Welding Consumables Guide Everything you need to know about flux core welding wire, flux welders, MIG welding wire composition, ER5356 aluminum wire, spool sizes, storage, and shelf life. Flux co...
READ MOREDate:Apr 10, 2026
Magnetic materials are broadly classified into two categories: hard magnetic materials and soft magnetic materials. The fundamental distinction lies in their coercivity — hard magnets resist demagnetization and retain their magnetism permanently, while soft magnetic materials magnetize and demagnetize easily with minimal energy loss. In practical engineering, soft magnetic alloys such as silicon steel, permalloy, and amorphous/nanocrystalline alloys are the backbone of transformers, inductors, motors, and sensors, precisely because they can cycle through magnetic states millions of times with very low core loss. Understanding which material to use — and why — is essential for optimizing electromagnetic device performance, efficiency, and cost.
Hard magnetic materials, also known as permanent magnets, are characterized by a high coercivity (Hc) — the resistance to demagnetization — and a large remanent magnetization (Br) after the external field is removed. Once magnetized, these materials maintain their magnetic state almost indefinitely under normal operating conditions.
The energy product (BH)max is the key figure of merit for hard magnets, representing the maximum magnetic energy that can be stored. Common hard magnetic materials include:
Hard magnetic materials are designed to resist changes in magnetization. Their microstructure — typically featuring single-domain particles or highly anisotropic crystalline structures — is engineered to pin magnetic domain walls, preventing flux reversal under moderate opposing fields.
Soft magnetic materials are defined by their low coercivity (typically below 1,000 A/m), high magnetic permeability, and low hysteresis loss. These properties allow them to respond rapidly and efficiently to changing magnetic fields, making them indispensable in AC electromagnetic devices.
The area enclosed by the B-H hysteresis loop of a soft magnetic material is very small, corresponding to very low energy dissipated as heat per magnetization cycle. For devices operating at 50 Hz or higher frequencies, these losses — referred to as core losses — accumulate rapidly, so minimizing hysteresis and eddy current losses is critical to efficiency.
Key properties used to evaluate soft magnetic materials include:
The table below summarizes the most important property differences between hard and soft magnetic materials, providing a clear reference for material selection decisions.
| Property | Hard Magnetic Materials | Soft Magnetic Materials |
|---|---|---|
| Coercivity (Hc) | High (10,000–1,000,000+ A/m) | Low (<1,000 A/m, often <10 A/m) |
| Remanence (Br) | High (0.5–1.5 T) | Low (near zero after field removal) |
| Permeability (μr) | Low (1–10) | High (200–100,000+) |
| Hysteresis Loss | Very high (large loop area) | Very low (narrow loop area) |
| Saturation Flux (Bs) | Moderate to high | High (0.5–2.4 T depending on alloy) |
| Primary Function | Permanent magnet, energy storage | Flux guide, transformer core, inductor |
| Typical Examples | NdFeB, SmCo, Alnico, Ferrite | Silicon steel, Permalloy, Amorphous alloy |
| Microstructure Goal | Pin domain walls, prevent reversal | Free domain wall motion, easy reversal |
Soft magnetic alloys represent a diverse family of engineered materials, each optimized for specific frequency ranges, flux densities, and loss requirements. The main categories are explored in detail below.
Silicon steel is by far the most widely used soft magnetic alloy in the world, accounting for the cores of virtually all power transformers and many electric motors. Adding silicon (typically 1–4.5 wt%) to iron serves two crucial purposes: it increases electrical resistivity (from ~10 μΩ·cm for pure iron to ~50–60 μΩ·cm for 3% Si steel), thereby reducing eddy current losses, and it reduces magnetocrystalline anisotropy, lowering hysteresis losses.
Grain-Oriented Electrical Steel (GOES) is produced by a controlled rolling and annealing process that aligns the [001] easy-axis grains in the rolling direction (Goss texture). This alignment results in extremely low core loss — as low as 0.8 W/kg at 1.7 T and 50 Hz for high-permeability grades — and is the standard core material for large power transformers. Non-Grain-Oriented (NGO) silicon steel, which has random grain orientation, is used in rotating machines where flux direction changes. NGO grades typically show losses of 2–5 W/kg under the same conditions but offer more isotropic behavior.
