Can You Weld a Magnet? The Honest Answer

Short answer: you can physically weld a magnet, but welding will almost certainly destroy its magnetism, and on most magnet materials it will also crack or chip the magnet while it is in your hands. The heat required to fuse steel runs thousands of degrees. Every permanent magnet material loses its magnetic properties at temperatures far below that.

There is a second, equally damaging effect that most guides never mention. It is not only that welding destroys the magnet. The magnet’s field also disrupts the arc, which is how you get lack of fusion, undercut and porosity in the weld you were trying to make. So a magnet near a weld degrades the joint from both sides.

Before we get into the physics, one clarification worth making, because it trips up a lot of searching. The phrase “welding magnets” means two completely different things:

• Welding a magnet — attaching a permanent magnet to a metal part using a welding process. This is what the rest of this article is about.
• Welding magnets — the tool category. Magnetic ground clamps, magnetic squares, magnetic angle jigs and pipe welding clamps. If you are shopping for gear rather than trying to preserve a magnet, skip to the section on welding magnet tools further down. Those products are built to survive welding, and the design tells you exactly why your magnet will not.

• Why heat destroys magnetism, explained properly
• The actual temperature numbers for every common magnet material
• Three kinds of magnet loss, and which ones you can undo
• Why magnets crack before they lose their pull
• What actually works: welding the case, brazing, bonding, re-magnetising
• How magnets cause arc blow in the weld you are making
• Welding magnet tools, and the one way they can hurt you
• Cobalt, fumes and safety when welding SmCo magnets
• Answers to the questions people actually ask

Why Heat Destroys Magnetism

First, a correction to something you will read almost everywhere online, including in most versions of this article. Magnets are not magnetic “because atoms have unpaired electrons that spin in the same direction.” That is a reasonable description of how some materials behave, but it leaves out the part that actually matters, and it misdescribes why iron behaves the way it does.

Here is a more useful model. A permanent magnet is a large block of microscopic regions called domains. Inside each domain, the magnetic moments of neighbouring atoms are already aligned with each other. A macroscopic magnet is simply a huge number of those domains all pointing the same way. Magnetise the magnet and you have lined the domains up. Let them get knocked out of line and the magnet’s external field collapses, even though every individual atom is still magnetic.

For iron, nickel and cobalt, the alignment mechanism is an exchange interaction, a quantum effect between neighbouring atoms that costs real energy to overcome. Heat is exactly what supplies that energy. Raise the temperature and the thermal agitation starts beating against the alignment. Below a certain point the domains wobble and get disordered, which costs you a few percent of your pull and recovers on cooling. Above a certain point the crystal structure itself reorganises into a non-magnetic phase, and the ability to be a magnet at all is gone.

That second threshold is the Curie temperature, and it is the number that matters for this question. Different materials lose their ferro-magnetism at wildly different temperatures, which is why a rare-earth magnet is destroyed by a hot cup of tea while an alnico magnet in a welding clamp shrugs off 450°C.

The Temperature Numbers That Answer the Question

Here is the data most articles omit, and it settles the debate in one table. Two numbers matter per material: the maximum working temperature above which you start losing magnetism permanently, and the Curie temperature above which the material cannot be a magnet at all.

Now compare that table against what a welding arc actually does. An arc temperature is roughly 6,000 to 10,000°C at the core, and even the heat-affected zone well away from the arc is far above 500°C. The filler metal in a steel weld melts somewhere around 1,450°C.

So the answer is not close. Every single common permanent magnet material has a Curie temperature an order of magnitude below welding heat. There is no grade of neodymium, no brand of rare earth, and no clever electrode that changes this. A welding arc will take a magnet well past the point where it can hold a magnetic domain, let alone keep them aligned.

One detail worth knowing: the highest-strength neodymium grades are actually the worst for temperature. N52, N50 and N50M are rated to only 60°C, below the standard N-series 80°C rating, because the extra coercivity comes with a lower working point. Buying the strongest magnet off the shelf is buying the one least able to survive being warm.

Three Kinds of Magnet Loss, Not Just One

“The magnet got demagnetised” covers three genuinely different situations, and only two of them are recoverable. This distinction is where most of the confusion in this topic lives.

