Diaphragm Wall Grab

What is a Diaphragm Wall Grab

A diaphragm wall grab is the excavating tool that opens the trench for a cast-in-place reinforced concrete diaphragm wall. The wall itself is a continuous underground structure used as a deep retaining wall for excavations, an alternative to driven walls installed by a pile driver where a stiffer, watertight, cast-in-place barrier is needed, as a hydraulic cut-off wall under and around dams and embankments, and as a permanent foundation element such as a barrette, an application it shares with the bored piles formed by a rotary drilling rig. Because the wall is cast in a trench supported by fluid rather than in an open hole, the grab must work blind, biting soil from a slurry-filled slot whose floor it cannot see, while holding the slot plumb to better than 1 percent of depth.

The grab is a clamshell: two hinged jaws that hang open while the tool is lowered, then rotate shut to scoop a bite of soil. It is suspended from a duty-cycle crane (a crawler crane built for repetitive free-fall and grab duty, not a lifting crane) by a main hoist rope. The working cycle repeats in four moves: lower the open grab to the trench floor, often letting the heavy bucket free-fall the last stroke to drive the jaws into the soil; close the jaws to capture a bite; hoist the closed, loaded bucket to the surface; swing clear and open the jaws to discharge spoil into a dump or onto a conveyor, where an excavator typically loads it out for disposal. A single bite removes roughly 0.5 to 2.5 cubic metres depending on grab width and length.

The method has a long industrial history. Continuous diaphragm walling under bentonite slurry was developed in Italy by ICOS in the early 1950s, and the rope-suspended clamshell grab was the original excavation tool. For three decades the cable grab was effectively the only choice. Hydraulic grabs with cylinder-actuated jaws appeared in the 1970s and 1980s and added controllable closing force and lower vibration; the hydromill, or trench cutter, introduced by Soletanche, Bauer, and Casagrande from the late 1970s onward, extended diaphragm walling into hard rock and to depths well beyond 100 m. Today the grab and the cutter are complementary: the grab owns the shallow-to-medium depth band on cost, and the cutter owns the deep and hard-rock band on capability.

Scale and economics favour the grab wherever the ground is workable. A diaphragm wall panel is typically 0.60 to 1.50 m thick, 2.8 to 7.0 m long, and 15 to 50 m deep, and a grab spread (one duty-cycle crane, one grab, a slurry plant, and a desanding unit) is a fraction of the capital cost of a cutter spread. The grab also tolerates obstructions and cobbles better than a delicate cutter wheel, and its teeth and jaws are field-replaceable. Its weaknesses are productivity in deep panels, where cycle time grows with trip depth, and an inability to excavate competent rock, where the jaws simply cannot bite.

Four engineering attributes determine whether a given grab suits a project: closing force (how dense a soil it can bite and how cleanly it cleans the trench bottom), trench width and bite length (which set the panel geometry), verticality control (steering plates and inclinometers that keep the slot plumb), and base-carrier compatibility (the duty-cycle crane class that can hoist the loaded bucket from full depth). These four, read together with the ground investigation, drive the selection that the rest of this guide develops.

Grab Types and Classification

Diaphragm wall grabs are classified first by how the jaws close (mechanical rope or hydraulic cylinder) and second by how the tool is guided in the trench (free-hanging on a rope, or rigidly guided on a kelly bar). These two axes produce the four practical configurations in the table below. Choosing the wrong configuration is the most common specification error: a free-hanging mechanical grab is cheap but drifts in deep or sloping ground, while a kelly-guided hydraulic grab holds plumb but costs more and needs a heavier carrier.

Mechanical (cable) grabs close their jaws by reeving a closing rope over sheaves on the grab head; the carrier winch tensions the closing rope while the main rope holds the body, and the jaws lever shut. Closing force is a function of bucket weight, free-fall energy, and rope mechanical advantage, so heavier buckets bite harder. The mechanical grab is simple, robust, cheap to buy and maintain, and forgiving of cobbles. Its limitation is that closing force cannot be increased independently of weight, and a free-hanging mechanical grab has no active way to correct drift, so verticality depends on operator skill and a heavy, well-balanced bucket. The Grab Specialist, Leffer, and many Chinese builders supply mechanical grabs in the standard 600 to 1,500 mm widths.

