Real Cycle Time: How to Calculate, Not Rely on the Catalog

Why the catalog shows a neat but incomplete number

Catalog data almost always reflects the best-case scenario. Tables usually list times close to pure cutting: the tool already in position, the blank clamped, the program running without pauses and the operator not wasting a second. On paper that number looks convincing. In a real shift it rarely repeats.

Between cuts there is time that is easy to underestimate. The machine finishes motion, ejects the tool, indexes the turret, brings the next tool in, clamps the part, waits for a command. The operator removes the finished part, fits a new blank, closes the door and starts the cycle again. Each action is short. Together they noticeably change the result.

Catalogs usually exclude tool changes, clamping and unclamping, loading and unloading blanks, short program pauses between transitions and the first check after start-up. Because of that the number looks tidy, but for planning it’s overly optimistic.

A simple example shows this well. Suppose the catalog lists 42 seconds. On a trial part you get 42 seconds of cutting, 6 seconds for tool changes, 5 seconds for clamping/unclamping, 7 seconds for loading/unloading and another 3 seconds of small pauses. The total becomes 63 seconds per part.

On paper the difference between 42 and 63 seconds seems tolerable. Over an 8-hour shift it becomes large. By the catalog you expect about 685 parts per shift; in reality you get about 457. That’s 228 parts fewer, not counting breaks, adjustments and possible insert changes.

This mistake affects more than the plan. Deadlines shift, costs rise, and a pre-agreed unit price starts to look bad. If a shop calculates output from a neat catalog number, it almost always promises more than it can actually deliver.

Therefore cycle time must be measured using a trial part and the whole sequence around it. Even a 15–20 second error per part turns into extra hours of work by the end of the week.

What a real shift cycle consists of

The full cycle time is not equal to the cutting time from the catalog. For planning output you must count everything that repeats per part. Only then do you see the working number, not the presentation figure.

Usually the cycle consists of five parts: pure cutting, approach and retract moves, idle movements, tool changes, possible re-fixturing, and loading/unloading. On paper the most visible part is cutting. On the shop floor the picture is different.

A machine may cut for 2 minutes, then spend 35 seconds moving idle, 18 seconds changing tools, and the operator another 30 seconds loading, clamping and removing the part. The catalog often lists only the 2 minutes. For planning you need almost 3 minutes 25 seconds per piece.

Approach and idle moves are often not logged at all. This is a common error. These include rapid axis moves, moving the tool to a safe point, opening and closing the chuck, waiting for a signal, turret or table indexing. Individually these are seconds. Over hundreds of parts they add up to hours.

Tool change time is part of the cycle if it occurs in every program or within several transitions. The same goes for re-fixturing. If the part is flipped, moved to another clamp or machined on the second side, that’s not a detail — it’s part of the calculation.

Record one-off setup separately. Tool referencing, chuck setting, a trial run, dimension corrections and the first-part check don’t belong to each piece. Count them once per batch and, if needed, divide them across the number of parts.

Simple rule: if an action repeats on every part, include it in the cycle. If you do it once before the series, treat it as setup. If an operation repeats not every part but every 50 or 80 pieces, convert it into an average time per piece. Then the estimate will match real shift behavior.

How to time a trial part

Use a real part, not a training blank or simplified test, and an approved program. If the route is still changing, the measurement will fluctuate — you’ll only get a rough estimate.

The best candidate is a part the shop is about to produce as a series. For turning or milling that means: an approved program, a known toolset, a defined material and a proper blank, not a “almost the same” substitute.

One run shows almost nothing. In the first cycle the machine, tool and operator often behave differently: the spindle comes up to speed, the operator watches dimensions more closely, an extra pause appears. Therefore run several identical parts in a row, at least five.

During measurement split the cycle into three parts:

• pure cutting, when the tool removes metal;
• automatic machine pauses, including approaches, retracts and tool changes;
• operator actions: loading, removing the part, blowing, pressing start.

This breakdown quickly shows where time is lost. Sometimes cutting is less than half the cycle; the rest is eaten by small stops and awkward loading.

It’s normal to drop the first cycle from the average. If it’s longer because of warm-up, dimension checks or minor korections, exclude it. For planning, the steady rhythm of subsequent cycles is what matters.

Small example: you turn a shaft. The first cycle was 4:10 because the operator checked the size twice. The next four cycles were 3:28, 3:31, 3:29 and 3:30. Their average is about 3 minutes 30 seconds. Use that number to calculate shift output.

If possible, time both with a stopwatch and by logging machine events from the CNC screen. Then compare the records. When they match closely, you have a working norm suitable for machine selection and shift planning.

