Key takeaways
- Four casters do not give four times one caster's rating. Divide the loaded weight by three.
- Wheel diameter is the biggest lever on obstacle climbing. A 6 in wheel needs about 40 percent less force than a 3 in wheel over the same lip.
- Softer treads protect floors and absorb shock. Harder treads roll easier and resist flat spotting.
- Total lock brakes stop the wheel and the swivel. A plain wheel brake lets the caster still steer.
Casters are where a tool box meets the building. Everything above them can be excellent and the box will still be unpleasant if the wheels are 3 in nylon on a floor with a seam every 12 ft. Casters are also, conveniently, the easiest part of a cabinet to upgrade.
Caster anatomy in the order it matters
Mounting. Two options. A plate caster bolts through a flat top plate with four fasteners, which spreads load into the cabinet frame and resists the bending moment a swiveling wheel generates. A stem caster threads or presses into a socket, which is compact but concentrates all load and moment into one small connection. Tool storage cabinets use plate casters almost universally, and any cabinet using stems for a heavy roll cab is worth a second look.
The rig, or yoke. The formed steel fork holding the axle. On heavy casters it is a thicker, deeper legged fork, often with a reinforcing gusset. The rig is what bends when a loaded cabinet drops off a threshold.
The swivel section. This is the part that separates a good caster from a cheap one.
- A kingpin swivel stacks a top plate, a load race, a thrust race, two ball sets and a central bolt or rivet that holds the assembly together. The kingpin is typically around 5/8 in diameter, so the whole swivel load and all shock loading pass through that one fastener. When a kingpin caster fails, the kingpin is usually what failed.
- A kingpinless swivel deletes the central bolt and instead runs two precision raceways, one carrying load and one carrying thrust, sealed inside a forged or heavy welded top plate. Because the raceway diameter is roughly 2 in instead of 5/8 in, the bearing circumference is three times larger and shock is spread over far more contact area. Kingpinless designs dominate the high capacity industrial range.
The axle and bearing. Plain bore is cheapest and stiffest but drags. Roller bearings are the shop standard. Precision ball bearings roll best and cost more. Sealed bearings matter in any environment with metal chips or grinding dust, which is most shops.
Brakes. A face brake or wheel brake presses a pad or cam onto the tread, stopping rotation only. The caster can still swivel, so a braked cabinet can still slowly walk under a side load. A total lock brake stops the wheel and locks the swivel simultaneously. On a heavy roll cab that gets leaned on, total lock is the meaningful upgrade.
Load rating math, and why four casters are not four times one
Publisher ratings are per caster. The temptation is to multiply by four. Do not.
Real floors are not flat. Concrete has slope for drainage, seams, and trowel variation. Cabinet frames flex. At any instant, one of the four casters is likely carrying much less than its share, and if the frame is stiff, potentially nothing at all. The industry sizing convention accounts for this directly.
Two refinements worth knowing:
- Uneven fore-aft loading. Tool cabinets are front-heavy when drawers are open and rear-heavy when they are not, but they are usually not symmetric side to side either. If one bank of drawers holds the impact sockets, bias the calculation toward the loaded corner rather than assuming an even split.
- Swivel casters generally need more capacity than rigid ones in the same position, because the swivel offset adds a bending moment the rigid caster does not see. Where a manufacturer publishes different ratings for the swivel and rigid versions of the same wheel, believe the lower one.
Feed your own numbers through the drawer load calculator to establish loaded weight first, and see tool box load capacity for how the caster limit interacts with the other three.
Wheel materials and what they do to your floor
| Wheel | Typical hardness | Floor friendliness | Rolling ease | Notes |
|---|---|---|---|---|
| Polyurethane on aluminum core | 85A-95A | Good | Very good | The premium shop default. Non-marking, quiet, high capacity for its size. |
| Polyurethane on steel core | 85A-95A | Good | Very good | Highest capacity poly option. Heavier and more expensive. |
| Polyurethane on nylon or polyolefin core | 85A-95A | Good | Good | The mainstream tool box caster. Good value, lower capacity than metal core. |
| Soft rubber | 65A-75A | Best | Poor | Quietest and gentlest. Low capacity, high rolling resistance, flat spots readily. |
| Thermoplastic rubber (TPR) | 70A-85A | Very good | Fair | Non-marking, quiet, common on light and medium cabinets. |
| Nylon or glass-filled nylon | Shore D range | Poor | Excellent | Rigid, no compression layer, point-loads coatings and marks epoxy. |
| Phenolic | Shore D range | Poor | Excellent | Very high capacity, heat tolerant, holds parked loads without flat spotting. Loud, and it marks finished floors. |
| Cast iron or steel | Rigid | Worst | Excellent | Foundry and mill duty. Has no place in a tool storage cabinet on a finished floor. |
Durometer is simply hardness. Elastomer treads are quoted on the Shore A scale, where 65A is a soft rubber and 95A is a firm polyurethane. Rigid wheels use Shore D. The trade is completely predictable:
- Harder tread: rolls with less effort, carries more load per inch of tread width, resists flat spotting, transmits more shock and does more to the floor.
- Softer tread: conforms to floor irregularities, absorbs shock, protects coatings and quiets the roll, at the cost of capacity and rolling resistance.
