Why Does Surface Finish Get Worse Near the Center When Facing on a CNC Lathe?

A light finishing pass can leave a rough center despite a smooth outer face. Cutting speed, actual allowance, and cutting-edge condition need checking together.

Center finish can worsen when cutting speed falls at the spindle’s RPM limit and the remaining stock is too thin for the insert to cut cleanly. A rounded or worn edge can rub and tear material. Feed mode, built-up edge, and incorrect tool height can add further defects.

A thin final pass may remove the intended dimension without producing a clean surface everywhere. The edge can cut near the rim and struggle as conditions change toward the axis. A useful starting point is the relationship between shrinking cutting diameter and the spindle speed available to keep the material moving past the tool.

Why Does Cutting Speed Fall Near the Center Even with Constant Surface Speed Active?

Constant surface speed can improve finish across a face¹, but its compensation ends at the RPM ceiling. A light finishing cut can become less stable beyond that point.

Constant surface speed requires increasing rpm as cutting diameter decreases. Once the spindle reaches its permitted limit, cutting speed falls with diameter. At the exact rotational axis, surface speed is zero at any finite rpm. A thin cut and unsuitable edge preparation can worsen the resulting low-speed finish.

The RPM limit changes the cutting conditions

The metric relationship is:

Vc = π × D × n / 1000²

Vc is cutting speed in m/min. D is the current cutting diameter in millimeters. n is spindle speed in rev/min.

Constant surface speed, commonly selected with G96, increases n as D decreases. The control can maintain the requested speed only within the available spindle range.

Region or condition	Spindle response	Effect on cutting conditions
CSS below the RPM ceiling	Rpm rises toward the center	Cutting speed stays near the command
CSS at the RPM ceiling	Rpm remains capped	Cutting speed decreases with diameter
Very small cutting diameter	Available rpm remains finite	Low-speed adhesion and rubbing may become more significant
Exact rotational axis	Further acceleration cannot solve the geometry	Rotational surface speed is zero

A light pass must still form a chip

A small finishing allowance is useful only if the selected edge can remove it cleanly. A heavily honed roughing edge may push or rub thin material instead of forming a stable chip³.

The required allowance depends on material, feed, edge preparation, and the surface left by roughing. A universal minimum such as 0.3 mm cannot describe every finishing setup⁴.

The tool nose radius and cutting-edge hone radius are different features⁵. The nose radius shapes the overall tip. The hone is the much smaller rounding at the cutting edge. Both affect light cutting, but they should not be treated as the same dimension.

How Can You Check Whether the Rough Area Begins Where the Spindle Reaches Its RPM Limit?

A roughness boundary can suggest a speed-related problem, but uneven finishing stock can create a similar pattern. The comparison needs actual rpm and allowance measurements.

The check compares the rough-area boundary with Dcap = 1000 × Vc / (π × ncap), the diameter where the commanded CSS reaches the RPM cap. Actual spindle speed and local finishing stock must also be checked. Similar boundaries support the diagnosis but do not establish the cause alone.

The calculation predicts where speed compensation ends

An illustrative example uses a CSS command of 180 m/min and an effective cap of 3000 rev/min:

Dcap = 1000 × 180 / (π × 3000) ≈ 19.1 mm

Below that cutting diameter, the spindle cannot maintain 180 m/min. At 10 mm diameter and 3000 rev/min, cutting speed is approximately 94.2 m/min.

These values demonstrate the calculation and are not recommended settings for a specific machine or workpiece.

Evidence to collect	What it helps establish
Active CSS command and effective RPM cap	Predicted limit diameter
Actual rpm against cutting diameter	Where spindle acceleration stops or lags
Roughness-boundary radius or diameter	Location of the surface change
Remaining axial stock across the face	Whether the finishing cut remains engaged
Insert condition and feed mode	Other causes of rubbing or tearing

The finishing allowance needs a local check

The operator should measure remaining axial stock relative to the final face datum at a planned stop before finishing. A measurement near the outside alone cannot establish the allowance across the whole face.

Roughing waviness, elastic recovery, or workpiece distortion can leave a nominal finishing pass with very little stock in some areas⁶. Those areas may be rubbed instead of fully recut.

A measured boundary radius must be doubled before comparison with Dcap. The machine’s X display also needs interpretation as a radius or diameter value.

A controlled trial can change the CSS command within the tool’s recommended range and retain the same RPM cap. Feed per revolution, actual stock, and tool condition should remain comparable.

A shifted roughness boundary provides supporting evidence. Surface deterioration may still begin farther inside the speed-limit diameter because the material and edge tolerate some speed reduction.