How to count tool changes and loading

Catalogs often show only cutting. In a shift a noticeable share of time is spent elsewhere: the turret indexes, the spindle stops, the operator removes the part, fits a new blank, clamps it and restarts the cycle. These seconds create the real cycle time.

Measure tool change time in practice, not from the machine’s data sheet. Don’t time the “ideal” turret rotation; time the whole interval from the command to change until the tool starts working again. That interval can include spindle stop, positioning, acceleration and a short program pause. If there are multiple changes in the cycle, measure each separately — they often differ.

It’s convenient to record 10–15 repeats and take the average. One measurement can be spoiled by distraction, a misaligned blank or chips interfering with clamping. The average is usually more honest than any neat table.

For loading count the entire repeating block between parts — not just the moment the operator brings the blank to the chuck. Time is spent removing the finished part, cleaning the clamping area, installing the new blank, clamping, closing guards and starting the program.

If the operator always blows out the chuck, checks a dimension quickly or enters a correction each time, include that. Look at frequency: if it happens on every part, take the full time. If it’s every fifth part, divide the time by five and add the average per piece.

A simple formula looks like this: block time = all tool changes in the cycle + part removal + cleaning + loading + clamping + recurring inspection.

Rare re‑setups aren’t included here. Chuck change, full batch re‑setup, installing a new tool, long warm-up and the first setup part belong to a different calculation. Otherwise you’ll mix per‑part time and batch setup time.

A short example sobers quickly. Cutting takes 54 seconds, plus 8 seconds for tool changes, 19 seconds to remove and install the part, 3 seconds for blowing, and 4 seconds for an inspection every second part (2 seconds average per piece). The total is already 83 seconds, not 54.

For machines chosen for a series this difference changes the whole output plan. So when evaluating equipment, ask not only for the spec number but also for a trial-part measurement that includes operator actions. If you work with a supplier like EAST CNC, discuss this format of the calculation from the start.

What to record in the measurement table

The same part can produce different times even on the same machine if you don’t fix initial conditions. The measurement table is not for a report, it’s for honest comparison. Without it the numbers quickly lose meaning.

First note the blank: material, diameter or section, length and allowance if it affects machining. Steel 45, stainless and aluminum behave differently, so a line “part 1” explains nothing.

Then note the batch size. Time per piece for a run of 5 often differs from a run of 200 because rhythm, inspection frequency and repeating actions change. If the operator deburrs manually each time or measures every second part, note that too.

A good table usually contains a few simple fields:

• blank description: material, size, allowance;
• batch size;
• tools used in the cycle;
• per-part times: loading, machining, unloading, inspection;
• pauses and stops: measurement, chuck cleaning, waiting, correction.

Use the same timing template for all parts in the measurement. If you timed the first part from clamp and the second from program start, comparison breaks. Choose one start and one end point and stick to it for the whole batch.

Record the tools separately. Not only the count matters but which positions are actually used, how many tool changes occur and whether wear corrections were needed. On a lathe the difference between 3 and 7 changes per cycle can eat several minutes per shift.

Make notes about pauses immediately, not afterward. A short note like “stop 4 min, chip wrap” is more useful than any averaged number. Later you’ll quickly see where time goes: cutting, loading or short stops that catalogs don’t show.

A simple notebook table is fine for a trial part. The key is consistency. Then you can honestly calculate the full cycle and shift output.

Example calculation for one part

Take a simple bushing in steel 45 from an order of 200 pieces. The catalog might list a spec time of 2 min 40 s — usually pure cutting without those seconds the machine and operator spend between passes.

On a trial part the picture changes. Measurement returned:

• loading the blank and clamping — 25 s;
• rough turning — 80 s;
• 3 tool changes at 6 s each — 18 s;
• finish turning — 40 s;
• drilling — 30 s;
• part removal and placing in a tray — 20 s.

Sum: 25 + 80 + 18 + 40 + 30 + 20 = 213 s. That’s 3 min 33 s per part. The catalog promised 160 s. The difference is 53 s.

For one bushing this may not seem critical. But shift output changes sharply. With 7.5 hours of productive time, the catalog predicts 168 parts per shift. In reality you get 126. That’s 42 parts fewer.

Over 22 work shifts that’s 924 parts per month. For a simple bushing or shaft that gap immediately hits lead times, operator load and cost. That’s why a plan exists but finished parts are fewer than expected.

And this is still a conservative estimate. We didn’t add the first-part setup, toolholder correction, insert change or short stops for chip issues. In live work the full cycle often ends up higher.

