Concave-grooved Pulley for Round Band: How It Works, Groove Sizing, Diagram & Uses

← Back to Engineering Library

A concave-grooved pulley for round band is a pulley whose rim carries a semicircular groove sized to cradle a round cross-section belt — typically a polyurethane or leather round cord. The profile dates to 19th-century textile machinery, with Singer adopting it widely on treadle sewing machines from the 1850s onward. The concave groove seats the belt by shape rather than by wedge friction, so the cord rolls into the groove and tracks naturally without climbing the flange. You see this drive on sewing machines, lab centrifuges, hobby lathes, and any low-power application below about 0.25 kW where simple, quiet, slip-tolerant transmission matters more than peak torque.

Concave-grooved Pulley for Round Band Interactive Calculator

Vary round belt diameter and groove sizing ratios to see the recommended concave groove radius, depth, and clearance.

Cord Radius
--
Groove Radius
--
Groove Depth
--
Radius Extra
--

Equation Used

r = d / 2; Rg = 1.05r; h = 0.7d

The calculator follows the article sizing rule for a concave-grooved pulley: first take the cord radius as half the belt diameter, then make the groove radius about 1.05 times that radius. Groove depth is shown as a fraction of cord diameter, with 0.7d matching the worked example guidance.

  • Round cord belt seats in a concave circular groove.
  • Default sizing follows the worked example: 4 mm belt gives Rg = 2.1 mm.
  • Groove depth is a tracking guideline, not a torque capacity calculation.
Concave Grooved Pulley Cross-Section Diagram Technical diagram showing the cross-section of a concave grooved pulley with a round belt seated in the semicircular groove. Rg 0.7d A A SECTION A-A Belt tracks in groove Concave groove Round cord belt Contact arc Friction surface Key Dimensions Groove radius Rg = 1.05 × cord radius Example: 4mm belt → Rg = 2.1mm Optimal depth: 0.7 × cord diameter Rg = 1.05r (r = cord radius) Why Sizing Matters: Too tight → cord pinches, overheats Too loose → point contact, rapid wear 1.05× ratio ensures proper wrap contact
Concave Grooved Pulley Cross-Section Diagram.

The Concave-grooved Pulley for Round Band in Action

The concave groove is cut as a half-circle whose radius matches the belt cord radius — the belt sits in the groove like a marble in a spoon. Drive force transfers through plain friction between the cord surface and the curved groove wall, not through wedging the way a V-belt does. That distinction matters: if the groove radius is too tight, the cord pinches and overheats; if it's too loose, the cord rides on a single contact line and wears flat in months. We size groove radius to roughly 1.05 × cord radius — so for a 4 mm polyurethane round belt, the groove radius wants to be 2.1 mm, never 2.0 and never 2.5.

Groove depth controls how the belt tracks. A shallow groove — depth less than 0.6 × cord diameter — lets the belt jump the rim under shock load or misalignment. A deep groove past 0.9° cord diameter traps the belt and adds bending loss every revolution, which you feel as heat and a measurable drop in efficiency on long runs. The sweet spot sits at roughly 0.7 × cord diameter, which is what you'll find on most Singer, Pfaff, and Juki sewing machine handwheel pulleys.

The common failure modes are predictable. Glazing happens when the cord slips under sudden load and the polyurethane melts a thin shiny layer onto the groove — once it glazes, friction drops and slip becomes permanent. Groove ovalisation shows up after years on hard-driven lathes where one quadrant of the groove sees most of the wrap angle. And cord-set takes hold if the belt sits tensioned for months without rotation, leaving a bend memory that slaps audibly every revolution. Round belts tolerate misalignment better than flat or V belts, but only up to about 2° of shaft skew — past that the belt walks out of the groove on the slack side.

