I. Not using rare earths does not mean not using magnetic steel
The drive motor does not require rare earth materials; it still employs a permanent-magnet configuration and a rotor with magnets—only the neodymium iron boron has been replaced with ferrite.
Left: Neodymium–iron–boron; Right: Ferrite; Note: Image generated with AI assistance
What is ferrite? It’s the kind of material used for refrigerator magnets, speaker magnets, and similar items. It’s inexpensive; the raw materials are everywhere, so there’s no need to scramble for them.
However, its magnetic energy product is only one-seventh to one-tenth that of neodymium-iron-boron magnets. This is a hard constraint determined by the material itself and cannot be changed.
How can the gap in the drive motor be closed? Public disassembly and analysis identify three key measures: redesigning the magnetic circuit of the rotor core to extract magnetic reluctance torque; arranging magnets in a Halbach array to concentrate the magnetic field on the working side; and using a hairpin stator to raise the copper fill factor to 70%, combined with robust temperature control to keep temperatures under control.
When these three measures are combined, performance is restored.
But pay attention: none of these three movements bypasses the magnet itself.
Ferrite magnets should still be mounted on the rotor. The essence of a Halbach array lies in arranging magnets in a specific orientation; the more precise the arrangement, the higher the requirement for positional accuracy.
In other words, the magnet has not disappeared; it has simply been replaced with a cheaper material. Once the magnet exists, the connection between the magnet and the rotor remains intact.
For the vast majority of motors, the answer to this question is a layer of adhesive 25 to 100 micrometers thick.
II. The bonding of ferrite is actually more troublesome
At this point, there is something that defies common sense.
Many people believe that switching from expensive neodymium–iron–boron magnets to cheaper ferrite magnets reduces costs, which should make the manufacturing process easier.
just the opposite.
Let us begin with surface energy.
Publicly available data provide a comparison: the surface energy of ferrite is approximately 40–50 mN/m, while that of NdFeB is 28–35 mN/m.
Theoretically, a higher surface energy makes a material more prone to wetting, which is good news. However, ferrite is a ceramic material with a denser and more inert surface; in practical bonding applications, the surface treatment therefore plays a more critical role. Moreover, ferrite is generally not nickel-plated. The nickel-plated surface of NdFeB is inherently highly adhesive; ferrite lacks this layer.
Without the plating layer, there is one fewer “interface,” meaning that the adhesive must directly bond to the ceramic.
Now, let’s talk about the heat.
Although ferrite has better temperature resistance than NdFeB and a considerably higher Curie temperature, its intrinsic thermal conductivity is very low. The thermal conductivity of NdFeB is approximately 7–9 W/m·K, whereas that of ferrite is only 3–5 W/m·K.
When the motor is operating, eddy-current losses occur within the magnet; this generated heat must be conducted away through the adhesive layer. Since the substrate already has poor thermal conductivity, if the adhesive layer is also inadequate, the heat becomes trapped inside.
When selecting the drive motor, the rotational speed should be increased to above 15,000 rpm to compensate for the torque shortfall at high speeds. As the speed rises, eddy-current losses increase accordingly. At this point, how the heat is dissipated becomes an unavoidable engineering issue.
There’s also something regarding the dimensions.
Ferrite has weak magnetic strength; to achieve the same magnetic flux, more and larger magnets are required.
Halbach array configurations are inherently sensitive to positional accuracy. If the magnet orientation is incorrect, the magnetic field is no longer concentrated but instead cancels out. As the number of magnets increases and their overall size grows, the cumulative tolerance becomes even more significant.
The adhesive layer is precisely the “softest” link in this tolerance chain. Positioned between the magnet and the rotor, it must withstand centrifugal force, accommodate thermal expansion differences, and ensure that its position does not drift. The criterion cited in the previously disclosed document is that Tg should be at least 20–30 °C higher than the maximum continuous operating temperature; otherwise, once the modulus drops, the magnet will gradually shift outward under the action of centrifugal force.
