• Add description, images, menus and links to your mega menu

  • A column with no settings can be used as a spacer

  • Link to your collections, sales and even external links

  • Add up to five columns

  • Front Hub Influence On Inverted Forks

    Mountain bikers riding through a forest with dust and sunlight.

    Abstract

    Bicycle front wheel hubs appear simple from the outside, yet their internal architecture has a significant influence on durability, bearing life, ease of service, manufacturing tolerance stack-up, and ride performance. A few key differentiators among modern hub designs are how bearing preload is established and maintained, and how end caps interface with the axle and fork. These differentiators are even more apparent when the hub is used in conjunction with an inverted fork. The inherent architecture of the inverted fork places more emphasis on the load path within the hub which translates directly to ride performance. This whitepaper outlines and compares four common front hub design types used in contemporary mountain, gravel, and road applications, focusing on preload strategy, structural implications, serviceability, and trade-offs.

    AUTHOR: Matt White (Engineering Team)

    Section 1: Design Overview

    The chart below is an overview of the 4 most common architectures of front MTB hubs. In the sections below the key characteristics, advantages and limitations will be reviewed in more depth.

    1. Internal Spacer & Internal Nested End Caps
    2. Internal Spacer & Flush End Caps
    3. Internal Axle & External Nested End Caps
    4. Solid Axle with External Preload Collar

    Internal Spacer & Internal Nested End Caps

    Internal Spacer & Flush End Caps

    Internal Axle & External Nested End Caps

    Solid Axle with External Preload Collar

    Section 2: Glossary of Terms

    2.1 Bearing Load Types

    Radial Load
    A force acting perpendicular to the axle. In a bicycle hub, radial loads are generated by rider weight, impacts, cornering forces, and braking reactions transmitted through the wheel. Radial loads are the primary operating load for hub bearings.

    Axial Load
    A force acting parallel to the axle. Axial loads arise from bearing preload, fork clamping force, and tolerance stack-up within the hub assembly. Depending on hub design, axial loads may be carried through the bearings or reacted directly by the axle.

    Combined Load
    The simultaneous presence of radial and axial loads acting on a bearing. Most bicycle hub bearings operate under combined loading during real-world use.

    2.2 Parts of a Rolling Element Bearing

    Inner Race
    The bearing race that interfaces with the axle or preload system. In most hub designs, axial preload is applied to the inner races.

    Outer Race
    The bearing race that is press-fit into the hub shell. The outer race transmits radial loads from the hub shell into the bearing.

    Rolling Elements
    Balls (in typical cartridge bearings) that roll between the inner and outer races, allowing low-friction rotation while transmitting load.

    Cage (Retainer)
    A component that spaces and guides the rolling elements, maintaining even load distribution and preventing contact between balls.

    Seals or Shields
    Elements that retain grease and exclude contamination. Seals increase friction slightly but are critical for durability in MTB environments.

    2.3 Bearing Clearances and Preload

    Internal Clearance
    The designed free movement between the rolling elements and raceways before preload is applied. Clearance can be radial, axial, or both.

    Radial Clearance
    The total permissible movement of the inner race relative to the outer race in the radial direction prior to preload.

    Axial Clearance
    The permissible axial movement between bearing races prior to preload.

    2.4 Parts Within a Mountain Bike Front Hub

    Hub Shell
    The outer structural component of the hub that supports the spoke flanges and houses the bearing outer races.

    Hub Axle
    The structural member that passes through the hub bearings and interfaces with the fork or end caps. Depending on design, the axle may also define bearing spacing and preload.

    End Caps
    Components that interface between the axle and the fork dropouts. End caps may be nested or flush and may or may not participate in setting bearing preload.

    Internal Spacer
    A fixed-length tube or sleeve located between bearing inner races that defines axial spacing and preload in spacer-based hub designs.

    Preload Collar / Adjuster
    An externally accessible threaded component used to set bearing preload in adjustable hub systems.

    Fork Clamping Interface
    The surfaces where the hub assembly contacts the fork dropouts.

    Fork Thru Axle
    Structural member of the fork chassis that passes through the hub and applies axial clamping pressure.

     

    Diagram of a hub assembly with labeled components.

