Session: 13-03-01: Computational Studies on MEMS and Nanostructures

Paper Number: 146010

146010 - A Framework for Surface Scanning to Additive Repairing Using Point Cloud Data and Surface Reconstruction Algorithms 

The utilization of Additive Manufacturing is not restricted solely to fabrication purposes; rather, it also possesses the potential to make a significant contribution to the restoration of damaged surfaces. To effectively address the increasing demand for additive repair methods, it is of utmost importance to emphasize the process of precise three-dimensional (3D) surface scanning, as it holds a pivotal significance. The identification of the geometry of the damaged surface plays a crucial role in determining the necessary nozzle print path and filler material volume. This research article introduces a technique that employs additive manufacturing as an alternative approach for repairing solid bodies through the accurate scanning and reconstruction of three-dimensional surfaces. The proposed methodology involves the utilization of a laser profiler, specifically the Keyence LJX 8200, to perform a comprehensive scan of the surface profile and obtain point cloud data. The laser profiler functions as a line scan device that can provide 3200 data points for a single two-dimensional plane. To extend the scanning capabilities beyond a single line, a gantry system driven by a stepper motor is thoughtfully integrated to offer additional scanning freedom over a specific area of interest. This integration of the laser profiler and the gantry system effectively facilitates the precise measurement of coordinate values within the designated region. The synchronization between the movements of the gantry mechanism and the data acquisition from the laser profiler is seamlessly performed using a central computing system with the assistance of the Python programming language. Furthermore, it is important to highlight that the laser profiler provides invaluable information regarding the X-axis and Z-axis, while the gantry mechanism offers significant insights into the Y-axis. By considering the information obtained from these three axes, Point Cloud Data (PCD) data of the scanned surface is meticulously gathered. It is imperative to ensure the quality of the acquired PCD through the implementation of various preprocessing techniques (e.g., outlier elimination and data sorting) before exporting the final PCD file. The PCD file effectively contains crucial information solely about the vertices of the scanned surface in terms of coordinates; however, it does not provide information about the faces of the actual surface. Effective preparation of the surfaces requires a key step involving the triangulation among the vertices and determination of the normal of the faces. This intricate task is expertly addressed through the utilization of a surface reconstruction algorithm that actively prepares a triangular mesh by carefully generating both triangles and normals from the PCD data. Several surface reconstruction techniques, including Delaunay Triangulation, Ball Pivoting Algorithm, and Poisson's Surface Reconstruction, are thoughtfully employed on the point cloud dataset. After careful evaluation, the study concludes that among the tested algorithms, it is Poisson’s Surface Reconstruction that consistently produces the highest level of accuracy, thus solidifying its status as the optimal choice for surface reconstruction in additive repair applications. The 3D surface produced through this process is exported in the form of a Standard Tessellation Language (STL) file, which can be printed via a 3D printer. The precision of the surface reconstruction algorithm is measured by comparing the dimensions of the scanned item with that of the 3D-printed item. The proposed approach demonstrates notable dimensional accuracy in generating a digital 3D representation of a physical object. The application of the presented algorithm is not limited to additive repairing, rather it can be successfully implemented in additive manufacturing of wearable sensors, protective gears, soft robotics, and customizable biomedical equipment at various length scales.

Presenting Author: Radif Uddin Ahmed Louisiana Tech University

Presenting Author Biography: Radif Uddin Ahmed is a graduate research assistant at Louisiana Tech University. He is working under the supervision of Dr. M Shafiqur Rahman and his research is in the field of additive manufacturing.

Authors:

Radif Uddin Ahmed Louisiana Tech University
Iftesam Nabi Louisiana Tech University
Chowdhury Sadid Alam Louisiana Tech University
Yun Chen Louisiana Tech University
M Shafiqur Rahman Louisiana Tech University