How to Choose Ball Screw Diameter and Lead Together: A Selection Framework

Author:    Date:2026-10-08 

How to Choose Ball Screw Diameter and Lead Together: A Selection Framework
 

1. Two numbers, two different jobs

Diameter — specifically the root diameter, not the nominal size — governs whether the screw physically survives and stays stiff: load capacity, axial stiffness, buckling resistance, critical speed, and the screw's own rotary inertia.

Lead governs the translation between motor and axis: linear speed per revolution, thrust per unit of motor torque, positioning resolution, and how the load's inertia reflects back to the motor.

Neither list contains the other's items. A "1605" and a "1616" share a diameter yet can behave like two different machines on the same axis. That is why this has to be a joint decision.


2. What the diameter decides

Load capacity. A larger diameter carries larger balls and can run more load-bearing turns, so the ratings rise — but C and C0 are properties of the complete nut assembly (ball size, loaded circuits, groove geometry, contact angle). A larger nominal size does not guarantee a higher rating; check the model suffix and the catalog row, and take C and C0 from the datasheet for the exact nut model you are ordering.

Axial stiffness. From the bar formula k = A · E / L, with A = π · d_r² / 4 and E ≈ 206,000 N/mm² for steel. Stiffness scales with the square of the root diameter, and it falls with loaded length — so the same screw is stiffer near the fixed bearing and softer far from it, producing position-dependent error along a long stroke.

Buckling. A slender screw under compression can bow before anything else fails. The catalog Euler form is F_k = n · π² · E · I / L², with I = π · d_r⁴ / 64. Root diameter enters to the fourth power, unsupported length to the square, and the end-fixity factor n runs from about 0.25 (fixed–free) to 4 (fixed–fixed) — so the mounting alone changes buckling capacity by more than an order of magnitude. Buckling is a compression effect: a screw in tension does not buckle, but a reversing axis must be checked in its compression phase.

Critical speed. The bending-resonance limit: n_c = f_c · (d_r / L²) · 10⁷. Diameter enters linearly, length squared. The end-fixity set most commonly reproduced in modern catalogs is:

Mounting arrangement

f_c

Fixed – Free

3.4

Supported – Supported

9.7

Fixed – Supported

15.1

Fixed – Fixed

21.9

Two conventions circulate, and they differ by a factor of 0.8. Some catalogs publish these f_c values as bare beam-theory factors and treat 0.8 × n_c as the allowable speed; others publish factors that already include the 0.8 margin, so the resulting n_c is itself the allowable speed. Confirm which convention your manufacturer's catalog uses, and apply the margin once — never twice. This article follows the second convention: the values below are allowable operating speeds, with the 0.8 margin already inside the factors shown.

Rotary inertia. The screw's own inertia scales with the fourth power of diameter (J ∝ d⁴). Going up a size to fix a speed or stiffness problem also makes the screw harder to accelerate — which reappears later as motor torque. That is the quiet cost of "just use a bigger screw."


3. What the lead decides

Speed. The basic relation V = lead × rpm / 1000 (V in m/min, lead in mm). Lead is a transmission ratio, not a speed setting: for a target feed rate it fixes the rpm, and the rpm is what meets the critical-speed ceiling.

Thrust per unit torque. Drive torque is T = F · lead / (2 · π · η), with η ≈ 0.9 for a ball screw. Invert it and the point is clear — for a fixed motor torque, thrust falls as lead grows.

Resolution. Linear distance per motor revolution equals the lead. Finer lead, finer resolution for a given encoder — but on a modern servo the dominant errors usually come from lead accuracy, backlash, bearing play, guide alignment and thermal growth, not from the lead itself.

Acceleration and reflected inertia. Moving mass reflects to the motor as J_reflected = m · (lead / 2π)² — scaling with the square of the lead. A large lead multiplies the inertia the motor must accelerate, which is why axis inertia matching can shift suddenly when only the lead changes.

Heat and efficiency. Ball screw efficiency is high (about 90%), but the remaining losses become heat, and heat grows with speed. A small lead chosen to hit a feed rate carries a thermal cost over a long duty cycle.


4. The coupling: why the pair matters

Put the halves together and the two extremes are not variations on a theme — they are opposite machines built into the same four digits.

Large diameter + small lead — high thrust, high stiffness and buckling resistance, fine resolution. But the small lead forces high rpm for any real feed rate, pushing toward the critical-speed ceiling and dragging a heavy screw's inertia behind it. Natural fit: vertical or heavy, slow axes that must push hard and hold.

Small diameter + large lead — high speed at low rpm, low screw inertia, smaller nut and bearings, lower cost. But low thrust per torque, low stiffness, poor buckling resistance, coarser resolution. Natural fit: long, light, fast horizontal axes.

The failure mode is optimizing one in isolation. Choosing the lead purely for speed, then "fixing" the stiffness problem by going up in diameter, fixes nothing about torque — it can make it worse, because a bigger, heavier screw now accelerates through the same large lead. The two numbers must be sized against the same load case at once.