    Section 3: Design Considerations

    3.1 Load Paths

    The load path within a bicycle front hub describes how forces generated at the tire–ground contact patch and at the fork interface are transmitted through the hub’s structural components. During riding, the hub is subjected to a combination of radial loads from rider weight, impacts, and cornering forces, as well as axial loads introduced by axle clamping, braking reactions, and bearing preload mechanisms. Radial loads are primarily carried from the hub shell into the bearing outer races and through the rolling elements to the axle, while axial loads may be carried either through the bearings, through internal spacers or axle shoulders, or directly through the axle itself, depending on the hub architecture. The manner in which these loads are routed and whether axial clamping forces are coupled to or isolated from the bearings has a significant effect on bearing life, friction, stiffness, and long-term durability. Understanding the hub load path is therefore essential to evaluating both performance and reliability of different front hub designs.

     

    3.2 Bearing Preload in a Bicycle Hub

    Bearing preload in a bicycle hub is the intentional application of a small axial force to the bearing inner races to eliminate internal clearance and ensure consistent contact between the rolling elements and raceways. Proper preload removes axial play, maintains precise bearing alignment, and allows loads to be distributed evenly across the bearing elements during operation. When correctly set, preload improves perceived hub stiffness and steering precision without significantly increasing friction. However, excessive preload increases rolling resistance, accelerates bearing wear, and can lead to premature failure, while insufficient preload allows micro-movement that reduces precision and can cause fretting or impact damage. The method by which preload is established—whether through spacers, axle geometry, or external adjustment can directly affect how sensitive the hub is to manufacturing tolerances and fork clamping forces.

    3.3 Effect of Tolerances on Hub Precision and Stiffness

    Manufacturing tolerances play a critical role in determining both the precision and effective stiffness of a bicycle hub. Variations in bearing bore diameter, bearing seat alignment, spacer length, axle shoulder location, and end cap geometry all contribute to tolerance stack-up that can alter bearing preload and component alignment. Excessive clearance can result in axial or radial play, reducing perceived stiffness and leading to imprecise steering and accelerated wear. Conversely, overly tight tolerances or unfavorable stack-up can introduce unintended bearing preload, increasing friction and reducing bearing life while giving a false impression of stiffness. In addition, misalignment between bearing seats can force bearings to operate under bending loads rather than pure radial loads, degrading efficiency and durability. Well-controlled tolerances allow the hub to achieve high structural stiffness through proper load sharing and alignment, rather than through excessive preload or clamping force, resulting in more consistent performance and predictable long-term behavior. 

     

    Section 4: Detailed Hub Architecture Review

    4.1 Internal Spacer with Internal Nested End Caps

    Design Description

    In this architecture, bearing preload is defined by an internal spacer located between the two hub bearings. The spacer length is fixed and sets the axial distance between the inner bearing races. End caps nest inside the spacer and/or bearing bores and press against the inner races when the hub is installed in the fork.

        Key Characteristics

    • Preload is determined entirely by spacer length and bearing tolerances
    • End caps are internally captured and often lightly retained by O-rings
    • Fork clamping force transfers through the end caps, spacer, and inner races

            Advantages

    • Simple external appearance with minimal exposed interfaces
    • No user adjustment required; preload is automatically set during assembly
    • Good protection of interfaces from contamination

            Limitations

    • Highly sensitive to tolerance stack-up (bearing width, spacer length, shell machining)
    • Limited ability to compensate for wear or bearing replacement variation
    • Incorrect spacer length can lead to either bearing drag or axial play

          Typical Applications

    • Mid- to high-end hubs balancing compatibility, and simplicity
    • Weight-sensitive designs where external adjustment hardware is undesirable

    Load Path Description

    In this design, axial load from the fork enters through the internally nested end caps and is transferred directly into the bearing inner races. The inner races react against the internal preload spacer, placing the spacer and both inner races in compression. The tolerance between the spacer width and the hub shell bearing seat width determines the axial preload on the bearings.

    Radial loads from riding inputs travel from the hub shell into the bearing outer races, through the rolling elements, into the inner races, and finally through the end caps and into the fork axle interface. Because axial and radial loads are coupled at the bearings, tolerance stack-up or excessive clamping force can significantly increase bearing stress.

