A crane accident scene is unstable in a way most failure scenes are not. The machine is often blocking a roadway or the only access to a project, several tons of load are lying where they fell, and the pressure to clear and restore service is immediate and legitimate. Recovery is also a heavy lift in its own right: it imposes new loads on already damaged structure and drives tracked equipment across the exact ground under examination. Most of what answers the causation question is perishable within days, some of it within hours.
The scene is a geometry problem first
Before anything moves, the record that matters most is spatial: where the machine sat, how far each outrigger was extended, the boom length and angle, where the load started and where it came to rest, and where the tipping axis ran relative to all of it. Those are the quantities a load chart is indexed by. Without them, capacity can only be assumed.
Total-station survey, photogrammetry or laser scanning captures it without touching anything, and aerial imagery taken before recovery is often the only record of pad positions.
The indicator memory has a clock on it
Rated capacity indicators and load moment indicators log load, radius, boom length and angle, and any alarm or cutoff activity. Their storage is finite and usually circular, so continued operation, repeated power cycling, or a technician connecting a service tool can overwrite or reset the record that describes the event.
Recovery almost always involves powering the machine, and diagnostic work on a damaged crane is normal before transport. Extracting the logger data, with the other parties on notice, belongs at the front of the sequence rather than after the machine reaches a yard.
What the indicator establishes, and what it does not
An indicator records what its sensors saw, which is not the same as what occurred. Load cells and pressure transducers drift out of calibration, an incorrectly entered configuration produces plausible readings against the wrong chart, and a bypassed or muted system says nothing about the warnings an operator actually received.
The data remains among the strongest evidence available, provided it is read alongside calibration history, the configuration entered, and whether any override was engaged.
Rigging is evidence: tag it, do not cut it
Slings, shackles, hooks, spreader beams and below-the-hook devices should leave the scene as found, with the assembly kept together and failed elements not cleaned, straightened or separated. Where something must be cut for extraction, the cut belongs well away from any fracture and should be marked as a recovery cut.
Fracture surfaces are the most fragile part of this. Handling, contact with surrounding debris, and corrosion over the days following an outdoor event degrade the features needed to distinguish an overload separation from a fatigue crack months in the making.
Wire rope: preserve the length, not the break
The parted ends are the obvious exhibit and the least informative in isolation. What carries the story is the distribution of broken wires along the rope, the pattern of wear and corrosion, crushing or improper spooling on the drum, and damage where the rope passed over a sheave or sharp edge.
Preserving substantial lengths on both sides of the failure, with the drum and sheaves, allows removal criteria to be applied to the rope as it actually was. A short sample cut around the break rarely supports a conclusion about retirement.
The ground is the most perishable item on site
Pad and mat impressions, the depth and shape of any settlement, soil moisture, and the condition of nearby excavations or backfill all change with weather and vanish under recovery equipment. Sampling and bearing capacity testing within the setup footprint have to happen before the area is regraded or staged on.
Where a machine tipped with no indication of overload, the ground is usually the question, and by then the setup area has often served as the working pad for the crane brought in to lift the wreck.
Recovery imposes loads of its own
Uprighting a machine, cutting a boom into transportable sections, and dragging a load off a structure all deform material that was already damaged. Later examination cannot reliably separate recovery damage from event damage unless the as-found condition was recorded first.
A documented recovery plan costs little: photographs at each stage, marked and logged cuts, and a record of what was lifted from where. It forestalls an argument about which deformation belongs to which event.
Records with expiry dates
Frequent and periodic inspection records, the lift plan, operator certification, assembly and disassembly documentation, maintenance history, and rental agreements are normally retained somewhere, but retention periods are finite and site paperwork is often held by a party with no stake in preserving it.
Electronic records age faster still: telematics from the crane or its carrier, site cameras, crew phone video, dispatch and delivery records, and meteorological data for the wind question. Preservation notices served early cost far less than a later fight over what was discarded.
Joint inspection and the order of work
Crane matters usually involve several parties at once: the operating employer, the crane owner or lessor, the general contractor, the rigging supplier and the manufacturer. Evidence examined by one before the others are on notice will be contested, whatever it shows.
A protocol fixing the order of work resolves most of that. Notice to all known parties, non-destructive documentation first, written agreement before any cleaning or sectioning, and a defined chain of custody. It also enforces the sequence a careful examination would have followed anyway.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.