Laying 4th Axis on It's Back?

[Myth Project]

I have found myself in a situation where I must lay my 4th axis down where it's at a 5° angle...  How would I go about making MX9 recognize that I've done that for some rotational milling?  As always, any and all help is much appreciated. 

[Guest]

Pretty sure you would need a 5 axis post to handle that 5 degree tilt, even though you don't need the output, you do need it to calculate the angles and rotations properly

[Myth Project]

I don't understand why I would.  I'm still only working on 4 axis', just not completely sure I know how to make mastercam understand that's what's going on. Lol.  (I am trying to figure it out while I'm waiting on ideas.  Thus far, I've noticed I can go in my machine definition and put a tilt angle in...  Not sure if this is going to land me where I need or not, but gotta try something.  )

[Colin Gilchrist]

The logic inside the 4 Axis Mastercam Posts is only setup to rotate about X or Y, at a 90 degree orientation. It checks for this orientation by looking at the Z Axis component of the Tool Plane Matrix. This behavior is coupled with the MP Post Engine, that "auto-magically" calculates the Rotary values for you by resolving each of the Tool Planes being used.

 

So that is why you would need a 5X post to do this. The 4th Axis post is not, in any way, setup to allow you to "tilt" the 4th axis "out of the box". Put another way, the 4th axis is only setup to be aligned with X+, X-, Y+, and Y-. It will not output correctly for any other orientation.

 

You will need to setup a 5 Axis post (configure the Generic Fanuc 5X Mill Post) to get the output you require. You can setup the axes like a normal "Table/Table" machine, and then just set a hard limit for the secondary axis to limit it to either +5/+5, or -5/-5 degrees for the secondary limit. Then it should not attempt to move the secondary, and this will allow you to get correct output for the tool planes you create. It is then a pretty simple matter to remove the 's_out' calls from the PST file, so that you only get 4th axis output.

 

For you to get a 4th axis post to do this, you would have to completely write the logic for handling the rotary output calculations from scratch, since the normal behavior of how MP calculates the 4th Axis position wouldn't work at all. You'd have to set 'rotaxtyp$' to '9999', and do all of the rotary calculations and position calculations yourself, inside 'pmx0$'. This is certainly an advanced post project, to say the very least.

 

To give you yet another example, I've got a fair amount of post experience. It would take me about 10 minutes to get the Generic Fanuc 5X Mill Post setup to handle this. (basic configuration only, debugging, troubleshooting, ect, would take longer...). In comparison, if I were to try and write a custom 4X post only that would handle this, it would be 10 hours minimum...

[Myth Project]

Hmmmm...  I suppose I see the delima now...  I've never played with 5 axis programming.  This project has turned from a big pain in the butt, to complete nightmare...  Not only do I have that issue to work through, but I also need to find tooling to cut this feature.  So far best candidate is solid carbide .750 x 9.0 oal.    I'm not sure how to modify everything I need to in MX9, but I guess it's trial by fire time.  Thanks for the explanation Colin. 

[Matthew Hajicek - Singularity]

Any chance you can get away with 3+1, or does the fourth actually have to move while cutting?

[Myth Project]

It rotates while cutting.  I'm basically making a 5° cone shaped trough in a part...

 

I attached a screen shot.

 

New Bitmap Image.bmp

[Matthew Hajicek - Singularity]

Yup.  That'll be five axis or a whole lot of duct tape and bubblegum.

[Myth Project]

I'm currently in the route of hand coding...  Decided it would be easier and faster than trying to figure out how to make MX9 do what I wanted.  Though I really wish I knew how to make it do what I wanted lol.

 

Gunna plunge, rotate, plunge, rotate, and rinse and repeate til I have the desired part lol.

[jlw™]

Could you do it from the top with a lollipop or some other undercut tool?

[JMahon]

Surface rough it, then finish it by hand coding the finish pass. Shouldn't be that hard on a cone

[Myth Project]

Can'r reach with Lolli, the exit width would be just over 3/8" and the reach would need to be around 8" from that side...  I'm going to do a helix bore for and .825" dia through the part at 0° and 260°, (the sweep of the cone), and the just hand code the rough to plunge in at the 0° and the rotate 260° the plunge rotate back.  That'll get me roughed in...  My thoughts for finishing are program a contour arc (simulating a lead in and lead out, nothing between the two) and then add the rotation by hand.  I think that'll get me close. (hopefully)

[SpacePilotPro]

we put our 4th axis on its back at 5 degrees, used a tooling ball to find the rotation center and z position. the part was essentially a disc, with a whole bunch of 5 degree facets on top. we indexed for each facet, with the facet being in the flat orientation when it was machined.

[Matthew Hajicek - Singularity]

This sounds like a job for Fanuc Macro B! (insert Superman song here)