All

A 9-roll 3-in-1 straightener feeder left 1.1 mm of edge wave on 180 mm SPCC while the centre stayed flat, and the toolroom checked the die twice. This separates edge wave from centre buckle, runs a two-hour sequence from the coil tail to the lead wire, and shows how 12 mm of off-centre tracking and a roll crown cut for wider strip each contributed. It also explains why slit-edge wave cannot be straightened out, only hidden.

A plant solved slip on oily cold-rolled strip by knurling its feed rolls, then moved 0.4 mm stainless onto the same NC servo feeder and lost 4 percent of output to marks. This compares smooth steel, polyurethane, knurled and carbide-coated surfaces on grip, marking threshold, wear and price, and converts controller bar readings into the kilonewtons the strip actually sees. It closes with a four-hour test that settles the choice on a strip sample.

A 10-tonne uncoiler was poured on the one-page drawing that came with it, and fourteen months later the rear anchors had lifted 3 mm. This separates bearing load, overturning moment and brake reaction, works the numbers for a 1,700 mm cantilever, and shows why the rear anchors carry about 3.3 tonnes each in tension. It also covers slab thickness, coil car rails, and the four cases where a bigger pad is the wrong answer.

An appliance plant reworked 3 percent of every shift on 0.5 mm prepainted steel, and both the coating supplier and the die shop were blamed before anyone measured the line. This traces the marks to four contact points, prices the fix at $7,500 against a $9,400 monthly rework bill, and separates the marking defect from a die clearance problem nobody had noticed. It also lists the four runs where coated strip should not be used at all.

A purchasing manager with three quotations spread across $60,000 had no way to rank them, because every bid was priced per machine and none per tonne of steel. This builds the per-tonne number from four inputs, works a 1,200-tonne example to about $56 a tonne, and shows which specification lines move it most. It also sets out the three cases where the figure misleads, and what to add to each bid before comparing prices.

A bracket die asked for 45 mm of feed at 190 SPM and the line delivered the speed with 3.2% scrap. A feed trace showed 0.09 mm of error at 190 and 0.04 mm at 140, all of it in the settle window rather than in the machine. Four free setting changes recovered 35 SPM, and the article closes with a three-term formula for calculating your own ceiling.

A fresh set of pilots lasted nine days because the feed unit sat 0.3 mm off the press centreline, a number nobody had measured since installation. The sequence runs from fixing a permanent datum to proving the result with a strip, with the tolerance for each check and what 0.2 mm of skew does to a pilot bush. It also sets out the four conditions where alignment is not the answer.

Two presses eleven metres apart and one decoiler between them: the drawing said shared, nobody drew the twenty minutes of re-datumming. This compares the three real arrangements, prices the rails and foundations that never appear on a machine quotation, and explains why an accumulator buys 63 seconds rather than an eighteen-minute coil change.

Coil weight looks like a purchasing detail and behaves like a throughput decision, because every coil change stops the press for the same eighteen minutes. Four inputs produce the number that matters: strip mass per metre, parts per coil, run time per coil and the downtime share. The article works a 5-tonne and a 10-tonne coil through the arithmetic, then shows where heavier coils start costing more than they save.

A 400-tonne press bay promised four weeks from delivery to first good part and took seven. The calendar was never one schedule but five overlapping clocks, and only some of them move for money. This breaks the six weeks into concrete, rigging, utilities, commissioning and process tuning, with the week-by-week plan and the five pre-shipment documents that keep the dates honest.

A cassette roll head turns a two-hour gauge change into a twenty-minute one, and puts a locating joint in the middle of the most highly loaded part of the machine. This article compares the two heads on the five decisions that change a production week, names the four cases where a cassette is the wrong purchase, and works the payback from one changeover a week to eight. It also covers what a 6 mm job does to the joint and how to test one before ordering.

A feed error that comes and goes with the weather is not an encoder problem. On three of the last five servo feeder slip complaints we traced, the clamp was not holding because supply pressure fell mid-stroke. This article covers the gauge logging that separates an air fault from everything else, four faults that look identical at the die, line sizing for 350 litres a minute, and the four observations that rule the air theory out.

A decoiler mandrel is asked to hold five tonnes of steel and expand into a bore it cannot see, and most of its damage happens in the ninety seconds before the coil turns. This article sets out the six coil checks to make before the crane moves, eight loading steps from floor to first turn, and the five failure modes behind almost every mandrel repair. It also names the coils that should never go on a standard mandrel.

An appliance bracket ran nine months before a 480-hour salt spray test failed on all three samples. Coating mass measurement found the answer: the incoming coil read 134 to 136 g/m2 per face, and strip after the straightener section read 118 to 125 g/m2 on the upper face only. This is a field case on zinc transfer to rolls, the three fixes that recovered 8 to 11 g/m2, and the four things the diagnosis cannot prove.