FB-MULTIPIER vs ADINA VALIDATION MODELING

1. INTRODUCTION

1.1 Purpose of FB-MultiPier Validation testing

Performing validation of structural analysis software delineates the capabilities and
limitations that a given structural analysis program possesses.

1.2 Other FEA software employed (ADINA)

The validation process was carried out by modeling given finite element models in both
FB-MultiPier and ADINA, and comparing the results of these models under various load cases.

1.3 Modeling Goal (Dynamic Validation Model)

The objective of this investigation was to develop and document a dynamic validation
model of somewhat realistic complexity and scale. Accordingly, a multiple pier model, with piers
replicating the structural characteristics of a physical bridge structure, were investigated. This
model validation was performed using linear analysis, while ignoring the extensive soil features
available in the FB-MultiPier program. Validation of the soil modeling features is presented in a
separate report.

1.4 Validation Approach

However, before a dynamic validation model of significant complexity was built, other
simpler models were created and tested in order to simplify certain anticipated structural
interactions under a given loading. More specifically, the validation process was conducted so that
certain individual structural characteristics of the models within FB-MultiPier were isolated; this
approach minimized any uncertainty regarding which structural responses were significantly
contributing to computed response quantities.
Hence, the validation process consisted of constructing and testing structural models of
increasing complexity. The separate models listed below constitute the structural models tested
during this validation process, and are discussed in detail later in this report. A single pile model
was created and tested first, followed by the creation and testing of a four-pile with pile cap model.
Next, a pier model was created and tested. Finally, a multiple-pier model was developed and tested.
The aforementioned models are discussed in order of increasing complexity below.

2. DESCRIPTION OF SPECIALTY MODELING COMPONENTS IN FB-

MULTIPIER

The following is a description of “special” provisions/elements that are automatically

implemented into FB-MultiPier models in an attempt to either more realistically transfer load from
one portion of a given structure to another, or to simulate stiffness found in physical structures that
may not be present in certain other finite element analysis software when using standard element
formulations.

1
2.1 Artificial Torsional Stiffness Introduced in Pile Cap Shell Element Corner Nodes of FB-
MultiPier
The pile caps in FB-MultiPier pier models consist of a grid of nine-node shell elements.
However, a drilling degree-of-freedom stiffness is not present in the standard shell element
formulation. Hence, shell elements FB-MutliPier are supplied with out-of-plane torsional stiffness
to the corner nodes (see Figure 1). The intent of this extra stiffness provision is based on the extreme
thickness of pile caps that typically occurs in real-world pier applications. Pile cap thicknesses can
exceed 10 ft, which is more than thick enough than that required to develop out-of-plane torsinal
stiffness in a structural member.

Figure 1. Out-of-Plane Torsional Stiffness Added to Four Corner Nodes of Each Shell
Element in Pile Cap
2.2 Quad-Pod Elements
Four beam elements extend from each FB-MultiPier pier column bottom node to four
adjacent pile cap nodes. The pier column end of these elements permits transfer of moments,
whereas the pile cap end of these elements does not. These “quad-pod” elements in FB-MultiPier
models form the sole structural connection that links the pier column to the underlying pile cap (see
Figure 2). More specifically, the bottom node of a given pier column in FB-MultiPier is not
structurally tied to the immediately underlying pile cap node. The purpose of the quad-pod elements
is to more evenly distribute pier column loads to the pile cap, or alternatively stated, prevent stress
concentrations.

2
 The exact connectivity (the exact designation of the beginning and end nodes) of the quad-
pods and the corresponding stiffness of the three member axes varies from model to model. Hence,
a user wishing to perform validation must examine the output file generated by FB-MultiPier upon
running analysis of a given model and search for “SUBSTRUCTURE MEMBER
CONNECTIVITY”. The quad pod elements are designated element numbers 1-n where n is four
times the number of pier columns. The quad-pod elements always occur in groups of four. The user
may then return to the interface and visibly verify the geometric configuration of the quad-pod
elements. Regarding the stiffness of the quad-pod elements, a user wishing to perform validation
using accurate stiffness values may search the aforementioned output file using the query
“PROPERTIES FOR CONNECTOR ELEMENTS”. This particular section of the output file
describes the assigned moments of inertia, modulus of elasticity, etc. of the quad-pod elements. In
general, the stiffness values of these elements are amplified relative to that of other ordinary
elements in typical pier structures. Finally, the mass of these elements is assigned as zero so as to
minimize dynamic effects due to mass contributions from “specialized” members, while still
providing stiffness similar to real-world pier structures.

