Why the Center of Gravity Decides the Whole Lift
Why this matters
Weight gets looked up. The centre of gravity gets assumed, and the assumption is almost always that it sits at the middle of the box. That assumption produces two different failures, and only one of them is the famous one. The horizontal error makes the legs share unevenly, which is a capacity problem. The vertical error decides whether the load is stable at all, which is not a capacity problem and cannot be fixed by heavier slings. A load rigged below its own centre of gravity does not hang crooked. It turns over.
Three coordinates, and the one nobody measures
The centre of gravity is a point in three dimensions, and a lift plan needs all three of them.
The two horizontal coordinates get attention because their consequence is visible: a load that hangs nose-down or tips to one side is announcing that the hook is not over the centre of gravity, and a sibling article covers exactly what that offset does to the legs and to the swing.
The vertical coordinate is the one that gets skipped, and it is the one that is hardest to establish. A weighed-reaction method gives you the horizontal position directly and gives you the height only if you can weigh the load tilted, which is rarely practical in the field. So the height typically comes from reasoning about where the heavy components sit, and it arrives with a much wider bound than the horizontal coordinates do. Say so in the plan rather than writing a single number that looks as solid as the other two.
Where the mass actually sits is the whole question, and it is rarely at the geometric centre. A skid with a motor and a gear reducer at one end, a cabinet with a transformer at the top, a vessel with an internal heat exchanger in its lower third, a truck body with a fuel tank down one side: every one of those puts the centre of gravity somewhere the outline does not suggest. Draining a vessel moves it. Removing a shipping cover moves it. A field-installed accessory moves it, and moves the weight too.
The effective suspension point, and the stability rule
legs converge upward legs converge downward
to the hook below the pick points
* hook ==== beam ====
/ \ \ /
/ \ \ /
[lug] [lug] [lug] [lug]
| x CG | | \ / |
+-------+ | x | meeting point
| x CG |
+--------+
A suspended load hangs with its centre of gravity directly below the point where the lines of action of the supporting forces meet. Call that the effective suspension point. For a bridle running up to a single hook, it is the hook. For a rig whose legs are not parallel, it is wherever those leg lines intersect when extended, and that point can be somewhere no piece of hardware occupies.
The rule that follows is exact and it is a rule about position, not about force:
- Centre of gravity below the effective suspension point: stable. Disturb the load and it swings and returns. This is a pendulum.
- Centre of gravity above the effective suspension point: unstable. Any disturbance, however small, grows. This is an inverted pendulum, and it does not hesitate.
The clearance between the two is the margin. More clearance means a load that resists rolling and that tilts less for a given horizontal offset of the centre of gravity. Less clearance means a load that is easy to upset with a tag line, a gust or a boom movement. Zero clearance is not a boundary you approach and stop at; it is a condition where the load is deciding for itself.
Where the clearance goes negative
Three configurations put the meeting point at or below the centre of gravity, and all three are things crews build without noticing.
A spreader beam wider than the load's pick points. The legs then angle inward as they go down, so their lines of action meet at a virtual point below the pick points rather than above them, as in the right-hand side of the drawing. On a tall or top-heavy load that virtual point can land under the centre of gravity. Beams that are too wide feel safer than beams that are too narrow, which is why this one gets built.
Slings choked or basketed low on a tall load. Wrapping a tall cabinet or a vertical vessel around its lower body puts the whole support system beneath most of the mass.
Forks, cradles and any support from underneath. These do not have an effective suspension point at all, which is the next section.
Where you cannot establish that the meeting point sits above the centre of gravity, that is a stop, not a caution. The determination belongs to a qualified person, and where the load is engineered or the configuration is unusual, to a registered professional engineer. It does not belong to whoever is holding the sling.
Supported from below is a different problem
A load resting on forks or on a cradle is not suspended and the pendulum rule does not apply to it. What governs there is a moment balance about a tipping line, and the same principle appears in a form every shop already handles: the load centre on a lift truck.
A powered industrial truck's rated capacity is stated at a rated load centre, meaning a specified horizontal distance from the face of the forks to the load's centre of gravity. Move the centre of gravity further out and the load's overturning moment about the front axle rises, so capacity falls. The moment arm is the load centre plus the distance from the axle to the fork face, so the fall is not quite proportional to the load centre alone; the forklift load-centre card carries that arithmetic and why the simple inverse ratio is the conservative one to use. The truck's own data plate carries the capacity and the load centre it was rated at, and where a truck is rated at more than one configuration the plate carries each. That plate is the answer, and it is a legal requirement rather than a courtesy: 29 CFR 1910.178(a) requires the nameplate information to be maintained legible, and requires the manufacturer's prior written approval before any modification that affects capacity or safe operation, with the plate updated to match.
