The ISO system of limits and fits, codified in ISO 286, provides the universal engineering language for specifying how mating cylindrical parts relate dimensionally. Every bolted assembly, bearing seat, and press-fit hub relies on precisely defined tolerance zones and fundamental deviations to guarantee function — whether the design demands free rotation, a snug location, or a permanent press.
Manual computation of limit dimensions from raw ISO tables is tedious and error-prone, particularly when switching between metric and US customary units or evaluating multiple candidate fits for a single joint. Automated tolerance analysis eliminates transcription mistakes, instantly classifies the resulting fit type (clearance, transition, or interference), and surfaces the maximum and minimum material conditions critical for process control and quality assurance.
Required Project Parameters
Before obtaining results, the following design specifications must be established:
- Measurement Standard — Metric (mm/μm) or US Standard (in/thou). Imperial values are internally converted to metric for ISO 286 processing and then converted back.
- Nominal Diameter ($D$) — The basic size of the mating bore and shaft, constrained to the ISO 286 range of 3–500 mm (approximately 0.125–20 in).
- Hole Tolerance Letter (A–Z) — The alphabetic code defining the fundamental deviation (position) of the hole's tolerance zone relative to the nominal size.
- Hole IT Grade (5–11) — The International Tolerance grade controlling the width of the hole's tolerance zone.
- Shaft Tolerance Letter (a–z) — The alphabetic code defining the fundamental deviation for the shaft.
- Shaft IT Grade (5–11) — The International Tolerance grade controlling the width of the shaft's tolerance zone.
Dimensional Tolerance Theory: From Size Bins to Fit Classification
Standard Size Bins and the Geometric Mean
ISO 286 does not calculate tolerances from arbitrary diameters. Instead, the standard partitions the entire 3–500 mm range into discrete size bins (e.g., 3–6 mm, 6–10 mm, 10–18 mm, and so on). Each bin has a lower boundary $D_{min}$ and an upper boundary $D_{max}$.
The representative diameter for any bin is its geometric mean:
$$D_{geom} = \sqrt{D_{min} \times D_{max}}$$
Using the geometric mean rather than the arithmetic mean provides a dimensionally balanced anchor that accounts for the nonlinear scaling of machining errors across the diameter range. A 10 mm shaft and a 100 mm shaft do not exhibit the same absolute dimensional variation, and the geometric mean feeds directly into the tolerance factor formula that models this relationship.
The Standard Tolerance Factor ($i$)
The fundamental building block of the entire ISO tolerance system is the standard tolerance factor:
$$i = 0.45 \times \sqrt[3]{D_{geom}} + 0.001 \times D_{geom}$$
where $D_{geom}$ is in millimeters and $i$ is in micrometers (μm).
This expression is not arbitrary — it encodes two distinct physical phenomena. The cube-root term ($0.45 \times \sqrt[3]{D}$) models the parabolic growth of machining error as workpiece diameter increases. Larger parts deflect more under cutting forces, introduce greater concentricity runout, and are harder to fixture rigidly.
The linear term ($0.001 \times D$) accounts for thermal expansion and measurement uncertainty at the standard metrology reference temperature of 20 °C. Together, these two components produce a tolerance function that closely matches empirical shop-floor capability data across the entire standardized diameter range.
IT Grade Multipliers: Scaling the Tolerance Band
Each International Tolerance (IT) grade is defined as a fixed multiple $k$ of the standard tolerance factor $i$. The tolerance magnitude for a given grade is simply:
$$T = k \times i$$
The multipliers are hardcoded by the ISO standard:
| IT Grade | Multiplier ($k$) |
|---|---|
| IT5 | 7 |
| IT6 | 10 |
| IT7 | 16 |
| IT8 | 25 |
| IT9 | 40 |
| IT10 | 64 |
| IT11 | 100 |
Higher IT numbers yield wider tolerance bands, corresponding to coarser manufacturing processes. IT5–IT6 are typical of precision grinding, while IT10–IT11 correspond to rough turning or stamping operations.
Fundamental Deviation: Positioning the Tolerance Zone
The IT grade determines the width of the tolerance zone. The fundamental deviation — encoded by the tolerance letter — determines where that zone sits relative to the nominal size. The deviation formulas are empirical power-law relationships derived from decades of manufacturing data.
