Concrete Armour Unit Mould: 2026 Engineering Guide

A concrete armour unit mould is not a box welded to a shape. It is the device that fixes the mass of every tetrapod, dolos and interlocking block that ends up on the slope. Armour stability is a function of unit mass, unit mass is a function of cavity volume, and cavity volume is a function of how accurately and how repeatably the mould holds its dimensions over hundreds of casting cycles. Tolerance error is therefore not a quality-report problem — it is transferred directly into the hydraulic performance of the breakwater.

This article walks through the full engineering chain: converting the coastal engineer's nominal mass into a mould cavity specification, understanding what a mass tolerance actually permits, designing the segmented steel architecture that a four-legged unit demands, and sizing the number of mould sets a precasting yard genuinely needs once low-carbon concrete mixes slow the demoulding cycle down.

Why Coastal Defence Work Is Redefining Armour Unit Mould Demand in 2026

The demand context matters because it changes what buyers ask for. The global precast concrete market is valued at roughly USD 149.5 billion in 2026, rising to about USD 186 billion by 2031 according to Mordor Intelligence's September 2026 assessment, while Grand View Research puts the 2026 base at USD 123.6 billion growing at a 6.4% CAGR to 2033. The narrower breakwater construction segment is smaller but faster: an October 2026 Business Research Insights estimate values it at USD 0.56 billion in 2026 heading to USD 0.97 billion by 2035, a 6.2% CAGR. More telling than any market-size figure is the Precast/Prestressed Concrete Institute's February 2026 construction outlook, which projects nonbuilding structures — ports, marine works, roads, utilities — to lead precast growth again, up around 4% in 2026, while nonresidential buildings sit flat. Precast demand growth is concentrating exactly in the infrastructure categories that consume armour unit and seawall block moulds.

Three technical forces inside that growth are stress-testing conventional steel mould design. First, design storm severity in coastal adaptation programmes is pushing nominal unit masses upward, and larger units magnify every distortion and assembly error. Second, binder replacement with ground granulated blast furnace slag, fly ash and limestone calcined clay systems slows early strength gain, which delays demoulding and multiplies the number of mould sets a yard must own. Third, armour geometry research has accelerated — the PentaPod configuration, the engineered armour units presented in 2026 peer-reviewed studies on rubble mound breakwaters, and bio-receptive eco-engineering units all introduce deeper undercuts, tighter cores and controlled surface texture. Each new geometry is a tooling problem before it is a design problem.

From Nominal Mass to Cavity Volume: The Calculation That Starts the Mould

The mould order does not begin with a drawing; it begins with a mass. Armour unit mass is conventionally derived from a Hudson-type stability expression of the form:

`` W = γ_r · H³ / [ K_D · (S_r − 1)³ · cot θ ] ``

where W is the nominal unit weight, γ_r the unit weight of the concrete armour unit, H the design wave height, K_D the stability coefficient for the unit shape, S_r the relative density of concrete against the surrounding water and θ the armour slope angle. The cubic dependence on H is the number that should shape procurement decisions: a 10% increase in design wave height raises the required unit mass by roughly 33%, not 10%. Projects that revise their design sea state late in the game do not buy slightly bigger moulds — they buy a new mould family.

The mass then converts to cavity volume by dividing by the achieved concrete density. For plain marine concrete this typically lands in the 2.35–2.40 t/m³ range depending on aggregate, but the value used in a cavity calculation must come from the project's approved mix design and a measured density, never from a rule of thumb. A 4,000 kg nominal unit at 2.40 t/m³ requires 1.667 m³ of compacted concrete; at 2.32 t/m³ the same unit needs 1.724 m³. That 3.4% spread comes from the mix, not the mould, which is the first thing to nail down before anyone critiques your dimensional accuracy.

