Recent talk on Geometric Shape Accuracy Control

Talks were given recently at University of Michigan-Ann Arbor and Georgia Tech. Slides can be downloaded here.

Methodology Extension from Stereolithography (SLA) to Selective Laser Melting (SLM)

     An initial study of SLM experimental data demonstrates that our prescriptive modeling methodology can be extended to metal based additive manufacturing processes.
 
   Based on four simple test shapes, we can predict the in-plane deformation profile of a freeform shape. Below is the predicted deformation (in solid line) comparing with CMM measurement (red dots). This is a joint study with Prof. Colosimo and Prof. Grasso at Politecnico di Milano.

                                      The plot was generated by PhD student Ms. HE Luan at USC.

Collaboration on Accuracy Control for Selective Laser Melting

Professor Colosimo and Professor Huang have initiated a collaborative effort to control geometric accuracy for Selective Laser Melting (SLM). Initial experimentation and data collection were conducted at Politecnico di Milano (Polimi), where Prof. Huang spent his sabbatical leave in spring 2016.

Professors Grasso, Colosimo, and Huang at Polimi

Talk on "Smart Calibration for Cyber-Physical Additive Manufacturing Systems" at ENS Cachan

Dr. Huang visited LURPA lab at ENS Cachan, France and exchanged research ideas with Dr. Anwer about AM related research such as FACAM and Skin Model Shapes. A talk on "Smart Calibration for Cyber-Physical Additive Manufacturing Systems" was given on July 21, 2016.

Prof. Anwer, Prof. Huang, and Mr. Zhu at LURPA (from left to right) 

Talk on Model Transfer - A Joint work with Dr. Arman Sabbaghi

STADIUS Seminar  Qiang Huang

Start date: 12/07/2016
Time: 16:00 - 17:00
Location: B02.35

Title: Model Transfer Through Engineering Effect Equivalence 

Abstract

A fundamental issue in science and engineering is whether a model developed in a study under one setting can be successfully transferred and applied to new settings. Bottlenecks to model transfer are introduced by lurking variables, or factors whose levels are completely unobserved due either to negligence, infeasibility of measurement, or insufficient knowledge (Box, 1966). This talk presents a scheme of model transfer based on engineering effect equivalence concept. The total equivalent amount of the lurking variables is represented in terms of an observed base factor with respect to a model feature. Bayesian methodology for modeling the total equivalent amount are developed under this condition. Applications in additive manufacturing are demonstrated.

https://www.esat.kuleuven.be/stadius/event.php?view=1&id=2060

Cybermanufacturing Systems (CM) - A New NSF program under CMMI Division

A new NSF program is established for Cybermanufacturing Systems.  Our FACAM research could contribute to this area.

