Why Shackle Side-Load Derating Tables Disagree With Each Other

Pull a shackle off-axis and its rated capacity drops. Everyone agrees on that. Ask by how much and you get different answers from sources that are all current, all published, and all correct.

The disagreement is not sloppiness in the secondary literature. It is written into the standard.

Scope: what this page covers, and what it does not

This page compares published side-load derating data for shackles and explains why the published sources differ. It does not tell you which number applies to your hardware, because that depends on who made it.

Three sourcing tiers appear below and they are labelled throughout, because the whole subject turns on the distinction:

• The standard. The figure and clauses quoted are from the 2015 edition of the rigging-hardware standard, read from the volume text. The current edition is newer and paywalled, and whether its numbers changed could not be checked. Both facts are stated again at the end.
• A manufacturer’s own catalog. Crosby’s tables are quoted from its own published catalog pages, which are linked and free to read.
• Distributors and suppliers. Named as such wherever quoted.

One clarification before the numbers. This page is about the shackle’s own capacity falling when loaded across the bow. It is not about redirect or snatch-block angle, which is a different mechanism concerning the force arriving at an anchor. This site covers that separately.

The three tables

What the standard publishes

Its side-loading figure gives four bands.

What Crosby publishes for shackles up to 3 inches

Its catalog gives seven bands for screw-pin and bolt-type shackles from 3/16 inch to 3 inches, up to 120 metric tons.

What Crosby publishes for its larger shackles

Above 3 inches, the same catalog page gives a different and much shorter table.

And above 125 metric tons with a side load greater than 5 degrees, the published instruction is to contact the manufacturer’s engineering department.

Where they actually diverge

The two main tables agree at the ends and disagree in the middle, which is where most real rigging sits.

Note the first row. At 8 degrees off-axis the standard has already applied its full 30 percent mid-band reduction while Crosby’s table says no reduction at all until past 10 degrees. That is a different shape of answer rather than a rounding difference: the standard uses coarse, conservative steps, and the manufacturer uses a finer curve fitted to hardware it designed and tested.

Why both can be right

The reason both tables can be correct is in the standard’s own wording.

The standard’s rigging-practices clause states that if a shackle is to be side loaded, the rated load shall be reduced according to its own figure or the recommendations of the manufacturer or a qualified person. That is an either/or, written into the requirement.

That clause alone is enough to explain the disagreement, and it is worth being precise that it does the work by itself.

A separate rated-load section says rated load shall be in accordance with the recommendation of the shackle manufacturer. That establishes where the baseline capacity comes from before any angle is applied. It is consistent with the picture, and it is not the clause that permits a manufacturer’s derating curve to replace the standard’s table; the rigging-practices clause is. Read together they point the same way, but only one of them is doing the permitting.

So the standard sets out a table and, in the same chapter, says the manufacturer’s recommendation may be used instead. Two published answers to one question, both sanctioned by the same document. That is the structure the literature was given rather than a failure of it.

The practical reading: the manufacturer’s data for your specific shackle is the more specific instruction, and the standard’s table is the fallback when you do not have it. What does not follow is averaging the two, or applying one maker’s curve to another maker’s hardware.

The position that says use neither generic table

A rigging distributor takes a firmer line and, notably, practices it. Its guidance is not to use generic derating factors, that only manufacturer-published derating applies, and to always confirm values against the specific manufacturer chart. Its page offers a single illustrative example rather than a table of its own.

A manufacturer states a compatible position: if a shackle is side loaded, the rated load must be reduced according to the manufacturer’s recommendation, applicable tables, or evaluation by a qualified person.

That was worth checking rather than assuming, because a source arguing against generic factors while publishing one would be a contradiction worth reporting. Neither does.

Meanwhile the compressed rule does circulate. A supplier’s shackle guide states that generally a 45 degree load implies a 30 percent loss and a 90 degree load implies 50 percent, and calls it a general guide only in the same sentence. Those two figures are real band boundaries from the standard. Dropping everything between them is what makes it a rule of thumb rather than a table.

What the off-road market publishes

Five off-road brands were checked. One publishes an angle derating for a specific product.

The Factor 55 figures are worth a second look. A drop from 12,000 to 8,400 lb is exactly 30 percent, which is the standard’s mid-band value applied across a 5-to-45-degree range. The bulletin also states that both figures are laser-engraved on the shackle for reference during rigging operations, which puts the derating on the part itself rather than in a document alone.

Four brands showing nothing is reported here as not found on the pages checked, not as proven absent. It was not an exhaustive audit of every technical manual.

What this page could not establish

The current edition’s numbers. The figure and clauses here were read from the 2015 edition. A newer edition exists and is paywalled, and whether the side-loading bands changed in it was not verified.

Whether Crosby’s newest catalog repeats page 94 unchanged. The 2017 and 2019 copyright editions were both read and match each other exactly. A 2023-copyright page from the same catalog was read and cross-references the side-loading section, confirming the doctrine still applies, but page 94 itself was not re-read in that edition.

One correction is worth recording, because it nearly became this page’s premise. A search summary attributed a five-band derating table to the distributor quoted above. Fetching the page directly showed no such table exists there; that source deliberately publishes none. The summary was wrong, and the page it described says close to the opposite.

How we sourced this

RiggingOps does not test hardware. Every number above is quoted from a published document and labelled with which tier it came from, because on this topic a standard’s table, a manufacturer’s catalog and a supplier’s rule of thumb are three different kinds of claim that look identical once someone retypes them into a blog post.

The standard’s figure was read from the volume text; because that copy circulates outside the publisher’s authorized distribution, it is not linked here, and the standard is cited by designation with the publisher’s own listing in the sources. Crosby’s tables come from Crosby’s own catalog pages, which are linked directly. Distributor and supplier positions are quoted and attributed as such.

If your shackle’s manufacturer publishes a side-load chart, that chart governs. If it does not, the standard’s table is the published fallback, and a qualified person is the answer the standard itself keeps pointing at.

Related reading

For what happens to a shackle that has been loaded past its rating, and why the published inspection criteria may not help you, see has my shackle been overloaded. For which volume covers which hardware in the first place, see which ASME standard covers a recovery hook. For the ratings themselves, WLL vs MBS. And for the different angle problem — force arriving at an anchor rather than capacity lost in a shackle — see winch redirect anchor force.

This page is part of the broader recovery gear ratings pillar.