Deepsoil User Manual v7

Nonlinear and Equivalent Linear Seismic Site Response of
One-Dimensional Soil Columns
Version 7.0
www.illinois.edu/~deepsoil
February 1, 2018
USER MANUAL
Youssef M. A. Hashash
Department of Civil and Environmental Engineering

University of Illinois at Urbana-Champaign
hashash@illinois.edu
When referencing the DEEPSOIL program in a publication (such as journal or conference papers, or
professional engineering reports) please use the following reference format
Hashash, Y.M.A., Musgrove, M.I., Harmon, J.A., Ilhan, O., Groholski, D.R., Phillips, C.A., and Park,
D. (2017) “DEEPSOIL 7.0, User Manual”.
DEEPSOIL User Manual V 7.0 Page 1 of 169 February 1, 2018

TABLE OF CONTENTS
1 Program Background and Installation .................................................................................. 14
1.1 About the Program ................................................................................................... 14
1.2 A Word of Advise and Caution ............................................................................... 15
1.3 Historical Development ........................................................................................... 15
1.4 Program Installation................................................................................................. 19
2 Program Organization ........................................................................................................... 21
2.1 Profiles Tab.............................................................................................................. 22
2.2 Motions Tab ............................................................................................................. 22
2.2.1 Baseline Correction ................................................................................................. 25
2.2.2 Response Spectra Calculation Methods................................................................... 27
2.2.3 Fourier Amplitude Spectrum Calculation and Smoothing Operation ..................... 29
2.2.4 Arias Intensity.......................................................................................................... 31
2.2.5 Significant Duration................................................................................................. 31
2.2.6 Housner Intensity ..................................................................................................... 31
2.2.7 Estimation of Kappa (κ) .......................................................................................... 32
2.2.8 Adding New Input Motions ..................................................................................... 33
2.3 Analysis Tab ............................................................................................................ 33
3 Analysis Flow ....................................................................................................................... 34
3.1 Analysis Definition: Step 1 of 5 .............................................................................. 34
3.1.1 Linear Analysis ........................................................................................................ 38
3.1.2 Equivalent Linear Analysis ..................................................................................... 38
3.1.3 Deconvolution via Frequency Domain Analysis ..................................................... 38
3.1.4 Non-Linear Analysis ................................................................................................ 40
3.2 Soil Profile Definition: Step 2 of 5 .......................................................................... 40

3.2.1 Single Element Test ................................................................................................. 43
3.2.2 Maximum Frequency (for Time Domain Analysis only) ........................................ 43
3.2.3 Implied Strength Profile (Step 2) ............................................................................. 45
3.2.4 Halfspace Definition (Bedrock) ............................................................................... 45
3.3 Step 3 of 5: Input Motion Selection......................................................................... 47
3.4 Step 4 of 5: Viscous/Small-Strain Damping Definition .......................................... 48
3.4.1 Frequency Independent Damping Formulation ....................................................... 50
3.4.2 Rayleigh Damping formulation types ...................................................................... 50
3.5 Step 5 of 5: Analysis Control Definition ................................................................. 51
3.5.1 Frequency domain analysis ..................................................................................... 52
3.5.2 Time domain analysis .............................................................................................. 53
3.6 Results ..................................................................................................................... 55
3.6.1 Time History Plots tab ............................................................................................. 56
3.6.2 Stress-Strain Plots tab ............................................................................................. 56
3.6.3 Spectral Plots tab ..................................................................................................... 57
3.6.4 Profile Plots tab ....................................................................................................... 58
3.6.5 Mobilized Strength tab ............................................................................................. 58
3.6.6 Displacement Animation tab .................................................................................... 60
3.6.7 Response Spectra Summary tab ............................................................................... 60
3.6.8 Check Convergence tab ........................................................................................... 60
3.6.9 Output data file ........................................................................................................ 61
4 Soil Models ........................................................................................................................... 62
4.1 Backbone Curves ..................................................................................................... 62
4.1.1 Hyperbolic / Pressure-Dependent Hyperbolic (MKZ) ............................................ 62
4.1.2 Generalized Quadratic/Hyperbolic (GQ/H) Model with Shear Strength Control ... 63

4.2 Hysteretic (Unload-Reload) Behavior ..................................................................... 64
4.2.1 Masing Rules ........................................................................................................... 64
4.2.2 Non-Masing Unload-Reload Rules.......................................................................... 64
4.3 Porewater Pressure Generation & Dissipation ........................................................ 66
4.3.1 Dobry/Matasovic Model for Sand ........................................................................... 66
4.3.2 Matasovic and Vucetic Model for Clays ................................................................. 72
4.3.3 GMP (Green, Mitcher and Polito) Model for Cohesionless Soil............................. 74
4.3.4 Generalized Energy-based PWP Generation Model................................................ 76
4.3.5 Park and Ahn Model for Sand ................................................................................. 76
4.3.6 Porewater Pressure Degradation Parameters ........................................................... 78
4.3.7 Porewater Pressure Dissipation ............................................................................... 79
5 Randomization of Site Profile Properties ............................................................................. 80
6 Database Output Structure .................................................................................................... 81
6.1 Database Structure for Analyses Output.................................................................. 81
7 Examples and Tutorials......................................................................................................... 84
7.1 Example 1: Undamped Linear Analysis with Resonance........................................ 84
7.2 Example 2: Undamped Linear Analysis with Elastic Bedrock ............................... 93
7.3 Example 3: Damped Linear Analysis with Elastic Bedrock ................................... 98
7.4 Example 4: Equivalent Linear Analysis with Discrete Points ............................... 101
7.5 Example 5: Nonlinear Analysis (MKZ Soil Model with Masing Re/Unloading
Behavior) 103
7.6 Example 6: Nonlinear Analysis (GQ/H Soil Model with Non-Masing Re/Unloading
Behavior) 106
7.7 Example 7: Nonlinear Analysis / Multi-Layer, Elastic Rock, Pore Water Pressure
Generation and Dissipation ..................................................................................................... 115

7.8 Example 8: Nonlinear Analysis / Multi-Layer, Elastic Rock, Pore Pressure
Generation and Dissipation: .................................................................................................... 122
7.9 Example 9: Equivalent Linear Frequency Domain Analysis / Multi-Layer, Elastic