Q. Huang Sabbatical at Politecnico Di Milano

Division of Civil, Mechanical and Manufacturing Innovation Cybermanufacturing Systems  (CM)  CONTACTS Name Email Phone Room Bruce  . Kramer bkramer@nsf.gov (703) 292-5348    PROGRAM GUIDELINES Apply to PD 16-018Y as follows: For full proposals submitted via FastLane: standard Grant Proposal Guide proposal preparation guidelines apply. For full proposals submitted via Grants.gov: the NSF Grants.gov Application Guide; A Guide for the Preparation and Submission of NSF Applications via Grants.gov Guidelines applies. (Note: The NSF Grants.gov Application Guide is available on the Grants.gov website and on the NSF website at: http://www.nsf.gov/publications/pub_summ.jsp?ods_key=grantsgovguide) Important Information for Proposers A revised version of the NSF Proposal & Award Policies & Procedures Guide (PAPPG) (NSF 16-1), is effective for proposals submitted, or due, on or after January 25, 2016. Please be advised that, depending on the specified due date, the guidelines contained in NSF 16-1 may apply to proposals submitted in response to this funding opportunity. DUE DATES Full Proposal Accepted Anytime Unsolicited full proposals may be submitted at any time.  There is no proposal submission window, deadline or due date. Proposals submitted to solicitations, such as the CAREER Program, must meet the respective deadlines of those solicitations. SYNOPSIS The Cybermanufacturing Systems (CM) Program supports fundamental research to enable the evolution of a wide range of network-accessed manufacturing services that: employ applications (or “apps”) that reside in the “cloud” and plug into an expansible, interactive architecture; are broadly accessible, guarantee reliable execution and have capabilities that are transparent to users; and are accessible at low cost to innovators and entrepreneurs, including both users and providers. Current manufacturing software applications are predominantly large, manufacturer-centric, general-purpose programs with the universal applicability needed to justify their development, marketing and acquisition costs.  They usually have broad capabilities, but are cumbersome to learn and often require expert intervention.  There is an opportunity for researchers to pursue research and educational efforts to accelerate the creation of an interoperating, cross-process manufacturing service layer that enables the rapid, bottom-up transformation of access to manufacturing services.  Such a service layer can allow creative entrepreneurs and companies to both furnish and access manufacturing apps that span the full spectrum from ideation to physical realization, giving rise to an era of “cybermanufacturing.” The cybermanufacturing service layer differs from existing Internet services in that it needs an architecture that can incrementally incorporate and organize the rich and deep semantic elements of manufacturing knowledge, requiring an almost unlimited capacity to expand the range and depth of content contributed in the form of partitioned, but interoperating, manufacturing applications.  Such efforts are well-suited to incubation in universities, where potential service layer architectures and application modules can be prototyped at low cost, used in coursework and tested by students and faculty. Of particular interest is the exploration of the tradeoffs between generality and tractability in algorithmic representations of manufacturing knowledge.  In the classic example, the automation of integrated circuit manufacturing depends on restricting device design options to those that can be produced with 100% reliability by a standardized set of manufacturing processes.  As a result, the problem of compiling manufacturing instructions is made tractable by limiting available design options to those that can be manufactured using proven methods.  In practice, the considerable design inefficiencies due to such limitations are more than compensated for by the cost savings due to dependable execution. Research areas of interest include, but are not limited to, the following: Frameworks for partitioning the mechanical design space to ensure tractability of design-to-manufacturing translation, possibly by part type or application domain; Computer-Aided Design (CAD) engines that facilitate the restriction of design options, possibly by facilitating the creation of generic part designs that can be customized by entering a limited number of dimensional parameters; Product- and domain-focused parametric design apps that connect to manufacturing resources and incorporate process constraints to enable part design and fabrication by users who lack detailed process knowledge; Software systems for generating and verifying machine instructions and providing guidance in design for manufacturability; Model-based process and machine controls that plug-and-play in a strongly integrated and networked environment; Methods for selecting and efficiently allocating networked manufacturing resources; Process and materials selection systems; Methods for establishing and maintaining evidence-based certification and controlled visibility of explicit and implicit assumptions; System architectures that are implementable using existing Internet protocols or that aim to identify the specific changes that are needed to existing Internet protocols to improve their effectiveness; Software and protocols for promoting and accommodating user-developed, interoperating manufacturing apps, including hardware computing platforms, operating systems, and middleware; and Methods for safeguarding the security and trustworthiness of cybermanufacturing system elements and integrating them to support end-to-end assurances. Collaborations between engineering and computer science faculty are strongly encouraged, as are collaborations with software, networking, internet service and industrial companies, including the partner institutes of the National Network for Manufacturing Innovation (NNMI, http://manufacturing.gov/welcome.html) and their member companies. Proposals with industry collaborations can be submitted to the CM Program as Grant Opportunities for Academic Liaison with Industry (GOALI) proposals.  GOALI proposals have special requirements, as specified in the most recent GOALI solicitation, https://www.nsf.gov/publications/pub_summ.jsp?ods_key=nsf12513.

