Heavy-Duty Caster Safety Factor: How to Build a Per-Caster Capacity Requirement from Load Distribution, De-Rates and Floor Conditions

Firs rule out any caster quoted on a static or single-wheel laboratory figure

Before you compare capacities, check what the number actually is. Published caster load ratings are laboratory values taken on a flat hard floor at roughly 70°F and about 3 mph with no impact^2,21.

Static and dynamic ratings are different quantities, and a caster rated high static can be rated much lower dynamic So a caster that looks adequate on a static figure will fail when you push it while loaded^15,19.

If the cart moves at all while carrying its working load, the dynamic rating is the one that governs, and it is the lower of the two^15,19. That single distinction is what makes the rest of the calculation necessary.

Every condition that moves your application away from the laboratory reference — speed, temperature, duty cycle, floor condition, swivel operation, side thrus Reduces the load the caster can carry, and the reductions are described as stacking rather than substituting for one another^2,21. A catalogue number is therefore a starting poin not a capacity.

Ask the supplier to state whether the quoted figure is static or dynamic, and at what reference speed and duty it was established, before you use it in any comparison.

Decide how many wheels share the load — three, not four

Dividing the total load by four assumes the weight sits equally on every wheel. That assumption holds on a new cart on a flat floor, where each of four wheels carries roughly a quarter of the load.

But real floors are not flat and during turns one or two wheels can briefly carry half the total or more²¹. Uneven ground, a shifting centre of gravity, steering and impact all push load onto individual casters⁷.

The conservative practice across the sources is to base per-caster capacity on three wheels rather than four, on the assumption that one caster is off the ground on an uneven floor^1,7,19.

One supplier frames this as an industrial convention: even with four casters installed, the load-bearing capacity of three is used as the estimate⁷. Dynamic load is calculated on one less wheel for the same reason¹⁹.

So your starting figure is total loaded weight — cart plus maximum payload, not empty weight — divided by three^1,21. Note what this does not give you: no source in this set quantifies the actual worst-case single-wheel share for a specific cart geometry.

So the divide-by-three rule is a screening assumption, not a measured load case.

Apply the published de-rates for your own floor, speed and duty cycle

These are the reductions that turn a catalogue number into an in-service capacity, and they multiply rather than add. A caster run at 5 mph, 180°F, 24/7 on a cracked floor is described as a fraction of its published rating once the de-rates stack².

That is why applying one large de-rate is not a substitute for working through each condition that applies to your route. Work through the table against your own operation and multiply the factors that apply.

The floor row is the one buyers most often skip: rough, cracked or uneven floors and thresholds introduce shock loading. And the instantaneous impact force when crossing a threshold or ditch may be several times the static load¹³.

The sources give no load-capacity-versus-floor-type table, so the 20-30% figure is a screening value rather than a measured penalty for your specific floor^2,21. Speed works the same way.

Dynamic capacity is specified at a reference speed, so a caster rated at 3 mph has lower capacity at 5 mph and lower still at 8 mph^2,21. The European standard scope stops at 1.1 m/s, about 4 km/h, and ball-bearing wheels can exceed that only with reduced load capacity²¹.

If your cart is towed or driven, the speed row is not optional.

Match wheel material and bearing type to the floor so you are not de-rated twice

Material and bearing choice change how much of the de-rated capacity you actually ge Rough workshop and warehouse floors favour hard polyurethane or nylon, which resist impact and wear; epoxy or noise-sensitive floors favour TPR or rubber, which absorb shock and reduce noise⁵.

Polyurethane is generally the stronger choice where wear resistance and load capacity matter, while rubber performs better where cushioning and floor protection are the priorities¹⁸. Matching the material to the floor you actually run on avoids paying for capacity you then lose to impact loading.

Bearing type is the second lever. Roller bearings are stated to carry more load than ball bearings because they distribute the load over a larger area, and roller or 6204 ball bearings are preferred for heavy loads^5,10.

One supplier reports more than 30% higher load capacity for roller or 6204 ball bearings than for ordinary bearings — treat that as that supplier's figure and verify it against the specific bearing offered⁵. Neither choice removes the need for the de-rates in the previous step.