High-silicon steel (6.5% Si) offers further loss reduction and near-zero magnetostriction — beneficial for reducing audible transformer hum — but is extremely brittle, requiring special processing techniques such as chemical vapor deposition (CVD) or rapid solidification.
Nickel-iron (Ni-Fe) alloys are the premier choice when ultra-high permeability and very low coercivity are the primary design requirements. The landmark composition is 78.5% Ni – 21.5% Fe (Permalloy), which achieves maximum permeability by sitting at the zero-crossing of magnetocrystalline anisotropy constant K1. With proper heat treatment in a hydrogen atmosphere, Permalloy can achieve initial permeability (μi) of 8,000–20,000 and maximum permeability exceeding 100,000 — approximately 500 times better than low-carbon steel.
Mu-Metal (77% Ni, 15% Fe, 4% Cu, 4% Mo) is a related alloy optimized for magnetic shielding applications, offering μr up to 80,000–100,000. It is commonly used to shield sensitive electronic instruments — such as electron microscopes, photomultiplier tubes, and MRI components — from stray magnetic fields.
The 50% Ni-Fe alloys (trade names include Deltamax, Orthonol) are optimized differently: they exhibit a nearly rectangular B-H loop, making them ideal for magnetic switches, pulse transformers, and saturable reactors. Saturation flux density for the 50% Ni alloys is around 1.5 T, while 78% Ni alloys saturate at about 0.75 T.
The principal disadvantage of Ni-Fe alloys is cost: nickel prices fluctuate significantly, and the precise processing (hydrogen annealing, controlled cooling rates) adds manufacturing complexity. As a result, their use is concentrated in high-value, precision applications rather than bulk power applications.
Iron-cobalt alloys — particularly the 49% Fe – 49% Co – 2% V composition known commercially as Permendur or Hiperco — possess the highest saturation magnetization of any soft magnetic alloy, reaching Bs values of 2.35–2.45 T. This exceptional saturation flux density enables transformer and motor cores to operate at much higher flux densities than silicon steel, allowing significant reductions in device size and weight.
The aerospace and defense sectors are the primary users of Fe-Co alloys. Aircraft generators, radar power supplies, and satellite power conditioning systems benefit greatly from the weight savings enabled by Permendur cores. A transformer core operating at 2.0 T with Fe-Co alloy can be roughly 30–40% lighter than an equivalent silicon steel design limited to 1.7 T.
However, Fe-Co alloys have significant drawbacks: they are extremely expensive (cobalt is a critical mineral with volatile pricing), mechanically brittle without the vanadium addition, and exhibit higher core losses than amorphous or nanocrystalline alloys at elevated frequencies. They are also difficult to stamp and machine.
Amorphous metal alloys (metallic glasses) are produced by rapid solidification of molten alloy at cooling rates exceeding 10⁶ K/s, typically via melt spinning onto a rapidly rotating copper wheel. The resulting ribbon (~20–30 μm thick) has no crystalline grain structure — hence no grain boundaries or magnetocrystalline anisotropy — which translates to dramatically lower hysteresis losses compared to crystalline materials.
The most commercially significant amorphous alloy is Metglas 2605SA1 (Fe-based: Fe₈₀B₁₁Si₉), produced by Hitachi Metals. Its core loss at 60 Hz and 1.4 T is approximately 0.125 W/kg — roughly one-third of the best grain-oriented silicon steel (~0.35–0.45 W/kg at comparable conditions). This has made it the preferred core material for distribution transformers in energy efficiency programs. The U.S. Department of Energy's efficiency standards for distribution transformers (DOE 2016 regulations, DOE 2016-based NEMA TP-2 standards) have accelerated the adoption of amorphous core designs.
Co-based amorphous alloys (e.g., Co₇₂Fe₅B₁₅Si₈) exhibit near-zero magnetostriction and extremely high permeability (μi > 100,000), useful for sensor cores, current transformers, and magnetic flux gates. However, the high cobalt content limits their use to precision applications.
