Development of slip partitioning within wet kaolin and dry sand oblique-convergence experiments

Kevin Toeneboehn1, Michele Cooke1, Karen Leever2

1University of Massachusetts - Amherst2Helmholtz Center Potsdam GFZ German Research Center for Geosciences

Slip partitioned systems can accommodate oblique convergence with different slip rake on two or more faults and are well documented; however, the evolution of slip partitioned crustal systems is unconstrained. Carefully scaled physical experiments in crustal analogs inform our understanding of fault evolution because we can control the loading and directly observe the ensuing deformation. Using experiments in both dry sand and wet kaolin clay we explore how slip partitioning evolves under different convergence angles. Previous cohesionless dry sand experiments documented that convergence angle and fault strength control slip partitioning. In contrast to dry sand, the non-zero cohesion

of wet kaolin produces long-lived fault structures that easily reactivate. Scaled experiments using both dry sand and wet kaolin with identical boundary conditions provide insights on the role of material properties in slip partitioning. We use motorized movement of rigid blocks overlaid by a layer of dry sand or wet kaolin clay to approximate crustal deformation due to oblique subduction zone convergence. Digital image correlation combined with stereovision techniques provides evolution of horizontal strain and uplift that constrain fault geometry and slip vectors along the faults. Additionally, force gauges record the evolution of fault-normal forces throughout slip partitioning, such

as stress drops associated with fault growth. Within the dry sand, an oblique-slip forethrust-backthrust pair forms first followed by a late stage through-going strike-slip fault. In contrast, at shallow convergence angles within the clay, the strike-slip fault forms first and a backthrust never develops. As convergence accumulates in the clay, a forethrust forms dipping toward the underlying discontinuity. The lack of cohesion in dry sand may prevent the concentration of mode III stresses that leads to the early vertical strike-slip fault growth observed in the clay. The cohesion of the clay, which is similar to crustal rock strength, may facilitate the maintenance of slip partitioned fault systems.

Abstract

GSA 2016 T220. Physical and Numerical Modeling of Geologic Processes, Paper No. 348-4

Slip Partitioning in the Crust

Experimental Design

Discussion/ Conclusions

Timeline of Slip Partitioning Evolution

Dry Sand Wet Kaolin

Clay/sandPlastic

Motors

Stationary Plate

Driving Plate

ROI

50 cm

2.5 cm

1.5 2 2.560

80

100

Shear Strain

Shea

r Str

ess

(Pa)

62%66%

fast fastslowstrain rate slow

Burger’s material

Maxwell

Kelvin

Three plastic blocks with abutting 30° contacts drive a 2.5 cm overly-ing layer of dry sand (320 µm) or wet kaolin. The blocks are positioned on two plates: one fixed and the other moved along the x- and y-axis by two stepper motors.

A calibrated stereo camera system mounted above the model captures high-resolution images of the region of interest (ROI). A force sensor embedded in the clay, or attached to the drive plate for sand models, records the variations in fault normal stresses throughout the experi-ments.

LENGTH SCALING3,4

Dry Sand: 1 cm = 5 km crustWet Kaolin: 1 cm = 0.75 - 1.4 km

Acknowledgements3Leever, K. A., Gabrielsen, R. H., Sokoutis, D. and Willingshofer E., 2011. The effect of convergence angle on the kinematic evolution of

strain partitioning in transpressional brittle wedges: Insight from analog modeling and high-resolution digital image analysis. Tectonics, 30, 1–25.

4Hatem, A. E., Cooke, M. L., Madden, E. H. (2015). Evolving efficiency of restraining bends within wet kaolin analog experiments. J. of Geophys. Res. Solid Earth, 120.

5Cooke & Van der Elst, 2010, etc. Cooke, Michele L and Nicholas J. van der Elst, 2012. Rheologic testing of wet kaolin reveals frictional and bi-viscous behavior typical of crustal materials, Geophysical Research Letters, vol. 39, L01308.

This work was made possible by a GSA student grant and funding from NSF and the UMass Geosciences Department. We would also like to thank John Sweeney and Brian Toeneboehn for their guidance and assistance in constructing the models and the staff at the Helmholtz Laboratory for Tectonic Modelling for their assistance in performing the sand experiments.