Finally, the cube law applies. Volume scales with the product of three axes, so a 1% dimensional error on each axis produces roughly a 3% mass error. A ±5% mass window therefore implies about 1.6% of freedom per axis (since 1.05^(1/3) ≈ 1.0164). On a two-metre leg that sounds generous — around 30 mm — which is precisely why gross dimensional error is rarely the real cause of mass drift in production. The genuine causes are surface build-up from repair welding, section thickening where concrete skin adheres, entrapped voids from under-vibration, and density swing. A mould that measures perfectly on first article and then drifts is a maintenance problem, not a fabrication problem.

What "Mass Tolerance" Actually Means for Concrete Armour Units

Buyers frequently cite EN 13383 in tenders, and it needs to be read carefully: EN 13383-1 is a product standard for armourstone — natural rock, and armourstone made from blast furnace and steelmaking slag — with its companion parts governing the specification of gradings, mass and size distributions for rock. It is not a specification for precast concrete armour units. For mould-made concrete units the governing document is almost always the project specification itself, which typically defines a nominal mass, an acceptable individual-unit mass window, a maximum proportion of units outside that window, and a gradation or size-distribution requirement for the consignment. Getting this distinction wrong in a tender response is one of the fastest ways to lose credibility with a coastal engineer.

What is consistent across jurisdictions is the placement consequence. The Rock Manual's guidance on concrete armour units notes that artificial armour units are usually cast from plain concrete of a standard grade, and that achievable placement rates vary with unit size and with the placement tolerance the designer allows. Mass and shape scatter therefore feeds straight into layer porosity and interlock, and an out-of-family unit cannot be corrected on the slope — it is either downgraded to underlayer or demolished.

A practical inspection regime that satisfies most specifications runs like this: weigh and dimensionally scan the first article from a new mould and record it against the nominal; weigh a defined percentage of production units per batch, with the sampling rate rising whenever the batch mean drifts toward a window edge; mark each accepted unit with mass, batch, mould set and casting date; and re-survey the mould itself after any structural repair and at a defined cycle interval, because internal abrasion and patch welding both change cavity volume. Traceability is what converts a rejected unit into a contained incident rather than a stop-work order.

Steel Mould Architecture for Tetrapod, Dolos and Interlocking Units

Segmentation and core strategy

A tetrapod's four legs point along axes that no single withdrawal direction can clear, so a tetrapod mould is inevitably a bolted assembly: a central core body, separate leg shells, and removable inserts at the leg tips. Design the parting lines around the demould sequence, not around whatever is easiest to cut plate for. Units with a single stem and a socket — Accropode-type geometry — instead require a mandrel core that must survive repeated impact striking without bending, which usually means a replaceable core sleeve rather than an integral core. Dolos units, with their two pedestal ends and relatively open profile, tolerate simpler parting but concentrate load on fewer faces, so they reward heavier stiffening rather than more segmentation. Where demould effort is the dominant cost driver, our easy-demould interlocking block mould approach — tapered faces, bolted flange interfaces and a defined striking sequence — is the pattern worth copying.

Draft, internal radii and demould force

Practical draft on every face that must release is the cheapest quality improvement available; where geometry fights it, radii do the rest. Sharp internal corners at leg roots are doubly expensive: they lock the unit onto the mould, they concentrate stress so the concrete cracks on stripping, and they are the first place a casting defect appears. Generous internal radii at leg-to-core junctions, matched by a radiused external transition, raise demould reliability, reduce chipping of the finished unit, and improve concrete flow into the leg tip. Lifting lug placement deserves equal attention — a unit lifted at a point that flexes the leg will show root hairline cracking days later, and nobody blames the mould.

Stiffness, weld discipline and assembly repeatability

Fresh concrete presses on a shell according to its fluid height, so the panel thickness and rib spacing that behave on a 500 kg mould become a drum on a 5 t mould. Stiffen against that pressure with a rib grid sized for the tallest pour, and pre-camber where asymmetric welding will otherwise pull faces out of true. Weld symmetrically, in a balanced sequence, with stress relief before machining mating faces — because the recurring culprit in out-of-round legs is not the operator but the accumulation of flange and bolt-up error each time the mould is reassembled. Machined mating faces, located dowels and a fabrication jig that sets leg axes once, in steel rather than on the shop floor each time, are what let cast number 400 match cast number 1. For units above a few tonnes this is where our custom breakwater formwork work concentrates its engineering.