For shaft deviations (lowercase letters), representative formulas include:
$$e_s(\text{c}) = -52 \times D^{0.2}$$
$$e_s(\text{f}) = -5.5 \times D^{0.41}$$
$$e_s(\text{g}) = -2.5 \times D^{0.34}$$
$$e_s(\text{h}) = 0$$
The special case of shaft 'h' has zero upper deviation, meaning the shaft's maximum material condition equals the nominal diameter exactly. Similarly, hole 'H' has zero lower deviation ($EI = 0$), placing the hole's minimum diameter at nominal.
This is precisely why standard manufacturing practice overwhelmingly favors the Hole Basis System. It is significantly more cost-effective to produce a standard 'H' hole using fixed-size reamers or broaches and then adjust the shaft's external diameter on a lathe or cylindrical grinder to achieve the desired fit class. The hole tooling remains constant across an entire family of fits.
Limit Dimensions and Fit Classification
Once the tolerance magnitude and fundamental deviation are known for both hole and shaft, the limit dimensions are computed:
$$D_{hole,max} = D_{nom} + ES$$
$$D_{hole,min} = D_{nom} + EI$$
$$D_{shaft,max} = D_{nom} + es$$
$$D_{shaft,min} = D_{nom} + ei$$
where $ES$ and $EI$ are the upper and lower deviations of the hole, and $es$ and $ei$ are those of the shaft.
The fit type is then classified based on the extreme material conditions:
- Clearance Fit — Minimum hole size exceeds maximum shaft size. Assembly is always free.
- Interference Fit — Maximum hole size is smaller than minimum shaft size. Assembly requires force or thermal methods.
- Transition Fit — Maximum clearance is positive and maximum interference is positive. The assembly may be either loose or tight depending on the actual sizes of individual parts within their tolerance bands.
Transition fits are particularly sensitive to thermal gradients. A fit calculated as transitional at the standard 20 °C reference temperature may shift into definite interference at elevated operating temperatures, or into clearance at sub-ambient conditions. This must be evaluated during detailed design review.
ISO 286 Tolerance Grades and Preferred Fit Designations
The following reference consolidates the most widely specified fits across general mechanical engineering, organized by fit type and typical application domain.
| Fit Designation | Fit Type | Typical Application | Shaft Tolerance |
|---|---|---|---|
| H7/g6 | Clearance | Sliding fits, precision bearings | g6 |
| H7/f7 | Clearance | Running fits, journal bearings | f7 |
| H8/f7 | Clearance | Loose running, general rotation | f7 |
| H11/c11 | Clearance | Very loose, agricultural machinery | c11 |
| H7/h6 | Transition | Location fits, spigot joints | h6 |
| H7/k6 | Transition | Keyed assemblies, light press | k6 |
| H7/n6 | Transition | Tight location, dowel pins | n6 |
| H7/p6 | Interference | Press fits, permanent assembly | p6 |
| H7/s6 | Interference | Heavy press, shrink fits | s6 |
The default pairing of Hole IT7 / Shaft IT6 reflects a core machining reality: internal operations such as boring and internal grinding are inherently less rigid and harder to control than external turning or cylindrical grinding. Assigning the shaft one IT grade tighter than the hole compensates for this asymmetry in manufacturing capability and balances production costs without compromising fit function.
The table below provides representative tolerance magnitudes (in μm) for selected IT grades across common size bins, illustrating how tolerance widens with both diameter and IT number.
| Nominal Range (mm) | $D_{geom}$ (mm) | IT6 (μm) | IT7 (μm) | IT8 (μm) | IT9 (μm) |
|---|---|---|---|---|---|
| 3–6 | 4.24 | 8 | 12 | 18 | 30 |
| 6–10 | 7.75 | 9 | 15 | 22 | 36 |
| 10–18 | 13.4 | 11 | 18 | 27 | 43 |
| 18–30 | 23.2 | 13 | 21 | 33 | 52 |
| 30–50 | 38.7 | 16 | 25 | 39 | 62 |
| 50–80 | 63.2 | 19 | 30 | 46 | 74 |
| 80–120 | 98.0 | 22 | 35 | 54 | 87 |
Interpreting Fit Outcomes for Assembly Engineering and Process Control
Clearance Analysis and Functional Implications
The maximum clearance occurs when the hole is at its upper material limit and the shaft is at its lower material limit. Conversely, the minimum clearance represents the tightest permissible assembly. These two values define the functional envelope of the joint.
For rotating assemblies, minimum clearance must exceed the combined thermal expansion of both parts at the maximum operating temperature, plus any lubricant film thickness requirements. Insufficient minimum clearance leads to seizure under thermal load, while excessive maximum clearance introduces vibration, noise, and reduced bearing life.