Rock, Bay Mud Profile ........................................................................................................... 126
7.10 Example 10: Nonlinear Analysis / Multi-Layer, Rigid Rock, Treasure Island Profile
130
7.11 Example 11: Nonlinear Analysis / Multi-Layer, Elastic Rock, MRDF ................. 136
7.12 Example 12: Nonlinear Analysis with Auto-Profile Generation Option: .............. 141
7.13 Example 13: Nonlinear Analysis with Randomized Soil Profile: ......................... 149
7.14 Example 14: Nonlinear Analysis by Soil Profile with various Soil Models at Different
Layers: 156
8 References ........................................................................................................................... 163
9 APPENDICES .................................................................................................................... 169

LIST OF TABLES
Table 3.1. Available analysis option in DEEPSOIL 7 .................................................................. 36

Table 3.2. IDs of different Soil Models in DEEPSOIL – Soil Model Descriptions ..................... 37
Table 4.1 Available Excess Pore Water Pressure Generation Models and Parameters ................ 66
Table 4.2 Description of Dobry/Matasovic Model Parameters .................................................... 67
Table 4.3: Material Parameters for Low Plasticity Silts and Sands for the Matasovic and Vucetic
(1993) pore pressure generation model (From Carlton, 2014) ..................................................... 70
Table 4.4 Description of Matasovic and Vucetic Model Parameters ........................................... 72
Table 4.5 Material parameters for the Matasovic and Vucetic (1995) clay pore pressure generation
model (From Carlton, 2014) ......................................................................................................... 73
Table 4.6 Description of GMP Model Parameters........................................................................ 74
Table 4.7 Description of Generalized Model Parameters ............................................................. 76
Table 4.8 Description of Park and Ahn Model Parameters .......................................................... 77
Table 7.1 Analyses properties and input motions for Example 1 ................................................. 84
Table 7.4 Analyses properties and input motions for Example 4 ............................................... 101
Table 7.7 Sand-Vucetic Dobry PWP model parameters for Sand layer ..................................... 116
Table 7.8 GMP PWP model parameters for Sand layer ............................................................. 122

LIST OF FIGURES
Figure 2.1. DEEPSOIL Main Window and Key Tabs as (a) Analysis Tab, (b) Motions Tab, and
(c) Profiles Tab ............................................................................................................................. 21
Figure 2.2. DEEPSOIL Options Window. .................................................................................... 22
Figure 2.3. Motion Viewer (right-click choices on Plots) ............................................................ 23
Figure 2.4. Motion Viewer (Motion Metrics and Tools) .............................................................. 24
Figure 2.5 Single Motion View .................................................................................................... 25
Figure 2.6. Baseline Correction. ................................................................................................... 26
Figure 2.7 Kappa Estimator .......................................................................................................... 32
Figure 3.1. Step 1/5: Choose type of analysis. ............................................................................. 34
Figure 3.2. Choose the Default Working Directory. ..................................................................... 35
Figure 3.3. Input Summary window. ............................................................................................ 37
Figure 3.4. Deconvolution analysis parameters. ........................................................................... 39
Figure 3.5. Complementary Equivalent Linear-Frequency Domain analysis............................... 40
Figure 3.6. Soil Profile Definition – Advanced Table View window. ......................................... 42
Figure 3.7. Layer Properties Tab. ................................................................................................. 42
Figure 3.8 Single Element Test Window ...................................................................................... 44
Figure 3.9. Soil Profile Plot. ......................................................................................................... 44
Figure 3.10. Halfspace Definition – “Bedrock”............................................................................ 46
Figure 3.11. Input Motion Selection. ............................................................................................ 47
Figure 3.12. Dropdown tools menu. ............................................................................................. 49
Figure 3.13. Viscous/Small-Strain Damping Definition............................................................... 49
Figure 3.14. Analysis Control Definition. .................................................................................... 51
Figure 3.15. Analysis Running. .................................................................................................... 55
Figure 3.16. Results - Time History Plots..................................................................................... 56
Figure 3.17. Results – Stress Strain Plots. .................................................................................... 57
Figure 3.18. Results – Spectral Plots. ........................................................................................... 57
Figure 3.19. Results – Profile Plots. ............................................................................................. 58
Figure 3.20. Results – Mobilized Strength. .................................................................................. 59
Figure 3.21. Results – Displacement Animation. ......................................................................... 59
Figure 3.22. Results – Response Spectra Summary. .................................................................... 60

Figure 3.23 Check Convergence Tab............................................................................................ 61
Figure 4.1 a) Carlton (2014), best fit correlating Vs (m/sec) to parameter F of Dobry pore water
pressure model for sands. b) Carlton (2014), best fit correlating FC (%) to parameter s of Dobry
pore water pressure model for sands ............................................................................................. 69
Figure 4.2 Proposed correlation to estimate curve-fitting parameter F (Mei et al. 2015) ............ 71
Figure 4.3 Comparison of the curves given by Matasovic (1993) and Vucetic (1992) (solid black
lines) for t, for different values of PI and OCR and the correlations presented (dotted red lines).
(Carlton, 2014) .............................................................................................................................. 73
Figure 6.1 DEEPSOIL V7.0 Output Structure.............................................................................. 83
Figure 7.1 Soil profile of Example 1 ............................................................................................ 84
Figure 7.2 Step 1: Analysis Type Definition for Example 1A ..................................................... 86
Figure 7.3 Step 2: Soil Profile Definition for Example 1A and 1B .............................................. 86
Figure 7.4 Step 2: Bedrock Profile Definition in Example 1A and 1B ........................................ 87
Figure 7.5 Step 2: Soil profile plot for Example 1A and 1B ........................................................ 87
Figure 7.6 Step 3: Input motion selection window for Example 1A and 1B ................................ 88
Figure 7.7 Step 4: Viscous/Small-Strain Damping Definition for Example 1B ........................... 88
Figure 7.8 Step 5: Analysis Control Definition Window for Example 1A ................................... 89
Figure 7.9 Step 5: Analysis Control Definition for Example 1B .................................................. 89
Figure 7.10 Results: Acceleration, Velocity, Displacement and Arias Intensity Time-Histories for
Example 1A .................................................................................................................................. 90
Figure 7.11 Results: 5% Damped Spectral Acceleration, Fourier Amplitude Spectrum (FAS) and
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 1A
....................................................................................................................................................... 90
Figure 7.12 Results: Exporting Analysis Results using Export to Excel Option.......................... 91
Figure 7.13 Results: Excel Output for Analysis Results for Example 1A .................................... 91
Example 1B................................................................................................................................... 92
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 1B
....................................................................................................................................................... 92
Figure 7.16 Soil profile of Example 2 .......................................................................................... 93

