Dynamic 4DCT Datasets for Compression of Additive Manufactured Parts. In Lawrence Livermore National Laboratory (LLNL) Open Data Initiative

Dynamic 4D time-space computed tomography (D4DCT) refers to reconstruction of moving or deformable objects over time while x-ray projections are acquired from multiple angles, in contrast to most conventional CT where objects are static during imaging. D4DCT offers broad applicability in examining object deformation and dynamics across various industrial and clinical applications. In material science and additive manufacturing (AM), the advanced capabilities of D4DCT facilitate the exploration of damage evolution resulting from dynamic thermal loads and mechanical stresses over time. This insight provides invaluable information for assessing the performance and safety of materials and structures.

We provide a real experimental D4DCT dataset of two AM parts and their compression using the Deben stage. The first 3D-printed lattice structure sample called "log-pile" was fabricated using an AM technique called direct ink writing (DIW). The second sample called "SiC cylinder" was made using another AM technique called binder jetting. The compression of those samples was performed using the Deben stage mounted in the Zeiss Versa scanner to observe in-situ deformation under mechanical loads over time. The projection data comprises 722 views over two full rotations, with each projection image having dimensions of 1024x1024 and a pixel size of 69.16 microns. The scanner operates under cone beam geometry.