A February 2026 construction-materials research synthesis (Patsnap Eureka) frames the trade-off explicitly: ferrite, rare-earth, and magnetic composites are the three families crossing into structural and energy-management use, with cost and energy-efficiency balance points now the gating design decision [S5].
For site engineers, this means the question is no longer "do we need a magnetic component" but "which family, in which geometry, at which operating point." The dominant families in current construction deployment are ferrites, nanocrystalline ribbons, and rare-earth permanent magnets, with the choice driven by frequency band, mechanical loading, and lifecycle cost rather than raw magnetic performance [S1][S5].
Three families doing real work on a jobsite
Construction sites now see four distinct material classes in play: amorphous iron, nanocrystalline, ferrite, and iron-nickel alloys, each characterized against common-mode choke design where dimensional resonance, eddy current loss, and core geometry dominate the design [S1]. The same classification maps directly onto construction shielding and energy-harvesting subsystems, where ferrite handles the bulk low-frequency shielding in building services, and nanocrystalline ribbons are reserved for higher-frequency power-quality filters inside elevator drives and HVAC variable-frequency drives (VFDs) [S1][S4].
For permanent-magnet functions (latches, vibration harvesters, magnetic bearing pre-loads in construction machinery and equipment), the relevant hard-magnetic families are NdFeB sintered grades and SrO-6Fe2O3 strontium ferrite; the latter is identified in MADM selection work as the optimal hard magnetic material under conflicting attribute weights of coercive field Hc, intrinsic coercivity Hic, energy product (BH)max, remanence Br, and maximum operating temperature Tmax [S2].
Selection criteria: permeability, saturation, loss, cost
For any inductive construction subsystem, the four evaluation parameters are relative permeability µr, saturation flux density Bsat, volumetric core loss Pv,core, and cost per kilogram, with the rule that "all else equal" the designer wants higher µr, higher Bsat, and lower Pv,core, but those three never align in a single material [S4]. In practical construction power-quality filter design, nanocrystalline ribbons deliver high µr (typically 10,000 to 80,000) but with strict limits on Bsat, while power ferrites offer Bsat near 0.4-0.5 T at much lower µr (1,500-5,000), and the geometry of the core window then drives whether the inductor physically fits the equipment cabinet [S1][S4].
Cost and lifecycle weight enter as the second decision axis.
For the soft-magnetic ranking the MADM analysis ranks attribute priorities Hc > µmax/µ0 > resistivity ρ > Br > Bs > Tc, and identifies the 79Ni-15Fe-5Mo-0.5Mn alloy (Supermalloy / Magnifer 7904) as optimal in the studied pool; in practice, Supermalloy is reserved for low-signal magnetic shielding in sensitive instrument rooms rather than structural building elements, where cost rules it out at scale [S2].
Who this is for, and who should not use it
Magnetic subsystems in construction are appropriate for four use cases: electromagnetic-interference (EMI) shielding of plant rooms and substation enclosures, power-quality filters on VFD-driven HVAC and elevator systems, vibration-energy harvesting embedded in floors or magnetic sensor networks for structural health monitoring, and magnetic bearings or couplings in select magnetic drive pump skids in water-treatment plants [S3][S5].
They are NOT appropriate as a substitute for structural steel, concrete reinforcement, or load-bearing members; the lifecycle cost analysis in the 2026 construction report is explicit that magnetic materials serve functional, not structural, roles, and that the ROI threshold is reached only when the magnetic function replaces a discrete electrical or electromechanical subsystem, not when it is added on top of conventional structure [S5].
Option comparison across the decision criteria
For construction-grade magnetic-material selection the four most common options stack up against four decision criteria as follows. MnZn power ferrite wins on cost and availability, Bsat near 0.4-0.5 T, µr roughly 1,500-5,000, used in bulk for low-frequency shielding and line-filter inductors. Nanocrystalline ribbon wins on permeability and size, µr in the 10,000-80,000 range, Bsat around 1.0-1.2 T, but at a per-kg cost 5-10x ferrite, justified in compact VFD common-mode chokes. Amorphous iron (Metglas-type) sits between the two on cost and on Bsat (around 1.4-1.6 T), with µr typically 5,000-10,000, and is used in distribution-transformer cores when 50/60 Hz losses dominate the bill. Iron-nickel alloys (50Ni-50Fe, 79Ni-Fe-Mo Supermalloy) win on temperature stability and very high µr (up to 100,000 for Supermalloy) but lose on cost and on saturation (Bs around 0.8-1.0 T), so they are restricted to instrument shielding and precision current-transformer duties [S1][S2][S4].
The cross-reference for adjacent selection problems is well covered: a spec-first map of magnetic material selection for electronics covers the same MnZn-versus-nanocrystalline trade at PCB scale, and aerospace magnetic material selection covers the higher-temperature end of the same alloy space.
Failure modes and real constraints on site
Three failure modes dominate construction deployment. First, dimensional resonance in nanocrystalline cores above 1 MHz turns the core from an inductive element into a radiating antenna, and the IEEE ECCE 2017 analysis shows that this resonance is geometry-dependent, not material-dependent, so two cores of the same alloy but different winding aspect ratios can behave differently at the same frequency [S1]. Second, eddy-current loss in laminated cores scales with lamination thickness squared, which is why amorphous and nanocrystalline ribbons ship in 18-25 µm thicknesses, and any field bending or stamping that damages lamination insulation degrades Pv,core disproportionately [S1][S4]. Third, Curie-temperature derating: MnZn ferrite loses usable µr above roughly 200-250 °C, while nanocrystalline stays useful to 500-600 °C, and site-installed cores near boiler flues or in unventilated plant rooms have been measured above the ferrite derating band during summer peak load [S2][S4].
Standards, sourcing, and what to put on the purchase order
Procurement specifications should reference the underlying material standard (IEC 60404 series for magnetic materials, ASTM A773 for DC magnetic properties) and the application standard where one exists: for building EMI shielding, EN 50147-1 governs honeycomb-shielded-room performance; for VFD line filters, the relevant product standard is IEC 61800-3 on adjustable-speed drives; and for power-quality inductor cores, the applicable measurement method is IEC 62044 for toroidal cores [S4][S5].
On the purchase order the four values to pin are: (1) minimum µr at 10 kHz and 100 kHz, (2) minimum Bsat at 25 °C and at the maximum site ambient (typically 40 °C or 55 °C for plant rooms), (3) maximum Pv,core at the design frequency and 200 mT for 50/60 Hz duty or at the design ripple for SMPS duty, and (4) lamination thickness and insulation class for any tape-wound core. A supplier that will not commit to all four on the certificate of conformance is one that should be de-listed from the construction-materials magnetic material qualified-vendor list [S1][S4][S5].
For construction tooling sites where magnetic material selection runs alongside hand-tool and consumable procurement, the selection logic shares more with the construction tools buy than with the electronics buy, and the lowest-cost ferrite grade is the correct default unless a documented EMI compliance or harmonics-mitigation target forces the move up to nanocrystalline.
Trackable next signals: the 2026 Patsnap report flags standardized test protocols for magnetic construction materials as still under development, so any release of an ASTM or IEC working-group draft on magnetic-material lifecycle testing in buildings between now and mid-2026 will be the next decision-altering data point for specifiers [S5].