Figure 2. Quad-Pod Elements Connectivity - Pier Columns to the Underlying Pile Caps.

2.3 Rigid Horizontal Link Elements along Pier Cap Beam

For every bearing defined in a given FB-MultiPier model, a rigid link (of beam element
type) spans from the pier cap beam centerline to the center-bottom of the bearing pad, as shown in
Figure 3. The horizontal rigid links are used to realistically place the base of the pad elements
(bridge bearing pads) within the model, with respect to the geometry found in real bridge structures.
The stiffness of the horizontal rigid links is assigned such that any load coming from the
superstructure (or described more locally, the bearing pads) will be transferred to the pier cap beam

3
with out contributing any additional displacement, and vice versa. The mass of the horizontal rigid
links is assigned as zero so as to minimize dynamic effects due to mass contributions from
specialized members, while still transferring any conceivable real-world loading to the other
portions of a given pier structure.

Figure 3. Pier Cap Beam Showing Horizontal Rigid Link Connectivity – Pier Cap Beam to Bearing
Pad Element

2.4 Bearing Pad Elements

Bearing pad elements act to transfer forces between the superstructure and substructure of
a given model, and vice versa (see Figure 4). More specifically, the bearing pad elements span
between the respective horizontal rigid link outer node and the respective, vertically aligned
transfer beam node (transfer beams are described in Section 2.5, below). Individual bearing pads
may either be fixed, released, or custom-defined with respect to any or all of the three translational
and any or all of the three rotational degrees of freedom. Additionally, because of the
aforementioned horizontal rigid link elements, up to two rows of separately configured bearings
may be placed (at realistic distances from the pier cap beam centerline) on a single pier cap beam;
it is implied here that such a structure contains three or more piers. For purposes of validation, these
elements were modeled as constraints in ADINA. For example, for a bearing set consisting of
translationally fixed and rotationally released bearings in FB-MultiPier; a set of geometrically
equivalent bearings would be created in ADINA constraining the top node of the bearing elements
to match the translations of the respective bottom node. Furthermore, because the bearings were

4
modeled using a system of constraints, no actual elements are defined between the top and bottom
bearing pad nodes in ADINA.

Figure 4. Pier Cap Beam Denoting Bearing Pad Element Connectivity – Horizontal Rigid Link to
Transfer Beam

2.5 Transfer Beam Elements

The transfer beam (see Figure 5) consists of beam elements that act to transfer forces to
each of the bearing pad elements from the superstructure. Conversely, the transfer beam may
accumulate all of the bearing pad element forces and transfer these forces to the superstructure.
Regarding connectivity, the transfer beam of a given FB-MultiPier model connects all of the
bearing pad elements in a given single row to a vertical link, which is described below. The stiffness
of a given transfer beam may be further delineated as a combination of the cross-sectional area,
moments of inertia, and modulus of elasticity. Alternatively, the transfer beam properties can be
custom-defined to promote modeling of the desired span-end (e.g., diaphragm) rigidity. The mass
of these elements is assigned as zero so as to minimize dynamic effects due to mass contributions
from specialized members while still transferring any conceivable real-world loading to the other
portions of a given pier structure.

5
 Figure 5. Pier Cap Beam Denoting Transfer Beam Connectivity – Bearing Pad Element to Vertical
Link

2.6 Vertical Link Elements

The vertical link consists of a beam element and acts to transfer forces from the
superstructure beam to the transfer beam (see Figure 6). The main purpose of the vertical link is to
transfer load from the center of the transfer beam up to the mid-height of the superstructure, where
it connects to the “spine” model of the superstructure. The moments of inertia of the vertical link
are assigned as the same values as those of the superstructure beam. However, the orientations of
the local material axes differ between these two members, as shown in Figure 7. The mass of this
element is assigned as zero so as to minimize dynamic effects due to mass contributions from
specialized members while still transferring any conceivable real-world loading to the other
portions of a given pier structure.

