A smaller thread pitch gives a round nut less axial travel for the same rotation, making finer nominal adjustment possible. It does not, by itself, guarantee a more accurate final position. The improvement is useful only when backlash, friction, thread geometry, structural deflection, and locking movement remain below the required positioning tolerance.

The essential distinction is between resolution, the smallest movement that can be commanded or read, and accuracy, how closely the final position matches the intended position. A fine thread round nut can offer excellent calculated resolution while delivering poor repeatability if it changes thread-flank contact, moves unevenly, or shifts when secured.

Pitch sets the displacement ratio—not the positioning tolerance

For a single-start thread, one complete revolution produces an ideal axial movement equal to the pitch. The displacement associated with a partial rotation is:

Δx = p × θ / 360°

Here, Δx is axial displacement, p is thread pitch, and θ is rotation in degrees. If pitch is expressed in millimetres, the calculated displacement is also in millimetres.

These are geometric calculations, not measured performance values. Reducing pitch from 1.00 mm to 0.50 mm halves the axial effect of any given angular movement. It also halves the displacement caused by a small angular setting error, provided the thread remains engaged against the same flank and the assembly moves without sticking.

Strictly, lead determines travel per revolution. Lead equals pitch for a single-start thread; for a multiple-start thread, lead equals pitch multiplied by the number of starts. An adjustment calculation based only on pitch is therefore incorrect unless the thread-start arrangement is known.

The useful pitch depends on how rotation is controlled

A pitch becomes “fine enough” only in relation to the angular movement that the mechanism can reliably produce. A hand-operated hook spanner, a graduated adjustment ring, and a geared actuator provide very different levels of angular control.

Suppose a mechanism must support nominal 5 µm adjustment increments, and its smallest reliably controlled rotation is 3°. For a single-start thread:

p ≤ 360° × 0.005 mm / 3° = 0.60 mm

A 0.50 mm pitch satisfies that geometric requirement. A 1.00 mm pitch produces approximately 8.33 µm for the same 3° rotation. However, this calculation allocates the entire 5 µm increment to angular control; it does not establish a ±5 µm positioning accuracy.

Tool access can be as important as graduation spacing. If a spanner must be repeatedly disengaged and repositioned, its nominal angular movement may be difficult to reproduce. Conversely, a larger pitch combined with controlled angular actuation and direct displacement feedback can outperform a finer thread adjusted by feel.

Backlash breaks the link between rotation and movement

Thread clearance allows the nut to rotate through some angle after reversal before the opposite flanks take up load. During this interval, rotation does not produce the expected axial displacement. This is backlash, and it is especially important when adjustment alternates between advancing and retracting.

Making pitch smaller does not automatically eliminate the axial clearance. If an assembly has an assumed 10 µm of axial backlash, its equivalent lost rotation is:

θ_b = 360° × b / p

That clearance corresponds to 3.6° at a 1.00 mm pitch and 7.2° at a 0.50 mm pitch. The finer thread increases the angular dead band associated with the same axial clearance; it does not reduce the clearance itself.

A consistent final approach from one direction can improve repeatability by maintaining contact on the same thread flank. Where the operating load can reverse, that procedure alone is insufficient. A spring bias, opposed nuts, or another preload arrangement may be needed to preserve contact throughout the operating cycle.

Preload brings its own constraints. It increases contact loading and can increase adjustment torque, wear, and heat generation. Its value must be sufficient to prevent flank separation under the relevant load, without creating unacceptable friction or deformation.

Friction determines whether fine rotation becomes smooth motion

The displacement equation assumes continuous sliding. Real threads can exhibit stick-slip: torque builds while the interface remains stationary, then stored elastic energy is released as a sudden movement. A small commanded rotation may therefore produce no measurable displacement, followed by a jump larger than the intended increment.

Surface finish, material pairing, lubricant, contamination, contact pressure, and adjustment speed all influence this behaviour. Finer pitch reduces nominal travel per degree, but it does not ensure that the smallest actual movement decreases proportionally.

A simplified, frictionless screw relationship illustrates why torque also needs attention:

T = F × L / (2π)

Here, T is torque, F is axial force, and L is lead. A smaller lead provides greater ideal mechanical advantage. Actual adjustment torque also includes thread friction and, where the nut bears against a stationary surface, bearing-face friction. These contributions can be substantial.

