verify / gouge

[k turk]

i have a part which has a surface rough pocket using a rotated plane. i have followed up the rough with a 5X flow toolpath to finish the surface. in verify the part looks perfect. off the machine, the rough pocket has cut into the finished surface about a sixteenth of an inch in one area. any ideas as to what is going on?

 

X2 MR2 SP1

(i tried to post this question about 10 minutes ago, but i haven't seen it. if you see this question twice . . . . well, you only have to answer one of them. )

[Greg Williams]

Sounds like a Post issue. Do you use a custom post?

[k turk]

yes. this part is on a 5X Onsrud router with an Osai control. Post was written by the fellows at Onsrud.

 

i was thinking that it might be a problem on the other side of MC and programming as the verify looks spot on. just wasn't sure if anyone has seen something like this and found the problem.

 

there is a lot of C and A five axis motion. the surface is steep and curtain-like with waves and undulations.

[Mr. Wizzard]

Have you looked at the cutter comp? Tip or Center?

[Travis_Buchanan]

Is the roughing and finishing done with the same tool? Is the C axis solution the same for both paths?

[k turk]

hello travis (been out of town, thus the long delay)

 

yes, using the same tool.

 

'C axis solution' . . . . hmmmmmm, don't follow the terminology there. is that a parameter setting somewhere?

[Travis_Buchanan]

For every A&C axis tool position there is another solution. For example A90 C90 is the same a A-90 C-90. If teh two paths were ran with C solutions 180 degrees apart this could indicate that the zero positions of the axis need adjustment. Have you bumped or crashed the head?

[k turk]

i see. no, to any bumping or crashing - she's still spotless . . . well, dusty, but undented.

 

the solutions for the axis are created in the post automatically? is this an integer pulled from the miscellaneous values tab? if so they appear to be the same.

[Travis_Buchanan]

Have you indicated the alignment? If you are positive of your programming, I would start troubleshooting by taking some measurement of the spindle alignment with an indicator.

[k turk]

the program seems good; or at least does not appear to corrupt the surface. toward the end of the program the limitation on the spindle rotation in conjunction with the flowline causes alot of unwinding. i don't know if some of this could be avoided using different axis solutions. you have pointed to something i have wondered about but didn't know how to adjust - a method to direct C rotation positive or negative at the beginning of a toolpath - i thought it might just be built into the post. is there a parameter in mastercam to set this direction?

 

 

- Have you indicated the alignment? -

 

of the tool for runout? no i haven't; if that is what you are asking. i will but it couldn't be that far out.

 

alignment of the spindle itself? in reference to what? the table?

[Zoober]

Turk, on't know your post, but in most 5x posts, there is are MIs for setting initial bias, and setting (forcing) positivwe or neg bias. That can alter the axis solution to bring more into primary/secondary travel.

I think the alignment Travis is talking about is the pivot position. pretty critical that you know exactly where it is....

[gcode]

One thing to keep in mind when using Verify..

It shows all rapids as straight lines and very few

machine tools rapid in straight lines.

If you are gertting random gouges that don't show up in Verify, they could be caused by rapid motion between cuts.

[k turk]

nice point gcode. seemed possible there could be some small rapids hiding in the display graphics of the tool vectors that i might have missed. unfortunately, i went back to double check the backplot and it is all feed moves.

 

well, at least two good things have come of this so far. had my eyes opened to some new integers to play with on 5X and 201 posts. maybe i'll make my avatar a giant question mark.

[Travis_Buchanan]

Indicate your spindle. Is it truly vertical at A0 C0? A 10" piece of drill rod in a collet, jog z +/- to see if the spindle is vertical. If so turn A to 90 and see if c is parallel to Y. Also check A for parallel to Y. There is much to be said about 5-Axis alignment, there are the mechanics, the home position, and the servo tuning. Also make sure that values for A&C are always on zero in your orgin table (fixture offsets). When using TCP you can not comp A&C with Orgins.

[k turk]

roger

[gcode]

another point to concider..

Mastercam Verify simulates the motion of the tool

at its tip.

It runs off NCI and has no knowledge of the actual gage length or pivot point of the tool.

A 5x positional motion that looks and acts like a straight line in Verify is actually a large arc centered on the pivot point of the axis... and it may travel right through your drive surfaces in real life.

The only sure way to know code like this is good

is to simulate in in software that can replicate

the tool's motion about its real life pivot points.

Mastercam verify cannot do this.