6
Figure 6. Pier Cap Beam Denoting Vertical Link Connectivity – Transfer Beam to Superstructure
Beam

7
Figure 7. Material Axes Orientation of Vertical Link Relative to that of the Superstructure Beam

8
3. VALIDATION MODEL DESCRIPTIONS/RESULTS

The following is a description of the various validation models that were tested during this
study, as well as a comparison of the displacement output between ADINA and FB-MultiPier
models. As aforementioned, the models are presented in order of increasing complexity. Wherever
feasible, a comparison is also made between the software output and theoretical displacements.

9
3.1 V001 Series Models; Single Pile Models

3.1.1 Model 1; Single Pile Model – Static Load - Flexure

Description: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity

Loading: A lateral load of 10 kips was statically applied to the top node to test pile
flexure.

Software Comparison:

20"x20" Prestressed Concrete Pile - X-Displacement Profile

0
Depth (-20 ft = Point of fixity) (ft)

10

15

ADINA
FB-MultiPier
20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Lateral Deflection (in)

(a) (b)
Figure 8(a). Physical Model Description and 8(b). Pile X-Displacement Profile

Percent Difference Comparison:

Table 1. Displacement Comparisons; Theoretical and Relative

Max Theoretical
Data Displacement Percent Software Percent
Source (in) Difference (%) Difference (%)
ADINA 0.7828 0.0000
FB- 0.0268
MultiPier 0.7830 0.0255
Theory* 0.7828 - -
*- Tip Deflection of a Cantilevered Bernoulli Beam (∆ = PL3/3EI); where ∆ is the tip displacement, P is the applied load, L is the
length of the member, E is the modulus of elasticity, and I is the moment of inertia.

10
3.1.2. Model 2; Single Pile Model – Static Load - Torsion

Description: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity

Loading: A Z-Axis moment of 100 kip-ft was statically applied at the top node to test
pile flexure.

Software Comparison:

Figure 9. Physical Model Description

Percent Difference Comparison:

Table 2. Displacement Comparisons; Theoretical and Relative

Max Theoretical
Displacement Percent Software Percent
Data Source (rad) Difference (%) Difference (%)
ADINA 6.9495E-03 0.0000
0.0000
FB-MultiPier 6.9495E-03 0.0000
Theory* 6.9495E-03 - -
*- Tip Deflection of a Pure Torsion Beam Member (∆ = TL/GJ); where ∆ is the tip rotation, T is the applied moment, L is the length
of the member, G is the shear modulus, and J is the torsional moment of inertia.

11
3.1.3 Model 3; Single Pile Model – Dynamic Load - Flexure

Description: A single pile with fixed base; 30’ length; 24” Diameter with 0.5”
thickness; steel; 29000 ksi modulus of elasticity

Loading: A lateral load of 10 kips was dynamically applied to the top node to test pile
flexure – dynamic behavior.

Software Comparison:

Pile 3 X-Displacement
3

2
Lateral Deflection (in)

2
ADINA
FB-MultiPier
3
0 2 4 6 8 10

time (sec)
(a) (b)
Figure 10(a). Physical Model Description and 10(b). Pile X-Displacement vs Time

Percent Difference Comparison:

Table 3. Displacement Comparisons; Relative

X-Displacement at Software Percent
Data Source Time = 10 sec (in) Difference (%)
ADINA 1.9638
1.382
FB-MultiPier 1.9872
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

12
3.2 V002 Series Models; Single Pile Cap with Four Piles Models

3.2.1 Model 4; Single Pile Cap with Four Piles Model – Static Load - Flexure

Description:
Piles: A four-pile model with fixed bases; 30’ length; 24” diameter pipe pile
with 0.5” thickness; Steel; 29000 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 6”x6” shell elements; torsional properties were enabled in the
FB-MultiPier model (see Sec 2.1)

Loading: A lateral load of 100 kips was statically applied at the pile cap mid-plane to
test pile group behavior.

Software Comparison:

(a) (b)
Figure 11(a). Physical Model Description and 11(b). Typical Pile X-Displacement Profile

Percent Difference Comparison:

Table 4. Displacement Comparisons; Relative

Software Percent
Data Source Max X-Displacement (in) Difference (%)
ADINA 1.4224
-0.2962
FB-MultiPier 1.4182
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

13
3.3 V003 Series Models; Experimental Models

3.3.1. Model 5; Experimental Model – Static Loading – Pile Torsion

Description:
Pile: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap; concrete; 4442000000 modulus of elasticity;
finite element mesh consisted of 1’x1’ shell elements; torsional properties were
enabled in the FB-MultiPier model (see Sec 2.1)
Pier Column: A single column; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
(see Sec 2.2)
Loading: A Z-Axis moment of 42 kip-ft was statically applied at the top node of the pile
to test pile torsion behavior.

