2.6 Angular Surfaces, Tapers, Radii, and Special Tolerance Cases

Angular Surfaces, Tapers, Radii, and Special Tolerance Cases

Quick Answer: Plus-minus angles create wedge-shaped zones that grow with distance; basic angles plus profile or orientation create uniform-width zones. R allows flats and reversals inside the crescent zone, CR does not. MIN/MAX single limits leave the other end unbounded by the drawing. Tabulated tolerances are a notation convenience, not a change in meaning.

The 10% Scope, General Dimensioning, and Symbology category is the one most candidates under-study, because it looks elementary next to datum referencing. It is not. The body of knowledge explicitly lists angular surfaces, conical tapers, flat tapers, radius, single limits, tolerance accumulation, units of measure, and tabulated tolerances — and with roughly fifteen questions riding on this category against a 50% floor, missing eight of them ends the exam regardless of your datum score.

Angular surfaces: the wedge problem

Dimension a 30-degree surface as 30° ± 1° and you have specified a wedge-shaped tolerance zone. The two limiting planes diverge from the vertex, so the permissible surface deviation is small near the vertex and large far from it. Two consequences follow:

1. The tolerance you actually get depends on where the surface lies relative to the vertex, which is rarely what the designer intended.
2. The zone is not uniform, so it cannot be verified with a simple indicator sweep against a fixed reference.

The Senior-correct alternative is a basic angle (boxed) combined with a geometric control:

• Angularity to a datum reference frame when you want a uniform-width zone that also constrains orientation to the datums.
• Profile of a surface when you want to control form, orientation, and location of the surface in one requirement.

Both produce two parallel planes separated by the tolerance value — a uniform zone at every point on the surface. When an exam question shows a plus-minus angle and asks what is wrong, "the tolerance zone is wedge-shaped and varies with distance from the vertex" is the answer.

Conical and flat tapers

A conical taper may be defined several ways:

The taper ratio is defined as the change in diameter divided by the length over which that change occurs. A flat taper uses the same logic on a flat surface, with the ratio expressed as the change in height over length.

Functional tapers — a Morse taper socket, a valve seat, a self-locking cone — should be defined with a basic taper plus profile. The plus-minus method allows the diameter error and the length error to accumulate into an angle error nobody specified.

Radius, controlled radius, and spherical radius

Three symbols, three different requirements:

This distinction has real consequences. A stamped fillet that dips and rises inside the crescent satisfies R and fails CR. Specify CR where a smooth transition carries the function — fatigue-critical fillets, sealing radii, surfaces a mating part must slide across. Specify R where only the envelope matters, because CR costs more to make and considerably more to inspect.

Single limits

A single limit dimension states only one end of the range using MIN or MAX:

• 5 MIN — the feature must be at least 5; the drawing places no upper limit.
• 0.8 MAX — the feature must be no more than 0.8; the drawing places no lower limit.

The unspecified end is governed by the design intent, the manufacturing process, and any other requirements the feature carries — not by "anything goes." Single limits are appropriate for depths of blind holes, thread runouts, chamfer breaks, and corner radii where only one direction can cause a failure. They are a poor choice on any feature that mates, because the unbounded end is exactly where an assembly problem hides.

Tolerance accumulation

Three dimensioning schemes produce three different accumulations between two features:

The Senior application is that you should dimension the relationship the function cares about directly, rather than letting it fall out of a chain. Geometric tolerancing with basic dimensions from a datum reference frame effectively behaves like baseline dimensioning, which is one reason it eliminates the classic stack-up surprises.

Units of measure and tabulated tolerances

Units of measure: a drawing is prepared in either millimetres or inches, stated once in the title block, and individual dimensions do not repeat the unit symbol. Millimetre dimensions omit a trailing zero on the tolerance and drop the leading zero on values less than one; inch dimensions historically retain the trailing zeros so that the tolerance's decimal places match the dimension's. Mixed-unit drawings must identify the exception explicitly.

Tabulated tolerances: when many features share a pattern of dimensions, the drawing may replace repeated values with letter designations — A, B, C — resolved in a table elsewhere on the drawing. This is purely a notation convenience. The tabulated value means exactly what a written value would mean, carries the same tolerance, and can be basic if boxed. A tabulated dimension does not become "reference" or "advisory" because it lives in a table.

Statistical tolerancing

Statistical tolerancing assigns a tolerance on the assumption that the feature will be produced under statistical process control, so that the assembly is evaluated against the statistical combination of the contributors rather than against their arithmetic worst case. The ST symbol placed next to the tolerance signals this.

Two forms appear on drawings:

• Statistical tolerance only. The stated tolerance is valid only when the process is under statistical control. The drawing carries a note to that effect, and a supplier without SPC may not use the tolerance at all.
• Statistical tolerance with an arithmetic alternative. Two tolerance values are shown — the larger one qualified with ST for a supplier running SPC, and a smaller arithmetic value for a supplier who is not. Both are legitimate; the supplier chooses which regime to work under and must satisfy the corresponding value.

The Senior application is a design-authority judgment: statistical tolerancing buys real tolerance back on a stack with many contributors, but it transfers a process-control obligation to manufacturing and is worthless if that obligation is not enforced. Never apply ST to a feature whose failure mode is a single worst-case interference, because the statistical argument does not protect against the one part that lands at the limit.

Scope and normative references

Finally, know what the standard claims for itself. Y14.5 establishes uniform practices for stating and interpreting dimensioning, tolerancing, and related requirements on engineering drawings and in related documentation. It is invoked by reference on the drawing, and it in turn invokes a family of companion standards that a Senior candidate should be able to name: Y14.5.1 for the mathematical definitions behind the tolerance zones, Y14.5.2 for the GDTP certification program itself, Y14.41 for digital product definition data practices, Y14.36 for surface texture symbols, and Y14.38 for abbreviations and acronyms. When a drawing and the standard disagree, the drawing governs — Y14.5 supplies the default rules that apply where the drawing is silent.