Dimensioning, Fits and Geometric Tolerancing

Introduction

An engineering drawing is a technical contract. It must communicate the geometry, allowable variation, surface condition and assembly relationships required for a component to function. A drawing that defines only nominal dimensions is incomplete because no manufacturing process produces exact geometry.

Dimensioning and tolerancing should balance function, manufacturability, inspection capability, interchangeability and cost. Excessively tight tolerances increase manufacturing and verification costs, whereas loose or ambiguous requirements may cause assembly failures, electrical misalignment, vibration, leakage or reduced service life.

Electrical engineers encounter these issues in motor shafts, bearing housings, connector panels, heat sinks, sensor mounts, switchgear enclosures, printed-circuit-board fixtures and electromechanical assemblies. The governing drawing standard must always be stated. This chapter principally uses concepts consistent with ASME Y14.5-2018 for geometric dimensioning and tolerancing (GD&T), while identifying relevant ISO practices where appropriate.

Learning Objectives

After completing this chapter, students should be able to:

• apply functional and unambiguous dimensioning;
• calculate limits, tolerance and allowance;
• distinguish strict clearance, strict interference, true transition and line-to-line boundary conditions;
• interpret hole-basis and shaft-basis fit systems;
• perform worst-case and statistical tolerance-stack calculations;
• interpret surface-texture symbols, parameters and lay;
• identify the 12 geometric characteristics retained in ASME Y14.5-2018;
• interpret feature-control frames and datum-reference frames;
• select appropriate material-condition modifiers; and
• avoid over-dimensioning and conflicting requirements.

1. Functional Dimensioning

1.1 Size, location and orientation

A dimension is a numerical value defining a physical requirement such as size, distance, angle or location. Good dimensioning begins with function rather than with the convenience of drawing construction.

Dimensions commonly control:

• size: diameter, width, thickness or length;
• location: position of a hole, slot or connector;
• orientation: parallelism, perpendicularity or angularity;
• form: straightness, flatness, circularity or cylindricity; and
• profile: the permitted variation of a line or surface.

A dimension should normally appear only once. Repeating the same requirement in different views can create a conflict after a design revision. Dimensions should also refer to visible features and stable functional references whenever possible.

1.2 Baseline, chain and coordinate dimensioning

In baseline dimensioning, related dimensions originate from a common reference. This limits accumulation of dimensional variation and is often preferred for hole patterns or connector locations.

In chain dimensioning, each feature is located from the preceding feature. Chain dimensioning may be useful when adjacent spacing is functionally important, but tolerances accumulate along the chain.

Coordinate dimensioning locates features by numerical coordinates from defined origins. Computer numerical control and coordinate-measuring systems use coordinates effectively, but the drawing must still communicate functional datum relationships.

Figure: Baseline dimensioning controls locations from one reference, whereas chain dimensioning accumulates variation.

1.3 Avoiding ambiguity and over-dimensioning

A drawing is over-dimensioned when it contains more controlling dimensions than are mathematically necessary. For example, if three collinear hole spacings are specified together with their total span, one dimension must be identified as reference information; otherwise manufacturing variation may make the requirements mutually incompatible.

Reference dimensions are commonly shown in parentheses. They provide information but do not establish an independent acceptance requirement. The governing standard determines exact presentation rules.

A robust drawing should:

1. dimension from functional interfaces;
2. avoid duplicate requirements;
3. distinguish basic, reference and toleranced dimensions;
4. use notes only when graphical specification is insufficient;
5. specify units and the governing standard; and
6. assign tolerances that can be manufactured and inspected.

2. Limits, Tolerances and Allowance

2.1 Limit dimensions

The limits of size are the largest and smallest permitted sizes. If a nominal size N has upper deviation e_u and lower deviation e_l, then:

where all quantities use the same unit. The size tolerance is:

A bilateral tolerance permits variation on both sides of nominal, such as 25.00 ± 0.05 mm. A unilateral tolerance permits variation in one direction, such as 25.00^+0.08_-0.00 mm. Unequal bilateral tolerances are also permitted.

Example: For 18.00^+0.03_-0.01 mm, the maximum size is 18.03 mm, the minimum is 17.99 mm and the total tolerance is 0.04 mm.

2.2 Hole–shaft relationships

Let:

• H_min, H_max = minimum and maximum hole sizes;
• S_min, S_max = minimum and maximum shaft sizes;
• C_min = minimum signed clearance; and
• C_max = maximum signed clearance.