So when choosing a machine for metalworking, prefer a trial on your own part. One proper test on a bushing typically gives more value than a spec time without loading and tool changes.

Where mistakes happen most often

The most common mistake is simple: they take the time from the CNC screen and assume it’s the full cycle. But the screen often only shows program run time. The operator still must remove the part, clean chips, load a new blank, clamp it and press start.

If the program runs 42 seconds and loading/unloading takes another 15–20 seconds, the shift output changes noticeably. On paper the number is fine; on the shop floor it doesn’t hold.

Second mistake — relying on one successful cycle. The first trial part or one “clean” pass is almost always better than average. For estimates you need a series, at least 10–20 parts in a row, to reveal small pauses, size checks, chip clearing and the operator’s normal rhythm.

Tool wear is often forgotten. At the start the cutter cuts briskly; over time speed drops, size adjustments appear and inserts may need replacing. If that’s not in the calculation, the result is too optimistic.

People also mix up setup time. Don’t add one-time setup to every part directly. But also don’t ignore it for small batches. For 20-piece batches setup heavily affects unit cost; for 2000-piece batches its share is small. These are different cases and must be treated accordingly.

Errors appear when conditions differ. One machine is timed on a short blank in an easy cut, another on a heavier part with conservative feeds. Comparing those numbers answers nothing.

For a fair comparison fix: the blank, material, accuracy requirements, loading method and batch size. When talking with suppliers, ask for a measurement under identical conditions. Otherwise you’re comparing scenarios, not machines.

A quick check before buying a machine

If a seller shows a single figure in seconds, that’s not enough. The full cycle is visible only when all options are measured on the same part and under the same conditions.

Use a simple trial part you actually plan to produce: a shaft, bushing or another common serial element. Then the comparison is honest and a machine won’t win just because it had an easier blank or smaller allowance.

Ask that all variants keep identical starting data:

• one part drawing;
• the same blank material;
• the same allowance;
• identical batch size;
• a clear cycle composition.

Batch size changes the picture a lot. For a run of 5, loading and setup take a big share; for 500 pieces machining dominates. If one supplier measures on a long series and another on a short one, the comparison loses meaning.

Then split the cycle into three parts: cutting, tool changes and loading. Catalogs usually show cutting or almost only it. But in real shifts the operator opens the door, removes the part, loads a new blank, starts the cycle, and sometimes blows and inspects. On short operations these seconds decide everything.

Don’t rely on one lucky run. Ask for an average over several consecutive cycles. The first is often longer, the second sometimes faster than usual, while the third and fourth are closer to real work. Averaging five cycles reduces error significantly.

A good calculation is self-explanatory. It shows what was included in cycle time and what wasn’t. If a table has just the row “27 seconds” with no notes, that number means little.

A proper table is simple: cutting time, tool changes, loading/unloading, then total time per part and shift output. This format makes it easy to check where a machine actually gains time and where the difference exists only on paper.

If the supplier calculates openly and shows the time composition, the discussion becomes substantive. With EAST CNC this is especially relevant, since the company provides selection, commissioning and service. That means the calculation can be tied to real shop conditions, not a single catalog row.

What to do with the numbers next

When measurements are collected, don’t keep them in scattered notes. Put everything into one simple table, otherwise the choice quickly becomes an argument of impressions.

A few columns are usually enough: machine model, pure cutting, average tool-change time per part, loading/unloading, small stops, full cycle and shift output. Add machine price, tool consumption and an estimated part cost. Then you see not only which model is faster but what that difference gives in money.

From such a table you can calculate three things easily: full cycle per part, pieces per shift and part cost per machine. You’ll also see where minutes are lost in real work.

The difference often feels small until you convert it to output. If one machine saves 12 seconds per cycle and the batch runs hundreds of parts, by day’s end that’s more than “a bit faster” — it’s a tangible volume advantage.

Compare models not by a neat catalog row but by two simple indicators: shift output and part cost. Spec speed alone doesn’t decide much. A machine that cuts quickly but spends more time on tool changes or is awkward to load can end up worse.

When selecting equipment for your part, discuss not only the model with the supplier. Share the drawing, material, required volume, loading method and shop conditions. That conversation is useful with EAST CNC too: since they offer consulting, selection, delivery, commissioning and service, the calculation can be linked to real production.

Ask for the calculation in explicit form. Have it show separately how much time goes to cutting, tool changes, loading and what safety margin is taken for normal pauses. A trial-part breakdown like this is far more useful than any neat catalog number.

If the supplier can’t break the total down into those parts, don’t use that result for comparison yet. If they can, you already have a solid basis for selection: clear shift output, clear part cost and clear launch risk.