Cylindrical Roller vs Deep Groove Ball Bearing: When the Load Case Decides

Is the load heavy and purely radial?

Start with the load case, not the price. The selection rule the sources give is blunt: light load and high speed point to a ball bearing, deep groove for general use; heavy radial load at moderate speed points to a cylindrical roller bearing for straight radial support without significant thrust^7,11.

The reason is contact geometry. A ball touches its raceway at a point, so the load sits on a very small area.

A roller touches along a line, and the loaded strip spreads that same force over far more material^4,15. That is why a roller bearing can carry roughly 2-3 times the radial load of a ball bearing of the same bore^4,15.

Two sources state that ratio, so treat it as the working figure rather than a marketing line. The practical consequence is envelope, not just capacity.

If a CRB carries 2-3x the radial load at equal bore, then for a given load it can be the smaller bearing, or the only feasible one where a DGBB of that bore would be undersized^7,11,15 (supplier-reported figure). Picture a conveyor or mill roll: a heavy radial load, no thrust worth mentioning.

The sources' rule sends that duty to a cylindrical roller bearing, and the same bore in a deep groove ball bearing is either the smaller option or simply not viable^7,11. Do not read this as a general verdict.

One source explicitly lists "assuming roller bearings are always better" as a common selection mistake⁴. The load case decides the family; everything else is a consequence you have to accept.

If your radial load is genuinely heavy and your thrust is negligible, this is the point where the CRB earns its place on the drawing.

Does any thrust load rule the CRB out?

A plain cylindrical roller bearing has no thrust path at all. It handles radial loads only¹⁵.

That single fact disqualifies it the moment a real axial component appears, and it is the check that most often kills the CRB after the load case has already pointed to it. A deep groove ball bearing is the opposite compromise.

Its grooves let it take axial load in both directions, but only a fraction of its radial capability: one source puts that at roughly 25-35% of the static radial rating C0^1,17. A second source quotes up to 70% of the radial load rating instead¹⁵, and the two figures do not agree (supplier-reported figure).

That disagreement matters because it decides whether a mixed load can stay on a DGBB at all. If your thrust sits between the two readings, you cannot settle it from this evidence; you settle it from the manufacturer's data sheet for the exact part.

Work a pump shaft with a heavy radial load plus residual axial thrust. The plain CRB cannot take the thrust, full stop.

The DGBB can take it only at roughly 25-35% of C0 on the conservative reading^1,17 (supplier-reported). So the buyer's job is to put the actual thrust figure next to that fraction before choosing either family.

If the thrust exceeds it, neither family alone is the answer and a second bearing or a different type enters the picture. Pull the thrust number off your load calculation and compare it to the DGBB's axial rating before you commit to a CRB.

Will speed or noise veto the roller bearing?

Speed is the second filter, and it is often the one that ends the discussion. Line contact raises friction, and friction raises heat at speed.

That is the mechanism one source gives for the roller family's lower ceiling: deep groove ball bearings can run to about 4 million DN, while roller bearings typically max out at 1-2 million DN¹⁵. DN is bore in millimetres multiplied by RPM, so the gap is not marginal.

A ball bearing can typically run at 2-3 times the speed of a roller bearing of the same size¹². Take an electric motor or a fan running near the top of the ball bearing's range.

At that duty the CRB is excluded on speed alone, whatever its load advantage, because it cannot reach the required DN¹⁵. The sources are consistent that ball bearings are the high-speed, low-friction family and that deep groove types are the standard choice for motors, fans, pumps, appliances and precision instruments^1,10,12.

Noise follows the same geometry. The sources put deep groove ball bearings at lower noise and roller bearings at higher noise¹², and describe the ball type as quiet and smooth because of how the balls meet the raceways¹.

One source notes that most real applications are settled by load and speed alone, before misalignment or cost are even considered⁷. That is the honest order of operations: if the duty is fast or quiet, the CRB never gets to the cost conversation.

Check your required RPM against the candidate's limiting speed and DN rating before you bank the load advantage.

Can the machine hold alignment?

Alignment is the constraint that surprises buyers, because it is a property of the machine, not the bearing. A deep groove ball bearing tolerates only about 0.5-1 degree of misalignment^3,17 (supplier-reported).

Beyond that limit the load stops being shared evenly around the raceway, and the uneven distribution drives premature failure¹⁷. Two sources give that 0.5-1 degree figure (supplier-reported); a third quotes 0.003-0.005 degrees instead¹², a difference of orders of magnitude that you must resolve against the manufacturer's data sheet rather than this page (supplier-reported).

When alignment cannot be held, the remedy is not a standard CRB. It is a self-aligning design, which compensates up to about 3 degrees open and 1.5 degrees sealed, or tighter machining and assembly³.

The trade is real: self-aligning bearings have extremely limited axial capacity, so the fix for misalignment can reopen the thrust question you already closed³. Consider a fabricated housing that cannot hold alignment to within about 0.5-1 degree (supplier-reported).

