Bearing Clearance Explained: C2, CN, C3, C4 — Which One Your Application Needs

Bearing clearance is the total distance the inner ring can move radially relative to the outer ring before load is applied. It is graded into groups: C2 (tighter than normal), CN (normal), C3 (larger — the default for electric motors), C4 and C5 (high temperature or heavy interference fits). The number stamped on the box is not the number that matters. What decides bearing life is operating clearance — what is left after the fits and the heat have taken their share.

What bearing clearance actually is

Radial internal clearance is the total radial distance one ring can travel relative to the other, measured with no load applied. It is not a manufacturing error and it is not a precision grade — it is a designed-in gap, specified and measured to ISO 5753 (China: GB/T 4604).

The point most buying guides skip: the clearance a bearing has in the box is not the clearance it has when it is running. Three things eat into it, in this order:

1. The shaft fit. Press the inner ring onto an interference-fit shaft and the ring expands. That expansion comes straight out of the clearance.
2. The housing fit. An interference fit in the housing squeezes the outer ring inward. Same effect.
3. The temperature difference. In most machines the inner ring runs hotter than the outer ring, because heat arrives through the shaft and leaves through the housing. The inner ring grows more than the outer one, closing the gap further.

What is left is the operating clearance, and that is the figure the bearing actually lives with:

Operating clearance
  = Initial (as-shipped) clearance
  − inner ring expansion from the shaft fit
  − outer ring compression from the housing fit
  ± change from the inner/outer temperature difference

Every clearance decision in this article is a decision about the number at the bottom of that sum, not the number on the box.

The clearance groups — C2 to C5

Reference values for a 6204 (20 mm bore), per GB/T 4604 / ISO 5753:

Two things to read off this table. First, the groups are ranges, not values — two bearings both correctly marked C3 can differ by 10 µm. Second, the whole CN band is 7 µm wide, which is the same order of magnitude as what an interference fit removes. That is why the fit matters as much as the group.

Why electric motors default to C3

An electric motor is the textbook case for every term in the operating-clearance equation at once:

• The inner ring is usually mounted with an interference fit on the rotor shaft.
• The rotor is where the heat is generated. The inner ring runs hotter than the outer ring, often by 5–10 °C in a steady-state motor.
• The housing is aluminium or cast iron and dumps heat to ambient.

Typical orders of magnitude for a small-to-medium motor bearing — treat these as illustrative reference values, not a specification:

Add those and you are looking at something like 7–11 µm gone. Start from CN (4–11 µm) and you can arrive at zero or negative operating clearance — a preloaded bearing that was never designed to be preloaded. The symptoms are familiar to anyone who has warrantied a motor: it runs hot, the grease degrades early, the cage carries loads it was not sized for, and the bearing fails long before its rated life.

Start from C3 (11–21 µm) instead and the same subtraction leaves a small positive clearance. That is the entire reason C3 is the motor default. It is not a quality upgrade. It is compensation.

What actually happens to life as clearance changes

This is where most clearance guides stop, and it is where the interesting part starts.

A 2024 study in Transmission Technology (Lü Haiting et al., Dalian Institute of Science and Technology, with Wafangdian Bearing's wind power division) modelled a 6308 deep groove ball bearing — 40 × 90 × 23 mm, ten 12 mm balls on a 65 mm pitch circle — across eight clearance values, using ISO/TS 16281, which accounts for operating clearance and internal load distribution rather than treating the bearing as a catalogue number.

⚠️ These are simulation results, not bench test results. The study did not run life tests. Use them for direction, not for acceptance criteria.

First finding — the number of balls carrying load collapses as clearance opens up. For the radially loaded bearing in the study:

At 0.3 mm — the kind of gap you get from a sloppy housing bore — one ball is carrying the entire radial load. That is the mechanism behind "too much clearance causes vibration and early failure", stated as a number instead of a warning.

Second finding — and this is the counterintuitive one — zero clearance gave the worst life of any case tested. Ten balls sharing the load, and the shortest life in the set.

More balls sharing a load, but a worse outcome. Both of these are moving at once:

• Contact stress. At zero clearance the balls are pinched between both raceways. Load sharing improves, but every contact is stressed even when it is not in the loaded zone.
• The ISO life modification factor a_ISO. Across the cases modelled it ranged from about 7.8 up to the 50 cap. Lubrication condition and internal load distribution feed into this factor, and it moves far more than the ball count does.

The net effect is that life versus clearance is not a straight line and not a simple "less is better" — there is an optimum, and it is not at zero.

Third finding — the optimum depends on how the bearing is loaded. The study modelled two bearings on the same shaft with different load splits:

Two consequences worth carrying into a real design:

• Two bearings on one shaft do not share an optimum. If you are specifying a matched set, optimise for the one with the shorter life, not for the average.
• Axial load flattens the curve. The axially loaded bearing kept all ten balls engaged across the entire range from 0 to 0.3 mm. If your application carries a meaningful thrust load or a designed preload, clearance group is a much less critical decision than it is for a purely radial one. That is a useful thing to be able to tell a customer who is agonising over C3 versus C4 on a thrust-loaded position.

The study's own summary recommendation for that 6308 was an operating radial clearance of 0.027 – 0.0374 mm.