The main limitations of amorphous alloys are: brittleness (the ribbon is not ductile and cannot be stamped like silicon steel), a relatively low saturation flux density (~1.56 T for Fe-based, ~0.5–0.8 T for Co-based), and the need for specialized core assembly techniques (wound toroidal or cut-core designs).
Nanocrystalline alloys represent the state of the art in soft magnetic performance for medium-to-high frequency applications. They are produced by partially crystallizing an amorphous precursor through controlled annealing, resulting in a two-phase microstructure: ultrafine α-Fe(Si) crystallites (~10–15 nm in diameter) embedded in a residual amorphous matrix.
The benchmark nanocrystalline alloy is FINEMET (Fe₇₃.₅Si₁₃.₅B₉Nb₃Cu₁), developed by Yoshizawa et al. at Hitachi in 1988. After optimal annealing (~540°C for 1 hour), FINEMET achieves: μi ≈ 100,000, Hc ≈ 0.5 A/m, Bs ≈ 1.23 T, and core loss at 100 kHz / 0.2 T of approximately 300 mW/cm³ — dramatically better than any crystalline alloy at this frequency.
The superior soft magnetic properties of nanocrystalline alloys arise from the random anisotropy model: when the grain size is much smaller than the magnetic exchange length (~30–40 nm in Fe alloys), the effective magnetocrystalline anisotropy averages to near zero across many grains, leaving almost no impediment to domain wall motion.
A second major nanocrystalline family is Nanoperm (Fe-M-B, where M = Zr, Nb, Hf), which achieves higher Bs (~1.5–1.7 T) at the cost of slightly higher Hc. Hitachi Metals' NANOMET alloy (Fe₈₃.₃Si₄B₈P₄Cu₀.₇), announced in 2012, pushes Bs up to 1.83 T — approaching grain-oriented silicon steel levels — while retaining nanocrystalline low-loss characteristics.
Nanocrystalline cores are now widely used in: high-frequency switching power supply (SMPS) transformers, common-mode chokes, power factor correction (PFC) inductors, EV on-board chargers, and ground fault circuit interrupters (GFCIs). Their outstanding combination of permeability, low loss, and reasonable Bs makes them the first choice for applications in the 10 kHz–1 MHz frequency range.
The following table provides quantitative benchmarks for the most important soft magnetic alloy families, enabling direct performance comparison for engineering selection.
| Alloy Type | Bs (T) | Hc (A/m) | μi (initial) | Core Loss @ 50 Hz, 1.5 T (W/kg) | Optimal Frequency |
|---|---|---|---|---|---|
| Low-carbon steel | 2.15 | ~80–200 | ~200 | ~8–15 | DC, very low freq. |
| NGO Silicon Steel (3% Si) | 2.03 | ~40–80 | ~1,000 | ~3–5 | 50–400 Hz |
| GO Silicon Steel (HiB) | 2.03 | ~4–10 | ~10,000 | ~0.8–1.0 | 50–60 Hz |
| 50% Ni-Fe (Deltamax) | 1.50 | ~4–16 | ~3,000–5,000 | ~0.5–1.5 | 50 Hz–10 kHz |
| 78% Ni-Fe (Permalloy) | 0.75 | <1 | ~20,000–100,000 | <0.3 | DC–100 kHz |
| Fe-Co (Permendur) | 2.40 | ~80–160 | ~800 | ~5–10 | 50–400 Hz |
| Fe-based Amorphous (Metglas 2605SA1) | 1.56 | ~2–4 | ~5,000–10,000 | ~0.125 | 50 Hz–20 kHz |
| FINEMET (Nanocrystalline) | 1.23 | ~0.5 | ~80,000–100,000 | <0.05 | 1 kHz–1 MHz |
| Soft Ferrite (Mn-Zn) | 0.35–0.50 | ~10–50 | ~1,000–15,000 | N/A (high freq.) | 10 kHz–1 MHz |
Understanding why soft magnetic alloys behave as they do requires examining the fundamental mechanisms of magnetization at the microstructural level.
