Automobile Transmission Powder Metallurgy Gears: How the Process Works

• 1
  What Happens Inside an Automobile Transmission Powder Metallurgy Gear
• 2
  Why Density Decides Load Capacity in Automobile Transmission Powder Metallurgy Gears
• 3
  Geometry Rules That Automobile Transmission Powder Metallurgy Gears Must Follow
• 4
  Tolerances You Can Hold Without Touch-Up Machining
• 5
  When Automobile Transmission Powder Metallurgy Gears Are the Wrong Choice
• 6
  Density Routes and What Each Buys You

An automobile transmission powder metallurgy gear begins as a blend of atomized iron powder, copper, nickel, molybdenum and a lubricant. The blend drops into a die cavity and two punches squeeze it from top and bottom at 400–800 MPa. The green compact that comes out holds its shape but has almost no strength. It is fragile enough that you can snap a thin web with your fingers.

Sintering is where the part becomes metal. The compact runs through a furnace at 1,120–1,150 °C in a nitrogen-hydrogen atmosphere for 20–40 minutes. Copper melts and spreads along the iron particle boundaries. Nickel and molybdenum diffuse more slowly and stay near their original positions. The necks between particles grow, the pores round off, and the part shrinks about 0.5–1.5% linearly.

The result is a gear with 10–12% residual porosity by volume. That porosity is the defining feature of the process. It is also the source of every limitation that follows. A pore is a stress raiser, a path for oil, and a spot where a cutting tool will interrupt its cut.

Involute helical gears are harder to press than spur gears because the helix has to be released from the die. Helix angles up to roughly 30° can be pressed with a rotating lower punch, but the tooling cost climbs fast beyond about 20°.

• 1
  Green densityTypically 6.8–7.2 g/cm³ before sintering
• 2
  Sintered density6.9–7.4 g/cm³ with single pressing and sintering
• 3
  Porosity10–12% by volume, mostly closed after sintering
• 4
  Shrinkage0.5–1.5% linear, predictable but die-specific

Bending fatigue strength scales with density almost exponentially. At 7.0 g/cm³, a sintered gear may show a bending endurance limit around 250–300 MPa. Push density to 7.4 g/cm³ and the same alloy can reach 400–450 MPa. That is why density is the first number to fix when you specify an automobile transmission powder metallurgy gear.

Single pressing and sintering gets you to roughly 7.1 g/cm³. Double pressing and double sintering reaches 7.3–7.4 g/cm³ but doubles the process steps and cost. Warm compaction, where the powder blend is heated to about 130 °C before pressing, buys another 0.1–0.2 g/cm³ without a second press cycle. Surface densification by rolling the flanks after sintering can push the tooth surface itself to near full density while the core stays porous.

Each step up in density tightens the dimensional window. High-density compacts spring back more, and die wear accelerates. A die that holds ±0.05 mm on the tooth profile at 7.0 g/cm³ may drift to ±0.08 mm after 100,000 strokes at 7.4 g/cm³.

The core stays porous even when the flanks are densified. That is not a defect. Oil held in the pores feeds the contact zone, which is why PM gears often run quieter and cooler than machined gears in lightly loaded applications.

Powder cannot flow uphill into a thin vertical wall the way molten metal can. A compacted gear needs draft on every wall that is not parallel to the pressing direction. Walls thinner than 2.5 mm are difficult to fill evenly, and walls under 1.5 mm usually show density gradients that crack during ejection.

Holes must be parallel to the pressing direction. A cross-drilled oil passage cannot be pressed in. You either press a straight hole and drill it later, or you change the oil feed design. Every feature that is not parallel to the press axis adds a secondary operation.

The length-to-diameter ratio of a hole matters too. A hole with a depth greater than about 8 times its diameter will not fill properly at the bottom, and the core rod that forms it will bend under load. Step the hole or accept a drilled finish.

Sharp internal corners are the enemy. A radius of 0.5 mm minimum on internal corners lets powder flow and gives the punch a chance to survive. Zero-radius corners are possible in a machined gear. They are not possible in a pressed one.

As-sintered gears typically hold ±0.05 mm on the tooth profile and ±0.10 mm on the outside diameter. That is not enough for a modern transmission input gear, where backlash and contact pattern are specified in tens of microns. So the sintered gear usually goes to a CNC operation for the bore, the face, and sometimes the flanks.

Boring the bore to H7 and grinding the faces brings the part into the tolerance band. At GreatLight we hold ±0.005 mm on machined features and Ra 0.8–1.6 μm on ground faces, using 5-axis centers that can locate the gear off the sintered tooth flanks rather than the pressed bore.

There is a common mistake here. Designers specify a sintered gear with a tight bore tolerance and expect the press to meet it. The press cannot. The bore is formed by a core rod that wears and springs back. Budget for a finish bore operation from the start.

Sintered gears also need a sizing or coining pass if the bore is used as a datum. Skipping it means the CNC setup has to find the part on the flanks, which costs cycle time.

The process loses to bar stock in three situations. First, low volume. A press die for a helical gear costs tens of thousands of dollars. Below roughly 20,000 pieces per year, machining from 4140 or 4340 bar is usually cheaper per part.

Second, high torque density. A transmission gear carrying 500 N·m or more at the tooth contact needs full density and clean steel. Sintered gears at 7.4 g/cm³ can work in this range with surface densification, but the cost advantage shrinks and the fatigue data is harder to defend. Machined or forged-and-cut gears are the safer call.

Third, geometry that cannot be pressed. Undercuts, cross-holes, and non-parallel features all break the pressing rule. If your gear has an internal oil gallery or a shoulder that blocks the punch path, PM is off the table.

There is also a supply chain point. PM die lead times run 8–14 weeks. If you need a prototype gear next month to validate a transmission design, you machine it. PM dies are for parts you already know will not change.