Hydraulic grabs carry their own hydraulic cylinders, fed from the carrier power pack through a hose reel, and develop high, controllable closing force regardless of bucket weight. Published cylinder forces run from roughly 1,200 kN to 1,800 kN on heavy units such as the Bauer DHG-V, and Liebherr lists a maximum closing force at the teeth on the order of 594 kN for mid-range slurry-wall grabs. Hydraulic grabs run quieter and with far less vibration than a free-falling mechanical grab, an advantage in cities and near sensitive structures. Most add steering plates (adjustable flaps on the grab frame) and an inclinometer so the operator can read and correct deviation in real time, which is why hydraulic grabs dominate deep and dense-ground work.

Kelly-guided grabs, such as the Casagrande KRC system, mount the hydraulic grab on the end of a telescopic kelly bar rather than letting it hang free. The kelly forces the grab to follow the rig mast, which makes verticality control far easier and lets the tool push through harder or layered strata that would deflect a free-hanging bucket. Casagrande publishes KRC working widths of 500 to 1,500 mm, bite lengths of 2,200 to 3,000 mm, and panel depths around 35 m for the kelly-guided range. The trade-off is a heavier, more complex rig and a limit on depth set by kelly length.

Low-headroom grabs pair a short kelly or compact grab with a low-mast carrier (for example a Bauer GB low-headroom grab carrier) for work under bridge decks, inside existing buildings, and beneath live infrastructure, where a full-height duty-cycle crane cannot stand. They trade reach and depth for the ability to work in a few metres of clear height. Across all four configurations the bucket bodies, jaws, and replaceable teeth share common wear-part logic, and grab width is always matched to one of the standard panel thicknesses described in Chapter 4.

Slurry Support and Ground Conditions

A grab cannot excavate a deep, narrow slot in soil unless the trench is held open by a support fluid, because the grab removes soil far faster than the ground can stand unaided and the slot would collapse. The support fluid, almost always a bentonite-water slurry or a synthetic polymer fluid, fills the trench as the grab digs and exerts hydrostatic pressure on the walls that exceeds the combined active earth pressure and groundwater pressure. Bentonite, a sodium-montmorillonite clay, also deposits a thin, low-permeability filter cake on the trench face that seals the soil pores so the fluid pressure can act on the wall rather than leak away. The slurry level is kept at least 1.0 to 1.5 m above the groundwater table for the whole dig, which is why guide walls are built first to contain and steer the fluid and the grab.

Ground condition is the single biggest control on whether a grab is the right tool at all. The table below maps common ground types to grab suitability and to the alternative where a grab struggles. It is a first-pass screen only: a site-specific ground investigation, including SPT or CPT data and a record of cobbles and boulders, must precede any final method choice.

Bentonite slurry management is as important as the grab itself. Fresh slurry is mixed to a target density (commonly around 1.03 to 1.10 g/cm3) and Marsh-funnel viscosity, then circulated and topped up as the grab digs. During excavation the slurry picks up sand and fines that increase its density and sand content; if not removed, this heavy, sandy slurry will not be cleanly displaced by tremie concrete and can leave sand pockets and inclusions in the finished panel. A desanding unit (cyclones over a vibrating screen) is therefore run continuously, and the slurry is tested for density, viscosity, and sand content before the trench is accepted for concreting. Polymer fluids are increasingly used in place of bentonite because they desand more easily and are simpler to dispose of, though they form a weaker filter cake and demand tighter level control.

Trench-bottom cleaning is the grab's final and most demanding job in each panel. After the slot reaches design depth, loose spoil and settled sand at the bottom must be removed so the reinforcement cage seats correctly and the toe of the wall bears on sound ground. A heavy grab with high closing force can grab-clean the base, but for critical work an airlift or suction pump is used to polish the bottom to a residue limit (EN 1538 references control of sediment at the base before concreting). Poor bottom cleaning is a leading cause of soft toes and base inclusions, defects that are expensive to remediate once the wall is cast.