Ferromagnetic materials are divided into magnetic domains — regions of uniform spontaneous magnetization — separated by domain walls (Bloch or Néel walls). In the demagnetized state, domains are oriented to minimize total magnetostatic energy, resulting in near-zero net magnetization. When an external field is applied, domains aligned with the field grow at the expense of misaligned domains through domain wall motion, and at high fields, domain rotation completes the magnetization process to saturation.
In soft magnetic materials, domain walls must move freely with minimal energy input. Any structural feature that pins a domain wall — grain boundaries, dislocations, precipitates, non-metallic inclusions, internal stresses — increases coercivity and hysteresis loss. The entire science of soft magnetic alloy processing (purification, annealing, composition control, grain size optimization) is ultimately aimed at removing or minimizing these pinning sites.
Magnetocrystalline anisotropy (quantified by anisotropy constant K1) describes the preference of magnetization to align along certain crystallographic directions (easy axes). In iron, the [100] direction is the easy axis; in nickel, it is [111]. Large K1 values mean the magnetization resists rotation away from easy axes, requiring more field energy to complete magnetization cycles and contributing to hysteresis loss.
The most effective soft magnetic alloys exploit compositions where K1 passes through zero. In the Ni-Fe system, K1 = 0 at ~78% Ni — exactly the Permalloy composition. In Fe-Co, K1 = 0 near 30–35% Co. At these "magic" compositions, the energy barrier to domain rotation vanishes, and permeability reaches its theoretical maximum. Silicon addition to iron similarly reduces K1, though it does not reach zero before the alloy becomes too brittle at ~6.5% Si.
Magnetostriction (λs) is the change in dimensions of a material upon magnetization. Non-zero λs means that magnetization cycles create internal stresses, which in turn create anisotropy and pin domain walls — increasing coercivity and hysteresis loss. Additionally, magnetostrictive forces cause the vibration responsible for the audible hum of transformers.
The optimal condition for soft magnets is λs ≈ 0. In the Ni-Fe system, λs = 0 occurs near 81% Ni, close to but not identical to the K1 = 0 composition. In practice, alloys like Supermalloy (79% Ni, 5% Mo, balance Fe) are designed to balance both K1 ≈ 0 and λs ≈ 0, achieving the highest permeabilities measured in any material. Co-based amorphous alloys exploit a similar compositional tuning to reach near-zero λs, giving them outstanding AC properties.
When a soft magnetic core is subjected to a time-varying magnetic field, circulating currents (eddy currents) are induced within the conductive material. These currents dissipate energy as resistive (Joule) heating. The classical eddy current loss per unit volume scales as:
Pe ∝ f² × B² × d² / ρ
where f is frequency, B is peak flux density, d is material thickness, and ρ is electrical resistivity. This relationship has three major consequences for soft magnetic alloy design:
This is why power transformer laminations (~0.3 mm thick) are adequate at 50/60 Hz, while high-frequency SMPS transformer cores must use amorphous ribbon (~25 μm), nanocrystalline ribbon (~18 μm), or ferrite (insulating ceramic).
The choice between hard and soft magnetic materials — and among soft magnetic alloys — is driven entirely by function. The following outlines the dominant application areas for each major category.
The global installed base of distribution transformers represents one of the largest consumers of soft magnetic core material. In the United States alone, there are an estimated 180 million distribution transformers in service. At 50/60 Hz, the dominant choice is grain-oriented electrical steel for large power transformers and amorphous metal (Metglas) for efficiency-premium distribution transformers.
The energy savings from amorphous core distribution transformers are substantial. A typical 25 kVA distribution transformer with an amorphous core has no-load losses of approximately 15–18 W, compared to 50–70 W for a conventional silicon steel core transformer of the same rating. Given that distribution transformers are energized 24 hours a day, 365 days a year, the lifetime energy savings justify the ~15–20% higher first cost of amorphous core units.
Electric motors consume approximately 45% of global electricity generation, making core loss reduction in motor laminations one of the highest-leverage energy efficiency opportunities available. The stator and rotor cores of AC induction motors, synchronous motors, and permanent magnet motors are almost exclusively made from NGO silicon steel.
