2 Ton Chain Hoist Load Chain Wear: P and d on a CB020

Two failures cost money here, and they pull in opposite directions. Pull a CB020 chain that still has life and you pay for a part plus the downtime of a hoist that was working; keep a chain whose wire has already thinned and the next lift runs on less steel than the design assumed. Both mistakes come from one habit, judging wear by how the chain looks instead of by the two dimensions the manual lists.

On a Harrington CB020, load-chain wear is a two-criterion judgment. P and d are separate numbers with separate limits, and a chain can sit inside the P limit with d already past its discard value. One pitch measurement is not a valid pass. Everything below stays inside the CB020 row of the Harrington CB manual, Table 5-8; those figures are not a rule for every 2-ton hoist sold in the US.

What a Visual Check Misses on a CB020 Chain

Stretch is what people watch for, and stretch is what P measures. On the CB020 row, P is a span taken across five links, with a standard figure and a discard figure. That number tracks elongation: links bed into each other, bearing surfaces lose material, and the span grows. d is a different animal. It is the wire diameter at the diagrammed link, again with a standard and a discard figure.

A not-stretched verdict fails because the two wear modes do not announce themselves the same way. Elongation shows up as a gap you can see against a known span; loss of wire diameter shows up as nothing at all until the wire is thin enough to matter. The reason that asymmetry is dangerous comes down to what the load actually rides on, the cross-section of the wire rather than the outside profile of the link. Which means a chain can look acceptable and still fail a d comparison.

Suppose a warehouse tech checks a CB020 by eye, sees no obvious stretch, and signs it off. That chain may pass P and still read below the discard value at the diagrammed link. The eye check produced an answer to a question nobody asked.

Confirm the CB020 Row Before You Measure

The sequence starts before any measurement: read the product code, not the capacity. Table 5-8 lists CB005, CB010, CB015 with CB030, and CB020, each row carrying its own P and d values. CB015 and CB030 share a single row, which means capacity alone cannot pick the row, because two different product codes land on the same numbers. The reason that matters is that the discard values are attached to the product code, so a 2-ton label plus a guess about which row applies gives you a number with no authority behind it.

Once the code matches, confirm the definition before the reading. P is a measurement across five links; d is wire diameter at the diagrammed link. Both definitions point to an illustration in the manual, and the excerpt available here does not include that drawing. Without it, the exact start and end of the five-link span, and the exact link called out for d, stay unconfirmed.

Treat that as a documentation gap rather than a pass. If the diagram page is missing from your copy, get the correct page first. A span you picked because it looked reasonable is not a CB020 P reading, and a d taken at a random link is not the dimension the table is talking about.

Measure P Across Five Links on the CB020

With the span confirmed, P is a single comparison. Standard is 4.76 in across five links, and the manual's direction is that P should not be greater than the discard value of 4.91 in. A reading past that line is the point where the manual's answer is replacement.

4.76 in is a baseline, not a second threshold. No interim action level is printed beside it. Wear moves P in one direction only, since the span grows as links bed and bearing surfaces lose material, so a number slightly above standard is not halfway to failure in any formal sense. It tells you the chain has been working.

Suppose a chain measures 4.88 in. That is inside 4.91, so P passes. It says nothing about d, nothing about gouges, nothing about whether the chain is reeved correctly, and nothing about the load sheave. One number, one gate.

Why d Is a Separate Failure Criterion

P can pass while d fails.

The reason d is a separate criterion is that it measures loss of wire cross-section at a single link, while P measures accumulated elongation across five links, which means the two numbers respond to different damage and neither can stand in for the other. The CB020 row puts d at 0.32 in standard and 0.28 in at discard, and the manual's direction runs opposite to P: d should not be less than the discard value.

The physics is easy to see. Load capacity rides on cross-sectional area, and area scales with the square of diameter. Compare 0.28 in against 0.32 in and the squared ratio is about 0.77, so a chain sitting at the d discard line has roughly a quarter less wire cross-section than a new one, across the link that carries the load. P cannot show that, because a link can thin at the bearing points without the five-link span growing much. The two effects are not locked together.

Because the manual gives each dimension its own discard value, record P and d as two independent results. There is no averaged wear score in Table 5-8, and averaging the two would hide exactly the case that matters: a chain long enough to look fine and thin enough to fail. Two numbers, two verdicts.

Surface, Lubrication, and Reeving Checks Beyond P and d

Dimensions are only part of the check. The CB020 manual also gates the chain on surface condition, lubrication, and reeving, and a clean P and d reading does not cancel a defect in any of them.

Surface condition should be free of gouges, nicks, dents, weld spatter, and corrosion; links should not be deformed or show signs of abrasion, and the surfaces where links bear on one another should be free of significant wear. The manual's response to a failed surface item is replacement, not dressing the damage out.

Lubrication is its own gate: the entire surface of each link should carry lubricant and be free of dirt and grime, and the chain should not emit a cracking noise while hoisting a load. The noise clause is worth keeping in the record because it is an auditory signal a caliper cannot produce, which means a chain can sit inside both dimensional limits and still be telling you something while it lifts. Where lubrication has failed, the manual's answer is to clean and lubricate.

Reeving closes the visual set. The chain should run properly through the load sheave, and on multiple-fall hoists it should be installed with no twists. The same passage points at the load sheave for inspection, with the idle sheave added on multiple-fall units. Why sheaves belong in this check comes down to contact: a worn sheave pocket changes how the chain sits and how it wears, so the chain and the sheave behave as one wear system rather than two.

How to Confirm the CB020 Check Is Complete

A wear check is complete when the record ties four things together: the product code that picked the row, both dimensional results against their own discard values, the visual and auditory items, and the reeving and sheave findings. Because the numbers only carry meaning when they are attached to CB020 and its diagram, a bare measurement is not a complete check. It is one cell of one.

Practical form: one line per hoist, carrying the model code, the P reading, the d reading, the surface result, the lubrication and noise result, the reeving result, the sheave result, and a note that the diagram was on hand when the span was laid out. If any cell is empty, the record is incomplete. Incomplete is not the same as pass, and it should not be filed as one.

Which brings back the question most readers will still have. My hoist is a 2-ton, so can I use these numbers? No. The values in Table 5-8 are the CB020 row and nothing wider. A capacity label does not select the row; the product code does, and CB015 sharing a row with CB030 is enough to show it.

If the code on your hoist does not match CB020, or the diagram page is missing from the manual you have, the practical move is to get the correct OEM row and illustration before you compare anything. Until then you do not own a pass or a fail. You own a documentation gap, and the chain stays unjudged.