .01 on the walls and floor and then copied the toolpath and left Zero on the floor and walls and change it to offset as the toolpath. Funny back then people thought you could only rough with it because of the name, but I have finished many parts using this toolpath on 4 and 5 Axis machines.

[Chally72]

Coming back to this with some new insight!

 

Ron, after being messaged by a fellow employee who caught this thread last night, I think I've narrowed down the issue with display discrepancies with axis sub. See second half of video below.

Flycut, the first half of the video below goes through Ron's example file and shows the proper unroll position for geometry. File also attached.

Axis Sub Example.mcam

 

[crazy^millman]

1 hour ago, Chally72 said:

Coming back to this with some new insight!

 

Ron, after being messaged by a fellow employee who caught this thread last night, I think I've narrowed down the issue with display discrepancies with axis sub. See second half of video below.

Flycut, the first half of the video below goes through Ron's example file and shows the proper unroll position for geometry. File also attached.

Axis Sub Example.mcam

 

Dylan, do me a favor and post that code. What good does backplot do us if the code is wrong? Move the Chain to center like Flycut talked about the backplot and the code will come out correctly. Like I said originally I hadn't  used it in years and he was correct and I had forgot about needing to put it on center when we want an A0 rotation and the Geometry is aligned with top. With the right settings in the Axis Sub the Toolpath backplots correctly and gives us good code. The funny thing is the backplot will show 180, but also give us good code when you use the opposite rotation in the Axis sub inside of the toolpath. All the different scenarios have been added the file I shared the link to and the corrected wireframe is on level 32. OP 16 is the 180 degrees in back plot that will make a good part and OP 15 is the toolpath that back plots correctly and posts correct code.

I have been using Axis sub since the V9 days 20+ years ago and to this day like I said still have to play with it to dial it in, but the code is superior in my humble option than what MultiAxis pocket is giving us. On a HASS run both sets of codes and see which that machine likes better. Higher end machines not much of a difference, but would wager that the Axis sub will still run smoother than the Multiaxis Pocket.

 

 

Edited September 21, 2022 by crazy^millman
Pictures removed to add space

[Chally72]

Yes, one of the benefits of axis sub is the extremely clean code, motion-wise. 

Ron, I shot you a PM as I'd like to explore this topic further with you.

[crazy^millman]

Bumping this topic since someone liked something mentioned in it. Hopefully this thread will help those of you learning some tricks to better your programming efforts.

Dylan congratulations and think Aaron would agree they made a good choice in filling in his shoes. I personally know everyone that has been in this position over the years and they are some big shoes to fill as each of them has my deepest respect. You will do each of them proud with how you carry this part of the product forward. You know how to get a hold of me and if you ever need anything please don't hesitate to reach out and I will do everything I can to help it would be my honor if ever needed.

[Chally72]

Thanks, Ron. As you well know, your thoughts and feedback are always welcome- as well as everyone else's here on the forum. 

 

Aaron made the mistake of relocating within driving distance, so he's bungled his clean escape here. 

[rchipper]

On 6/23/2021 at 12:02 PM, crazy^millman said:

No one ever owes me anything. I am just honored to be able to help anyone I can help. I had to fight my way through the ranks to gain a lot of the knowledge I have. The better we all are the more we can help each other. All I ever ask is to pass it on to someone who is trying to learn.

Much respect. Thank you!!!!