CNC machining tolerances define how much a machined feature is allowed to vary from its nominal dimension. A tolerance is not simply a number added to a drawing. It is a statement about what the part must still do when real manufacturing variation occurs. If a hole accepts a pin, two plates must align, a shaft must rotate in a bearing, or several parts must assemble without force, the tolerance should be based on that function.
The better question is not “How tight can a machine hold?” but “How tight does this feature need to be?” A tolerance tighter than the functional requirement can add machining, inspection, scrap risk, and cost without improving the product.
Manufacturing guidance consistently shows that tighter tolerances can increase cost, change the preferred process, require more specialized inspection, and behave differently depending on material and geometry. Practical engineering discussions also show that designers struggle with default tolerances, hole positioning, tolerance stack-up, GD&T, fit selection, and making drawings easy for machinists and inspectors to interpret. This guide from Tuofa CNC Germany brings those questions together.
What Is a CNC Machining Tolerance?
A tolerance is the permitted variation around a nominal dimension. For example, 20.00 ±0.05 mm means the acceptable size is 19.95 to 20.05 mm. A unilateral tolerance such as 20.00 +0.10/-0.00 mm allows the feature to range from 20.00 to 20.10 mm.
The nominal CAD dimension is not the complete manufacturing requirement. The drawing must communicate acceptable limits. Tolerancing does not eliminate variation; it defines enough control for the part to remain functional and interchangeable.
How Tight Should a CNC Tolerance Be?
There is no single “best” CNC tolerance for every feature. The required value depends on function, feature size, material, geometry, process, quantity, and inspection method.
A useful design approach is to divide dimensions into three groups.
Critical dimensions directly control function. Examples include bearing seats, mating diameters, locating pins, sealing surfaces, press fits, and critical hole patterns. These features may need tighter dimensional or geometric controls.
Important but non-critical dimensions affect assembly or appearance but have more room for variation. These can often use moderate tolerances.
Reference or non-functional dimensions do not directly affect fit or performance. These should not automatically receive the same tight tolerance as a bearing bore.
This avoids a common drawing problem: applying a very tight general tolerance to every dimension simply because a template uses it. For example, a bracket may only need a 50 mm outside dimension for clearance, while a bearing bore on the same part may need much tighter control.
General Tolerance vs. Individual Tolerances
Designers often ask whether every dimension needs its own tolerance. Usually, no.
A drawing can define a general tolerance in the title block or notes for dimensions that do not have individual requirements. Critical dimensions can then receive specific tolerances.
This keeps the drawing readable and makes design intent clearer. A general tolerance should not be treated as a shortcut for avoiding engineering decisions. The designer still needs to decide which features are functional and whether the default tolerance is appropriate.
In many engineering environments, standards such as ISO 2768 may be used for general tolerancing. The exact class should be stated rather than assuming that one number applies to every feature. General tolerances also do not replace geometric controls when flatness, perpendicularity, position, or orientation is important.
The practical rule is simple: use a general tolerance for ordinary dimensions and override it where function requires more control.
Why Do Tight Tolerances Cost More?
A tighter tolerance reduces the acceptable manufacturing window and usually requires more process control.
- More precise machining strategies: roughing may need controlled finishing passes.
- Additional setups or fixtures: the feature may need stronger support to reduce movement and vibration.
- Tool wear control: offsets and tool changes may need closer monitoring.
- Longer cycle time: conservative cutting parameters may be needed for dimensional stability.
- Specialized measurement: a critical feature may require a bore gauge, optical system, or CMM instead of a simple caliper.
- Higher scrap risk: a small process shift can push a feature outside a narrow tolerance window.
- Secondary processes: very tight bores or surfaces may require boring, reaming, grinding, lapping, or honing.
Tighter tolerances can significantly increase manufacturing cost depending on the tolerance and geometry. Forum discussions show the same practical pattern: machinists frequently recommend loosening non-functional tolerances when the design allows it.
The point is not that tight tolerances are bad. They are valuable when they protect function. The problem is applying them where they provide no functional benefit.
Can a CNC Machine Hold the Same Tolerance on Every Feature?
No. Tolerance capability depends on the feature, material, geometry, setup, process, and inspection method. A short external diameter may be easier to control than a deep bore. Thin plates can move, long shafts can deflect, and small deep holes can be affected by tool runout. Plastics, titanium, and nickel alloys can also behave very differently during cutting. Therefore, a machine’s advertised accuracy should not be treated as a universal tolerance for every feature.