Low-Carbon Concrete Is Quietly Changing What a Mould Must Do

Binder substitution is the least visible and most expensive change now reaching armour unit yards. Replacing a large share of Portland cement with slag or fly ash improves long-term durability and cuts embodied carbon — the reason a 2025 life cycle assessment of tetrapod concrete armour units in Frontiers in Sustainable Cities could treat production, transport and placement as one optimisable chain — but low-early-strength mixes also delay the hour at which a unit can be stripped without chipping its edges, particularly below 15 °C.

That delay is a mould-count problem, and mould count is capital. The calculation a yard must run before ordering is:

`` mould sets = daily output target × cycle hours / 24 × (1 + rotation float) ``

Worked example. A yard targeting 120 units per day with a 24-hour cast-to-strip cycle needs 120 sets; with 10% rotation float for maintenance and repair, 132 sets. Introduce validated low-temperature steam curing and cut the cycle to 16 hours, and the same output needs 80 sets, or 88 with float. That is a 44-set difference in capital, floor space and crane moves — created entirely by a curing decision, not by the mould design. Buyers who quote mould quantities off a historical cycle time for a 100% clinker mix routinely under-order tooling by a third.

The mix change also loads the mould itself. Dense, low water/cement-ratio concrete needs more vibration energy to reach the leg tips, so stiffener layouts must be coordinated with external vibrator clamping points instead of accidentally damping them, and venting at high points becomes mandatory rather than optional. Slag-rich mixes are also more sensitive to surface carbonation and blowholing, which raises the value of a genuinely smooth, ground-flush mould skin. Marine exposure classes are unforgiving here: BS EN 206 and its UK companion BS 8500 drive splash and spray zone work (XS3) toward a high specified strength class, a low free water/cement ratio and a minimum binder content, and the durability rules for the splash zone are set out in the industry guidance on durability rules for the waves. Confirm the exact combination for the project's design life with the specifying engineer, then design the mould around the resulting stiff mix.

Surface Finish, Release Agents and the New Bio-Receptive Texture Requirement

An armour unit's surface is not cosmetic. Skin porosity governs how aggressively chloride and moisture enter at the outer layer, and surface geometry governs how the units interlock when placed. Both start in the mould. Welds should be ground flush to the parent plate, because a proud bead leaves a fin on the unit and a groove in the mould; the mould interior needs deliberate venting at points where air cannot escape, since trapped air produces the blowhole clusters that buyers later call "casting defects" and that are in fact air-management defects.

Release agent practice is where consistent process control beats product choice. Over-application does not just stain the unit, it pools in leg roots and radii and leaves pinholes across the finished face. What matters is a thin, uniform film, a release chemistry matched to a slow-curing slag mix rather than a generic oil, and one method applied the same way every shift.

The genuinely new requirement in this space is eco-engineering. Bio-receptive armour units — the Coastalock programme validated at model scale with Canada's National Research Council and the University of Ottawa, involving more than 600 model-scale replicas — are engineered to be colonised by marine life, which means their surface must carry controlled texture and micro-habitats rather than being smooth. That reverses a rule this industry has followed for decades. A mould that was "good" because it produced glassy concrete is now the wrong tool, and the texture has to be created by machined panels, inserted texture liners or moulded inserts held to a defined pattern and depth. Buyers specifying these units should expect to approve a texture sample panel before the production mould is released.

Steel vs Plywood vs FRP: A Cost-per-Cast Framework

For armour units the material decision is a durability and dimensional-stability decision, not a headline price decision.

Compare them on cost per cast, never on purchase price:

`` cost per cast = (mould cost + lifetime refurbishment cost) / total casts + demould labour per cast + (rework rate × rework cost per unit) ``

Illustrative inputs, to be replaced with your own quotations: a steel mould at 100 cost units that delivers 900 casts with 25 units of refurbishment costs 0.139 per cast before labour; a plywood mould at 18 cost units that degrades after 60 casts with no refurbishment already costs 0.30 per cast before labour — and it typically loses the mass tolerance window long before it physically fails. The steel mould only looks expensive on the purchase order; the plywood mould is expensive in reject units, which for a 4 t armour unit means the loss of one or two finished castings erases the difference outright.