Then:

Positive signed clearance means a gap; negative signed clearance means interference. Because H_min ≤ H_max and S_min ≤ S_max, it follows that C_min ≤ C_max.

The traditional allowance is the intentional difference between mating parts at maximum material condition. For the hole–shaft pair it is numerically H_min-S_max=C_min. It is positive for a strict clearance fit and negative for a strict interference fit.

2.3 Correct fit classification

Equality at zero must not automatically be called a transition fit. A true transition fit requires both positive and negative signed clearances to be possible.

Worked Example 1: Classifying a fit

A hole has limits 20.010–20.030 mm and a shaft has limits 19.990–20.015 mm.

Since -0.005<0<0.040, both interference and clearance are possible. This is a true transition fit.

If the shaft maximum had instead been 20.010 mm, then C_min=0, with positive maximum clearance. That would be a clearance-to-line boundary condition, not a true transition fit.

3. Standard Fit Systems

3.1 Hole-basis and shaft-basis systems

In a hole-basis system, the basic hole is held at a standard tolerance position and shaft deviations are selected to obtain the required fit. In the ISO system, an H hole has zero lower deviation. Hole-basis practice is common because standard drills, reamers and gauges make holes economical to standardize.

In a shaft-basis system, the shaft is standardized and hole deviations are varied. An ISO h shaft has zero upper deviation. This system can be advantageous when commercial shafting is used without further size adjustment.

An ISO fit designation such as 40 H7/g6 identifies:

• 40 mm basic size;
• H tolerance position for the hole;
• tolerance grade 7 for the hole;
• g tolerance position for the shaft; and
• tolerance grade 6 for the shaft.

The numerical limits must be obtained from the applicable ISO fit tables for the specified basic-size interval. Letter position and grade alone should not be converted using invented rules.

3.2 Selecting a fit

Fit selection depends on function rather than terminology alone.

Thermal effects must be considered when mating parts use different materials. For uniform free expansion:

where ΔL is length change, α is the coefficient of linear thermal expansion in K⁻¹, L₀ is initial length and ΔT is temperature change in K or °C difference. The formula assumes approximately constant α over the temperature interval and unconstrained expansion.

4. Tolerance Stack-Up

A tolerance stack is the cumulative variation of dimensions that determine a functional result such as a gap, connector position or insulation distance.

For a linear relationship:

where a_i are signed sensitivity coefficients, the nominal result is found by substituting nominal dimensions.

4.1 Worst-case analysis

If each input may independently reach either limit, the worst-case bilateral deviation bound is:

where w_i is the bilateral half-width or maximum absolute deviation of input X_i. This method guarantees assembly if every component remains within its specified limits, but it may be conservative.

4.2 Statistical root-sum-square analysis

For independent, centered variations expressed at a consistent statistical coverage level, an approximate combined half-width is:

RSS is not a guaranteed-limit method. It is inappropriate without evidence that processes are stable, approximately centered and sufficiently independent. Correlation requires covariance terms and can substantially change the result.

Worked Example 2: Connector clearance stack

A required clearance is:

Given L=100.00±0.20 mm, A=40.00±0.10 mm and B=25.00±0.08 mm, the nominal clearance is:

The worst-case half-width is:

Thus the guaranteed limit result is G=35.00±0.38 mm, or 34.62–35.38 mm.

Under justified independent, centered statistical assumptions:

The approximate RSS result is therefore 35.00±0.24 mm. It must not replace worst-case acceptance without process-capability evidence and an explicit risk decision.

5. Surface Texture

5.1 Functional significance

Surface texture affects friction, wear, sealing, electrical contact resistance, coating adhesion, fatigue strength and bearing performance. A very smooth surface is not automatically superior: lubricant retention, bonding or coating may require controlled texture.

Common profile parameters include:

• Ra: arithmetic mean absolute profile deviation;
• Rq: root-mean-square profile deviation;
• Rz: a profile-height parameter whose exact evaluation definition depends on the cited standard; and
• waviness: longer-wavelength variation than roughness.

Over evaluation length L, the idealized continuous form of Ra is:

Here z(x) is the filtered profile deviation from its mean line. Both z and R_a have length units, commonly μm. Real instruments use standardized filtering, sampling and evaluation procedures; the formula alone is not a complete inspection specification.

