GreatLight CNC Machining Factory logo
CNC Machining
Rapid Prototyping
Materials
Industries
News
About GL

Get Instant Quote

Concrete formwork guide

How Can 3D Printing Alleviate Formwork Waste and Labor Strain?

This guide is for engineers and procurement teams who keep hearing about 3D printed concrete and want to know where the process actually pays off. We walk through the questions to ask, the geometry that suits printed formwork, the print and casting parameters that matter, and the point where conventional CNC mold work is still the better call.

Formwork first±0.005 mm CNC trimNo minimum order
How can 3D printing alleviate formwork waste in construction
Quick answer

Key takeaways

Formwork is the leverPrinted polymer or cementitious molds cut plywood waste and the carpentry hours behind it. Printed structural concrete itself is still limited in span.
Complexity is cheap on the print bedUndercuts, ribs and text cost the same to print. They cost plywood carpenters real time.
One-off or low volume onlyAbove roughly 20 identical units, a machined steel or aluminum mold wins on unit cost and surface finish.
Print, then machinePrinted molds still need CNC-trimmed mating faces and sealing surfaces if you want tight concrete tolerances.
Check the release agent firstPolymer molds need compatible release chemistry or the surface pulls and the part is scrapped.
Scope

What This Answer Covers, and What It Does Not

Ask how can 3D printing alleviate construction waste and most answers jump to printing a whole house. That is the wrong place to start. The process that is already repeatable and already costed is printed formwork: molds for columns, beams, wall panels, precast segments and decorative elements. The mold is small, the geometry is free-form, and the waste is measured in kilograms instead of skips of plywood.

This guide stays on that ground. We cover polymer and cementitious molds, the print parameters that hold a concrete surface, and the handoff to casting. We do not cover structural certification of printed load-bearing concrete, reinforcement design, or fire ratings. Those need a licensed structural engineer and local code review, and no machining supplier can sign them off.

The reader we write for is a project engineer, a precast plant manager, or a buyer comparing a printed mold against a plywood or steel alternative. By the end you should be able to read a part drawing, decide whether the mold belongs on a printer or a CNC, and name the parameters you need to specify.

  • 1
    In scopePrinted formwork for precast and cast-in-place, small to medium units, complex geometry, tight surface requirements.
  • 2
    Out of scopeStructural approval of printed concrete, mix design, reinforcement detailing, fire and acoustic ratings.
Mechanism

Where the Savings Actually Come From

Three mechanisms do the work. The first is material: a printed mold is a shell, not a solid block. You can print a 20 mm to 40 mm wall with a lattice or ribbed backing and still take the hydrostatic pressure of a 1.5 m concrete pour. A plywood equivalent is a full box of sheets, battens and ties, and most of it is cut to waste.

The second is labor. Plywood formwork is a carpentry job. Every compound curve, chamfer and recess is measured, cut, fitted and braced by hand, and the hours scale with geometric complexity. On a printer, a curve and a straight line cost almost the same. A shape that takes a carpenter two days can print overnight.

The third is iteration. When a client changes a corner radius or a logo recess, a printed mold is edited in CAD and reprinted. The old mold goes to recycling. With plywood, the whole panel is scrapped and rebuilt. On projects with two or three design revisions, that is where the cost curve separates.

  • 1
    Shell vs solidPrinted molds are hollow shells with ribs, so material scales with surface area, not volume.
  • 2
    Complexity is freePrint time changes little between a plain panel and a sculpted one.
  • 3
    Revision costA CAD edit plus a reprint is cheaper than rebuilding a timber panel set.
Geometry

Which Parts Suit a Printed Mold

Printed formwork pays off when the mold is complex, the batch is small, or the surface has to be reproduced exactly. Good candidates: a column capital with 60 mm deep flutes, a precast facade panel with a 15 mm negative relief, a tunnel segment with a matching keyway, a bench or planter with an undercut that would need a split plywood box.

It is a poor fit for long straight runs and repetitive flat panels. A 6 m straight wall panel does not need printing; standard steel or aluminum molds are cheaper per unit and give a better surface. It is also a poor fit for very high pours. Above roughly 2 m of fresh concrete head, the pressure on a polymer shell gets hard to control without a steel backing frame.

Size is the other gate. Most industrial polymer printers used for molds top out around 1 m to 2 m in one axis, and larger parts are printed in segments and bonded or bolted. That is workable, but the joint lines have to be planned into the drawing. If the concrete surface is architecturally exposed, decide where the seams land before you print, not after.