Key Components

  • Pulley Hub & Bore: The hub carries the bore that fits the shaft. For a 6 mm shaft the bore tolerance wants to be H7 — 6.000 to 6.012 mm — paired with a setscrew on a flat or a keyway. Loose bores are the number-one cause of pulley wobble that shows up as a once-per-rev belt thump.
  • Concave Groove Profile: A semicircular groove cut into the rim, radius 1.05 × cord radius. For a 4 mm round belt the groove radius is 2.1 mm. The groove finish should be Ra 0.8–1.6 µm — too smooth and the belt slips, too rough and the cord abrades.
  • Rim Flanges (when present): Some concave-grooved pulleys add a small lip on each side of the groove to prevent belt walk-off during start-up shock. Flange height typically 0.3–0.5 mm above the groove edge — enough to catch the belt, not enough to cut it.
  • Round Cord Belt: The mating element — usually 85A–95A shore polyurethane, sometimes leather on antique machinery. Standard diameters are 3, 4, 5, 6, and 8 mm. Belt diameter must match groove radius within 0.1 mm or contact geometry collapses to a point.
  • Crown or Flat Bottom: The groove bottom is either a true semicircle or a slightly flattened arc. A flattened bottom increases contact patch area by about 15% but reduces self-tracking — a design tradeoff most sewing machine builders resolve in favour of the true semicircle.

Industries That Rely on the Concave-grooved Pulley for Round Band

Concave-grooved round-belt pulleys show up wherever you need quiet, low-power, slip-tolerant drive between shafts that aren't perfectly aligned. The geometry handles quarter-turn drives — input shaft horizontal, output vertical — better than any other pulley type, which is why textile mills used round belts for spindle banks for over a century. Below about 0.25 kW the round-belt-and-concave-pulley combination beats V-belts on cost, beats timing belts on noise, and beats flat belts on tracking. Push past 0.5 kW and you'll start cooking the belt — the cord can't dump enough heat to survive sustained higher loads.

  • Sewing Machines: Singer 201, Pfaff 130, and Juki DDL-8700 industrial lockstitch machines all use a concave-grooved handwheel pulley driving a 4 or 5 mm polyurethane round belt to the motor pulley below the table.
  • Hobby Lathes: Sherline 4400 and Taig Micro Lathe headstocks use a concave-grooved spindle pulley with a 4 mm urethane round cord, giving 70–2800 RPM range with quick belt-position changes.
  • Laboratory Centrifuges: Older IEC and Sorvall benchtop centrifuges drove the rotor head from a fractional-horsepower motor through a round-belt-and-concave-pulley pair — the slip behaviour acted as a built-in overload protection.
  • Textile Spinning: 19th and early 20th century cotton spinning frames drove vertical spindle banks from a horizontal line shaft through quarter-turn round-cord drives over concave-grooved tin-plated wood pulleys.
  • Office & Business Machines: Mechanical typewriters, early adding machines, and Teletype Model 15 carriages used small concave-grooved pulleys with 2–3 mm round belts to transfer motion between offset shafts.
  • Vending & Conveyor Tooling: Light-duty product-orienting conveyors at confectionery lines, like those at Nestlé tabletop chocolate-wrapping cells, use round-belt drives over concave pulleys to handle frequent jams without snapping the belt.

The Formula Behind the Concave-grooved Pulley for Round Band

The driving relationship for a round-belt concave-pulley system is the capstan equation, which tells you the maximum torque you can transmit before the belt slips. Tension ratio depends on the friction coefficient between cord and groove, and the wrap angle around the pulley. At the low end of typical operating range — 90° wrap, dry urethane on aluminium, μ ≈ 0.5 — you'll get a tension ratio of about 2.2, plenty for a sewing machine but borderline for a hobby lathe under cut load. At the high end — 270° wrap on an idler-tensioned drive — the ratio jumps past 30, and the belt acts almost as if it's pinned. The sweet spot for most builds is 180° wrap with μ around 0.4, giving a ratio of 3.5 and a margin against momentary load spikes without cooking the belt with heat.

