�ݺ�ߣ

Pro-active Management of Visual
Appearance of Products: from the
Automotive Sector to other Industries
Fco M. Martínez-Verdú
Color & Vision Group: http://web.ua.es/en/gvc
University of Alicante (Spain)
verdu@ua.es

 Visual appearance of products
 Color & Texture managed currently in the automotive sector
 Challenges for its optimal and efficient management
 Multi-scale approach (bottom-up vs. top-down)
 Foundations for pro – active Quality Management:
 Visual and instrumental correlation
 Multivariate statistics: visual psychophysics, DoE, etc.
 Conclusions
OUTLINE

 Great variety of visual attributes in daily products
VISUAL APPEARANCE OF PRODUCTS

VISUAL APPEARANCE OF PRODUCTS
Dyes & Pigments
New visual appearance
attributes
Multi-functional properties
Coloration processes
Market forces: performance – cost balance,
customer preferences, etc.
Continuous loop

 Color & Texture
 Reflection & Transmission
 Goniochromatism: BRDF
 Sparkle & Graininess
VISUAL APPEARANCE IN AUTOMOTIVE
® Wikipedia

 MSc degree in Color Technology for the Automotive Sector
VISUAL APPEARANCE IN AUTOMOTIVE

• Bottom – up:
• Many variables
• Impracticable
• Top – down:
• Feasible
• How?
CHALLENGES: MULTI – SCALE APPROACH
Color, Texture
Radiative
transfer theory
Particles
interaction
Light – Matter
interaction
particle models
Light sources tech.,
Pigments, dyes
Gloss, sparkle, etc.
Color differences
Visual appearance
Emission SPD(l)
Reflection r(l)
Transmission t(l)
Coefficients:
Absorption K
Scattering S
Substrate
Coloration
application
processes:
no. layers, etc.
Phys. + Chem.
Particles & Substrate:
Size, Shape, Thickness
Refraction index,
Extinction index,
Roughness, etc.
THEORETICALAPPROACH
EXPERIMENTALAPPROACH

• But, in this case (empirical approach = top – down), the
typical challenge is how we can understand and manage
by a pro-active way the relevance and interplay of
nano/micro (structural) parameters, and other ones
(coloration application processes, optical, etc.), on final
visual appearance attributes (color, texture, etc.).
• HOW?
• Metrology, Visual Psychophysics, and Statistics
• inter and multi-disciplinary (hybrid) approach
CHALLENGES: MULTI – SCALE APPROACH

• IDEAL CONTEXT:
• BiRD motto:
• What You See Is What
You Measure Rightly = WYSIWYMR
• ICC profile format (Graphics Arts)
• Objective: WYSIWYG
VISUAL & INSTRUMENTAL CORRELATION
Instrumental scaling
Visualassessment
?

Visual appearance of materials
DT = f(DE, DG, DS, ...) is the “GOAL”

• Human visual perception tasks:
• Detection
• Influence of viewing distance and geometry
• Spatio-chromatic dithering
• Scaling (ordering: from less to more)
• Color (from spectral data to 3 dim.),
• Sparkle (2 dim.), Graininess (1 dim.), etc.
• Color & Texture palettes
• Discrimination (differences):
• Perceptibility vs. Acceptability
• FAIL vs. PASS controls by tolerance ellipses

 Special equipment:
 Tele-spectro-radiometer
 Radiometric, photometric and colorimetric measurements
without contact, and adjusted to the target size
 Spectrofluorimeter
 Multi-angle spectrophotometers
 Lighting cabinets for visual assessments

 UA – Research Technical Services:
 XPS, WDX, FRX, SEM, FT-IR, ATR, Raman, etc.
 Pending advanced instrumentation
 multi – angle spectroscopic ellipsometry
 spectral constants of absorption (K) and scattering (S) to different measurement
geometries (irradiation / observation)
 multi-angle micro-spectrophotometer
 X-CT (tomography)
 (3D) transversal scanning of nanomaterials, etc.
 interferometric microscopy using white light
 3D surface contactless profilometer

• Current challenges in color industries:
• Gonio – appearance: color & texture
• Spectral BRDF  own color palette
• Formulation of new colors outside Rösch – McAdam solid
• Tolerances  Total Visual Appearance (color, gloss, sparkle, etc.)
• Measurement without contact (by tele – spectroradiometer, etc.)
• Reversible or irreversible electro / thermo- chromism, etc.
• Real colored products vs. its efficient digital simulation
• Color gamut of displays technologies
• Pro – active prediction models for visual quality of products

 Products: why?
 Earn money being competitive (Porter)
 by differentiation:
 and better than … , impossible to be copied, etc.
 faith perceptually digital simulation to the original
 specific colors & textures
 functional (added value from color: resistance, etc.)
 gonio – apparent
 fluorescent, thermochromic, etc.
 viewing distance effect: spatio – chromatic dithering
 near vs. far
 lighting conditions changes effect:
 type of light source (wLED, etc.)
 type of measurement geometry: diffuse vs. directional (gonio - )
MULTIVARIATE STATISTICS

 Processes: why? how? when?
 Design and production easy to be managed
 Feasibility & stability of original product model (std. or master)
 Ease for creativity & innovation
 Repeatability & accuracy of batches
 Measure to save time & money:
 Comparison with error range  TOLERANCE
 Multi – scale process: nano/micro  visual
 From bottom – up approach  top – down
 Predictive model of pro – active management by:
 Statistical design of experiments (DoE)
 Regression models
MULTIVARIATE STATISTICS
DEAUDI2000 <  2 = 1.41 OK
DEAUDI2000  [ 2 ,  3 = 1.73] cOK
DEAUDI2000 > 1.73 FAIL