For high-efficiency (IE4, IE5 class) motors, premium NGO grades with silicon content up to 3.5% and carefully controlled grain size are specified, reducing core loss by 15–25% compared to standard grades. Thin-gauge (0.2–0.27 mm) laminations are increasingly adopted for high-speed motors (above 3,000 rpm) or variable frequency drive applications to manage the elevated harmonic content.
In aerospace electric motors, Fe-Co Permendur is used specifically for its ultra-high Bs, enabling the lightest possible motor designs. A Permendur-core motor can potentially reduce total magnetic core weight by 30–50% versus silicon steel at equivalent power output — critical in aircraft and spacecraft where every kilogram of mass carries a fuel or payload cost.
Switch-mode power supplies (SMPS) operate at 20 kHz–2 MHz, where silicon steel is completely unsuitable (eddy current losses would be enormous). The dominant core materials in this frequency range are:
High-permeability Ni-Fe alloys (Permalloy, Mu-Metal, Supermalloy) find their niche in applications requiring extreme sensitivity to low-level magnetic fields. Examples include:
Electric vehicles (EVs) represent one of the fastest-growing application areas for advanced soft magnetic alloys. Three main subsystems consume soft magnetic material:
The properties of soft magnetic alloys are extremely process-sensitive. The same alloy composition can have vastly different magnetic performance depending on thermomechanical processing history.
Annealing is the single most important processing step for soft magnetic alloys. The primary goals of annealing are to relieve internal stresses (which pin domain walls), promote grain growth (reducing grain boundary pinning), and establish the correct crystallographic texture (for GOES) or phase transformation (for nanocrystalline alloys).
For Ni-Fe permalloy, a hydrogen-atmosphere anneal at 1,100–1,200°C followed by controlled slow cooling through the ordering temperature (~600°C) is essential to achieve maximum permeability. The hydrogen atmosphere serves two purposes: it prevents oxidation and removes dissolved carbon and sulfur, both of which are potent domain wall pinners even at ppm concentration levels.
For nanocrystalline FINEMET, the annealing protocol is precise and critical: heating the as-spun amorphous ribbon to ~540°C causes nucleation and growth of α-Fe(Si) nanocrystals. Annealing temperature must be controlled within ±10°C; too low leaves the alloy partially amorphous with suboptimal properties, while too high causes excessive grain growth beyond 50 nm, rapidly increasing coercivity. Magnetic field annealing can additionally induce a uniaxial anisotropy in the ribbon plane, flattening the B-H loop for inductor applications.
Laminated cores are the standard construction method for silicon steel and Ni-Fe alloy cores operating at power frequencies. Individual laminations are coated with an electrically insulating layer (typically 1–5 μm of phosphate or oxide coating, or organic varnish) to interrupt eddy current paths. The stacking factor (the fraction of the core cross-section occupied by active magnetic material rather than insulation) is typically 0.95–0.97 for modern laminations.
Joint design in laminated cores is critical for power transformer performance. Conventional butt joints introduce large air gaps that degrade permeability and increase magnetizing current. Step-lap joint configurations — where laminations are offset by one or more steps at each joint — reduce effective gap length and are standard in modern high-efficiency power transformers, reducing no-load losses by 3–7% compared to single-step butt joints.
Soft magnetic powder cores are made by compacting alloy powder (iron, Fe-Si, Fe-Ni, Fe-Ni-Mo, or amorphous/nanocrystalline) with an insulating binder under high pressure (600–1,500 MPa), followed by a low-temperature cure or sinter. The insulating matrix between particles provides a distributed air gap — radically different from the localized air gap of a gapped ferrite core — which gives powder cores their characteristic ability to maintain high permeability under significant DC bias current without abrupt saturation.
Key powder core families include MPP (Molypermalloy Powder, 79% Ni – 17% Fe – 4% Mo), High Flux (50% Ni – 50% Fe), and Kool Mμ (Fe-Si-Al, also known as Sendust powder). MPP cores offer the lowest core loss among powder types and are used in precision inductors for audio and instrumentation. High Flux cores tolerate the highest DC bias levels, making them preferred for flyback and boost converter inductors. Kool Mμ cores offer a good cost-performance compromise for mainstream power electronics inductors.
