Prepositioning XYZ ABC on head table / head head kinematics

[AchimN]

Some users might have noticed that if they have a head/table or a head/head 5 axis machine that the prepositioning to the inital XYZ positions looks strange in some cases on the machine since the machine is moving in 3 axis simultaneously to that position which is a risk for having collisions.

We have started to develop some code to avoid this behaviour, but we would like to get some feedback on this.

Imagine following:

You have a machine like this and you need to machine your part on 2 opposite sides:

In worst case, the movements on the machine will look like shown with the red lines.

The new method is using TCP for prepositioning and will look like the green lines.

This is done for heidenhain only for now, but it is possible to transfer this method to other posts also.

So, if you have a machine with this type of kinematic and you want to test it, please send an email to cam.posts@autodesk.com and we will send you a post which includes this new method.

Here is a snippet of this movement out of the nc program:

44 ..........
45 L X-0.4 Z+65 FMAX
46 PLANE RESET STAY
47 L Z-20.415 R0 FMAX M91 << move to safe Z height
48 * - 2D Contour2
49 M3
50 PLANE SPATIAL SPA+0 SPB+90 SPC+0 TURN FMAX SEQ+ << Indexing to the new orientation
51 PLANE RESET STAY << reset the workplane (ABC will stay into the new orientation)
52 M128 << enable TCP
53 L X+65 Y+26.2 R0 FMAX << move to global inital XY (ABC is already into the new orientation)
54 L Z-59.6 R0 FMAX << move to global inital Z
55 M129 << disable TCP
56 PLANE SPATIAL SPA+0 SPB+90 SPC+0 TURN FMAX SEQ+ << recall the workplane
57 L X+59.6 R0 FMAX
58 L Z+65 R0 FMAX
59 L Z+55 FMAX
60 ..............

Thanks in advance.

[ArjanDijk]

Hi Achim,

Although I like the effort to have a very nice solution in the post, this is something that in my opinion belongs in the CAM software. Maybe in a manual move operation or a special retract operation.

In Inventor HSM you can see the tool move, you can import your fixtures and you can have an idea if the retracts is sufficient. Putting this in the posts makes it very hard to check for collision and optimal toolpath.

I think your ideas are good and the CAM development team should really implement these.

[Rob_Lockwood]

@ArjanDijk - since machine simulation is probably the only place to accurately check for this, and since machine configuration would have knowledge of the machines kinematics... The logical next step is to implement a routine into machine simulation that causes it to reproduce the move accurately.

[Laurens-3DTechDraw]

@Rob_Lockwood agreed.