scholarly journals Quantifying Geometric Accuracy With Unsupervised Machine Learning: Using Self-Organizing Map on Fused Filament Fabrication Additive Manufacturing Parts

2017 ◽  
Vol 140 (3) ◽  
Author(s):  
Mojtaba Khanzadeh ◽  
Prahalada Rao ◽  
Ruholla Jafari-Marandi ◽  
Brian K. Smith ◽  
Mark A. Tschopp ◽  
...  
Keyword(s):  
Machine Learning ◽  
Additive Manufacturing ◽  
Process Conditions ◽  
Complex Geometries ◽  
Self Organizing Map ◽  
Geometric Accuracy ◽  
Fused Filament Fabrication ◽  
Unsupervised Machine Learning ◽  
Geometric Deviations ◽  
Self Organizing

Although complex geometries are attainable with additive manufacturing (AM), a major barrier preventing its use in mission-critical applications is the lack of geometric accuracy of AM parts. Existing geometric dimensioning and tolerancing (GD&T) characteristics are defined based on simple landmark features, and thus, need to be customized to capture the subtle difference in parts with complex geometries. Hence, the objective of this work is to quantify the geometric deviations of additively manufactured parts from a large data set of laser-scanned coordinates using an unsupervised machine learning (ML) approach called the self-organizing map (SOM). The central hypothesis is that clusters recognized by the SOM correspond to specific types of geometric deviations, which in turn are linked to certain AM process conditions. This hypothesis is tested on parts made while varying process conditions in the fused filament fabrication (FFF) AM process. The outcomes of this research are as follows: (1) visualizing and quantifying the link between process conditions and geometric accuracy in FFF and (2) significantly reducing the amount of point cloud data required for characterizing of geometric accuracy. The significance of this research is that this unsupervised ML approach resulted in less than 3% of over 1 million data points being required to fully quantify the part geometric accuracy.

scholarly journals Multi-Objective Accelerated Process Optimization of Part Geometric Accuracy in Additive Manufacturing

2017 ◽  
Vol 139 (10) ◽  
Author(s):  
Amir M. Aboutaleb ◽  
Mark A. Tschopp ◽  
Prahalad K. Rao ◽  
Linkan Bian
Keyword(s):  
Experimental Data ◽  
Additive Manufacturing ◽  
Process Optimization ◽  
Process Parameters ◽  
Process Conditions ◽  
Optimization Approach ◽  
Geometric Accuracy ◽  
Multi Objective Optimization ◽  
Fused Filament Fabrication ◽  
Multi Objective

The goal of this work is to minimize geometric inaccuracies in parts printed using a fused filament fabrication (FFF) additive manufacturing (AM) process by optimizing the process parameters settings. This is a challenging proposition, because it is often difficult to satisfy the various specified geometric accuracy requirements by using the process parameters as the controlling factor. To overcome this challenge, the objective of this work is to develop and apply a multi-objective optimization approach to find the process parameters minimizing the overall geometric inaccuracies by balancing multiple requirements. The central hypothesis is that formulating such a multi-objective optimization problem as a series of simpler single-objective problems leads to optimal process conditions minimizing the overall geometric inaccuracy of AM parts with fewer trials compared to the traditional design of experiments (DOE) approaches. The proposed multi-objective accelerated process optimization (m-APO) method accelerates the optimization process by jointly solving the subproblems in a systematic manner. The m-APO maps and scales experimental data from previous subproblems to guide remaining subproblems that improve the solutions while reducing the number of experiments required. The presented hypothesis is tested with experimental data from the FFF AM process; the m-APO reduces the number of FFF trials by 20% for obtaining parts with the least geometric inaccuracies compared to full factorial DOE method. Furthermore, a series of studies conducted on synthetic responses affirmed the effectiveness of the proposed m-APO approach in more challenging scenarios evocative of large and nonconvex objective spaces. This outcome directly leads to minimization of expensive experimental trials in AM.

scholarly journals Data-Driven Microstructure and Microhardness Design in Additive Manufacturing Using a Self-Organizing Map

Engineering ◽  
2019 ◽  
Vol 5 (4) ◽  
pp. 730-735 ◽  
Author(s):  
Zhengtao Gan ◽  
Hengyang Li ◽  
Sarah J. Wolff ◽  
Jennifer L. Bennett ◽  
Gregory Hyatt ◽  
...  
Keyword(s):  
Additive Manufacturing ◽  
Data Driven ◽  
Self Organizing Map ◽  
Self Organizing

Geometric Accuracy Prediction for Additive Manufacturing Through Machine Learning of Triangular Mesh Data

2019 ◽  
Author(s):  
Nathan Decker ◽  
Qiang Huang
Keyword(s):  
Machine Learning ◽  
Additive Manufacturing ◽  
Three Dimensional ◽  
Triangular Mesh ◽  
Geometric Accuracy ◽  
New Approach ◽  
Machine Learning Model ◽  
3D Printed ◽  
3D Shapes ◽  
Complex Parts