Figure 7.17 Step 2: Bedrock Profile Definition for Example 2A and 2B ..................................... 95
Figure 7.18 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for
Example 2A .................................................................................................................................. 95
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 2A
....................................................................................................................................................... 96
Example 2B................................................................................................................................... 96
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 2B
....................................................................................................................................................... 97
Figure 7.22 Soil profile of Example 3 .......................................................................................... 98
Figure 7.23 Result: Computed Time-Histories (Acceleration, velocity, displacement, and Arias
Intensity) for Layer 1 for Example 3A ......................................................................................... 99
Figure 7.24 Result: Computed 5% Damped Spectral Acceleration for Layer 1 along with that for
Input Motion for Example_3A...................................................................................................... 99
Intensity) for Layer 1 for Example 3B ........................................................................................ 100
Input Motion for Example_3B .................................................................................................... 100
Figure 7.27 Soil profile of Example 4 ........................................................................................ 101
Intensity) for Layer 1 for Example 4 .......................................................................................... 102
Input Motion for Example_3A.................................................................................................... 102
Figure 7.31 5% Damped Spectral Acceleration and Fourier Amplitude Spectrum computed at
Ground Surface for Nonlinear and Equivalent-Linear Analyses using DC, MR and MRD fitting
procedures along with those for Input Motion ............................................................................ 105
Figure 7.33 Step 1: Analysis Type Definition for Example 6. ................................................... 109

Figure 7.34 Step 2: Soil Profile Definition for Example 6 ......................................................... 109
Figure 7.35 Step 2: Shear Strength Input for Soil Layer 1 ......................................................... 110
Figure 7.36 Step 2: Definition of Darendeli (2001) Dynamic Curve ......................................... 110
Figure 7.37 Step 2: Application of “MRDF with UIUC Reduction Factor” Option .................. 111
Figure 7.38 Step 2: Fitting of GQ/H Model ................................................................................ 111
Figure 7.39 Step 2: GQ/H Model Fit and Model Parameters for Layer 1 .................................. 112
Figure 7.40 Step 5: Analysis Control Definition for Example 6 ................................................ 112
Intensity) for Layer 1 via Nonlinear Analysis (blue line) and Equivalent-Linear Analysis (red line)
methods ....................................................................................................................................... 113
Input Motion via Nonlinear Analysis Method ............................................................................ 113
Input Motion via Equivalent-Linear Analysis Method ............................................................... 114
Figure 7.44 Step 1: Analysis Type Definition for Example 7 .................................................... 117
Figure 7.45 Step 2: Soil Profile in Nonlinear Analysis with PWP Generation and Dissipation 117
Figure 7.46 Step 2: Definition of Parameters for PWP Generation and Dissipation Model of Sand-
Vucetic/Dobry Model ................................................................................................................. 118
Figure 7.47 Step 2: Halfspace Definition ................................................................................... 118
Figure 7.48 Step 2: Soil Profile Plot ........................................................................................... 119
Figure 7.49 Step 3: Time Histories, FAS and 5% Damped Spectral Acceleration for Kobe motion
..................................................................................................................................................... 119
Figure 7.50 Results: Profile Plots after analyses ........................................................................ 120
Example 7 for Layer 1 (blue line) and Layer 3 (red line). .......................................................... 120
Figure 7.52 Results: Stress-Strain plots for Layer 1 (blue line) and Layer 3 (red line) .............. 121
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Layer 1
(blue line) and Layer 3 (red line) ................................................................................................ 121
GMP ............................................................................................................................................ 123

Figure 7.55 Results: Profile Plots ............................................................................................... 123
..................................................................................................................................................... 124
Figure 7.57 Results: Stress-Strain plots for Layer 1 (blue line) and Layer 3 (red line) .............. 124
(blue line) and Layer 3 (red line) ................................................................................................ 125
Figure 7.59 Step 2: Soil Profile Definition ................................................................................. 127
Figure 7.60 Step 2: Bedrock Profile Definition in Example 9 .................................................... 127
Figure 7.61 Step 2: Soil profile plot............................................................................................ 128
Figure 7.62 Results: Computed Time-Histories (Acceleration, velocity, displacement, and Arias
Intensity) for Layer 1 .................................................................................................................. 128
Figure 7.63 Results: 5% Damped Spectral Acceleration for Layer 1 (blue line) and for Input
Motion (black line) ..................................................................................................................... 129
Figure 7.64 Step 1: Analysis Type Definition ............................................................................ 131
Figure 7.65 Step 2: Soil Profile Definition (Layer 1) ................................................................. 131
Figure 7.66 Step 2: Bedrock Profile Definition .......................................................................... 132
Figure 7.67 Step 2: Soil profile plot............................................................................................ 132
Figure 7.68 Step 3: Properties of Kobe Motion .......................................................................... 133
Figure 7.69 Step 4: Viscous/Small-Strain Damping Definition ................................................. 133
Intensity) from Nonlinear (blue line) and Equivalent-Linear Analysis (red line) for Layer 1 ... 134
Figure 7.71 Results: 5% Damped Spectral Acceleration for Layer 1 via Nonlinear Analysis ... 134
Figure 7.72 Results: 5% Damped Spectral Acceleration for Layer 1 via Equivalent-Linear Analysis
..................................................................................................................................................... 135
Figure 7.74 Step 2: Soil Profile Definition (Layer 1) ................................................................. 137
Figure 7.75 Step 2: Halfspace Definition ................................................................................... 138
Figure 7.76 Step 2: Soil profile plots .......................................................................................... 138
Figure 7.77 Step 4: Viscous/Small-Strain Damping Definition ................................................. 139

Intensity) from Nonlinear (blue line) and Equivalent-Linear Analysis (red line) for Layer 1 ... 139
Figure 7.79 Results: 5% Damped Spectral Acceleration for Layer 1 via Nonlinear Analysis ... 140
..................................................................................................................................................... 140
Figure 7.82 Step 1: Inputs to Automatic Profile Generation ...................................................... 143
Figure 7.83 Step 2: Mean Soil Profile Definition (for Layer 1) ................................................. 144
Figure 7.84 Step 2: Advanced Table View ................................................................................. 144
Figure 7.85 Step 2: Bedrock definition as Elastic Half-space .................................................... 145
Figure 7.86 Step 2: Subdivision of each Layer Thickness and GQ/H Model Fit to Layer 1 ...... 145
Figure 7.87 Step 2: Soil Profile Plots.......................................................................................... 146
Figure 7.88 Step 5: Parameters for Nonlinear and Equivalent-Linear Analyses ........................ 146
Nonlinear Analysis (blue line) and Equivalent-Linear Analysis (red line) ................................ 147
Figure 7.90 Results: 5% Damped Spectral Acceleration for Layer 1 (blue line) along with for Input
Motion (black line) from Nonlinear Analysis ............................................................................. 147
Motion (black line) from Equivalent-Linear Analysis................................................................ 148
Figure 7.92 Step 1: Specification of Maximum Frequency and Definition of Parameters for
Thickness, VS and Dynamic Curve Randomization..................................................................... 152
Figure 7.93 Step 2: Properties of Layer 1 for first Realization after Randomization ................. 152
Figure 7.94 Step 2: Export Options for Site Profile Realizations ............................................... 153
Figure 7.95 Randomized VS Profiles along with their logarithmic mean and 95% Confidence
Interval on Logarithmic Mean .................................................................................................... 153
Figure 7.96 Step 2: Soil Profile Plot for Profile Realizations ..................................................... 154
Nonlinear Analysis...................................................................................................................... 154
Fourier Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Nonlinear
Analysis....................................................................................................................................... 155

