GreatLight CNC Machining Factory logo
CNC Machining
Rapid Prototyping
Materials
Industries
News
About GL

Get Instant Quote

CNC turning setup

6 problems that be paid when CNC turning is set up in one pass

Six setup problems that be paid with a scrapped batch if nobody checks them before the cycle starts. Written for process engineers and shop programmers who set tools, offsets and cut data on turning centers. Read it to decide which parameters to lock before the first part and which ones to leave adjustable.

±0.005 mmRa 0.8–1.6 μm127 CNC machinesNo minimum order quantity
Scope

What this page covers

Turning and turn-mill setup on continuous cycles, not single-part lathe work.

Problem 1

Depth of cut and feed are chosen as a pair, not separately

A turning center that runs a whole family of parts in one automatic cycle gets its depth of cut (ap), feed (f) and cutting speed (v) set once, then trusted for hours. The common mistake is to pick ap from the drawing allowance and pick f from a table. Those two numbers interact. Too small an ap rubs the insert instead of cutting it. Too large an ap with a fine feed loads the turret and the spindle in a way the tool holder was never meant to absorb.

Roughing should remove the bulk of the stock with a tough insert and a heavy feed. That is what breaks the chip. A fine feed on a heavy depth of cut makes long, stringy chips that wrap the tool and drag on the finished surface. A light ap with a heavy feed does the opposite: the insert skates, work-hardens the skin and the next pass cuts into harder metal than the first.

Set the pair together and prove it on the first part. If the chip comes off in short coils and the spindle load stays flat, the pair is right. If the load needle climbs through the pass, reduce ap before you touch the speed.

Feed per revolution also sets the surface finish on turning, roughly f²/8r for a round-nose insert. Doubling the feed quadruples the theoretical scallop height. On a finish pass that is the difference between Ra 0.8–1.6 μm and a surface that fails inspection.

  • 1
    RoughingTough insert, heavy feed, ap up to the insert nose radius limit.
  • 2
    FinishingSmall ap, moderate feed, nose radius matched to the finish callout.
  • 3
    Warning signSpindle load that rises steadily through one pass.
Problem 2

Tool selection is made for the whole job, not for one operation

A turning setup is a chain. The rougher, the finisher, the groover and the threader all share the same turret and the same offsets. Choosing each tool on its own merit looks reasonable until the second operation. A long boring bar picked for reach will chatter at the speed the finisher needs. A heavy roughing holder that clears the chuck jaws in one orientation may foul them in the next.

Pick roughing tools for rigidity and edge strength first, then let the finishing tools take the precision. A coated carbide insert with a strong edge and a small nose radius handles interrupted cuts better than a sharp, positive tool. For finishing, a larger nose radius and a sharper edge give the surface finish, but they limit the ap you can take without vibration.

Balance the turret as well. Tools with very different lengths hang out at different stiffness. If a long tool runs next to a short one at the same speed, the long one will chatter first. Group operations so each tool works at a speed it can actually hold.

On parts that need both turning and milling in one setup, a mill-turn center lets the same program finish bores and faces without re-clamping. That removes a whole class of tool-access problems, and it removes the error stack from a second fixture.

  • 1
    Roughing toolStrong edge, coated grade, short overhang.
  • 2
    Finishing toolLarger nose radius, sharp edge, light ap.
  • 3
    Boring barsKeep length-to-diameter under about 4:1 where possible.
Problem 3

Cutting speed is set from the material, not from habit

Speed is the parameter most often copied from the last job. That works until the material changes. Aluminum 6061 runs happily at surface speeds that burn a 316 stainless insert in minutes. Titanium and Inconel sit at the opposite end. They need lower surface speed, higher feed per tooth and a rigid setup, because the heat goes into the tool rather than the chip.

The physical limit is the temperature at the cutting edge. Carbide holds its hardness to roughly 800 °C. Push past it and the edge deforms, the insert wears on the flank and the surface finish drifts. On stainless and high-temperature alloys, coolant delivery matters as much as the number. High-pressure coolant through the tool reaches the edge; a flood nozzle aimed at the outside often does not.

Work hardening is the trap with austenitic stainless and titanium. A light pass that rubs the surface leaves a harder layer behind. The next pass cuts the hardened skin, the tool wears faster, and it never quite settles. Keep the feed high enough to get under the skin on every pass.

For plastics such as POM and PEEK the limit is different again. Heat builds in the part and it grows, so the finished bore comes out undersize after cooling. Sharp tools, high speed and air blast beat coolant here.

  • 1
    AluminumHigh surface speed, large rake, watch for built-up edge.
  • 2
    Stainless and titaniumLower speed, heavier feed, never rub the surface.
  • 3
    PlasticsSharp edge, air blast, allow for thermal growth.
Starting points

Cut data starting points by material

Verify on the first part; these are starting points, not a guarantee.

MaterialSurface speedFeed per rev (rough)Insert noteCoolant
Aluminum 6061-T6300–600 m/min0.2–0.5 mmPolished, high rakeFlood or air
Stainless 304 / 316L120–200 m/min0.15–0.3 mmCoated, strong edgeHigh pressure
Steel 1045 / 4140150–250 m/min0.2–0.4 mmCVD coatedFlood
Ti-6Al-4V40–70 m/min0.1–0.25 mmUncoated, sharpHigh pressure
Inconel25–45 m/min0.1–0.2 mmWhisker or PVDHigh pressure
POM / PEEK200–500 m/min0.1–0.3 mmSharp, polishedAir blast
Problem 4

Finishing allowance is left uniform, or it is not left at all

A finish pass removes what the roughing pass left. If the roughing pass leaves 0.1 mm on one side of the diameter and 0.6 mm on the other, the finishing insert sees a varying load. It will chatter on the heavy side and rub on the light side, and the tolerance will drift across the part. Uniform allowance is the single most useful thing a programmer can leave behind.

