Coil Yield

Managing Coil Yield to Protect Material Margin

Material is 60 to 75 percent of stamping cost, so coil yield is where margin lives or dies. This playbook covers honest yield math, benchmarks by part family, and the weekly routine that recovers points.

In most stamping shops material is 60 to 75 percent of product cost, which makes coil yield the highest leverage number in the building. A plant buying 10,000 tons of steel a year at $900 per ton spends $9 million on coil; moving yield from 68 to 71 percent cuts purchases by about 4.2 percent, worth $380,000 straight to margin, minus a modest scrap credit you were already getting. No labor project in the plant returns money that fast. Yield decides quotes too: a competitor running 6 points better can underbid you and still make more per part.

The math is simple and easy to fudge. Coil yield equals shipped part weight divided by purchased coil weight. Worked example: a 10,000 kg coil produces 8,200 good parts at 0.85 kg each, so 6,970 kg ships and yield is 69.7 percent. The remaining 3,030 kg is skeleton, offal, slugs, coil ends, and scrapped parts. Run the numbers through the Coil Yield calculator per coil, not per month, because monthly averages hide the one die that runs 8 points below its strip layout. Weigh scrap hoppers; do not back calculate from theoretical part weight on a part that gets trimmed.

Separate engineered yield from actual yield. Engineered yield comes from the strip layout: part area divided by pitch times width. Actual yield subtracts real world losses: coil ends of 3 to 8 meters per end, threading scrap, setup parts, and quality rejects. A die engineered at 74 percent that delivers 67 percent actual has a 7 point execution gap, and on a die consuming 500 tons a year at $900 per ton that gap costs $31,500. Track both numbers; the engineered number is a die design problem, the gap is a shop floor problem, and they have different owners.

Benchmarks vary hard by geometry. Simple rectangular blanks nest at 85 to 95 percent. Progressive dies with carriers and webs typically run 55 to 75. Deep drawn round shells land 40 to 60 before trim, and irregular brackets can be worse. Coil end and setup losses should stay under 2 percent of coil weight on a disciplined line; sloppy lines give up 4 to 6. Scrap credit softens the blow but never fixes it: offal sells for 15 to 25 percent of what you paid, so every kilogram of skeleton still burns 75 to 85 percent of its purchase price.

The design levers pay the biggest single hits. Tighten pitch and web: cutting a 62 mm pitch to 60 with the same part lifts yield 3.3 percent on that die forever. Order coil width to the layout instead of nesting into a stock width; a layout needing 147 mm run on 152 mm stock donates 3.3 percent to the scrap hopper on width alone. Consider two out layouts, mirrored nesting, and trapezoid blanks for tapered parts. On draws, blank development trials that shave 2 mm off the blank diameter routinely save 1.5 to 3 percent of material.

Execution levers are cheaper and faster. Cut coil end loss by running coils to the tail with an end joiner or by sequencing short jobs onto remnants. Reduce setup scrap from 40 parts to 10 with recorded feed and pilot settings. Fix edge camber and width variation with your supplier; a mill delivering plus or minus 1.5 mm on width forces you to order wider than the layout needs. Track scrapped parts per coil change: at 12 changes a day, saving 20 parts per change at 0.85 kg is 204 kg daily, roughly 50 tons a year.

The failure modes are predictable. Scrap credit accounting that nets material cost makes yield look like a small number nobody owns. Weighing nothing and trusting the ERP standard hides drift for months. Purchasing switching to a cheaper wider coil wipes out a die room yield project without anyone connecting the two. And chasing yield into the web can crack a die section or cause strip wander misfeeds; below about 1.5 times material thickness on web width, downtime and burr costs usually exceed the material saved. Every yield change needs a die maker and a quality tech in the room.

Cadence: daily, log yield per coil on the top five material consumers and flag any coil 2 points below standard. Weekly, Pareto scrap kilograms by die, pick one, and assign an engineering or purchasing action. Monthly, reconcile purchased tons versus shipped tons versus scrap sold; if the three do not close within 1 percent you have a measurement problem, not a yield problem. World class plants hold actual yield within 2 points of engineered, keep coil end and setup loss under 1.5 percent, and review every new die layout for yield before steel is cut, because the cheapest scrap is the scrap you never design in.

Published 2026-07-02.