Test weights, chains and billets: the gear your calibration depends on

The three references sit on a ladder. A static weight is the quickest and most repeatable. A calibration chain lays a load more like real material. A billet system builds a weight into the scale so a span check needs nothing lifted on and off. Each is a good tool, and each has care and handling that decides whether the number it gives you is worth anything.

This page is about the gear itself: what each reference is for, how to look after it, and the traps we see ruin real calibrations in the field. For how the three methods compare and when to use each, see belt weigher calibration methods.

When are static test weights the right reference?

A static test weight is a known mass hung on the weigh frame to apply a known force. It is the workhorse reference for routine calibration: quick, repeatable, and honest on a sound, well located scale. The whole idea only works if the weight is genuinely the mass stamped on it, of known and verified mass, because everything downstream is set to match it.

That is where care matters more than people expect. A test weight leads a hard life around a plant, and small changes to its mass are invisible until they show up as a span error:

• Corrosion. Rust flakes away metal over time, so a rusted weight is lighter than its stamp. Water sitting in a pitted surface adds mass one week and drops it the next.
• Paint build-up. A weight repainted every few years to keep the rust off slowly gains a coat of extra mass. A little each time adds up.
• Damage. A weight that has been dropped, chipped, or ground on can lose or gain metal. Once it is no longer the number on its side, it is not a reference any more, it is just a lump.
• Wear from use. Lifting eyes, edges and seating faces wear with handling and repeated application. A worn seating face can change how the weight sits on the frame even when its mass is still right.
• Storage. Kept off the ground, out of the weather, dry and supported so it cannot fall. A weight stored in mud or left in the rain is corroding while it waits for the next visit.

There is a second, quieter care point: the weight has to be applied the way the scale expects. A weight hung off the right point on the frame acts through the frame's levers, so it has to sit where the geometry intends, not wherever is convenient. A right weight in the wrong position is as misleading as a wrong weight.

The rule of thumb is simple. If a weight has been dropped, rusted, or repainted enough to doubt it, stop trusting the stamp and get it checked. A calibration against a wrong weight looks perfect on the day and is wrong in production. Because the stamp carries so much of the work, weights are worth re-checking against a known standard now and then, not just eyeing them for rust. A reference of known, verified mass is the anchor for everything the scale reads, and confirming it is cheap next to a season of drifted tonnes.

What is a calibration chain and what ruins one?

A calibration chain is a rolling reference. It is a length of chain of known mass per metre, laid across the weigh span so it loads the belt the way spread out material does, and it rolls along with the belt instead of sitting on one point. That makes it a closer stand-in for a real distributed load than a weight hung on the frame, which is exactly why some sites prefer it.

A chain earns that only if it rolls freely and cleanly. Its traps are physical:

• Bent or damaged links. Chains are heavy and awkward, and most of the damage happens in transport and handling, not in service. A dropped or kinked chain no longer lies or rolls the way it should.
• Seized rollers. If the rollers do not turn, the chain drags instead of rolling, and it stops behaving like the distributed load it is meant to imitate.
• Mud and build-up. Caked material adds mass the chain is not supposed to have, so the reference itself is heavier than its rating. Clean it before you trust it.
• The right rating for the span. A chain is only a reference at its stated mass per metre laid across the correct span. The right chain has to be matched to the scale, not just any chain long enough to reach.
• Storage and handling. Stored coiled and supported, off the ground and dry, and moved with the weight and the pinch points respected. It is easy to hurt both the chain and the person carrying it.

Before we trust a span set with a chain, we confirm the chain hangs and rolls free across the whole span. A chain that binds or drags is giving the frame a load it was never meant to see, and a chain caked in material or missing a working roller is not the mass its rating promises.

Why do billet systems need watching so closely?

A billet system is a test weight that lives on the scale permanently and is applied to the frame by a lever or cam arrangement, so a span check can be done without anyone lifting weights on and off. It is a genuinely good idea, and on a well set up scale it makes routine calibration faster and safer. It also comes with a handful of specific traps we see cause real calibration damage in the field, and they are worth knowing because the scale will not warn you about any of them.

These traps are inherent to the underhung style of billet, where the weight is applied to the frame through a chain and a cam. They are not a fault of any one brand. Arrangements like this appear on a number of Australian scales, including some CST designed systems, which are widespread, well regarded, and which we service and calibrate ourselves (see servicing CST belt weighers). The point below is not that the gear is bad. It is that this style has quirks a technician has to watch for, and a site that does not know them can end up trusting a bad number.

Three traps come up often on this style, and all three leave the reading looking perfectly normal while the calibration quietly goes wrong:

• The cam can twist as it applies. On an underhung billet on a chain, the cam can rotate slightly as it engages and put tension into the chain. That tension pre-loads the weigh frame, and from that day the scale sits on a higher zero than it should. The zero looks stable because it is consistently wrong, so nothing flags it until someone checks against a clean reference.
• The chain can foul while the billet is applied. If the chain is catching on the structure, on build-up, or on the frame itself as the billet comes on, the force reaching the frame is not the force it should be. The span then shifts significantly. If the site does not know what a normal span looks like for that scale, the bad number gets accepted, and the calibration stays ruined until someone who knows better checks it.
• The frame collects what falls on it. This style often runs long, flat frame sections sitting under the belt, and those sections collect spillage and build-up constantly. That extra mass loads the frame and walks the zero between services, so the scale drifts even when nothing electrical has changed. More on that in why the zero keeps drifting.

The common thread is that a billet can hide its own faults. The reading looks calm whether the applied load is right or wrong, so the only way to catch a twisting cam, a fouling chain or a loaded frame is to watch the scale as the billet goes on and comes off, and to know what a healthy span looks like for that machine. That is field knowledge, not a menu setting, which is why an experienced set of eyes on the billet is worth having.

What do we do about the reference gear on a visit?

The reference gear is the first thing we inspect, not the last, because a calibration is only as good as what it is measured against. On a visit that involves any of this gear, we:

• Inspect the reference before we use it. Test weights checked for rust, damage and heavy paint. Chains checked for bent links, seized rollers and build-up. Suspect gear is called out, not quietly trusted.
• Prove a chain hangs and rolls free before we believe any span it sets, so the reference is actually behaving like a distributed load.
• Match the reference to the scale. The right test weight or the right chain rating for that frame and span, so the number we measure against is the number the scale should actually see.
• Watch the zero when a billet is applied and released. A zero that shifts as the billet comes on or off is telling you something is binding, twisting, or fouling. That is a signal, not a nuisance, and we chase it before setting anything. We also check the billet's chain path is clear before we lean on any span it sets.
• Clean the frame before we calibrate. Setting a span against a dirty or loaded frame just calibrates the build-up in. First we get it clean and free, then we measure.
• Record which reference was used. Noting whether a visit used weights, a chain, or the billet lets drift between visits be traced back to the gear or the scale, instead of guessing. If the reference changed, the comparison changed with it. That note lives with the calibration figures in AIMM, our asset management system, for every scale we look after.

None of this is exotic. It is the discipline of treating the reference as seriously as the scale, because the reference is what decides whether the calibration means anything. A site that knows what a normal span looks like on its own scales, and keeps its weights and chains in good order, is a site whose tonnes hold up between visits. For the wider set of things that quietly throw a scale off, see common belt weigher faults.

Frequently asked questions