Catia

Microfluidic systems were made of polydimethylsiloxane (PDMS). For the purpose of experimental flexibility, we constructed individual devices that had 5 linearly connected chambers, each containing 24 uniformly sized 125 nL cubic traps for spheroid formation (Fig. 2A). The dimension of each device is about 7.5 × 1 cm². The microfluidic device layout was designed using CATIA (Dassault Systemes, France), based on previously published design rules for spheroid culture chips to ensure proper oxygenation and nutrient supply^{31 (link)}. The mold was fabricated on poly-methyl methacrylate (PMMA) using micromachining. The device is made of two layers of PDMS: the bottom layer contains spheroid culture chambers with dimensions 500 × 500 × 500 µm³, and the top layer contains a straight channel covering all the culture chambers. Each device could be treated with a particular systemic agent, while each chamber could be treated with a particular RT dose (Fig. 2B).

All stent geometries were generated on Abaqus (v2020, Simulia, Dassault Systemes) with the wrap mesh plugin application to possess eight unit cells in the circumferential directions and seven unit cells in the longitudinal direction as observed in Figure 1. All stents were modeled using 2-node linear B31 beam elements (21 (link)). An element size of 0.1 mm was selected for all stent designs after a preliminary mesh sensitivity analysis and the corresponding number of mesh elements for each stent design has been enlisted in Table 2. Sharp edges observed on the stent geometries were modified using CATIA (Dassault Systèmes) software under feedback with Electro Optical Systems (EOS GmbH, Germany) AM engineers and exported in STL format for AM fabrication.

In this study, a non-parametric subject-specific solid CAD model of the human femur obtained from a CT scan, described in [33 ], was used to represent bone geometry. Materialise Mimics (ver. 17 Materialise, Leuven, Belgium) and Dassault Systèmes CATIA (V5R21, Dassault Systèmes, Paris, France) were used to create the surface model of the bone from medical images. The bone was modeled as a homogenous solid. This approximation was introduced in order to simplify the model and focus on the parameterization of SIF configuration and placement. The CAD model of the bone was prepared for assembly with the parametric CAD model of the fixator, by creating the necessary landmarks, namely anatomical points, axes, curves, and planes. All those objects were created on the femur model in order to serve as geometrical references for positioning the fixator and for the definition of loads and boundary conditions in subsequent finite element analysis (FEA). In the current study, this preparation was performed manually according to a predefined procedure, while the ultimate goal is to have the whole procedure automated. A fully parametric CAD model of SIF was used, and its position on the femur was defined via parametric constraints. SolidWorks (ver. 2015, Dassault Systèmes, Paris, France) was used to create the SIF model and the femur–SIF assembly.