Adhesive bonding interface profile. Note: Image generated with AI assistance
III. The real challenge has never been getting it to spin.
When a new technology emerges, everyone is primarily concerned with whether it can actually be implemented successfully.
But in the motor industry, the old rule is: getting it to start spinning is not the hard part; the real challenge is getting it to run for 100,000 hours.
I have seen far too many projects where prototype testing looks impressive, but mass production ramp-up fails spectacularly. They run smoothly for 200 hours on the test rig without any issues, yet once installed on the customer’s equipment, problems begin to emerge after just three months.
The cause often lies precisely in those “unimportant” details: the torque applied to a layer of sealant, a sealing ring, or a single screw.
The year before last, I visited a repair workshop; the very moment I entered, I was met with that distinctive smell. Burnt enameled wire mixed with charred adhesive was pungent enough to irritate the throat; after just ten minutes, the smell had already permeated my clothes. The shelves were piled high with motors that had already clocked several thousand operating hours.
Replacing rare earths with ferrite constitutes a substitution at the material level; its impact is evident and readily verifiable.
However, the effects associated with the layer below it—adhesive interfaces, thermal expansion mismatch, and long-term creep—are latent; verification takes a long time, and any problems emerge only later.
What should domestic manufacturers look for?
The reason is straightforward: Tesla defines the complete vehicle itself, defines the motor itself, and defines the production-line cycle time itself; it can completely redesign the rotational speed, magnetic circuit, windings, and cooling system from the ground up. In contrast, the vast majority of domestic motor manufacturers provide supporting components for third-party equipment, which imposes constraints on rotational speed, dimensions, and interfaces, leaving them with far less flexibility for such overhauls.
However, one factor can be directly leveraged: the level of emphasis placed on the bonding interface.
Among the projects I have worked on, discussions regarding the “no rare earths” approach generally stop at the stages of electromagnetic design and magnetic material selection. Bonding solutions are often finalized only at the final stage; in many cases, procurement personnel simply select a product based on price or continue using the previous-generation model, on the grounds that “that’s how it was used before.”
What worked in the past does not necessarily mean it will work now.
When the magnetic material is changed, its coefficient of thermal expansion changes, the operating temperature changes, and the dimensions of the magnet also change. These variables are transmitted to the adhesive layer, making the original set of parameters unlikely to remain applicable.
There is a readily available example, based on a case study disclosed by the manufacturer: a certain new-energy vehicle manufacturer previously used silicone rubber to bond magnetic steel; at 120 °C, the detachment rate was 12%. After switching to an epoxy-based system, the failure rate dropped from 3.2‰ to 0.5‰. The same magnetic steel and the same rotor were used; only the intermediate layer was replaced.
A specific figure
There’s another piece of data that I think is quite worth pondering.
According to publicly disclosed case studies from the manufacturer, after an 800 V electric drive platform adopted a magnetic steel bonding solution, the failure rate caused by magnetic steel detachment was reduced to zero, and NVH performance improved by 7 dB.
What does 7 dB mean?
Don’t rush to say “halfway.” In acoustics, two values are often confused: a reduction of 6 dB halves the sound pressure; a reduction of 10 dB is required for the human ear to perceive a halving in loudness. 7 dB falls somewhere in between: the pressure is not reduced by more than half, yet the change is still perceptible to the listener.
That characteristic “humming” high-frequency whistling of the motor falls precisely within the range most sensitive to human hearing. Reducing the volume by 7 dB produces a genuinely perceptible improvement in subjective comfort.
I suspect that when most car owners complain about motor noise, they do not think about that layer of adhesive; they simply perceive the vehicle as quiet.
Different types of magnetic steel bonding have a variety of adhesive options, which depend on the bonding process and performance requirements of the magnetic steel. If you have specific requirements for magnetic steel adhesives when bonding magnetic materials such as magnets, magnetic steel, or ferrites, please feel free to contact us; JOME is always happy to provide consultation.
