negative skin friction downdrag dragload
TRANSCRIPT
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 1
NEGATIVE SKIN FRICTIONDOWNDRAGDRAGLOAD
General Framework and back analysis of a real case
Augusto Lucarelli, Derrick Blanksma, Ryan Peterson
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 2
Terminology and general framework
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 3
Is it a bearing capacity problem?
What happens when Q+QNSF > QP+QPSF ?
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 4
Applied Load Qc at the top of the pile
Drag load: the difference between the max axial force and the load at the top of thepile. It is maximum when Qc is zero and goes to zero when geotechnical capacity isreached.
0 1
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 5
So…….
Negative skin friction is a soil-structure interactionproblem.• It doesn't change the geotechnical bearing capacity;• It changes the pile stiffness and produces settlements
(downdrag);• It changes the axial load distribution along the pile shaft
(dragforce). Check structural capacity of the pile.
From a geotechnical point of view, it is a Service LimitState issue. It might be a structural capacity problem atthe neutral plan elevation although usually there is nosignificant bending moment.
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 6
The ultimate bearing capacity is around 10 MN
Let’s work out an example….Drilled Shaft, D=1.0 m; L=30m
Layer 1: soft claytlim = 25 kPa
Layer 2: medium sandtlim = 70 kPa
Layer 3: dense sandtlim = 110 kPa
Base:qblim = 5000 kPa
Ground level
-10.0
-20.0
-30.0
Settlement profile
200 mm
50 mm
-15.0
Pile element
Displacements imposed on the non-liner springs
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 7
Pile-Soil interaction…
Pile-soil interaction (along the shaft and at the base) is accounted for by means of non-linear t-z springs. The effect of negative skin friction is evaluated by imposing boundary displacements to the spring.
Base curve:displacementat full capacityfor drilledpiles 0.25-0.30D
Curve at 5 mdepth withnegative skinfriction
t-z curve at 5 m depth
tau
[kP
a]
Soil-pile relative displacements100
25
-25
Curve at 27m depthwithoutnegative skinfriction. Itgoes throughthe origin
t-z curve at 27 m depth
Soil-pile relative displacements
110
-110
Displacements at the base
qb
[kP
a]
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 8
Load curve without NSF
Load Curve with NSF
Load settlement curve at the head of the pile
NSF doesn’t effect the ultimate bearing capacity of the pile-soil system.NSF does effect the stiffness of the pile-soil system and axial loaddistribution along the shaft.
Let’s consider a Service Load of 4000 kN: without negativeskin friction the settlement would be around 5 mm. With skinfriction the settlement would be around 15 mm. If the lastvalue is not tolerable, the Service Load must be reduced.
Failure Load: 10000 kNA
xial
Lo
ad a
pp
lied
at
the
to
p [
kN]
Displacements [mm]
4000
2000
5 1540 100
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 9
Axial force distribution along the pile
Downdrag force 1800 kN
Sand SandDe
pth
[m
]
Axial Force [kN]
Qc = 0.0Qc = 2000 kN
Downdrag force 1700 kN
4000200020001000
Axial Force [kN]
0
10
20
10
20
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 10
Sand
Axial Force [kN]
Qc = 5000 kN Qc = 8000 kN
Axial Force [kN]
Sand
De
pth
[m
]
10
20 20
10
Downdrag force 1200 kN
Downdrag force 600 kN
5000 8000
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 11
Axial Force [kN] Axial Force [kN]D
ep
th [
m]
Sand20 20
1010
Qc = 10000 kN
Downdrag force 0All NSF has become PSF
Qc = 2000Qc = 5000
Qc = 8000
Qc = 0 Qc = 10000
Sand
Ne
utr
al p
lan
po
siti
on
4000 10000
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 12
BACK ANALYSIS OF A REAL CASE USING FLAC3D
Steele County Highway 7 in Owatona
Bridge 74551
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 13
Project Site
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 14
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 15
ShapeAccelArray (SAA) Profile
25
0
-25
-50
-75
-100
Deformation profile measured by the SAA after the bridge deck was placed.
SAAN
South Abut.
North Abut.
SAA
1
0
-1
-2
-3
-4
0 20 40 60 80 100 120Length [ft]
Ap
pro
ximate D
isplace
me
nt [m
m]
Dis
pla
cem
ent
[in
]
Maximum SAA Deflection = 3.6 in (91 mm)
Approximate Length [m]0 6 12 18 24 30 36
Plan Location
Elevation Location
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 16
SAA Time History
• Surcharge loads induced roughly 1.7 inches of vertical displacement.
• Surcharge removal resulted in 0.9 in of rebound.• Construction loads resulted in a little over 2 in of
vertical displacement.
SAAN
South Abut.
North Abut.
SAA
0
-1
-2
-3
-4Time (August, 2010 – June, 2014)
Ap
pro
ximate D
isplace
men
t [mm
]
Dis
pla
cem
ent
[in
]
0
-25
-50
-75
-100
Plan Location
Elevation LocationSettlements of the soil around the pile
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 17
Soil Stratigraphy
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 18
Model Setup: single pile interaction
FILL 9 m
SOIL 15 m
BEDROCK
Simplified model: only one embedded pile. The objective is to simulate the local interaction with soil considering the main construction phases.
