6.1 Flatness and Straightness
Key Takeaways
- Flatness is the only geometric tolerance in ASME Y14.5-2009 that by definition never references a datum; its zone is two parallel planes free to orient to the surface.
- Straightness has two modes: line-element straightness (two parallel lines, applied to a surface element) and axis straightness (a cylindrical zone applied to the derived median line of a feature of size, with a diameter symbol).
- Only straightness applied to a feature of size overrides Rule #1 for that feature; flatness and line-element straightness do not.
- Flatness controls shape only; it cannot control orientation (parallelism, perpendicularity) because it has no datum to relate to.
- Flatness deviation is the minimum-zone peak-to-valley separation, not a plus-or-minus swing from a single zero.
6.1 Flatness and Straightness
Quick Answer: Flatness and straightness are form controls in ASME Y14.5-2009 that govern shape without reference to any datum. Flatness confines a surface between two parallel planes; straightness confines a line element between two parallel lines, or—when applied to a derived median line/axis of a feature of size—within a cylinder. Flatness is unique among geometric tolerances: it never references a datum. Both controls appear small on the Senior exam (Form is roughly 5% of the 150 questions) but are application-level, so the traps are about selection and Rule #1 interaction, not definitions.
Flatness — the only datum-free geometric tolerance
Flatness is specified by the geometric characteristic symbol (a parallelogram) followed by a tolerance value. The tolerance zone is two parallel planes separated by the tolerance value t, within which the entire considered surface must lie. The zone may be oriented anywhere in space to fit the actual surface—there is no datum to constrain its orientation, which is why flatness cannot also control orientation (parallelism, perpendicularity, angularity) or location.
Key properties:
- Flatness applies to a real (planar) surface, not to a feature of size and not to a derived median plane.
- The tolerance zone is free to rotate and translate to minimize departure; a coordinate measuring machine (CMM) or surface plate check uses a best-fit (minimum-zone) plane.
- It is the only geometric tolerance that by definition never references a datum. Even profile and runout, which can be used without datums, are not strictly barred; flatness is structurally datum-free because a plane has no intrinsic reference.
- Flatness does not control size, does not establish a datum, and does not constrain orientation. A perfectly flat part can still be tilted relative to another surface.
When flatness matters at the Senior level
Senior GDTP questions rarely ask "what is flatness." They ask when to apply it and what it does not do:
- Sealing surfaces on flanges, valve faces, and pump housings need local flatness so a gasket compresses uniformly.
- Datum establishment: a planar primary datum feature is typically controlled with flatness so the datum simulator (surface plate) contacts the high points reproducibly.
- Mating lands: a flatness callout finer than the general form control ensures two parts seat without rocking.
Worked example — flange face flatness
A pump-cover flange is shown with a flatness tolerance of 0.05 mm on the planar face that mates to the pump body. The face is 120 mm across. The inspector sweeps the surface with a dial indicator on a surface plate and records:
- Peak reading: +0.03 mm
- Valley reading: −0.06 mm
The flatness deviation is the difference between the highest peak and lowest valley after the best-fit plane is established: 0.03 − (−0.06) = 0.09 mm, which exceeds the 0.05 mm zone. The part is nonconforming. The indicator readings are relative to the plate; flatness is computed as a minimum-zone envelope, not as a plus-or-minus deviation from a single zero.
Straightness — two distinct applications
Straightness has two application modes, and the Senior exam distinguishes them sharply:
-
Straightness of a surface element (line element). Applied to a planar or cylindrical surface element. Tolerance zone = two parallel lines separated by t, within which each longitudinal line element must lie. The zone floats independently for each line element; the tolerance does not constrain the feature as a whole. Because it controls surface elements, it is treated as a form control.
-
Straightness of a derived median line (axis) of a feature of size. When straightness is applied to a feature of size (a shaft or a width) with a diameter modifier, the tolerance zone is a cylinder of diameter t within which the derived median line must lie. This is one of the few form controls that can be applied to a feature of size and that overrides ASME Rule #1 for that feature (see §6.3). The material condition modifier (at MMC, at LMC, or RFS by default) is meaningful here.
Selection table
| Control target | Symbol form | Zone shape | Datum allowed? | Overrides Rule #1? |
|---|---|---|---|---|
| Flatness of a surface | t (parallelogram) | Two parallel planes | No (never) | No |
| Straightness of a line element | t (no diameter) | Two parallel lines | No | No |
| Straightness of an axis (FOS) | diameter t | Cylinder | No | Yes |
| Straightness of a center plane (FOS) | t (width) | Two parallel planes | No | Yes |
Trap: "flatness controls parallelism"
A common Senior distractor claims a flatness tolerance on the top face of a block keeps it parallel to the bottom face. It does not. Two faces can both be perfectly flat and not parallel. Parallelism is an orientation control and requires a datum; flatness has no datum, so it cannot simultaneously enforce orientation.
Trap: line-element straightness is not axis straightness
On a shaft, a straightness callout applied to the surface (leader to the cylindrical surface, no diameter symbol) controls line elements only. The shaft can still banana-bow along its axis because each longitudinal generator is straight even though the axis curves. To control axis bow you must apply straightness to the feature of size with a cylindrical zone. Mixing these two is a classic exam error.
Worked example — shaft straightness override of Rule #1
A shaft is dimensioned Ø20 ±0.1 mm with straightness applied to the feature of size: diameter 0.2 at MMC. MMC = 20.1, LMC = 19.9. At MMC the axis must lie in a diameter 0.2 cylinder; as the shaft departs toward LMC, the straightness tolerance grows by the amount of departure, reaching diameter 0.4 at LMC. Because straightness is applied to a feature of size, Rule #1 (perfect form at MMC) is overridden for this feature, and the 0.2/0.4 boundary replaces the perfect-form boundary.
Senior application takeaways
- Choose flatness when you need a planar datum to be reproducible or a face to seal.
- Choose line-element straightness when only one direction of a surface matters (e.g., a rail, a way).
- Choose axis straightness with a cylindrical zone when a bent shaft must still assemble, and accept the Rule #1 override that follows.
- Never specify a datum reference frame on a flatness callout—the drawing is invalid if you do.
Under ASME Y14.5-2009, which statement about the four form tolerances (flatness, straightness, circularity, cylindricity) is correct?
A shaft is dimensioned Ø20 ±0.1 and controlled with straightness of its derived median line, diameter 0.2 at MMC. What straightness tolerance applies at LMC (Ø19.9)?
An inspector measures a flange face with a sweep indicator and records a peak of +0.03 mm and a valley of −0.06 mm after best-fit. The flatness callout is 0.05. The part is:
Which callout would control bow (axis curvature) of a cylindrical shaft AND override Rule #1 for that feature?