We measure the clay’s shear strength by fall cone method and adjust to 90-115 Pa by varying the water content (70% by weight).

Wet Kaolin clay deforms as a bi-viscous Burger’s material ex-hibiting rate and state behvior at failure5 (below).

Slip partitioning can occur at multiple scales within the crust and ranges from local convergence within restraining bends (e.g., the Transverse Ranges of the southern San Andreas Fault) to thousands of kilometers across along subduction zones such as the Great Sumatra Fault (right) or median tectonic line in Japan (e.g., Bowman et al., 2003; Jones & Wesnousky, 1992; Fitch, 1972).

At the subduction zone scale, slip partitioning occurs along two margin parallel faults with a characteristic geometry: a dipping oblique slip fault along the trench and a continental vertical strike-slip fault (Fitch, 1972).

Due to their crustal scale, these faults are capable of producing some of the largest magnitude earthquakes in

recorded history including the devastating Sumatra-Andaman earthquake of 2004 and the more recent 2011 Tohoku Oki earthquake in Japan.

Above, a three-dimensional schematic of the Sumatra subduction zone depicts the oblique-thrust fault along the Sunda trench where the Australian and Indian plates subduct beneath the Sunda plate. Map view shows the parallel trending strike-slip fault bisecting Sumatra. Modified from Nature 435,756-757 (9 June 2005).

Curl at 22 mm

Curl at 109 mm

Shear Strain (rad/min)

CCCCuuuuurrrrlrlll aaaaattttt 222222222 mmmmmmmmCCCuCuuuurrrrrllll aaaaattttt 222222222 mmmmmmme displacement = 22 pppppPPPPPlalaaaatatttteteeee e dddddidiisisssspsppppplalaaaaaccceceeeeemmmmmmmemeeeenennnnnttt t 22222222 mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteteee mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteteee mmmmmmmmmmmmmmmmPPPPPPll ttttt

2222222p

CCCCuuuuurrrrlrlll aaaaattttt 11110000999 mmmmmmmmmCCCuCuuuurrrrrllll aaaaattttt 11110009999 mmmmmmmmmpppppPPPPlPllalaaaatattteeee dddddidiisisssppppplpllalaaaacacccceceeeememmmmmmmeeeeenennnntntttt t 11110000009999 mmmmmmmmmmmmmmmmPPPPPlPllalaaaatattteeee dddddidiisisssppppplpllalaaaacacccceceeeememmmmmmmeeeeenennnntnttttt ====== 11110000009999 mmmmmmmmmmmmmmmmPPPPPlPllalaaaatattteeee dddddidiisisssppppplpllalaaaacacccceceeeememmmmmmmeeeeenennnntnttttt ====== 11110000009999 mmmmmmmmmmmmmmmmPPPPPlll tttt dddddiii lll ttttt 11110000009999

lll 1111000999p

05° Convergence Angle 0 +0.1

4 cm

S S

F

S S

Curl at 20 mm

Curl at 81 mm

e displacement = 20 pppppPPPPPllaaaattteteee mmmmmmmmmmmmmmmmPPPPPPll mmmmmmmmmmmmmmmmPPPPPPll tttt mmmmmmmmmmmmmmmPPPPPlCCCCCCuCuuuuururrrrrlCCCCCuCCuCuuuuuuruururrrrrrlrlrlrlCCCCCuCuCuuuuuuururururrrrrrlrllllate displacement = 20 mmlate m

e displacement = 81PPPPPlalaaaatatttteteeee e dddddidiisisssspsppppplalaaaaacccceceeeeemmmmmmmemeeeenennnnnttt t 8888881111 mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteeeee dddddidiisisssspppppplalaaaacacccceceeeeemmmmmmmemeeeenennnnntttt ====== 8888881111 mmmmmmmmmmmmmmmmPPPPPPll mmmmmmmemeeeenennnnn mmmmmmmmmmmmmmmmPPPPPPll mmmmmmmemeeeenennnnn mmmmmmmmmmmmmmml at 81l at 81CCCCCCCuCuuuuuuuururrrrrrlrlllll aaaaaaaaatatttttttt 8888881111111 1 mmmmmmmmmmmmmmCCCCCuCuCuCuuuuuuruururrrrrrlrlrlrl 1111 mmmmmmmmmmmmmmmmCCCCCuCuuuuruurrrrll 1 mmmmmmmate displa ement 81 mmt di l

