Surface roughness cannot be inspected away after the fact. If the finish on a turned shaft, a milled mould pocket or a bored hydraulic bore comes out above the drawing limit, the part is either reworked or scrapped. Yet in most shops Ra is still treated as something that “happens” once the insert grade is chosen — and operators often respond by lowering the feed until cycle time doubles.
In reality, roughness is a geometric result first and a metallurgical result second. Feed per revolution, nose radius and wiper geometry determine roughly 70–80% of Ra on a stable machine; speed, wear, runout and coolant decide whether you reach that number.
Ra, Rz and What the Drawing Actually Asks For
Under ISO 4287, Ra is the arithmetic mean deviation of the assessed profile from its mean line, while Rz is the maximum height of the profile within a sampling length. Ra averages the whole profile, so a surface with a few deep grooves but a generally flat plateau can still report an acceptable Ra while sealing badly. For sealing faces, bearing surfaces and fatigue-critical parts, Rz or Rmax is the more meaningful number.
For turned and milled surfaces produced by a defined tool geometry, the practical conversion is:
- Rz ≈ 4–6 × Ra for conventional turning and milling
- Rz ≈ 6–7 × Ra for rough, torn or built-up-edge surfaces
- Rz ≈ 3–4 × Ra for wiper-finished or ground-like surfaces with a plateau profile
This matters because a drawing that specifies Ra 1.6 µm is effectively demanding Rz 8–10 µm, and a shop that measures only Ra on a portable stylus can easily ship a surface that leaks. Always confirm the evaluation length, the cut-off (usually 0.8 mm for Ra 0.4–2.0 µm) and the measurement direction — roughness must be measured across the feed marks, not along them.
Turning: The Theoretical Ra Formula
In longitudinal turning with a round or rhombic insert, the tool leaves a helical thread-like profile whose peak-to-valley height h depends on feed and corner radius:
h ≈ f² / (8 · rε) → Ra ≈ f² / (32 · rε)
where f is feed per revolution in mm and rε is the corner radius in mm. The result is in mm; multiply by 1000 for µm. The quadratic dependence on feed is the single most important fact in finishing: doubling the feed quadruples the theoretical roughness, while doubling the nose radius only halves it.
Table 1 — Theoretical Ra (µm) for Turning, by Feed and Nose Radius
| Feed f (mm/rev) | rε 0.4 mm | rε 0.8 mm | rε 1.2 mm | rε 1.6 mm |
|---|---|---|---|---|
| 0.10 | 0.78 | 0.39 | 0.26 | 0.20 |
| 0.15 | 1.76 | 0.88 | 0.59 | 0.44 |
| 0.20 | 3.13 | 1.56 | 1.04 | 0.78 |
| 0.25 | 4.88 | 2.44 | 1.63 | 1.22 |
| 0.30 | 7.03 | 3.52 | 2.34 | 1.76 |
| 0.40 | 12.50 | 6.25 | 4.17 | 3.13 |
| 0.50 | — | 9.77 | 6.51 | 4.88 |
Two practical warnings follow from this table. First, the theoretical value is a floor, not a target: a real insert with edge honing, runout and vibration will typically measure 20–60% above the calculated figure. Second, achieving Ra 0.8 µm with rε 0.4 mm requires a feed below 0.10 mm/rev, which is unproductive and often produces rubbing rather than cutting. The correct answer is a larger radius or a wiper edge, not an ever-smaller feed.
Choosing the Nose Radius for the Job
Nose radius is a compromise between finish, force and vibration:
- rε 0.2–0.4 mm — slender shafts, thin walls, small-diameter boring bars, high-chatter setups. Poor finish capability, low radial force.
- rε 0.8 mm — the general-purpose finishing choice for steel and stainless. Stable enough for f 0.15–0.25 mm/rev.
- rε 1.2 mm — rigid short parts, castings and interrupted cuts where a strong edge is needed; allows f 0.25–0.35 mm/rev at Ra 1.6–2.0 µm.
- rε 1.6 mm and above — roughing and heavy-duty finishing, or combined with wiper flats for high-feed finishing.
A useful rule is to keep the depth of cut at least half the nose radius, and preferably close to the full radius, so the chip is generated along the curved portion of the edge. A finishing pass of ap 0.05 mm on a 0.8 mm radius insert does not cut cleanly — it rubs, work-hardens the surface and produces worse Ra than a 0.3 mm pass.
