SEOUL NATIONAL UNIVERSITY

Transcription

1 Fashion Technology 7. Manufacturing Technology Sungmin Kim SEOUL NATIONAL UNIVERSITY Printing Process Conventional Silk-Screen Printing 2

2 Printing Process Conventional Rotary-Screen Printing 3 Digital Textile Printing Digital Textile Printing (DTP) Features Unlimited range of expression Small lot printing for sampling and production Needs pre and post treatment Slower than rotary-screen printing 4

3 Digital Textile Printing Digital Textile Printing Ten Reasons for using DTP Short run printing advantage Run lengths as low as one yard of fabric without the need for screen changes Lower water and power consumption Elimination of the substantial amount of water and electrical energy for rotary screen printing Less chemical waste Results in significantly less ink usage and waste relative to rotary screen printing Large repeat sizes Without the usual screen printing limitation in pattern repeat size Reduced production space requirements The production footprint for digital printing is a fraction of the size of screen print facility 5 Digital Textile Printing Digital Textile Printing Ten Reasons for using DTP Less printed inventory needed Reduction of the need for pre-printed inventory of fabric that may or may not be used Sampling and production done on same printer Samples of designs will exactly match the final printed material Print flexibility Small quantity printing for market test before high volume rotary screen printing Variety of creative design choices for printing Capable of photographic/continuous tone images Low capital investment Lower capital investment compared to rotary printing production 6

4 Digital Textile Printing Digital Textile Printing Application Fields Apparel Upholstery Transportation Footwear 7 Digital Textile Printing Digital Textile Printing Application Fields Rapid Prototyping 8

5 3D Printing 3D Printing Technology Stereolithography (SLA) Patented by Charles Hull, co-founder of 3D Systems, Inc in D model is converted into an STL file (up to ten layers per each millimeter) SLA machine exposes the liquid plastic and laser starts to form each layer of the item After plastic hardens the platform drops down in the tank a fraction of a millimeter Printed object is rinsed and placed in an ultraviolet oven for finish processing 9 3D Printing 3D Printing Technology Digital Light Processing (DLP) Created in 1987 by Larry Hornbeck of Texas Instruments Uses digital micro mirrors laid out on a semiconductor chip The same technology applied for movie projectors One section of object is built simultaneously The printing speed is pretty impressive Other method DLP method 10

6 3D Printing 3D Printing Technology Selective Laser Sintering (SLS) Developed by Carl Deckard and Joe Beaman (Texas University) in 1980s Doesn t need any support structure More spread among manufactures rather than 3D amateurs at home Use of high-powered lasers, which makes the printer to be very expensive 11 3D Printing 3D Printing Technology Selective Laser Melting (SLM) Developed in Fraunhofer Institute ILT in 1995 The fine metal powder is intensively fused by applying high laser energy Metal powder melts fully and forms a solid object (stainless steel, titanium, chrome, aluminum) Applied to parts with complex geometries and structures Thin walls, hidden voids or channels 12

7 3D Printing 3D Printing Technology Fused Deposition Modeling (FDM) Developed by Scott Crump, Stratasys Ltd. founder, in 1980s Slower than SLA or DLP Simple-to-use and environment-friendly Different kind of thermoplastic can be used to print parts 13 3D Printing 3D Printed Garment Advanced 3D Printing (4D Printing) Life-size Garment 14

8 3D Printing 15 3D Printing 16

9 3D Printing 3D Printed Garment 17 3D Printing 3D Printed Garment by FDM 18

10 Sewing Robot Sewbo Prototype Sewing Robot (Jonathan Zornow, 2016) Using water soluble thermoplastic to stiffen fabric to be as sturdy as cardboard Fabric manipulation using universal robot Sewn garment is put into hot water for the plastic to melt off 19 Objective Evaluation Application Fields Quality Evaluation Visual appearance Seam pucker, wrinkle, smoothness appearance, pilling, color fastness, etc. Physical property Drape coefficient, etc. Structure Analysis Extraction of fabric design parameters Technology Non-contact 3D Measurement Image Analysis Numerical Analysis Artificial Intelligence 20

11 Objective Evaluation Fabric Surface Quality Evaluation Conventional Method Comparison of specimen with standard replica by human experts Newly Developed Method Objective evaluation of specimen with 2D/3D measurement Wrinkle Replicas Smoothness Replicas Seam Smoothness Replicas 21 Objective Evaluation Fabric Surface Quality Evaluation Non-Contact 3D Measurement Laser Scanning Stereovision Object Shift 22