Software Comparison:

Figure 12. Physical Model Description

Percent Difference Comparison:

Table 5. Displacement Comparisons; Theoretical and Relative
Pile Head Theoretical Percent Software Percent
Data Source Displacement (rad) Difference (%) Difference (%)
ADINA 0.002896 0.0000
0.0345
FB-MultiPier 0.002895 0.0345
Theory* 0.002896 - -
*- Tip Deflection of a Pure Torsion Beam Member (∆ = TL/GJ); where ∆ is the tip rotation, T is the applied moment, L is the length
of the member, G is the shear modulus, and J is the torsional moment of inertia.

14
3.3.2 V003-2 Series Models; Experimental Models – Pier Column Top Node Loading

Model 6; Experimental Model – Static Loading – Axial on Pier Column

Description:
Pile: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap; concrete; 4442 modulus of elasticity; finite
element mesh consisted of 1’x1’ shell elements; torsional properties were
enabled in the FB-MultiPier model (see Sec 2.1)
Pier Column: A single column; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model

Loading: A Z-Axis load of 100 kips was statically applied at the top node of the pile to
test proper load transfer in system flexure.

Software Comparison:

Figure 13. Physical Model Description

Percent Difference Comparison:

Table 6. Displacement Comparisons; Relative
Data Source Pier Column Top Node Displacement (in) Software Percent Difference (%)
ADINA 3.8118
0.1915
FB-MultiPier 3.8191
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

15
Model 7; Experimental Model – Static Loading – Lateral on Pier Column

Description:
Pile: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap; concrete; 4442 modulus of elasticity; finite
element mesh consisted of 1’x1’ shell elements; torsional properties were
enabled in the FB-MultiPier model (see Sec 2.1)
Pier Column: A single column; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model

Loading: An X-Axis load of 1 kips was statically applied at the top node of the pile to
test proper load transfer in system flexure.

Software Comparison:

Figure 14. Physical Model Description

Percent Difference Comparison:

Table 7. Displacement Comparisons; Relative

Pier Column Top Node X- Software Percent
Data Source Displacement (in) Difference (%)
ADINA 0.6298
0.1431
FB-MultiPier 0.6289
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

16
Model 8; Experimental Model – Static Loading – Longitudinal on Pier Column
Description:
Pile: A single pile with fixed base; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap; concrete; 4442 modulus of elasticity; finite
element mesh consisted of 1’x1’ shell elements; torsional properties were
enabled in the FB-MultiPier model (see Sec 2.1)
Pier Column: A single column; 20’ length; 20”x20” Prestressed Concrete;
4415 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model

Loading: A Y-Axis load of 100 kips was statically applied at the top node of the pile to
test proper load transfer in system flexure.

Software Comparison:

Figure 15. Physical Model Description

Percent Difference Comparison:

Table 8. Displacement Comparisons; Relative
Pier Column Top Node Displacement Software Percent
Data Source (in) Difference (%)
ADINA 0.6823
0.0000
FB-MultiPier 0.6823
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

17
Model 9; Experimental Model – In-plane Pile Cap Shear without Torsional Provisions

Description:
Pile Cap: A 4’ thick pile cap; concrete; 4442 modulus of elasticity; finite
element mesh consisted of 1’x1’ shell elements; torsional properties were not
enabled in the FB-MultiPier model (see Sec 2.1)
Special Boundary Conditions: A line of supports was applied along Line I
(as shown in Figure 18 below), fixing all degrees of freedom (DOF). A line of
supports was applied along Line II fixing all DOF except for Y-Axis deflection.
These boundary conditions were applied in an attempt to isolate shear within
the pile cap.

Loading: A Y-Axis line consisting of twenty-one 100 kip loads was statically applied at
each node along the Line II support system.

Software Comparison:

Figure 16. Physical Model Description

Percent Difference Comparison:

Table 9. Displacement Comparisons; Relative and Theoretical
Line II Average Displacement Theoretical Percent Software Percent
Data Source (in) Difference (%) Difference (%)
ADINA 0.01092 15.49
0.3663
FB-MultiPier 0.01096 15.92
Theory* 0.009455 - -
*- Pure shear deflection is calculated using ∆ = Vh/Gdw; where ∆ is the displacement, V is the applied shear, h is the span of the
member between boundary conditions, G is the shear modulus, d is the depth of the member, and w is the width of the member.