Consequently, wrench torque is not a reliable substitute for position measurement. The same applied torque can produce different axial forces as friction changes, and different axial forces can produce different elastic deflections even when the nominal nut position is unchanged.

Thread tolerances do not establish complete adjustment accuracy

A specified thread size and tolerance class establish requirements for the mating thread geometry. They do not, on their own, certify the displacement accuracy of the assembled mechanism.

ISO 261 addresses the general plan for ISO general-purpose metric screw threads, while the ISO 965 series addresses tolerances for ISO general-purpose metric screw threads. These references help define thread compatibility and permissible dimensions. They should not be interpreted as a complete precision-positioning specification.

A tighter fit can reduce available clearance, but it does not necessarily provide a controlled relationship between rotation and axial travel over the required adjustment range. Local pitch variation, accumulated lead error, and variation around the circumference can affect that relationship.

The nut’s bearing face also matters. If it is not suitably perpendicular to the thread axis, rotating the nut can change the contact pattern against the supported component. The resulting tilt or seating movement can exceed the displacement calculated from a small angular increment. Thread-axis alignment, mating-shaft geometry, and the stiffness of the supported component must therefore be evaluated together.

Where micrometre-level adjustment is required, the drawing needs requirements beyond “fine thread”: the relevant displacement error, bearing-face geometry, datum relationships, and verification conditions must be defined explicitly.

A nut can control position—or create preload

The meaning of axial adjustment changes when a round nut contacts a bearing ring, spacer stack, or another constrained component. Before contact, rotation may produce largely free axial travel. After contact, further movement is divided among clearance take-up, contact seating, and elastic deformation.

For a simplified linear load path:

ΔF = k_eq × Δx

Here, k_eq is the equivalent axial stiffness of the loaded assembly. A small additional advance can produce a large force change in a stiff stack. A finer pitch can make that force change easier to control through rotation, but it does not establish the resulting preload without information about stiffness, seating, and friction.

Bearing arrangements require particular care because their stiffness can vary with load. The same angular increment need not produce the same preload increment throughout adjustment. If the functional requirement is bearing preload or endplay, acceptance should be based on that quantity rather than nut angle alone.

Position under operating load can also differ from position during adjustment. In a linear approximation, a load change ΔF produces displacement Δx = ΔF/k_eq. A finer thread does not remove this compliance error. Nor does it remove thermal movement of the shaft, housing, or spacer stack.

Locking can impose a coarser limit than the thread

An accurately adjusted nut is useful only if it retains its position after locking. Jam nuts, tab washers, radial locking screws, and clamping features interact differently with the adjustment interface. They can introduce additional rotation, alter flank loading, or change the bearing-face contact.

For a locking arrangement with N equally spaced permissible angular positions, the nominal axial spacing between lock positions is:

Δx_lock = p / N

With a 0.50 mm single-start pitch and 12 permissible positions per revolution, the spacing is approximately 41.7 µm. That is much coarser than the 1.39 µm nominal travel per degree. This calculation applies to the effective locking positions of the actual design, which may differ from the visible number of slots or tabs.

Continuously positionable locking avoids discrete indexing, but it still needs verification for locking-induced movement. Position should be checked after the specified locking sequence, not only while the nut remains free to adjust.

Specify and test the assembled result

A meaningful adjustment requirement identifies the controlled quantity: unloaded axial position, loaded position, endplay, or preload. It also distinguishes absolute accuracy from repeatability. A mechanism can return consistently to the same position while retaining a systematic error relative to the nominal displacement equation.

Verification should measure the functional output against a suitable reference, with instrument resolution and measurement uncertainty appropriate to the required tolerance. Measuring a convenient exposed nut face may miss movement or deformation elsewhere in the load path.

An informative test records displacement through the working adjustment range, includes reversals to reveal backlash, and repeats the specified approach-and-lock sequence. Load, lubrication, temperature, and settling conditions should be controlled or recorded. If service loads reverse, testing only under a constant one-direction load can conceal the most important source of error.

The most useful pitch is therefore not automatically the smallest available. It is the pitch that provides sufficient angular controllability while leaving an achievable margin for reversal error, friction, geometric variation, deformation, and locking shift. Once those effects dominate the error budget, reducing pitch adds turns and manufacturing demands without necessarily improving the final axial position.