Abstract While additive manufacturing has seen tremendous growth in recent years, a number of challenges remain, including the presence of substantial geometric differences between a three dimensional (3D) printed part, and the shape that was intended. There are a number of approaches for addressing this issue, including statistical models that seek to account for errors caused by the geometry of the object being printed. Currently, these models are largely unable to account for errors generated in freeform 3D shapes. This paper proposes a new approach using machine learning with a set of predictors based on the geometric properties of the triangular mesh file used for printing. A direct advantage of this method is the simplicity with which it can describe important properties of a 3D shape and allow for predictive modeling of dimensional inaccuracies for complex parts. To evaluate the efficacy of this approach, a sample dataset of 3D printed objects and their corresponding deviations was generated. This dataset was used to train a random forest machine learning model and generate predictions of deviation for a new object. These predicted deviations were found to compare favorably to the actual deviations, demonstrating the potential of this approach for applications in error prediction and compensation.

scholarly journals Cataloguing the radio-sky with unsupervised machine learning: a new approach for the SKA era

2020 ◽  
Vol 497 (3) ◽  
pp. 2730-2758 ◽  
Author(s):  
T J Galvin ◽  
M T Huynh ◽  
R P Norris ◽  
X R Wang ◽  
E Hopkins ◽  
...  
Keyword(s):  
Machine Learning ◽  
A Priori ◽  
Radio Continuum ◽  
Host Galaxy ◽  
Wide Field ◽  
Self Organizing Map ◽  
Unsupervised Machine Learning ◽  
New Approach ◽  
Som Algorithm ◽  
Infrared Sources

ABSTRACT We develop a new analysis approach towards identifying related radio components and their corresponding infrared host galaxy based on unsupervised machine learning methods. By exploiting Parallelized rotation and flipping INvariant Kohonen maps (pink), a self-organizing map (SOM) algorithm, we are able to associate radio and infrared sources without the a priori requirement of training labels. We present an example of this method using 894 415 images from the Faint Images of the Radio-Sky at Twenty centimeters (FIRST) and Wide-field Infrared Survey Explorer (WISE) surveys centred towards positions described by the FIRST catalogue. We produce a set of catalogues that complement FIRST and describe 802 646 objects, including their radio components and their corresponding AllWISE infrared host galaxy. Using these data products, we (i) demonstrate the ability to identify objects with rare and unique radio morphologies (e.g. ‘X’-shaped galaxies, hybrid FR I/FR II morphologies), (ii) can identify the potentially resolved radio components that are associated with a single infrared host, (iii) introduce a ‘curliness’ statistic to search for bent and disturbed radio morphologies, and (iv) extract a set of 17 giant radio galaxies between 700 and 1100 kpc. As we require no training labels, our method can be applied to any radio-continuum survey, provided a sufficiently representative SOM can be trained.

scholarly journals Design and additive manufacturing of brushless electric motor components

2021 ◽  
Vol 343 ◽  
pp. 01007
Author(s):  
Camil Lancea ◽  
Valentin-Marian Stamate ◽  
Lucia-Antoneta Chicoş ◽  
Sebastian-Marian Zaharia ◽  
Alin-Mihai Pop ◽  
...  
Keyword(s):  
Additive Manufacturing ◽  
Electric Motor ◽  
Fabrication Process ◽  
Complex Geometries ◽  
Fused Filament Fabrication ◽  
Additive Process ◽  
Manufacturing Costs ◽  
Manufacturing Accuracy

The electric motor components are manufactured through the additive process of fused filament fabrication in order to verify the functionality of the electric motor assembly. This process was chosen due to the advantages it confers: fast obtaining of components, low manufacturing costs, no tools required for processing or for moulds manufacturing. Through the fused filament fabrication process, parts with complex geometries, which cannot be obtained by classical machining, can be manufactured. Due to the above-mentioned advantages, this technology is extremely useful for the manufacture and testing of prototypes. The paper aims to manufacture components of a brushless electric motor in order to verify the assembling compatibility and manufacturing accuracy.

scholarly journals SEM-based automated mineralogy (SEM-AM) and unsupervised machine learning studying the textural setting and elemental association of gold in the Rajapalot Au-Co area, northern Finland

2021 ◽  
Vol 93 (2) ◽  
pp. 129-154
Author(s):  
Jukka-Pekka Ranta ◽  
◽  
Nick Cook ◽  
Sabine Gilbricht ◽  
◽  
...  
Keyword(s):  
Machine Learning ◽  
Late Phase ◽  
Processing Technique ◽  
Geochemical Data ◽  
Self Organizing Maps ◽  
Unsupervised Machine Learning ◽  
Automated Mineralogy ◽  
Relationship Of ◽  
Elemental Associations ◽  
Self Organizing

SEM-based automated mineralogy (SEM-AM) techniques allow fast and effective way of studying the textural settings of gold in hydrothermal deposits. Unsupervised machine learning (e.g. self-organizing maps) is an intuitive way of processing multi-dimensional geochemical datasets in order to reveal hidden patterns potentially represent different mineralization stages. We combined these two methods for studying the relationship of gold and cobalt within different prospects in a Paleoproterozoic gold-cobalt mineralized area known as Rajapalot. Gold is found as a texturally late phase, occurring in fractures of silicates and sulfides. Based on the elemental associations observed from the whole-rock geochemical dataset using self-organizing-maps, Co-only, Au-Co and Au associations can be inferred relating to either different mineralization stages or different fluid-host rock interactions. Also, the dominant mineralization-related alteration in different occurrences within the Rajapalot Au-Co prospects are reflected as elemental associations with gold in the geochemical data. Our study shows the effectiveness SEM-AM methods for studying distribution of valuable minerals. Unsupervised neural networks provide for easy and intuitive processing technique that can be validated with the mineralogical observations.

Sign in / Sign up

Export Citation Format

Share Document