Dimensioning, Fits, Tolerances, and Surface Texture

Introduction

An engineering drawing is a contractual description of a product. It must communicate not only the nominal size and shape of a component but also the permissible variation, the surfaces from which measurements are made, the required surface condition, and the functional relationships between parts. A dimension such as 25 mm is incomplete unless the drawing also establishes how much the size may vary and how that variation affects assembly and performance.

Dimensioning and tolerancing therefore connect design intent with manufacturing and inspection. Excessively tight tolerances increase machining time, inspection effort, scrap, and cost. Excessively loose tolerances may cause leakage, vibration, loss of accuracy, or failure to assemble. The engineer's task is to specify the least restrictive requirements that guarantee function.

This chapter follows internationally common principles associated with ISO GPS practices and ASME Y14.5. Particular company, national, or customer standards must always be consulted because symbol conventions and default rules can differ.

Learning Objectives

After studying this chapter, the reader should be able to:

• create functional and unambiguous dimensions;
• distinguish size tolerance, geometric tolerance, allowance, and fit;
• select and interpret clearance, transition, and interference fits;
• calculate limits, tolerances, allowances, and simple tolerance stacks;
• apply datum concepts and basic geometric dimensioning and tolerancing (GD&T);
• interpret surface-texture requirements and relate them to manufacturing processes;
• make design decisions that balance interchangeability, cost, capability, and inspection.

1. Functional Dimensioning

1.1 Dimensions as design requirements

A dimension defines a measurable characteristic of a feature. Dimensions may specify:

• Size, such as a hole diameter or plate thickness;
• Location, such as the distance between hole axes;
• Form, such as straightness or circularity;
• Orientation, such as perpendicularity or angularity;
• Location control, such as position;
• Surface condition, such as roughness or lay;
• Manufacturing or inspection information, such as datums, notes, and material condition.

A dimension should be placed where its meaning is clearest and should normally be shown only once. Repeating a dimension can create conflicting requirements. Dimensions should be arranged from functional references rather than from arbitrary edges whenever possible.

1.2 Dimensioning methods

Chain dimensioning places dimensions end to end. It is easy to read but may accumulate manufacturing variation. Baseline dimensioning locates several features from one common reference and is preferred when a functional datum controls assembly. Coordinate dimensioning gives feature locations from two perpendicular reference axes. Tabular dimensioning is useful for repeated features, while ordinate dimensioning uses a common origin and numbered coordinates.

A dimension should be selected according to the function of the part. For example, if a shaft must align with a bearing and a gear, the bearing seat and gear seat should be related to a common axis or datum rather than dimensioned through a long chain of intermediate steps.

1.3 Basic dimension terminology

The nominal size is the stated size used to identify a feature. The actual size is the measured size. The limits of size are the maximum and minimum permitted sizes. The upper deviation is the algebraic difference between the upper limit and the nominal size; the lower deviation is the corresponding difference for the lower limit.

For a feature specified as 40.00 ± 0.05 mm:

The tolerance is a magnitude and is therefore non-negative. The deviations may be positive, negative, or zero.

1.4 Direct, reference, and basic dimensions

A direct dimension is a required value subject to a tolerance. A reference dimension is informational, often enclosed in parentheses, and is not normally used as an independent manufacturing requirement. A basic dimension is theoretically exact; its permissible variation is established by a geometric tolerance such as position.

Dimensions should describe design intent rather than prescribe an unnecessary manufacturing sequence. A drawing may specify the required result without forcing a particular machine or operation, unless the process itself is essential to function.

2. Tolerance Systems

2.1 Why tolerances are necessary

No manufacturing process produces a perfectly exact size. Variation arises from tool wear, thermal expansion, machine accuracy, material properties, operator technique, measurement uncertainty, and environmental conditions. A tolerance converts an ideal dimension into an acceptable interval.

The total size tolerance is:

where T is the total tolerance, U is the upper limit, and L is the lower limit. Both limits must have the same units.

A unilateral tolerance varies in only one direction from nominal, such as 20.00 +0.02/−0.00 mm. A bilateral tolerance permits variation in both directions, such as 20.00 ± 0.02 mm. A limit dimension gives the two boundary values directly, such as 19.98–20.02 mm.

2.2 Standard tolerance grades and deviations

The ISO system of limits and fits defines tolerance grades and fundamental deviations. A grade expresses the magnitude of the tolerance interval; a smaller grade number generally means a tighter tolerance. A letter identifies the position of the tolerance zone relative to the zero line. Capital letters are normally used for holes and lowercase letters for shafts.