Verified JOME Engineering Grades & Specification Cross-Reference
To qualify an industrial-grade adhesive, potting compound, or thermal interface material for this application, compare verified Shanghai Jome Industrial Co., Ltd. (JOME) formulations below or explore our primary category hubs: Anaerobic Threadlockers & Retaining Adhesives and Industrial Structural Adhesives.
| JOME Grade / SKU | Polymer Chemistry & Cure | Key Verified Specification | Primary Industrial Application |
|---|---|---|---|
| JOM 3107 | Anaerobic Magnet Structural Bonding Adhesive | High impact & centrifugal shear (>20 MPa) | +150°C rated | Servo motor, traction motor & humanoid robot rotor magnet bonding |
| JOME 1767 / JOME 2029A/B | 1K Heat-Cure & 2K High-Temp Motor Epoxy | Continuous service up to +180–200°C | High Tg & chemical resistance | High-speed turbine rotors, rare-earth & ferrite magnet bonding |
| JOM 638 | High-Strength Cylindrical Bearing Retaining Compound | Fills 0.05–0.25 mm radial clearances | Prevents fretting wear | Motor shafts, harmonic drive bearings, gears, bushings & sleeves |
| JOM 263/277 & JOM 204/93 | Liquid Anaerobic & Microcapsule Pre-Coated Threadlockers | Breakaway torque 20–42 N·m | Oil-tolerant & dry-to-touch film | Chassis bolts, robot joint fasteners, mass pre-coated screws |
Interactive Engineering Tools: Master Substrate & Thermal Cross-Reference Chart · Potting Volume, Mix Ratio & ΔT Calculator · Industrial Adhesive Manufacturer Comparison
Need Technical Data Sheets (TDS) or Trial Samples for Line Qualification?
Share your substrate pair, operating temperature, or target cycle time—only Email and Application / Part Code are required. Our R&D engineers reply within 24 hours.
Request free sample
-Speak with our material engineer
WhatsApp +86 18918756805
Matched JOME Industrial Formulations: For production bonding, potting, and sealing qualification in this application, evaluate 2K polyurethane structural adhesive, Single-component low-temperature heat-curing epoxy structural adhesive, 2K High-Temperature Resistant Epoxy Structural Adhesive, and High-strength 2-part epoxy structural adhesive (ISO 9001:2015 and CTI RoHS 2.0 verified; free engineering samples shipped via UPS freight collect).
Frequently Asked Engineering Questions
Q: Which JOME adhesive or potting grades are recommended for the drive motor does not require any rare earth materials at all,?
Depending on substrate pair, continuous operating temperature, and curing line constraints, Shanghai Jome Industrial Co., Ltd. (JOME) recommends evaluating JOM 3107, JOME 1767 / JOME 2029A/B, and JOM 638. Review our full specification table on this page or request a customized formulation sample.
Q: Can JOME customize viscosity, cure speed, or thermal conductivity for our automated production line?
Yes. JOME's Shanghai RandD laboratory customizes A/B mix viscosity, thixotropic index, pot life, Shore hardness, and thermal filler loading (0.75 to 25.0 W/m·K) to match your dispensing equipment and environmental test standards.
