The slump test, and why water is the real argument

A truncated steel cone is filled with fresh concrete in layers, each layer rodded, the top struck off level. The cone is lifted straight up. The concrete settles. The distance the top of it drops, measured in millimetres, is the slump.

That is the entire test. It takes about two minutes and needs a cone, a rod, a flat plate and a ruler. It is the cheapest quality check on any construction site and one of the most frequently skipped.

What it is measuring#

Workability. How easily the concrete can be placed and compacted without segregating.

Slump is not strength, and it does not measure strength. What makes it important is that the easiest way to increase slump on site is to add water, and adding water reduces strength. So slump is the observable that stands in front of the invisible thing you actually care about.

A specification will name a slump range appropriate to the work: stiffer for something that must hold its shape, wetter for a heavily reinforced section where the concrete has to flow between bars. Both the top and the bottom of that range matter. Too low and it will not compact, leaving voids. Too high and either water has been added or the mix has been designed wet, and in the first case the strength has already gone.

The argument, in its usual form#

Concrete arrives. Somebody looks at it and says it is too stiff to place. A hose appears. Two buckets of water go in, the drum turns, and the load becomes much easier to handle.

Everybody in that conversation is behaving reasonably. The person placing the concrete has a real problem — stiff concrete in a congested section genuinely will not compact, and badly compacted concrete is also weak. The driver wants to discharge and leave. Nobody involved is trying to weaken the building.

What has happened is that a decision about strength has been made by the person holding the hose, on the basis of how the mix looks, with no record.

The correct answer is almost never water. It is either an admixture, which is what admixtures exist for, or a mix designed for the placement conditions in the first place, or a different pour sequence. All of those require somebody to have thought about it before the lorry arrived.

Where and when to test#

At the point of discharge, from the load you are about to place, before it goes in. Not at the plant, not from the first spill, and not after water has been added — which sounds obvious and is the most common way the test is defeated.

There is a timing subtlety worth knowing: concrete loses slump as it stands. A load tested an hour after batching will read lower than the same load at the plant, without anybody having done anything wrong. If the site is a long way from the plant or the truck queued, that loss is real and expected. The specification and the mix design have to accommodate it, which is a conversation with the supplier and not a reason for a hose.

Reading the shape, not just the number#

The way the concrete slumps tells you something the ruler does not.

• A true slump, where the mass settles evenly and keeps its shape, is what you want to see.
• A shear slump, where one side slides off, suggests a mix that is not cohesive. The number is unreliable and the test should be repeated.
• A collapse, where it spreads out flat, means the mix is far too wet for this test to say anything useful about it.

Two loads can produce the same reading with completely different behaviour in the shutter. Somebody who has watched a hundred slump tests learns more from the four seconds after the cone lifts than from the measurement.

Why it is skipped#

Because it is the only quality check that costs the site something at the exact moment it is performed.

Everything else about concrete quality is checked later or elsewhere. The cube test happens now and reports in four weeks, so it costs nothing today. The slump test happens now, reports now, and its result may require somebody to reject a load with a queue of lorries behind it and a gang standing idle.

That is the honest reason it does not get done. Not ignorance. A test that can stop work is a test people find reasons not to run, and the reasons are always credible on the day.

Making it survive contact with a real site#

The only version of this that works is one where recording a slump is easier than not recording it, and where somebody other than the person under pressure sees the number.

That means the slump belongs on the same record as the load itself — the same row that carries the challan number, the vehicle, the grade, the volume and the receiving officer. That row is written at the barrier, which is why it matters what the gate has been told before the lorry arrives — why a lorry surprises the gate. Not in a separate register that a separate person keeps. A field on a form that already has to be filled costs nothing to fill; a second register costs a walk to the office, and the walk is what actually decides whether it happens.

It also means the number has to be visible afterwards. A slump reading that disappears into a file is a reading that only ever gets looked at during a dispute. Held beside the grade and the cube results for the same load, it becomes something else: the first place anybody looks when a twenty-eight-day result comes back low. High slump on arrival and a poor cube result four weeks later is a pattern that explains itself, and you can only see the pattern if the two numbers live in the same place.

This is a small instance of a much more general point about site records — that a record which requires an extra journey is a record that exists in theory. We have written about the same problem in the context of a gate register somebody will actually keep.

The short version#

Slump is a two-minute measurement of workability that stands in for the thing you cannot see, which is how much water went into the concrete.

Test at discharge, before anything is added. Watch the shape as well as the ruler. Write the number on the same row as everything else about that load.

And when somebody reaches for a hose, understand what is actually being decided, and by whom.