     

    Internal Spacer with Nested End Caps

    Manufacturer Model Axle x Spacing Bearing

    I9

    1/1

    15x110

    18x30x7

    Hadley

    Boost

    15x110, 20x110

    25x37x7

     

    Hope

    Pro 4&5

    15x110

    20x32x7

    Onyx

    Vesper

    15x110

    20x32x7

    4.2 Internal Spacer with Flush End Caps

    Design Description

    This design also relies on a fixed internal preload spacer setting the distance between bearings but the spacer, bearing and end cap all have the same inter diameter. The end caps still press against the inner races when the hub is installed in the fork, however the end caps are not radially controlled by the bearings. Instead, the end caps are mostly radially controlled by the fork thru axle.

        Key Characteristics

    • Fixed preload set by internal spacer
    • Spacer, Bearings, End Cap all have the same ID
    • Load paths are interrupted compared to nested end cap designs

        Advantages

    • Simple construction and low weight
    • In some cases the 15mm axle nested designs can be converted to a 20mm flush design

        Limitations

    • Still sensitive to manufacturing tolerances for preload
    • No external adjustment for preload
    • Interruptions in load paths can reduce robustness

        Typical Applications

    • OEM-focused hubs prioritizing simplicity and low part count
    • Weight-sensitive designs for 15mm axles where smaller bearing can reduce weight

    Load Path Description

    Similar to the previous design, axial load enters through the end caps, passes into the bearing inner races, and is reacted by the internal spacer. As before, the tolerance between the spacer width and the hub shell bearing seat width determines the axial preload on the bearings.

    Radial loads follow a similar path; however, the load path from the bearing inner races into the end caps is not as well defined as in the nested design. Because the end caps are not radially captured by the bearings, this architecture relies solely on axial face pressure at the bearing inner races to transfer load to the fork axle interface. As a result, tolerances in the bearings, end caps, and even the fork axle can allow axial float within the system and significantly reduce perceived hub stiffness.

    Internal Spacer with Flush End Caps

    Manufacturer Model Axle x Spacing Bearing

    DT Swiss

    240 & 350, 15mm

    15x110

    15x28x7

    DT Swiss

    240 & 350, 20mm

    20x110

    20x37x9

     

    Hunt

    Front Hub

    15x110

    15x28x7

    Onyx

    Vesper

    20x110

    20x32x7

    4.3 Internal Hub Axle with External Nested End Caps

    Design Description

    In this architecture, preload is set by a stepped or shouldered axle that defines bearing spacing internally. End caps nest externally over the axle ends and are typically non-load-setting components. The axle itself controls inner race spacing and preload.

        Key Characteristics

    • Axle length and shoulder geometry define preload
    • End caps are primarily adapters for axle standards and fork interfaces
    • Reduced reliance on separate internal spacers

        Advantages

    • Improved control of preload through a single machined component
    • More robust against tolerance variation during bearing replacement
    • End caps can be easily swapped for different axle standards

        Limitations

    • Axle is a more complex and higher-cost component
    • Damage to axle shoulders can directly affect preload
    • Slightly higher weight compared to spacer-based systems

        Typical Applications

    • Performance-oriented hubs prioritizing consistency and durability
    • Designs where modular end cap systems are required

    Load Path Description

    In this architecture, the load path differs from the previous two examples, while the bearing preload concept remains similar with one key distinction. The hub axle incorporates integral shoulders that define bearing spacing and preload, as well as extensions that capture the bearing inner races to maintain axial alignment.

    Axial load from fork axle clamping enters through the external nested end caps and is transferred into the bearing inner races. The inner races react against the shoulders of the hub axle, placing the axle and both inner races in compression. The tolerance between the axle shoulder spacing and the hub shell bearing seat spacing determines the resulting axial preload on the bearings.

    Radial loads pass from the hub shell into the bearing outer races, through the rolling elements and inner races, into the hub axle, through the nested end caps, and finally into the fork interface. Because axial and radial loads are coupled at the bearings, tolerance stack-up or excessive clamping force can significantly increase bearing stress. However, because the hub axle promotes axial alignment within the system, bearing loads are generally more stable and less sensitive to external fork flex.