This is the same idea as everything else in the subject, stated in a different geometry: a rated capacity is a statement about one loading condition. The sling's rating assumes a vertical pull. The shackle's assumes an in-line pull. The truck's assumes a stated load centre. Every real job departs from at least one of them.
The centre of gravity block, filled in
Below is the centre of gravity block from a lift plan for a packaged skid, completed. The angle convention is stated because half the trade quotes it from the vertical: all sling angles here are measured from HORIZONTAL.
| Field | Entry | Source and bound |
|---|---|---|
| Load | packaged skid, 12.0 ft long, 5.0 ft wide, 7.0 ft tall | measured on site |
| Weight | 6,000 lb | weighed on load cells, not estimated |
| CG longitudinal | 7.2 ft from the left end | weighed reactions, bound plus or minus 0.1 ft |
| CG transverse | 2.5 ft from the left side, so centred | assumed from frame symmetry, bound plus or minus 0.4 ft |
| CG height above base | 3.2 ft | reasoned from component positions, bound plus or minus 0.8 ft |
| Pick points | two top-rail lugs, 5.0 ft apart, at 6.5 ft above base, both at the 7.2 ft longitudinal station | measured |
| Sling legs | 5.0 ft each, giving 60 degrees from horizontal | 2.5 ft half-spacing over a 5.0 ft leg |
| Apex height above the pick plane | 2.5 x tan 60 = 2.5 x 1.732 = 4.33 ft | geometry |
| Apex height above base | 6.5 + 4.33 = 10.83 ft | added |
| CG height used for the check | 4.0 ft | the HIGH end of the bound, 3.2 plus 0.8, because the high end shrinks the clearance |
| Apex-above-CG clearance | 10.83 - 4.0 = 6.83, carried as 6.8 ft | rounded down, because rounding a clearance down is the conservative direction |
| Stability determination | stable, with substantial clearance | apex above CG |
Now what the bounds cost, which is the reason a bound is recorded rather than a single value.
- Vertical share per leg with the hook exactly over the centre of gravity: 6,000 / 2 = 3,000 lb.
- Leg tension at that share: 3,000 / sin 60 = 3,000 / 0.866 = 3,464, carried as 3,470 lb because rounding a tension upward is the conservative direction.
- Now apply the transverse bound. The centre of gravity may sit up to 0.4 ft off the midpoint of a 5.0 ft pick spacing. The near leg's share is the total times the distance from the centre of gravity to the far leg, over the spacing: 6,000 x 2.9 / 5.0 = 3,480 lb. The far leg gets 6,000 x 2.1 / 5.0 = 2,520 lb. The mechanics of that split, and what it does to the load's attitude and its swing, belong to the sibling article on an off-centre hook; what matters here is the number it hands back.
- Leg tension at the bound: 3,480 / 0.866 = 4,018, carried as 4,020 lb.
- Sizing decision: the slings get sized on 4,020 lb, not on 3,470 lb. That is about 16 percent more sling than the centred case, and it is the price of a transverse coordinate that was assumed from symmetry rather than measured.
That last line is the whole argument for measuring. The bound is not a footnote on the plan; it is a capacity input, and it is the one that goes on the ticket. Narrow the bound and the rig gets lighter. Leave it wide and you either buy the margin or you find out where the centre of gravity was during the pick.
Do the trial lift with the load at a small clearance, everybody outside the arc the load can swing through, nobody under it and nobody touching it, and confirm the attitude matches what the block predicted. In construction that clearing duty for cranes and derricks is at 29 CFR 1926.1425; in general industry the load-handling requirements at 29 CFR 1910.179 carry it for overhead and gantry cranes. If the attitude does not match, land the load and block it before anyone approaches, because the block got something wrong and you do not yet know which line.
What the block does not settle
The block establishes where the mass is and whether the configuration is stable. Two things it deliberately leaves alone.
It does not establish that the lugs are adequate. A lug's capacity belongs to whoever designed or installed it, and where no rating exists the question routes to a registered professional engineer.
It does not stay true if the load changes. Drain a tank, pull a shipping brace, remove a panel, or add a field accessory and the block is stale in both weight and position. Date it, and re-run it when the load is not the load you measured.
References
- 29 CFR 1910.178, Powered industrial trucks (general industry): nameplate and capacity marking requirements at (a), including manufacturer approval for modifications affecting capacity
- 29 CFR 1926.1425, Keeping clear of the load (construction cranes and derricks), and 29 CFR 1910.179 for general industry overhead and gantry cranes
- ASME B30.20, Below-the-Hook Lifting Devices, in the edition adopted by your jurisdiction, your contract or your employer's programme: marking and rating of spreader beams and similar devices
- Equipment manufacturer's documentation: shipping weight, designated lifting points, and any published centre of gravity for the unit as configured
- See related: How to Find a Load Center of Gravity Without a Drawing; What Happens When the Hook Is Not Over the Center of Gravity