Metals and Materials Manufacturing Case Studies
Industrial5 Case StudiesUpdated Sep 2026
Metals and materials manufacturing sits where chemistry, mechanics, and scheduling meet on the same shop floor. The five cases below come from plants and fabrication shops of different sizes, and each one turns on a decision that looked reasonable at the time it was made. Read them for the evidence each team used, what they chose to measure, and where the records fell short. The analysis paragraphs point to the constraints that mattered most. The discussion questions are meant for a team meeting, so bring the data tables and argue from them.
Case Study 1
Cracked hydraulic pump housings in cold climate service
Brandt Ridge Forge, a closed-die forging plant of about 240 people that supplies crankshaft blanks to two engine assemblers.
The situation
Brandt Ridge received field returns over several months from a hydraulic pump housing cast in ductile iron and machined in the plant's own cell. Units installed on mobile equipment working in sub-freezing yards were cracking at a bolt boss near the inlet port. The part had passed its drawing tolerances at final inspection, and the material test reports from the foundry were in order. The housing had been in production for about four years with a warranty return rate that stayed low until the equipment fleet moved its duty cycle toward longer idle periods in cold weather. Engineering had a service life target written into the customer specification, and the returns were eating into the margin on that contract.
The challenge
The team had to decide whether the fault lay in the casting, the machining sequence, the bolt torque, or the duty cycle itself. Each explanation predicted a different pattern in the returns, and the plant had records for only some of them. Maintenance logs showed the torque checks were done at install but not at the first service interval, which left a gap nobody had flagged.
Analysis
Teams grounded in Metals and Materials Manufacturing usually start by separating what the drawing requires from what the part actually experiences in service. Dimensions were already cleared, so the real constraint was the load spectrum. Telematics data showed the housings spent long stretches at low temperature with thick fluid and cold pump starts, which drove pressure spikes into the boss area. The foundry test report confirmed tensile strength, but the test bar had been cast at a different section thickness than the housing wall. That gap in the evidence mattered more than any single test result.
The corrective options were narrower than the team first assumed. Raising the wall thickness at the boss would have fixed the stress concentration but added mass to a part that already sat close to a packaging limit on the fleet. Changing the boss fillet radius was cheaper and could be validated with the existing fixture, though it needed a tolerance change on the drawing and a customer sign-off. Maintenance records added a third path: a torque check at the first service interval, which cost little and closed the loss-of-preload pathway that had been masking the stress. Ranking used the service life target, not cost alone.
Actions taken
- Restore torque verification at the first service interval for all units in cold duty, and log the readings against the housing serial numbers so the next review has field data.
- Request a fillet radius change at the boss, backed by a finite element run that uses the measured telematics load spectrum rather than the static design case.
- Cast a witness coupon from the production pour at the housing wall thickness and send it for tensile and fatigue testing before the drawing change is released.
- Hold the foundry to a section-thickness requirement on its test reports, and reject lots whose test bars were not poured to match the housing wall.
Results
Torque verification at the first service interval went in during the first month, and cold-duty returns fell sharply over the next two quarters. The customer approved the fillet change once coupon fatigue results matched the baseline within test scatter, and it reached production by the end of the quarter. The foundry reissued its test reports against the section requirement. No new returns have appeared in the cold-duty group, though the team still reviews it monthly because the record is short.
Questions for discussion
- Which of the four corrective options would you have tested first, and what field data would have changed that choice?
- How would you set the sampling plan for the witness coupons if the foundry pours only a few lots per month?
- What records should a plant keep at install and first service so that a later failure can be traced to a single cause?
Case Study 2
Press line throughput drops as burr defects climb
Ostrander Precision Stampings, a progressive die stamping plant of about 310 people serving appliance and HVAC makers.
The situation
Ostrander runs eleven mechanical presses on a progressive die line that stamps mounting brackets from cold-rolled steel coil. Over several weeks, output per shift slipped while the scrap bin filled faster. Burrs on the pierced holes were measured above the drawing limit on a rising share of parts, and the downstream welding fixture began rejecting brackets that would not seat. The line had run the same tooling for about three years, but a coil supplier change had moved the material to a slightly harder grade in the same nominal gauge. Operators were adjusting clearance and stopping the press to clear jams, and the stoppages were not all being logged in the same way.