10° Convergence Angle Shear Strain (rad/min)0 +0.1

S S

F

S S

1. Dry sand: an oblique-slip forethrust-backthrust pair forms first followed by a late stage through-going strike-slip fault

2. Wet Kaolin clay: the strike-slip fault forms first and a backthrust never develops, a forethrust forms dipping toward the underlying discontinuity

3. The lack of cohesion in dry sand may prevent the concentration of mode III stresses that leads to the early vertical strike-slip fault growth observed in the clay.

4. The cohesion of the clay, which is similar to crustal rock strength, may facilitate the maintenance of slip partitioned fault systems.

5. Pre-existing crustal weaknesses, such as from back-arc volcanism along a subduction zone, play an important role in the timing of slip partitioning.

6. Due to the high obliquity of the tested convergence angles, these experiments however, likely more appropriately model smaller scale transpressional systems such as the West Spitsbergen fold-and-thrust belt or Transverse Ranges of the San Andreas Fault system rather than larger scale subduction zones.

PPPPPPPPPPPPPPPPPPPPPlatePlate

05° Convergence Angle Shear Strain (rad/min)0 +0.1

F

F

S S

Plate displacement = 18 mmmmmmmmmmmPPPPPPPPPPPPPPPPPPPPPlate mmPlat 8 mm

Plate displacement = 65 mmPPPPPPPPPPPPPPPPPPPPPPlatePlat

15° Convergence Angle Shear Strain (rad/min)0 +0.1

F

S S

B

Uplift From Stereovision

Measuring Stress Changes with Pressure Transducers

Stereovision records the 3D topography throughout the experiments providing the incremental uplift. Uplift maps show where vertical strain accumulates between faults in addition to providing uplift rates across the faults, which can be converted to dip slip rates with knowledge of fault dip. Insets of incre-mental uplift show how the pattern of uplift rate between the strike-slip and oblique-thrust faults varies as slip partitioning begins to subside.

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

uplift (mm)

-2

-1

0

1

2

3

4

5

6

uplift (mm)

-2

-1

0

1

2

3

4

5

6

Plate displacement = 75 mm

Plate displacement = 80 mm

Plate displacement = 115 mm

Plate displacement = 120 mm

uplift (mm)

-2

-1

0

1

2

3

4

5

6

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elev

atio

n (m

m)

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elev

atio

n (m

m)

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elev

atio

n (m

m)

elevation (mm)

0

2

4

6

8

10

12

14

16

18

20

22

24

elev

atio

n (m

m)

uplift (mm)

-2

-1

0

1

2

3

4

5

6

Uplift from 75 to 80 mm Uplift from 115 to 120 mm

10° Clay Experiment 10° Clay Experiment

10° Clay Experiment 10° Clay Experiment

∆ uplift

-2

6

-2

6

uplif

t (m

m)

uplif

t (m

m)

∆ uplift

Dry Sand 15° Convergence Angle Wet Kaolin Clay 05° Convergence Angle

Oblique-slip Fault

Shear strain at 25 mm

Shear strain at 85 mm Shear strain at 100 mm

Shear strain at 25 mm

��

��

���

���

���

���

���

���

���� ����� ����� ����� ����� ������ ������ ������ ������

For

ce (

g)

Displacement (mm)

Wet Kaolin Clay 05° Convergence Angle Oblique-slip fault forms

A

C

B

D

E

A Tekscan® FlexiForce A301 (0-1 lb) force sensor and Tekscan® ELF handle record pressure changes in the clay at 8 Hz. The force sensors are oriented with the sensing area parallel to the vertical strike-slip fault plane to maximize the signal from fault normal stress changes as faults initiate. Results from the 05° clay experiment (below) show a distinct increase in force im-mediately preceding the formation of the oblique-slip fore-thrust suggesting an “un-clamping” effect facili-tates strike-slip faulting.