Industrial Adhesive Categories, Matched Grades and Engineering Guides
Product and Application Categories: Thermally conductive adhesive · Thermal Conductive Sealant · Thermal Conductive Pad · Silicone Potting Compound · Structural Adhesive · Silicone Structural Adhesive · Anaerobic Structural Adhesive · Silicone Sealants · Epoxy Sealants · UV Curing Adhesives · UV Silicone Sealant · Anaerobic Adhesive · Anaerobic threadlocker adhesive · Thread sealant · UV Curable Acrylic · Lubricating grease · Medical Grade Silicone gel · Company News Guides · Applications Guides · Electrical Appliances / Household Appliances / Consumer Electronics Guides · Wind power Guides · Medical Devices Guides
Matched Industrial Adhesives: Thermal Conductive Polyurethane Potting Compound（0.82W） · Medical-grade 2-Part Silicone Gel Scar Repair · Pre-coated thread sealant Water-based environmentally friendly（White, Red, Blue） · 2K Thermal Conductive Polyurethane Structural Adhesive（1.3W） · UV light rapidly cures thin-layer Potting Compound · Thermal conductive single-component RTV silicone adhesive（0.82W and 1.5W） · Single-component silicon-based thermal conductive gel (14 W/m-K) · 2-part silicone bonding adhesive high elasticity · UV silicone dual-curing high elasticity sealant can replace rubber sealing rings · Single-component RTV fluorosilicone sealant · Single-component heat-curing epoxy structural adhesive · 204 Microcapsule pre-coated threaded locking adhesive · 567 Anaerobic metal pipe thread sealant · Thermal conductive single-component RTV silicone adhesive（0.82W and 1.5W） · Micro Capsule Epoxy Thread Locking Agent · Thermal Conductive potting Silicone Compound(1.5W) · General-purpose epoxy potting compound (black, white, red, clear) · SMT epoxy base surface mount red adhesive · 10:1 Condensation type high bonding strength silicone potting compound
Technical Application Guides: Silicone vs Polyurethane vs Polysulfide vs Acrylic vs MS-Polymer Sealant: A Comparison Table of Common Sealant Types and Their Characteristics · Anaerobic Threadlocker, Pipe Thread Sealant & Retaining Compound Cross-Reference Guide (263 / 277 / 567 / 638 Equivalents) · Adhesive & Potting Compound Volume, A/B Mix Ratio Dosage, Bead Length & Thermal Impedance Calculator · How should you choose the surface-mounted magnetic steel adhesive for the inner rotor of a servo motor, drive motor, or high-speed turbine motor? · Adhesive selection for LiDAR transmission modules · Cost share of adhesives in medium-frequency transformers · Adhesive Manufacturer & Supplier for Machinery & Equipment · Zero VOC Adhesive and Low VOC Sealant Selection Guide · One Component Epoxy: Latent-Cure Structural Bonding Guide · Advanced Battery Cell Fixing & Potting Solution: JOMS 5230 Series A/B Silicone Encapsulants · High-Temperature Adhesive Solutions for Modern Home Oven Appliances · Comprehensive Guide to IGBT Potting Protection: Enhancing Reliability with High Thermal Conductive Epoxy · Introduction to Adhesives: Do You Really Understand Them All? — Adhesives Knowledge (Part 1) · A Brief Introduction to Polymer Adhesives · OCA Optical Adhesive: You've never compared the differences across these five scenarios · Silicone adhesive for hardware · What adhesive should be used to bond silicone sealing rings? · Which materials can be bonded using liquid hot-melt adhesives? · Can powdered fillers be added to adhesives? · Applications of Silicone Materials in UV-Curing · Understanding UV Viscosity in One Article: Core Insights into Resins, Coatings, Inks, and Adhesives · "Hidden Materials" in the Photovoltaic Industry: UV Coatings and UV Adhesives · Adhesive Solution for In Vitro Medical Devices (Monitors, Infusion Pump Housings, Sensors) & Endoscopes · Wind Turbine Whole Machine Adhesive Solution (Onshore + Offshore Wind Power)-Cowlings, Interior Assembly of the Machine · Sealing materials for vehicle oil circuit components – fluorosilicone rubber and fluorosilicone gel · New energy vehicle thermal management solution · The scheme for the adhesive application areas in car body assembly · Recommended adhesive solution for power controller · The application of high-performance adhesives in lithium battery modules · Analysis of Adhesives for Power Batteries in New Energy Vehicles · Medical Device Adhesive Solutions · A comprehensive bonding solution for the wind power industry (including onshore and offshore wind power) · Complete Adhesive Solutions for Industrial & Consumer Electronics