Table 2 — Starting Finishing Parameters by Workpiece Material
| ISO group | Example material | Vc (m/min) carbide | Feed (mm/rev) | ap (mm) | Typical achievable Ra |
|---|---|---|---|---|---|
| P | 42CrMo4 / AISI 4140, C45 | 200–280 | 0.10–0.25 | 0.2–0.5 | 0.8–1.6 µm |
| M | 304 / 316L stainless | 140–200 | 0.08–0.20 | 0.2–0.4 | 0.8–1.6 µm |
| K | GJL 250 cast iron | 180–260 | 0.10–0.25 | 0.2–0.6 | 1.6–3.2 µm |
| N | 6061 / 7075 aluminium | 400–1000 | 0.10–0.30 | 0.2–0.5 | 0.4–1.6 µm |
| S | Ti-6Al-4V | 40–70 | 0.08–0.15 | 0.3–0.8 | 0.8–1.6 µm |
| H | 55 HRC tool steel (CBN) | 80–140 | 0.05–0.12 | 0.1–0.3 | 0.4–0.8 µm |
For steel, raising Vc from 160 to 240 m/min usually improves finish by suppressing built-up edge. Beyond the point where flank wear accelerates, Ra deteriorates quickly. In stainless and titanium, coolant pressure and direction often change Ra more than a 20 m/min speed change.
Wiper Inserts: Breaking the Feed-Roughness Trade-Off
A wiper insert carries a small flat or a second, much larger radius behind the main corner radius. This geometry does not generate the surface — it presses and smooths the feed marks left by the previous revolution. The effective radius can be two to four times the nominal rε, so the same feed produces a considerably lower Ra.
Wipers are not free performance. They require:
- Rigidity — the smoothing action needs stable depth of cut; a weak setup converts wiper contact into chatter.
- Sufficient ap — usually at least 0.3 mm, so the wiper flat is fully engaged. Below that, only the main radius works and the benefit disappears.
- Low runout — radial error should stay under 0.02–0.03 mm, otherwise one edge wipes and the other scrapes.
- Higher radial force — slender parts may deflect and produce taper even though the local Ra looks good.
Table 3 — Conventional vs Wiper Feed Capability at Equal Ra (rε 0.8 mm, steel)
| Target Ra | Conventional insert, feed | Wiper insert, feed | Productivity gain |
|---|---|---|---|
| 3.2 µm | 0.40 mm/rev | 0.65–0.75 mm/rev | +60–85% |
| 1.6 µm | 0.23 mm/rev | 0.40–0.50 mm/rev | +70–115% |
| 0.8 µm | 0.16 mm/rev | 0.28–0.35 mm/rev | +75–120% |
| 0.4 µm | 0.11 mm/rev | 0.18–0.22 mm/rev | +60–100% |
Values are starting estimates for rigid setups in ISO P steel at Vc 200–260 m/min; the actual gain depends on edge preparation and machine condition. On a flexible shaft, the conventional insert at low feed may in fact deliver a better and more consistent result than a wiper fighting deflection.
Milling: Feed per Tooth and Cusp Height
In milling, the surface is generated by the cutter rotation and the feed per tooth, not the feed per revolution of the spindle. The relevant relationship for a corner-radius or toroidal cutter is:
h ≈ fz² / (8 · rε) (per tooth, with the same rε convention as turning)
For a ball-nose cutter, the dominant feature is the step-over scallop:
h ≈ ae² / (8 · R)
where ae is the step-over and R is the ball radius. This is why ball-nose finishing is a trade-off between step-over density and cycle time: halving the step-over reduces cusp height by a factor of four.
Table 4 — Face Milling Finishing Guidelines (Ra, µm)
| fz (mm/tooth) | Corner rε 0.4 | Corner rε 0.8 | Corner rε 1.2 | Wiper insert |
|---|---|---|---|---|
| 0.05 | 0.6 | 0.4 | 0.3 | 0.2 |
| 0.08 | 1.6 | 0.8 | 0.6 | 0.3 |
| 0.10 | 2.5 | 1.3 | 0.9 | 0.4 |
| 0.15 | 5.6 | 2.8 | 1.9 | 0.8 |
| 0.20 | 10.0 | 5.0 | 3.3 | 1.3 |
Milling adds a factor that turning does not have: cutter runout and insert mismatch. One insert standing 0.02 mm proud of the others carries most of the chip load and leaves a periodic deep mark. On a 100 mm face mill with eight inserts, axial runout above 0.02 mm typically limits Ra to 1.6–2.5 µm regardless of feed reduction. Dial in every insert on a granite plate before chasing parameters.
