Page 35 - MSAM-3-4
P. 35

Materials Science in Additive Manufacturing             Gyroid non-pneumatic tires through additive manufacturing



            generated using the solid wheel as the bounding geometric   published literature. The compression testing was
            body. Three separate TPMS designs were created with   performed in accordance with the ASTM D695 standard
            varying sheet thicknesses. Table 1 shows design properties   for compressive testing of rigid polymer tensile bars and
            for different tire designs. Tire 1 had a constant sheet   rigid cellular polymers.  Force and displacement were
                                                                                   51
            thickness of 1.0 mm, while Tire 2 and Tire 3 had radially   recorded by the load frame.
            ramped wall thickness, increasing linearly to 1.5 mm and   To accurately measure the deformation of each TPMS
            2.0 mm, respectively. This choice of thicknesses was used to   layer, we utilized the VIC-3D DIC system (Correlated
            keep volume fractions low while allowing for comparison   Solutions, USA). Each tire was speckled with a pattern of
            between the performance of constant and ramped lattices.   white dots, which were captured by the DIC cameras during
            The TPMS unit cell (UC) measured 5  mm in height by   compression testing. Using VIC-3D post-processing
            10 mm in the radial direction, repeated 16 times around the   software, these speckles were individually tracked. The
            central axis. Once the TPMS lattices were generated, they   positions of the top and bottom points of each layer were
            were added between the hub and the tread to complete the   recorded, allowing for the calculation of the layer thickness
            tire design. This operation ensured that the rims remained   at any given time. Changes in layer thickness were used to
            solid, and the TPMS lattice filled the regions between the   determine the deformation of the layer. Local deformation
            two while forming one part, resulting in a complete wheel   values for each band (L0, L1, L2, and L3) were recorded for
            with a robust support structure. The tire designs were   global displacement measurements across all tire variants
            manufactured through  a DLP  AM  process,  using  a  3D   (Tire 1, Tire 2, and Tire 3). These localized deformation
            printer (3D Systems Inc., United States of America [USA])   values for different bands indicate local stiffness in specific
            and a flexible, impact-resistant resin with high toughness,   band regions.
            FLEX-BLK 20. The DLP printer uses an ultraviolet (UV)
            lamp with a 405  nm wavelength and a high-definition   2.3. Finite element analysis
            digital micromirror device (DMD) chip with a pixel size
            of 65 μm. The layer thickness used for printing was 50 μm.   Numerical FEA for simulating the compression
            The resin used has a density and viscosity of 1.11 g/cm³   behavior of the NPTs was conducted using the static
            and 2250 cps, respectively. 50                     structural module of ANSYS software (version  2024
                                                               R1) to validate the experimental findings. The objective
            2.2. Experimental testing                          was to accurately constrain the problem  and simulate

            Compression testing was performed on the tire samples   compressive loading in the radial direction along a flat
                                   ®
            using the MTS Criterion  Electromechanical Test    road surface. Computer-aided design (CAD) of the tire
            System (equipped with a 50 kN load cell) to evaluate the   from nTopology was imported into Ansys Workbench,
            vertical deformation of the tires. A custom test rig was   and a tetrahedron mesh was used for all models.
            designed  and manufactured  to  replicate the boundary   A  convergence study was conducted on mesh sizes
            conditions of a tire under operation. The fixture was   ranging from 0.2 to 1.0 mm to determine an appropriate
            made of a 1040 steel shaft that was turned down to fit   element size for this FEA study with convergence at
            the D-clevis on the load frame. The axel that supports
            the tire was then attached to the shaft using a 3/4 inch
            pin. A  crosshead displacement rate of 0.02  mm/s was
            used, which is equivalent to 10% of the sample thickness
            per minute. The tire moves down and is compressed
            against the bottom platten, which is analogous to the
            tire surface in contact with the road, being compressed
            by the weight of the vehicle. No standardized methods
            for testing sub-scale tires were found in previously


            Table 1. Local band regions for different tire designs
            Tire design  Wall thickness (mm)  Layer thickness (mm)
                                        L0   L1   L2   L3
            Tire 1     1.0              3.4  4.5  4.3   4.4
            Tire 2     1.0 – 1.5        3.4  4.4  4.0   3.9
                                                               Figure 2. Boundary condition constraints (with an applied displacement
            Tire 3     1.0 – 2.0        3.4  4.1  3.8   3.6    of 12 mm) in the finite element analysis


            Volume 3 Issue 4 (2023)                         4                              doi: 10.36922/msam.5022
   30   31   32   33   34   35   36   37   38   39   40