Verticality, finally, is harder to hold in layered or sloping ground because a free-hanging grab tends to follow the path of least resistance and drift toward softer strata or down a dipping rockhead. This is precisely where hydraulic grabs with steering plates, kelly guidance, or a switch to a cutter earn their cost: in interbedded soils, near a sloping bedrock surface, or where adjacent panels must key together within a few centimetres, drift that a mechanical grab cannot correct becomes a quality and watertightness problem at the panel joints.

Standards, Tolerances, and Panel Construction

Diaphragm walling is a governed activity, and the grab must be specified to meet the tolerances the wall standard demands. The controlling execution standard in Europe is BS EN 1538:2010+A1:2015, Execution of special geotechnical works, diaphragm walls, which applies to walls built in a slurry-supported or dry trench with a wall thickness of at least 0.40 m. Geotechnical design sits under EN 1997-1 (Eurocode 7). North American practice references contractor specifications and bodies such as the DFI and ADSC. The key numeric requirements that a grab and its instrumentation must satisfy are summarised below.

Verticality is the headline tolerance. EN 1538 requires panels, including their ends, to be plumb to within 1 percent of depth in both directions for retaining walls, unless the project specifies tighter, and allows the tolerance to be relaxed where boulders or obstructions make it unachievable. One percent over a 40 m panel is 400 mm of permissible drift at the toe, which is generous for a cut-off wall but often too loose for a deep basement where panels must interlock, so contractors routinely specify 0.5 percent or better and rely on hydraulic grabs with inclinometers and steering plates, or on cutters, to achieve it.

The panel construction sequence frames everything the grab does. Work begins by casting two parallel reinforced-concrete guide walls along the wall line; these guide the grab on its first bites, contain the slurry, support the stop-end pipes, and carry the reinforcement cage before concreting. The grab then excavates the panel, usually as a primary panel dug in two or three bites with a soil pillar left between bites and trimmed last, all under slurry. Stop-end pipes (round or shaped joint formers the same diameter as the wall thickness) are lowered at the panel ends to form the construction joint with the next panel and are extracted after the concrete has begun to set.

Concreting follows trench cleaning and cage installation. The reinforcement cage, fabricated from deformed rebar, is lowered into the slurry-filled trench, then concrete is placed from the toe upward through one or more tremie pipes in a single continuous pour, often fed by a truck-mounted concrete pump, displacing the slurry as it rises so the fresh concrete never mixes with the fluid. A retarded, highly workable, self-compacting tremie mix is used so it flows around the cage without vibration. As panels are built in a primary-secondary or continuous sequence, the stop-end joint and good bottom cleaning are what give the finished wall its watertightness, which is why the grab's verticality and base-cleaning performance translate directly into wall quality.

Key Specification Parameters

Reading a grab datasheet means decoding a small set of parameters that together fix what the tool can dig, how plumb it can hold the slot, and which carrier it needs. The same eight parameters recur across Bauer, Liebherr, Casagrande, and Leffer sheets: trench width, bite length, closing or cylinder force, grab weight, open and closed height, grab volume, working depth, and verticality instrumentation. Each is explained below, with representative figures from published datasheets.

Trench width is the panel thickness the grab produces, and it is built to one of the standard values 0.60, 0.80, 1.00, 1.20, or 1.50 m; Leffer, for example, lists grabs and matching stop-ends in exactly these five widths. Special builds go narrower (down to about 0.40 m, the EN 1538 minimum) or wider. Bite length is the long dimension of the jaw, typically 2.2 to 2.8 m, reaching 3.0 to 3.8 m on heavy units; the Bauer DHG-V covers bite lengths of 2,400 to 3,800 mm. Width times length sets how many bites make a panel.