Embedded pile
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 19
Embedded pile: lateral interaction with the soil
Yield Criteria: effective stress approach
FILL
SOIL
BEDROCK
fslim
ks
knfnlim
𝑓𝑠𝑙𝑖𝑚
ҧ𝑙= 𝜏𝑠
𝑙𝑖𝑚𝑃
𝜎𝑧𝑧′ is the effective vertical stress
𝛽∗ factor ….function of soil type, installation method…
𝑘𝑠 =𝑓𝑠𝑙𝑖𝑚
𝛿𝑠𝑙𝑖𝑚
𝜏𝑠𝑙𝑖𝑚 = 𝜎𝑧𝑧
′ 𝛽∗ 𝑘𝑃𝑎
Linear elastic beam element, EA, EJ…
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 20
𝑓𝑠𝑙𝑖𝑚
ҧ𝑙= 𝜏𝑠
𝑙𝑖𝑚𝑃
P is the perimeter of the pile
ҧ𝑙 is unit length along the pile
𝑓𝑠𝑙𝑖𝑚
ҧ𝑙
𝑓𝑠ҧ𝑙
ks
𝑘𝑠 =𝑓𝑠𝑙𝑖𝑚
𝛿𝑠𝑙𝑖𝑚
𝐸𝑙𝑎𝑠𝑡𝑖𝑐 − 𝑝𝑒𝑟𝑓𝑒𝑐𝑡𝑙𝑦 𝑝𝑙𝑎𝑠𝑡𝑖𝑐 𝑐𝑜𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑣𝑒 𝑏𝑒ℎ𝑎𝑣𝑖𝑜𝑟…𝑓𝑟𝑜 𝑛𝑜𝑤
𝑆ℎ𝑒𝑎𝑟 𝑆𝑡𝑖𝑓𝑓𝑛𝑒𝑠𝑠
Shear Response along the shaft of the pile
𝛿𝑠𝑙𝑖𝑚 = 1 − 5𝑚𝑚 𝛿𝑠
𝑘𝑁/ ҧ𝑙
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 21
Loading
BACKFILL
SOIL
BEDROCK
(1) Load induced by backfilling was modeled with a density “ramping” procedure of the back fill.
(2) Additional loading (Pile cap, beams, etc…) was simulated by applying an axial force directly to the pile head.
SURCHARGE
(3) Additional surcharge loading after the deck was placed, was simulated by increasing the density in the zones above the pile.
(1)
(2)
(3)
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 22
Soil Profile & Properties
BACK FILL
A) CLAY
B) SANDY CLAY
C) CLAYEY SAND
D) SANDY CLAY
E) SAND
F) SANDY CLAY
G) BEDROCK
STRATA BF A B C D E F G
Bottom elev. [m] 9.0 10.0 11.5 13.5 20.5 22.0 24.0 30.0
Young’s modulus, [MPa]
35 35 35 35 52.5 65 65 100
Poisson ratio [] 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Angle of Friction [°] 26 26 26 26 26 26 26 30
Cohesion [MPa] - - - - - - - 520
Density [g/cc] 1.8 1.7 1.7 1.7 1.8 1.8 1.8 2.5
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 23
Settlement
4 cmCalibrated with the SAA
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 24
Initial Shear Response to Loading
𝜏𝑠𝑙𝑖𝑚 = 𝜎𝑧𝑧
′ 𝛽∗
BACKFILL
SOIL
BEDROCK
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 25
Final Shear Response to Loading
𝜏𝑠𝑙𝑖𝑚 = 𝜎𝑧𝑧
′ 𝛽∗
BACKFILL
SOIL
BEDROCK
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 26
Loading Process
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 27
Load History
• Case 3: 100 MPa Base Layer
Place Fill
Load Pile
Additional Load After Deck Placed
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 28
Sensitivity analysis: Cases
• Five cases were run to simulate the response to changing base layer stiffness– Case 1: The base layer is 10 GPa representing bedrock– Case 2: The base layer is reduced to 1 GPa, representing weathered
bedrock– Case 3: The base layer is reduced to 100 MPa, representing gravel or
very soft bedrock– Case 4: The base layer is reduced to 10 MPa, representing loose
sand– Case 5: The base layer is reduced to 1 MPa, representing soft clay
• In all cases, the mean soil modulus was kept constant at 52.5 MPa
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 29
Axial Force as a function of bedrock stiffness
(Case 2)
(Case 3)
(Case 4)
(Case 5)
Mean soil modulus = 52.5 MPa
Base layer modulus varies by case
Case 3 (100 MPa base layer) correlates fairly well to the strain gage data.
(Case 1)
h = Ln/Ls = 10/15 = 0.67
FILL
SOIL
BEDROCK
Ls
Ln
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 30
Neutral plane:
relative displ.
is zero
Bottom of fill
Relative displacements
Relative displ. = soil displ. – pile displ.
