Caster Load Ratings: 5 Checks Before You Derate for Speed, Floor and Shock

Is the catalogue number static or dynamic?

A static rating describes a caster standing still under prolonged pressure. A dynamic rating describes a caster rolling under load.

Movement adds stresses from temperature, floor deviation, obstacles, speed, load distribution and direction changes, so the dynamic value is always the lower of the two and is the one that governs a moving application^8,12. Casters are in practice rated and designed on dynamic load capacity^8,12.

One source notes that buyers must still ask whether a published value is static or dynamic, which implies catalogue pages may carry either¹. The distinction matters because a static figure can look more impressive and tempt a buyer into sizing a moving application on it¹.

A server rack left in place for years is a static case. The same rack on a cart pushed daily is a dynamic case.

The catalogue may publish only one number, so the buyer must ask which one it is. If your application moves, treat the published number as dynamic until the supplier confirms otherwise.

Rule out ratings with no test basis

A rating is only meaningful if the test that produced it is known. A dynamic endurance test runs the caster on a track with obstacles under full load, exercising the wheel, bearings and housing in a way a static compression test never does⁸.

A caster that passes static pressure with no irreversible deformation of the housing and no flat spots on the tread may still fail a dynamic obstacle test under simulated fully loaded conditions¹⁶. The two tests answer different questions.

BS EN 12532 covers industrial wheels and castors up to 1.1 m/s, roughly 4 km/h, and its scope includes dynamic testing, fatigue testing, performance testing and load capacity¹. One source reports that published ratings assume lab conditions: smooth floor, 70°F, 3 mph, no impact¹⁰.

A supplier claim that all casters are tested to industry benchmarks still does not tell the buyer whether the published value is static or dynamic¹⁴. Ask for the test basis in writing before you accept the number.

Calculate on one less wheel

On an uneven surface the load is not shared equally and one caster may lose contact, so the remaining casters must carry the whole load^8,12. Both static and dynamic load are calculated on one less wheel than installed^8,12.

This is the mechanism that prevents a single wheel from being overloaded the moment the floor deviates. A four-caster cart carrying 800 lb must be specified as if three casters carry the load, giving a required per-caster capacity of about 267 lb before any de-rating, not 200 lb.

That is a 33% higher starting point, computed from the cited one-less-wheel rule. The same logic applies to a 180 kg medical cart: one source reports a required single-wheel dynamic capacity of at least 60 kg using the one-less-wheel method¹⁶.

If your drawing shows four casters, divide by three, not four.

Apply the speed de-rate

Higher travel speed increases stress on the wheel, bearings and tread, builds heat inside polyurethane and other elastomers, and multiplies bearing operating cycles⁴. The permissible dynamic load falls as speed rises.

One source reports that some industrial wheels carry a higher dynamic load at 4 km/h than at 6 km/h⁴. Another source reports a speed de-rate of roughly 25% per mph above 3 mph for standard polyurethane¹⁰ (supplier-reported figure).

BS EN 12532 covers applications up to 1.1 m/s, about 4 km/h, so above that speed the standard's rating basis does not cover the application at all¹.

A standard polyurethane wheel rated at 3 mph loses roughly 25% of its dynamic capacity for each additional mph, so a 5 mph application cannot use the catalogue figure unchanged (supplier-reported figure). That is a 50% loss at 5 mph, computed from the cited per-mph de-rate.

Specify actual travel speed rather than assuming walking pace⁴. If your cart or AGV runs above 4 km/h, the catalogue number is not your rating.

Stack the floor, swivel, side-thrust and duty de-rates

Each real-world condition removes capacity for a different reason. Uneven floors cause shock loading and unequal distribution¹⁰.

Swivelling under load loads the raceway and wheel¹⁰. Side thrust concentrates stress on the kingpin and raceway¹⁰.

Continuous duty accumulates cycles¹⁰. Because the de-rates stack, the order of magnitude matters more than the exact figure.

One source reports a 20-30% floor-condition de-rate for cracked or uneven floors, a 15-20% swivel de-rate, a 25-40% side-thrust de-rate, and a 30-40% duty-cycle de-rate for 24/7 AGV operation versus an 8 hr/day reference¹⁰.

The same source reports that a 1,000 lb caster run at 5 mph, 180°F, 24/7 on a cracked floor becomes a 300-400 lb caster once the de-rates are stacked¹⁰ (supplier-reported). That is a 60-70% loss from the catalogue figure, computed from the cited stacked example.

For a 200-piece aluminium bracket cart, the same stacking applies. If your route has cracked concrete and sharp turns, the catalogue number is not your capacity.

