Concrete block dimension tolerance: mould wear, or a compaction problem?
Measure length and width before you look at anything else. If the plan dimensions are holding and only the height is wandering, the mould is almost certainly not your cause — and a replacement mould set will arrive months later and reproduce the same variation. Height is not a dimension the mould cuts. It comes out of feed volume, compaction energy and the height-control system, on whatever pallet happened to sit under the mould that cycle.
That one split — plan dimensions against height — settles most of these arguments before anybody writes a purchase requisition.
Length and width are a mould question. Block height variation usually is not
Cavity walls and division plates cut length and width. They wear in one direction only: outward, off the product face, slowly. So plan-dimension drift has a signature — it is monotonic. A block cast this month is never smaller than one cast last year out of the same cavity. Growth, not scatter.
If your length figures scatter both ways week to week, you have a measurement problem or a moisture and curing problem, not a worn cavity. Steel does not grow back.
Height behaves nothing like that. Each cycle it is set by how much material the feed drawer drops, how far that material consolidates under vibration, and where the machine decides to stop the tamper head. All three move during a shift. Height scatters in both directions by nature; the only questions are how wide the band is and whether the band is drifting.
Map the pallet before you map the mould
Number the cavities. Measure every unit off one pallet at the same three points, then repeat for ten consecutive pallets — same cure age, height gauge, not a tape. Two patterns fall out of that data, and they point in opposite directions.
Variation tied to a position — two cavities over one corner always low, one long edge always low — is an energy-distribution problem. Table, clamping, pallet support, vibrator condition. The mould is innocent.
Variation tied to a cavity, following that cavity across every pallet regardless of where it sits, is tooling. Now you have something worth photographing.
Measure the production pallets themselves while you are at it. Thickness variation and bow transfer into block height one for one. Steel pallets take a set, timber pallets swell wet and shrink dry, and a mixed-age pallet fleet gives you a height band that no machine setting will ever close. It is the cheapest check in the building and the one that gets skipped.
Which symptom points where
| What the measurements show | Where it points | What the wrong call costs you |
|---|---|---|
| Length and width have crept larger over months; scatter unchanged | Cavity liner wear — genuine mould job | Retuning vibration cannot pull back a dimension the machine does not control. You scrap for another year. |
| Only height varies; plan dimensions steady | Feed, compaction or pallets | A new mould reproduces the same height band, and the lead time is gone. |
| The same cavity is out on every pallet | That cavity, its division plate or its shoe | Machine retuning drags the good cavities off target to average out one bad one. |
| The same pallet position is out regardless of cavity | Clamping, pallet flatness, energy distribution | Tooling money spent; the corner cavities stay low. |
| Band widens through the shift, tightens after a batching change | Mix moisture and feed control | Nothing you buy fixes it. |
| Torn top faces and escaping fines alongside height movement | Tamper head to cavity clearance | A new mould against a worn head reopens the gap within weeks. |
| Flash or fins along a division plate joint | Joint faces worn, or a cracked liner seat | Longer vibration times make the flash worse, not better. |
What mould wear looks like once the tooling is on the bench
Pull it, clean it, and look for five things.
- Cavity faces polished bright with directional scoring, and arrises that should be sharp carrying a radius.
- Flash at division plate joints — the joint faces have lost contact with each other.
- Clearance between the tamper head shoes and the cavity walls, feeler-gauged against the builder's figure. When that gap opens, fines blow up past the shoe, the top face tears, and height starts to move. This is the one case where a real tooling fault presents as a height fault — but it never presents alone. The torn face comes with it.
- The mould's contact face to the vibrating table. If it is no longer flat, or the clamping surfaces are hammered, the mould is not rigidly coupled to the table and you lose amplitude whatever the vibrator is doing.
- Cracks at liner seats and welds, particularly where a hard aggregate has been running.
Mould geometry is builder-specific: clamping, pallet size, table interface. Hess Group tooling does not drop onto another builder's table, and the spec you quote from has to match the machine standing on your floor rather than the group standard your head office holds. Which platforms and tooling run on what is set out on the equipment pages.
Where the compaction energy goes missing
Assume the vibrator is healthy until proven otherwise. It usually is. Energy gets lost between the vibrator and the mix.
- Mould clamping pressure. A mould held by tired hydraulic clamps or hammered clamp jaws absorbs vibration instead of transmitting it. This is the most common source of a height band that has "always been there".
- Table rubber or spring elements. They harden and settle. Amplitude falls quietly across years, and nobody notices because nothing ever changed suddenly.
- Vibrator bearings, belt slip, and unbalance weights that somebody adjusted two years ago and did not write down.
- Feed drawer timing and agitator wear. An uneven fill front to back becomes an uneven height front to back — and on a position map it looks exactly like a mould problem.
- Mix moisture. Height tracks moisture more tightly than it tracks any machine setting. If your band opens in the afternoon and closes after a batching correction, stop looking at steel.
Put the shift log next to the height data: vibration time, feed time, moisture reading, pallet batch. If the height curve follows the moisture curve, no tooling purchase changes anything. Tamper head refurbishment, clamp overhaul and vibrator work are service jobs and take a fraction of a mould's lead time — which matters when head office approval is the long pole in your schedule.
The figures a mould enquiry needs, and the two that are always blank
The first missing figure is cycle count on the existing mould. Without it nobody can say whether the tooling died early or lived a normal life, and that decides whether you re-order the same specification or change it.
The second is the abrasion character of your aggregate. Ordering like-for-like after a short mould life, when the wear pattern says the liner grade was under-specified for what you are running, buys you another short mould life. Wear pattern, aggregate and cycle count read together tell you whether the specification failed rather than the steel.
The rest of what a quotation needs, and what usually has to be chased: machine builder, model and year; pallet dimensions and thickness, including the measured spread across the fleet rather than the nominal figure; the product height range to be run in this mould; division plate layout; tamper head condition and date of last refurbishment; and which tolerance basis you are certifying to — ASTM C90 works to a fixed allowance on the standard dimensions, while EN 771-3 works from declared tolerance categories, so "out of tolerance" means different things depending on the answer.
One misjudgement is worth naming on its own: replacing the mould and leaving the tamper head alone. Fresh cavity walls against a worn head give you a clearance mismatch on day one, and the height band you paid to close is back inside a month. Send your tolerance measurements — cavity by cavity, with pallet position noted — together with the mould's service history and cycle count, and you get back a read on whether this is a tooling job or a machine job: /en/contact.