Powder Metallurgy for Humanoid Robot Components: Gears, Bearings, Joints, and Structural Parts

# Powder Metallurgy for Humanoid Robot Components: Gears, Bearings, Joints, and Structural Parts

Humanoid robots are leaving the lab. Tesla Optimus, Figure, 1X, Agility Robotics, and a dozen Chinese manufacturers are converging on a similar architecture: 28–40+ rotary joints per unit, each packed with gears, bearings, structural carriers, and a compact electric motor. At scale — thousands of robots per year, hundreds of thousands of identical gears — the component supply chain will look less like aerospace job-shop machining and more like automotive tier-one production.

That is exactly the manufacturing regime where powder metallurgy is strongest.

This guide surveys the powder metallurgy components inside humanoid robot joints and actuators, the materials and processes that apply, and the design and quality considerations specific to this emerging product category. It complements our existing pages on robotics and automation PM parts and the robot planetary gearbox case study.

The Humanoid Robot Joint: A Component Map

A typical rotary joint in a humanoid robot contains:

Each of these has a direct analog in automotive, power tool, or electric motor production — the PM industry has been making gears, bushings, brackets, and SMC cores for decades. The humanoid robot application changes the packaging, performance targets, and documentation requirements, not the fundamental process capability.

PM Gears for Robot Joints

Gear Types and Sizes

Robot joint transmissions — usually planetary or harmonic-drive configurations — use gears that are smaller and finer-pitch than automotive equivalents:

• Module: 0.3–1.0 (automotive is typically 1.0–3.0)
• Outer diameter: 10–50 mm for planet and sun gears
• Face width: 3–15 mm

This is at the fine end of conventional PM gear capability. Tooling quality, powder fill uniformity, and sintering atmosphere control become more critical as tooth size decreases. However, PM has a significant advantage over hobbing at small modules: the compaction die forms the tooth profile in one step, while hobbing a module-0.5 gear requires tiny cutters, slow feeds, and high tool wear.

Material and Heat Treatment

Typical robot PM gear specification:

• Grade: FN-0205 (MPIF), density 7.0–7.2 g/cm³ minimum
• Heat treatment: Case hardening (carburizing at ~870°C, oil quench, temper to 150–200°C)
• Surface hardness: 55–62 HRC equivalent on tooth flanks
• Core hardness: 25–35 HRC for toughness
• Tooth quality: AGMA 8–10 / JIS 5–7 as-sintered, with sizing or grinding for higher precision when needed

Tooth-Root Fatigue

This is the life-limiting criterion for robot gears. A humanoid knee or hip joint cycles millions of times at varying loads — the gear train must survive not just peak torque but cumulative fatigue. PM gears at 7.0+ g/cm³ density with proper heat treatment have tooth-root bending fatigue limits competitive with machined low-alloy steel gears, but the design must account for residual porosity at 90–93% density. For the highest-demand joints, PM gear blanks plus finish grinding of the tooth profile is a practical path.

For gear design background, see our PM gear design guide.

Oil-Impregnated Bearings: Self-Lubricating Joints

One of PM's quieter advantages for humanoid robots: oil-impregnated sintered bushings that provide lubrication for the life of the joint without external grease, oil reservoirs, or maintenance access.

In a robot with 40 rotary joints, eliminating even 10 grease fittings or sealed bearings simplifies assembly, reduces part count, and saves weight. PM bearings made from bronze (Cu-Sn) or iron-graphite materials hold 18–25% porosity by volume — after vacuum impregnation with oil, the pores act as a built-in lubricant reservoir that releases oil gradually as the bearing warms during operation.

Applications:

• Planet gear bores running on needle rollers or pins
• Pivot bushings for fingers and wrist joints
• Low-speed oscillating bearings that see frequent reversing motion

For bearing specification details, see our guides on oil impregnated bearings and bronze self-lubricating bearings.

SMC Motors for Robot Joints

Humanoid robot joint motors are pushing toward axial flux topologies — the pancake form factor fits the short axial space behind knee and hip joints, and high pole counts plus direct-drive or quasi-direct-drive operation mean high electrical frequencies where soft magnetic composites outperform laminated steel.

Key SMC advantages for robot motors:

• 3D flux paths in claw-pole and segmented stator topologies that laminations cannot follow
• Lower eddy-current loss at the 400–1,000+ Hz fundamental frequencies typical of high-pole-count robot actuators
• Net-shape pressing eliminates stamping scrap from complex stator shapes
• Segmented SMC teeth simplify winding insertion for small-slot robot motors

Tradeoffs: lower permeability and higher hysteresis loss at low frequency versus laminations, and cured SMC has lower mechanical strength than sintered structural PM parts.

For the full SMC story, see our soft magnetic composites guide and the axial flux motor blog.

Structural Brackets and Carriers

PM structural parts for robot joints share the same DFM constraints as any PM part: the geometry must eject along one pressing axis, wall sections should be above ~1.5 mm, and undercuts require secondary machining. The advantage is material efficiency — a complex carrier plate with gear pockets, bearing seats, and mounting bosses can be pressed to near-net shape instead of machined from a solid block.

For robot programs concerned about corrosion (washdown environments, close-to-skin contact), stainless PM grades like 316L or 304 are available. These cost more than ferrous grades but eliminate the need for post-process plating or coating.

Production Economics: Why PM Fits the Humanoid Robot Scale-Up

Humanoid robot production is at an inflection point. At prototype volumes — dozens to low hundreds per year — CNC machining from bar stock or billet is the pragmatic choice. Tooling amortization favors PM only when annual program volumes cross roughly 5,000–10,000 parts per part number.

But the math changes fast:

• 1,000 robots/year × 40 joints/robot × 3 planet gears/joint = 120,000 planet gears/year
• Add sun gears, ring gears, bushings, brackets — easily 300,000–500,000 PM parts per year from one robot program

At that scale, PM unit cost undercuts machining for most gear and structural part geometries, and the quality of PM production — statistical process control, batch traceability, PPAP documentation — is already proven in automotive and industrial supply chains.

For cost comparison background, see PM vs CNC cost comparison and PM cost factors.

Getting Started: From Drawing to Feasibility Review

If you are designing a humanoid robot joint and considering PM components, send:

1. Gear data: tooth count, module, pressure angle, face width, torque rating
2. Bearing data: bore, OD, length, load, speed, lubrication requirement
3. Structural part data: STEP file, material preference, critical tolerances
4. Program data: target annual volume, timeline, quality documentation requirements

Engineering will respond with a DFM review within 24–48 hours, including a process recommendation and an honest assessment of whether PM, machining, MIM, or a hybrid route is the best fit.

👉 Submit a quote request

Related Resources

• Robotics & Automation PM Parts — Industry Overview
• Robot Planetary Gearbox Case Study
• PM Gear Design Guide
• Soft Magnetic Composites Material Guide
• Axial Flux Motors and SMC
• Oil-Impregnated Bearings Overview