A thread about a vertical conveyor that moved five millimetre a minute while it claimed to be stopped, and how the brake torque margin got checked against the real duty cycle. Names trimmed.
From: J. Groves, Safety Officer — To: R. Almeida, Design; cc plant manager
Subject: Incident note VCL-9 — load crept 300 mm in a power-off stop, brake did not smoke
Ricardo, recording this before it becomes folklore. On the 7.5 kW vertical conveyor the pallet platform crept about 300 millimetres down over ten minutes while the machine sat in a powered-down state, brake engaged, no alarm, nobody inside the cage at the time, and the night shift found it by the taped mark. The spring-applied brake on the motor holds the datasheet torque of 45 Nm, the load plus platform and chain weighs the output at about 18 Nm referred through the gearbox, so by the arithmetic the brake should hold four times the load, and I want to know where the factor went, because a creep that slow with a brake that is not slipping tells me the brake is engaged, the torque is there, and something in the chain is still turning anyway, which is a worse story than a worn brake disc.
From: R. Almeida, Design — To: J. Groves, M. Sorensen, Brake Supplier; cc plant manager
Subject: RE: VCL-9 — brake torque is not the only hold point
Josh, good instinct, the arithmetic said hold and the platform moved, so the slip is downstream of the brake. A spring-applied motor brake holds the motor shaft, and if anything between the motor and the load can wind, creep or ratchet through a backstop, the brake torque is irrelevant to the gross motion, it only holds its own shaft. The suspects on this drive train: the brake on the motor, then a flexible coupling, then the worm gearbox, then the chain to the carriage. A worm gear holds the load through friction when stationary, so if your unit is not a self-locking worm but a helical, the load is held only by the brake, and if the gearbox has a backstop it may be worn or deliberately slipped. I have asked the brake supplier three questions and I want answering before we touch anything: what the actual static holding torque of this brake measures at this air gap, what the gearbox drive back-drive torque is, and whether the creep we saw can pass through the coupling. The brake datasheet is a number printed in a catalogue; the creep is a fact printed on a tape mark, and the tape wins.
From: M. Sorensen, Brake Supplier — To: R. Almeida, J. Groves; cc plant manager
Subject: RE: VCL-9 — measure the gap, then read the catalogue again
Ricardo, three answers, and the first one is a habit more than a product detail. Static holding torque of a spring-applied brake is quoted at a specific air gap and a fresh liner, and the gap on your unit after a motor run-in has almost certainly opened past the fresh setting, because every energise-and-release cycle wears the friction liner and the armature sets the gap. A 45 Nm datasheet brake fitted with the gap at the wear limit can deliver materially less than its printed torque, and the creep you saw is consistent with the brake holding most of the time and yielding slowly under a load that is nowhere near its rating. On the gearbox question, if your drive is a helical unit, the back-drive is real, the load pushes through the gear train to the low-speed shaft and the brake has to hold the full reflected torque with nothing helping it; a worm box at under, say, a 60 to 1 ratio is not reliably self-locking either once worn. So the brake torque margin is not your belt and braces, it is your only line, and on a vertical axis it should be two to three times the reflected load torque, not the comfortable 2.5 times the datasheet says if the real torque has quietly halved with wear. Get me two measurements: the actual air gap on the brake, and the reflected back-drive torque at the motor shaft with a torque wrench and the carriage at mid-travel, and we will know which door the creep walked through.
From: R. Almeida, Design — To: J. Groves, M. Sorensen; cc safety committee
Subject: RE: the two measurements came back, and the answer is both doors
Measurements are in, and neither of us gets to be smug. Air gap on the brake measured 0.62 millimetre against the 0.3 millimetre fresh spec, so the delivered static torque is somewhere in the low thirties, I will put a bench figure in later, well under the 45 Nm nameplate. The back-drive torque at the motor shaft with the carriage loaded, measured with the brake released, came to 21 Nm on a still morning with the chains cleaned, which is above my 18 Nm estimate because the platform guide friction is less than the drawing assumes when everything is greased and warm. So the brake is delivering perhaps 30 to 32 Nm against a back-drive of 21 Nm that the 45 Nm datasheet was supposed to swallow, and the margin has gone from three times the load to one and a half. The creep is both a worn-gap brake and a lightly-rated margin, and the fix is not one thing, it is regap the brake, set the maintenance scheduled check, and recompute the margin with the real numbers so we stop quoting the datasheet in the safety file.
From: J. Groves, Safety — To: R. Almeida; cc plant manager, maintenance superintendent
Subject: RE: I want a bigger brake and a scheduled proof, same week
Ricardo, I am not comfortable holding a vertical axis on a brake that delivered 30 Nm against a nameplate of 45 Nm without at least a proof test in the same week as the regap. A vertical conveyor is the classic place where a brake failure is not an incident, it is a drop, and I am asking for three things in writing: regap and reline the brake now, add the brake holding test to the weekly checklist with a torque wrench pass/fail at 1.8 times reflected load, and rework the safety file to quote the measured 32 Nm expected torque and the 21 Nm back-drive as the basis for the margin, not the catalogue page. If the rework will not fit before the scheduled line window, I will park the conveyor and eat the line outage, because a tape-marked creep that nobody believes is a warning I am not going to re-gift the night shift.