For a nominal size D, an ISO tolerance value is commonly represented by:

where i is the standard tolerance unit in micrometres and D is the geometric mean diameter of the relevant diameter step in millimetres. This relationship is used within the ISO limits-and-fits system; it is not a universal formula for every manufacturing tolerance. A grade tolerance is expressed as a multiple of i, with the multiplier determined by the selected grade and size range.

The hole-basis system keeps the lower deviation of the hole at the zero line, commonly designated H. Different shaft tolerance positions are then selected to obtain the required fit. The shaft-basis system keeps the upper deviation of the shaft at the zero line, commonly designated h, and varies the hole position. Hole basis is widely used because standard drills, reamers, and hole-making tools make standard hole sizes economically convenient.

2.3 Allowance and fit

The allowance is the intentional difference between mating parts at maximum material condition. For a clearance fit, it is the minimum clearance. For an interference fit, it is the maximum interference, considered with its sign convention or stated explicitly as interference.

For a hole and shaft:

where H_max and H_min are the maximum and minimum hole sizes, and S_max and S_min are the maximum and minimum shaft sizes. Positive values indicate clearance; a negative value indicates interference.

The fit tolerance, representing the possible range of clearance or interference, is:

This worst-case relation assumes independent size extremes and is applicable to limit-based assembly analysis.

2.4 Types of fit

A clearance fit always has a positive clearance. It is used where relative motion, lubrication, or easy assembly is required. A transition fit may produce either a small clearance or a small interference. It is used for accurate location with possible pressing or light tapping. An interference fit always produces interference and requires force, heating, cooling, or both for assembly.

Fit selection depends on load, speed, alignment, temperature, material, assembly method, service environment, and maintenance requirements. A press fit that is acceptable for a steel hub on a steel shaft may be unsuitable for a thin aluminium housing because the housing may split or permanently deform.

3. Worked Example: Limits and Fit Selection

A shaft and hole have the following limits:

• Hole: 30.000 to 30.021 mm
• Shaft: 29.980 to 29.994 mm

The maximum clearance is:

The minimum clearance is:

Because the minimum clearance is positive, the assembly is a clearance fit. The fit tolerance is:

The design permits clearance from 0.006 to 0.041 mm. If the shaft must rotate continuously, the designer should verify lubrication, thermal expansion, expected speed, and manufacturing capability rather than selecting the fit from size limits alone.

4. Geometric Dimensioning and Tolerancing

Size tolerances alone do not control all conditions that affect assembly. A hole can have an acceptable diameter but be tilted, misplaced, or out of round. GD&T controls geometric variation using symbols, tolerance zones, and datums.

4.1 Form controls

Form controls apply to individual features and generally do not require a datum:

• Straightness controls a line element or derived axis;
• Flatness controls a surface between two parallel planes;
• Circularity controls each circular cross-section between concentric circles;
• Cylindricity controls the entire cylindrical surface between coaxial cylinders.

4.2 Orientation controls

Orientation controls establish relationships to a datum:

• Parallelism controls orientation at 0° to a datum;
• Perpendicularity controls orientation at 90°;
• Angularity controls a specified basic angle other than 0° or 90°.

4.3 Location and profile controls

Position controls the location of a feature relative to basic dimensions and datums. It is often more informative than separate coordinate plus size tolerances because it creates a cylindrical or, for certain features, planar tolerance zone. Profile of a line and profile of a surface control the boundary of a feature relative to a theoretically exact profile. Circular runout and total runout control variation during rotation about a datum axis.

A geometric tolerance should be chosen according to the function it protects. Applying many tight controls without functional justification increases inspection and manufacturing cost.

4.4 Datums and the datum reference frame

A datum is a theoretically exact reference derived from a physical feature. A primary datum establishes the first constraint, a secondary datum establishes additional orientation or location, and a tertiary datum completes the reference frame. The familiar three-plane model constrains six degrees of freedom: three translations and three rotations.

Datum selection should reflect assembly and measurement practice. The primary datum is commonly the surface that seats against another component; the secondary datum may locate an axis or side; the tertiary datum may fix clocking or final translation. A datum chosen merely because it is convenient to draw can cause functional errors.

A datum feature is the actual surface, axis, or center plane used to establish a datum. The distinction matters: the datum is theoretically exact, while the physical datum feature contains manufacturing variation.

