Cobod Builds Great Things: The Future of 3D Printing

• 1
  How Cobod Builds Great Things Layer by Layer
• 2
  What the BOD2 Process Window Allows
• 3
  Material Choices and Why Certification Matters
• 4
  Where Subtractive Machining Still Wins
• 5
  Design Rules That Keep a Printed Build Out of Trouble
• 6
  Printed Concrete vs Machined Metal: Fit by Requirement

Cobod builds great things with a gantry printer, not a robot arm. The BOD2 uses Cartesian motion: a frame carries the print head in X, Y and Z over the slab, and the frame bolts together in modules. That is why the same machine can be shipped in containers and rebuilt at a different length.

The nozzle traces a closed path for each wall, then the gantry lifts by one layer height. A typical layer runs 30–50 mm wide and 10–20 mm tall. The pump pushes mortar through a hose; the motion system decides where it lands. Speed and pump rate have to stay matched, or the bead necks down and the wall gets a thin spot.

Setting happens fast because the mix is formulated for it. Accelerator is dosed at the nozzle, so the bead must carry its own weight within minutes. That is the whole trick. Without rapid green strength, a 3 m wall would slump before the next pass.

• 1
  Gantry over robot armSquare corners and long straight runs stay accurate over a 30 m frame.
• 2
  Nozzle-dosed acceleratorSets the bead fast enough to stack layers without formwork.
• 3
  Pump and motion are coupledBead width is a process parameter, not an operator habit.

The published envelope for this platform reaches roughly 30 m in length, 10 m in height and 8 m in width. Those are maximums, not a typical job. Most walls land between 2.5 m and 4 m tall, because height drives frame stiffness and pump pressure more than floor area does.

Wall thickness is set by nozzle choice and pass count. A single 40 mm pass suits a partition. A 200 mm structural wall needs multiple adjacent passes, and the joints between them become the weak plane. Shear and tension across a cold joint are where engineers spend their review time.

Openings are just skipped paths. A window or door is left as a gap and bridged later with a lintel or a printed arch. Arches work well because the printer can lay a curved path. Long flat spans do not, so most designs put a steel or precast lintel over anything wider than about 1 m.

• 1
  Height is the real limitPressure and frame stiffness bite before floor length does.
• 2
  Cold joints are the weak planePlan reinforcement across every pass boundary.
• 3
  Curves are cheap, flat spans are notUse printed arches, buy lintels for wide openings.

The mix is a cement-based mortar, not ordinary ready-mix. It needs a yield stress high enough to hold shape and a setting time short enough to build fast. Sand grading, water-to-binder ratio, and admixture dose all move together. Change one and the bead changes.

Printing insulation in the same pass is a real advantage. Foam can be extruded into a cavity between two printed shells, so the wall arrives with a thermal break already inside it. That saves the separate insulation trade and the cavity ties that go with it.

Structural certification is the slow part. A printed wall is a new material system, and building codes in most markets want test data for the specific mix, nozzle and cure regime. Expect a testing program before approval, not a datasheet. Local material sourcing also matters, since shipping dry mix across an ocean rarely pays.

• 1
  Formulated, not standardYield stress and set time are tuned per application.
• 2
  Foam in the cavityInsulation prints in the same pass as the shells.
• 3
  Code approval takes testingBudget for a mix-specific test program.

Printing gives you a rough near-net wall. It does not give you a bearing seat. Anything that must locate, seal, or rotate needs a machined face, and that is where a CNC shop enters the project. Embed plates, anchor pockets and MEP sleeves are typical printed-in features that get finished on a mill.

GreatLight machines those interface parts from aluminium, stainless and steel. A printed wall can hold a cast-in plate within a few millimeters, but the plate itself is held to ±0.005 mm on our 5-axis centers. The printer sets the coarse geometry; the mill sets the datum everything else references.

The same split applies to prototypes. If a design team wants to test a connection detail before committing to a print run, machining a small batch of brackets in 6061 or 17-4PH takes days, not weeks. Print the wall, machine the node. That is the practical division of labor today.

• 1
  Print coarse, machine preciseEmbed plates, sleeves and anchor pockets get a machined finish.
• 2
  Datum controlOne machined face defines the reference for the whole assembly.
• 3
  Prototype the nodeTest the connection in metal before the print run.

Keep the wall path continuous. Every stop leaves a cold joint, and every cold joint needs reinforcement or a mechanical key. Corners should be rounded, not sharp, because the nozzle decelerates into a corner and the bead gets fat. A 50–100 mm radius costs nothing and removes a defect.

Layer time matters more than most teams expect. If a pass takes too long, the layer below is already set and the bond is weak. If it is too fast, the layer below is still soft and the wall bulges. Cycle time per layer is a real design constraint, and it is set by wall length divided by print speed.

Plan the services before the print. Conduit, water lines and drainage are easiest as printed-in chases, but the diameter and slope have to be right the first time. Cutting a chase into a printed wall afterwards is slow and weakens the section.

• 1
  Round the cornersA 50–100 mm radius avoids bead buildup at direction changes.
• 2
  Watch layer cycle timeSet time and green strength must match the pass duration.
• 3
  Design the chases firstRetrofit cutting damages the printed section.