A double ball nut uses two nut halves to create preload, but it does not double axial load capacity: under one-directional thrust, one preloaded half normally carries the applied load [S3].
The practical choice depends on load direction, required stiffness, travel length, speed, and the acceptable loss of preload with wear. This guide applies to recirculating ball screws with matching helical grooves and closed ball paths [S1].
Single-circuit and double-circuit nut architecture
A ball screw converts rotation into linear travel through rolling contact between balls and matching screw and nut raceways. The balls circulate through the nut and return system in a closed path as either the screw or nut rotates [S1]. The nut controls the assembly’s load capacity and life because its number of active threads relative to the screw determines when the nut reaches fatigue failure before the screw [S1].
A single-circuit nut places one ball circuit around the screw. It is mechanically compact and less expensive, and its circuit can carry axial load in both directions [S6]. A double nut consists of two single nuts tensioned against each other to establish preload [S3]. The terminology must be handled carefully: a two-start screw with the same lead as a one-start screw has twice as many rolling elements but is not automatically a double-nut preload arrangement [S2].
Ball screws are appropriate where high thrust, precision positioning, continuous operation, or high speed is required. Their rolling contact can deliver efficiency above 90%, compared with roughly 40% input-power transfer for a conventional sliding ACME screw [S4]. A lower-cost lead screw can be adequate where efficiency, wear life, and high-speed capability are not dominant requirements [S4][S7].
Preload changes stiffness and failure behavior
Preloaded double nuts create two-point ball-to-raceway contact, while a preloaded single nut has four-point contact. Four-point contact can permit differential slip or partial sliding, increasing heat and contributing to premature wear under some loading conditions [S3]. This contact geometry is a central reason to specify a double nut when backlash, elastic deflection, and continuous-load behavior matter more than envelope size.
Three preload methods are established for a double-nut assembly: a spacer ring between the halves, a spring between the halves, and a manufacturing-induced lead offset [S3]. A spacer sets preload from its thickness and can produce consistent friction torque when the spacer and nut faces are machined to high tolerance, but a non-flat spacer or a non-perpendicular interface can make the ball contact angle inconsistent [S3].
A spring removes the spacer-flatness problem and can compensate for wear because it changes position as the nut halves deflect. Its limit is the spring’s maximum thrust capacity: once the spring stiffness is overcome and the spring collapses, preload is lost [S3]. One method for inducing preload in a double ball nut is creating an offset in the lead.
Decision criteria for single versus double nuts
Compare the designs on axial capacity, motion quality, envelope, cost, and direction reversal rather than assuming that two nuts mean twice the performance. A double nut adds parts, preload controls, assembly length, and an additional nut interface, while a single nut reduces package length and procurement complexity [S3][S6].
Single nuts have ball circuits that carry load in both directions and are generally more compact and less expensive. Double ball nuts do not have higher load capacities than single ball nuts because, when preloaded, the tension between the nuts causes only one nut to take the load from a given direction.
Matching the nut to load direction and motion profile
For bidirectional axes, confirm that the selected single-circuit nut is rated for axial loading in both directions. Do not infer bidirectional capacity from a preloaded double nut, because the two halves react against one another and a one-direction external force primarily loads one half [S3]. This makes the double nut different from simply adding parallel load paths.
For a continuously loaded positioning axis, the double nut reduces clearance and deflection, but the selected preload method still sets the operating limit. Spacer preload requires controlled machining and flatness; spring preload is limited by collapse; lead offset trades active grooves and dynamic life for a shorter housing [S3]. If the axis is exposed to reversing loads, include reversing friction, heat, and preload stability in the acceptance test rather than selecting from thrust capacity alone.
For long-travel axes, separate nut preload from screw buckling and support stiffness. A related engineering analysis of ball screw buckling limits for long travel axes is required because nut selection cannot correct excessive screw deflection. Likewise, the screw and nut should be evaluated as an assembly, not as independent components [S1].
Speed, accuracy, and life constraints
Double-nut construction does not remove this shaft-dynamics limit.
Accuracy selection should be tied to positioning error rather than nut count alone. The referenced product range includes C0 through C10 accuracy grades and screw diameters from 4 mm to 80 mm, illustrating that accuracy and size are separate procurement variables [S4]. The ball screw should be specified with its accuracy grade, diameter, lead, stroke, lubrication method, support arrangement, and required life rating as one package.
Ball return architecture also affects performance. External return tubes increase radial envelope, internal return systems conserve space, and tangential return systems are used where smooth ball flow, high speed, or high loading is required [S1]. Life calculations for a nut normally account for the number of loaded balls and raceway geometry, while the return-system material must be considered separately [S1].
Preload selection method and factory checks
Specify the target preload, the expected axial-load range, the motion direction, allowable friction torque, and the required service life before choosing the nut architecture. For a spacer design, the spacer thickness controls preload, so define spacer thickness tolerance, nut-face perpendicularity, and the permitted friction-torque band [S3]. For a spring design, define the permissible external force and confirm that the spring cannot collapse within the load range [S3].
For lead offset, confirm the actual lead variation across the nut and retain the manufacturer’s active-groove and dynamic-life calculation [S3]. The terms single circuit, double circuit, double nut, two start, and double start must be placed on the drawing with the actual geometry described, because a two-start screw with the same lead is not equivalent to a preloaded double nut [S2][S3].
Factory acceptance should verify no-load friction torque, backlash or lost motion in both directions, axial rigidity, running resistance, and ball return operation. Repeated full-stroke cycling can reveal sticking, uneven preload, heat, or abnormal torque. The purchase specification should identify the screw grade, nut configuration, preload method, accuracy grade, lubrication, support bearings, and calculated dynamic life so that the linear actuator supplier is not left to substitute an equivalent architecture.
For most precision, one-direction-loaded axes, specify a double nut with a controlled spacer or spring preload. For compact, lower-cost, reversible axes with modest stiffness requirements, specify a single nut and verify its bidirectional rating. Track measured friction torque and lost motion across the duty cycle, because preload retention, not the label “double nut,” determines long-term performance.