4.5 Maximum material condition and bonus tolerance

For a shaft, maximum material condition (MMC) is its largest permitted size. For a hole, MMC is its smallest permitted size. If a position tolerance is applied at MMC with the appropriate modifier, departure from MMC may provide bonus tolerance.

For a hole:

For a shaft:

These expressions give a non-negative bonus when the feature departs from MMC. The concept is useful when assembly clearance increases as a hole becomes larger or a shaft becomes smaller. Modifiers must be interpreted according to the applicable standard; they should not be added casually to a feature-control frame.

5. Tolerance Stack-Up

A tolerance stack-up predicts how variations in several dimensions affect a functional requirement such as a gap, alignment, or travel.

5.1 Worst-case method

If a functional dimension is the sum of independent contributors, the worst-case tolerance is the sum of their absolute tolerances:

The resulting total range is conservative because it assumes every dimension simultaneously reaches the combination that produces the largest error. It is appropriate where interchangeability and guaranteed limits are essential.

5.2 Statistical root-sum-square method

When contributors are independent, approximately centered, and statistically characterized, a statistical estimate may be calculated as:

This is a standard root-sum-square estimate when each t_i represents a one-standard-deviation contribution or a consistently defined statistical tolerance component. It is not a guarantee of the absolute limit. Correlation, bias, non-normal distributions, drift, and non-centred processes can make the estimate misleading.

5.3 Worked stack-up example

A required assembly gap is nominally 2.00 mm and is calculated as:

G = A − B − C, where A = 50.00 ± 0.10 mm, B = 30.00 ± 0.05 mm, and C = 18.00 ± 0.03 mm.

The nominal gap is:

The worst-case tolerance is:

Therefore the worst-case gap is 1.82 to 2.18 mm. If the dimensions are independent and the stated values are compatible statistical tolerances, the RSS value is:

The RSS interval would be approximately 1.884 to 2.116 mm under its statistical assumptions, but it must not be presented as a guaranteed limit.

5.4 Design response to excessive stack-up

If a stack-up violates a functional requirement, possible responses include changing the dimensioning scheme, reducing only the most influential tolerances, using a functional datum, introducing adjustment, redesigning the interface, or applying selective assembly. Tightening every dimension is usually the least economical response.

6. Surface Texture

Surface texture describes fine deviations from an ideal surface. It is distinct from larger-scale waviness and from form error. Texture results from cutting marks, abrasive action, vibration, tool geometry, material behaviour, and finishing processes.

6.1 Roughness parameters

The arithmetic mean roughness, commonly denoted R_a, is:

where y(x) is the signed profile deviation from the mean line, L is the evaluation length, and the result has units of length, commonly micrometres. The absolute value is required because positive and negative deviations would otherwise cancel.

The root-mean-square roughness is:

R_q is more sensitive to high peaks and valleys than R_a. Roughness values are meaningful only with the specified measurement standard, cut-off, evaluation length, filtering method, and direction of measurement.

Other parameters, such as maximum profile peak height, valley depth, total height, and bearing-area parameters, may be more relevant for sealing, fatigue, lubrication, or wear than R_a alone.

6.2 Surface-texture requirements

A surface-texture indication may specify a roughness limit, machining allowance, lay direction, or a process restriction. The symbol and notation must follow the governing drafting standard. A general roughness note may apply to unspecified surfaces, while critical surfaces receive individual requirements.

The required texture should be functional. A bearing seat may need controlled roughness and waviness; a sealing surface may require limits on peaks and valleys; a decorative surface may be judged visually; a nonfunctional cast surface may need no unnecessary machining.

6.3 Manufacturing and cost implications

Typical tendencies are as follows:

These are general tendencies, not guaranteed values. Capability must be established for the actual machine, material, tool, operator, and inspection method.

7. Inspection, Capability, and Interchangeability

Interchangeability means that any acceptable part can assemble with any mating acceptable part without hand fitting. It requires compatible limits, a stable process, correct datums, and a measurement system capable of resolving the requirement.

A process capability index is often used when a stable, approximately normal process is demonstrated:

where USL and LSL are upper and lower specification limits and σ is process standard deviation. The index C_p measures potential capability assuming the process is centred. A related centred capability index is:

where μ is the process mean. These indices are meaningful only after checking statistical stability and ensuring that the measurement system is adequate. They do not replace functional tolerance analysis.