18
Model 10; Experimental Model – Pile Cap – Through Cap Bending without Torsional
Provisions – Thin Cap
Description:
Pile Cap: A 6” thick pile cap; concrete; 4442 modulus of elasticity; finite
element mesh consisted of 1’x1’ shell elements; torsional properties were not
enabled in the FB-MultiPier model (see Sec 2.1)
Special Boundary Conditions: A line of supports was applied along Line I
(as shown in Figure 18 below), fixing all degrees of freedom (DOF). A line of
supports was applied along Line II fixing all DOF except for Z-Axis deflection
and Y-Axis rotation. These boundary conditions were applied in an attempt to
isolate bending about the Y-axis within the pile cap.
Loading: A Z-Axis line consisting of twenty-one 100 kip loads was statically applied at
each node along the Line II support system.
Software Comparison:

Figure 17. Physical Model Description

Percent Difference Comparison:

Table 10. Displacement Comparisons; Relative and Theoretical
Average Pile Cap Z Theoretical Percent Software Percent
Data Source Displacement (in) Difference (%) Difference (%)
ADINA 7.866 2.504
0.2161
FB-MultiPier 7.883 2.293
Theory* 8.068 - -
*- Tip Deflection of a Cantilevered Bernoulli Beam (∆ = PL3/3EI); where ∆ is the tip displacement, P is the applied load, L is the
length of the member, E is the modulus of elasticity, and I is the moment of inertia.

19
3.3.4 V004 Series Models; Old St. George Island – Pier 3 – Single Pier Models
Model 11; Single Pier Model – Static – Lateral Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were disabled in the
FB-MultiPier model (see Sec 2.1)
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Loading: An X-Axis load of 500 kips was statically applied to the top pier column node
to test pier lateral behavior.

20
Software Comparison:

Pier Column Displacement Profile

Distance Above Pile Cap Midplane (ft)

40

30

20

10
ADINA
FB-MultiPier
0
0 1 2 3 4 5 6

Pile Displacement Profile

20
Dist from Point of Fixity (ft)

15

10

5
ADINA
FB-MultiPier
0
0 1 2 3 4 5 6

Lateral Deflection (in)

(a) (b)
Figure 18(a). Physical Model Description and 18(b). Typical Pile and Pier Column X-Displacement
Profiles

Percent Difference Comparison:

Table 11. Displacement Comparisons; Relative
Data Source Maximum X-Displacement (in) Software Percent Difference (%)
ADINA 1.5256
0.8128
FB-MultiPier 1.5380
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

21
Model 12; Single Pier Model – Static – Longitudinal Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were disabled in the
FB-MultiPier model (see Sec 2.1)
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Loading: A Y-Axis load of 500 kips was statically applied to the top pier column node
of each column to test pier longitudinal behavior.

22
Software Comparison:

Pier Column Displacement Profile

Height above Pile Cap midplane (ft)

40

30

20

10
FBMultiPier
ADINA
0
0 5 10 15 20 25 30

Pile Displacement Profile

20
Height above Pile Point of Fixity (ft)

15

10

FB-MultiPier
ADINA
0
0 5 10 15 20 25 30

Displacement (in)
(a) (b)
Figure 19(a). Physical Model Description and 19(b). Typical Pile and Pier Column X-Displacement
Profiles

Percent Difference Comparison:

Table 12. Displacement Comparisons; Relative
Data Source Maximum X-Displacement (in) Software Percent Difference (%)
ADINA 26.002
2.873
FB-MultiPier 26.749
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

23
Model 13; Single Pier Model – Static – Longitudinal Twist Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were disabled in the
FB-MultiPier model (see Sec 2.1)
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Loading: Two Y-Axis loads, oppositely directed, with a magnitude of 500 kips each were
statically applied to the top pier column node of each column to test pier longitudinal twist
behavior.