1. Reversible loss

Magnet strength falls as temperature rises, along a predictable coefficient, and comes straight back when it cools. Standard neodymium loses roughly 0.12% of its output for every degree Celsius. A magnet sitting 20°C above ambient has given up about 2.4% of its strength, and gets it all back. Nothing is damaged. Nothing needs fixing.

2. Irreversible but recoverable loss

Exceed the maximum working temperature and the internal operating point is driven down past the “knee” of the demagnetisation curve. Some domains flip permanently out of alignment. When the magnet cools, they stay flipped, so the flux does not return.

The useful part: the material keeps its intrinsic coercivity. It is still capable of being a magnet, and re-magnetising it can restore its strength. From the outside the magnet simply looks weaker, because its remanence is lower. Once this has happened the magnet is also considered thermally stabilised, meaning repeating that temperature will not lose anything more.

3. Irreversible and irrecoverable loss

Go far enough above the Curie point, or long enough at high temperature, and you get a metallurgical phase change. The material is permanently altered. It may oxidise, grain-coarsen, or simply be a different microstructure when it cools. Re-magnetisation will not bring it back. This is the state a heated neodymium magnet ends up in, and it is the reason a cooked magnet is scrap.

Cracking and chipping land in this last category too. If a magnet fractures, the pieces may each still show local magnetism, but the assembly no longer works as one magnet, and no amount of magnetising fixes a broken part.

The Problem That Bites Before the Heat Does

If you ignore magnetism entirely and only ask “can I fuse this thing to a steel bracket,” you are still in trouble, because permanent magnet materials are mechanically hostile. A neodymium magnet is a hard, dense, brittle ceramic-like alloy with almost no ductility.

• Density: about 7.5 g/cm³, similar to steel, so it does not float away on the slag.
• Vickers hardness: around 570 HV, far harder than most weld metals. A softer filler will smear rather than wet the surface.
• Compression strength: around 780 MPa, which sounds reassuring.
• Tensile strength: roughly 78 MPa. That is the number that matters. A magnet has roughly a tenth of the tensile strength of mild steel and essentially no capacity to stretch.

Translation: any tensile stress, any impact, any sudden thermal gradient will fracture it. A weld bead contracts as it cools. That shrinkage is a load the magnet cannot accommodate, and it cracks. Ferrite is worse still, being a true ceramic. Samarium cobalt is described by its own trade as prone to chipping. Ferrite and NdFeB fragments are also sharp and hazardous, and neodymium’s brittleness makes “it exploded into pieces” a genuinely common workshop outcome when clamping or striking one.

This is why the honest answer is that welding a magnet is not a heat problem with a clever workaround. It is a heat problem and a fracture problem at the same time.

What Actually Works Instead

There are several approaches that genuinely solve this, and they are ordered roughly by how good an engineer thinks they are.

Weld the case, not the magnet

Most real magnet assemblies are already surrounded by steel. Assemblies use a soft-iron yoke or return path that closes the magnetic circuit and concentrates the field where it is needed. Potting protects the magnet. The correct move is to weld the yoke or the casing, leaving the magnet itself untouched and cold enough.

This is how industrial magnetic chucks, sensor housings, lifting devices and motor fixtures are all attached. If you have a magnet in a steel can, weld the can. Heat conduction still reaches the magnet, so keep your heat input low, use several small passes rather than one big one, and clamp your total heat cycle well under the magnet’s working temperature.

Use alnico when heat is unavoidable

This is the part that makes the “you cannot weld near magnets” argument fall apart, because the welding industry already solved it. Alnico magnets are used as welding fixtures, specifically because they are the only common permanent magnet material that tolerates heat near an arc.

Real products exist: pipe welding angles with an alnico core rated to 450°C, and alnico magnetic holders with M3 threads rated to 450°C holding 200 to 400 grams. One documented advantage of these alnico holders is that they can be easily demagnetised and re-magnetised, which matters for shipping, since they ship with a removable steel yoke closing the magnetic circuit to protect them from demagnetising in transit.