The challenge
The plant manager needed a recovery plan that could be agreed within the week, but the causes were tangled. Die wear, material hardness, lubrication, and press speed all pointed to the burr pattern, and each had a different owner. The data on hand came from a quality log that recorded rejects but not the press state at the time, so the team could not yet say which factor dominated.
Analysis
Mapping the line from coil feed to bracket transfer showed three places where variation could enter: the coil itself, the die and its clearance, and the press stroke rate that sets how the material is sheared. The drawing called for a burr height under a set limit at the pierce, with clearance held to a percentage of material thickness. Those tolerances were written for the original steel grade, so the supplier change had quietly moved the design constraint without touching the drawing. Checking hardness on incoming coil and clearance on the die at each shift start gave the team a first look at which input had shifted.
Ranking the fixes scored each on burr impact and how fast the result could be verified. Die clearance was the most direct lever, since tighter clearance for the harder coil reduces burr, but it also raised die wear. Coil hardness could not be changed on the floor, only gated at receiving, so it became a control rather than a fix. Lubrication and stroke rate had smaller effects and were cheaper to check, so they went to the back of the queue. Acceptance was set as burr height inside the drawing limit on at least 95 percent of sampled parts across two consecutive shifts.
Actions taken
- Reset die clearance at the pierce station to the low end of the range for the current coil, then verify burr height on twenty consecutive parts at each shift start.
- Add a hardness check at receiving with a hold limit, so lots outside the agreed band are segregated before they reach the press line.
- Start logging every stoppage with press state, die number, and a cause code, and review the log at the end of each shift with the tooling lead.
- Set the acceptance test for the change as burr height inside the drawing limit on at least 95 percent of parts over two consecutive shifts, with stoppages under the agreed count per shift.
Results
The clearance reset lowered burr rejects during the first shift of running, and the receiving hold kept two suspect lots off the press in the first week. Throughput came back to within a few percent of its earlier rate by the end of the second week. Scrap per shift fell by roughly a third. The stoppage log also showed that most lost time came from jams in the first hour after a coil change, which became the next project.
Questions for discussion
- If the coil supplier could guarantee a hardness band but not a mean, how would you set the receiving gate and the die clearance together?
- Which stoppage causes would you want coded separately in the log before anyone draws conclusions about press speed?
- At what point would you stop tuning clearance and open a discussion with the supplier about the grade itself?
Case Study 3
Wet scrap charging hazard before a new furnace commissioning
Harrow & Vell Remelt, a secondary aluminum remelter of about 190 people that buys clean and painted scrap from regional collectors.
The situation
Harrow & Vell planned to add a tilting rotary furnace to its remelt building, replacing a smaller unit that had run for many years. The new furnace would take a higher share of painted and oily turnings, and the scrap intake would move from a covered bay to an open-sided yard with a conveyor feed. The engineering package had been prepared by the furnace vendor and reviewed in-house for mechanical fit and burner sizing. The safety file was thin. It contained a general hazard list copied from the old unit, and the vendor's drawings showed no burst panel details, no water-exclusion measures for the charge area, and no gas detection layout for the burner train.
The challenge
The review had to decide whether the project could move on to procurement with the gaps open, or whether it had to stop until the evidence existed. The site safety lead wanted a hard stop on the wet charge. The project lead argued that the conveyor design already excluded standing water and that the vendor's standard package met the general requirements. Nobody had yet checked which standards the new furnace fell under once the scrap intake changed.
Analysis
The most serious hazard in a remelt is water or wet charge meeting molten metal, which can flash to steam and eject metal from the furnace. Oily and painted turnings add a second hazard, since their combustion products and any hydrogen released during heating can accumulate around the burner train. Moving the intake into an open yard brings rain and washdown water onto the charge pile, so the old controls no longer hold. The controls that matter are a dry, preheated charge, an exclusion zone that keeps water out of the feed, and a relief path sized for the worst case rather than the old unit's duty.
Applicable requirements depend on jurisdiction and furnace rating, so the review had to name them rather than assume them. Pressure equipment rules govern the vessel and its relief device, combustion safeguard requirements govern the burner train and its purge logic, and fire and dust provisions govern the charge area. The file also lacked a traceable hazard analysis, meaning no link from each cause to a control and a test. Missing items were the burst panel calculation, the gas detection layout, the interlock description, and a written wet-charge rejection procedure. The review set those four items as a gate before any furnace purchase order could be released.
