FANUC high-speed cycle machining: expanding the data chain and L repetitions at two levels
Using the 30i-A high-speed cycle machining rules, distinguish repetitions in each cycle header from repetitions specified by call L. Three original data groups illustrate actual execution counts and time, and why the call line alone does not describe the entire sequence.
Contents
One high-speed cycle machining call may execute a chain of pregenerated data groups. The call repetition count and each group's header repetition count act separately. Looking only at the call line can underestimate the number of executions.
The executed objects are dedicated cycle data, not ordinary subprogram names
The FANUC 30i/31i/32i-A user's manual B-63944EN/03, §20.5, pages 689–690 (public mirror of the original manual) explains that high-speed cycle machining uses a macro executor to generate data groups suitable for high-speed pulse distribution, then calls them with G05. P selects the starting cycle and L specifies call repetitions. Header linkage can also chain several cycles, each with its own repetition count. This differs from associating G05.1 with AI contour control simply on seeing the code.
Expand one chain before applying outer repetitions
The following original conceptual data are not a real importable cycle file. Assume the link sequence A→B→C→end:
| Data group | Header repetition count | Next group | Assumed time per execution |
|---|---|---|---|
| A | 2 | B | 0.04 s |
| B | 1 | C | 0.08 s |
| C | 3 | End | 0.02 s |
One complete chain expands to A, A, B, C, C, C. With call L=2, the whole chain executes twice: A, A, B, C, C, C; A, A, B, C, C, C. Thus A executes 4 times, B 2 times and C 6 times, for 12 data-group instances in total.
What is missed by repeating only the starting group?
Incorrectly treating L as applying only to starting group A gives four executions of A, one of B and three of C, totalling 8. This loses 4 executions and changes the relative counts of B and C compared with A.
Ignoring linkage overhead, the correct model takes 2×(2×0.04+0.08+3×0.02)=0.44 s; the incorrect model takes 4×0.04+0.08+3×0.02=0.30 s. The 0.14 s difference does not represent a slower servo: it comes from different interpretations of the execution sequence. Actual time still depends on real data and control execution; table times are calculation assumptions only.
Correct counts do not prove correct data linkage
Sequence review must expand the starting number, next item in each header, and each item's repetition count together. If the generating-data version changes the linkage, actual execution can change even when the NC call line stays the same. No automatic ability to repair incorrect links is assumed here.
The same manufacturer section states that single-block stop and manual-handle interruption are ineffective during high-speed cycle machining. Debugging practices for ordinary NC blocks therefore cannot be transferred directly to this data execution. Verification must follow the machine and cycle-generation software procedures. This article supplies neither macro-executor generation formats nor executable data files. More high-speed machining resources are in the OPTION section.