Alnico does have one weakness: low coercivity. It holds flux better than anything else but demagnetises more easily from shock and opposing fields. The trade is deliberate. For a fixture that lives near a hot weld, tolerating 450°C matters more than resisting a bump. Its reversible temperature coefficient of about −0.02% per °C is roughly six times gentler than standard neodymium.

Weld the steel first, magnetise last

This is the design answer, and it eliminates the problem rather than managing it. Build and weld the entire steel assembly with all brackets, bosses and mounts in place, complete every heat treatment and stress relief, finish the machining and coating, and only then fit or magnetise the magnet.

Brackets that later need to be welded should be bolted or riveted so the magnet can be removed before any hot work happens. A magnet on a removable threaded boss, or bonded into a pocket, means future repair work never has to cook it. Cost more at design stage, free forever afterwards.

Mechanical fastening

Bolts, rivets, screws, threaded inserts, press fits. These preserve the magnet completely because no heat reaches it. If the joint does not need to be hermetic or high strength, this is simply the right answer, and it is reversible, which is a bonus for maintenance.

Adhesive bonding

Structural epoxy bonds magnets to metal well and keeps heat out of the magnet entirely. Two-part epoxies handle metal-to-metal shear loads fine. Watch for two things: keep the bond line thin, because thick adhesive layers creep under sustained load and a magnet that creeps is a magnet that has moved, and prefer a bond that fails before the magnet does. A magnet is a consumable part. Design it to be replaceable.

Brazing

Brazing runs well below melting temperatures, typically 450 to 700°C for common silver brazes, so it is far kinder than welding. But be honest about the limit: 450°C is still above the working temperature of every neodymium grade, and above the Curie point of nothing useful. Brazing works for ferrite and alnico. For NdFeB it trades a total loss for a partial permanent loss. A brazing alloy with a low liquidus, in the 400°C range, is the best case.

Soldering

A typical iron sits at 330 to 400°C, which is above the 80°C working limit of standard neodymium by a wide margin. So soldering to a neodymium magnet causes irreversible loss. Soldering to its steel casing or yoke is completely fine and is the normal way this is done. Soldering directly to ferrite is also fine, since ferrite’s 250°C limit comfortably clears a low-temperature solder.

Re-magnetising afterwards

Technically possible, rarely practical. Remagnetising requires a field strong enough to overcome the material’s coercivity, applied in the correct direction, after the magnet has cooled. Specialist pulsed coils or superconducting magnets are used in industry. The important caveat is that a strong permanent magnet cannot do the job. Even neodymium-to-neodymium contact will not reliably re-magnetise a weakened magnet; you need controlled field parameters, not just a bigger magnet stuck on the end.

There is one more theoretical route: heat the magnet above its Curie point and cool it inside a strong field. The domains re-form aligned to whatever field is present. This is standard manufacturing practice, and it is exactly why a magnet that simply cools with no field applied stays dead. In practice, if the magnet survived the heat without cracking and you have access to a magnetiser, you can recover it. For most jobs replacing the magnet is cheaper.

How a Magnet Wrecks Your Weld, Not Just Itself

This is the effect that catches experienced fabricators, and it is the reverse of the question. A welding arc creates its own magnetic field around the electrode. If an external field disturbs that balance, the arc deflects away from the shortest path between electrode and work. That is magnetic arc blow.

When the arc blows, you get a specific and recognisable set of defects: lack of root penetration, lack of fusion, undercut, and pores. Lincoln Electric notes that arc blow also produces beads of irregular width, crooked beads, wavy beads and excessive spatter. Small amounts are sometimes even useful for shaping bead profile, but once it is contributing to defects it must be controlled.

Magnetic flux density in the field around a workpiece is the diagnostic. Steel industry guidance puts the threshold for meaningful arc disturbance at roughly 40 Gauss, and notes that some materials react at 20 Gauss. Highly permeable materials such as 9% nickel steel are notoriously sensitive. In pipe welding the effect is amplified badly: a field measured in open air at a few tens of Gauss can become concentrated in the root gap and reach 1,000 Gauss.