Motion (black line) from Equivalent-Linear Analysis................................................................ 155
Figure 7.100 Step 2: Basic Soil Properties and GQ/H Model Fit to Rockfill Layers (Layer 1 as
representative) ............................................................................................................................. 158
Figure 7.101 Step 2: Assignment of Linear Material to Bedrock Layers ................................... 158
Figure 7.102 Step 2: Removal of Reduction Factor Formulation for linear Bedrock layers ...... 159
Figure 7.103 Step 2: Basic Soil Properties of Bedrock Layers (Layer 45 as representative) ..... 159
Figure 7.104 Step 2: Rigid half-space definition ........................................................................ 160
Figure 7.105 Step 2: Soil Profile Plots........................................................................................ 160
Figure 7.106 Results: Stress-strain plots for Soil Layer 1 (GQ/H Model with Non-Masing
Un/Reloading Formulation (red line) and Soil Layer 43 (linear elastic soil material (blue line) 161
Layer 1 (red line) and Layer 43 (blue line) ................................................................................. 161
(blue line) and Layer 43 (yellow line) ........................................................................................ 162

1 Program Background and Installation
1.1 About the Program
DEEPSOIL is a one-dimensional site response analysis program that can perform: a) 1-D nonlinear
time domain analyses with and without pore water pressure generation, b) 1-D equivalent linear
frequency domain analyses including convolution and deconvolution, and c) 1-D linear time and
frequency domain analyses.
DEEPSOIL was developed under the direction of Prof. Youssef M.A. Hashash in collaboration
with several graduate and undergraduate students including Duhee Park, Chi-Chin Tsai, Camilo
Phillips, David R. Groholski, Daniel Turner, Michael Musgrove, Byungmin Kim and Joseph
Harmon, Okan Ilhan, and Guangchao Xing at the University of Illinois at Urbana-Champaign. This
manual was developed in collaboration with these students as well as Maria Kontari and Hua Shao.
When referencing the DEEPSOIL program in a publication (such as journal or conference papers,
or professional engineering reports) please use the following reference format:
Hashash, Y.M.A., Musgrove, M.I., Harmon, J.A., Okan, I., Groholski, D.R., Phillips, C.A., and
Park, D. (2017) “DEEPSOIL 7.0, User Manual”.
The program is provided as-is and the user assumes full responsibility for all results. The use
of the DEEPSOIL program requires knowledge in the theory and procedures for seismic site
response analysis and geotechnical earthquake engineering. It is suggested that the user reviews
relevant literature and seek appropriate expertise in developing input of the analysis and
interpretation of the results.
Initial development of DEEPSOIL was based on research supported in part through Earthquake
Engineering Research Centers Program of the National Science Foundation under Award Number
EEC-9701785: the Mid-America Earthquake Center. Any opinions, findings, and conclusions or
recommendations expressed in this material are those of the authors and do not necessarily reflect
the views of the National Science Foundation. The authors gratefully acknowledge this support.

By using this program, the user(s) agree to indemnify and defend Youssef Hashash and the
University of Illinois against all claims arising from use of the software and analysis results by the
user(s) including all third-party claims related to such use.
Please see the program license for additional information.
DEEPSOIL implements the Armadillo C++ linear algebra library (Sanderson, 2010; Sanderson,
2016). Armadillo is an open-source software released under the Mozilla Public License 2.0. A
copy of this license is available at https://www.mozilla.org/MPL/2.0/. You may obtain a copy of
the Armadillo source code at http://arma.sourceforge.net/download.html.
1.2 A Word of Advise and Caution
The appropriate use of this software program requires significant expertise and knowledge in a
number of areas including ground motions, site characterization, seismic behavior of soils and
numerical modeling. Often a team effort maybe required to develop the appropriate input for the
analysis and to interpret analysis results. The appropriate use of the software is exclusively the
responsibility of the user.
1.3 Historical Development
DEEPSOIL has been under development at UIUC since 1998. The driving motivation for the
development of DEEPSOIL was, and continues to be, making site response analysis readily
accessible to students, researchers and engineers worldwide and to support research activities at
UIUC.
In DEEPSOIL we maintain that it is always necessary to perform equivalent linear (EL) in

conjunction with nonlinear (NL) site response analyses. Therefore, DEEPSOIL, since its inception,
has incorporated both analysis capabilities. Version 6 of DEEPSOIL gives the user the option to
automatically obtain EL analysis results whenever an NL analysis is selected, without the need to
separately develop an EL profile.
As with any development, DEEPSOIL has benefited from many prior contributions by other
researchers as well as current and former students at UIUC. For the interested reader, a detailed
description of many of the theoretical developments and the background literature can be found in
the following publications:

▪ Hashash, Y. M., and Park, D. (2001). Non-linear one-dimensional seismic ground motion
propagation in the Mississippi embayment. Engineering Geology, 62(1), 185-206.
▪ Hashash, Y. M., and Park, D. (2002). Viscous damping formulation and high frequency
motion propagation in non-linear site response analysis. Soil Dynamics and Earthquake
Engineering, 22(7), 611-624.
▪ Hashash, Y. M., Tsai, C. C., Phillips, C., and Park, D. (2008). Soil-column depth-dependent
seismic site coefficients and hazard maps for the upper Mississippi Embayment. Bulletin
of the Seismological Society of America, 98(4), 2004-2021. .
▪ Hashash, Y.M.A., Phillips, C. and Groholski, D. (2010). "Recent advances in non-linear
site response analysis", Fifth International Conference on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics, Paper no. OSP 4.
▪ Park, D. (2003) "Estimation of non-linear seismic site effects for deep deposits of the
Mississippi Embayment", Ph.D. Thesis. Department of Civil and Environmental
Engineering. Urbana: University of Illinois, p 311 p.
▪ Park, D., and Hashash, Y. M. (2004). Soil damping formulation in nonlinear time domain
site response analysis. Journal of Earthquake Engineering, 8(02), 249-274.
▪ Park, D., and Hashash, Y. M. (2005). Evaluation of seismic site factors in the Mississippi
Embayment. I. Estimation of dynamic properties. Soil Dynamics and Earthquake
Engineering, 25(2), 133-144.
▪ Park, D., & Hashash, Y. M. (2005). Evaluation of seismic site factors in the Mississippi
Embayment. II. Probabilistic seismic hazard analysis with nonlinear site effects. Soil
Dynamics and Earthquake Engineering, 25(2), 145-156.
▪ Tsai, C. C. (2007). Seismic site response and extraction of dynamic soil behavior from
downhole array measurements (Doctoral dissertation, University of Illinois at Urbana-
Champaign).
▪ Tsai, C. C., and Hashash, Y. M. (2008). A novel framework integrating downhole array
data and site response analysis to extract dynamic soil behavior. Soil Dynamics and
Earthquake Engineering, 28(3), 181-197.
▪ Tsai, C. C., & Hashash, Y. M. (2009). Learning of dynamic soil behavior from downhole
arrays. Journal of Geotechnical and Geoenvironmental Engineering, 135(6), 745-757.

▪ Phillips, C., & Hashash, Y. M. (2008). A simplified constitutive model to simultaneously
match modulus reduction and damping soil curves for nonlinear site response analysis. In
Geotechnical Earthquake Engineering and Soil Dynamics IV (pp. 1-10).
▪ Phillips, C., & Hashash, Y. M. (2009). Damping formulation for nonlinear 1D site response
analyses. Soil Dynamics and Earthquake Engineering, 29(7), 1143-1158.
▪ Groholski, D. R., Hashash, Y. M., Kim, B., Musgrove, M., Harmon, J., and Stewart, J. P.
(2016). Simplified model for small-strain nonlinearity and strength in 1D seismic site
response analysis. Journal of Geotechnical and Geoenvironmental Engineering, 142(9),
04016042.
▪ Musgrove, M., Harmon, J., Hashash, Y. M., & Rathje, E. (2017). Evaluation of the
DEEPSOIL Software on the DesignSafe Cyberinfrastructure. Journal of Geotechnical and
Geoenvironmental Engineering, 143(9), 02817005.
The executable version of DEEPSOIL was originally (circa 1998-1999) developed as a MATLAB
program and (circa 1999) was later redeveloped as a C based executable to improve computational
efficiency. A visual user interface was added soon afterwards. Since then, numerous developments
have been added. Listed below are some important milestones:
• DEEPSOIL v1.0: First version of DEEPSOIL with both an equivalent linear analysis
capability and a new pressure dependent hyperbolic model in nonlinear analysis:
o The equivalent linear capability was based on the pioneering work of Idriss and Seed
(1968), and Seed and Idriss (1970) as employed in the widely used program SHAKE
(Schnabel, et al., 1972) and its more current version SHAKE91 (Idriss and Sun, 1992).
o The new pressure dependent hyperbolic model introduced by Park and Hashash (2001)
is employed in nonlinear analysis. This model extended the hyperbolic model
introduced by Matasovic (1992) and was employed in the nonlinear site response code
D-MOD, which was in turn a modification of the Konder and Zelasko (1963)
hyperbolic model. The hyperbolic model had been employed with Masing criteria
earlier in the program DESRA by Lee and Finn (1975, 1978). The hyperbolic model
was originally proposed by Duncan and Chang (1970), with numerous modifications
in other works such as Hardin and Drnevich (1972) and Finn et al. (1977).
• DEEPSOIL v2.0-2.6:

o Full and extended Rayleigh damping is introduced in DEEPSOIL (Hashash and Park,
2002; Park and Hashash, 2004) with a user interface. This was in part based on Clough
and Penzien (1993) and the findings of Hudson et al. (1994) as implemented in the
program QUAD4-M. Additional developments and modifications are made in
DEEPSOIL benefited greatly from the PEER lifeline project “Benchmarking of
Nonlinear Geotechnical Ground Response Analysis Procedures (PEER 2G02)”.
• DEEPSOIL v3.0-3.7: Additional enhancements are made to the user interface as well as
inclusion of pore water pressure generation/dissipation capability.
o Current pore water pressure models employed include the same model introduced by
Matasovic (1992), Matasovic and Vucetic (1993, 1995) and employed in the program
D_MOD.
o The current dissipation model used in DEEPSOIL is derived from FDM considerations.
• DEEPSOIL v3.5: A new soil constitutive model is introduced to allow for significantly
enhanced matching of both the target modulus reduction and damping curves (Phillips and
Hashash, 2008).
o A new functionality in the user interface is implemented that allows the user to
automatically generate hyperbolic model parameters using a variety of methods
(Phillips and Hashash, 2008).
o DEEPSOIL v3.7: A new pore water pressure generation model for sands is added – the
GMP Model (Green et al., 2000), in addition to various improvements in the user
interface, as well as the capability to export output data to a Microsoft Excel file.
• DEEPSOIL v4.0: Complete rewrite of DEEPSOIL user interface.
o DEEPSOIL was made multi-core aware, leading to much faster completion of batch-
mode analyses.
o An update manager was added to notify the user when updated versions of DEEPSOIL
were available.
o A motion processor and a PEER motion converter were added.
• DEEPSOIL v5.0: Updates of DEEPSOIL user interface and computational engine.
o Introduced a new dynamic properties window with significant usability enhancements.
o First version of DEEPSOIL to natively support 64-bit Windows, enabling faster
analyses and the ability to use very long motions.

• DEEPSOIL v6.0: Complete rewrite of DEEPSOIL computational engine and user
interface from the ground up resulting in significantly faster software. Numerous new
capabilities are introduced. A new analysis workflow is introduced.
• DEEPSOIL v6.1: The GQ/H nonlinear model is added to DEEPSOIL, allowing the user
to specify soil strength in a Generalized Hyperbolic Model.
• DEEPSOIL v7.0: The last version of DEEPSOIL overhauls the program workflow and
adds new capabilities. The main updates in version 7.0 include:
o Automatic subdivision of input soil profile, to achieve proper discretization based on
maximum target propagated frequency.
o Ability to perform layer thickness, shear wave velocity and dynamic curve
randomizations for soil profile.
o Output provided in relational database format, which allows the user to retrieve large
data sets efficiently.
o Overhauled user interface.
o Multiple user language selection.
o Simultaneous view/plot of multiple motions.
1.4 Program Installation
Installing DEEPSOIL Using Setup:
• System Setup:
DEEPSOIL uses “.” as the symbol for the decimal. For most users outside the USA please
change "," to "." for the decimal mark in your system when using DEEPSOIL.
• Hardware Requirements:
2 GHz or faster processor*
2 GB or more available RAM
250 MB available on hard drive for installation
Parallel analyses require a multi-core processor
• Software Requirements:
Windows 7 or later

Microsoft .NET Framework 4.5.2 or later
Administrator privileges are required for installation
• Installation:
Run “DEEPSOIL.msi”
The DEEPSOIL installer will automatically detect if your system supports 64-bit
installations and install the appropriate libraries.