Interference Analysis and Assembly Method Selection
When a fit is classified as interference, the maximum interference governs the required assembly force or the temperature differential for shrink fitting. The minimum interference determines the residual grip force that holds the joint together under operational loads.
A common engineering guideline for shrink fitting is to heat the outer part (or cool the inner part) until the thermal expansion exceeds the maximum interference by at least 25% to ensure reliable assembly without galling. For a steel-on-steel joint, the approximate heating differential is:
$$\Delta T = \frac{\delta_{max}}{D_{nom} \times \alpha}$$
where $\delta_{max}$ is the maximum interference and $\alpha$ is the coefficient of thermal expansion (approximately $12 \times 10^{-6}$ per °C for carbon steel).
Unit Conversion Considerations
When working in US customary units, all dimensions supplied in inches are divided by 25.4 to convert to millimeters before entering the ISO 286 computation pipeline. After processing, outputs are multiplied by 25.4 and rounded: three decimal places for dimensions in inches, two decimal places for deviations in thousandths of an inch (thou). This rounding convention balances practical shop-floor readability with adequate numerical precision.
Frequently Asked Questions
The convention of specifying the shaft one IT grade finer than the mating hole (e.g., H7/g6 rather than H7/g7) is rooted in manufacturing economics, not geometric theory. Internal machining operations — boring, reaming, internal grinding — are performed with tools that overhang from the spindle into an enclosed space. This geometry reduces tool rigidity, limits chip evacuation, and makes real-time dimensional feedback more difficult.
External turning and cylindrical grinding, by contrast, allow the operator or CNC system direct visual and tactile access to the workpiece. Tooling is shorter and stiffer, and in-process gauging is straightforward. As a result, achieving a given tolerance grade on a shaft costs less than achieving the same grade inside a bore.
By assigning the tighter (and thus more expensive) grade to the easier-to-machine feature, the total manufacturing cost of the assembly is minimized without sacrificing the required fit quality.
A transition fit means that, within the specified tolerance bands, some individual assemblies will exhibit slight clearance while others will show slight interference. The classification is strictly valid only at the ISO reference temperature of 20 °C and assumes zero form errors (perfect roundness and straightness).
In practice, thermal gradients introduce systematic bias. For a steel bore-and-shaft combination with a nominal diameter of 50 mm, a 30 °C rise above the reference temperature produces approximately 18 μm of differential expansion (assuming the housing heats first). This can shift a marginal transition fit firmly into interference territory, potentially preventing assembly or causing unexpected seizure.
Designers working with transition fits should always perform a thermal stack-up analysis at the extremes of the expected operating envelope and add a form-error budget derived from the GD&T (Geometric Dimensioning & Tolerancing) callouts on the part drawings.
The standard tolerance factor $i = 0.45 \times \sqrt[3]{D} + 0.001 \times D$ was empirically derived during the development of the ISA (predecessor to ISO) tolerance system in the early 20th century, but it has a solid physical rationale.
The cube-root dependence arises from elastic beam deflection theory. A turning tool or boring bar behaves as a cantilevered beam; its deflection under a constant cutting force scales with the cube of the overhang length. As workpiece diameter increases, so does the required tool reach, and achievable dimensional accuracy degrades following a cube-root relationship.
The linear term ($0.001 \times D$) represents the metrological limit imposed by thermal expansion. At the standard reference temperature of 20 °C, steel expands at roughly 12 μm per meter per degree Celsius. For a diameter $D$ measured with a temperature uncertainty of approximately 1 °C, the measurement ambiguity is on the order of $0.012 \times D$ μm — well approximated by $0.001 \times D$ after rounding and empirical calibration.
Together, the two terms produce a tolerance function that closely matches observed shop-floor capability data across the full 3–500 mm diameter range.
Automated Tolerance Computation: Engineering Precision Without Manual Risk
The ISO 286 framework involves nested lookups — size bin identification, geometric mean computation, tolerance factor calculation, deviation formula selection, and limit stacking — each step compounding the risk of transcription and arithmetic errors. A single misread row in a printed tolerance table can shift a fit class from clearance to interference, with consequences ranging from seized bearings to scrapped production batches.
Automated computation eliminates these failure modes entirely. It enforces ISO-compliant rounding, handles metric-to-imperial conversion with proper decimal precision, and instantly evaluates alternative fit designations during the parametric design phase. For manufacturing engineers and quality teams, this translates directly into faster design iterations, fewer inspection escapes, and traceable, auditable tolerance records.