If a magnet is anywhere near the arc, assume you have a second source of magnetism feeding this. The fixes that actually work:

• Use AC. The field direction reverses continuously, and induced eddy currents oppose external fields. Less susceptible to arc blow than DC.
• Move the magnet. Simplest option available and usually sufficient.
• Move the workpiece earth point. The return current through the work creates its own field. For back blow, position the connection away from the direction of welding; for forward blow, towards it.
• Incline the torch against the direction of the blow.
• Degauss the steel. An AC field slowly reduced to zero. Note that line-frequency degaussers have poor skin penetration, so they may clean the surface layer while leaving magnetism underneath.
• Check before you start with a cheap hand-held gauss meter. This is the fastest way to know whether you have a problem.

Welding Magnet Tools

If you arrived here looking for gear rather than for a way to weld a magnet, this is your section. These are the products industry actually uses, and they explain the physics nicely.

Magnetic ground clamps. These replace bolted ground clamps and, importantly, remove the need to tack on temporary ground tabs before a large sheet or pipe job. Switchable designs hold 90 to 200 pounds of breakaway force and mount to flat or round steel. The design detail worth noting is that the magnet must be able to turn completely off. An always-on magnet collects swarf and chips; a switchable one releases them, which is both a cleanliness win and a hand-safety win. Modern fin designs drop the temperature at the clamp by roughly 27%.

Magnetic squares and angle jigs. For holding pieces at 45 to 135 degrees, locating pipe, and positioning sheet. Because there is nothing to clamp, they stay clear of the joint so the torch can reach and see the work. Multi-angle products commonly cover 45, 75, 90, 105, 120 and 135 degrees with detents.

Butt weld clamps. Panel clamps using rare-earth magnets with swivelling feet to match curved surfaces, typically holding a 0.03 inch gap and working up to around 3/8 inch material. These are for fit-up, not for welding through.

Heat-resistant clamps. As described above, alnico-cored pipe welding angles rated to 450°C. This is the product category that proves the point about material selection.

The one way these tools can bite you

Here is a genuine, specific warning. Steel industry guidance explicitly names magnets as a cause of the problem they then have to solve. Sources that enhance the magnetism of steel include magnets used in jacking equipment, or storage in a place with high levels of magnetic flux density.

So a magnetic clamp sitting clamped on a joint does two things: it holds the work, and it magnetises it. If your clamp is positioned near the root pass or on heavy plate where the magnetic path is already concentrated, you have created the exact conditions for arc blow. Fixtures also pick up magnetism over time through repeated welding. The practical rules: keep strong magnetic clamps off the root and away from where you are about to lay the root pass, check with a gauss meter if the bead looks odd, and remove the clamp before welding if it is not needed for access.

Safety: Welding Samarium Cobalt Magnets

If your magnet is SmCo rather than NdFeB or ferrite, the hazard conversation changes. SmCo is cobalt-based, and cobalt is a genuinely dangerous material to weld or grind.

Chronic cobalt exposure is associated with interstitial lung disease, respiratory sensitisation, allergic dermatitis, and cardiomyopathy. When cobalt appears combined with tungsten carbide it is specifically associated with hard metal lung disease, and IARC classifies cobalt metal as Group 2B, possibly carcinogenic, moving to Group 2A when exposure involves tungsten carbide. Fatalities have historically been reported from cobalt exposures as low as 1 to 2 mg/m³. In one documented series, among 12 workers in tungsten carbide tool manufacture who developed interstitial lung disease, eight died.

The exposure limits are low. OSHA’s current permissible limit for cobalt and inorganic compounds is 0.02 mg/m³ as an 8-hour time-weighted average for inhalable particulate. The ACGIH threshold limit value is tighter still at 0.005 mg/m³, with cobalt designated a respiratory sensitiser. NIOSH documented a magnetic manufacturing facility where a worker developed work-related asthma from cobalt exposure while wearing a paper mask, and where 19% of co-workers reported breathing symptoms attributable to work at that plant.

SmCo also presents a physical hazard, because it is brittle and chips, and those chips are sharp. Add to that the fact that cobalt dust can form explosive mixtures in air and that powdered cobalt is pyrophoric, burning brilliantly at room temperature. Water is explicitly not the extinguishing medium for cobalt fires.