How Should Designers Tolerance Holes?
Hole tolerancing is one of the most common areas of confusion in CNC drawings.
A hole has at least two separate questions: how large should it be, and where should it be?
For example, a drawing may specify a Ø10.00 mm hole with a size tolerance. That controls diameter, but it does not automatically define the exact location of the hole relative to other features.
If the hole is used for a locating pin, bearing, dowel, or precision assembly, its position may be just as important as its diameter. Designers should therefore consider both size and location.
For a hole pattern, ask what actually matters:
- Does each hole need a precise location from the part edges?
- Does the distance between holes matter more than the location of the entire pattern?
- Does the pattern need to align with a mating part?
- Can the holes have additional clearance?
- Does hole size variation within the same pattern matter?
A hole pattern can have acceptable diameters and still fail to assemble because the holes are in the wrong locations.
GD&T and Positional Tolerance
GD&T can be useful when simple ± dimensions do not clearly communicate functional intent.
A positional tolerance defines a controlled zone around the theoretically exact location of a feature. Datum references establish the reference framework used to evaluate the feature. This is especially useful for hole patterns, locating features, perpendicular features, and assemblies where several geometric relationships matter at the same time.
However, GD&T should not be added simply to make a drawing look more professional. Forum discussions show designers struggling with unnecessary datums, missing basic dimensions, overly tight position tolerances, and geometric controls that do not match the real function.
Start with the function:
- What feature locates the part?
- What surface establishes orientation?
- Which hole or surface must align with the mating component?
- Which variation can the assembly tolerate?
Then select the geometric control that communicates that requirement. If a simple dimensional tolerance fully communicates the requirement, GD&T may not be necessary. If several features must maintain a controlled relationship, GD&T can make the drawing much clearer.
What Is Tolerance Stack-Up?
Tolerance stack-up occurs when variations from multiple dimensions combine and affect a final functional dimension.
Consider two holes located using separate dimensions from a common edge. If each location has ±0.05 mm tolerance, the relationship between the holes may have a different worst-case variation than either individual dimension.
The important lesson is to evaluate the complete functional chain, not each tolerance independently.
A simple stack-up can be estimated by adding worst-case dimensional variations when that method matches the design requirement. Statistical methods may be appropriate in other production situations, but the method should be chosen deliberately.
A drawing can look fully dimensioned while still allowing too much accumulated variation for mating parts to assemble.
Before releasing a drawing, ask:
- What is the maximum possible gap?
- What is the minimum clearance?
- What happens if dimensions shift toward the worst functional direction?
- Which dimensions are part of the same functional chain?
- Can one datum or locating feature control the relationship more directly?
Tolerance stack-up is especially important when multiple machined parts are assembled together.
Bilateral, Unilateral, and Limit Tolerances
Bilateral tolerance allows variation in both directions. For example, 25 ±0.05 mm gives a range of 24.95 to 25.05 mm.
Unilateral tolerance allows variation primarily in one direction. For example, 25 +0.10/-0.00 mm gives a range of 25.00 to 25.10 mm.
Limit dimensions state the upper and lower acceptable sizes directly, such as 25.10 / 25.00 mm.
Choose the form that matches the function. A unilateral limit can be clearer when a shaft must not exceed a maximum size; bilateral tolerance is simpler when variation is acceptable in both directions.
For mating parts, designers should think about the combined tolerance of both components rather than choosing shaft and hole limits independently.
Clearance, Transition, and Press Fits
A common mistake is to specify a tight dimensional tolerance without defining the intended fit.
Suppose a shaft and bore must rotate freely. The design needs a clearance range, not simply a small tolerance on both diameters.
If the shaft must be pressed into a hub, the design needs a controlled interference range. Too little interference may allow movement; too much may damage components during assembly.
Standard fit systems such as ISO 286 can help define cylindrical mating relationships. Designers should select the fit based on material, temperature, load, assembly method, lubrication, and service conditions rather than simply choosing the smallest possible clearance.
Temperature and Material Effects
A dimension measured at room temperature is not necessarily the same dimension under operating conditions.
