Heavy-Duty vs Light-Duty Casters for 500–2000 kg: What to Verify Before You Buy

Rule Out Any Caster Rated Only on Static Load

Static capacity is the higher number and is not the service-life figure for a rolling application. Dynamic rating is the number that governs a caster in motion, and it is typically 60–75% of static 1 .

A catalog that quotes only static therefore overstates what the caster will carry once the cart rolls. One supplier's polyurethane wheel shows how far the two can diverge: 661 lb static against 331 lb dynamic at roughly 1.86 mph, about 50% of static 12 .

That sits below the general 60–75% band, so the ratio is not a fixed rule you can apply to every wheel. It is a reason to read the actual dynamic figure rather than derive one.

Before you compare anything else, ask for the dynamic rating at your operating speed and the basis on which it was measured. A caster whose only published number is static has not been qualified for your application.

Set the Per-Caster Load on the Conservative Basis

On uneven or worn industrial floors, one of four casters is routinely off the ground because of frame racking or floor slope, and the remaining three absorb the full load 1 . Dividing total weight by 4 assumes all four stay in contact, which is optimistic and leads to early failure 1 .

Dividing by 3 is the procurement-safe assumption 1 9 18 25 . That basis is a load-distribution convention, not a standard.

The sources do not establish that any governing standard requires it, nor do they quantify how often a caster lifts off 1 9 18 25 . Treat it as the conservative starting figure for your calculation.

For a 500–2000 kg cart, the difference is material: a 1,500 kg cart yields 500 kg per caster on the divide-by-3 basis against 375 kg on divide-by-4. Size the caster to the higher number, then apply your safety margin on top.

Choose a Safety Margin for the Actual Duty

Reported safety factors range from 1.2 to 5x depending on source and application, and no single value is established as mandatory 2 8 9 10 18 19 22 32 33 . The lowest figures are associated with smooth-floor or general use; the highest with industrial and AGV duty 9 10 22 .

One supplier describes 2x as a marketing minimum that fails within 6–12 months under real duty, and 4x as the industrial standard for 3–5 year service life, with a 20–30% up-front cost delta 2 .

That is a supplier position, not a standard, but it frames the trade: the margin you buy is service life, not just capacity. Duty cycle modifies the margin.

For 24/7 use, capacity ratings — which assume intermittent service — should be derated by 25–33% 28 . AGV and 24/7 constant-speed duty is reported to need a 5x factor plus fatigue verification, because standard 8 hr/day ratings under-spec that duty by 30–40% 2 .

A worked instance from one source: a 1,000 kg load on four casters is sized at 600 kg each, using a 2x multiplier on a 500 kg dynamic rating plus a 20% margin 19 .

That is one supplier's arithmetic, not a formula to copy — but it shows how the per-caster figure, the multiplier and the margin stack.

Rule Out Wheel Materials That Cannot Reach the Range

Wheel material caps the per-caster figure before any bracket or bearing question matters. Rubber tops out well below the 500–2000 kg band, so it is out of scope for this duty regardless of how the caster is labeled 25 .

The families that reach the range are nylon, cast iron or steel, and heavy polyurethane 25 22 26 . The choice among them trades floor protection against push and pull effort: cast iron and steel carry the highest loads but demand floor protection and generate noise, while polyurethane sits between load capacity and floor care 25 .

Check the material ceiling against your per-caster figure before you look at price. A caster whose wheel cannot reach the number is not a heavy-duty caster for your application, whatever the catalog says.

Match Bearing Type to Load and Duty Cycle

Bearing mis-spec is reported more often than wheel mis-spec, and it shows up as axle wear rather than wheel failure 4 . One supplier reports a 1,000 lb die cart on plain bearings, pushed 200 ft per cycle, wearing its axle out in under 90 days — the wheel was not the problem 4 .

Bearing families carry different load ceilings per wheel. Plain bearings reach up to 1,200 lb, roller (needle or tapered) up to 3,500 lb, and precision sealed ball up to 2,200 lb 4 .

Roller bearings are the heavy-duty default for high load and shock; precision sealed ball bearings suit long-run and AGV duty where push force matters 28 11 . For a 500–2000 kg per-caster application, the bearing must be rated for the duty cycle, not just the load.

Ask what bearing family is fitted and what load it is rated to carry at your speed.

Confirm Mounting and Bracket Construction

For loads over 1 ton, the mounting method carries the attachment strength, and threaded stem insertion is not acceptable at that load 13 . Plate mounting with welded or bolted attachment is the preferred route 13 .

Check these before accepting a quote.

• ✓Confirm the mounting type is plate, not threaded stem, for loads above 1 ton
• ✓Verify bracket plate thickness is at least 5.5 mm steel
• ✓Confirm double-sided welding on the bracket, reported 30% stronger than single-sided
• ✓Check the wheel core uses a double-bearing design for even load distribution
• ✓Match plate dimensions and bolt-hole pattern to your frame before ordering

Demand Test Evidence Before Releasing the Order

Named standards and test methods exist, but no source in this set provides an actual test report or certificate for any caster 7 15 20 35 22 . You have to request the evidence rather than assume compliance.

The test methods themselves are defined: a static load test applies 150% of rated load for 24 hours, and a dynamic test cycles 100,000+ times at 5 km/h 20 . Those are the numbers to ask a supplier to demonstrate against, not to assume were run.

Standards designations differ between sources — EN 1253, EN 12530 and ISO 22883 all appear in connection with caster requirements, and no source establishes which is mandatory for a given market 7 15 35 . Ask your supplier which standard applies to your target market and request third-party reports such as SGS or TÜV 15 .

Separate the Technical Rating From the Commercial Request

The load calculation produces a rating, not a purchase order 38 .

If price, volume breaks, minimum order quantity and lead time are not requested alongside the rating, you select on capacity alone and discover at order stage that the caster has a minimum order far above your build or a lead time that stops the project 38 .

Custom caster MOQ is reported at 2,000–3,000 pieces, with simple modifications as low as 500 pieces, and sample lead times of 7–10 or 15–25 business days depending on whether new tooling is required 36 . Those are one supplier's terms, not an industry schedule.

Ask for unit price at the quantity you actually need, volume breaks at the next order size up, MOQ for the specific wheel and bearing configuration, and lead time on the required quantity 38 . Ask for the rating basis in writing alongside the price so you can compare suppliers on the same terms 38 .

Where the Sources Disagree

The safety-factor figures range from 1.2 to 5x and no single value is established as mandatory 2 8 9 10 19 22 25 32 33 . The dynamic-to-static ratio and the standard designation also differ between sources.

Verify the side that matches your duty before you commit.

What the sources do not establish

• No source provides an actual third-party test report or certificate for any caster.
• No source gives a numeric brake holding-capacity rating.
• No source gives a numeric rated load per mounting style (plate vs stem vs bolt-hole).
• No source gives volume unit price at 500–2000 kg duty, warranty terms, or spare-wheel availability.
• No source establishes which standard is mandatory for a given market or application.
• No source gives a standardized bearing load rating method or speed limit in rpm.
• No source gives kingpin or raceway construction details with numeric specifications.
• No source gives a standard test method for the wheel material capacity bands.