     

    Internal Hub Axle with External Nested End Caps

    Manufacturer Model Axle x Spacing Bearing

    I9

    Hydra 1, 6 Bolt

    15x110

    20x32x7

    I9

    Hydra 1 CL

    15x110

    1830x7

    I9

    Hydra 2, 15mm

    15x110

    17x30x7

    I9

    Htydra 2, 20mm

    20x110

    25x37x7

    One Up

    Front Hub

    15x110

    18x28x7

    Race Face

    Vault 414

    20x110

    25x37x7

    4.4 Solid Axle with External Preload Collar

    Design Description

    This design uses a solid, continuous axle that passes through both bearings, with preload set by an externally accessible threaded collar or adjuster. The collar applies axial load directly to one bearing’s inner race, allowing fine preload adjustment. A variation to this design would include externally nested end caps on the solid axle for the sole purpose of adapting to different fork interface standards. However, the additional end caps would not change the overall function of this design.

            Key Characteristics

    • User- or technician-adjustable preload
    • Preload set independently of fork clamping force
    • Often includes a locking mechanism to maintain adjustment

            Advantages

    • Highly tolerant of bearing wear and manufacturing variation
    • Enables precise preload tuning for minimal friction and maximum bearing life
    • Well-suited for serviceability and long-term use

        Limitations

    • Increased part count and assembly complexity
    • Potential for misadjustment if not properly secured
    • Slight weight and cost penalty

        Typical Applications

    • High-end or service-focused hubs
    • Designs emphasizing longevity, tunability, and consistent performance

    Load Path Description

    This design most clearly separates fork clamping loads from bearing preload. An external adjustment collar on the solid axle allows bearing preload to be set independently of the axial clamping load applied by the fork.

    Axial load from the fork is reacted almost entirely by the solid axle spanning between the dropouts. Once preload is set via the external collar, additional clamping force do not impact the bearing load. The preload collar applies a controlled axial force to one bearing inner race, while the opposite bearing reacts against a fixed shoulder.

    Radial loads are transferred conventionally from the hub shell to the bearing outer races and into the axle. Because axial and radial loads are largely decoupled, the bearings operate in a more predictable and lower-stress environment.

     

    Solid Axle with External Preload Collar

    Manufacturer Model Axle x Spacing Bearing

    Chris King

     

    MTN Boost

    15x110, 20 x 110

    30

    Onyx

    Classic 15mm

    15x110

    25x37x7

    Onyx

    Classic 20

    20x11025x37x7

    17x30x7

    Section 4.5 Hub Architecture Summary

    The four hub architectures discussed each offer distinct advantages depending on design priorities and application. For inverted fork systems in particular, the solid axle with external preload collar provides the most consistent and mechanically decoupled solution, as it isolates fork clamping loads from bearing preload and delivers predictable stiffness and durability. At the opposite end of the spectrum, the internal spacer design with flush end caps is the least desirable for inverted applications due to its reliance on axial face contact and higher sensitivity to tolerance stack-up; however, it remains acceptable for many riders depending on riding style and performance expectations. The internal nested end cap design and the axle-defined preload architecture both represent high-quality solutions and are generally recommended, offering strong alignment control and consistent load transfer when properly manufactured.

    Beyond architectural type, overall hub quality, including machining precision, bearing selection, material strength, and tolerance control is a critical factor in performance and longevity. There are many high-quality hubs available on the market today. In general, PUSH recommends brands such as Hope, Industry Nine, and Chris King, which have demonstrated strong design execution, manufacturing consistency, and durability in demanding applications.

     

          Section 5: Hub–Fork Interface Effects on Inverted Fork Flex

    The interface between the front hub and the fork plays a more influential role in the flex behavior of an inverted fork than in a conventional fork design. Conventional forks use both a structural arch and an axle to tie the lower legs together, creating a rigid, box-like structure that limits the influence of the hub–dropout interface on overall system flex. In contrast, inverted forks do not use an arch, placing greater importance on the hub and axle interface as the primary structural connection between the fork legs.