12 Objective Evaluation Fabric Surface Quality Evaluation Fractal Dimension Index of Shape Complexity λ : 2.5 N( λ) : N ( λ) ln N λ D = ( ) L = D ln λ + D ln L ( λ) = D ln λ + C 10 λ : 5 N( λ) : 14 ln N(λ) 8 6 λ : 10 N( λ) : y = x R 2 = lnλ 23 Objective Evaluation Fabric Surface Quality Evaluation Suggestion of New Evaluation Criteria Linear relationship between grade and actual visual appearance y = x r 2 = Fractal Dimension New linear evaluation criteria 24

13 Objective Evaluation Image Analysis Image acquisition Digital camera Image processing Grayscale conversion Binary conversion Topology modification Opening, closing, thinning, skeletonization, etc. Filtering Average, median, Gaussian, edge detection, etc. Frequency domain analysis Fast Fourier transformation (FFT) Quantitative analysis Counting, measurement 25 Objective Evaluation Image Analysis Fabric Pilling Evaluation ASTM 3512 photographic pilling standards FFT, Gray Scaling, Binarization, Counting Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 26

14 Objective Evaluation Image Analysis Fabric Structure Analysis Extraction of Design Parameters Warp/weft density Fabric cover factor Fabric thickness 27 Objective Evaluation Image Analysis Fabric Structure Analysis Extraction of Design Parameters Evaluation of crimp Orthogonality Defect detection 28

15 Objective Evaluation Image Analysis Yarn Crimp Evaluation of yarn crimp under zero-tension condition 29 Objective Evaluation Image Analysis Fabric Drape Coefficient Quantitative analysis of drape phenomenon 30

16 Objective Evaluation Physical Measurement Garment Fit Strain/pressure simulation and measurement 31 Objective Evaluation Physical Measurement Evaluation of Human Factors Interaction between Human body and garment Garment and environment Physiological analysis Garment function Garment comfort Safety factors Elimination of questionnaire 32

17 Artificial Intelligence Negative View? Evil Self-Conscious A. I. in Movies HAL 9000 (2001 Space Odyssey, 1968) Sky Net (Terminator, 1984) Ava (Ex Machina, 2015) 33 Artificial Intelligence Current Status Problems Easy for Computer Hard for human Simple-repetitive-precise calculation Huge data analysis Problems Hard for Computer Easy for human Voice, image, text recognition 34

18 Artificial Intelligence Killer Robots? DARPA Robotics Challenge Most robots could not cross a few hurdles made of brick pieces Just climbing a ladder or opening a door were difficult The key to success is the seamless interaction with the real world 35 Solution of NP-Hard Problems Nondeterministic Polynomial Time Problem Artificial Intelligence 3!=6 4!= !=

19 Artificial Intelligence Optimization Definition Determination of optimum set of independent variables y = f ( x, x2, x3,..., x 1 n ) Examples Determination of the area, height, and cost of building construction Consideration of the price, function, and after-service of goods Determination of the optimal route for overseas travel planning Determination of an index from a number of variables without any visible relationship 37 Artificial Intelligence Traditional Solution Numerical Analysis Simulated Annealing Conjugated Gradient NFL (No free lunch) Theorem! 38

20 Artificial Intelligence A. I. Solution Artificial Neural Network Imitation of neuron-synapse system Error back propagation network 39 Artificial Intelligence A. I. Solution Genetic Algorithm Calculation based on Natural Evolution Process Simulated evolution Similar terms are used : Population, Generation, Fitness, Selection, Cross-Over, Mutation % % % % Initial Population Fitness Evaluation Selection Cross-Over Mutation 40

21 Artificial Intelligence A. I. Solution Fuzzy Inference Logic for the Description of Vague Concepts Meaningless and ambiguous properties such as person s appearance etc. Level of Middle Agedness Application Fields Control Subway Robot arm Furnace temperature Social Science Fuzzy questionnaire Information Science Fuzzy database Age Artificial Intelligence Applications in Fashion Technology Material Selection Fusible Interlining Input» Physical properties of fabric and adhesive» Fusing condition (temperature, pressure) Output» Optimum interlining Neural Network 42

22 Artificial Intelligence Applications in Fashion Technology Categorization Multivariate nonlinear regression Body type Insole type 43 Artificial Intelligence Applications in Fashion Technology Pattern Nesting Classic NP-Hard problem 44

23 Research Topics Improvement of Unstructured Data Processing Method Why do I think so? Understanding the mechanism of human thinking Image or text recognition is particularly important in fashion sectors Artificial Intelligence 45