T1 / T2 = e(μ × θ)

Variables

Symbol Meaning Unit (SI) Unit (Imperial)
T1 Tight-side belt tension N lbf
T2 Slack-side belt tension N lbf
μ Coefficient of friction between cord and groove dimensionless dimensionless
θ Belt wrap angle around the pulley rad rad
e Base of natural logarithm (≈ 2.718) dimensionless dimensionless

Worked Example: Concave-grooved Pulley for Round Band in a benchtop watchmaker's polishing spindle

You are building a benchtop watchmaker's polishing spindle for a small horology workshop in Geneva. The motor sits below the bench at 1400 RPM, the polishing spindle is mounted on the bench top with a 60 mm concave-grooved aluminium pulley, and you're driving it with a 4 mm polyurethane round belt over a 180° wrap. You need to know what tight-side tension you can run before the belt slips at 0.4 N·m of polishing load.

Given

  • Dpulley = 60 mm
  • dbelt = 4 mm
  • μ = 0.4 dimensionless
  • θnom = 180 (π rad) degrees
  • Torque = 0.4 N·m

Solution

Step 1 — at the nominal 180° wrap, calculate the tension ratio from the capstan equation:

T1 / T2 = e(0.4 × π) = e1.257 ≈ 3.51

Step 2 — the torque transmitted equals the difference in belt tensions times the pulley radius. With pulley radius 0.030 m and required torque 0.4 N·m:

T1 − T2 = 0.4 / 0.030 = 13.33 N

Combined with the ratio from Step 1, T2 = 13.33 / (3.51 − 1) = 5.31 N, and T1 = 18.64 N. That's the nominal operating point — comfortable for a 4 mm urethane cord rated to about 80 N working tension.

Step 3 — at the low end of the typical range, 90° wrap (a quarter-turn drive), the ratio collapses:

T1 / T2 = e(0.4 × π/2) ≈ 1.87

Now T1 climbs to 28.7 N to deliver the same 0.4 N·m — still within belt rating but you've cut your slip margin in half. At the high end, 270° wrap with an idler:

T1 / T2 = e(0.4 × 3π/2) ≈ 6.59

The required T1 drops to 15.7 N, and the belt could in theory handle five times the polishing load before slipping. In practice though, the extra wrap from an idler bends the cord through more flexion cycles per revolution and you'll halve belt life. 180° is the sweet spot.

Result

Run the belt at a tight-side tension of about 18. 6 N — roughly the pull of a 1.9 kg weight hung off the cord. That tension feels firm but not drum-tight when you pluck the belt; if it sings a clear note above 100 Hz on a 200 mm span, you're over-tensioning. The 90° quarter-turn variant works but gives you almost no slip margin under polishing-stick snags, while the 270° idler-wrapped version transmits more torque at the cost of halving belt life from cord flexion. If you measure persistent slip below the predicted 0.4 N·m capacity, check three things in order: (1) groove glazing — look for a shiny ring around the pulley which drops μ from 0.4 to under 0.2; (2) bore-to-shaft slop letting the pulley wobble and shed wrap angle on each rev; (3) belt cord-set from sitting under tension overnight, which you'll feel as a once-per-rev pulse in the spindle.

Choosing the Concave-grooved Pulley for Round Band: Pros and Cons

Round belt and concave pulley sits in a specific niche — quiet, cheap, slip-tolerant, low-power. Compare it honestly against V-belts and timing belts and you'll see when to pick which.

