Short answer: the induced-currents rule in the NEC targets ferromagnetic metal, and the annealed 304 or 316L sheet your stainless box is made from is essentially non-magnetic — 316L measures about 1.01 relative permeability against roughly 200 for carbon steel. An as-delivered austenitic stainless enclosure barely magnetizes, so the induction heating the rule was written to stop has nothing to grip. Group all conductors of each circuit through one entry anyway: you satisfy the rule's letter, it costs nothing, and it works in every metal. The zones to watch are the ones that have been cold worked or welded, because 304 can pick up a real magnetic pull there.
What the induced-currents rule asks of you
The NEC's rule on induced currents covers alternating-current conductors in ferrous metal enclosures and ferrous raceways. It requires all phase conductors, and where used the grounded conductor and all equipment grounding conductors, to be grouped together. The stated purpose is to avoid heating the surrounding ferrous metal by induction. A feeder whose conductors pass through a steel wall in separate holes forms a one-turn transformer with that wall; the circulating current and the alternating flux heat the steel. At 60 Hz the flux reverses 120 times a second, and a separated high-current set — say 200 A per conductor — can push a knockout rim past the 90 C [194 F] class of its insulation and above the 120 C [248 F] continuous limit of a typical neoprene gasket. Grouping the whole set in one entry drives the net field at the wall toward zero and removes the mechanism.
So the action is simple: every conductor of an AC circuit goes through the wall in the same hole or gland, whether that hole is an M20 or a 1 in NPT hub. The open question for a stainless box is whether the metal is ferrous in the sense that matters.
Why magnetic beats ferrous in this rule
Ferrous means the alloy is based on iron, and stainless steel is ferrous by that definition. Ferromagnetic means a high relative permeability, written μr, and that is what the heating needs. A high μr pulls the alternating flux out of the air and into the wall.
The concentrated flux turns to heat two ways: hysteresis loss as magnetic domains flip each cycle, and eddy currents induced in the conductive steel. Aluminum makes the contrast plain. It conducts, so eddy currents do circulate — the same effect that brakes a pendulum swinging between magnets — but at μr of 1.0 the flux stays at the air value and there are no domains to flip. The useful question about a stainless enclosure is therefore not whether it contains iron; it is what the sheet actually measures for μr. For how a conductive wall behaves around alternating fields in general, see how metal enclosure shielding attenuates magnetic fields.
Where annealed 304 and 316L land, and what a magnet shows
Published permeability numbers
Annealed austenitic stainless sits just above air. One published measurement set lists 304L at about 1.1, 316L at 1.01 and 316LN at 1.005; another records 1.003 for as-built 316L with zero cold work. Carbon steel runs orders of magnitude higher: handbook curves for low-carbon steel span roughly 100 to well over 1,000 across the working range, and the worked example below takes 200 as a mid value. On those numbers, an annealed stainless wall carries about one percent more flux than air, while a carbon-steel wall carries 100 to 1,000 times more. Even a 10 percent cold-work raise in 304L takes it only to about 1.2 — still two orders of magnitude below any carbon steel.
What the magnet test shows
A magnet is your shop-floor check, and it takes seconds per entry. Annealed 304 and 316L sheet will not hold a magnet that grabs a file, while heavily cold-worked zones will. Read it as a pass-fail scan, not a measurement: it shows ferromagnetic response, it cannot tell 304 from 316L, and a weak magnet over a thin worked skin — a flange turned on a 1.5 mm [0.06 in] door edge — can mislead you. Scan the wall around every entry, hinge and seam, and apply the carbon-steel arrangement anywhere it pulls. When you need the alloy itself confirmed before welding or repair, identify the stainless alloy before you weld it.
A 100 A feeder 25 mm from the wall: carbon steel against annealed 316L
This is the case the rule exists for: one conductor of a circuit passing a wall, with its return path far away, so the wall sees the full field. The inputs are worked-example assumptions — swap in your own circuit and geometry. The circuit carries 100 A. The conductor runs 25 mm [about 1 in] from the inner wall face, a typical standoff for lugged feeders in a compact wall-mount box. The permeabilities are the published values above: 1.01 for annealed 316L and 200 for carbon steel as a handbook mid value. The field of a long straight conductor is H = I / (2πr), and the flux density in a material is B = μ0 · μr · H with μ0 = 4π×10⁻⁷ H/m — the same relation behind worked inductor calculations from relative permeability.
Field strength at the wall: H = 100 A / (2π × 0.025 m) = 100 / 0.157 m = 637 A/m.
Air and annealed 316L: B = 4π×10⁻⁷ T·m/A × 637 A/m = 8.0×10⁻⁴ T = 0.80 mT, and 1.01 × 0.80 mT = 0.81 mT. The stainless wall sits one percent above air — nothing for the flux to concentrate in.
Carbon steel: 200 × 0.80 mT = 160 mT (0.16 T), two hundred times the flux. Eddy heating tracks roughly the square of flux density at a fixed frequency and wall thickness. The steel therefore carries about 200² = 40,000 times the eddy loss per unit volume, and hysteresis loss, which follows about the 1.6 power of flux density (the usual Steinmetz exponent for iron), about 200^1.6 ≈ 4,800 times.