⚠️ Do not copy that range onto a drawing. It is an operating clearance for one bearing size under one set of loads. To turn it into a purchasing specification you would have to add back the fit losses and the thermal losses for your own machine.

C3 vs C4 — when to step up

C4 exists for two situations, and neither of them is "we want a better bearing":

• High operating temperature or a large inner/outer temperature differential — oven fans, hot-air blowers, anything where the inner ring runs far hotter than the housing.
• Heavy interference fits — k5/m5 and above, where the fit alone eats an unusual amount of the clearance.

Outside those cases, C4 is not an upgrade, it is a mistake with a bigger number. Go back to the ball-count table: opening the clearance shrinks the loaded zone, concentrates load onto fewer balls, and raises vibration. A C4 bearing in a normal motor position is a bearing running with fewer balls under load than the designer intended.

The honest one-line version: C3 compensates for what the fit and the heat take away. C4 compensates for more of it. Neither buys you precision.

Clearance and precision are two different things

These get confused constantly in RFQs, so it is worth being blunt about it.

• Clearance (C2/CN/C3/C4/C5) governs the internal gap — how much the rings can move relative to each other.
• Precision grade (P0/P6/P5/P4/P2, or ABEC) governs dimensional tolerance and running accuracy — bore and outside diameter tolerance, radial and axial runout.

They are independent specifications. A P0 bearing can be C3. A P5 bearing can be CN. The most common combination in general motor production is P0 with C3 — ordinary tolerance, enlarged clearance.

Asking for "a more precise bearing" when the real problem is that the shaft fit closed up the clearance will not fix the machine, and it will raise the price for nothing.

How clearance is measured and verified

Clearance is measured to ISO 5753, before mounting, under a specified measuring load. At NLHB we check radial internal clearance on the assembled bearing before shipment, and we re-check after any operation that could disturb it.

Two questions worth putting to any bearing supplier, ours included:

1. Do you measure clearance before assembly, after assembly, or both? The answer tells you whether closing operations — shield pressing, seal fitting — are being verified or assumed.
2. Can the measured values travel with the shipment? For a specified clearance requirement, batch measurement records should be available rather than a group marking alone.

(We supply to the clearance specification agreed on the order; measured values are reported per batch.)

Common clearance mistakes buyers make

1. Drawing says C3, order gets filled with CN. CN is usually unmarked, so a bearing with no clearance suffix looks like "standard". If the drawing calls out a group, the group belongs in the order line, not just the drawing.
2. C3 specified for a high-temperature application. C3 compensates for a normal temperature differential. An oven fan is not normal.
3. C4 requested because it sounds better. It is not a grade. See above.
4. Clearance confused with precision. Two different specifications, two different problems.
5. Ordering a clearance group without stating the fits. A k5 shaft and a g6 shaft take very different amounts out of the same bearing. The clearance group cannot be chosen without knowing the fit.
6. Optimising a matched pair on the average. Two bearings on one shaft, loaded differently, have different optima.

Quick selection table

Working out a clearance specification is not a catalogue lookup. Send us the bearing size or drawing, the shaft and housing tolerances, the expected operating temperatures, the speed and the load direction, and we will work back from the operating clearance you need to the group you should be ordering.

FAQ

What is bearing clearance C3?
C3 is a radial internal clearance group larger than normal (CN), defined in ISO 5753 / GB/T 4604. For a 20 mm bore bearing such as a 6204 it is roughly 11–21 µm against 4–11 µm for CN. It is the usual default for electric motors, because the shaft fit and the running temperature difference both reduce clearance once the bearing is installed.

Can I use C3 instead of CN?
For most rotating equipment with an interference fit on the shaft, yes — that is exactly what C3 is for. For a precision or low-noise position with a light fit, going to C3 may leave too much clearance and increase runout and noise. Check the fit before substituting.

Is C4 better than C3?
No. C4 is not a higher grade, it is a larger gap. It is correct for high temperature or heavy interference fits and harmful elsewhere — a larger clearance puts fewer balls under load, which raises vibration and shortens life.

How is bearing clearance measured?
To ISO 5753 (GB/T 4604), before mounting, under a defined measuring load. Values after mounting differ, because the fits and the operating temperature both reduce the clearance.

What clearance should an electric motor bearing have?
C3 in the general case. The reason is thermal: the inner ring runs hotter than the outer ring, and the interference fit on the rotor shaft expands the inner ring. Both effects reduce clearance once the motor is running.

Does clearance affect noise?
Yes, in both directions. Too much clearance concentrates the load onto fewer balls and produces rattle and vibration; too little produces drag, heat and a squealing or whining character. Noise grade and clearance group are separate specifications, however.

Related resources

• Bearing Suffixes Explained: 2RS, ZZ, NR, ET and More
• 6204 Bearing: Dimensions, Load Ratings and Interchange Guide
• Bearing Precision Grades P0/P6/P5/P4 — What Tolerance Class Actually Buys You
• Bearing Noise Levels Z1-Z4: dB Limit Tables and What to Specify
• Double Row Angular Contact Bearings: 2RS vs ZZ
• Sealed Bearing Speed Limits: Four Published Figures and Why They Disagree