     

    Figure 1: Standard End Cap Diameter 21mm

    As a result, changes in the engagement area of the hub end caps and the dropout interface directly affect how loads are shared between the hub, axle, and fork legs. Increasing the interface diameter, such as through the use of larger-diameter end caps or Torque Caps, alters the flex characteristics of the front-end system. These characteristics govern how the fork is perceived under riding inputs such as braking, jumping, and cornering. The ability to adjust this interface allows riders and designers to tune front-end feel, with larger interfaces generally improving steering precision and consistency by reducing unwanted flex, while smaller interfaces can introduce additional compliance and grip that may enhance comfort or rider confidence.

     

    Cross-section of a mechanical assembly with seals and a threaded component.

    Figure 2: Torque Cap End Cap Diameter 31mm

    A further consideration with Torque Caps is that not all designs are equivalent, and some are incompatible with floating-axle systems used on certain forks. Figure 3 illustrates two Torque Caps with the same nominal outer interface diameter of 31 mm; however, in the red cap, much of the interface surface has been removed, likely for weight savings. This reduced engagement area is not compatible with the floating axle used on most inverted forks and can compromise the axial tension required by the system. 

     

    Figure 3: Torque End Caps

    Section 6: Conclusion

    Front hub design plays a critical role in how loads are transmitted between the wheel and fork, directly influencing bearing life, structural stiffness, and front-end feel. Differences in preload architecture—whether defined by internal spacers, axle geometry, or external adjustment—fundamentally change the hub load path and the degree to which fork clamping forces are coupled to the bearings. Designs that rely on fixed spacers demand tight tolerance control to avoid unintended preload and variability, while axle-defined and externally adjustable systems offer greater consistency and robustness across manufacturing variation and service life.

    In addition to internal architecture, the hub–fork interface itself is a key contributor to overall system behavior, particularly in inverted fork designs where the absence of a structural arch places greater importance on the axle and end cap interface. Changes in interface diameter and engagement area can meaningfully alter flex characteristics, steering precision, and grip, enabling a level of front-end tuning not possible in more arch-dominated systems.

    Ultimately, no single hub architecture is universally optimal. Each represents a trade-off between simplicity, adjustability, tolerance sensitivity, and structural control. A clear understanding of load paths, bearing preload, and interface mechanics allows designers, manufacturers, and riders to make informed decisions that balance performance, durability, and ride feel for their intended application.

    Appendix A: Consumer Preference for Front MTB Hubs

    Vital 2025 Reader Survey

    WHEELSET: Will purchase within 12 months - 27.2%

    IF "YES," PRE-BUILT OR CUSTOM: Pre-built - 58.7% Custom - 41.3%

    IF “PRE-BUILT,” MATERIAL: Aluminum - 42.5% Carbon - 38.8% Undecided - 18.7%

    IF “PRE-BUILT,” WHAT BRAND:

    IF “CUSTOM,” WHAT BRAND OF HUBS:

    Undecided - 22.9%

    DT Swiss - 15.8%

    Race Face - 8.6%

    Reserve - 6.7%

    Industry Nine - 5.7%

    Hunt - 4.1%

    Roval - 3.2%

    We Are One - 3.2%

    Crankbrothers - 2.8%

    Hope - 2.8%

    e*thirteen - 2.5%

    Mavic - 2.1%

    Bontrager - 2.0%

    Shimano - 1.4%

    ENVE - 1.3%

    Zipp - 1.3%

    Newmen - 0.9%

    Halo - 0.9%

    Spank - 0.9%

    Stan's No Tubes - 0.9%

    NOBL - 0.7%

    Ibis - 0.5%

    Reynolds - 0.5%

    WTB - 0.4%

    Tenet - 0.3%

    American Classic - 0.3%

    FSA - 0.3%

    Misc. / Fill-In - 2.0%

    DT Swiss - 18.9%

    Industry Nine - 17.4%

    Hope - 14.8%

    Undecided - 12.4%

    Onyx - 7.2%

    Chris King - 6.3%

    One Up -2.6%

    Race Face1.9%

    Project 3211.5%

    e*thirteen1.4%

    Profile1.3%

    Spank1.3%

    Shimano1.2%

    Hadley0.8%

    Crankbrothers0.7%

    Misc. / Fill-In 3.3%