The decision follows directly. Annealed 316L at 0.81 mT gives the field no reason to concentrate, so there is no induction-heating threshold to trip. Carbon steel at 0.16 T is the case that scorches washers and cooks gaskets — and that is why the grouping rule exists. Group the same 100 A set through one gland and, assuming a 5 percent residual imbalance, the field at the wall falls to 0.80 × 0.05 = 0.04 mT and the heating to about 0.05² = 0.25 percent of the single-conductor case, in any metal. When the wall at an entry holds a magnet, put every conductor of that circuit through one hole. When it does not, grouping still costs you nothing and keeps the arrangement valid in every metal, including the zones that might go magnetic later.
Where a stainless box turns magnetic
Cold work converts austenite to martensite, and that is where your box picks up magnetism. Flanged door edges, punched knockouts, rolled threads and bent feet on 304 take a strong magnetic response after substantial cold working. 316 stays essentially non-responsive through the same work. Welding can convert the seam and its heat-affected zone — a few millimetres either side of the bead — the same way, and deep cold below about -50 C can do it spontaneously, which matters in cryogenic service rather than on a panel.
The magnetism is not a corrosion signal: cold working changes the structure, and the composition and corrosion behavior stay put. Keep weld zones cleaned and protected as described in passivation for stainless steel enclosures. Any zone that pulls the magnet gets the carbon-steel treatment — conductors of a circuit through one entry.
How to group entries in a real panel
Group by circuit: each AC circuit's phases, its grounded conductor and its equipment grounding conductors leave the box through the same hole or gland. That one arrangement satisfies the NEC's letter for a ferrous enclosure and kills the heating mechanism in magnetic and non-magnetic walls alike. It also stops parallel conductor sets from circulating current between separate entries; running parallel conductors through a steel enclosure without induction covers that side of it. Steady DC builds a static field that cycles nothing, so a DC feed adds no induction heating, and a 24 V DC control pair at a few hundred milliamps carries too little current for the effect to register.
Water and dust ratings say nothing about any of this: a Type 4X / IP66 rating covers hose-directed water and corrosion, and ours is rated NEMA 4X / IP66 by the manufacturer; not UL listed. When you add an entry later, de-energize, lock out and verify absence of voltage first, then see drill bit speed and feed for stainless enclosures before you cut. A standard austenitic box needs no anti-induction hardware, no spacers and no split entries; the grouped gland is the whole job.
What to put on the order
Set the grouping on the drawing instead of in the field. List each AC circuit's entry as one cutout with a gland sized for the whole bundle, and cluster related circuits on one face so spare conductors keep their grouping — for example, one M32 gland for a three-phase set of 10 mm² [7 AWG] conductors plus its ground. Choose the grade by environment: 304 for indoor washdown and most outdoor sites, 316L for salt air, brine and chlorides, as laid out in the 304 vs 316L grades guide. A custom layout is built from one piece with your cutouts and delivered in 10 business days from your approval of the drawing. Put the magnet scan and the one-entry-per-circuit rule on your commissioning sheet, and the arrangement holds for the life of the panel.
Catalog dimensions to compare
Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 12 × 10 × 6 in | 316L stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 20 × 16 × 8 in | 316L stainless | 20 x 16 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 24 × 20 × 8 in | 316L stainless | 24 x 20 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Is 316L stainless steel magnetic?
Not in the annealed sheet a box is made from: published measurements put 316L near 1.01 and 316LN near 1.005 relative permeability, essentially the air value of 1.0. Heavy cold working leaves 316 essentially non-responsive, while 304 picks up a real pull after the same work. Scan the wall around each cable entry with a magnet, and group that circuit's conductors through one hole anywhere it holds.
Does the NEC induced-currents rule apply to a stainless box?
By the letter it can: the rule names ferrous metal, and stainless is ferrous because iron is its base. The heating it guards against needs a ferromagnetic wall, which annealed 304 or 316L is not, at roughly 1.0 relative permeability against about 200 for carbon steel. Grouping each circuit's conductors through one entry satisfies both readings and costs nothing.
Why does a magnet stick to my stainless enclosure door edge?
Cold work. Flanging, punching and rolling turn austenite into martensite, and 304 takes a strong magnetic response after substantial cold working; weld seams can do the same, while annealed 316 stays essentially non-responsive. The flat sheared face of the wall usually holds nothing. Treat any zone that pulls the magnet like carbon steel: pass that circuit's phases, neutral and grounds together through one gland.
Do aluminum boxes have the induced-current problem?
Aluminum is non-ferrous and non-magnetic, so the NEC rule does not name it, but eddy currents still circulate in an aluminum wall. The worked example above puts 100 A at 25 mm from a wall at about 0.80 mT of flux in air, and an aluminum wall tracks that air value while carbon steel carries 200 times more. Grouping the conductors removes the heating in every metal.
What arrangement keeps a stainless box cool at the entries?
One entry per AC circuit. Pass the phases, the grounded conductor and the equipment grounding conductors through the same hole or gland, sized for the bundle, and run spare conductors with the set they belong to. A 5 percent residual imbalance then leaves about 0.25 percent of the single-conductor heating, and the same arrangement stays valid if a cold-worked or welded zone goes magnetic.