The size of the allowance depends on the material and the insert. Hardened steel and titanium need a larger allowance because the roughing tool deflects more. Aluminum and brass can run with very little. A common range is 0.2–0.5 mm on diameter for the finishing pass, taken in one or two light cuts.

Do not leave so little that the finishing tool skips over a hard skin. That is the same work-hardening problem from the other direction. If the roughing pass left 0.05 mm, the finishing insert may rub instead of cut.

For parts held to ±0.005 mm, the finishing pass should be predictable. Measure the first part, adjust the wear offset, then let the cycle run. Watching the load meter through the pass tells you if the allowance is even before you measure anything.

  • 1
    Even stock0.2–0.5 mm on diameter is a workable finishing allowance.
  • 2
    Too littleThe insert rubs, work-hardens and wears fast.
  • 3
    First partMeasure, adjust wear offset, then release the cycle.
Problem 5

Workholding decides which of the six problems you actually have

Chuck pressure, jaw position and part length change the stiffness of the whole system. A thin-wall tube held in three hard jaws will go out of round when the jaws close and spring back when they release. The diameter measures right on the machine and wrong after it comes out. Soft jaws bored to the part diameter, or a collet closer, remove most of that error.

Long parts need support, but the support has to match the operation. A tailstock center adds rigidity for the turning pass; a steady rest is better when the middle of the part needs to be cut. Choosing wrong is expensive. A steady rest that contacts a finished surface will mark it. A tailstock center on a part with a shallow center hole will not seat.

Bar work has its own limit. The bar feeder pushes the stock through the spindle, and if the bar is undersize or the collet is worn, the part creeps in Z. Odds and ends show up as a length that drifts over the run. Check the collet grip and the bar end before blaming the program.

For small batches, one fixture setup is usually enough. For 10,000-part runs, the fixture is worth designing properly. Dowel pins, a positive stop and a torque-controlled clamp pay for themselves in the first week.

  • 1
    Thin wallSoft jaws or collets, never hard jaws on a finished diameter.
  • 2
    Long partTailstock for turning, steady rest for mid-part features.
  • 3
    Bar workCheck collet grip and bar straightness when length drifts.
Problem 6

Lighting and inspection points get treated as an afterthought

The last of the six problems that be paid is the one most often ignored. On a turning center the operator cannot see the cutting edge. Chips, coolant and the tool holder are in the way. Setting a general light over the machine and a small lamp at the control gives the operator enough to spot a broken insert, a chip jam or a part that has come loose. A dedicated inspection lamp aimed at the part is worth more than a brighter ceiling.

Inspection has to be planned with the cycle, not after it. If the first part is pulled for a full check, the machine stops. If it is checked while the next part runs, there is a window where the process is unverified. On tight-tolerance work, stop and measure the first part. After that, in-process gauging or a periodic check every set number of parts keeps the run honest.

Record what you find. A log of insert changes, offset adjustments and measured sizes turns a drifting process into a known one. When the same problem shows up on three shifts, the log tells you whether it is the tool, the material or the fixture.

For parts that ship to aerospace, automotive or medical programs, inspection reports are part of the deliverable. Plan the measurement points into the setup so the data is available without stopping the cycle.

  • 1
    Machine lightLight the part and the tool, not the whole bay.
  • 2
    First partStop, measure, adjust, then run.
  • 3
    LogInsert changes and offsets per shift.
FAQs

Common questions

What does “problems that be paid” mean in a turning setup?

It means the setup decisions that get paid for later, either in scrap or in a stable run. Depth of cut, feed, speed, tool choice, finishing allowance, workholding and inspection are all decided before the cycle starts. Get them wrong and the cost shows up in rejected parts.

How do I know if my finishing allowance is even?

Watch the spindle load through the finishing pass on the first part. A flat trace means even stock removal. A trace that climbs means the allowance is heavy on one side.

Measure the part at several points along the length. A diameter that tapers suggests the workpiece is deflecting, not that the allowance is uneven.

Should I use high-pressure coolant on stainless?

Yes, where the machine supports it. Through-tool coolant reaches the cutting edge and controls the heat that causes flank wear. External flood coolant often misses the edge on a deep turning pass.

For titanium and Inconel the difference is larger. Without pressure at the edge, the insert fails early and the surface finish suffers.

When is a mill-turn center better than two setups?

When the part has features that must stay in one relationship, such as a bore and a face that share a datum. One setup removes the re-clamping error and the second fixture.

For simple parts with one turned diameter, a lathe with a tailstock is faster and cheaper.

How often should I check parts during a long run?

It depends on the tolerance and the tool life. On ±0.005 mm work, check the first part properly, then at a fixed interval driven by the insert life.

A shorter interval early in the run catches a process that is still settling. Once the sizes are stable across several checks, the interval can be extended.

What causes a length that drifts over a bar-fed run?

The usual cause is the bar creeping in the collet, not the program. Check the collet grip, the bar diameter and the bar end condition.

Thermal growth in the spindle and the part also shifts length over a long run. Let the machine warm up before the first part is measured.

Send the drawing before the setup is fixed

We review your turning part, flag the setup problems that would be paid for later, and quote with a DFM note.

12-hour quote100% inspection±0.005 mmNo minimum order quantity

Trusted by engineers and manufacturers worldwide

Tesla Ford Motor Company BYD Auto Denso Magna International Boeing Airbus Medtronic KUKA FANUC