FACAM Workshop Photos

Prof. Huang (USC)

Prof. Chen (USC)

Prof. Sabbaghi (Purdue)

Prof. Plumlee (U of Michigan)

Prof. Wang (FSU)

Dr. Zeng (HP)

Participants from Purdue, FSU, HKUST, NSF, USC, HP,  U. of Michigan, and USC(left to right)

Bambo (Harvard)

Andi (HKUST)

HE (USC)

Afternoon Session with PhD students in fronts (Andi, Raquel, Megan, Yuan, HE)

Before Workshop Dinner at RPV, CA

After Workshop - LA Live

After workshop - Staples Center @ LA Live

Fun Time After Workshop @Tom's Urban

Fun Time after Workshop at LA Live

Upcoming FACAM 2016 Workshop Agenda - Updated

Theoretical Foundations for Accuracy Control in Additive Manufacturing (FACAM 2016)

USC

January 18th, 2016

Synopsis: Additive Manufacturing (AM) fundamentally changed the premise of quality engineering. This workshop aims to discuss the research needed to establish the theoretical foundations for accuracy control in AM.

Participants

Dr. Qiang Huang, Associate Professor

Gordon S. Marshall Early Career Chair in Engineering

Epstein Department of Industrial and Systems Engineering, USC

Dr. Yong Chen, AssociateProfessor

Epstein Department of Industrial and Systems Engineering, USC

Dr. Fugee Tsung, Professor, Head

Dept. of Industrial Eng & Logistics Mgt 

HKUniv. of Science & Technology

Dr. Arman Sabbaghi, Assistant Professor

Department of Statistics

Purdue University

Dr. Matthew Plumlee, Assistant Professor

Dept of Industrial & Operations Engineering

University of Michigan

Dr. Zhijian Pei, Program Director

Manufacturing Machines and Equipment (MME) Program Engineering/CMMI

National Science Foundation

Dr. Hui Wang, Assistant Professor

Industrial & Manufacturing Engineering

Florida State University

Dr. Jun Zeng

Senior Scientist / Engineer at HP Labs

PhD Student Participants:

USC: He Luan, Yuan Jin, Yanqing Duanmu, Ahmed Alzanki 

HKUST: Andi Wang

Harvard: Sobambo Sosina

Purdue: Megan Endress, Raquel De Souza Borges Ferreira

Workshop Agenda

Major events

• Jan 18th, 2016:   FACAM 2016 Workshop (8:30am - 5:00pm)

   Refreshment in the morning and box lunch will be provided

   Morning session is open to public for published work and afternoon session is closed to collaborators for on-going research.

• Jan 18th, 2016: Welcome dinner at LA Live (6:00pm)

• Jan 19th, 2016: Dr. Zhijian Pei will give a NSF talk at GER 206 10-11:00 a.m.

Workshop Schedule on Jan 18th, 2016

Morning Session @ GER 206 (8:30-12:00pm) 

• Dr. Qiang Huang: Foundation of Optimal Compensation of Three-dimensional Shape Deformation in Additive Manufacturing
• Dr. Yong Chen: High-Precision Additive Manufacturing Processes – Design and Processing Perspective
• Dr. Arman Sabbaghi: Model Building from Small Samples of Disparate Data in 3D Printing
• Dr. Matthew Plumlee: Uncertainty quantification for computational models of physical processes
• Dr. Hui Wang: Effect Equivalence Transformation and Transfer Learning in Engineering Systems
• Dr. Jun Zeng (HP): Fulfillment fidelity challenges in 3D printing and computing and computational research opportunities 

Afternoon Session @ GER 206 (1:00-5pm):

• Dr. Matthew Plumlee: Model Uncertainty Quantification of 3D Printing Process
• Sobambo Sosina: Missing Data Treatment of Shape Deformation Modeling
• Dr. Arman Sabbaghi and Dr. Qiang Huang:  Transferability of Causal Mechanisms via Equivalent Effects of Lurking Variables
• Andi & Fugee:  Predictive Modeling and Compensation of Shape Deviations for Fused Deposition Modeling Processes under Different Printing Parameters.
• Yuan Jin, He Yuan: On-going research on 3D shape deformation modeling and control

Remaining Issues in "An Analytical Foundation for Optimal Compensation of Three-Dimensional Shape Deformation in Additive Manufacturing"