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 31
Results from other real cases around the world
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 32
Results from other real cases around the world
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 33
Results from other real cases around the world
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 34
Results from other real cases around the world
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 35
Results from other real cases around the world
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 36
Neutral plan position – end bearing in clay
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 37
Neutral plan position – end bearing in sand & rockIn our case Nspt>50….but is an H pileand there is a sand layer just on top ofthe bedrock….100 Mpa is not a verystiff bedrock.
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 38
Axial Force and Neutral Plane Position
Neutral Plane
Dep
th [m
]
Case 1Axial Force [kN]
Case 2Axial Force [kN]
Case 3Axial Force [kN]
Case 4Axial Force [kN]
Case 5Axial Force [kN]
As the base layer stiffness is decreased, the neutral plane position moves from the bottom of the pile (Case 1) up towards the top (Case 5).
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 39
Sensitivity AnalysisCase 1 Case 2
Case 3
Case 4Case 5
Case 1Case 2
Case 3
Case 4
Case 5
Case 1 Case 2Case 3
Case 4
Case 5
• Relative stiffness is calculated as the ratio of the mean elastic modulus of the soil to the elastic modulus of the base layer.
• The soil modulus was kept constantat 52.5 MPa and the base layermodulus was decreased by an orderof magnitude for each case. Theinitial base layer modulus was 1,000MPa.
• With a relatively stiff base layer, downdrag forces increase, the neutral plane is near the bottom of the pile and pile displacement is minimal.
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 40
Comparison of Axial Forces and Neutral Plane Depth
Case 5
Case 4
Case 3
Case 2Case 1
Case 5
Case 4
Case 3
Case 2Case 1
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 41
Considerations• Relative stiffness between the soil and base layer
influences the amount of dragload, the axial force distribution and the position of the neutral plane.
• At relative stiffness below 0.1 (very stiff base layer), the neutral plane is at the bottom of the pile and maximum possible dragload forces are realized.
• At a relative stiffness above 10 (very soft base layer), the neutral plane is near the top of the pile and the drag load forces are minimal.
• Between a relative stiffness of 0.1 and 10, The neutral plane position and drag load forces are functions of several factors.
• This region is a transition zone where 1) the drag load increases with increasing base layer stiffness and 2) the position of the neutral plane decreases with increasing base layer stiffness.
Case 1 Case 2
Case 3
Case 4Case 5
Case 1Case 2
Case 3
Case 4
Case 5
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 42
Downdrag Force Matrix (kN)
0.00525 0.0525 0.525(Base Case)
5.25 52.5
0 630 610 500 405 365
275 538 506 390 246 192
550(Base Case)
430 385 225 40 0
825 320 263 65 0 0
RSPileLoad (kN)
RS is relative stiffness i.e., the ratio of the mean soil modulus to the base layer modulus.
Pile Load is applied directly to the top of the pile.
The base case is the scenario calibrated to the field data i.e., RS = 0.525 and Pile Load = 550 kN
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 43
Relative to Max Downdrag Force Matrix
0.00525 0.0525 0.525(Base Case)
5.25 52.5
0 1.0 1.0 1.0 1.0 1.0
275 0.85 0.83 0.78 0.61 0.53
550(Base Case)
0.68 0.63 0.45 0.01 0
825 0.51 0.43 0.13 0 0
RSPileLoad (kN)
RS is relative stiffness i.e., the ratio of the mean soil modulus to the base layer modulus.
Pile Load is applied directly to the top of the pile.
The base case is the scenario calibrated to the field data i.e., RS = 0.525 and Pile Load = 550 kN
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 44
Downdrag Force Contour
Downdrag
force [kN]
Increasing base stiffness
High
Downdrag
force
Low
Downdrag
force
Base case
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 45
Load test simulation: effect of NSF
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 46
250 kN Applied load 500 kN Applied load
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 47
750 kN Applied load 1000 kN Applied load
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 48
1250 kN Applied Load 1500 kN Applied Load
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 49
As the axial load applied atthe pile head increases, themaximum axial movestoward the pile headvanishing the effect ofnegative skin friction. All theavailable friction along thepile becomes positive.
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 50
Future Developments: pile group
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 51
Project overview
Application of FLAC3D on a non-standard foundation design
A4 Arc Bridge and its interferences with other bridgesIssues:1. Narrow piles at the perimeter2. Mutual interaction between foundation
piles of neighboring structures
Expo ViaductAbutment
A4 Arc: Plinth #2
A4 Arc: Plinth #1
Stephenson ViaductAbutment
A4 Arc: Plinth #4
Expo Viaduct Pier #1
A4 Arc: Plinths #1 and #2 A4 Arc: Plinths #3 and #4
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 52
Application of FLAC3D on a non-standard foundation design
A4 Arc, Plinth n.4: Vertical interaction between adjacent foundations
Expo Viaduct Pier #1
A4 Arc: Plinth #4
Tensional axial stress piles,subjected to negative skin friction
January 10, 2016 Best Practices in Numerical Modeling – TRB 2016 Slide 53