Full-field kinematic measurements by Digital Image

3D Tomographic reconstruction Feldkamp algorithm for cone beam geometry (whole sample) or FBP for fan beam assumption (central layer) Same slice of re...

0 downloads 115 Views 3MB Size
Department of Civil and Environmental Engineering (DICA), Politecnico di Milano, June 8th, 2015

Full-field kinematic measurements by Digital Image Correlation for mAterial characterization at different scales

Roberto Fedele

Civil and industrial applications of Digital Image Correlation (DIC) to be discussed

1

Assessment of innovative CFC/Cu joints and Finite Element Model Updating

2

Adhesive joints for aerospace engineering

3

Study of delamination tests on FRP-reinforced masonry pillars

4

X-ray microtomography of foam samples under in situ loading

1 Characterization of innovative CFC/Cu joints by full-field measurements and finite elements

Roberto Fedele*, A. Ciani, L. Galantucci, Valentina Casalegno#, Monica Ferraris#

*Dept. of Civil and Environmental Engineering (DICA) Politecnico di Milano, Milan, Italy

# Dept

of Applied Science and Technology (DISAT) Politecnico di Torino, Turin, Italy

ITER (International Thermonuclear Experimental Reactor)

Single-lap shear tests on flat-tile joined samples

glue (Araldites AV 119, Ciba-Geigy)

Macroscopic response intrinsic CFC shear strength 15-20 MPa !

 max

Pmax   23.6 [MPa] Ajoint uncertain boundary conditions due to the glue layers and compliant grips!

comparative assessment by different tests

 max  34  4

[MPa]

2D-surface Digital Image Correlation “passive” advection of the local texture (optical flow conservation) displacement vector

local form

f (x)  g (x  u [x])

reference image deformed image

weak form similarity measure

arg min u

  (u)    f (x)  g (x  u) dx 

variational problem

2



ui 1 = ui   ui 1

incremental form

pixel

Joint collapse

pixel

Cu CFC

Local approach: FE simulations driven by boundary displacements plane stress schematization

(anisotropic) elastic-plastic behavior finite thickness joint

Displacement fields measured by DIC and computed

Tangential traction and 95% confidence strip

Normal traction and 95% confidence strip

2 Simultaneous assessment of interface properties & boundary conditions through Digital Image Correlation Roberto Fedele*, R. Santoro, F. Hild#, S. Roux

*Dept. of Civil and Environmental Engineering (DICA) Politecnico di Milano, Milan, Italy

# Laboratoire de Mécanique et Technologie (LMT) École Normale Supérieure (ENS), Cachan (France)

Non-conventional debonding test

[ mm ] 1.4

t  19

1.3

1.2

1.1

1

0.9

0.8

y, u y 0.7

[ mm ] 0.8

x, u x

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

3 Delamination tests on FRP-reinforced masonry experiments and modelling Roberto Fedele*, G. Rosati, P. Condoleo, C. Tedeschi, M. Scaioni§, L. Barazzetti, S. Cattaneo

*Dept. of Civil and Environmental

§ Dept. of Architecture, Built Environment

Engineering (DICA) Politecnico di Milano, Milan, Italy

and Construction Engineering (ABC) Politecnico di Milano, Milan, Italy

CFRP-reinforced pillar: historical bricks (XVII century) and high strength mortar

single-lap shear test

4 X-ray tomography and motion estimation by 3D-Volume DIC

Roberto Fedele*, L. Galantucci, A. Ciani, L. Andena#, M. Bettuzzi§

*Dept. of Civil and

# Dept. of Chemistry,

Environmental Engineering Material and Chemical Engineering “G. Natta” (DICA), (CMIC) Politecnico di Milano Politecnico di Milano

§ Dept.

of Physics, X-ray Radiation Group INFN

Photon spectrum and X-ray region speed of light in vacuum

E  h 

hc



h  4.13  10 18 keV  sec

Planck-Einstein equation

wave/particle duality

p 

h



De Broglie’s linear momentum

h  x  p  2 Heisenberg’s uncertainty principle

Accurate scaling for metrology aluminum reference object

Linear measurements at the macroscale by high-accuracy gages Cube Side = (9.9984 ± 0.00163) mm

XY slice

estimated voxel size = (9.123 ± 0.0183) μm

Viscoelastic relaxation (prescribed loading) closed cell, polymer foam AIREX C70

 (t ) E 

L0  10 [mm]

L  7% L0 apparent Young modulus

1.0 [hour]

3.3 [hour]

3D Tomographic reconstruction Feldkamp algorithm for cone beam geometry (whole sample) or FBP for fan beam assumption (central layer)

Same slice of reconstructed tomos at different loading stages

3D-Volume Digital Image Correlation “passive” advection of the local texture (resulting from 3D tomographic reconstruction) displacement vector

local form reference image

f (x)  g (x  u [x]) deformed image

weak form similarity measure

 2 arg min   (u)    f (x)  g (x  u) dx u   variational problem

  

Local cross-correlation as block matching  arg max  b (u)    f (x) g (x  u) dx u b 

  

convolution via Fourier Transform

g

f b-th voxel block

f ( - x)*g  F 1[Ff Fg ]

translation of block centroids by integer displacements

x  u3

blockwise constant field: not continuous !

x  u2

x  u1

2D Debonding process of an adhesive joint 100 128

200  256

image pyramid

3

2

400  512 200  256

1

800  1024

400  512

sn

sn 0

scale

Vertical displacement along a vertical section