5. A selection sequence you can run

Order matters, because some checks are cheap to fail and others expensive to discover late.

Step 1 — Worst-case axial load. Not the payload — the force. Horizontal: acceleration force + friction + process force. Vertical: gravity + acceleration + process force, plus the load must be held when stopped. Include nut preload drag.

Step 2 — Motion profile. Target rapid speed, required acceleration, duty cycle. "As fast as possible" is not a number.

Step 3 — Accuracy and preload. Positioning tolerance, repeatability, preload class. These set the accuracy-grade end of the problem, not the diameter.

Step 4 — Speed → rpm, then the ceiling. For each candidate lead, rpm = V / lead. Screen against the critical speed n_c for your span and mounting, and against the nut's permitted speed (DN limit). This is where long axes usually eliminate a lead first.

Step 5 — Thrust → torque. Apply the worst-case force to T = F · lead / (2π·η), then add the acceleration torque from the reflected inertia relation in Section 3. Compare to the motor's continuous and peak ratings.

Step 6 — Diameter from load, buckling and stiffness — not speed alone. Between the surviving leads, pick the diameter that gives adequate buckling resistance and the stiffness the accuracy target implies. Verify life: L10 = (C / F)³ × 10⁶ revolutions, with a shock-load factor.

Step 7 — Iterate once. Changing diameter moves the critical speed and the inertia; changing lead moves the torque and the resolution. One pass is usually enough — but do the pass.

Quick allowable-speed check. For a Ø16 rolled screw over a 1,000 mm span with a class root diameter of about 13.2 mm (a class reference, not a specification), the term (d_r / L²) × 10⁷ ≈ 132 — giving an allowable speed of ≈ 1,990 rpm with fixed–supported ends and ≈ 2,890 rpm with fixed–fixed ends, using the f_c set above (the 0.8 margin is already inside those factors).


6. Application matrix (class reference)

A class reference for first-pass thinking, not a specification. Confirm the real combination against a load case and the datasheet.

Application

Typical diameter

Typical lead

What usually governs

Desktop / light gantry CNC

Ø12–16

4–10 mm

moving mass, cost, motor size

Pick-and-place, small robot

Ø16–20

5–10 mm

acceleration and cycle time

CNC machining centre, X/Y

Ø32–40

10–16 mm

stiffness and rapid speed together

CNC Z axis

Ø25–32

5–10 mm

buckling, holding torque, resolution

Semiconductor inspection / metrology

Ø16–25

5–10 mm

smoothness, straightness, repeatability

Medical and lab automation

Ø12–20

2–10 mm

mixed short-fast and long-gantry strokes

Long router / plasma / gantry

Ø40–50

20–25 mm

critical speed on long spans, buckling

Vertical press / lift

Ø32–50

5–10 mm

buckling and thrust

Typical industry ranges, shown for first-pass orientation only — they are not specifications. Confirm the diameter and lead against your load case and the manufacturer's datasheet before ordering.


7. Three common selection errors

Choosing diameter from the load rating alone. On a long axis the rating is often not the binding constraint — critical speed is. Illustrative example: a 1,000 mm axis sized for load lands on a Ø16 screw that is fine on capacity but out of its speed range at the required feed. The capacity check passed; the axis still fails.

Treating the lead as a pure speed knob. Lead also sets thrust per torque, resolution and reflected inertia. Illustrative example: an axis reaches its rapid speed but faults during acceleration, because the inertia reflected by a large lead was never included in the motor sizing.

Ignoring span and mounting when checking limits. Buckling and critical speed depend on the unsupported length — and that is not the overall screw length. For a preloaded nut it is the greatest nut-to-support distance that occurs during operation. Illustrative example: checking between the two bearings rather than the nut-to-support span overstates the allowable speed.

None of these is exotic — they are the Section 5 checks, skipped or mis-fed.


8. Worked example: a 1,000 mm horizontal axis at 20 m/min

The sequence run end to end. All figures are illustrative, to demonstrate the method.

Given: horizontal axis, linear-guide supported; moving mass 150 kg; stroke 1,000 mm, unsupported span ≈ 1,000 mm; target rapid 20 m/min; acceleration 0.3 g (≈ 2.94 m/s²); friction coefficient ≈ 0.02.

Step 1 — worst-case axial force.

  • Acceleration force: m·a = 150 × 2.94 ≈ 441 N
  • Friction: μ·m·g = 0.02 × 150 × 9.81 ≈ 29 N
  • Process force: assumed zero for a positioning axis
  • Running total ≈ 470 N, plus nut preload drag, taken from the datasheet for the preload class you select.