It changes which caster can meet the de-rated requirement at an acceptable wheel diameter and cos not the requirement itself.

Set the safety factor as a range, and make the supplier state the rating basis behind it

This is the step where the sources part company. Quoted safety factors run from 1.25-1.5 up to 4x, and the higher values are tied to dynamic, towing or uneven-floor conditions rather than to a single industry figure^1,3,6,13,24.

One source recommends 4x on dynamic capacity for standard industrial use after the de-rates are applied¹. Another states 2x minimum for heavy-duty, rising to 3x when equipment is towed up inclines or over uneven floors³.

Others use a 1.25 dynamic factor plus a 25-30% margin, raising the factor to 1.5 or higher for thresholds, ramps or elevator gaps⁶, a 1.3-1.5 per-wheel factor for uneven force distribution¹³, or a 1.25-1.5 range that varies with speed, floor and movement frequency²⁴.

What the sources agree on is the direction, not the number. Higher travel speed reduces allowable dynamic capacity, so towing at speed or over rough surfaces is advised to raise the factor to 2-3x the calculated load^2,17,21.

Towing up inclines or over uneven floors is advised to move from 2x to 3x³. Dynamic applications such as thresholds, ramps or elevator gaps are advised to raise the dynamic factor from 1.25 to 1.5 or higher⁶.

No source in this set establishes a mandatory safety factor, and the differing values cannot be reconciled to one multiplier without a governing standard or a specific manufacturer's rating basis^1,3,6,13,24. So do not pick a number off a blog and write it into the RFQ.

Instead, state the factor you intend to apply, state whether it is applied to a static or a dynamic rating, and require the supplier to confirm the rating basis their quoted capacity rests on.

The magnitude depends on which rating the factor multiplies, so a 1.5 on a dynamic rating and a 1.5 on a static rating are not the same requirement²¹.

Check the specification against the standards and the commercial consequences before the RFQ goes out

Standards set a methodology floor, not a quality bar, and a higher margin buys heavier construction. Run these checks before you send the enquiry.

• ✓Confirm whether the quoted capacity is static or dynamic, and at what reference speed and duty it was established.
• ✓Confirm the per-caster figure was built on three wheels, not four, if the floor is uneven or the cart turns under load.
• ✓List every de-rate that applies to your route and multiply them — do not substitute one large de-rate for the stack.
• ✓State the safety factor you intend to apply and whether it multiplies a static or a dynamic rating.
• ✓Ask whether the caster cites ANSI MH28.3, and note that compliance standardises the rating methodology rather than guaranteeing service life — two casters passing at the same rated capacity can differ in bearing quality, wheel compound and swivel durability.
• ✓Check the travel speed against the European scope limit of 1.1 m/s (~4 km/h), above which ball-bearing wheels need reduced load capacity.
• ✓Normalise quotations on load, wheel diameter, brake, mounting, quantity and documentation before comparing price, since higher capacity means heavier construction, larger wheels and better bearings.

The competing safety-factor figures, side by side

The sources state different figures for the same quantity, and the buyer cannot take a single number as the industry figure from these sources alone^{1,3,6,13,15,17,24}. The table keeps both sides visible so you can see what evidence would settle the number for your application: a governing standard, or a specific manufacturer's stated rating basis.

Neither is quoted in this se

Before you request a quote

• Public sources do not settle this — put it to the supplier: no manufacturer's own published rating basis (static vs dynamic, single-wheel vs set of four) or stated derating guidance is quoted.
• Public sources do not settle this — put it to the supplier: no quantified dynamic load amplification factor for drops or obstacles — only qualitative statements that impact may be several times static load.
• Public sources do not settle this — put it to the supplier: no speed/load curve or table for any wheel type.
• Public sources do not settle this — put it to the supplier: no load-capacity-versus-floor-type data.
• Public sources do not settle this — put it to the supplier: no quantified worst-case single-wheel load share for a specific cart geometry.
• Public sources do not settle this — put it to the supplier: no allowable stress, impact strength or temperature-limit data for wheel materials.
• Public sources do not settle this — put it to the supplier: no standard text specifying a minimum caster safety factor.
• Request a written quote with price basis, MOQ, tooling, payment terms and production lead time.