Practical rule: do not arc-weld samarium cobalt magnets. Use local exhaust ventilation at minimum, wear suitable respiratory protection, handle with gloves so chips do not end up in your hands, and treat any grinding or cutting of SmCo as a serious exposure requiring proper containment and a written assessment of the exposure limits above.

Choosing a Lower-Temperature Process

If you must do something hot near a magnet, the process choice is your main lever, because heat input is what you are trying to minimise.

• Brazing or soldering to the case — lowest heat, preferred where the design allows.
• TIG — the most controllable heat input of the arc processes. You can make small passes, control where the puddle sits, and stop the moment the joint is made.
• Spot or resistance welding — very local, very short heat cycles, and the right answer for thin sheet assemblies.
• Stud welding — for attaching fasteners rather than continuous joints.
• Induction heating — targeted, repeatable, and useful if you need to heat a specific area without involving the magnet at all.
• MIG — fast and productive but higher heat input per pass than TIG. Also relevant for flux-cored options, since flux-cored self-shielded wires are the standard answer for dirty or outdoor field work. Our guide to what dual-shield welding is covers those consumable choices.
• Stick — a heavy thermal footprint. If your priority is protecting a nearby magnet, this is the worst option.

Whatever you pick, the discipline is the same: smallest possible bead, shortest possible arc length, allow the part to cool, and track the total heat cycle against the magnet’s maximum working temperature rather than the temperature of the arc. If you need guidance on holding heat down on thin material, our walkthrough of welding thin metal with an arc welder is worth a read.

Myths Worth Dropping

“Magnets work because unpaired electrons all spin the same way.” Not quite, and not usefully so. Domains are the operative concept, and exchange interaction is the reason iron holds on to its ordering. “All the atoms spinning together” does not explain why a demagnetised magnet’s atoms are still individually magnetic.

“Magnets cannot be welded under any circumstances.” Too absolute. Alnico-cored welding clamps rated to 450°C are a commercial product category. What is true is that you cannot weld a neodymium or ferrite magnet and keep its magnetism, and that welding any magnet risks cracking it.

“All magnets are ruined at the same temperature.” Nothing like it. Maximum working temperature ranges from 80°C for standard neodymium to around 500–540°C for alnico. SmCo sits in between with a much gentler temperature coefficient. The right question is never “will it survive” but “which material, and what is its limit.”

“A welding magnet will protect your magnet.” A magnetic clamp does not shield anything. It holds work, and it magnetises the work. Near the root pass it can create the arc blow you were trying to avoid.

“Magnets are made from iron, cobalt and nickel.” Those are the ferromagnetic elements, and iron, nickel and cobalt are soft magnetic materials not normally used as strong permanent magnets. Practical permanent magnets are engineered alloys: NdFeB, SmCo, alnico, or ferrite.

Frequently Asked Questions

Can I weld a magnet to a steel bracket?

Not if you need the magnet to keep working. The arc destroys the magnetism, and the shrinkage cracks the magnet. Weld a steel case or yoke if the assembly has one, or redesign so the magnet is fitted after all welding is finished.

What temperature destroys a neodymium magnet?

Standard N-grade neodymium starts losing magnetism permanently above 80°C, and 60°C for N50, N52 and N50M. It is fully demagnetised at the Curie temperature of roughly 310°C. Higher grades reach 150 to 230°C working limits, still far below the Curie point and far below welding heat.

Can a demagnetised magnet be re-magnetised?

Often yes. If the magnet has not been heated past its Curie point, has not cracked, and has not corroded, its crystalline structure and coercivity survive, and a sufficiently strong field in the right direction restores it. Remagnetising needs pulsed coils or superconducting magnets; another permanent magnet will not reliably do it. In most workshop situations replacing the magnet costs less than the magnetising service.

Can you solder to a magnet?

Yes, if you solder to the metal casing or yoke rather than the magnetic material. Soldering directly to ferrite is also reasonable. Direct soldering to neodymium is a bad idea because iron temperature exceeds its 80°C limit, and you get irreversible loss. Use a low-temperature solder to keep the margin as wide as possible.

Can I use a magnet as a welding ground?