Plate Displacement (mm)

0 10 20 30 40 50 60 70 80 90 100 120

15° Clay

05° Sand

08° Sand

15° Sand

110

05° Clay

15° Clay

F

F

F

B

B

08° Clay

Slip Partitioning

10° Clay

PRE-

CU

T

Slip Partitioning

Slip Partitioning

Slip Partitioning

Slip Partitioning

Slip Partitioning

F

F

F

S S

S S

S S

S S

S S

S S

S S

F

Landsliding

Landsliding

LEGEND = Strike-slip fault forms (S S) = Oblique-thrust fault forms F = Forethrust B = Backthrust = Strike-slip = Oblique-slip

Summary of fault timing reveals how convergence angle, analog material, and the presence of a pre-existing weakness (pre-cutting a vertical fault) affect the development and maintenance of slip partitioning. The wet kaolin experiments form a strike-slip fault (red arrowheads) early before the oblique-thrust (blue arrowheads). In contrast, the sand experiments form an oblique-slip forethrust (F) backthrust

(B) pair before a through-going strike-slip fault facilitates slip partitioning. Only the pre-cut clay 15° develops slip partitioining, the other 15° experiments only form a single oblique slip fault. The onset of slip partitioning in the clay is later for lower convergence angles (opposite for sand) and earlier in pre-cut clay experiments than uncut.

ate displacement = 85 mmmmmPPPPPPPlPl mmPPPPPPPlPl mPPPPPPPllPPPPPPPlPlPPPPPPPllPlate cement = 85 Plate cement = 85 mm

08° Convergence Angle Shear Strain (rad/min)0 +0.1

F

F

S S

B

B

10 cm

10 cm

10 cm

4 cm4 cm

Incremental shear strain maps overlaid by horizontal displacements for experi-ments in dry sand and wet kaolin. Shear strain is calcu-lated using the curl of dis-placement. A critical shear value for faulting is deter-mined emperically and used to create a fault map. Over-laying the displacement vec-tors from PIV on top of the fault maps reveals the slip style of faulting in the evolv-ing slip partitioned system.

Note early slip partitioning (85 mm) in the sand exper-iments along active oblique fore- and back-thrusts (blue arrows) and through-going strike-slip fault. Slip parti-tioning develops later in the 05° clay experiment (109 mm) as an oblique fore-thrust (F) and strike-slip fault (red arrows-SS).

Slip partitioning fails to develop in either 15° exper-iment over the tested range of total strain. At higher total strain (>65 mm plate dis-placement) the sand experi-ments begin to develop un-stable slopes and landsliding obscures detailed analysis of fault activity (e.g. noisey vectors in the 15° sand fault-map on bottom left).

displacement = 24 PPPPPllaaaattteteeee e ddddddiissspspppppllaaaacccceeeemmmmmmmemeeeeennnnttt t 22244444 4 mmmmmmmmmmmmmmmmPPPPPPllaaaattteteee mmmmmmmmmmmmmmmmPPPPPPllaaaattteteee mmmmmmmmmmmmmmmmPPPPPPllaaaattteteee mmmmmmmmmmmmmmmpate displacement 24 mm

displacement = 109PPPPPllaaaatatttteteeee e ddddddiisisssspspppppllaaaaacccceeeeemmmmmmmemeeeeennnnnttt t 1111000099999 9 mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteteeeee dddddidiisisssspsppppplalaaaaaccceceeeeemmmmmmmemeeeenennnnntttt ====== 11110000999999 mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteteee aaccceceeeeemmmmmmmemeeeenennnnn 9999 mmmmmmmmmmmmmmmmPPPPPPlalaaaatatttteteee aaccceceeeeemmmmmmmemeeeenennnnn 9999 mmmmmmmmmmmmmmmlate displacement 109 mate displacement 109 m

15° Convergence AngleShear Strain (rad/min)0 +0.1

F

F