  • 1
    Good fitSculpted panels, capitals, keyways, undercuts, one-off architectural elements, 1 to 20 units.
  • 2
    Poor fitFlat repetitive panels, pours over 2 m head, tight dimensional tolerances on the concrete itself.
Parameters

The concrete copies the mold, so the mold surface is the concrete surface. For a polymer mold, a 0.4 mm to 0.8 mm nozzle with a 0.2 mm to 0.3 mm layer height gives a surface that reads as a fine horizontal texture after casting. If the specification calls for a smooth finish, the printed shell is sanded and coated, and that step adds hours. Print it rough and you will cast it rough.

Wall thickness should follow the pour pressure. As a rule of thumb, a 4 mm shell with ribs every 100 mm holds a 1 m head; a 6 mm shell with 150 mm ribs holds 1.5 m. These are starting points, not guarantees. Fresh concrete behaves differently at 5 °C and 30 °C, and a slow pour raises pressure on the lower panels. Run a trial pour on the first mold before you commit the batch.

Draft and release matter more than print quality. A 1.5° to 3° draft on vertical faces, no sharp internal corners below 3 mm radius, and a compatible release agent are what keep a part from tearing on stripping. We have seen good prints fail at demolding because the release chemistry was chosen for steel. Check compatibility with the polymer before the first pour.

  • 1
    Layer height0.2–0.3 mm for a fine texture; sand and coat if the drawing calls for smooth.
  • 2
    Shell and ribs4 mm shell with 100 mm ribs for 1 m head; 6 mm shell with 150 mm ribs for 1.5 m.
  • 3
    Draft1.5°–3° on vertical faces; 3 mm minimum internal radius.
Hybrid route

Where CNC Still Beats Printing

A printed mold is not a finished tool. The face that meets the casting table, the bolt pattern, and any sealing groove usually need machining. Printing leaves layer lines and a soft surface; bolting a mold down on layer lines leaks. That is why the practical route is hybrid: print the free-form shell, then machine the mating faces, the flange and the locating features.

This is where a shop with both processes saves a step. At GreatLight we print the shell and then trim it on a 3-axis or 5-axis machine to ±0.005 mm on the critical faces, with the flange held to Ra 0.8–1.6 μm so a gasket seats. The same plant machines the backing frame. The mold arrives as one assembly instead of three suppliers arguing about who is out of tolerance.

Choose all-CNC when the mold is simple and the batch is larger. A machined aluminum mold for a flat 1.2 m panel is fast, durable and gives a better surface than a printed one. Above roughly 20 identical units, the machined tool usually wins on unit cost. Below that, and especially with compound curvature, printing wins.

  • 1
    Print the shellFree-form surface, ribs, undercuts, logos, draft.
  • 2
    Machine the interfacesFlange, bolt pattern, sealing groove, datum faces at ±0.005 mm.
Workflow

Seven Steps From Drawing to First Cast

Each step names the parameter to fix and the mistake that kills the mold.

  • 1
    1. Fix the concrete surface spec firstDecide finish class before any CAD work. A smooth architectural finish means a coated and sanded mold; an as-cast finish tolerates visible layer texture. Write the target in the drawing. Printing first and arguing about finish later is the most common expensive mistake.
  • 2
    2. Split the mold around draft and seamsApply 1.5°–3° draft on all vertical faces and place parting lines where they will not cross an exposed architectural surface. Plan segment joints for printers with a 1 m to 2 m build envelope, and put the joint on a natural shadow line.
  • 3
    3. Set shell thickness against pour pressureUse 4 mm shell with 100 mm ribs up to a 1 m head, 6 mm shell with 150 mm ribs up to 1.5 m. Add a steel backing frame above 2 m. Keep fresh concrete temperature and pour rate in the calculation, not just height.
  • 4
    4. Print with the right layer and nozzleA 0.4–0.8 mm nozzle at 0.2–0.3 mm layer height is a good default. Orientation matters: print so layer lines run horizontally on vertical faces, which reads as a regular texture after casting instead of a stepped pattern.
  • 5
    5. Machine the flange and sealing facesTrim the mold base, flange and bolt holes on a CNC. Target ±0.005 mm on mating faces and Ra 0.8–1.6 μm on the seal groove. Print-then-machine in one shop avoids tolerance arguments between two suppliers.
  • 6
    6. Seal, coat and test the release agentSeal the printed surface, apply the release agent and run a small trial pour. Confirm the release chemistry is compatible with the polymer. A 300 mm cube trial pour costs little and catches tearing before the production batch.
  • 7
    7. Cast, strip and inspectTrack pour rate and concrete temperature against the trial. Strip at the specified age, inspect the first part, and record where the surface failed. Feed that back into the next mold revision rather than adjusting the mix blindly.
Decision table

Printed Mold vs Plywood vs Machined Mold

Use this as a first screen before quoting. Unit counts are starting points, not rules.