Research in soft magnetic materials is driven by the demands of electrification — higher efficiency, higher power density, higher operating temperatures, and reduced reliance on critical minerals.
6.5% Si steel has long been recognized as an ideal composition — it has near-zero magnetostriction, lower core loss than 3% Si steel, and higher resistivity — but its extreme brittleness prevented practical manufacturing. JFE Steel's CVD process applies Si vapor to pre-rolled 3% Si steel, diffusing Si content up to 6.5% in the surface layers, and has been in commercial production since the 1990s. A similar approach using rapid solidification (melt spinning followed by hot rolling) has been developed by various research groups. High-silicon steel at 6.5% Si has core loss approximately 30–40% lower than 3% Si steel at 400 Hz, making it attractive for aircraft and high-speed drive applications.
A major research thrust is developing nanocrystalline alloys that combine high saturation flux density (>1.7 T) with low core loss — essentially bridging the gap between silicon steel (high Bs, moderate loss) and FINEMET (low Bs, ultra-low loss). Hitachi's NANOMET alloy (Fe₈₃.₃Si₄B₈P₄Cu₀.₇) achieves Bs = 1.83 T with nanocrystalline structure and low loss, representing a significant advance. Research groups in Germany, China, and Japan are actively pursuing alloys in the Fe-Si-B-P-Cu system with Bs approaching 2.0 T.
Soft Magnetic Composites (SMCs) are iron powder particles coated with an inorganic insulating layer and compacted into three-dimensional near-net shapes. Unlike laminated silicon steel, SMCs can be pressed into complex geometries (e.g., claw-pole motor stators, axial flux motor cores) that would be impossible or prohibitively expensive to laminate. Their isotropic properties also make them ideal for 3D flux paths in transverse flux and claw-pole machines. Current SMC technology has higher core loss than silicon steel at 50 Hz, but this disadvantage shrinks at frequencies above 1 kHz and is outweighed by the manufacturing freedom for complex geometries.
3D printing of soft magnetic components is an active research area, particularly for prototype and specialty motor cores with optimized topology. Selective laser melting (SLM) of Fe-Si powders has been demonstrated for complex motor stator geometries, though the high residual stress and microstructural damage from the laser process typically result in higher coercivity than conventionally processed material. Post-printing stress relief annealing is essential. The capability to 3D print topologically optimized magnetic circuits — minimizing material usage while maintaining or improving flux paths — could be transformative for high-performance motor design.
Choosing between hard and soft magnetic materials — and selecting among the available soft magnetic alloys — requires a systematic evaluation of the device's operating requirements. The following decision framework captures the most important considerations:
The growing emphasis on energy efficiency is reshaping the soft magnetic materials market. Several regulatory and policy drivers are accelerating the transition from standard silicon steel to advanced amorphous and nanocrystalline alloys:
The fundamental division between hard and soft magnetic materials reflects two opposing engineering needs: permanence versus responsiveness. Hard magnets store magnetic energy and resist change; soft magnets conduct and transform magnetic flux with minimal loss.
Within the soft magnetic family, the hierarchy is clear:
As global electrification accelerates — driven by EV adoption, renewable energy expansion, and grid modernization — the demand for advanced soft magnetic alloys will grow substantially. The combination of tightening efficiency regulations and falling prices for advanced processing methods suggests that amorphous and nanocrystalline alloys will progressively displace conventional silicon steel in an expanding range of applications, reducing electromagnetic energy losses at a global scale.
Recommended Articles
Content
Welding Consumables Guide Everything you need to know about flux core welding wire, flux welders, MIG welding wire composition, ER5356 aluminum wire, spool sizes, storage, and shelf life. Flux co...
READ MOREWelding Wire Guide A practical, spec-driven comparison of MIG welding wire and flux-core welding wire, covering material choice, wire diameter, welding cable setup, and when each process actually ...
READ MORE90/10 vs 70/30 Copper-Nickel: The Direct Answer Choose 90/10 copper-nickel for general seawater piping, condenser tubing, and budget-sensitive marine projects with moderate flow rates (under 3.5 m/s)...
READ MORE