I've found Fusion 360's estimation for slots and clearing pretty accurate -- well done, all! But drill op estimations are really bad, and I would like to see Fusion 360 have better estimation for the various canned cycles.
Example:
I'm using a Tormach PCNC 1100 which has a rapid of 110 ipm X/Y and 90 ipm Z. My part has 64 holes. I used a peck cycle (drill/deep drilling - full retract) with these parameters:
Plunge feedrate: 12 ipm
Retract feedrate: 50.4 ipm
Clearance height: 1.37 in
Retract height: 0.97 in
Feed height: 0.814 in
Surface height: 0.614 in
Bottom height: -0.01 in
Cycle: Deep drilling - full retract
Pecking depth: 0.0455 in
Fusion 360 estimated time: 5m03s
Actual time: 24m11s
Here's the G-code setup:
G0 Z0.97
G98 G83 X15.4756 Y0.3476 Z-0.0647 R0.8142 Q0.0455 F12.
Tormach says that G83 rapids to "the current hole bottom, backed off a bit", feeds downwards by Q, rapids back to R, and repeats. To give Fusion 360 the benefit of the doubt, I'll assuming that "backed off a bit" is zero.
My guesstimate per hole:
From feed height to bottom height = 0.824 in in 0.0455 in steps = 18 steps (I rounded down to give Fusion 360 the benefit).
Step N starts at the retract height, rapid feeds to (N-1)*0.0455, drills to N*0.0455, and rapid feeds back to the retract height. So we have:
From retract height to feed height = 0.814 - (N-1)*0.0455 in at 90 ipm = 0.57 - N*0.03 sec
From feed height to drill height = 0.0455 in at 12 ipm = 0.23 sec
Retract from drill height to retract height = 0.97 - N*0.0455 in at 90 ipm = 0.65 - N*0.03 sec
Thus, step N takes a total of 1.45 - N*0.06 seconds.
Sum this up from step 1 through step 18, and we get 15.84 seconds per hole.
Times 64 holes = 16m54s (vs Fusion 360's 5m03s and actual 24m11s)
Now, this ignores travel time, but my point is that I removed time from the total wherever I could (travel time = 0, "back off a bit" = 0) and still came up with a number over three times greater than Fusion 360 did.
So somewhere Fusion 360 is taking a lot of liberties with drill calculation.