2 Program Organization
The DEEPSOIL graphical user interface is composed of several steps to guide the user throughout
the site response analysis process as illustrated in the Navigation box shown in Figure 2.1,
presented to the user upon starting DEEPSOIL.
(a) (b) (c)
Figure 2.1. DEEPSOIL Main Window and Key Tabs as (a) Analysis Tab, (b) Motions Tab, and (c) Profiles
Tab
At the top left, the user has the option of choosing the “Analysis,” “Motions,” or “Profiles” tab.
These tabs are discussed in the following section.
Figure 2.2 shows the Options window. This window can be accessed by clicking on the Options
menu. The window allows the user to set the default working directory, the directory containing

the input motions for use in the analyses, the default directory in which profiles are saved, the
default units and languages, enable or disable multi-core support, choose the graph color and the
line thickness.
Figure 2.2. DEEPSOIL Options Window.
2.1 Profiles Tab
Saved profiles are shown in this tab. The user can directly select a profile and start a new analysis
or modify a saved analysis file.
2.2 Motions Tab
DEEPSOIL contains a Motions tab which can be used to view/process input motions. To
view/process a motion, simply click on the check box for the related motion. DEEPSOIL will

generate acceleration, velocity, displacement and Arias intensity time histories, as well as the
response spectrum and Fourier amplitude spectrum for the selected motion (Figure 2.3). In
DEEPSOIL V7.0, multiple motions can be viewed at the same time (Figure 2.3). Once the new
motion is selected, the color of the plot for the previous motion becomes gray and the new motion
is plotted with blue color. The relative size of the plots can be adjusted by clicking on the gray
vertical line and dragging it to the left or right. Additionally, zoom in and out options for each axis
to arrange the axis limits are available for each plot, such that scroll bar appears once the user
moves the cursor on each axis, and the axis limits can be arranged using this scroll bar. Figure 2.3
shows that linear and log scales on the axes of Fourier Amplitude Spectrum and Response
Spectrum can be selected through the pop-up menu via right-click.
Figure 2.3. Motion Viewer (right-click choices on Plots)
The ground motion parameters (PGA, PGV and PGD), timestep, Nyquist frequency and number
of points of the selected ground motion can be viewed under Motion Metrics and Tools at the
bottom of window (Figure 2.4). The pop-up menu of “Process” under Motion Metrics and Tools
provides the option of (i) kappa estimation, (ii) baseline correction, (iii) timestep reduction, (iv)

tripartite graph view, and (iv) single motion view. The selected motion can be monitored in a
separate window using Single Motion View (Figure 2.5). This window also provides to the user
the option to linearly scale the selected input motion under Time History Plots tab. The user is
provided two options for scaling: scale the original motion by a specified factor (scale by) or scale
the original motion to a specified maximum acceleration (scale to). The desired method can be
selected using the drop-down list in the upper right corner of the window. Click on the Apply
button to scale the motion and recalculate the other data. After scaling, the user can save the new
motion by pressing the Save As button. In Spectral Plots tab, Log-triangle and Rectangular with
user specified window width FAS smoothing types are provided, and Response Spectrum can be
calculated with five different methods: (i) Frequency Domain, (ii) Duhamel Integral, (iii) Duhamel
Integral with Timestep Correction (zero-padded in frequency domain), (iv) Newmark Beta
Method, and (v) Newmark Beta Method with Timestep Correction (zero-padded in frequency
domain). Fourier amplitude and phase angle plots are presented in Fourier Plots tab. The
calculated data is also provided in data tables which can be accessed by selecting the Time History
Data or Spectral Data tabs on the top of the window.
Figure 2.4. Motion Viewer (Motion Metrics and Tools)

Figure 2.5 Single Motion View
2.2.1 Baseline Correction
Similarly to the motion viewer, the baseline correction can be used by selecting a motion in the
list and clicking on the “Baseline Correction” option under Motion Metrics and Tools.
DEEPSOIL can perform baseline correction for any input motion (Figure 2.6). By selecting an
input motion and clicking on the Baseline Correction option, a new window appears, which
shows the acceleration, velocity, and displacement time-histories corresponding to this motion.
Motions which exhibit non-zero displacement time-histories for the latter part of the motion should
be corrected. The corrected time-histories are also calculated and presented to the user. The
response spectra and Fourier amplitude spectra for the original motion and the baseline-corrected

motion are also provided for the user. The spectra should be carefully examined by the user to
ensure that the baseline correction process did not greatly alter the input motion. The baseline-
corrected motion can then be stored as a file defined by the user. The relative size of the plots can
be adjusted by clicking on the gray vertical line and dragging it to the left or right. Dragging to the
left causes the response spectra and Fourier amplitude spectra plots to increase in size, while
dragging to the right causes the time-histories plots to increase in size, respectively.
The baseline correction routine in DEEPSOIL is adapted from the baseline correction routine
included in the USGS motion processing program BAP (USGS Open File Report 92-296A). The
baseline correction is accomplished using the following steps:
1. Truncate both ends of the motion using the first and last zero-crossings as bounds.
2. Pad the motion with zeros at both ends.
3. Process the motion with a second order, recursive, high-pass (0.1 Hz cutoff frequency)
Butterworth filter with convolution in both directions in the time domain.
4. Truncate the new motion using the last zero-crossing as bound.
Figure 2.6. Baseline Correction.

