Search

Additive manufacturing processes show potential and utility in the construction industry by enabling new design freedoms and increasing the degree of automation. Accurate prediction of structural deformations and optimization of process parameters is a crucial consideration to ensure better component quality, reduced test effort and fewer buildability failures. This paper introduces an approach to build a nonlinear model that maps the interlayer waiting time in the building process to the corresponding structural deformation based on FEM results obtained from a process based FEM simulation. The model is utilized to optimize the interlayer waiting time during the printing process using the maximum deformation of the structure as the objective function and the allocated printing time together with the allowed maximum interlayer waiting time as constraints. Finally, the validity of the approach is numerically verified using a test component.

In an effort to reduce prototypes in the industrial manufacturing of sheet metal components and their assembly, numerical simulation is well established. Where each single process transfers into its virtual counterpart to ensure quality assurance, disregarding the influence of their preceding processes. In order to further the accuracy of a virtual process, links between sequenced methods are established to create a process chain considering all influences on a manufactured part in its early design stages. If a finite element solver change occurs within this linkage, it is necessary to transfer and/or transform source mesh bound data to fit the target mesh, if said data may influence any simulation further down the chain. This data transfer, called “mapping”, depends on the utilized allocation and interpolation algorithm. This contribution aims to link the difference in mesh discretization and position to the resulting data quality depending on both meshes as well as various approximation and interpolation procedures. The considered geometry represents a S-shaped profile common in sheet metal forming and subsequent joining whereas the mapped data consist of scalar and tensorial values such as thickness and stresses. It is expected that a large difference in the meshes coupled with certain interpolation methods lead to significant errors during mapping between source- and target-data.

The bond strength between a thermoformed fibre reinforced thermoplastic sheet and an injected polymer is the limiting factor for the structural integrity of overmoulded thermoplastic composites. In this contribution, a simulation based digital twin of the thermoforming process is presented. From numerical parametric studies a reduced order model based on Proper Orthogonal Decomposition (POD) is developed. The combination with machine learning methods enables the real-time computation of arbitrary physical reliable temperature fields with sufficient accuracy to be used for design purposes and as inline quality gates.

Based on experiences with common additive manufacturing processes, the application in the construction industry opens up new design freedoms and cost-effective production of complex structures. However, the time-dependent yield strength of fresh concrete leads to deformations of underlying layers during the printing process, especially when using conventional path planning methods in combination with material extrusion or jetting methods. This paper presents a finite element model based approach to minimize the resulting deviations from the target geometry by iteratively adjusting the process parameters according to simulated deformations. To achieve more detailed modelling, the utilized finite element model is derived from the printing path instead of the CAD data.

In this contribution, a computational approach for the manufacturing of multi-material lightweight parts is presented. For an efficient online optimization procedure, a proper orthogonal decomposition on offline obtained numerical results is carried out to construct a surrogate model. The functionality of the proposed framework is demonstrated on a use case study of a multi-material component consisting of a sheet metal basic structure and a plastic reinforcement structure. The manufacturing process chain consists of a deep drawing process followed by an injection molding process of short fiber reinforced plastics. The proposed methodology provides a fast and accurate computational model for structural properties with respect to the process settings.

Based on their properties, PPS fibers are a promising material for reinforcing elastomeric components that are subjected to high mechanical and thermal loads. The use of this material is at present hindered because of the low adhesion between the fiber and matrix. Atmospheric pressure plasma treatments based on the dielectric barrier discharge were performed on PPS fibers using air as reactive gas for different treatment durations in order to improve the adhesion. The effects of these treatments have been characterized by determining the surface energy, and the residual tensile strength as well as by analyzing the surface chemistry. Required conditions for an improved wetting behavior and a significant increase in the polar component of the surface energy could then be identified.