This matters for precision assemblies because aluminum, steel, stainless steel, plastics, and other materials expand differently. Material also affects machining stability: thin plastic walls can move, aluminum can expand with heat, and hard stainless alloys can create higher cutting forces.
For critical assemblies, designers should consider both machining temperature and operating temperature.
Surface Finish and Tolerance Are Different Requirements
Surface roughness and dimensional tolerance control different aspects of a part. A surface can have the correct size but an unsuitable finish, while a smooth surface does not guarantee correct geometry.
Surface treatments can also affect dimensions. Anodizing, plating, powder coating, polishing, or passivation may add or remove material, so critical threads, bores, and mating diameters should state whether the tolerance applies before or after finishing.
Design for Inspection, Not Just Machining
One important gap in basic CNC tolerance guides is inspection.
A difficult-to-machine feature may also be difficult to measure. Tight geometric requirements can therefore require additional equipment, fixtures, or measurement time.
An ordinary external diameter may be checked with a micrometer. A complex hole pattern may require a CMM. A thread may require an appropriate thread gauge. A surface roughness requirement needs a suitable roughness measurement method.
The designer should ask whether the requirement can be verified reliably. A tolerance that cannot be measured consistently is not a strong manufacturing requirement.
A Practical CNC Tolerance Workflow for Designers
Start with function, not machine capability.
- Identify mating and functional features.
- Define the minimum clearance, maximum interference, or required alignment.
- Separate critical dimensions from non-critical dimensions.
- Apply reasonable general tolerances to ordinary features.
- Add individual tolerances only where function requires them.
- Use GD&T when geometric relationships are important.
- Perform a tolerance stack-up for critical assemblies.
- Check whether the chosen tolerance can be manufactured with the intended process.
- Check whether the feature can be inspected with a suitable method.
- Review the effect of heat treatment and surface finishing.
- Confirm material, size, quantity, and drawing revision.
- Ask the machining supplier for DFM feedback when a tolerance is unusually tight or a feature is difficult to access.
This helps avoid drawings that are either too loose to protect function or unnecessarily tight and expensive.
Common CNC Tolerance Mistakes to Avoid
Do not use a tight tolerance simply because the CAD model shows many decimal places. CAD precision and manufacturing tolerance are not the same thing.
Do not apply the same tight tolerance to every dimension. Functional features usually deserve more attention than clearance or non-functional dimensions.
Do not control hole diameter while ignoring hole position when assembly depends on the pattern.
Do not assume that a basic dimension with a GD&T callout has the same meaning as an ordinary ± dimension.
Do not build a long chain of dimensions without checking tolerance accumulation.
Do not specify a tight tolerance without considering how the feature will be inspected.
Do not ignore material deformation, thermal effects, or surface treatment.
Do not assume that a CNC machine’s advertised accuracy means every feature on every material can automatically be produced to that value.
How Tuofa CNC Germany Reviews CNC Tolerances
Tuofa CNC Germany approaches tolerance requirements as part of the complete manufacturing plan rather than as isolated numbers on a drawing. During technical review, critical dimensions can be considered together with part geometry, material, tool access, setup, machining process, finishing requirements, and inspection needs.
For prototype and low-volume CNC parts, this early review can help identify tolerances that may be unnecessarily restrictive as well as features that need stronger control for assembly.
When a drawing includes tight bores, precision shafts, hole patterns, threads, thin walls, or complex geometric relationships, the manufacturing process and inspection method should be considered before production. If a requirement appears unclear or difficult to manufacture as specified, the engineering team can raise the issue before machining begins rather than after a part has failed inspection.
Conclusion
Good CNC tolerancing is not about making every dimension as precise as possible. It is about controlling variation where it affects function.
The best drawing separates critical features from ordinary ones, defines mating conditions, controls hole position when needed, prevents harmful stack-up, and gives the manufacturer a practical way to inspect the result.
The core manufacturing guidance is clear about the cost, process, inspection, and material effects of tighter tolerances. Machinist and engineer discussions add practical issues such as general tolerances, GD&T, fits, hole patterns, stack-up, measurement, and design intent.
By connecting those factors before production, designers can create CNC parts that are easier to manufacture, easier to inspect, and more likely to work as intended. Tuofa CNC Germany can review CAD models and technical drawings for custom CNC machining projects and help align tolerance requirements with the actual function of the part.