122 C. Casavola et al. Some papers have dealt with experimental measurements of residual stress distribution in plastic parts [12–14] but few works in FDM parts [11, 15]. Turnbull et al. [12] carried out a comparison among several techniques to measure residual stresses in ABS, Polycarbonate, and Nylon. They concluded that the hole drilling can be employed as a valid measurement method to measure residual stresses in plastic materials. Nau et al. [13] highlighted that the process parameters and procedures applied for stress analysis in metallic materials cannot be employed in polymers. They pointed out that the surface preparation of specimens, the strain gauge bonding, and the drilling speed are critical issues in order to obtain a correct measure. However, both Turnbull et al. [12] and Nau et al. [13] did not consider the local reinforcement effect that the installation of a rosette produces in materials that have a low Young’s modulus. Indeed, Magnier et al. [14] studied the influence of material viscoelasticity, room temperature and local reinforcement of the strain gauge on the measure of deformation by HDM of plastic materials. They highlighted that the use of strain gauge to measure the deformation on plastic materials can produce a difference up to 30% between the results recorded by strain gauge and DIC. Casavola et al. [11, 16–17] studied the effect on residual stresses of the raster orientation in FDM parts. They found that the stacking sequence C45ı/ 45ı shows the lowest values of residual stresses. Moreover, they highlighted that there is not a clear difference between the bottom and the top of the printed specimens. Only one paper has tried to deal with the residual stress issues in FDM part by numerical simulation. Zhang and Chou [18], using simplified material properties and boundary conditions, have simulated different deposition patterns and have demonstrated the feasibility of using the element activation function to reproduce the filament deposition. They found that there was a modification of the residual stress distributions changing the tool-path pattern. However, they did not validate their model using residual stress measurements but only by comparing the distortion of the printed part and the numerical prediction. The aim of the present work is to measure the residual stresses in several points of printed parts, both on top and bottom, in order to verify if the constrain conditions employed during the printing produce a substantial variation from a point to another. The residual stresses have been measured in ABS parts employing the hole-drilling method. In order to avoid the local reinforcement of the strain gage, an optical technique, i.e. ESPI (electronic speckle pattern interferometry), is employed to measure the displacement of the surface due to the stress relaxation and, consequently, to calculate the residual stresses. 18.2 Materials and Methods A RepRap Prusa i3 equipped with a marlin firmware and a nozzle with a diameter of 0.4 mm has been employed to produce the specimens. These have a rectangular shape and the dimensions of 80 40 7 mm. Four stacking sequences have been studied, i.e., the raster angles are ˙30ı, ˙45ı, 0ı/90ı and 0ı only. A layer with a 0ı raster angle has the deposited beads parallel to the major side of the specimen. Moreover, the samples have been manufactured with the minimum dimension of the part perpendicular to the build platform. The Fig. 18.1 shows the coordinate system for the deposition and for the residual stresses. The parameters reported in Table 18.1, such as the layer thickness or the number of contour lines, have been kept constant for every specimen. In Table 18.1, the air gap is the distance between two, adjacently deposited, beads of the same layer; the layer thickness and the bead width are respectively the height and the width of a deposited filament. The number of contours represents how many edges have been deposited before filling the inner part by inclined beads. The bed temperature has been set to 90 ıC and some glue on the bed has been employed to reduce the warping effect. The solid model, created using a 3D CAD, has been sliced using the open source software Slic3r. Fig. 18.1 Specimen examples with˙45ı (a) and 0ı/90ı (b) stacking sequence
RkJQdWJsaXNoZXIy MTMzNzEzMQ==