Property Concave-Grooved Pulley + Round Belt V-Belt Drive Timing Belt Drive
Power capacity Up to ~0.25 kW practical, 0.5 kW absolute 0.5–150 kW typical 0.1–100 kW typical
Speed ratio accuracy ±2–5% (slip) ±1–2% (slip) 0% (positive engagement)
Maximum belt speed 20 m/s 30 m/s 60 m/s
Tolerance to shaft misalignment Up to 2° skew, handles quarter-turn drives 0.5° max, parallel shafts only 0.25° max, parallel shafts only
Cost per metre of belt $2–8 (urethane cord) $5–25 (V-belt) $15–80 (toothed belt)
Noise at 1500 RPM 50–55 dB(A) 60–70 dB(A) 70–80 dB(A)
Typical lifespan 2,000–8,000 hr 5,000–25,000 hr 10,000–40,000 hr
Best application fit Sewing machines, hobby lathes, small instruments Industrial fans, compressors, machine tools CNC, robotics, timing-critical drives

Frequently Asked Questions About Concave-grooved Pulley for Round Band

Round belts self-track on a concave groove only when the groove radius matches the cord radius within about 0.1 mm. If the groove is too wide — say a 2.5 mm groove on a 4 mm cord (radius 2.0 mm) — the belt rides at one edge of the groove and the asymmetric contact pushes it sideways every revolution. It walks until it hits the flange or jumps the rim.

Check groove width with a pin gauge or a fresh piece of cord pressed into the groove. If the cord seats below the rim by less than 0.6 × its diameter, the groove is too shallow or too wide and re-cutting the pulley is the only fix. Shaft parallelism matters too, but it's the second thing to check, not the first.

No — and the failure mode is sneaky. A 5 mm cord in a groove cut for 4 mm rides on the rim edges instead of seating into the groove. You get high contact stress on two narrow lines, the cord deforms into an oval cross-section under tension, and within 50–100 hours you'll see flat-spotting on the belt and accelerated groove edge wear.

The right move is to either change the pulley to a matching 5 mm groove or stay with the 4 mm cord and add wrap angle with an idler if you need more torque capacity.

For 0.18 kW with frequent speed changes by moving the belt to different pulley diameters, the round belt wins on convenience — you can pop the belt off and reposition it in seconds, which is exactly why Sherline and Taig went that route. A V-belt at that power is overbuilt and a pain to swap.

If the mill runs at one fixed speed and you cut steel rather than brass and aluminium, a 3L-section V-belt gives you better slip margin under interrupted cuts and lasts 3–5× longer. Decide on cut interruption frequency, not headline power.

Intermittent slip with theoretical margin almost always traces to dynamic effects the static capstan equation doesn't capture. The two big ones are: belt cord flex stiffness fighting the wrap (especially on small pulleys under 30 mm diameter), which lifts the belt off the groove on entry and exit, and pulley runout converting steady tension into a pulsing tension that briefly drops below the slip threshold every revolution.

Quick diagnostic: dust the belt with chalk and run it for 30 seconds. If the chalk wears off in a band rather than uniformly, you have a runout or alignment problem. If it wears off uniformly but the belt still slips, the cord is too stiff for that pulley diameter — go down a cord size or up a pulley size.

Glazing leaves a visible shiny band on the cord and a matching shiny ring inside the pulley groove — under a loupe the surface looks like polished plastic instead of the matte texture of fresh urethane. Stretch shows up as the belt simply being looser, with no surface change.

The fix differs. Stretched belts can be shortened (cut and re-weld for spliced belts, or replaced for endless belts). Glazed belts and grooves cannot be reliably restored — the heat-affected layer is harder than the surrounding rubber and even sanding it off changes cord diameter outside the 0.1 mm tolerance window. Replace both belt and pulley together once glazing has set in.

A round belt is symmetric in cross-section, so it doesn't care which way it twists between pulleys — the cord rolls into the concave groove the same regardless of approach angle. A V-belt has a defined top and bottom; twist it 90° between pulleys and the wedge faces present diagonally to the V-groove, contact patch collapses, and the belt either jumps out or burns through in hours.

This is exactly why every 19th-century textile mill ran round cotton or leather cords on quarter-turn drives off the line shaft to vertical spindles, and why anyone restoring that machinery should not substitute a V-belt no matter how convenient the modern pulley looks.

References & Further Reading

  • Wikipedia contributors. Round belt. Wikipedia

Building or designing a mechanism like this?

Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.

← Back to Mechanisms Index
Share This Article
Tags: