Cycle Time

The Milling Cycle Time Reduction Playbook

Cycle time is capacity you already paid for. This playbook shows how to split a cycle into cut, air, and overhead, where the benchmark splits sit, and the daily cadence that takes minutes out one job at a time.

Cycle time is capacity you already paid for. A machining cell running a 12 minute cycle produces 5 parts an hour; take 90 seconds out and it produces 5.7, which is 14 percent more output from the same machine, the same operator, and the same electricity. At $95 per hour and 5,500 spindle hours per year, a 10 percent cycle reduction on one bottleneck machine is worth roughly $52,000 in freed capacity, and unlike overtime it repeats every year. The reason most shops do not capture it is that nobody owns the number: cycle time gets set at first article and then calcifies for the life of the contract.

Start with honest math. Cycle time equals toolpath length divided by cutting feed, plus an allowance for approach, retract, positioning, and in-cut feed variation. A part with 240 inches of cutting at an average 80 in/min is 3.0 minutes of cut; with a typical 40 percent allowance the estimate lands at 4.2 minutes. The Milling Cycle Time calculator gives you that baseline in a minute, and the diagnostic power is in the gap: when the machine actually runs 6.5 minutes against a 4.2 estimate, you have 2.3 minutes of air, dwells, slow rapids, and tool changes to hunt, and every one of them is cheaper to fix than cutting faster.

Benchmark the split before touching anything. Typical shops run 40 to 60 percent of cycle actually in the cut; well-run cells hit 70 to 80 percent. The rest divides into rapids and positioning at 10 to 20 percent, tool changes at 2 to 6 seconds each and often 15 to 25 changes per part, and canned cycle overhead like full retracts on every drilled hole. Measure it once with a stopwatch and the control's distance-to-go screen over three cycles. A part showing 55 percent in-cut has more recoverable time in the other 45 percent than any feed increase will ever deliver, and recovering it carries zero tool life risk.

The air is where the easy money hides. Retract planes set at 1.0 inch instead of 0.1 add about 0.9 inch times two, times every positioning move; across 200 moves at 400 in/min rapids in Z that is nearly a minute. Ordering operations to cut tool changes from 22 to 14 saves 30 to 50 seconds at 4 seconds each. Drilling with chip-break pecks instead of full-retract pecks on a 1 inch deep hole cuts hole time 40 to 60 percent across 80 holes. Then optimize the cut itself: the slowest 20 percent of toolpath usually produces 5 percent of the material removal, so fix the corners and slotting moves first, not the healthy passes.

Failure modes: chasing seconds in the cut while ignoring a 40 percent non-cut share; buying cycle time with scrap, where a feed increase saves 30 seconds and creates a 2 percent scrap rate that costs triple; and optimizing machines that are not the constraint, which improves a report and nothing else. The quietest one is standard erosion: the router says 12.0 minutes, the machine runs 13.8, and everyone has stopped noticing. That 15 percent gap is pure schedule padding, and it compounds into missed promise dates and phantom capacity shortfalls that get answered with capital requests instead of a stopwatch.

Cadence: daily, the cell lead compares actual cycle to standard on every active job and writes down any gap over 5 percent with a cause. Weekly, run a Pareto of the top five jobs by total machine hours, pick one, and spend two focused hours on its air time; a 5 to 10 percent reduction per attempt is a realistic batting average. Monthly, re-baseline standards for anything improved so the schedule sees the gain, and review scrap on every job that got faster. Quarterly, retire the worst legacy program on the bottleneck machine and reprogram it from zero with current tooling and strategies.

World class is boring and specific: actual cycle within 3 percent of standard on 95 percent of runs, in-cut share above 70 percent on the top jobs, a visible board showing seconds removed this month per cell, and standards that get updated within a week of any improvement. Shops that run this system typically pull 10 to 20 percent capacity out of an existing machine base in the first year, which defers a $400,000 machine purchase with a stopwatch, a distance-to-go screen, and an hour a day of somebody's disciplined attention.

Published 2026-07-02.