Measurement uncertainty must be small relative to the tolerance. Calibration, traceability, temperature control, probe access, datum simulation, and repeatability are essential. A digital display with many digits does not necessarily provide accurate measurement.

8. Applied Case: Pump-Shaft and Bearing Housing

Background

This illustrative case concerns a small centrifugal pump. A steel shaft rotates in a rolling bearing, and the bearing outer ring is located in a cast housing. The assembly must rotate freely, resist creep under operating load, and remain serviceable.

Problem

An initial design used a very tight shaft-to-bearing fit and a rough housing bore. Assembly required excessive force, several bearings were damaged during pressing, and operating temperature caused occasional seizure.

Analysis

The design team identified three interacting requirements: the shaft seat must prevent unwanted ring movement, the bearing must not be damaged during assembly, and the housing bore must provide adequate seating and heat transfer. They reviewed shaft and bore limits, surface texture, roundness, runout, datum references, and thermal expansion. They also examined actual process data rather than relying only on nominal dimensions.

The shaft fit was changed to a controlled interference appropriate to the bearing manufacturer's recommendation and the expected rotating load. The housing fit was selected to locate the outer ring without excessive assembly force. A shoulder datum and a controlled face-to-axis perpendicularity requirement were added so that the bearing seat and locating face established the functional reference frame.

The drawing specified a practical bore texture and form control, rather than an unnecessarily small roughness value. Inspection instructions defined the measuring datum, temperature condition, bore gauge method, and sampling plan. A capability study was conducted after the machining process was stabilized.

Outcome

Assembly force became predictable, bearing damage was reduced, and running temperature remained within the design range. The cost reduction came not from eliminating tolerances but from assigning tight control only to the shaft seat, bore, face relationship, and surface conditions that affected function.

Learning points

1. Fit selection must include load, speed, temperature, material, and assembly method.
2. A size tolerance cannot substitute for control of orientation, location, or form.
3. Surface texture must be specified with regard to function and measurement method.
4. Datums should represent how the component is assembled and inspected.
5. Capability evidence should guide tolerance decisions; nominal values alone are insufficient.

9. Best Practices and Common Errors

Effective practice includes defining functional requirements first, using one clear dimension for each requirement, selecting a datum reference frame based on assembly, analysing critical tolerance chains, consulting standard fit tables, and communicating inspection conditions.

Common errors include repeating dimensions, mixing units, tolerancing a nonfunctional feature tightly, using chain dimensions for a critical locating pattern, treating a reference dimension as a requirement, assuming a surface symbol guarantees a manufacturing process, ignoring thermal effects, and using RSS results as guaranteed limits. Another serious error is specifying a geometric tolerance without understanding the datum simulator or the inspection equipment required.

Computer-aided design systems can automate tolerance annotations, hole tables, and interference checks. Model-based definition can associate product manufacturing information with 3D geometry. However, software cannot decide whether a tolerance is functionally justified, whether a process is capable, or whether an inspection method represents actual assembly. Engineers remain responsible for standards compliance and design judgement.

Chapter Summary

Functional dimensioning communicates the size, location, geometry, and surface requirements needed for a part to work and to be manufactured economically. Limits define acceptable size variation; tolerance is the difference between limits; allowance describes the intended relationship between mating features. Clearance, transition, and interference fits are selected from functional, material, thermal, load, and assembly considerations.

GD&T controls form, orientation, location, profile, and runout more directly than size dimensions alone. Datums establish a meaningful reference frame for assembly and inspection. Tolerance stack-up may be evaluated by worst-case addition when guaranteed limits are required, or by root-sum-square methods when justified statistical assumptions apply. Surface texture affects friction, wear, sealing, fatigue, and manufacturing cost, and must be specified and measured using an appropriate standard. Reliable drawings balance interchangeability and performance against process capability, inspection effort, and cost.

Key Terms

Nominal size: The stated size used to identify a feature.

Actual size: The measured size of a manufactured feature.

Limit of size: The maximum or minimum permitted size.

Tolerance: The difference between the upper and lower limits.

Deviation: The algebraic difference between a limit and the nominal size.

Allowance: The intentional difference between mating features at maximum material condition.

Fit: The relationship between the sizes of mating features.

Clearance fit: A fit that always provides positive clearance.

Transition fit: A fit that may provide clearance or interference.

Interference fit: A fit that always provides interference.

Datum: A theoretically exact reference established from a datum feature.

GD&T: A symbolic system for controlling geometric variation.

