WORK ORDER — Maintenance and Improvement Record
Work order no: WO-A-2026-01 Date issued: 2026-09-11 Category: Pneumatic fixture
Machine / cell: Drill Cell DC-4, vertical drilling station Status: Closed
Engineer: J. Radek (fixtures & tooling) Total labour: 9.5 hours Parts cost: GBP 214.00
Reason for work: recurrent part movement and scrap on the flange hole drilling operation
1. Fault description as logged by the shift team
Tuesday 06:10, operator record on WO ticket: the DC-4 fixture is letting the aluminium flange walk on the third clamp, part position drifts enough that the drilled hole runs 0.35 mm off the drawing centre, and the flange has scraped the drill bush twice this week. The clamp that walks is the rear pneumatic swing clamp, a 32 mm bore unit, and the rest of the fixture, two fixed end stops and a side pusher, holds the part fine. The process engineer has put the cell on reduced rate until the clamp recovers its holding truth.
Wednesday I arrived with the fixture drawing and the clamp datasheet, because a clamping complaint is usually two stories, the clamp’s rated force and the force it actually delivers at the pressure the plant runs. The swing clamp tag reads 1,050 N at 6 bar, but the regulator on this cell is set at 5.2 bar to share a header with a neighbouring ring crimper, and at 5.2 bar the same clamp promises about 900 N, already a twelve percent slip from nameplate. That alone should still hold a 0.9 kg aluminium flange, so the force number did not explain the walk, and I went looking for the geometry.
Evening note to the file: never trust a clamping force off a tag, the regulator pressure the plant actually runs is the force that matters, and I have the feeling the real story is parked on the clamp stem and the fixture pad, not the barrel.
2. On-site inspection record (Day 1 afternoon)
I pulled the DC-4 fixture apart on the bench, not before logging the air pressure at the clamp port with a test gauge: 5.2 bar at idle, dropping to 4.6 bar at the moment of clamp due to the ring crimper upstream cycling on the same header, so the swing clamp was seeing a jumping supply, and its effective force was flexing between roughly 800 and 900 N across the cycle. The part is a 6061-T6 aluminium flange, 190 mm diameter, 8 mm thick, mass 0.9 kg, machined on both faces, with a clean surface that should sit easily on three pads. Third, I checked the clamp geometry against the drawing and found the first honest fault: the rear swing clamp stem had 1.7 mm of end play at the bush, worn slop from two years of over-cycling, so the clamp foot was landing off-square on the part rim, tipping the flange and letting it creep out from under the pad under the drill thrust.
The second fault came off the pad faces: the rear pad under the swing clamp had been ground flat but the drawing calls a 0.4 mm radius on the pad corner, and the missing radius had bitten a burr into the flange rim on the previous setup, so the part was resting on a proud burr instead of the pad plane, rocking on a high spot roughly 0.15 mm proud and working loose under the 420 N drill feed thrust. Two faults, one worn clamp stem and one over-sharp pad corner, both small, both exactly the kind of detail a tag force figure cannot predict. The pad corner radius is the sort of thing a draughtsman draws at 0.4 millimetres and a grinder forgets at the wheel, and the workshop had faithfully ground the pad flat for two years while the drawing begged for the radius all along.
I measured the actual clamp force on the bench at the plant’s real pressures, because the datasheet numbers are a laboratory fiction until the bench confirms them: at 5.2 bar the 32 mm swing clamp delivered 896 N on the load cell, at 4.6 bar 812 N, and the drawing’s required minimum for this flange under 420 N feed thrust, working from the friction coefficient of 0.22 for aluminium on steel, came out at 640 N single-clamp, so even the weakest measured case had headroom, which confirmed the force was never the whole story and pushed the fix onto the geometry.
Evening note: the clamp was not weak, it was lying on a chattering pad and a worn stem, and a load cell at true plant pressure beats a datasheet table every time.
3. Corrective action taken
With both faults on the bench, the fix list wrote itself, and I kept the scope tight so the fixture went back tomorrow morning. First, the rear swing clamp stem and bush were replaced, new stem with a hardened bush pressed square to drawing, end play recovered from 1.7 mm to a measured 0.03 mm, all slop gone and the clamp foot now lands flat on the part rim. Second, the rear pad was re-ground and the drawing’s 0.4 mm corner radius put back on the stone, so the part cannot sit proud on a burr again, and I touched the two fixed pads with the same radius for consistency. Third, I re-levelled the fixture on its sub-plate, checking the three pad heights with a 25 micron feeler parked against the datums so the flange sits on one plane and not on a diagonal.
Fourth came the pressure hygiene, because a clamp force that jumps with the neighbour’s crimper is a clamp force doing half a job: the cell header was re-balanced so the drill cell gets a dedicated 6.0 bar feed from the dryer, the swing clamp now sees a steady 5.9 bar at the port instead of a fluttering 4.6 to 5.2, and at that pressure the load cell records 1,030 N, comfortably above the 896 N the fixture was making at its best before. Five, I set the clamp’s clamping force factor in the fixture instruction sheet to 1.3 times the calculated requirement, the plant’s standard safety margin for machining fixtures with interrupted cuts, so the 640 N minimum now sits against a 1,030 N installed force with the margin the drawing always meant.