5.2 Symbol structure and value placement

ISO 1302 and ASME surface-texture practices use related but not completely interchangeable conventions. The drawing must cite its governing standard. The basic check-mark-like symbol establishes a surface-texture requirement. Adding a bar indicates that material removal is required; adding a circle indicates that material removal is not permitted.

Figure: Basic surface-texture symbol forms and an illustrative complete callout. Exact placement and parameter notation must follow the cited drawing standard.

In the complete example, GRIND states the process, Ra 0.8 μm states the roughness parameter and limit, λc 0.8 mm identifies a cutoff where required, ⊥ specifies lay, and a = 2 mm indicates machining allowance. Designers should not specify a manufacturing process unless the process itself is functionally necessary.

5.3 Direction of lay

Lay is the predominant direction of surface pattern. Common standardized concepts include:

The meaning is interpreted relative to the drawing view and the cited standard. Lay may influence sealing, sliding wear and electrical contact behavior.

6. Fundamentals of GD&T

GD&T is a symbolic language for controlling geometric variation according to function. It does not eliminate size dimensions; rather, it supplements size control with explicit requirements for form, orientation, location and runout.

6.1 The 12 geometric characteristics in ASME Y14.5-2018

ASME Y14.5-2018 contains 12 geometric characteristic symbols:

Concentricity and symmetry were removed from ASME Y14.5-2018. Legacy drawings may contain them, but new drawings governed by the 2018 edition should use functional alternatives such as position, runout or profile as appropriate.

6.2 Datums and datum reference frames

A datum feature is a physical feature identified on the part. A datum is the theoretically exact plane, axis or point established from that feature through simulated contact. Datum letters in a feature-control frame establish precedence.

A common planar 3-2-1 setup uses:

• a primary datum plane to constrain three degrees of freedom;
• a secondary datum plane to constrain two additional degrees; and
• a tertiary datum plane to constrain the remaining degree.

Actual contact and restraint depend on datum-feature geometry, material-boundary modifiers and the governing standard.

6.3 Feature-control frame example

Consider a rectangular mounting plate with a four-hole pattern. The mounting face is datum feature A, the left side is datum feature B, and the lower side is datum feature C. A position tolerance may be specified as:

| ⌖ | ⌀0.20 Ⓜ | A | B Ⓜ | C |

This means that each controlled hole axis must lie within a cylindrical tolerance zone 0.20 mm in diameter when the hole is at maximum material condition (MMC). The zone is oriented and located from datum reference frame A–B(M)–C. The circled M after B indicates that datum feature B is referenced at maximum material boundary, potentially permitting datum shift when applicable.

Figure: Symbolic position feature-control frame and datum-feature attachments for a mounting plate. Boxed dimensions represent theoretically exact basic locations.

6.4 Basic dimensions and tolerance zones

A basic dimension is theoretically exact and is commonly enclosed in a rectangular box. Its permissible variation is controlled by a geometric tolerance rather than by a plus/minus tolerance applied directly to the basic value.

Position applied to a cylindrical feature generally creates a cylindrical tolerance zone. Profile of a surface creates a three-dimensional boundary around the true profile. Orientation tolerances require a datum when they control orientation relative to another feature; form tolerances generally do not require a datum.

6.5 Material-condition modifiers

For a feature of size:

• MMC is the size containing the most material: smallest hole or largest external shaft;
• LMC is the size containing the least material: largest hole or smallest shaft; and
• RFS means the geometric tolerance applies regardless of feature size.

For an internal feature controlled at MMC, departure from MMC can provide bonus tolerance. If a hole has MMC size D_MMC and actual mating-envelope size D_a, the size departure is:

This expression applies to an internal feature and assumes D_a≥D_MMC. For an external feature, the corresponding departure is MMC size minus actual size. The total available geometric tolerance is the stated MMC tolerance plus valid bonus tolerance, subject to the standard’s envelope and feature requirements.

7. Illustrative Case: Encoder-Motor Mounting Assembly

Background

An illustrative manufacturer is designing an optical encoder module for an industrial motor. The encoder bearing fits into an aluminum housing, while its coupling mounts on a steel shaft. Four holes locate the sensor board relative to the shaft axis.

Problem

Prototype assemblies show three problems: some bearing housings are difficult to assemble, the sensor board occasionally exceeds permitted optical misalignment, and a sealing face is specified merely as “smooth.” The original drawing chains dimensions from several unrelated edges and repeats one hole location.