Closing force separates the two grab families and is the parameter that decides how dense a soil the grab can bite and how well it cleans the trench bottom. Hydraulic grabs quote a cylinder force directly: the Bauer DHG-V is offered at 1,200 kN or 1,800 kN, the older Bauer DHG C develops a cylinder force of about 120 t (roughly 1,180 kN) at 300 bar, and Liebherr lists a maximum closing force at the teeth in the order of 594 kN for mid-range models. Mechanical grabs do not quote a cylinder force; their closing capability is implied by bucket weight and free-fall, which is why a heavier mechanical bucket bites harder.

Grab weight matters twice: it drives a mechanical grab's bite, and it sizes the carrier for every grab. The Bauer DHG-V spans 17 to 37 t across its widths and lengths; the DHG C runs from about 12,000 to 19,200 kg and the DHG E from 19,000 to 25,000 kg. A heavier grab needs a duty-cycle crane with enough line pull and drum capacity to hoist the loaded bucket from full depth, so weight and working depth together set the minimum carrier class. Open and closed height determine the headroom the grab needs and, with kelly length on guided units, the achievable depth.

Working depth and verticality instrumentation close the spec. Rope-suspended grabs are good to roughly 40 to 50 m, instrumented hydraulic grabs beyond that, while kelly-guided ranges such as Casagrande KRC publish depths around 35 m limited by kelly length. The instrumentation line on a modern sheet lists steering plates (often with a stroke range, for example Bauer's 0, 4, 8, 12 settings), an onboard inclinometer with a deviation readout in the cab, and an optional turning device to rotate the grab in the slot. The decision factors below put these parameters into a working sequence.

Selection Decision Factors

To turn the preceding five chapters into a specific grab and carrier, work through the decision sequence below in order. Most selection mistakes come not from one wrong number but from deciding the tool before the ground investigation and the wall tolerance are known. These seven steps can serve as a fixed RFQ template for a diaphragm walling package.

1. Ground and method screen: Read the ground investigation first. Soft-to-medium soils with manageable obstructions suit a grab; competent rock or required depths past 50 to 60 m point to a hydromill cutter. Decide grab-versus-cutter before anything else, because it changes the whole spread.
2. Panel geometry: Fix the wall thickness (0.60 to 1.50 m, never below the EN 1538 minimum of 0.40 m), the panel length, and the design depth. These set the grab width, the bite length, and the number of bites per panel.
3. Closing force and grab family: Match closing capability to soil density. Soft to medium soils run on a mechanical grab; dense gravels, cobbles, and deep panels need a hydraulic grab developing 1,000 kN or more, and tight or layered ground favours a kelly-guided grab.
4. Verticality target and instrumentation: Take the wall tolerance from the design (EN 1538 default 1 percent, often tightened to 0.5 percent for keyed basements). A tight target requires a hydraulic grab with steering plates and an inclinometer, kelly guidance, or a cutter.
5. Base carrier match: Size the duty-cycle crane to the grab weight and working depth. A 17 to 37 t hydraulic grab needs a heavy duty-cycle crane (commonly the 80 to 110 t class) with free-fall main winch, separate closing or hose winch, and adequate drum capacity for full-depth hoisting.
6. Slurry and desanding plant: Specify the support fluid (bentonite or polymer), the mixing and storage capacity, and a desanding unit sized to the dig rate. Trench stability, bottom cleanliness, and final concrete quality all depend on slurry that is kept within density, viscosity, and sand-content limits.
7. Site constraints and headroom: Check overhead clearance and access. Work under bridges or inside buildings calls for a low-headroom grab and a compact carrier, trading reach and depth for the ability to stand in a few metres of clear height.

One last commonly overlooked dimension is serviceability and wear-part supply. Grabs are abrasive-duty tools: teeth, jaw lips, hoses, and steering plates wear, and a panel cannot wait days for a part. Confirm that the supplier holds local stock of replaceable teeth and jaws, that the grab matches a duty-cycle carrier already in the contractor's fleet or available for hire nearby, and that field service can rebuild the bucket on site. Bauer, Liebherr, Casagrande, and Leffer all maintain spare-part and service networks across major foundation markets, which makes them low-risk choices on long programmes; a cheaper grab with no nearby parts support can idle an entire walling spread when a tooth set wears out.