Table 5 — Ball-Nose Cusp Height (approx. Rz, µm)
| Step-over ae (mm) | Ø6 (R3) | Ø10 (R5) | Ø16 (R8) | Ø20 (R10) |
|---|---|---|---|---|
| 0.05 | 0.10 | 0.06 | 0.04 | 0.03 |
| 0.10 | 0.42 | 0.25 | 0.16 | 0.13 |
| 0.20 | 1.67 | 1.00 | 0.63 | 0.50 |
| 0.30 | 3.75 | 2.25 | 1.41 | 1.13 |
| 0.50 | 10.4 | 6.25 | 3.91 | 3.13 |
| 0.80 | — | 16.0 | 10.0 | 8.00 |
Treat these values as Rz; apply an Rz/Ra ratio of about 4 for a clean ball-nose surface. A Ø10 cutter at 0.2 mm step-over therefore gives roughly Ra 0.25 µm in theory and 0.4–0.6 µm in practice.
Drilling, Reaming and Boring
Hole finishing follows a different logic. A drill generates its surface with the outer corner and the margin, and the resulting roughness is strongly influenced by margin contact, coolant and hole straightness rather than by feed alone. Reaming removes a small stock allowance with many flutes at low speed, producing a burnished, plateau-like surface. Boring with a single-point wiper insert is the most controllable route to a fine finish in a large bore.
Table 6 — Typical Achievable Ra by Hole-Making Operation
| Operation | Typical Ra | Typical parameters | Notes |
|---|---|---|---|
| HSS / carbide twist drill | 3.2–6.3 µm | f 0.10–0.25 mm/rev | Coolant and point geometry dominate |
| Indexable insert drill | 2.5–6.3 µm | f 0.08–0.20 mm/rev | Outer insert corner radius sets finish |
| Reaming (6–8 flutes) | 0.4–1.6 µm | Vc 8–20 m/min HSS, f 0.4–1.0 mm/rev | Stock 0.1–0.4 mm; never ream a rough hole |
| Single-point boring | 1.6–3.2 µm | f 0.08–0.15 mm/rev | Bar deflection limits finish on deep bores |
| Boring with wiper insert | 0.4–0.8 µm | f 0.15–0.30 mm/rev, ap 0.3–0.5 mm | Requires rigid bar, low runout |
| Fine boring / roller burnishing | 0.1–0.4 µm | Vc 100–180 m/min | Burnishing displaces rather than cuts |
Troubleshooting: What the Surface Is Telling You
| Symptom | Likely cause | Corrective action |
|---|---|---|
| Regular, evenly spaced feed marks too deep | Feed too high for radius, or wrong rε | Reduce f by 20–30%, or move to rε 1.2 / wiper |
| Periodic marks at tooth frequency | Insert runout or damaged edge | Re-index inserts, check axial runout < 0.02 mm |
| Smeared or torn surface on stainless | Built-up edge, speed too low, poor coolant | Raise Vc 20–30%, increase coolant pressure and aim at the edge |
| Chatter pattern, variable Ra along the part | Overhang, weak workpiece, radial force too high | Reduce rε, shorten overhang, lower ap, adjust speed to a stable lobe |
| Good Ra but visible waviness | Machine or workpiece deflection, not roughness | Measure with a longer cut-off; support the part; check tailstock pressure |
| Ra fine at the start, poor at the end of the pass | Progressive flank wear | Change grade or reduce Vc; check wear land at 0.2 mm VB |
How the Major Brands Differ in Finishing Geometry
All major manufacturers offer wiper geometries, but the way they are applied differs. The table below compares the practical finishing approach rather than catalogue numbers, since the exact chipbreaker designation varies with insert shape and size.