Analysis

The proposed bearing seat has limits 35.018–35.034 mm. The bearing outer diameter has limits 34.992–35.000 mm.

This is a strict clearance fit. Whether it is acceptable also depends on bearing load, thermal growth and manufacturer guidance.

For a coupling bore of 12.000–12.006 mm and shaft of 12.008–12.015 mm:

This is a strict interference fit, ranging from 0.002 to 0.015 mm interference.

Approach and Solution

The design team:

1. establishes the motor mounting face as primary datum A;
2. uses the bearing-seat axis as a functional secondary datum where the part geometry and standard permit;
3. locates the board-hole pattern with basic dimensions and position rather than independent coordinate tolerances;
4. removes the repeated hole dimension;
5. specifies a measurable sealing-face texture, including Ra value and lay where functionally necessary;
6. verifies the bearing and coupling fits at both size extremes; and
7. evaluates optical alignment through a worst-case tolerance stack before considering RSS optimization.

Outcome

The revised drawing provides one unambiguous datum structure, measurable surface requirements and independently verifiable fits. Inspection can use functional gauging or coordinate measurement aligned to the same datum reference frame used by design.

Learning Points

• Fits must be calculated from limits, not inferred from nominal sizes.
• Functional datum selection can reduce tolerance accumulation.
• GD&T does not compensate for an incorrect datum strategy.
• Surface texture requires a parameter, value and standard—not adjectives such as “smooth.”
• Statistical tolerancing requires process evidence and risk control.

8. Professional Practice, Limitations and Governance

Tolerances should reflect product risk. Safety-critical switchgear, rotating machinery and high-voltage clearances demand documented design verification. Drawing revisions must be controlled so that suppliers and inspectors use the same standard edition.

Inspection capability is a design constraint. A tolerance smaller than measurement uncertainty cannot be reliably verified. Measurement strategy should therefore be considered during design, including temperature, fixturing, datum simulation, probe accessibility and surface filtering.

Model-based definition can embed dimensions and GD&T in a three-dimensional product model. Automated inspection planning and AI-assisted drawing checks can detect missing datums, duplicate dimensions or unusual tolerance values. Such tools do not replace engineering judgment: they may identify syntax without understanding load paths, assembly sequence or failure consequences.

Ethically, engineers should not loosen requirements merely to accept nonconforming parts without a documented functional assessment. Conversely, unjustified tight tolerances waste energy, tooling, material and inspection resources. Formal deviation approval, configuration control and traceable risk analysis are essential.

Chapter Summary

Functional dimensioning communicates design intent through necessary, nonconflicting requirements referenced to meaningful interfaces. Limits define permissible size variation, while signed minimum and maximum clearances provide a rigorous basis for classifying fits. A true transition fit requires both clearance and interference to be possible; zero-clearance boundaries require separate treatment.

Tolerance stacks may be evaluated by guaranteed worst-case analysis or, under justified statistical assumptions, by RSS analysis. Surface-texture specifications communicate roughness parameters, manufacturing constraints and lay using standardized symbol structures. GD&T controls form, profile, orientation, position and runout through feature-control frames and datum-reference systems. ASME Y14.5-2018 retains 12 geometric characteristics and excludes concentricity and symmetry. Successful application requires coordination among design, manufacturing, inspection and configuration management.

Key Terms

Allowance: Intentional signed difference between mating features at maximum material condition.

Basic dimension: Theoretically exact dimension used with geometric tolerances.

Baseline dimensioning: Dimensioning multiple features from a common reference.

Bonus tolerance: Additional geometric tolerance obtained through permitted departure from MMC.

Clearance fit: Fit for which positive clearance exists at all size combinations.

Datum: Theoretically exact reference established from a datum feature.

Datum feature: Physical feature identified to establish a datum.

Feature-control frame: Rectangular compartmented specification of a geometric tolerance.

Fit: Relationship between the sizes of mating features before assembly.

Interference fit: Fit for which interference exists at all size combinations.

Lay: Predominant direction of a surface pattern.

Limit dimension: Maximum or minimum permitted size.

Maximum material condition: Feature size containing the greatest amount of material.

Position: Geometric characteristic controlling feature location relative to true position.

Reference dimension: Informational dimension that does not independently govern acceptance.

Surface texture: Small-scale topographic characteristics of a manufactured surface.

Tolerance: Permissible variation in a dimension or geometric characteristic.