24
Software Comparison:

Pier Column Displacement Profile

40

Height above Pile Cap midplane (ft)

30

20

10

FBMultiPier
ADINA
0
0 2 4 6

Pile Displacement Profile

20
Height above Pile Point of Fixity (ft)

15

10

FB-MultiPier
ADINA
0
0 2 4 6

Displacement (in)
(a) (b)
Figure 20(a). Physical Model Description and 20(b). Typical Pile and Pier Column X-Displacement
Profiles

Percent Difference Comparison:

Table 13. Displacement Comparisons; Relative
Data Source Maximum X-Displacement (in) Software Percent Difference (%)
ADINA 5.4622
0.6664
FB-MultiPier 5.4986
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

25
Model 14; Single Pier Model – Dynamic – Lateral Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were disabled in the
FB-MultiPier model (see Sec 2.1)
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Loading: An X-Axis load of 500 kips was dynamically applied and suddenly released at
the top pier column node to test pier lateral dynamic behavior.

26
Software Comparison:

Pier 3 - Lateral Dynamic Load Response

6

Displacement (in) 2

6
0 2 4 6 8 10

Time (sec)
ADINA
FB-MultiPier
(a) (b)
Figure 21(a). Physical Model Description and 21(b). Pier Column Top Node Displacement vs Time

Percent Difference Comparison:

Table 14. Displacement Comparisons; Relative
Data Source Peak X-Displacement at 2 sec (in) Software Percent Difference (%)
ADINA 4.0729
2.301
FB-MultiPier 4.1666
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

27
Model 15; Single Pier Model – Dynamic – Longitudinal Twist Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were disabled in the
FB-MultiPier model (see Sec 2.1)
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were applied to the ADINA model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Loading: Two Y-Axis loads, oppositely directed, with a magnitude of 500 kips each were
dynamically applied and suddenly released at the top pier column node of each column to
test pier longitudinal twist behavior.

28
Software Comparison:

Pier 3 - Longitudinal Twist Dynamic Response

6

Displacement (in)

6
0 2 4 6 8 10

Time (sec)
ADINA
FB-MultiPier
(a) (b)
Figure 22(a). Physical Model Description and 22(b). Pier Column Top Node Displacement vs Time

Percent Difference Comparison:

Table 15. Displacement Comparisons; Relative
Data Source Peak Y-Displacement at 2 sec (in) Software Percent Difference (%)
ADINA 5.5153
1.002
FB-MultiPier 5.4606
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

29
3.3.5 V005 Series Models; Old St. George Island – Pier 3 – Two Pier Models

Model 16; Two Pier Model – Static – Lateral Load

Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were enabled in the
FB-MultiPier model (see Sec 2.1)
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were not applied to the ADINA
model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Superstructure Beam: Modeled using moments of inertia representative
of the gross moments of inertia of the actual Old St. George Island Bridge
superstructure; Reinforced concrete; 4442 ksi modulus of elasticity; all of the
special elements described in Sect 2.3-2.6 were applied to the ADINA model in
order to emulate the superstructure to substructure interactions that occur in FB-
MultiPier
Loading: Two X-Axis loads with a magnitude of 500 kips each were statically applied
at the top pier column node of one pier column of each pier to test multiple pier lateral
static behavior.

30
Software Comparison:

Pier 1 Pile X-Displacement

20

Depth (ft) 15

10

ADINA
FB-MultiPier
0
0 0.5 1 1.5 2

Lateral Deflection (in)

(a) (b)
Figure 23(a). Physical Model Description and 23(b). Pier 1 – Typical Pile Displacement

Percent Difference Comparison:

Table 16. Displacement Comparisons; Relative
Data Source Peak X-Displacement (in) Software Percent Difference (%)
ADINA 1.5232
1.031
FB-MultiPier 1.5389
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

31
Model 17; Two Pier Model – Dynamic – Lateral Load
Description:
Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap per pier column; 8’x10’ X-Axis and Y-Axis
dimensions, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were enabled in the
FB-MultiPier model (see Sec 2.1)
Pier Strut: 48”x30” X-Axis and Z-Axis dimensions, respectively; Reinforced
concrete; 4442 Modulus of Elasticity; placed 4’ above pile cap midplane
Pier Column: Two pier columns spaced 20.5’ center to center; 47.5’ length;
50”x42” X-Axis and Y-Axis dimensions, respectively; Reinforced concrete;
4442 ksi modulus of elasticity; Quad-Pods were not applied to the ADINA
model
Pier Cap Beam: 50”x48” X-Axis and Y-Axis dimensions, respectively;
Reinforced concrete; 4442 ksi modulus of elasticity; connected pier columns
Pier Superstructure Beam: Modeled using moments of inertia representative
of the gross moments of inertia of the actual Old St. George Island Bridge
superstructure; Reinforced concrete; 4442 ksi modulus of elasticity; all of the
special elements described in Sect 2.3-2.6 were applied to the ADINA model in
order to emulate the superstructure to substructure interactions that occur in FB-
MultiPier
Loading: Two X-Axis loads with a magnitude of 500 kips each were dynamically applied
and suddenly released at the top pier column node of one pier column of each pier to test
multiple pier lateral dynamic behavior.