Step 2 — speed → rpm (20 m/min = 20,000 mm/min):

Lead

Required rpm

4 mm

5,000

5 mm

4,000

10 mm

2,000

16 mm

1,250

20 mm

1,000

25 mm

800

Step 3 — allowable-speed check. Ø16 class root diameter ≈ 13.2 mm — confirm the root diameter on the datasheet for the exact part ordered; (d_r / L²) × 10⁷ ≈ 132, so the allowable speed is:

  • Fixed–supported (f_c = 15.1): ≈ 1,990 rpm
  • Fixed–fixed (f_c = 21.9): ≈ 2,890 rpm

Screen the leads: 5 mm → 4,000 rpm, fails both mountings outright. 10 mm → 2,000 rpm, right at the fixed–supported limit and inside the fixed–fixed limit — workable, but only with fixed–fixed ends. 16 mm → 1,250 rpm, inside the fixed–supported limit — but see torque below.

Step 4 — thrust → torque (T = F · lead / (2π·η), η ≈ 0.9, F = 470 N):

Lead

Drive torque (≈)

Load reflected inertia J_reflected (≈)

10 mm

0.83 N·m

3.8 × 10⁻⁴ kg·m²

16 mm

1.33 N·m

9.7 × 10⁻⁴ kg·m²

Then add the acceleration term. Required angular acceleration α = a · 2π / lead is higher for the smaller lead (≈ 1,850 rad/s² at a 10 mm lead). Add the screw's own inertia — J = (π/32)·ρ·L·d_r⁴, or the datasheet value for your screw — and the coupling, then compare peak to the motor's peak rating and the RMS to its continuous rating. The lead that survives the speed check is not automatically the one that suits the motor.

Step 5 — the diameter decision. Two routes out of the corner the Ø16 screw is in:

  • Keep Ø16, raise the lead to 16 mm, mount fixed–fixed. Speed is fine (1,250 rpm), but torque rises to ≈ 1.33 N·m plus inertia, and resolution coarsens to 16 mm per motor revolution.
  • Go to Ø25 with a 10 mm lead. The larger root diameter lifts the ceiling: with a class root diameter ≈ 22 mm — confirm on the exact part's datasheet — and fixed–supported mounting, the allowable speed is ≈ 3,320 rpm, comfortably above the 2,000 rpm the 10 mm lead needs. Stiffness and buckling resistance improve too; the cost is a larger nut, larger bearings, more inertia and more space.

Route chosen (illustrative): Ø25 × 10 mm, fixed–supported, subject to a final L10 life check and a motor torque check. The Ø16 × 5 mm request from the opening paragraph is the wrong pair for this axis — not because it is weak, but because its lead forces 4,000 rpm through a 1,000 mm span.

(Illustrative example. Formulas and coefficients are general; every catalog-specific value must come from the datasheet for the exact part ordered.)


9. Frequently asked questions

Does a larger diameter always make for a better screw? No. It raises load capacity, stiffness and buckling resistance, but also rotary inertia and cost, needs a larger nut and bearings, and does nothing to reduce the torque a large lead demands. Size the diameter to load, buckling and stiffness — not as a blanket fix.

Can I choose the lead independently of the diameter? Not safely. They are coupled through critical speed, drive torque, reflected inertia and resolution. The lead fixes the rpm for your feed rate; the diameter decides whether the screw can turn that rpm without whipping.

What diameter do I need for a 1,000 mm stroke? No single answer — it depends on target speed and mounting. A slower axis can use a smaller diameter than a fast one over the same span, and a fixed–fixed mounting raises the allowable speed for the same shaft. Work the Section 5 sequence rather than picking a size.

How do I turn my target speed into an rpm? rpm = V / lead, with V in mm/min and lead in mm (or V in m/min × 1,000). If you need 20 m/min on a 10 mm lead, the screw must turn 2,000 rpm — and that must be inside the critical-speed limit for your span.

Should a vertical axis use a different diameter and lead? Yes. It must hold load against gravity and is subject to buckling, so it usually wants more mechanical advantage, a holding-torque check, and an explicit buckling calculation a horizontal axis does not need.


10. Data sources

  • Critical-speed formula and end-fixity factors: standard ball screw selection practice, widely reproduced in ball screw manufacturer catalogues and cited alongside ISO 3408 and JIS B1192; confirm the factor set and its margin convention against the catalogue for the screw you use.
  • Buckling (Euler form), axial stiffness and reflected-inertia relations: standard mechanical-design and ball screw selection formulas; E ≈ 206,000 N/mm² for steel.
  • Drive torque T = F · lead / (2π·η): standard ball screw selection guidance; η ≈ 0.9.
  • Life equation L10 = (C/F)³ × 10⁶ revolutions, load factor, dynamic C and static C0: ISO 3408-5 (load ratings and operational life); accuracy grades: ISO 3408-3.
  • Root diameters, ratings and permitted speeds in the worked example are class references. Take every catalog-specific value from the datasheet for the exact screw ordered.

This article is general technical guidance, not a substitute for a project-specific calculation. All values must be confirmed against the respective manufacturer's official specifications before use in design.

Last updated: 2026-10-08


11. Sizing a specific axis?

If you are working through diameter and lead together for a particular stroke, load and speed target, we can help you check the pair against the actual screw and mounting you plan to use — including custom lengths and end machining. See our ball screws  with your stroke, moving mass, target speed, acceleration and mounting arrangement.

top