For a coated or polished concrete garage floor, polyurethane in the 90A to 95A range on a metal or nylon core is the balanced answer. For a wood shop floor or an upstairs installation, drop toward 85A. For a heavy cabinet that lives in one place and moves twice a year, a harder wheel is genuinely better because it will not take a compression set.
Rolling resistance, seams, cords and wheel diameter
The complaint "this box does not roll" is nearly always about obstacle climbing rather than steady rolling.
The force needed for a wheel to climb a step is a geometry problem. For a wheel of radius R meeting an obstacle of height h under load W, the required horizontal push is approximately:
F = W x sqrt(h x (2R - h)) / (R - h)
That table explains a lot of shop behavior. Expansion joints, air hose runs, extension cords, welding leads and door thresholds are all half-inch class obstacles. A cabinet with 6 in casters crosses them. A cabinet with 3 in casters gets shoved, stalls, and then gets shoved harder, which is exactly the tipping scenario worked through in tool box load capacity.
Two secondary effects:
- Tread width matters on soft surfaces and coated floors. A wider tread spreads pressure and reduces both rolling resistance on soft floors and coating damage. On hard, smooth floors it makes little difference.
- Swivel lead (the offset between the swivel axis and the wheel contact patch) determines how readily a swivel caster turns to follow. Longer lead swivels more willingly and wanders more.
Swivel and rigid layouts
| Layout | Behavior | Best for |
|---|---|---|
| 2 rigid rear, 2 swivel front | Tracks straight, steers from the handle end, predictable | The tool box standard. Corridors, long bays, repeated routes. |
| 4 swivel | Moves sideways and rotates in place, no fixed heading | Tight bays, cabinets parked parallel to a wall, service carts |
| 2 rigid center, 2 swivel each end (six wheel) | Pivots about the center pair, very short turning radius | Long 60 in and 72 in cabinets, covered in tool chest sizes |
| 4 swivel with two total locks | Full maneuverability, positive parking | Cabinets that get leaned on or bumped |
The frequent complaint about four-swivel cabinets is that they wander when pushed in a straight line. The frequent complaint about two-rigid cabinets is that they will not go sideways into a parking spot. Both are true. Choose based on whether your floor time is mostly travel or mostly parking, and read that decision alongside your garage shop layout.
Flat spotting, and what parking does to a wheel
An elastomer tread under continuous load takes a compression set. The tread deforms at the contact patch, and given months and a warm shop it stops springing back. The cabinet then thumps once per revolution and the swivels fight it.
Risk factors, in order:
- Soft tread. Rubber at 65A to 75A is the most vulnerable. Polyurethane at 90A and above is much more resistant.
- High load per wheel. A cabinet loaded to its caster rating and parked is the worst case.
- Heat. Elastomers set faster warm. A cabinet parked in an uninsulated garage through a hot season is exposed.
- Time without movement. Months, not days.
The mitigations are trivial: move the cabinet a few inches every month or two, keep the heaviest drawers loaded no higher than they need to be, and for a cabinet that genuinely never moves, specify phenolic or high durometer polyurethane, which hold parked loads without setting.
Upgrading casters on a cabinet you already own
This is one of the highest-value upgrades in tool storage, and it is mostly a measuring job.
- Measure the existing top plate and the bolt hole pattern. Common industrial plate sizes cluster around 4 in x 4-1/2 in for medium duty and larger rectangles for heavy duty, but hole patterns vary by manufacturer. Measure center to center in both directions rather than trusting a catalog category.
- Measure overall caster height, floor to the top of the plate. Changing this changes drawer heights, work top height, and stability geometry. A 1 in taller caster raises the whole center of gravity by 1 in and reduces the tipping margin slightly.
- Match or exceed the plate footprint. A smaller plate on the same holes concentrates load into less frame metal.
- Check what is behind the mounting holes. If the cabinet base is formed sheet with no backing plate, add fender washers or a backing plate so the new fasteners do not pull through. This is the point where steel gauge becomes practical rather than theoretical.
- Keep at least two braked positions, and prefer total lock over wheel-only brakes.
- Buy all four the same. Mixing wheel diameters or durometers guarantees that one caster carries more than its share permanently.
Caster selection by floor and load
| Floor | Loaded cabinet weight | Recommended wheel | Diameter | Rig notes |
|---|---|---|---|---|
| Sealed or coated concrete, few seams | Under 500 lb | Polyurethane 90A on nylon core | 5 in | Standard kingpin fine, two total locks |
| Sealed or coated concrete, seams and cords | 500-900 lb | Polyurethane 92A-95A on metal core | 6 in | Kingpin acceptable, kingpinless preferred |
| Bare or rough concrete | 500-1,200 lb | Polyurethane 95A on steel core | 6-8 in | Kingpinless, sealed bearings |
| Epoxy or polished concrete, appearance matters | Under 700 lb | Polyurethane 85A-90A, wide tread | 5-6 in | Avoid nylon and phenolic entirely |
| Wood shop floor or upper story | Under 600 lb | Soft polyurethane or TPR, 80A-88A | 5-6 in | Wide tread to spread point load |
| Trailer or service van deck | Any | Polyurethane on metal core plus tie-downs | 4-5 in | Casters are for positioning, straps carry transit loads |
| Cabinet that essentially never moves | Any | Phenolic or 95A polyurethane | 5-6 in | Immune to flat spotting, keep it braked |