32
Software Comparison:

Two Pier - Lateral Dynamic Response

2

1
Lateral Displacement (in)

FB-MultiPier
ADINA
2
0 2 4 6 8 10 12 14 16 18

Time (sec)
(a) (b)
Figure 24(a). Physical Model Description and 24(b). Pier 1 - Pile 8 Top Node Displacement vs Time

Percent Difference Comparison:

Table 17. Displacement Comparisons; Relative
Data Source Peak X-Displacement at 8 sec (in) Software Percent Difference (%)
ADINA 1.518
2.503
FB-MultiPier 1.556
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

33
3.3.6 V006 Series Models; Miscellaneous Models; Brick vs Shell Element Pile Caps
Model 18; Pile Cap Models – Static – Lateral Load
Description:
The V005-5 Pile Cap Models consist of a comparison of 3 models within FB-
MultiPier and ADINA, with varying aspect ratios with respect to cap length and depth.
The specific aspect ratios and accompanying dimensions are described below.

Piles: Four piles per pier column with fixed bases for each pile; 20’ length;
20”x20” Prestressed Concrete; 4415 ksi modulus of elasticity
Pile Cap: A 4’ thick pile cap for each of three models in each program; with
X-Axis and Y-Axis dimensions of 16’x8’, 24’x8’, and 40’x8’ for aspect ratios
of 4, 6, and 10, respectively; concrete; 4442 modulus of elasticity; finite element
mesh consisted of 1’x1’ shell elements; torsional properties were enabled in the
FB-MultiPier model (see Sec 2.1)
Loading: One X-Axis load with a magnitude of 100 kips was statically applied at the
mid-plane of the pile cap to test lateral static behavior of brick and shell element pile caps.
Software Comparison:

(a)

34
 Pile Displacements in the X-Direction for Aspect Ratio = 4
30
28
26
24
22
20
18
Depth (in)

16
14
12
10
8
6
4 shell element aspect=4
2 brick element aspect=4
0
0 0.11 0.21 0.32 0.43 0.53 0.64 0.75 0.85 0.96 1.07 1.17 1.28 1.39 1.49 1.6

X-Displacement (in)
(b)
Pile Displacements in the X-Direction for Aspect Ratio = 6
30
28
26
24
22
20
18
Depth (in)

16
14
12
10
8
6
4 shell element aspect=6
2 brick element aspect=6
0
0 0.093 0.19 0.28 0.37 0.47 0.56 0.65 0.75 0.84 0.93 1.03 1.12 1.21 1.31 1.4

X-Displacement (in)
(c)

35
 Pile Displacements in the X-Direction for Aspect Ratio = 10
30
28
26
24
22
20
18
Depth (in)

16
14
12
10
8
6
4 shell element aspect=10
2 brick element aspect=10
0
0 0.093 0.19 0.28 0.37 0.47 0.56 0.65 0.75 0.84 0.93 1.03 1.12 1.21 1.31 1.4

X-Displacement (in)
(d)
Figure 25(a). Physical Model Description (Shell Element Pile Cap with an Aspect Ratio of 6 is shown)
and 25(b)-25(d). Typical Pile Displacements for Aspect Ratios 4, 6, and 10, respectively

Percent Difference Comparison:

Table 18. Displacement Comparisons; Relative
Peak X-Displacement (in) Software Percent Difference (%)
Data Source Aspect 4 Aspect 6 Aspect 10 Aspect 4 Aspect 6 Aspect 10
ADINA 1.3492 1.29978 1.29907
5.091 5.603 5.309
FB-MultiPier 1.4182 1.37261 1.36804
*- For software comparison, the value of displacement with the smallest absolute value is taken is the denominator in the percent
difference calculation.

36