FactorPrinted polymer moldPlywood formworkMachined aluminum mold
Best batch size1 to 20 units1 to 5 units20 to 10,000+ units
Complex curvatureNo extra costHigh carpentry hoursNeeds 5-axis machining
Typical concrete tolerance±2 to ±5 mm±5 to ±10 mm±0.5 to ±1 mm
Mold surfaceLayer texture or coated smoothBoard texture, seam linesSmooth, Ra 0.8–1.6 μm
Max pour headAbout 2 m with backing frameAbout 3 m with bracing3 m and above
Design change costCAD edit plus reprintRebuild the panelRe-machine or weld repair
Reuse cycles10 to 50 cycles3 to 8 cycles500 to 2,000 cycles

Print the shape, machine the fit

Use printed formwork where geometry is complex and the batch is small. Machine the flange, seal and datum faces so the mold seals and repeats. If the batch passes about 20 identical units, switch to a machined tool.

FAQs

Questions Engineers Ask Next

Can a printed polymer mold take normal concrete pour pressure?

Yes, within limits. A 4 mm shell with ribs every 100 mm handles roughly a 1 m head; a 6 mm shell with 150 mm ribs handles about 1.5 m. Above 2 m, add a steel backing frame.

Pressure depends on pour rate and concrete temperature as much as height. A slow pour at 30 °C loads the lower panels harder than the same height poured fast in cool weather. Run a trial pour before the batch.

What concrete tolerance should I expect from a printed mold?

For cast-in-place work with a printed shell, expect ±2 to ±5 mm on the concrete, driven by mold stiffness and how well the mold is braced to the casting table.

If the drawing needs tighter than that, machine the critical faces of the mold to ±0.005 mm and stiffen the backing. The concrete will still move during curing, so mold accuracy is only one term in the stack.

How many times can a printed mold be reused?

Ten to fifty cycles is a realistic band for a sealed polymer mold with a compatible release agent. The failure mode is usually surface wear and edge damage at the flange, not the printed shell cracking.

Cementitious printed molds behave differently and are often treated as sacrificial. If you need hundreds of cycles, a machined aluminum tool is the correct choice.

Do I still need CNC machining if I print the mold?

Usually yes, for the interfaces. The flange, bolt pattern, sealing groove and datum faces are machined so the mold sits flat and seals. Layer lines left on a mating face will leak grout.

Printing and machining in one shop keeps the tolerance stack in one place. We hold mating faces to ±0.005 mm and seal grooves to Ra 0.8–1.6 μm.

Can you machine and print the same mold in one order?

Yes. Our custom 3D printing service and the CNC cells are in the same plant, so a printed shell can be trimmed, fitted with its backing frame and shipped as one assembly.

No minimum order quantity applies. A single prototype mold and a 10,000-part production run go through the same inspection process, with 100% inspection before shipment and reports on request.

What file format and lead time should I plan for?

Send STEP or STL for the shell and a 2D drawing for the machined interfaces. We return a quotation and a free DFM analysis within 12 hours, and production can start within 24 hours after that.

Printed and machined parts typically ship in 3 to 5 days. For a first mold, add time for the trial pour and any surface coating, since those steps are on your side.

Send the mold drawing and we will screen it

Upload the shell geometry and the machined interfaces. You get a quotation and a free DFM analysis within 12 hours, with the print and the CNC trim quoted as one assembly.

12-hour quotePrint and machine in one plantNo minimum order

Follow

More process notes

We publish setup notes, tooling trials and inspection data from the factory floor.

FacebookTikTokYouTubeLinkedInInstagramThreadsPinterest

Trusted by engineers and manufacturers worldwide

Tesla Ford Motor Company BYD Auto Denso Magna International Boeing Airbus Medtronic KUKA FANUC