2.2.2 Response Spectra Calculation Methods
The frequency-domain solution, the Newmark β method and Duhamel integral solutions are the
three most common methods employed to estimate the response of Single Degree of Freedom
(SDOF) systems and therefore to calculate the response spectra. A brief description is presented
for each method to calculate the response of SDOF systems and to solve the dynamic equilibrium
equation defined as (Chopra, 1995; Newmark, 1959):
𝑚𝑢̈ + 𝑐𝑢̇ + 𝑘𝑢 = −𝑚𝑢̈ 𝑔 (2.1)
where m, c and k are the mass, the viscous damping and the system stiffness of the SDOF system
respectively. 𝑢̈ , 𝑢̇ and 𝑢 are the nodal relative accelerations, relative velocities and relative
displacements respectively and 𝑢̈ 𝑔 is the exciting acceleration at the base of the SDOF.
2.2.2.1 Frequency-domain solution

In the frequency-domain solution, the Fourier Amplitude Spectra (FAS) input motion is modified
by a transfer function defined as:
−𝑓𝑛2
𝐻(𝑓) = 2 (2.2)
(𝑓 − 𝑓𝑛2 ) − 2𝑖𝜉𝑓𝑓𝑛
1
where fn is the natural frequency of the oscillator calculated as 𝑓𝑛 = 2𝜋 √𝑘⁄𝑚 and 𝜉 is the damping
𝑐
ratio calculated as 𝜉 = 2√𝑘𝑚. Use of the frequency-domain solution requires FFTs (Fast Fourier
Transforms) to move between the frequency-domain, where the oscillator transfer function is
applied, and the time-domain, where the peak oscillator response is estimated. Over the frequency
range of the ground motion, the frequency-domain solution is exact.

2.2.2.2 Duhamel integral solution
The second method to compute the response of linear SDOF systems interpolates –commonly
assuming linear interpolation– the excitation function (−𝑚𝑢̈ 𝑔 ) and solves the equation of motion
as the addition of the exact solution for three different parts: (a) free-vibration due to initial
displacement and velocity conditions, (b) a response step force (−𝑚𝑢̈ 𝑔 ) with zero initial
𝑖
conditions and (c) response of the ramp force [− 𝑚 (𝑢̈ 𝑔 − 𝑢̈ 𝑔 )⁄𝛥𝑡 ]. The solution in terms of
𝑖+1 𝑖
velocities and displacements is presented in the following equations:
𝑢̇ 𝑖+1 = 𝐴′𝑢𝑖 + 𝐵′𝑢̇ 𝑖 + 𝐶′(−𝑚𝑢̈ 𝑔 ) + 𝐷′(−𝑚𝑢̈ 𝑔 ) (2.3)

𝑖 𝑖+1
𝑢𝑖+1 = 𝐴𝑢𝑖 + 𝐵𝑢̇ 𝑖 + 𝐶 (−𝑚𝑢̈ 𝑔 ) + 𝐷 (−𝑚𝑢̈ 𝑔 ) (2.4)

𝑖 𝑖+1
where:
𝜉
𝐴 = 𝑒 −𝜉𝜔𝑛Δ𝑡 ( 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡) + 𝑐𝑜𝑠(𝜔𝐷 Δ𝑡)) (2.5)
√1 − 𝜉 2
1
𝐵 = 𝑒 −𝜉𝜔𝑛Δ𝑡 ( 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡)) (2.6)
𝜔𝐷
1 2𝜉 1 − 2𝜉 2 𝜉 2𝜉
𝐶= { + 𝑒 −𝜉𝜔𝑛 Δ𝑡 [( − ) 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡) − (1 + ) 𝑐𝑜𝑠(𝜔𝐷 Δ𝑡)]} (2.7)
𝑘 𝜔𝑛 Δ𝑡 𝜔𝐷 Δ𝑡 √1 − 𝜉 2 𝜔𝑛 Δ𝑡
1 2𝜉 −𝜉𝜔𝑛 Δ𝑡
2𝜉 2 − 1 2𝜉
𝐷 = [1 − +𝑒 ( 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡) + 𝑐𝑜𝑠(𝜔𝐷 Δ𝑡))] (2.8)
𝑘 𝜔𝑛 Δ𝑡 𝜔𝐷 Δ𝑡 𝜔𝑛 Δ𝑡
𝜔𝑛
𝐴′ = −𝑒 −𝜉𝜔𝑛Δ𝑡 ( 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡)) (2.9)
√1 − 𝜉 2
𝜉
𝐵 ′ = −𝑒 −𝜉𝜔𝑛Δ𝑡 (𝑐𝑜𝑠(𝜔𝐷 Δ𝑡) − 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡)) (2.10)
√1 − 𝜉 2

1 1 𝜔𝑛 𝜉 1 (2.11)
𝐶′ = {− + 𝑒 −𝜉𝜔𝑛 Δ𝑡 [( + ) 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡) + 𝑐𝑜𝑠(𝜔𝐷 Δ𝑡)]}
𝑘 Δ𝑡 √1 − 𝜉 2 Δ𝑡√1 − 𝜉 2 Δ𝑡
1 𝜉
𝐷′ = [1 − 𝑒 −𝜉𝜔𝑛Δ𝑡 ( 𝑠𝑖𝑛(𝜔𝐷 Δ𝑡) + 𝑐𝑜𝑠(𝜔𝐷 Δ𝑡))] (2.12)
𝑘Δ𝑡 √1 − 𝜉 2
2.2.2.3 Newmark β time integration method in time-domain SDOF analysis

The third method is the Newmark β method. The Newmark β method calculates the nodal relative
velocity 𝑢̇ 𝑖+1and 𝑢𝑖+1 displacements at a time i+1 by using the following equations:
𝑢̇ 𝑖+1 = 𝑢̇ 𝑖 + [(1 − 𝛾)Δ𝑡]𝑢̈ 𝑖 + (𝛾Δ𝑡)𝑢̈ 𝑖+1 (2.13)
𝑢𝑖+1 = 𝑢𝑖 + (Δ𝑡)𝑢̇ 𝑖 + [(0.5 − 𝛽)(Δ𝑡)2 ] 𝑢̈ 𝑖 + [𝛽(Δ𝑡)2 ]𝑢̈ 𝑖+1 (2.14)
The parameters β and γ define the assumption of the acceleration variation over a time step (Δt)
and determine the stability and accuracy of the integration of the method. A unique characteristic
of the assumption of average acceleration (β = 0.5 and γ = 0.25) is that the integration is
unconditionally stable for any Δt with no numerical damping. For this reason, the Newmark β
method with average acceleration is commonly used to model the dynamic response of single and
multiple degree of freedom systems.
The Newmark β method has inherent numerical errors associated with the time step of the input
motion (Chopra, 1995; Mugan and Hulbe, 2001). These errors generate inaccuracy in the solution
resulting in miss-prediction of the high-frequency response. To determine if a motion’s time step
is too large to be used directly, the response spectrum calculated with the Newmark β method can
be compared with the response spectra calculated by other means, with and without a time step
correction in the motion viewer/processor (see section 2.2).
2.2.3 Fourier Amplitude Spectrum Calculation and Smoothing Operation
One of the most important factors to consider when evaluating ground motions is frequency
content. The most common measure of frequency content is the Fourier amplitude spectrum, which

