Metal Additive Manufacturing Open Repository

LMD dataset

This dataset gathers data from different parts of the Laser Metal Deposition process (Additive Manufacturing). The dataset covers not only the process data, but also the design, NDT (Non-Destructive Testing) and dimensional inspection.

Motivation

The industrialisation of Additive Manufacturing (AM) requires a holistic data management and integrated automation. INTEGRADDE aims to develop an end-to-end Digital Manufacturing solution, enabling a cybersecured bidirectional dataflow for a seamless integration across the entire AM chain.

The goal is to develop a new manufacturing methodology capable of ensuring the manufacturability, reliability and quality of a target metal component from initial product design via Direct Energy Deposition (DED) technologies, implementing a zero-defect manufacturing approach ensuring robustness, stability and repeatibility of the process.

To achieve this aim, INTEGRADDE addresses following key innovations:

•  Development of an intelligent data-driven AM pipeline.
•  Combination of automatic topology optimisation algorithms for design, multi-scale process modelling, automated hardware-independent process planning, online control and distributed NDT for the manufacturing of certified metal parts.
•  A self-adaptive control is adopted focused on the implementation of non-propagation of defects strategy. Moreover, Data Analytics will provide a continuous refinement by acquiring process knowledge to assist in the manufacturing of new metal components, improving right-first-time production by adopting a mass customization approach
•  Cybersecurity ensures data integrity along the AM workflow, providing a novel manufacturing methodology for the certification of metal AM parts.

INTEGRADDE implements a twofold deployment approach for the pilot lines: both in application-driven at five industrial end-users (steel, tooling, aeronautics, and construction) and open-pilot networks at RTOs already owning AM infrastructure (AIMEN, IREPA, CEA, WEST). This will allow a continuous validation and deployment of specific developments towards industrialization, boosting definitive uptake of AM in EU metalworking sector.

Authors

• Carlos Gonzalez-Val: Main contact (carlos.gonzalez@aimen.es)
• Baltasar Lodeiro
• Marcos Diez

Entities

This dataset was collected under the INTEGRADDE project. Attributions:

• AIMEN: Process data collection and manufacturing of P1, P2, P3 and P4
• CEA: Tomography analysis.
• DATAPIXEL: Dimensional inspection.

Structure

The dataset follows this structure:

• Dataset
  • [SAMPLE 1 NAME]
    • README: metadata and information about the sample. Format: txt.
    • Photo: a photo of the manufactured sample. Format: jpg.
    • Design: a 3D design file of the piece before manufacturing (original design). Format: stl.
    • Trajectories: the trajectories followed for the manufacturing. Format: gcode.
    • Process data: data recorded from the process. Format hdf5.
    • Tomography: data from a 3D tomographic reconstruction. Format: raw.
    • Dimensional inspection: A comparison
  • [SAMPLE 2 NAME]
    • ...

Further information and metadata is contained in each stage's subdirectory.

Note that not all the samples contain all the stages.

Software

To open the different files that conform the dataset, we recommend the following Open softwares:

•  hdf5 -> HDF5 Viewer: https://www.hdfgroup.org/downloads/hdfview/
•  stl/amf -> Slic3r: https://slic3r.org / OpenJScad: https://openjscad.org/
•  stp -> ShareCad: https://beta.sharecad.org/
•  gcode -> Text editor / Slic3r: https://slic3r.org/
•  raw -> ImageJ: https://imagej.net/