• Statistical Design of Experiments (DoE)
• Statistical technique used in quality control for planning,
conducting, analyzing, and interpreting sets of experiments
aimed at making sound decisions without incurring a too high
cost or taking too much time
• Qualitative and quantitative variables  optimization objective
• Selection of the minimal number of samples
• Non-linear / linear multidimensional regression models
• Increasing sampling for an optimal prediction model
• even combining qualitative and quantitative (measureable) variables
MULTIVARIATE STATISTICS: DoE

• Problem formulation
• Aim (reproducible and measurable)
• Relevant factors (qualitative and quantitative)
• Screening design
• Selection of levels for each factor
• Experiments (no. of samples)
• Analysis of the raw data
• Data analysis (Pareto, regression, etc.)
• Optimization & Robustness studies
MULTIVARIATE STATISTICS: DoE BASICS

1 – Sparkle detection distance vs. metallic pigment size & shape
2 – Sparkle detection distance vs. concentration, achromatic
background, illuminance level & pigment type
3 – Sparkle detection distance vs. colored background
4 – Color matching vs. silver finishing process on a coated plastic
5 – Gonio-appearance of 3D printed parts vs. 3D printing technology
and its sub – processes
FIVE DoE EXAMPLES

• Relevance and interplay of colored backgrounds by CIE-L*C*abhab
• Fixed structural and environmental data (factors)
• Color mix: variable solid pigment + fixed effect pigment
• L*: 3 levels
• C*ab: 3 levels
• hab: 4 levels
SPARKLE DETECTION DISTANCE
Complete multi-level factorial table of experiments (samples)
Sample no. C L h Sample description [Hue / Lightness / Chroma]
1 0 1 1,00 RED / LIGHT / MEDIUM
2 1 -1 1,00 RED / DARK / STRONG
… … … … …
13 -1 1 -1,00 GREEN / LIGHT / WEAK
14 -1 -1 0,33 BLUE / DARK/ WEAK
… …
23 0 1 0,33 BLUE / LIGHT / MEDIUM
24 0 -1 -0,33 YELLOW / DARK / MEDIUM
… … … … …
34 1 0 1,00 RED / GRAY / STRONG
35 0 0 0,33 BLUE / GRAY / MEDIUM
36 -1 1 1,00 RED / LIGHT / WEAK

• Goal: color matching (DEab = 0), L* = 82 , & maximum transparency
• Initial DoE proposal: Taguchi L16 (215-11) Matrix, before analysis
COLOR MATCH vs. SILVER FINISHING
Worksheet MEASURED RESPONSES
Nº experim. Material
PVD
Thickness
PVD
Conc.
Topcoat
Topcoat
Robot
Basecoat
Basecoat
Robot DEab L* Transparency (T)
1 Metal A
Low
Low Low
translucent
white
Low Low Low
2
Metal B
High
High High
3
High
Low
4 Metal A
High Low Low
5 Metal C
Low
High
translucent
white
6
Metal D
Low
High High
7
High
High
8 Metal C
Low Low
Low
9 Metal A
High
Low
High
10
Metal B
High
High Low
11
High
Low
12 Metal A
High Low High
13 Metal C
Low Low
translucent
white
14
Metal D
Low
High Low
15
High
High
16 Metal C Low Low High

• Can 3D printed parts for cars (body or interior) equal or better
color & texture without losing phys – chem performance?
• DoE aims: high sparkle, flop, chroma, colorfastness, etc.
• Factors:
• Qualitative:
• Technologies: FFF or FDM, MultiJet Fusion, ColorJet, Powder-bed, living AM, etc.
• Materials: (bio)polymers, pigments, additives, process sequence, etc.
• Quantitative:
• Temperature, irradiation, speed, layer height, infill, head size, etc.
GONIO-APPEARANCE IN 3D PRINTED PARTS

• FFF experiment table (Taguchi L9): PLA fixed, simple interactions
• Head size (mm): 3 levels
• 100, 200 & 300
• Speed (mm/s): 3 levels
• 20, 40 & 60
• Infill (%): 3 levels
• 0, 20 & 100
• Color: 3 levels
• Without pigment
• Solid or special-effect pigment
Sample no. HEAD SPEED INFILL COLOR
1 1 3 2 3
2 3 2 2 1
3 1 2 3 2
4 3 1 3 3
5 3 3 1 2
6 2 1 2 2
7 2 3 3 1
8 1 1 1 1
9 2 2 1 3
Plane printed samples for measuring flop

• FFF experiment tables:
• Complete multi-level factorial:
• All previous factors with 2 levels, except color  set = 24, all possible interactions
• Multi-level factorial + D – optimal design
• Only speed with 2 levels  complete set = 54, but optimally reduced to 21
• Multi-level factorial + D – optimal design:
• All factors with 3 levels + new factor (polymer: ABS & PLA)  complete set = 162, but
optimally reduced to 21, and simple interactions well detected
• Multi-level V2 factorial + D – optimal design:
• Only speed and polymer with 2 levels  from 108 to 21, quadratic interactions

 Hybrid multi – scale approach for visual appearance of materials
applied successfully in automotive can be extended to other
industries as ceramics, coatings, cosmetics, plastics, printing, etc.
 Structural elements (pigments, etc.), advanced instrumental techniques,
visual and instrumental correlation methods, statistics (DoE, etc.), can save
time and money to implement new color & texture quality controls
successfully, etc., and even to make easy new competitive advantages for
companies.
CONCLUSIONS

�ݺ�ߣ

Pro active management of visual appearance of products

More Related Content

Pro active management of visual appearance of products