- 1Jurisdiction reviewConfirms which pressure, combustion, and fire rules apply
- 2Hazard analysis rebuiltTraces wet charge, hydrocarbon, and hydrogen causes
- 3Charge area controlsDry feed, exclusion zone, and drainage design
- 4Relief and purge designBurst panel sized for worst case, interlocks verified
- 5Safety lead sign-offRequired before any furnace purchase order is released
Actions taken
- Stop the wet-charge path until a dry, covered intake or an enclosed drying step is in place, and record the decision in the project file.
- Ask the furnace vendor for a traced hazard analysis that links each cause to a control, a verification test, and the owner responsible for it.
- Size the burst panel and relief path for the worst charge case, using the calculation method the pressure equipment rules require, and have it checked by an independent engineer.
- Hold the purchase order until the site safety lead and the insurer's engineer have reviewed and signed the four missing documents as a single package.
Results
The revised package showed the relief path had been undersized for wet-charge events, and the vendor added a larger panel and a dedicated drying step to the intake. Procurement was paused for a short period while the vendor reissued its analysis. The insurer's engineer asked for a purge interlock test on the burner train before commissioning, which the vendor agreed to include in the acceptance tests. The furnace order went forward about one quarter later than planned, and the scrap intake was moved under cover before the furnace arrived.
Questions for discussion
- Which of the four missing documents would you have asked for first, and what would you have accepted as a substitute if the vendor could not produce it quickly?
- How should the review treat a hazard that the vendor says is handled by an operator procedure rather than an engineered control?
- If the scrap mix will change over the life of the furnace, what evidence should be required before the charge limits are relaxed?
Case Study 4
Late plate delivery threatens a fixed plant shutdown date
Tarrant Vessel Works, a pressure vessel and skid fabricator with about 420 people across two shops.
The situation
Tarrant Vessel Works had contracted to deliver a skid-mounted heat exchanger package to a customer whose plant would shut down for a planned turnaround. The delivery milestone was fixed by that outage window, and a late skid would push the customer's restart by a full cycle. The package used duplex stainless shell plate and tubesheets, and the order had been placed on the assumption that plate stock would be available from a distributor. Partway through fabrication, the distributor reported that the plate heat lot had been allocated to another buyer and that the next lot was months out. Design had also changed once already, after the customer asked for a larger nozzle on the shell.
The challenge
The project manager had three weeks to present a recovery plan that the customer could accept. Each option moved cost, schedule, and quality in different directions. Expediting plate from another mill would cost more and needed a new chemistry review. Splitting the package would meet the milestone for the critical section but leave the rest for the next outage. Resequencing the shop around a substitute plate would hold the schedule only if the nozzle redesign was frozen today.
Analysis
The recovery options are bounded less by the calendar than by what the code and the customer specification permit. Duplex plate must carry heat-traceable mill reports, and the ferrite and impact requirements limit which substitute lots can be used. A plate from a new source has to pass chemistry and heat treatment checks before cutting, and that takes time even when the mill is willing. The nozzle pad was designed against the original thickness, so any substitute had to be rechecked against the pad calculation first. The customer's records from its last turnaround also showed a restart that had slipped, which made the schedule risk specific.
Three plans were compared on cost, schedule risk, and quality exposure, using the same units so the comparison could survive review. Expediting plate from a second mill added about 9 percent to package material cost and carried the lowest schedule risk, though the chemistry review cost roughly two weeks of engineering time. Splitting the package met the milestone for the critical exchanger only, and the remainder would have carried a delay penalty of a little under 4 percent of contract value. Resequencing around substitute plate cost nothing in material, but any later nozzle change would reopen the pad check and erase the saving.
- First weekPlate search widens and nozzle design is frozen for the pad check
- Second weekSecond-mill lot certified after chemistry and heat number review
- Third weekCutting and fit-up run on a second shift for the critical exchanger
- Final week before shutdownSkid tested and shipped ahead of the outage window
Actions taken
- Accept the second-mill plate offer on the condition that each lot ships with heat-traceable mill reports and a chemistry analysis before it leaves the mill.