On that machine a DGBB fails early, and the answer is a self-aligning bearing or a better housing, not a cylindrical roller bearing^3,17. Note what the evidence does not give you: no source states a misalignment tolerance for cylindrical roller bearings, and none says whether a self-aligning CRB variant exists or is needed.

That gap belongs in your RFQ, not in your assumption. Measure or estimate the alignment your shaft and housing can actually hold, and put that number against the 0.5-1 degree limit before you specify (supplier-reported).

Can the housing take the lubrication change?

The roller family asks more of its lubrication and sealing than the ball family does. One source contrasts them directly: deep groove ball bearings need less grease and less maintenance, while roller bearings need more grease and commonly run in an oil bath¹².

The same source describes the ball type as simple to mount and low maintenance, generally not needing constant lubrication or frequent adjustment once installed¹. That difference is not a footnote on a data sheet; it changes the housing.

A design built around a greased-for-life DGBB has seals, a grease cavity and a maintenance interval sized for a small grease charge. Move to a CRB and the lubrication provision changes, which means the sealing arrangement and the maintenance plan change with it¹².

If your housing is already tooled, that is a redesign cost that arrives before any bearing price is quoted. Take a sealed-for-life appliance duty against a roller bearing running in an oil bath.

The DGBB needs less grease and less attention; the CRB needs more grease and different sealing, and the housing has to accommodate that^1,12. The sources do not give relubrication intervals or filtration requirements for either family, so you cannot size the maintenance plan from this evidence.

Ask the supplier for both. Confirm your housing can accept the CRB's lubrication and sealing provisions before you treat the family change as a like-for-like swap.

Is the roller premium acceptable here?

The cost direction is settled even though no price is. Deep groove ball bearings are the most cost-effective rolling bearing type, mass-produced in thousands of sizes and widely available^1,3.

Roller bearings cost more at the same bore: one source puts the premium at typically 20-60%¹². That is a single-source figure, so treat it as a directional band to verify, not a quotation.

The mechanism behind the premium is structural rather than negotiable. The DGBB's cost leadership comes from mass production and standardisation across a huge size range^1,3.

A roller bearing at the same bore sits outside that volume, so the premium reflects production scale and content rather than a supplier's pricing mood. You will not negotiate it away at the same bore and grade.

Run the arithmetic on a general-duty application at a standard bore. If the DGBB is the baseline, a CRB at the same bore lands roughly 1.2 to 1.6 times that baseline, computed from the cited 20-60% premium¹² (supplier-reported).

That is the number to carry into your cost comparison, and it is a derived range, not a quoted price. The sources give no unit prices, MOQs, tooling charges or precision-grade premiums for either family, so the real figure has to come from quotes.

The premium only makes sense if the CRB is buying you something the DGBB cannot deliver at that bore. If it is not, the cheaper family is the correct answer.

What must you confirm before the RFQ?

These checks are what separate a family decision from a custom tooling order. A non-catalogue size adds lead time and setup charges, and a selection made on peak load instead of equivalent dynamic load picks the wrong bearing entirely.

The sources are explicit that selection should run on equivalent dynamic load and an L10 life calculation against the candidate's Cr, not on peak load or size alone^2,7,16. The life relationship is cubic, so doubling the equivalent dynamic load leaves only about one-eighth of the L10 life¹⁴.

That is why the load figure you feed the calculation matters more than the family label on the box.

• ✓Confirm the load direction and magnitude, and separate the radial and thrust components.
• ✓Check the required RPM against the candidate's limiting speed and DN rating.
• ✓Run the L10 life calculation using the candidate's Cr and your equivalent dynamic load, not peak load.
• ✓Verify misalignment tolerance, lubrication and mounting against your shaft, housing and environment.
• ✓Confirm the chosen size exists as a standard catalogue item before considering a custom design.
• ✓Compare Cr, limiting speed and friction data across the shortlisted part numbers.

Where do the sources disagree?

The published limits for these two families are quoted differently across sources, and the spread is wide enough to change a decision. A buyer who takes one figure as fact can rule a bearing in or out on a number another source contradicts.

The two rows below are the disputes that matter most for this choice: the DGBB's misalignment tolerance and its axial capacity. Both must be resolved against the manufacturer's data sheet for the exact part number, not against a general article (supplier-reported).

What the sources do not establish

• No Cr values, reference speeds or limiting speeds for any specific CRB or DGBB bore or series, so the load ratio at which the CRB wins cannot be calculated
• No grease-versus-oil speed ratings for either family
• No published vibration or sound level data for CRB versus DGBB
• No misalignment tolerance stated for cylindrical roller bearings, and no statement on whether a self-aligning CRB variant exists
• No unit prices, MOQs, tooling or setup charges, or precision-grade premiums for either family
• No lead times, stock levels or counts of qualified suppliers for CRB versus DGBB
• No relubrication intervals or filtration requirements for CRB versus DGBB
• No dimensional data confirming or denying drop-in interchangeability between a CRB and a DGBB of the same bore