MMC: Maximum material condition of a feature.

Bonus tolerance: Additional geometric tolerance obtained by departure from MMC when permitted.

Tolerance stack-up: The combined effect of variations in several dimensions.

Surface texture: Fine deviations of a surface from its ideal form.

Ra: Arithmetic mean roughness over a specified evaluation length.

Waviness: More widely spaced surface variation than roughness.

Interchangeability: The ability of acceptable mating parts to assemble without individual fitting.

Process capability: The demonstrated ability of a stable process to meet specification limits.

Review Questions

Short Answer Questions

1. Distinguish nominal size, actual size, limit of size, and tolerance.
2. What is the difference between a clearance fit and an interference fit?
3. Why are datums required in geometric dimensioning and tolerancing?
4. Define maximum material condition and bonus tolerance.
5. Why can a surface-texture value not be interpreted independently of its measurement conditions?

Long Answer Questions

1. Explain chain, baseline, coordinate, and ordinate dimensioning, and identify an appropriate application for each.
2. Describe the ISO hole-basis and shaft-basis systems and explain why the hole-basis system is widely used.
3. Explain the major classes of GD&T controls and relate them to manufacturing and assembly functions.
4. Compare worst-case and root-sum-square tolerance stack-up methods, including their assumptions and limitations.
5. Discuss how tolerance, surface texture, process capability, inspection, and product cost should be balanced during design.

Analytical / Application Questions

1. A hole is 20.000–20.018 mm and a shaft is 19.990–20.006 mm. Calculate the maximum and minimum clearance and classify the fit.
2. A functional dimension is X = A − B + C, with A = 80.00 ± 0.08 mm, B = 25.00 ± 0.04 mm, and C = 10.00 ± 0.03 mm. Determine the nominal value and worst-case limits of X.
3. A machined plate contains four holes located from two edges, but the mating component is located from a central bore and a side face. Propose a datum scheme and dimensioning strategy, and justify how it could improve assembly and inspection.

Multiple Choice Questions with Answers

Question 1

A hole has limits of 30.000 mm and 30.021 mm, while its mating shaft has limits of 29.980 mm and 29.994 mm. What is the minimum clearance?

A. 0.006 mm

B. 0.014 mm

C. 0.021 mm

D. 0.041 mm

Correct Answer: A. 0.006 mm

Question 2

Which fit always has positive clearance?

A. Interference fit

B. Transition fit

C. Clearance fit

D. Press fit

Correct Answer: C. Clearance fit

Question 3

For a dimension 40.00 ± 0.05 mm, what is the total tolerance?

A. 0.05 mm

B. 0.10 mm

C. 39.95 mm

D. 40.05 mm

Correct Answer: B. 0.10 mm

Question 4

Which GD&T control primarily controls the location of a feature relative to basic dimensions and datums?

A. Position

B. Flatness

C. Circularity

D. Straightness

Correct Answer: A. Position

Question 5

For a hole, maximum material condition is the:

A. Largest permitted hole size

B. Smallest permitted hole size

C. Nominal hole size

D. Average measured hole size

Correct Answer: B. Smallest permitted hole size

Question 6

The worst-case tolerance stack-up for independent contributors is generally calculated by:

A. Multiplying all tolerances

B. Adding their absolute tolerances

C. Averaging the tolerances

D. Subtracting the smallest tolerance from the largest

Correct Answer: B. Adding their absolute tolerances

Question 7

Which statement about root-sum-square tolerance analysis is correct?

A. It always guarantees absolute assembly limits

B. It assumes no contributors are present

C. It is a statistical estimate requiring appropriate assumptions

D. It is used only for surface roughness

Correct Answer: C. It is a statistical estimate requiring appropriate assumptions

Question 8

In the roughness parameter R_a, why is the absolute value of profile deviation used?

A. To make the result dimensionless

B. To prevent positive and negative deviations cancelling

C. To measure only waviness

D. To convert millimetres to micrometres

Correct Answer: B. To prevent positive and negative deviations cancelling

Question 9

A primary datum in a datum reference frame primarily serves to:

A. Establish the first functional constraint

B. Specify the material grade

C. Define the surface colour

D. Set the cutting speed

Correct Answer: A. Establish the first functional constraint

Question 10

Which capability index considers both process spread and displacement of the process mean from specification limits?

A. Ra

B. Cp

C. Cpk

D. Tfit

Correct Answer: C. Cpk