Tensile Tests of Different Families of Ultra-Lightweight Photosensitive Polymer Micro-Lattices

Micro-lattice manufacturing and post-processing

The following is a detailed description of the manufacturing and post-treatment process used in this study:

For all manufacturing processes, particular attention must be invested in post-processing the parts. The highest risks of damaging the lattice occur during the handling phases before the part is completely hardened. The samples are fragile, and their transportation must be carefully planned.

The design (nTopology software), 3D model (.stl) and manufacturing files (sliced elements for the printers used) are available online on Lab INIT Robots GitLab

With the SLA Form 3, the printing preparations are carried out on the Preform software. Formlabs allows the user to access more customization parameters thanks to which layer heights as small as 25 μm can be reached. All samples were produced with Standard Grey Resin (Product code: RS-F2-GPGR-04). After printing, the parts were cleaned in an isopropyl alcohol bath for 10 minutes before being post-cured with UV light in the Form Wash and Form Cure stations. The part was exposed for 60 minutes at 60C with a 10-minute break in the middle of this process to remove excess print media from the cleaning solution. This gives the mesh enough rigidity to withstand the removal of the support beams without over-curing them either. The times are those recommended by Formlab with a 72% gain in strength versus 65% after 30 minutes according to the manufacturer: FormCurePost-CureSettings.pdf

With the MSLA Elegoo Mars, printing was configured on Chitubox Basic (Chitu Systems - Version 1.9.1). The parameters are adjustable to the same extent as with the Formlabs Preform software. The parts were made with a black resin from Nova with a 50 μm layer height. This resin was cleaned with water and the curing process was identical to the samples made with Form 3.

For the PolyJet Objet30 Pro, the print was prepared with the Objet Studio slicer. The printer relies on proprietary software that strongly limits the parameters that the user can alter. The position of the part on the plate is the only parameter that the user can edit. These limitations by Stratasys are intended to ensure that parts will be produced reliably with a high rate of success.
The printer is equipped with two tanks of resin: a tank that contains the resin used to print the model itself - we select the Veroclear resin, and a tank that contains the resin used for the support - we used the SUP706B resin. VeroClear resin allows for the use of the smallest layer height in the high-definition mode in the slicer: 16 μm.
With the PolyJet technique, the support material wraps completely the sample with a ''matte" finish or we can choose to only use supports under the part for a ''glossy" rendering of the top layers.
To keep the same conditions for all struts of the lattice, the ''matte" finish was applied to all specimens. Then the supports are removed in a post-process step, most commonly (SUP705 resin) using a cleaning station equipped with a pressure jet.
Due to the fragility of our lattice struts, this post-treatment method was not feasible: the beams would have been torn off by the pressurized water jet.
Stratasys also provides a soluble support material that decreases the need for pressure jet cleaning: the SUP706B resin. This material is soluble in a water-based solution containing 2% caustic soda and 1% sodium metasilicate. Nevertheless, the micro-lattice's density prevented the solution from penetrating the core of the network.
In the end, to clean the smallest areas, we developed a post-process in four cycles, each divided into two steps: (1) dipping the sample in the solution for ten minutes and (2) rinsing it with water to removing pieces of support.

The solvent can interact with the polymer in several ways. First, the solvent can dissolve the polymer, making it softer and allowing it to flow more easily. Secondly, solvent-polymer interactions can also affect the physical properties of the polymer, such as viscosity, density, and tensile strength. In addition, the presence of a solvent can change the degree of polymerization of the polymer by adding or removing monomers to the polymer chain.

Finite Element Analysis of Micro-lattices

Today, the most common approach to computing and validating the mechanical resistance of non-trivial structures is FEA. Unfortunately, the macro, meso, and micro differences between the digital model and the manufactured part are very important. It is sometimes possible to geometrically rebuild [1] the final model thanks to CT-scan. However, the inter-layer adhesion (microlevel) does impact the mechanical performances of the parts. Several research works are beginning to propose models like voxel-based simulation [2] for inter-layer adhesion. More specifically, there is no work for modelling defects in parts printed from resin. The authors focus on the effects of FDM and powder-based processes even though this work summarises a very interesting methodology for studying micro-lattices using FEA [3]. There is still a lot to be done in adapting these techniques to other processes, as the micro-structure changes with each method and machine. Despite these limitations, FEA can be applied to small sections of the lattice pattern. It provides unique information: studying sections of that scale is not possible with physical specimens since the lattice is too small to be mechanically loaded with the equipment available. Nevertheless, the results are indicative at best as the micro-structure is not modelled. We first conducted a tensile analysis on a group of 4x4 BCC micro-lattice cells of 2.5 mm size in the Autodesk Fusion 360 “simulation” module. Figure 1 illustrates the displacement distribution between cells. One grip section was set as fixed and a load of 100N was applied to the other. The architectural construction of a lattice splits the load at each node along the geometry. Figure 1 shows a large discontinuity depending on the placement of the cell in the micro-lattice. For instance, edge effects result from fewer neighboring cells as illustrated by the red outline at the border of the network. These differences in displacements will have an impact on the location of the first failure which will propagate under the constraints, as demonstrated by [4]. The authors quantified the loss of stiffness on the free edges of the mesh by computation: the size of the cells' network and its arrangement will also have a great impact on the stress and displacement distribution. They are useful tools to quickly highlight critical failure areas.

The second set of tests focused on 2x2x2 cells of each pattern illustrated in Fig.2. Large differences in stress distribution are observed between TPMS meshes (gyroid) where stress areas are widely distributed and other strut meshes where the nodes of the lattices show strong stress gradient areas in Fig.2. The constitutive model is elastoplastic with an isotropic material whose characteristics are taken from the preliminary tests in the Section of bulk material characterization. The solver is Abaqus with a tetrahedral free mesh.

[1] Xiao, L., Song, W., Xu, X., 2020. Experimental study on the collapse behavior of graded Ti-6Al-4V micro-lattice structures printed by selective laser melting under high speed impact. Thin-Walled Structures 155, 106970. https://doi.org/10.1016/j.tws.2020.106970
[2] Park, S., Rosen, D.W., 2016. Quantifying effects of material extrusion additive manufacturing process on mechanical properties of lattice structures using as-fabricated voxel modeling. Additive Manufacturing, Special Issue on Modeling & Simulation for Additive Manufacturing 12, 265–273. https://doi.org/10.1016/j.addma.2016.05.006
[3] Dong, G., Tang, Y., Zhao, Y.F., 2017. A Survey of Modeling of Lattice Structures Fabricated by Additive Manufacturing. Journal of Mechanical Design 139. https://doi.org/10.1115/1.4037305
[4] Yoder, M., Thompson, L., Summers, J., 2019. Size effects in lattice-structured cellular materials: edge softening effects. Journal of Materials Science 54. https://doi.org/10.1007/s10853-018-3103-9