Sixth was the little disipline that keeps a fixture honest: I wrote the expected clamp force and the measured value into the PM card, with the note re-check the pad radius and stem end play at the six-month service, because a fixture that is re-ground flat without its corner radius will spend the next two years quietly making scrap again. The whole corrective scope took 9.5 hours including the bench fixture pull-down and re-assembly, and the part cost, one stem, one bush, and a re-grind, came to GBP 214.00 against a scrap bill that had already cost the month’s production budget three times that number.
4. Acceptance test and result
Friday morning the fixture went back into the DC-4 cell and I ran the acceptance test the way the drawing asks: thirty flanges drilled in a row at full feed rate, 420 N thrust, 0.9 kg parts, and every drilled hole checked with a plug gauge for position against the datum. All thirty holes came in on the drawing centre, the worst run-out measured 0.03 mm against the previous 0.35 mm drift, and the last flange sat dead flat on the three pads with no burr rocking and no creep under the drill. The cell returned to full rate at 08:45, the process engineer signed the acceptance sheet with the note, fixture holding true, and the shift team has stopped logging the walk in the shift log.
5. Parts and labour summary
| Item | Description | Qty | Cost GBP |
|---|---|---|---|
| 1 | Swing clamp stem, 32 mm bore, hardened | 1 | 86.00 |
| 2 | Clamp bush, pressed, reamed | 1 | 34.00 |
| 3 | Pad re-grind with 0.4 mm corner radius | 3 | 94.00 |
| Total | 214.00 |
Labour: strip and bench inspection 2.5 h, corrective work 4.5 h, re-install and acceptance 2.5 h. Total 9.5 h.
6. Key measurements before and after
| Parameter | Before | After |
|---|---|---|
| Clamp port pressure | 4.6–5.2 bar | 5.9 bar steady |
| Clamp force at load cell | 812–896 N | 1,030 N |
| Clamp stem end play | 1.7 mm | 0.03 mm |
| Hole position drift | 0.35 mm | 0.03 mm max |
| Pad corner radius | none (flat) | 0.4 mm |
7. Foreman’s review note
The root cause was never a weak clamp. The fixture had a worn stem that tipped the pad off-square and a burr-bitten corner that rocked the part, and both were hiding behind a tag number that promised 1,050 N the workshop never actually saw at the plant’s real pressures. The lesson for the fixture file is written in the PM card: measure the clamp force at the pressure the cell truly runs, keep the pad geometry to drawing, and check the stem end play at service intervals, because a fixture that holds to 0.03 millimetre today is only a stem replacement away from drifting again. The DC-4 cell is back at full rate, the flange holes sit on centre, and the shift team has stopped adding walk entries to the log, which is the quietest sign of all that the fix took.
Attachment: fixture audit checklist
1. Confirm actual supply pressure at the clamp port, not the tag. 2. Load-cell the clamp force at that pressure. 3. Check stem and bush end play against drawing. 4. Inspect pad faces for corner radius and burr bite. 5. Verify pad heights sit on one plane. 6. Apply the plant clamp force safety factor of 1.3. 7. Log measured force and geometry onto the PM card.
Glossary
Swing clamp: a pneumatic clamp whose arm swings over the part then clamps down. Pad: the fixed support area the part rests on. End play: axial clearance in a stem/bush joint. Load cell: a force transducer used to verify clamp force. Feed thrust: the axial force of the drill during cutting.
Field calculation backing sheet — why 640 N was enough and 900 N still failed
For the record, the clamping force arithmetic that shaped this job: the drill applies a 420 N axial thrust on the flange, and the flange resists sliding by friction on the pads with a coefficient of 0.22 for aluminium on steel, so the required clamp load to hold the part stationary comes from dividing the worst-case tangential load by the coefficient and applying the 1.3 safety factor, giving the 640 N minimum written in the drawing. The measured swing clamp at its weakest moment delivered 812 N, so by the arithmetic the part should have held comfortably, and yet it walked, which is the whole point of the work order. The numbers failed because the model assumed two things the bench proved false: that the clamp foot lands flat on the rim, and that the part sits on the pad plane. With 1.7 mm of stem play the foot landed off-square, turning the 812 N into an angled contact that put only a fraction of the force through the pad, and with the missing corner radius the part rocked on a 0.15 mm proud burr, so the effective holding force dropped below the friction needed for a 420 N thrust even while the load cell said the barrel was pushing its rated number. The geometry turned a 1.3-margin clamp into a 0.8-margin clamp, and the fixture leaked position at the same rate a genuinely under-sized clamp would have. The fix did not add clamping capacity, it restored the geometry the force arithmetic assumes, and once the stem was true and the pads were flat with their radius back, the same barrel at a steadier pressure held the flange to 0.03 mm without a single walk across thirty parts. Any shop that chases a walking fixture with a bigger cylinder without checking the pad plane and the stem wear is spending money on the one link of the chain that was never weak.