First-Article Acceptance Checklist

Before releasing a mould to production, put these ten items in the sign-off record. It takes one day and prevents a rework loop measured in weeks.

1. Cavity volume verified against the approved mix design density, with the calculation attached.
2. Individual unit mass recorded for the first article, inside the project's mass window with margin, not at the edge.
3. Three-dimensional scan or full dimensional survey of the cast unit, leg lengths and leg axes included.
4. Concrete density measured from a companion cube set, so mass error can be attributed correctly.
5. Draft angles and internal radii confirmed as-built against the drawing, including leg roots.
6. All parting lines opened and re-bolted, checking that dowels — not bolts alone — locate the shells.
7. Demould timed and the striking sequence written down, with the number of operators recorded.
8. Blowhole and fin inspection under raking light, with a photographic standard for acceptance.
9. Lifting lug positions confirmed against a lift test on a sacrificial unit.
10. A mould re-survey interval defined, and triggered explicitly after any structural repair weld.

Sign the checklist, attach it to the mould set number, and archive it with the batch records — that file is what defends you when a consultant audits the yard. If your project needs units built to a documented acceptance trail, send the drawings and design wave height to our engineers and we will return a technical proposal rather than only a price.

Frequently Asked Questions

What mass tolerance should I specify for tetrapod moulds?

Specify it as a window on the finished unit, not on the mould, and give yourself headroom: if the project allows ±5% on the unit, the mould should be released at roughly ±2–3% so that abrasion, repair welding and density variation have somewhere to go. Also state the maximum share of units permitted outside the window, and a sampling rate, because a tolerance nobody measures is a tolerance nobody meets.

Why do my armour units vary in mass even though the mould measures correctly?

Because mould geometry is only one of four variables. Achieved density, entrapped air from under-vibration, surface build-up inside the cavity, and section thickening from adhered skin all change mass without changing a single measured dimension. Weigh a sample series, then scan the mould, and you will know within an hour which of the four you are fighting.

Can one tetrapod mould produce two different unit sizes?

Not reliably. Leg axes, core geometry and cavity volume are interdependent, so a scaled unit normally needs a new core body and leg shells. What can be shared is the base frame, lifting arrangement and flange standard between sizes — which is where a family-based tooling programme reduces cost without touching the tolerance chain.

How many mould sets do I need for a coastal defence project?

Run the cycle calculation: daily output target multiplied by cast-to-strip hours divided by 24, plus 10–15% float for maintenance. Then challenge the cycle hours with your actual mix. On slag-blended mixes at winter temperatures the honest number is frequently 1.5 times what was assumed from catalogue data, and discovering that after the first mould shipment arrives is the expensive version of this question.

References. Mordor Intelligence, Precast Concrete Market (September 2026) — mordorintelligence.com/industry-reports/precast-concrete-market. Grand View Research, Precast Concrete Market Report 2026–2033 — grandviewresearch.com/industry-analysis/precast-concrete-market. Precast/Prestressed Concrete Institute, 2026 Construction Outlook for the Precast Concrete Industry (February 2026) — precast.org/blog/2026-construction-outlook-for-the-precast-concrete-industry. Business Research Insights, Breakwaters Market (October 2026) — businessresearchinsights.com/market-reports/breakwaters-market-109552. Life cycle assessment of tetrapod concrete armour units, Frontiers in Sustainable Cities (2025) — frontiersin.org/journals/sustainable-cities/articles/10.3389/frsc.2025.1592987/full. The Rock Manual, Chapter 3, concrete armour units — kennisbank-waterbouw.nl/DesignCodes/rockmanual/BWchapter%203.pdf. Related tooling pages: Tetrapod Mold, Breakwater Block Mold, Breakwater moulds for coastal defence, Breakwater category.