0
Research Papers

Online Real-Time Quality Monitoring in Additive Manufacturing Processes Using Heterogeneous Sensors

[+] Author and Article Information
Prahalad K. Rao

Department of Systems Science and
Industrial Engineering,
State University of New York at Binghamton,
Binghamton, NY 13702

Jia (Peter) Liu

Grado Department of Industrial and
Systems Engineering,
Virginia Polytechnic Institute and
State University,
Blacksburg, VA 24061

David Roberson

Grado Department of Industrial and
Systems Engineering,
Virginia Polytechnic Institute and
State University,
Blacksburg, VA 24061

Zhenyu (James) Kong

Grado Department of Industrial and
Systems Engineering,
Virginia Polytechnic Institute and
State University,
Blacksburg, VA 24061
e-mail: zkong@vt.edu

Christopher Williams

Department of Mechanical Engineering,
Virginia Polytechnic Institute and
State University,
Blacksburg, VA 24061

1Corresponding author.

Contributed by the Manufacturing Engineering Division of ASME for publication in the JOURNAL OF MANUFACTURING SCIENCE AND ENGINEERING. Manuscript received September 17, 2014; final manuscript received February 6, 2015; published online September 9, 2015. Assoc. Editor: Robert Gao.

J. Manuf. Sci. Eng 137(6), 061007 (Sep 09, 2015) (12 pages) Paper No: MANU-14-1480; doi: 10.1115/1.4029823 History: Received September 17, 2014

The objective of this work is to identify failure modes and detect the onset of process anomalies in additive manufacturing (AM) processes, specifically focusing on fused filament fabrication (FFF). We accomplish this objective using advanced Bayesian nonparametric analysis of in situ heterogeneous sensor data. Experiments are conducted on a desktop FFF machine instrumented with a heterogeneous sensor array including thermocouples, accelerometers, an infrared (IR) temperature sensor, and a real-time miniature video borescope. FFF process failures are detected online using the nonparametric Bayesian Dirichlet process (DP) mixture model and evidence theory (ET) based on the experimentally acquired sensor data. This sensor data-driven defect detection approach facilitates real-time identification and correction of FFF process drifts with an accuracy and precision approaching 85% (average F-score). In comparison, the F-score from existing approaches, such as probabilistic neural networks (PNN), naïve Bayesian clustering, support vector machines (SVM), and quadratic discriminant analysis (QDA), was in the range of 55–75%.

FIGURES IN THIS ARTICLE
<>
Copyright © 2015 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Two parts made under identical FFF process conditions yet showing markedly different results. The part on the left was made from an ABS filament which was briefly exposed to light machine oil and thus attracted dirt due to static activity. The part on the right was made from unaffected ABS filament.

Grahic Jump Location
Fig. 2

(a) Schematic of the FFF process. (b) Schematic of the FFF setup instrumented with multiple in situ sensors used in this work for measuring process conditions in real-time (the instrumented setup is shown in Fig. 5).

Grahic Jump Location
Fig. 3

Summary of the research approach integrating experimental tests, sensor data, and DP modeling for real-time monitoring of FFF process

Grahic Jump Location
Fig. 4

Test part resembling the NAS 979 standard part used for experimental trials (note: dimensions are in inches)

Grahic Jump Location
Fig. 5

Heterogeneous sensor array for real-time monitoring of FFF process. (a) Noncontact IR temperature sensor for measuring meltpool temperature, borescope camera for real-time video capture capability, and thermocouple to measure extruder temperature. (b) MEMS accelerometer to measure extruder vibration. (c) MEMS accelerometer to measure table vibration, and thermocouples (×4) to measure table temperature.

Grahic Jump Location
Fig. 6

Diagram showing the average surface roughness (Ra, μm) for each of the thirteen TCs tested (averaged over three replications)

Grahic Jump Location
Fig. 7

Schematic representation of failure due to nozzle clog at low feed to flow ratio (v)

Grahic Jump Location
Fig. 8

(Top) The three observed process states in FFF, demarcated as normal, abnormal, and failure. (Bottom) The transition between normal, abnormal, and failure states is prominently evident from the time series pattern of the noncontact IR temperature sensor.

Grahic Jump Location
Fig. 9

The summary of the DP mixture model and ET approach developed in this work for real-time monitoring of process states in FFF

Grahic Jump Location
Fig. 10

Histogram with fitted probability density functions of IR sensor data at normal, abnormal, and failure process states. The signal patterns have a non-Gaussian probability density function, which is particularly pronounced for the normal state.

Grahic Jump Location
Fig. 11

Comparison of F-scores for various sensor combinations using DP + ET approach. The error bars are one standard deviation (σ) long.

Grahic Jump Location
Fig. 12

Comparison of F-scores for the three predefined process states in FFF using high-dimensional DP mixture modeling and our proposed DP + ET approach. The error bars are one standard deviation (σ) long.

Grahic Jump Location
Fig. 13

Comparison of F-scores for the three predefined process states in FFF with four widely used classification approaches. The error bars are one standard deviation (σ) long.

Grahic Jump Location
Fig. 14

The ten locations on the workpiece where the arithmetic mean surface roughness (Ra) values are measured

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In