Producing fractures with a 3D-printer for flow experiments – the price to pay for the easy way out

Recently, the technology of 3D-printing has been applied to literally look into flow and transport processes in fractures and fracture system using transparent material for printing. Preparing some own experiments with printed fractures it was noticed, though, that very little is known about dimensional accuracy of printed components. There are indications that fractures, as a rule of thumb, require an aperture of at least ten times of the printing resolution that is claimed by the manufacturer in order to be water conducting at all. Seemingly little attention has also been paid to the roughness of the printed fracture surfaces that would affect flow in an allegedly smooth plane fracture. Furthermore, the majority of resins  that are used for 3D-printing take up water which leads to swelling of the printed components. This has an up to now unknown influence on fracture aperture.

For these reasons a list of simple tests has been developed to check and compare the geometrical and physical properties of printed components right after production and under subsequent influence of water. Varied were printers, materials, sample orientations, sample geometries and postprocessing parameters like curing time. The dimensions of dry and wet samples were meticulously measured. Exemplarily on one sample, also the water uptake dynamics under the influence of water vapour under different degrees of saturation have been determined. A further test concerned the penetration depth of water.

The results form a little data base. Sample geometries and testing methods have been kept simple deliberately to allow for a meaningful comparison, not only for ourselves but also for other potential testing parties that might want to contribute data in exchange for the whole data base.

The major general conclusion that can be drawn from the tests is that contact time of 3D-prints with water should be minimised if dimensional accuracy is of importance. Tests concerning fracture flow in printed samples thus need to be performed as quickly as possible.