Not as an earth connection. A magnet is not an electrical conductor and is not rated to carry welding current. Use a proper ground clamp sized for the process and amperage. That said, a purpose-built magnetic ground clamp is a legitimate tool, and a good one; it just makes contact through a designed steel path rather than using a raw magnet as a conductor. Be aware of the arc blow issue described above when positioning it.

Can you glue a cracked magnet back together?

Structurally, yes, with a good two-part epoxy and proper surface prep. Magnetically, expect a partial recovery at best. Fractured pieces do retain local magnetism, so stacking them with their poles correctly aligned often restores a useful fraction of the original pull, commonly less than the original. Because a crack also weakens the joint and a magnet under clamping load sees real shear, glue is a temporary repair. If the assembly is safety critical, replace it.

Does a magnet near the weld damage the weld?

Yes, through magnetic arc blow. A strong external field disturbs the arc’s own field and deflects the arc, producing lack of fusion, lack of penetration, undercut and porosity. Check with a gauss meter, move the magnet, switch to AC if you can, or reposition your earth point.

Are magnetic welding clamps safe for my welds?

They are widely used and generally an improvement on tacking temporary ground tabs. The one caution is that they magnetise the steel they clamp, which can contribute to arc blow near the root pass on thick or high-permeability material. Keep them clear of the root where access allows, and keep them switched off when not in use so swarf does not build up.

Which magnet material can survive the most heat?

Alnico, with a maximum working temperature around 500 to 540°C and a Curie point of roughly 750 to 860°C. That is why alnico is the standard choice for welding clamps and fixtures that live near an arc. Samarium cobalt 2:17 is next best among the rare-earth grades, usable to around 250 to 350°C depending on grade, with a Curie point around 800°C.

Why does welding a magnet sometimes crack it?

Tensile strength, not heat. Neodymium has roughly 78 MPa tensile strength against about 350 to 400 MPa for mild steel, and effectively zero ductility. A weld bead contracts significantly as it cools, and the magnet cannot stretch to accommodate it. Combined with high hardness at the surface and the thermal shock of arc-on-cold-metal, the magnet fractures. Ferrite is worse, being a true ceramic.

Conclusion

Can you weld a magnet? You can put an arc to it. What you cannot do is keep a magnet. The arithmetic is not close: an arc runs at thousands of degrees, while standard neodymium gives up its magnetism permanently at 80°C and loses the ability to be a magnet at all around 310°C. Alnico, the best performer, manages around 500 to 540°C and still cannot touch welding heat. And even if the magnetism somehow survived, the magnet’s tensile strength of roughly 78 MPa means the weld’s own shrinkage cracks it.

The productive way to think about this is that the design problem is solvable and the welding problem is not. Weld the steel case or yoke rather than the magnet. Fit the magnet after all hot work is finished. Use bolts, adhesive, or a low-temperature braze where the design allows. Specify alnico when the fixture must live near an arc. And when a magnet must sit near a weld for any reason, remember that it is also acting on your weld through arc blow, so check the field with a gauss meter before you blame your technique.

And keep an eye on the right half of the phrase. If you wanted a welding magnet rather than a magnet you could weld, you were probably after a magnetic ground clamp or a magnetic square. Those exist precisely because the alternative, tacking tabs and clamping near your own work, is worse.

Sources and Further Reading

• Eclipse Magnetics — NdFeB standard grade datasheet (grade suffixes, working temperatures, mechanical properties)
• Radial Magnetics — Magnet temperature derating by material (max operating and Curie temperatures for NdFeB, SmCo, ferrite, alnico)
• SDM Magnetics — Maximum operating temperature (definitional distinction between working temperature and Curie point)
• TME — Can a magnet be recharged if it has lost strength?
• Duramax — Top causes of demagnetization
• Lincoln Electric — Prevent arc blow
• SSAB TechSupport #46 — Magnetic arc blow (residual magnetism, gauss thresholds, fixtures and magnets as sources)
• OSHA — Cobalt metal, dust and fume (exposure limits)
• NIOSH — Prevent work-related asthma and hard metal lung disease from cobalt exposure
• OSHA — Cobalt dust health hazards
• Magswitch — Welding and fabrication magnetic fixturing (magnetic ground clamps and workholding)
• Magna-C — Heat-resistant alnico pipe welding angle, 450°C