For noise-sensitive interiors, the evidence points to soft treads — rubber, soft polyurethane or TPR — as the quietest option, with rubber reported 5–8 dB below comparable polyurethane on tile floors. Hard treads (nylon, phenolic, steel) are consistently identified as the loudest and should be ruled out before price comparison.

The sources do not publish absolute dB levels per material or per floor, so the choice between rubber and soft polyurethane remains a supplier-verification question rather than a settled one.

Rule out hard treads before comparing prices

Tread hardness is the one noise decision the sources treat as settled: soft elastic materials absorb vibration from floor joints, thresholds and surface imperfections, while hard materials transmit that energy into the floor and the frame 1 , 8 , 10 , 17 , 22 , 31 . Rubber and soft polyurethane or TPR are named for hospitals, offices and quiet workshops 1 , 8 , 10 , 17 , 22 , 31 .

Nylon, phenolic and steel are the loudest 10 . A quote built on a hard tread is not a quiet caster, whatever its price, so filter those out before you compare unit costs.

The one comparative figure available is that rubber casters run typically 5–8 dB below comparable polyurethane on tile floors 10 ; that is a tile-floor comparison, not a universal rule.

For a hospital or office cart, the shortlist is rubber or soft-tread polyurethane, and the supplier should be asked to state the tread compound and hardness on the datasheet.

Match tread hardness to the floor, not to the caster alone

Hard floors — concrete, tile, hardwood, epoxy — reflect rolling noise and need a soft tread to damp it and protect the surface 24 , 26 , 27 . Carpet behaves the opposite way: a soft tread sinks and plows, so a harder tread such as nylon rolls more easily 3 , 24 .

The same caster cannot be specified for a tiled corridor and a carpeted ward without accepting a compromise on one of them. The guiding principle the sources give is that the harder the floor surface, the softer the caster wheel should be 26 .

On ultra-smooth polished tile, a softer TPR wheel may be needed for grip 26 . Before the RFQ, list every floor type the cart will cross — including thresholds and lift entrances — and ask the supplier to confirm the tread recommendation for each.

Choose bearing and rig features that keep rolling and swiveling smooth

Precision ball bearings are recommended for quieter operation in hospitals and office settings 18 , and sealed precision ball bearings roll 15–25% easier than plain bushing bearings 4 . Ball bearings at the wheel axle matter more than the swivel raceway for rolling resistance 4 .

On the swivel side, a trail under 1.25 inches causes shimmy and heavy steering, while 1.5–2 inches is the optimal range for manual carts 5 . Above 800 lb per caster, tapered roller raceways drop swivel force 30% against double-ball 5 .

Kingpinless rigs are recommended for shock, thermal cycling or tugger duty, with kingpin rigs acceptable for light manual use only 7 . Optional raceway seals are available and ideal for healthcare applications 28 .

None of these features carries a measured dB figure in the sources; they are recommended for smoother, quieter operation, so the RFQ should ask the supplier to state the bearing type, seal type and trail dimension rather than a noise number.

Fix wheel diameter and tread profile for the load and turning duty

Larger wheels roll over surface imperfections instead of pushing through them, and they spread load over a larger area, reducing floor pressure 4 , 23 . Going from 4-inch to 6-inch wheels typically drops push force 25–35% 4 .

Diameter recommendations in the sources are load-based, not noise-based: up to 300 lb per caster, 3–4 inch wheels; 300–900 lb per caster, 4–5 inch wheels 3 .

Tread profile is a separate trade-off — crowned tread turns easier but rolls slightly harder than flat tread, and crown matters more on turning carts than on straight-line tuggers 4 , 25 .

For a quiet indoor cart, the diameter should be set by the load and the obstacle height, then the profile by how much turning the cart does. The sources do not publish a diameter range specifically for quiet indoor use, so the load table is the only sizing basis available.

Decide on shock absorption, rig type and mounting interface

Spring-loaded shock-absorbing casters cushion loads and reduce vibration when moving across uneven floors 29 , and for floors with seams or thresholds a larger-diameter rubber wheel or shock-absorber wheel is recommended for its buffering effect 31 .

Mounting types include top plate, stem (threaded, grip ring, grip neck) and kingpinless designs, and the choice must match the equipment interface 27 .

The sources describe these mounting types but do not differentiate them by vibration or noise performance, so the mounting decision should be driven by the equipment frame and the duty cycle, not by a noise claim.

Pre-RFQ checklist: normalize the specification and the evidence

Cost is driven by wheel size and load, tread and core material, brake or lock system, mounting design, bearings and steering structure, OEM customization, and testing or compliance 14 , 30 .

Quotes are only comparable when those lines are normalized, so ask every supplier to quote against the same specification and to state what documentation, sample cost, tooling and lead time come with it 14 , 30 .

  • ✓State the floor types the cart will cross and the required tread material.
  • ✓Specify total equipment weight, number of casters, safety margin and wheel diameter.
  • ✓Name the exact brake or lock mode and how often the equipment is parked or repositioned.
  • ✓Provide the mounting drawing, height, hole pattern or thread details.
  • ✓State the required maneuverability, noise target and movement frequency.
  • ✓Ask for the bearing type, seal type and swivel trail dimension on the datasheet.
  • ✓Request test reports covering load capacity, material specs and internal QC.

Where the sources disagree on noise

The disagreement is about which soft tread wins, not whether soft treads win. One source ranks rubber above polyurethane for hospitals, offices and hardwood 1 , another reports rubber 5–8 dB below comparable polyurethane on tile 10 , and a third markets polyurethane's softer tread as the low-noise choice for hospitals 22 .

The evidence does not resolve which soft tread is quieter for a given project, so the buyer should ask each supplier for a comparative noise test on the actual floor type.

Disputed item (with unit)One source reportsAnother reportsWhat the buyer should do
Noise level of rubber vs polyurethane on hard floors (dB)Rubber is quieter than polyurethane for hospitals, offices and hardwoodRubber is typically 5–8 dB below comparable polyurethane on tileAsk each supplier for a comparative noise test on the project floor type
Quietest tread for hospitals (material)Rubber wins for hospitals, offices and hardwoodPolyurethane's softer tread reduces rolling noise, ideal for hospitalsVerify the tread compound and hardness on the datasheet before quoting

What the sources do not establish

  • No absolute dB levels per material on hard versus resilient flooring.
  • No durometer-to-noise mapping; the Shore A 85–92 range is a floor-protection recommendation, not a noise curve.
  • No measured dB contribution of bearing type to rolling or swiveling noise.
  • No recommended diameter range specifically for quiet indoor use; the diameter figures are load-based.
  • No tread-width-to-noise relationship.
  • No quantified noise or floor-marking effect of specific tread profiles or edge geometry.
  • No measured dB data per floor type (tile, concrete, vinyl, laminate, epoxy versus carpet, rubber, cork).
  • No quantified noise reduction from raceway type, kingpin versus kingpinless, seals or offset.
  • No quantified vibration or noise differences between mounting types.
  • No ISO 22878/22879 certification or specific noise test reports.
  • No volume price breaks or tooling/MOQ for custom treads beyond general statements.
Sources · 21

Technical references cited for verifiability — not supplier recommendations. Browse the research library.