ASME B17.1-1967 (2017) scan — Keys and Keyseats

🔑 ASME B17.1-1967 is a foundational document in the ASME Mechanical Power Transmission (B17) family. Its title, Keys and Keyseats, understates its importance: it answers a set of small but decisive questions that govern the reliability of an entire drive train. Given a shaft, how large should the key be? What tolerances apply to the key’s width, height, and length? How deep must the keyseat be cut in the shaft, and how deep in the hub? Which fit class applies to the assembly? And how can a machinist verify keyseat depth and side-face parallelism on the shop floor with the simplest possible instruments? The standard was approved as an American National Standard on 15 September 1967 by the USA Standards Institute (predecessor of ANSI) and designated B17.1-1967; after successive reaffirmations, the 2017 confirmation is the currently applicable text. The source material for this article is a scanned copy, so the discussion follows the standard’s table of contents, chapter structure, and established mechanical power transmission practice.

1. Scope and General Framework 📌

The standard specifies the size, type, and tolerances of parallel and taper keys and their keyseats, and it supplies the supporting rules for key-to-keyseat fits, depth control of keyseats, chamfers and fillets on keys and keyseats, set screws used over keys, and alignment tolerances of keyseats relative to the shaft axis. Its subject is the keyed joint between a shaft and a hub — the means by which torque is transmitted from a shaft to a gear, sheave, sprocket, or coupling, or in the reverse direction.

The standard has a clear lineage. Individual USA Standards on keys were first issued in the late 1920s and were consolidated with the shafting standards in the 1934 publication Standard Shafting and Stock Keys. A revised issue appeared in 1943 and was withdrawn in 1955 because it was not being sufficiently supported. A separate standard covering Woodruff keys, keyslots, and cutters (B17f-1930) was reaffirmed in 1955. In 1962 the USA Standards Committee B17 was reactivated, and a subcommittee developed this standard based on current industry practice; it was approved on 15 September 1967 and designated B17.1-1967.

Equally important is what the standard does not cover: involute and straight-sided splines (governed by the B92 series), pin joints and interference fits, fatigue assessment of keyed shafts, and material selection for high-temperature, corrosive, or very high-speed duty. Strength verification must be performed separately, while the dimensions and tolerances are taken from this standard.

2. Key Types and Definitions 🔧

The definitions clause of the standard separates terms that are very often confused, and everything that follows depends on getting them right:

• Key: a machine element placed between shaft and hub to transmit torque; it normally carries no axial load.
• Keyseat: the slot cut in the shaft.
• Keyway: the slot cut in the hub bore. Together they form the key–keyseat assembly.
• Parallel key: a straight key of square or rectangular cross section whose side faces are the working faces; torque is carried by side bearing and shear, and a clearance is left between the top of the key and the bottom of the hub keyway.
• Taper key: a key with roughly a 1:100 taper on its top face; when driven in, it wedges tightly and adds friction to the side bearing action. It may be plain or gib-headed, the gib head easing removal.
• Woodruff key: a half-moon key seated in a semicircular seat in the shaft, able to self-align with the hub keyway; common on small-diameter and tapered shaft ends.

Other variants encountered in practice include the Pratt & Whitney key (semicircular ends, tolerates some axial adjustment) and the feather key (fixed to the shaft while the hub slides along it). All of them inherit the width–height–depth relationships established by this standard.

3. Key Size versus Shaft Diameter (Table 1) 📏

Table 1 of the standard establishes the “key size versus shaft diameter” relationship and is by far its most frequently used content. The principle is straightforward: key width increases in steps as shaft diameter increases, and within a given diameter band either a square key (width equal to height) or a slightly shallower rectangular key may be used, giving designers flexibility when hub wall thickness is limited or when the depth of the keyseat has become the fatigue-critical feature of the shaft. The table below lists the customary inch bands together with approximate metric equivalents.

Metric drawings, which dominate in many regions, use the same stepping pattern as ISO 773 and DIN 6885-1, and can be cross-referenced to the inch system of B17.1. The following table gives the customary metric correspondence so that key and keyseat depths can be taken quickly for a metric shaft diameter.

4. Tolerances and Fit Classes for Keys and Keyseats ⚙️

The standard treats the three key dimensions very differently, and that asymmetry is the key to understanding the whole fit system:

• Key width b: the only fitting dimension, toleranced most tightly; all clearance or interference between key and keyseat is carried here.
• Key height h: a non-fitting dimension controlling only top clearance. A clearance must remain between the top of a parallel key and the bottom of the hub keyway (typically about 0.005–0.015 in, or 0.13–0.38 mm), ensuring load travels through the side faces rather than being carried on the top face.
• Key length L: a free dimension, taken as roughly 0.8–0.9 of hub length, with allowance at each end for chamfer or radius.