Optimal Compensation Policy to Reduce Maximum Shape Deformation

      Optimal compensation policies established in "An Analytical Foundation for Optimal Compensation of Three-Dimensional Shape Deformation in Additive Manufacturing" satisfy the Minimum Area Deviation (MAD) Criterion or the Minimum Volume Deviation (MAD) Criterion respectively. The polices will reduce the average shape deformation for 2D or 3D shapes. 
     If maximum shape deformation of a product should be less than a pre-specified value, a MinMax criterion will be more appropriate. However, currently no MinMax optimal compensation policy has been established for 3D shape deformation.
    A more complicated case is the combination of MAD and MinMax criteria, which is more realistic for practical case.  More research is needed to address the remaining issues left in "An Analytical Foundation for Optimal Compensation of Three-Dimensional Shape Deformation in Additive Manufacturing" (ASME-JMSE, in press).

-QH on 12/06/2015

Journal of Quality Technology: Special Issue on “Quality Engineering in Advanced Manufacturing”

Journal of Quality Technology

Special Issue on “Quality Engineering in Advanced Manufacturing”

Emergence of advanced manufacturing technologies along with significant progress in computing and communication has fundamentally transformed manufacturing to a smart operation scheme that facilitates better management and control of quality and productivity under a broadly connected and accessible environment.  Future manufacturing has increasingly been characterized as complex, networked cyber-physical systems.

At the product level, complex shapes can be now efficiently manufactured and product customization allows for short-run or even one-of-a-kind manufacturing.

At the process level, smart sensing allows for high-frequency, multi-stream data collection, mixing signal data to (usually contactless) in-line inspection. At the design level, large-scale computer experiments can be used to simulate performances of complex products and processes.

With this dramatic transformation of manufacturing environment, Quality Engineering, which has a strong connection with traditional mass production scheme, faces fundamental challenges. On the other hand, the “data-rich”, dynamic, smart manufacturing environment also offers tremendous opportunities to develop new-generation methodologies and techniques to manage and improve quality.

In an attempt to disseminate cutting-edge developments in quality engineering resulting from the paradigm shift in manufacturing, the Journal of Quality Technology seeks submissions for a special issue on “Quality engineering/techniques for advanced manufacturing” on the following topics:

• ·       Smart sensing scheme for efficient data collection
• ·       Non-contact data modeling, monitoring and control
• ·       Pervasive monitoring and control
• ·       Data analytics for information capture, representation, extraction, and utilization
• ·       High fidelity real-time and predictive modeling and simulation of advanced manufacturing processes (e.g., computer experiments design, modeling, calibration and optimization)
• ·       Robust mathematical models to simulate advanced manufacturing processes and enable complex control algorithms.
• ·       Design of experiments and statistical quality monitoring for low volume and large product varieties
• ·       Uncertainty quantification of manufacturing processes using computer experiments
• ·       Designing efficient experiments to augment information from high-resolution and low-resolution models
• ·       Complex shapes modeling, monitoring, inspection and optimization
• ·       Data fusion for quality engineering

Paper Submission

All papers must be prepared in accordance with the Journal of Quality Technology standards and guidelines, and will be reviewed following the regular review process of the journal. Submission of a manuscript implies that the work has not been published before, and is not currently under consideration for publication elsewhere.

Authors are required to prepare a cover letter indicating the submission is for the special issue and to submit the paper in PDF or MS Word via email to: Valerie Funk, Manuscript Coordinator, ASQ (Email: manuscripts@asq.org).

Editorial Team:

·         Bianca Colosimo (Co-guest Editor), Politecnico di Milano, Email: biancamaria.colosimo@polimi.it

·       Tirthankar Dasgupta (Co-guest Editor), Harvard University, Email: dasgupta@stat.harvard.edu

·       Qiang Huang (Co-guest Editor), USC, Email: qiang.huang@usc.edu

·       Fugee Tsung (Editor-in-Chief), Hong Kong University of Science and Technology, Email: season@ust.hk

Important Dates:

·          Aug 31, 2016: Paper submission deadline

·       July 31, 2017: Final decision on paper acceptance

·       Late 2017 or Early 2018: Publication of the special issue