- Freeze the nozzle redesign and run the pad calculation against the substitute thickness before any cutting, so the fabrication sequence does not reopen.
- Add a second shift in the shell bay for the critical exchanger only, funded from the expediting allowance, and keep the remaining packages on the original sequence.
- Agree with the customer in writing that a partial shipment will be accepted only if the critical section passes the outage acceptance test.
Results
The second-mill lot arrived in the second week and passed review after one mill report was reissued with a corrected heat number. Cutting began on the substitute plate in the third week, and the second shift held the critical exchanger on the revised path. The skid was tested and shipped before shutdown, meeting the milestone. Material cost ran about 9 percent over the original estimate, and the engineering review added roughly two weeks of effort, which the customer accepted as a variation. Other packages slipped a few days but were recovered later.
Questions for discussion
- What would have to be true about the second-mill lot for you to accept it without a further chemistry review?
- How would you weigh a few days of slip on the remaining packages against the premium paid for the faster plate?
- Which records from the customer's last turnaround would you request to test whether the restart risk is structural or a one-time event?
Case Study 5
Cutting annealing gas use without losing part properties
Greaves Heat Treat and Coil, a contract annealing shop of about 260 people serving automotive stamping customers.
The situation
Greaves ran three continuous annealing furnaces on steel coil and a batch bell furnace for smaller lots. Gas consumption per tonne had climbed over the past year, and waste heat from the main exhaust was vented to the stack with no recovery. Scrap from edge trim and off-spec coil had also grown, because a few customers had tightened their flatness and surface requirements. Management wanted a plan to cut energy and scrap intensity by a measurable amount without putting customer approvals at risk. Every change would have to pass the same metallurgical checks as the current process, since a shift in hardness or surface finish would send coil back.
The challenge
The team had three candidate changes: a recuperative air preheater on the main furnace, a switch of the bell furnace to a shorter cycle with a different atmosphere, and a change in the trim shear setup to reduce edge scrap. Each had a payback estimate, but the estimates were based on the vendor's nominal numbers. The maintenance records showed the existing burners were overdue for tune-ups, which affected baseline gas use.
Analysis
Comparing the options required one baseline that everyone agreed on. The gas meter was reliable but not normalized for gauge, line speed, or batch mix, so a heavy month would look worse even if the furnace ran well. The team tracked that intensity beside the metallurgical pass rate, so no saving could cost hardness or surface finish. Lifecycle cost also mattered, since a preheater adds fan wear, refractory upkeep, and its own electrical load, and those belonged in the comparison.
Each option was scored on gas and power saved per tonne, scrap reduction, expected effect on pass rate, and added maintenance load. The preheater promised the largest gas saving on the main furnace, though its benefit fell at low line speed and the fan load took back part of the gain. The shear setup change cut edge scrap with almost no energy cost and no metallurgical effect. The bell furnace atmosphere change was the most uncertain, since a shorter cycle could alter surface finish on thin lots and needed customer trials. The team kept the pass rate as the tiebreaker.
Actions taken
- Complete the overdue burner tune-ups on the main furnace and bell furnace, then measure gas per tonne for four consecutive weeks as the baseline.
- Reset the trim shear setup and add a width check at the exit, with a limit agreed with the metallurgical lab so that edge scrap falls without touching the coil profile.
- Scope the air preheater as a trial on one furnace zone with a fan sized for the lower line speeds, and require the maintenance plan and fan wear data before purchase.
- Run customer trial coils through the bell furnace shorter cycle, measuring surface finish and hardness against the approved sample, before any change to the recipe is released.
Results
Tune-ups lowered gas per tonne on the main furnace, and the revised baseline showed that part of the earlier climb came from degraded burners rather than product mix. The shear change cut edge scrap within the first month with no change in pass rate. The preheater went to a single-zone trial, with early numbers encouraging but short of the payback case. The bell furnace recipe passed on finish and hardness for most lots, and the thinner gauges were held back for another round of trials.
Questions for discussion
- Which of the normalizing factors for gas per tonne would you add first, and how would you check that it actually removes the product mix effect?
- What would make you accept a longer payback on the preheater in exchange for lower maintenance exposure?
- How would you decide when a customer trial result is strong enough to change a recipe for a gauge band that the customer buys only occasionally?
Learning Resources
Industrial CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.