What a Chain Fall Tells You Before It Fails

Why this matters

A hand chain hoist is unusually honest equipment. Its load chain stretches measurably before it breaks, its hook opens measurably before it releases, its brake slips measurably before it lets go, and every one of those is a number you can take with a tape, a caliper and ten minutes. A chain fall that fails in service almost never failed without notice. It failed a measurement nobody took, on a tool that was judged by how it looked. This card is about the four quantities that carry the warning, and about why the visual judgement that people substitute for them is not just weaker but frequently backwards.

The four things that are gaugeable

Load chain pitch elongation. Chain under load yields before it fractures, and yielding shows up as pitch growth. Measure across a stated number of links, compare to the manufacturer's as-new gauge length for that hoist, and express the difference as a percentage of the as-new length. The gauge length and the allowable percentage both belong to the hoist manufacturer's manual, and ASME B30.16 for overhead hoists addresses inspection criteria in the edition your jurisdiction, contract or employer programme has adopted. Take both numbers from those documents. A remembered percentage is worth nothing here because the limits differ by chain grade and by manufacturer.

Link wear at the bearing surfaces. Where two links contact, material goes away, and the link section thins at the crown. This is measured as a reduction in the link's wire diameter at the worn point compared to the unworn part of the same link, again against the manufacturer's stated limit. It is a separate failure path from elongation and it can be at limit while the pitch is still fine, which is why both get measured.

Hook throat opening and twist. A hook that has been overloaded opens, and a hook that has been side-loaded twists. Both are measured against the manufacturer's criteria, and ASME B30.10 covers hooks under the same adoption condition. The reason this one matters out of proportion to its size is that an opened throat is direct evidence the hoist has been loaded beyond what it should have been, whatever the person who used it says.

Brake holding. A load-holding brake either holds a static load without creep or it does not. The test is to take a load, raise it a few inches, mark the position, hold it, and re-check after a stated interval. The acceptance criterion is the manufacturer's, and the test itself puts a load in the air, so everything the exclusion-zone card says applies while it is up: nobody under it, nobody in the line it would swing through, and it comes back down to the deck before anyone approaches. Do the test at a fraction of rating with a load whose weight you know, over ground that will take it, not over a bench and not over a customer's equipment.

Two hoists off the same rack

One gate, applied to two tools: the tool goes back on the rack only if every gaugeable quantity is inside the manufacturer's stated limit, whatever the tool looks like. The two resolve oppositely, and the direction is the surprise.

Hoist A came back from a coastal site. Surface rust across the housing, the load chain dull and discoloured, the hand chain grubby, the tag faded but legible. Every experienced person who walked past it wanted it off the rack.

• Load chain, 11 links measured: manufacturer's as-new gauge length for this model, taken from the manual, 12.60 in. Measured 12.72 in.
• Elongation: (12.72 - 12.60) / 12.60 = 0.95%, rounded up to 1.0% because a wear figure is never rounded down.
• Link wire diameter at the worn crowns: within the manufacturer's stated reduction limit at every link checked.
• Hook throat: unchanged against the manufacturer's criteria. No twist.
• Brake: held the test load with no measurable creep over the manual's stated interval.
• Call: back in service. It is ugly and it is inside limits. Surface rust on a housing is not a load-path defect, and the chain's discolouration is oxide, not section loss, which is exactly what the diameter measurement establishes.

Hoist B came back from an indoor plant job. Clean housing, bright chain, everything looked as it did the day it was issued.

• Load chain, 11 links measured: same model, same as-new gauge length of 12.60 in. Measured 12.85 in.
• Elongation: (12.85 - 12.60) / 12.60 = 1.98%, rounded up to 2.0%.
• Hook throat: opened against the manufacturer's criteria, and one link near the hook sits at a visible angle to its neighbours.
• Brake: held.
• Call: out of service, tagged, and the elongation figure written on the tag rather than "worn." Whether 2.0% is over or under the removal limit is the manufacturer's number, not one this card will supply, but the opened throat settles it independently, and the two findings together say the tool has seen a load it should not have.

