Scribd
Upload
44 views15 pages

SEM 2010 (Presentation)

This document describes a method for determining isoclinic maps from complex photoelastic fringe patterns using phase-stepping photoelasticity. The method uses a regularization algorithm and cost function to resolve ambiguities in the isoclinic fringe maps. It is demonstrated on experimental data from a fatigue crack tip, producing high-quality isoclinic and isochromatic fringe maps even in dense, complex fringe regions near the crack tip. The algorithm provides a robust solution using only a small number of digital images, improving on existing techniques that often mask dense fringe areas or require many more images.

Uploaded by

Hassan Ali
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
44 views15 pages

SEM 2010 (Presentation)

This document describes a method for determining isoclinic maps from complex photoelastic fringe patterns using phase-stepping photoelasticity. The method uses a regularization algorithm and cost function to resolve ambiguities in the isoclinic fringe maps. It is demonstrated on experimental data from a fatigue crack tip, producing high-quality isoclinic and isochromatic fringe maps even in dense, complex fringe regions near the crack tip. The algorithm provides a robust solution using only a small number of digital images, improving on existing techniques that often mask dense fringe areas or require many more images.

Uploaded by

Hassan Ali
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

Determination of isoclinic map for complex

photoelastic fringe patterns


Philip Siegmann, Universidad de Alcal, Spain
Chiara Colombo, Politecnico di Milano, Italy
Francisco Daz, Universidad de Jan, Spain
Eann Patterson, Michigan State University

Outline
Introduction to photoelastic stress analysis
Basics of phase-stepping photoelasticity
Ambiguity of isochromatic fringe maps
Regularization algorithm and cost function
Fatigue crack propagation exemplar
Conclusions

Introduction
All transparent materials subject to stress exhibit temporary
birefringence
 Isochromatic fringes related to stress magnitude
 Isoclinic fringes related to direction of principal stresses

Variety of digital fringe processing


 RGB photoelasticity
 No user-input for calibration of fringe map
 Limited to isochromatic fringe orders less than 3

 Fourier transforms
 High-quality isochromatic and isoclinic data
 Requires tens of images
 Fringe unwrapping and user-calibration

 Phase-stepping
 4 to 6 images for isochromatic & isoclinic data
 Fringe unwrapping and user-calibration

Mathematical description of polariscope


/4

polariser
1/4 wave plate

specimen
retardation,

x
y

1/4 wave plate


analyser
camera

In a circular polariscope, light intensity at a point (x,y):

i(x,y) = im+ iv [sin2().cos - sin().cos().sin]


Wang, Z.F., Patteson, E.A., Use of phase-stepping with demodulation and fuzzy sets for
birefringence measurement, Optics & Laser Engineering, 21:91-104, 1995.

Approach to phase-stepping solution


In a circular polariscope, light intensity at a point (x,y):

i(x,y) = im+ iv [sin2().cos - sin().cos().sin]


where

= 2N = 2 (1-2)t/f

Unknowns:

Variables

, relative retardation;

, output /4 plate angle;

, isoclinic angle

, analyser angle,

im stray light term;


iv constant of
proportionality.
Wang, Z.F., Patteson, E.A., Use of phase-stepping with demodulation and fuzzy sets for
birefringence measurement, Optics & Laser Engineering, 21:91-104, 1995.

Fringe processing

Photoelastic
model in a
polariscope

Isoclinic & fringe


order maps

Digital images collected with phase-steps


i1 = im + iv cos
i2 = im iv cos
i3 = im iv sin sin 2
i4 = im + iv sin cos 2
i5 = im + iv sin sin 2
i6 = im iv sin cos 2

Periodic isoclinics & isochromatics calculated


i i

= arctan 5 3
2
i4 i6

i4 i6

(i1 i2 ) cos 2

= arctan

Wang, Z.F., Patteson, E.A., Use of phase-stepping with demodulation and fuzzy sets for
birefringence measurement, Optics & Laser Engineering, 21:91-104, 1995.

Ambiguity
Small modification to algorithm
 Use 4-quadrant arctangent function in MatLab
 Accounts for sign to yield arctangent between and -
( I 5 I 3 ) sin 2 a + ( I 4 I 6 ) cos 2 a
I I
1
= arctan 5 3
= arctan
a

I4 I6

w, a

I1 I 2

Q
-w,a

+w,a
Q

Regularization
Selection of +w,a or -w,a at each pixel using a regularization
algorithm
 Start at an arbitrary pixel
 At following pixel selection depends on w calculated for neighboring
pixels
 If - < w,a < 0 or > w,a > 0 then w
must be continuous relative to its neighbors
Q
 If w,a= 0 or w,a= then gradient of w
must be continuous relative to its neighbors

Decision implemented using a cost function


10
U (i; ) = Q(i ) A (i ) +
B (i )
Q (i )
where is the window of pixels at i(x,y) being considered
is a regularization parameter
Siegmann, P., Diaz-Garrido, F.A., Patterson, E.A., Robust approach to regularize an isochromatic
fringe map, Applied Optics, 48(22):E24-E34, 2009

Application of regularisation
w
+w,a

-w,a

fringe order, N

theory
experiment

Crack tip fringe patterns

i2

i3
i1
i4

i5

i6

Crack tip data: pre-regularization


+w,a

Graphic User Interface

Crack tip fringe orders

Plastic shielding investigation

10N

55N

100N

Constant amplitude fatigue cycle

145N

190N

110% overload fatigue cycle

Colombo, C., James, M.N., Patterson, E.A., Vergani, L., On crack tip shielding due to plasticityinduced closure during an overload, FFEMS, in press

Conclusions
Robust and fast algorithm demonstrated on troublesome fringe pattern
 Extensive interaction between isochromatic & isoclinic parameters
 Density and complex of fringes around the crack tip

Better than existing algorithms using a small number of images


 Previously tended to mask such areas which is not appropriate when data
needed close to crack tip
 More images not viable when analysis a propagating fatigue crack

Solution combines robust algorithms of Quiroga & Gonzlex-Cano1 and


fast algorithms of Siegmann et al2
 Novel application of cost function using quality map
 Elimination of the minimization of the cost function at every pixel
1. Quiroga, J.A., Gonzalex-Cano, A., Separation of isoclinics and isochromatics from photoelastic
data with a regularized phase-tracking technique, Applied Optics, 39:2931-2940, 2000.
2. Siegmann, P., Backman, D., Patterson, E.A. , A robust approach to demodulating and
unwrapping phase-stepped data, Experimental Mechanics, 45(3):278-289, 2005.

You might also like