Pick the wheel and bearing on the trade-off

Harder wheels carry more load with lower rolling resistance but transmit shock and noise and can scratch soft floors. Softer wheels deform, absorb energy and protect floors but roll harder and generate more push force^7,13.

Bearing type changes rolling resistance far less than wheel material or diameter, so it is a secondary lever¹⁵. A 70A soft rubber wheel has a rolling resistance coefficient of 0.06-0.08 against 0.03 for 95A polyurethane because it deforms and absorbs energy¹⁵ (supplier-reported).

Precision ball bearings cut rolling resistance 20-30% versus roller bearings, and tapered roller bearings add 10-15% versus ball bearings in exchange for heavier load capacity¹⁵ (supplier-reported). If your floor is tile or hardwood, the wheel material choice matters more than the bearing.

Set the safety factor before the RFQ

The safety factor is the margin between the calculated load and the rated capacity. It must rise with the severity of the application because manual indoor movement on smooth floors produces small, slow loads while motorised outdoor movement on poor surfaces produces impact and side thrust that the rating never saw^3,11.

One source reports a range of about 1.0-1.5 for manual indoor smooth floors up to 3.0 for motorised outdoor poor surfaces³. Another source breaks the same territory into four bands: 1.0-1.5 indoor manual, 1.5-2.2 outdoor manual, 1.4-2.0 indoor powered, and 2.0-3.0 outdoor powered, with 1.33 commonly used as a general allowance¹¹.

A manual indoor cart on a smooth floor sits at the bottom of the range around 1.0-1.5, while a motorised outdoor unit on poor surfaces needs up to 3.0. That is roughly a 2x difference in required capacity between the two extremes, computed from the cited bands.

If your application is powered and outdoors, do not accept a supplier's default 1.33.

Ask the supplier these questions

A catalogue number without its test conditions cannot be derated correctly, and the supplier is the only party who can confirm the basis of the rating.

Skipping these questions leaves the buyer applying de-rates to a figure that may already be a static value or may apply only to the wheel rather than the assembled castor¹. Different manufacturers use slightly different calculation methods, and approaches should not be mixed automatically³.

Write the answers into the RFQ so the quote is against the right rating basis.

• ✓Which standard supports the rating?
• ✓Is the value static or dynamic?
• ✓At what speed does it apply?
• ✓Which wheel and bearing configuration was tested?
• ✓Does temperature reduce capacity?
• ✓Has the complete castor or only the wheel received the rating?

Get commercial terms separately

The technical decision fixes the required capacity and the wheel and bearing configuration. Only then can the supplier quote a price, MOQ, lead time and tooling cost for that configuration.

Asking for commercial terms before the derating is settled invites a quote against the wrong rating basis³. A buyer who has calculated a required per-caster capacity of 400 lb after stacking de-rates can ask for a quote on a specific wheel and bearing combination (supplier-reported).

A buyer who asks for a price on the catalogue 1,000 lb caster will receive a quote that does not match the application. No source in the evidence provides price, MOQ, lead time or tooling cost data, so these must be obtained from the supplier and treated as a separate commercial negotiation.

The calculation method must match the manufacturer whose rating you intend to specify, because different manufacturers use slightly different methods³. If your RFQ goes out before the derating is done, you are comparing quotes on different bases.

Where the sources disagree

Two sources give different safety factor bands and two sources disagree on whether casters are always rated on dynamic load or whether published values may be either. The buyer who sees the disagreement can ask the supplier to state the basis instead of assuming one source is correct.

One source gives a single range of about 1.0-1.5 for manual indoor smooth floors up to 3.0 for motorised outdoor poor surfaces, while another breaks the same territory into four bands of 1.0-1.5, 1.5-2.2, 1.4-2.0 and 2.0-3.0, with 1.33 commonly used as a general allowance^3,11.

The third dispute is about the basis of the dynamic load rating itself: one source says it is established by validation testing with defined factors, another says BS EN 12532 scope includes dynamic, fatigue and performance testing and load capacity, and a third says published ratings assume lab conditions^1,8,10.

Ask the supplier which basis applies to the caster you are buying.

What the sources do not establish

• No source gives numeric pass/fail criteria for dynamic endurance tests (cycles, wear limits, deformation limits).
• No source provides a manufacturer-published speed derating curve or table.
• No source provides a manufacturer-published floor-condition derating factor table by floor type.
• No source defines a shock load capacity or shock safety factor.
• No source gives consolidated load-speed-temperature limits per wheel material and bearing combination.
• No source cites ISO 22883 or ANSI/ICWM, and no source gives EN 12532 test clauses with numeric criteria.
• No source provides price, MOQ, lead time or tooling cost data.
• No source states whether the de-rate percentages in S10 are validated or vendor-specific.