Table 7 — Finishing Geometry Comparison
| Brand | Finishing geometry | Typical grade family | Feed capability vs conventional | Practical note |
|---|---|---|---|---|
| Sandvik Coromant | Wiper inserts in CoroTurn 107 / 111 and CoroTurn 300 ranges | GC4325, GC4315, GC4425 | 1.6–2.0× | Very stable wiper flat; sensitive to ap below 0.3 mm |
| Iscar | Wiper chipformer inserts (W / WM designations) | IC8250, IC907 | 1.5–1.9× | Strong offering in small and medium insert sizes |
| Seco | Wiper inserts, Duratomic-coated | TP2501, TP3501 | 1.5–2.0× | Good balance of wiper feed and edge security in stainless |
| Walter | Wiper cutting edges on Tiger·tec Silver inserts | WPP05S, WPP10S | 1.5–1.8× | Benefits from high-pressure coolant in P steel |
| Kyocera | Wiper inserts with MEGACOAT NANO coating | PR1535, CA025P | 1.4–1.8× | Good wear resistance for long finishing runs |
| Tungaloy | Wiper inserts in turning and milling ranges | AH725, T9125 | 1.4–1.8× | Wide coverage from steel to cast iron |
These differences are smaller than the difference between a rigid and a flexible setup. A conventional rε 0.8 insert on a well-supported shaft can still reach Ra 0.8 µm at f 0.12–0.15 mm/rev, at the cost of cycle time.
Worked Example: 42CrMo4 Shaft, Target Ra 0.8 µm
Part: Ø80 × 400 mm shaft, 42CrMo4 (AISI 4140), 28 HRC, supported by tailstock, finishing pass.
Target: Ra 0.8 µm, Rz ≤ 4 µm.
Tool: CNMG 120408 wiper insert, rε 0.8 mm, PVD grade for steel.
Machine: 25 kW turning centre, high-pressure coolant available.
Starting point from Table 2: Vc 240 m/min, f 0.2 mm/rev, ap 0.4 mm. The conventional theoretical Ra at f 0.2 and rε 0.8 is 1.56 µm; a wiper typically reduces this by 50–60%, giving an expected 0.6–0.8 µm. If the first part measures 1.1 µm, the sequence should be:
- Check radial runout and insert seating; correct anything above 0.02 mm.
- Confirm ap is at least 0.3 mm — a 0.1 mm finishing pass will not engage the wiper flat.
- Increase coolant pressure and direct it at the cutting zone; BUE on 4140 is a common Ra limiter.
- Reduce feed to 0.16 mm/rev before touching speed; feed has the strongest and most predictable effect.
- Only then adjust Vc within 220–260 m/min, watching for chatter rather than chasing speed.
If Ra is acceptable but cycle time matters, the same insert at f 0.30 mm/rev lands near Ra 1.0–1.3 µm with the wiper — sufficient if the drawing allows Ra 1.6 µm, and roughly 50% faster than the 0.2 mm/rev pass.
Common Mistakes That Inflate Ra
- Trusting the formula on a flexible setup. The formula assumes zero vibration and perfect geometry. Real machines add 20–60%.
- Reducing feed instead of increasing radius. Feed reduction is quadratic but quickly reaches a rubbing regime where Ra stops improving and tool life collapses.
- Using a wiper with ap below 0.3 mm. The wiper flat never engages, and the insert behaves like a conventional tool with higher force.
- Ignoring runout. On multi-insert cutters and indexable drills, runout above 0.02 mm sets a hard Ra floor no parameter change can beat.
- Measuring along the feed marks. A stylus traced parallel to the grooves reports artificially low Ra. Always measure transverse to the dominant lay.
- Chasing finish with speed alone. Speed affects BUE and wear, but the geometric contribution comes from feed and radius. Fix geometry first.
Key Takeaways
- Ra ≈ f² / (32 · rε) for turning; feed is quadratic, nose radius is linear.
- Wiper inserts allow 1.5–2× higher feed at the same Ra, but demand rigidity, ap ≥ 0.3 mm and runout below 0.02–0.03 mm.
- In milling, use feed per tooth for corner-radius cutters and ae² / (8R) for ball-nose cusp height.
- Hole finishing is a separate discipline: drilling 3.2–6.3 µm, reaming 0.4–1.6 µm, wiper boring 0.4–0.8 µm.
- Measure Rz as well as Ra on sealing and fatigue-critical surfaces; the Rz/Ra ratio reveals whether the profile is plateau-like or torn.
- Insert brand matters far less than setup rigidity, edge condition and coolant delivery once the geometry is chosen correctly.
A finishing operation should be designed from the surface backwards: decide the required Ra and Rz, calculate the feed that geometry allows, choose the largest nose radius or wiper the setup can support, and then tune speed and coolant to reach the theoretical floor. That sequence turns surface roughness from a trial-and-error problem into a predictable part of the process plan.
Shop Related Products at HOOGUU
