PM Benchmarks

Powder Metallurgy KPIs and Benchmark Ranges for Sintering Plants

The KPIs that matter in a sintering plant, world-class versus typical benchmark ranges, how to measure them, and the levers that move each one.

First-pass yield is the headline KPI. World-class structural PM plants hold 97 to 99% good parts from pressed greens; typical shops run 93 to 96%, and anything below 92% signals a tooling or density-window problem. Measure it as good shipped over green pressed, per part number and per shift, not blended across the plant. The biggest levers are green density stability, punch and die wear tracking, and pre-sinter handling. Track yield alongside your Part Yield output so a 2-point gap converts directly into the press strokes and powder you are burning without revenue.

Green density consistency, not just its average, drives quality. Target a standard deviation under 0.03 g/cm3 around your setpoint; world-class cells hold under 0.02 while typical cells drift 0.05 or more, which shows up as shrinkage scatter and out-of-tolerance bores. Sample 5 parts per hour and plot Cpk against your density spec; aim for Cpk above 1.33. Levers are fill-shoe consistency, powder apparent-density control lot to lot, and press repeatability. Monitor the setpoint with Green Density and react to lot changes before they reach the furnace.

Furnace loading efficiency is the most under-watched cost lever. Belt or tray utilization should sit at 80 to 90% of usable area; many plants run 55 to 70% and pay full energy for half a load. Measure loaded footprint over available hot-zone footprint each run. Because a half-empty furnace roughly doubles per-part energy, moving from 60 to 85% loading cuts energy cost per part by nearly 30%. Check planned loading against Sinter Furnace Capacity so scheduling batches parts of similar residence time instead of starving the belt.

Powder loss rate separates disciplined cells from wasteful ones. Total loss including lubricant burn-off runs 1.5 to 3.0%; world-class handling keeps recoverable spill and setup loss under 0.5%, while sloppy cells lose 2% or more just to spillage and mixed reclaim. Weigh powder issued against good-part mass plus documented burn-off monthly. Levers are closed fill systems, disciplined reclaim segregation, and setup-scrap reduction. Trend it with Powder Loss Rate, because a half-point of avoidable loss on a 500-tonne-per-year cell is 2.5 tonnes of premix walking out the door.

Overall equipment effectiveness ties availability, performance, and quality together. Press-line OEE of 85% is world-class, 60 to 75% is typical, and under 55% means changeover or downtime is eating you. On a mechanical press, performance loss usually comes from running below rated strokes per minute; a line rated at 18 strokes/min but averaging 12 is a 33% performance hit before any downtime. Track availability, speed, and yield separately so you fix the real constraint. Furnace availability above 95% is the target, since a cold restart from ambient costs 6 to 10 hours.

Scrap and rework rate should be measured by stage, not blended. World-class total scrap is under 2%; typical is 3 to 5%. Green cracks and edge chips before sinter should stay under 1%, and final-gauge rejects under 1.5%. Because late rejects carry all upstream cost, weight your scrap KPI by the stage it occurs, not by piece count. The levers are density control for green cracks, atmosphere and dew-point control for sinter defects, and tool maintenance for dimensional rejects. A dew point held below minus 30 C in the hot zone prevents most oxidation-driven rejects.

Dimensional capability and energy intensity round out the scorecard. Sintered dimensional Cpk should exceed 1.33 on critical features, with as-sintered tolerances near +/- 0.3 to 0.5% of dimension and sized features holding IT6 to IT8. Energy intensity, measured as kWh per kg of good product, benchmarks at 1.5 to 3.0 kWh/kg for iron systems on a well-loaded belt furnace; poorly loaded lines exceed 5 kWh/kg. Track kWh/kg monthly against Furnace Energy Cost so loading and yield improvements show up as a falling energy-per-good-part number, which is the cleanest single proxy for plant discipline.

Published 2026-07-01.