The two results are the point. Appearance and gauge disagreed, in both directions, on the same rack, in the same week. The reason appearance runs backwards this often is that the conditions that make a tool look bad, weather and dirt, are not the conditions that damage a load path, while the condition that does damage it, a single overload by someone who did not report it, leaves a clean tool with a stretched chain.

What each measurement tells you about the tool's history

The measurements are worth more than a pass or a fail, because each one points at a different cause and therefore at a different fix.

Elongation spread evenly along the chain says the whole chain has been working near its limit for a long time, which is a sizing conversation about the jobs this hoist keeps getting sent to.

Elongation concentrated in a few links says one event, not a career. Those links spent time somewhere the others did not, which usually means an overload while a short section was in the sheave, or a shock. Find the event and the person, because whatever caused it will happen again.

Wear at the link crowns with little elongation says the chain is being worked over a worn load sheave, and a new chain in a worn pocket will wear at the same rate. Replacing the chain without looking at the sheave buys one cycle.

A twisted link, a link riding wrong, or a chain that will not run freely says the chain was reeved wrong or has been capsized. On a double-fall hoist a capsized chain will bind under load, which is a jam rather than a break, but a jam under load is a load you cannot control.

An opened hook throat with everything else in limits is the clearest single finding in the set: this tool was overloaded and nobody said so.

What is deliberately not on the list

Three things people check that do not decide anything, and one that would if it were possible.

Hand chain condition does not carry the load. It should run and it should not be worn to the point of jumping the wheel, but a scruffy hand chain is a nuisance, not a removal.

Housing paint, dents and surface corrosion are not load-path findings unless the damage is at a lug, a suspension point or a hook shank, in which case it is a structural finding and the tool goes to the person who inspects rigging rather than into a judgement on the rack.

The look of the chain is the one to strike explicitly, because it is the one everybody uses. Bright is not strong and dull is not weak.

The chain's remaining strength is the thing you would actually like to know, and there is no field method for it. Nothing you can do at the rack tells you how much margin is left; the measurements tell you only whether the tool is inside the limits its maker set. That is a real limitation and it is why the limits are conservative and why a tool that has been shocked comes out of service on the event rather than on a measurement.

Setting up so the measurement actually gets taken

The failure mode is never that somebody measured and got it wrong. It is that nobody measured, because the gauge length is in a manual that is in a truck that went to a different job.

Three things make it happen. Write the as-new gauge length and the number of links it spans on the hoist's own tag, at receipt, in permanent marker, so the comparison does not require the manual. Keep the last recorded elongation figure with the tool rather than in a file, so the person on the rack sees a trend rather than a single number, and a chain that moved from 1.0% to 1.6% in a quarter is a finding even if 1.6% is inside the limit. And require the figure in writing rather than a tick, because a tick can be applied from across the shop and a number cannot.

Then there is the part that costs nothing and catches the most: make the person who overloaded a tool able to say so without it going badly for them. Every opened hook throat on this rack arrived attached to a story nobody told. A tool that comes back tagged "I think I overdid it" is a tool that never reaches the next crew, and that is worth more than any inspection interval.

References

• 29 CFR 1910.184 (slings) and 29 CFR 1926.251 (rigging equipment for material handling) for the rigging in the same load path
• ASME B30.16 (overhead underhung and stationary hoists) and ASME B30.21 (lever hoists), in the edition your jurisdiction, contract or employer programme has adopted
• ASME B30.10 (hooks), same adoption condition, for throat opening and twist criteria
• The hoist manufacturer's manual, which is the governing source for as-new gauge length, allowable elongation, allowable wire-diameter reduction and brake acceptance
• See related: What a Come-Along and a Lever Hoist Are Rated to Do; What Shock Loading Does That a Static Load Does Not