Pull Force & Holding Force Explained

A published pull force is the breakaway force measured under deliberately ideal conditions: the magnet in direct, flat contact with a thick, ground, low-carbon steel plate, pulled slowly and squarely away in the direction perpendicular to the contact face. Change any of those conditions and the number changes — usually downward, often dramatically.

Ratings answer the question "what is the most this magnet can ever hold?" They are honest maximums, comparable across a catalog, and reproducible in a lab — and they are the starting point of a holding design, not the design itself.

At contact, pull force scales with the square of the flux density crossing the interface, times the contact area (F ≈ B²A/2µ₀). The practical consequences:

• Geometry beats grade. Because force goes as B², and grade steps move Br only a few percent each (see the grades chart), an N52 pulls only ~15–20% harder than an N42 of identical size — while a modestly larger magnet, or a better circuit, can double the force.
• Thickness matters up to a point. For a given diameter, added thickness raises pull force with diminishing returns — beyond roughly an aspect ratio of one, more length adds little at contact.
• The circuit multiplies. A magnet in a steel cup (pot magnet) channels flux that would otherwise leak into the working face — the same magnet can deliver 2–4× its open-face pull. This is the highest-leverage design move in holding applications.

Everything that separates the magnet face from the steel — paint, powder coat, plating, anodize, dirt, surface roughness, plastic housings, "just a label" — is an air gap magnetically, and pull force falls steeply with it. Representative behavior for a typical small disc magnet (exact curves depend on geometry — squat magnets fall faster than long ones):

Indicative ranges for small disc geometries at contact-dominated scales; use the calculator or test hardware for design values.

• Thin steel saturates. The rating assumes steel thick enough to carry all the magnet's flux. Sheet metal can't: a strong 1" disc on 22-gauge (0.8 mm) sheet may deliver only a fraction of rated pull because the saturated sheet lets flux leak past. Rule of thumb: mild steel roughly as thick as ~1/2 the magnet's smaller face dimension is safe; when in doubt, test or ask.
• Alloy matters. Low-carbon steels (1006–1018) are the reference. Hardened and high-carbon steels hold less; 400-series stainless holds noticeably less; 300-series stainless (304/316) is essentially non-magnetic — the most common "our magnet doesn't work" support call in existence.
• Surface condition: mill scale, rust, and rough finishes act as gap; a ground or smooth cold-rolled surface performs best.
• Cast iron holds less than mild steel; aluminum, brass, zinc die castings hold nothing.

Ratings are a straight pull-off. Most real products load magnets in shear — gravity dragging a hook down a refrigerator door — where the magnet doesn't detach, it slides. Sliding resistance is friction: roughly the friction coefficient times the magnetic normal force, typically 15–30% of rated pull for smooth nickel-plated magnets on smooth steel.

• A 20 lb-pull magnet on a vertical smooth surface may reliably support only 3–6 lb before creeping downward.
• Fixes: a rubber-coated magnet face or friction pad (raises µ dramatically — rubber-jacketed pot magnets often hold more in shear than bare ones despite a lower rated pull), a mechanical ledge or hook feature that converts shear to bearing, or simply more magnets.
• Peel is worst of all: lifting one edge concentrates the entire detachment on a thin line — a magnet that resists 20 lb of flat pull can be peeled off with a finger. Products that get peeled (closures, covers) should be sized on peel testing, not pull ratings.

• Reversible loss: NdFeB loses roughly 0.11–0.13% of Br per °C above room temperature — and since force goes as B², pull falls about twice that fast: ~11% down at 80 °C, recovering on cooling.
• Irreversible loss: exceed the grade's real temperature capability (rating + load line — see Magnets 101) and some loss stays after cooling. A holding design that lives warm should be specified per the grades chart temperature classes.
• Cold is friendly to NdFeB (slightly stronger) but hostile to ferrite, which can demagnetize under load in deep cold — relevant for outdoor ferrite holding assemblies.

These are the levers a supplier reaches for when your envelope is fixed and the force target isn't met — often cheaper than more magnet. Assemblies (pots, channels, rubber-coated systems) exist precisely because raw magnets are rarely the optimal holding device.

Work the chain in order, then add margin:

• Step 1 — Start from rated pull for the candidate size (catalog or calculator).
• Step 2 — Derate for the interface: gap/coatings, steel thickness and alloy, surface condition.
• Step 3 — Derate for load direction: shear ≈ 15–30% of the derated normal force unless friction is engineered; peel gets its own test.
• Step 4 — Derate for temperature at the hottest expected condition.
• Step 5 — Apply the safety factor to what remains: ≥2× for benign static holding, 3–5× where vibration, bumps, or user handling occur, and higher (with mechanical capture as backup) wherever a release could injure someone or drop something valuable — the same fail-safe principle as in the mounting guide.

• Estimate with the pull force calculator for size and gap studies — calculators assume the ideal conditions of section 01, so treat outputs as the top of the derating chain.
• Verify on real hardware: your steel, your coating, your load direction, your temperature. A luggage-scale pull test on the actual product surface teaches more than any table on this page.
• For production acceptance, don't use pull tests to accept magnets — use helmholtz flux measurement and keep pull testing for design verification, as covered in How Magnets Are Tested & Measured.