Tolerance stack-up: Accumulated variation of dimensions affecting a functional result.

Transition fit: Fit in which both clearance and interference are possible.

Review Questions

Short Answer Questions

1. Distinguish nominal size, limits of size and tolerance.
2. Why is a zero minimum clearance not sufficient to establish a true transition fit?
3. What does the bar added to a surface-texture symbol indicate?
4. What is the difference between a datum and a datum feature?
5. Name the four form characteristics in ASME Y14.5-2018.

Long Answer Questions

1. Explain how functional dimensioning reduces ambiguity and tolerance accumulation.
2. Compare hole-basis and shaft-basis fit systems, including their manufacturing implications.
3. Explain the assumptions, benefits and limitations of worst-case and RSS tolerance analysis.
4. Describe the structure and interpretation of a feature-control frame containing position, MMC and three datum references.
5. Discuss how surface roughness, lay and manufacturing process can affect an electromechanical product.

Analytical / Application Questions

1. A hole is 24.020–24.045 mm and a shaft is 24.000–24.025 mm. Calculate the minimum and maximum signed clearances and classify the fit.
2. A functional gap is G=A-B-C, where A=80.00±0.15 mm, B=30.00±0.08 mm and C=20.00±0.05 mm. Determine the nominal gap, worst-case limits and RSS half-width under justified assumptions.
3. Develop a datum scheme and position control for a four-hole connector flange. Justify the primary, secondary and tertiary datum features based on assembly function.

Multiple Choice Questions with Answers

Question 1

A dimension is specified as 30.00^+0.04_-0.02 mm. What is its total size tolerance?

A. 0.02 mm

B. 0.04 mm

C. 0.06 mm

D. 0.08 mm

Correct Answer: C. 0.06 mm

Question 2

A hole has limits 20.010–20.030 mm and a shaft has limits 19.990–20.015 mm. How is the fit classified?

A. Strict clearance fit

B. Strict interference fit

C. True transition fit

D. Exact line fit

Correct Answer: C. True transition fit

Question 3

For a hole of 10.000–10.010 mm and a shaft of 9.990–10.000 mm, which classification is correct?

A. True transition fit

B. Clearance-to-line boundary condition

C. Strict interference fit

D. Interference-to-line boundary condition

Correct Answer: B. Clearance-to-line boundary condition

Question 4

How many geometric characteristic symbols are retained in ASME Y14.5-2018?

A. 10

B. 12

C. 14

D. 16

Correct Answer: B. 12

Question 5

Which pair of geometric characteristics was removed from ASME Y14.5-2018?

A. Circularity and cylindricity

B. Position and profile of a line

C. Concentricity and symmetry

D. Circular runout and total runout

Correct Answer: C. Concentricity and symmetry

Question 6

In a surface-texture symbol, what does adding a bar to the basic symbol indicate?

A. Material removal is required

B. Material removal is prohibited

C. The surface must have circular lay

D. The value is a reference requirement

Correct Answer: A. Material removal is required

Question 7

For G=L-A-B, where L=100.00±0.20 mm, A=40.00±0.10 mm and B=25.00±0.08 mm, what is the worst-case result?

A. 35.00±0.18 mm

B. 35.00±0.24 mm

C. 35.00±0.30 mm

D. 35.00±0.38 mm

Correct Answer: D. 35.00±0.38 mm

Question 8

Which statement is required for valid use of a simple RSS tolerance stack?

A. Every component must always be at its dimensional limit

B. The variations should be independent, stable and approximately centered

C. All dimensions must have identical nominal values

D. The stack must contain exactly three dimensions

Correct Answer: B. The variations should be independent, stable and approximately centered

Question 9

In the feature-control frame | ⌖ | ⌀0.20 Ⓜ | A | B | C |, what does ⌀0.20 specify for a cylindrical feature?

A. A cylindrical positional tolerance zone 0.20 mm in diameter

B. A radial clearance of exactly 0.20 mm

C. A surface roughness of 0.20 μm

D. A hole size tolerance of ±0.20 mm

Correct Answer: A. A cylindrical positional tolerance zone 0.20 mm in diameter

Question 10

Which dimensioning method usually minimizes accumulation of location variation across a hole pattern?

A. Chain dimensioning from each preceding hole

B. Repeated dimensioning in multiple views

C. Baseline dimensioning from a common functional reference

D. Omitting the feature locations

Correct Answer: C. Baseline dimensioning from a common functional reference