How Does Feed per Revolution Differ from Feed per Minute During Facing?

Reducing feed may appear to improve a finishing pass, but excessive reduction can promote rubbing. The active feed mode determines what happens as rpm rises.

Feed per revolution maintains the programmed tool advance for each spindle turn. Feed per minute maintains travel speed, so advance per revolution decreases as rpm increases. During light finishing, that decrease can produce chips too thin for the cutting edge. Neither feed mode is inherently more accurate.

The same travel speed can produce different feed per revolution

The feed relationship is:

F = f × n⁷

F is travel feed in mm/min. f is feed per revolution in mm/rev. n is spindle speed in rev/min.

The following arithmetic example shows how the two commands behave.

Feed command	At 1000 rev/min	At 3000 rev/min
0.10 mm/rev	100 mm/min travel	300 mm/min travel
100 mm/min	0.10 mm/rev	Approximately 0.033 mm/rev

With fixed feed per minute, the tool advances less per revolution as CSS raises rpm. Finer geometric feed marks may result, but an unsuitable edge can also start rubbing.

Feed per revolution keeps the tool advance consistent during the speed increase. It does not correct insufficient allowance or restore a damaged edge.

Feed and edge preparation must work together

A sharp finishing geometry can form thin chips with lower forces than an edge designed for heavy cutting. A controlled small hone can still provide useful edge strength. Heavier honing is not automatically suitable for a very light pass.

The chip thickness depends on feed and local cutting geometry as well as depth of cut⁸. Nominal allowance alone cannot establish whether the edge will cut cleanly.

A larger nose radius can reduce theoretical feed-mark height⁹, but it can also increase forces and vibration in a flexible setup. The selection should balance finish requirements with available rigidity.

Under constant radial feed per revolution, spiral spacing remains approximately constant. A dull or smeared surface with fine marks needs inspection for rubbing, adhesion, and wear. Further feed reduction should follow that evidence.

How Should Surface Finish Be Compared at the Outer Diameter, Mid-Radius, and Center?

Surface brightness alone cannot distinguish low-speed tearing, rubbing, and tool wear. A recorded comparison across the face helps connect the defect with the cutting conditions.

Surface finish should be compared at defined radial positions using consistent lighting and measurement settings. The record should include cutting diameter, actual rpm, feed per revolution, and local allowance. Tool inspection then helps distinguish adhered material from wear. A central nub also requires a tool-height and toolpath check.

The three regions need comparable records

The outer measurement should avoid an edge chamfer or burr. Mid-radius lies halfway between the axis and the outer edge. A central measurement needs a stated radius because a roughness trace cannot represent a mathematical point.

Region	Surface evidence to record	Related checks
Near the outer diameter	Baseline feed marks, tearing, or smearing	Edge condition and actual finishing stock
Mid-radius	Whether defects begin or remain unchanged	CSS response, feed mode, and stock variation
Near the center	Rough patches, burnishing, or waviness	Low cutting speed, thin-chip cutting, and tool height
Exact center	Raised nub or remaining material	Center-height setting and programmed end position

The finish does not have to worsen steadily from rim to center. Similar poor finish across all three regions points toward a broader problem with allowance, edge selection, wear, or stability.

Tool inspection separates deposits from material loss

Built-up edge is adhered workpiece material near the cutting edge¹⁰. Its formation and detachment can leave intermittent torn patches and change the effective cutting geometry.

Flank wear is loss of tool material behind the edge¹¹. A growing wear land can contribute to rubbing, but rubbing can also occur with a fresh edge when chip thickness is too small.

The stopped tool should be inspected under magnification before its condition is altered. Deposits, missing carbide, and wear may coexist. Surface patches or noise alone cannot establish which mechanism dominates.

A suitable finishing geometry and consistent lubricant delivery can help adhesion-prone materials. Wear replacement limits should follow the insert and part requirements.

Comparative trials should preserve the useful evidence

A trial should change one factor and repeat the same inspection positions. A feed-direction comparison also needs a holder and insert suitable for the proposed path; outward finishing is not a universal cure.

Remaining vibration calls for checks of overhang, clamping distortion, and machine condition. Suspected axial spindle movement needs a specified service check.

Roughness measurements should use consistent trace direction and evaluation settings. A central area too small for the normal trace needs a separate local assessment.

Conclusion

A rough center needs checks of actual cutting speed, local allowance, feed mode, and edge condition. Controlled comparisons help identify whether speed loss, rubbing, adhesion, or wear dominates.

---