For parallel keys the standard recognizes two assembly fit classes. The Class 1 fit is looser, with a definite clearance between key width and keyseat width; it eases hand assembly, tolerates slight misalignment, and suits general drives, frequently dismantled parts, and axially sliding hubs. The Class 2 fit is closer, with minimal clearance or a transition condition; it is used for reversing duty, shock loads, precision drives, and cases where fretting must be suppressed. Taper keys have their own wedging provisions: after driving, the top face should seat against the bottom of the hub keyway.

5. Keyseat Depth Control and Shop Measurement 📐

Depth control is the most practically valuable part of the standard. Two symbols describe depth: T for the keyseat depth in the shaft and S for the keyseat depth in the hub. Together with key height h they must satisfy S + T = h + top clearance. For square keys, T = S = h/2 is customary; for rectangular keys the split follows Table 3, generally with T slightly larger and S slightly smaller so that adequate bearing height remains on the shaft side.

Because the shaft is cylindrical, measuring from the bottom of the slot to the shaft axis is both awkward and error-prone. The standard therefore introduces the concept of a depth-control dimension: the distance from the bottom of the keyseat to the opposite outside surface of the shaft. For the shaft the depth-control dimension is d − T; for the hub it is d + S. A micrometer or height gauge then suffices, and no tool setting on a curved surface is required.

Three inspection methods are common. First, a vernier or micrometer depth gauge reads T and S directly. Second, a cylindrical gauge pin sized to the key width b is laid in the seat and the distance from the top of the pin to the opposite side of the shaft is measured and converted; this method also reflects the quality of the bottom radius. Third, in batch production, a dedicated keyseat caliper or limit gauge gives a go/no-go verdict. The pin method best reflects the assembled condition and is preferred for taper keys and tapered shafts.

6. Machining, Fillets, and Alignment Tolerances 🛠️

Keyseat quality directly affects shaft fatigue life. The standard addresses keyseat edge chamfers, key end radii, and the parallelism of keyseat side faces to the shaft axis (Tables 7 and 6):

• Chamfers and fillets: keyseat edges are deburred, and key ends are chamfered or radiused, typically with a radius of about 0.005–0.015 in, to reduce stress concentration and to ease assembly.
• Parallelism: the keyseat side faces should be parallel to the axis, normally held within about 0.02 mm per 25 mm of length (0.001 in/in). Electric motor and generator shaft extensions have their own more explicit requirements in Table 6, and motors should be accepted against that table.
• Symmetry: the keyseat center plane must be aligned with the datum of the shaft; offset forces the key to carry a bending moment and is a frequent cause of early shear failure.
• Set screws (Table 8): where a key must be retained in its seat, or a feather key must be limited axially, a set screw over the key may be used; the standard gives recommended screw sizes by key width.

7. Engineering Practice and Strength Checks 📊

Dimensional conformance is only the starting point; a strength check is still required. Keyed joints are normally verified against two failure modes, shear and bearing (crushing):

F = 2M / d   →   τ = F / (b · L)   →   σ = F / (t · L)

where M is transmitted torque, d shaft diameter, F the tangential force on the key, b key width, L effective key length, and t the effective bearing height (h/2 for a square key). Worked example: shaft d = 50 mm, torque M = 500 N·m, key 14 × 9 × 60 in 45 steel. Then F = 2 × 500 / 0.05 = 20 kN; shear area A = 14 × 60 = 840 mm², so τ ≈ 23.8 MPa; bearing area A_b = 4.5 × 60 = 270 mm², so σ ≈ 74 MPa. With allowable bearing stress of roughly 100–150 MPa and allowable shear of roughly 60–90 MPa for a 45-steel key under static load, the key is adequate; for shock duty, reduce the allowables by a factor of about 0.7 and recalculate.

• 🔍 Selection workflow: read Table 1 for key width and height from shaft diameter → set key length from hub length → choose the fit class for the duty → take T and S from Table 3 and dimension the depth control → check shear and bearing → specify chamfers, fillets, and parallelism.
• 🔧 Material and heat treatment: cold-drawn carbon steel (AISI 1020/1045) is typical and is normally not heat treated. The key should be slightly softer than shaft and hub so that failure occurs in the inexpensive, replaceable part rather than in the shaft.
• 📋 Drawing callouts: state keyseat width and tolerance, the depth-control dimension (d − T or d + S), symmetry and parallelism, and the fit class. A note reading “keyseat per B17.1” alone does not give a machinist enough information.
• 🧰 Assembly: deburr the seat, apply a thin film of grease against corrosion, and drive taper keys only to the specified amount; over-driving eccentrically displaces the hub and adds load to the bearings.

8. Frequently Asked Questions ❓