Digital Twin Series · Level 2 · Case study
3D Concrete Printer Digital Twin
What can a digital twin of a gantry concrete printer tell me before I print?
A gantry 3D concrete printer has three things to get right: the motion of the gantry, the flow of concrete from the plant to the nozzle, and the way the layers stand up while the concrete is still soft. The twin lets you test each of these on a model, then set the speed, layer height and nozzle that print fastest without a blocked pump or a fallen wall. The material and machine values on this page are assumptions for teaching, not measurements of any real mix or machine: put in your own.
The move uses a jerk-limited profile: acceleration is ramped, not stepped, which is gentler on the structure. The drive force is mass x acceleration plus rolling friction. The mast is treated as a cantilever carrying the print head; its stiffness, the masses and the pinion data are assumptions. Take yours from the CAD model and the drive data sheets. The vibration rule (the move transition should last at least three natural periods) is a common rule of thumb, not a standard.
The printable concrete is modelled as a Bingham fluid in laminar flow: it does not move until the pressure beats its yield stress, then it flows with a plastic viscosity. The pressure comes from the exact Buckingham-Reiner relation for the hose and for the nozzle conduit. Real mixes also change with time, shear and pump pulsation, so measure the yield stress and viscosity of your own mix. These settings feed the sections below.
Fresh concrete gains strength as it rests. A simple plastic-collapse check compares the weight of the layers above with a capacity that grows with the age of the lower layers. If the layer time is too short, the load grows faster than the strength and the wall collapses. The yield stress, its growth rate and the collapse factor are assumptions of this teaching model; different simple models use different factors, so the factor is a setting here.
The plant has to supply at least what the nozzle takes, and the concrete must reach the nozzle before it starts to set. The demand comes from the print settings above. The cement content and silo stock are assumptions.
If a layer waits too long before the next one goes on, the joint between them gets weaker. The decay below is an assumed shape for teaching; the real curve has to be measured on your mix, for example with bond tests at different intervals.
| Rank | Nozzle (mm) | Speed (mm/s) | Layer (mm) | Print time (h) | Pressure (bar) |
|---|
The search tries every combination of nozzle, speed and layer height in a grid and keeps those that the pump can deliver, the plant can supply and the layer shape allows (width between 1.5 and 4 times the layer height). If a layer would be too quick for the wall to stand, a waiting time is added, so that time is counted too. The result is the fastest that passes all checks, not a guarantee.
| Change from the settings above | Print time (h) | Pressure (bar) | Demand (m³/h) | Verdict |
|---|
Each row changes one setting and keeps the others. Nozzle sizes, speeds and layer heights are the ones in the case. Use this to see which setting matters before you try it on the printer.
A simulated print feeds the dashboard. The twin knows what pressure to expect at the set speed. When a blockage builds, the pressure runs above that and the twin warns and suggests a lower speed that stays inside the pump limit.
The starting values are an example of typical improvements, not measurements. Replace them with what you measured before and after on your own jobs. The saving is the cost difference per job times the jobs per year.
| Stage | Tools shown in the case | What they do here |
|---|
The tools are examples, not recommendations. Choose by what your team owns and what your machines and PLCs support.
The simulators are simplified teaching models. Measure your own mix and machine, validate your own model, and have structures and safety reviewed by a qualified engineer.