From: L. Fontaine, Maintenance Supt. — To: J. Groves, R. Almeida; cc plant manager
Subject: RE: prove the fix, not the drama — regap is a four-hour job, new brake is a twelve-day lead
Josh, the regap, reline and proof can be done in one four-hour window on Saturday, and I have booked it. A bigger spring-applied brake is a twelve-day lead and a motor rework because the bracket and shaft extension change, and it buys the same thing the regap buys if the gap is the disease, which our measurement says it is. The one number I want on the record before Saturday: after regap, the brake should measure back near its nominal at fresh gap, and if the bench figure stays under 38 Nm we will start the brake replace conversation with the supplier immediately, because at that point it is a worn liner speaking, not a gap. My position is prove the 45 Nm unit is a 45 Nm unit again before we order a bigger one to make up for a schedule we can all count.
From: R. Almeida, Design — To: J. Groves, L. Fontaine, M. Sorensen; cc plant manager
Subject: RE: agreed — regap plus proof, and holding vs dynamic on the drawing
Agreed with both of you, and one clarification so the drawing records what this brake does and does not do. A spring-applied brake on a vertical drive is a holding device for the stopped condition; it is not guaranteed to stop a moving load from speed and it is not a dynamic brake unless the application says so, because the energy of a loaded descent can exceed what the friction face can absorb in one engagement. If we ever add an emergency stop from motion, the drive must first ramp and the brake engages after, not during, and I want that sequencing note on the brake detail so a future electrician does not wire the brake to the E-Stop as the sole stopper. The torque margin stays the design rule: two and a half to three times the reflected back-drive at the motor shaft, verified by test at the working air gap, quoted with the real numbers, and re-checked whenever the liner or the gap moves.
From: L. Fontaine, Maintenance — To: J. Groves, R. Almeida; cc M. Sorensen
Subject: RE: Saturday proof — 43 Nm at fresh gap, creep gone, log attached
Regap done in the four-hour window, liner inspected with two-thirds life remaining, and the bench torque after regap measured 43 Nm at the 0.31 millimetre gap against the 45 Nm nominal, which takes the margin to 2.0 against the measured 21 Nm back-drive, adequate and documented. Static holding test passed at 1.8 times reflected load as agreed, the platform sat dead for twenty minutes through a powered-off night stand with the taped mark untouched, and the creep has not reappeared across two shifts. The weekly holding test is in the checklist with the torque wrench setting written on the sheet, and the safety file now carries the measured numbers instead of the catalogue page, so next year’s auditor gets the same statement this week’s crew got: 43 Nm tested, 21 Nm required, every Thursday, pass or park.
From: R. Almeida, Design — To: all — Subject: RE: close-out, the numbered lessons
Closing VCL-9 with four lines for the file. First, a spring-applied brake torque is a condition, not a constant, and the air gap is the temperature it is measured at, so gap checks belong on the PM schedule, not the fault list. Second, on a vertical axis the brake is the only hold point if the gearbox back-drives, so verify the back-drive with a wrench and size the margin against the measured figure, not the drawing estimate. Third, holding and dynamic braking are different jobs; the brake stops the stopped machine from moving, the drive stops the moving machine from staying moved, and the sequence is not optional. Fourth, a tape mark is a measurement, a datasheet is a hope, and when the two disagree the tape has already told you which one to trust.
Reference data — 7.5 kW vertical conveyor; reflected load 18 Nm drawing, 21 Nm measured; brake nominal 45 Nm; fresh gap 0.30 mm, wear limit ~0.7 mm; measured pre-gap ~30-32 Nm; post-gap 43 Nm at 0.31 mm; margin 2.0 against measured back-drive; holding test 1.8 x load; creep 300 mm over 10 min.
Glossary — spring-applied brake: a brake held closed by springs, released by an electromagnet, so it engages on power loss, the fail-safe direction; reflected torque: load torque referred through the gearbox to the motor shaft; self-locking worm: a worm set whose friction holds the driven load at rest; back-drive: the tendency of a non-locking gearbox to transmit load torque back to the high-speed shaft; holding torque: the static torque a brake can sustain while stopped, distinct from dynamic stopping torque.
Field note — the brake that held, and the one that only thought it would
A vertical axis that creeps with the brake engaged has lost its margin somewhere between the catalogue and the shaft. In this case it was the air gap, a wear-driven growth that quietly cut a 45 Nm brake to thirty-something while everyone quoted the nameplate, and the real load pulled another three Nm over the drawing estimate, so the comfortable three-fold margin turned into a false one and a half. The fix is boring and it is the whole job: measure the gap, measure the back-drive, regap, and put the holding test on a schedule, because the difference between a safety brake and a brake that only thinks it is a safety brake is a torque wrench and a written number renewed on a calendar.