indicates how the amplitude of the ground motion is distributed across different frequencies.
Calculation of the spectrum requires a transformation of the ground motion from the time domain
to the frequency domain. This transformation is called a Fourier transform. In DEEPSOIL, the
transformation is completed using a Fast Fourier Transform (FFT). The resulting Fourier spectrum
is then used to calculate the Fourier amplitude spectrum using the following equations:
i
fi = (2.15)
time step ∗ n
|F|i = √(real(Ci ))2 + (imag(Ci ))2 ∗ time step (2.16)
where fi is the i-th frequency, n is the number of points in the FFT, |F|i is the Fourier amplitude at
the i-th frequency, and Ci is the i-th amplitude and phase (in complex number representation) of
the FFT. The maximum frequency that can be contained in the motion is dictated by the motion’s
time step. This maximum frequency is called the Nyquest frequency and is calculated using the
following equation:
1
fNyquest = (2.17)
2 ∗ time step
DEEPSOIL can also smooth the calculated Fourier amplitude spectrum to make interpretation
easier by providing a clearer view of the overall frequency content. DEEPSOIL uses a triangle
smoother in log space (also called a log-triangle smoother). The smoothing routine in DEEPSOIL
uses a sliding triangular smoothing window in log-space and is adapted from a routine developed
by David Boore. The weights assigned to each point are based on the log distance from the point
of interest. Currently, the maximum smoothing width is set to 0.2. At each frequency of the
spectrum the weights of the smoothing window are calculated as follows:
• For frequencies below the current frequency:
log10 (i ⁄ lower bound index)
Wi = (2.18)
log10 (current index ⁄ lower bound index)
• For the current frequency:
Wi = 1 (2.19)
• For frequencies above the current frequency:

log10 (i ⁄ current index)
Wi = 1 − (2.20)
log10 (upper bound index ⁄ current index)

where the upper and lower bound indices are determined using the desired window width and
index of the current frequency.
2.2.4 Arias Intensity
The Arias intensity provides a measure of the intensity of the motion as a function of acceleration.
It is plotted as a function of time and is calculated using the following equation:
𝑡
𝜋
𝐼𝑎 (𝑡) = ∫[𝑎(𝑡)]2 𝑑𝑡 (2.21)
2𝑔
0
where g is the acceleration due to gravity and a(t) is the acceleration time history.
2.2.5 Significant Duration
The significant duration is defined as the timespan (in seconds) between the occurrence of 5% and
95% of the total Arias Intensity (section 2.2.4). The significant duration, and its location in the
motion time histories, can be shown by checking the box at the lower left of the motion viewer.
2.2.6 Housner Intensity
The Housner intensity (also referred as spectral intensity) provides a measure of the intensity of
the motion as a function of spectral velocity. It is plotted as a function of time. The Duhamel
integral method is used in calculation of the acceleration response spectra for computational
efficiency, and converted to velocity spectra by multiplying the spectra by the corresponding
angular frequency. The Housner intensity is often reported as a single value, however, DEEPSOIL
is able to provide the Housner intensity as a time-history by calculating the response spectrum at
each point of an acceleration record. The Housner intensity is calculated using the following
equation:
𝑡 2.5
𝐼ℎ (𝑡) = ∑ ∫ 𝑆𝑣 (𝑇, 𝜉) (2.22)

0 𝑇=0.1
where T is the period and ξ is the damping ratio. In DEEPSOIL, the Housner intensity is calculated
assuming a damping ratio of 5%.

2.2.7 Estimation of Kappa (κ)
DEEPSOIL includes a tool to aid in the estimation of the high-frequency attenuation parameter κ.
This tool is accessed by clicking on the Estimate Kappa under Process of Motion Metrics and
Tools on the motion processor window. To estimate κ, the user defines two bounding frequencies.
DEEPSOIL will then average the Fourier amplitude spectrum (as described in section 2.2.3) and
perform a linear regression over the range of frequencies chosen by the user. The plot is then
updated to reflect the chosen range of frequencies and the resulting κ and amplitude intercept.
The user can also plot a fixed κ value. The resulting line can be moved vertically by specifying an
amplitude intercept.
Once a line of constant κ is plotted (either by estimation or user-specification), it can be
interactively positioned vertically using the scroll-wheel on the mouse. The user can also
show/hide the averaged Fourier amplitude spectrum and plot a legend by right-clicking on the plot.
Figure 2.7 Kappa Estimator

2.2.8 Adding New Input Motions
Motions may be added to DEEPSOIL by clicking File Menu to select New and them Motion. This
tool is designed to convert motions from the PEER “.AT2” format to the DEEPSOIL format,
through a fully automated process. DEEPSOIL will read through the PEER file and determine the
number of data points and the time step. Additional options are provided for reading non-PEER
motions and should be set as needed. If DEEPSOIL cannot complete the conversion, a message
box is used to notify the user of the failure. Upon successful conversion, the user is notified by a
message box and the motion is added to the Motion Library.
Motions can also be added manually. This is done by using a text editor capable of producing .TXT
files. To add an input motion, enter the necessary data in the format described below and save as
a .TXT file in the “Input Motion” directory. The default input motion directory is: C:\Users\[User
Name]\Documents\DEEPSOIL\Input Motions\. If the user has specified a different directory, the
input motion file should be placed in the user-specified directory. The motion will then appear in
the motion list under Motions tab. The added motion should have the properties as:
• Units of the ground motion should be seconds and g’s.
• The format should be as follows:
o 1st row: Number of data points & time step (separated by 1 space)
o 2nd and subsequent rows: time & acceleration (separated by 1 space)
2.3 Analysis Tab
The analysis tab options are discussed in detail in the next section.

3 Analysis Flow
3.1 Analysis Definition: Step 1 of 5
The first step in the analysis requires the selection of analysis type. Figure 3.1 illustrates the form
for Step 1.
In the introductory tab (Analysis Definition), the user is required to choose the Analysis Method,
the Solution Type (Frequency of Time domain), the Default Soil Model for all newly generated
layers and the Default Hysteretic Re/Unloading Formulation for the analysis of DEEPSOIL. In
addition, the user can choose whether DEEPSOIL will automatically generate profiles for the
given input data (Automatic Profile Generation on/off), the Unit System (English or Metric) as
well as the type of Complementary Analyses that may be requested (Equivalent Linear-Frequency
Domain, Linear-Frequency Domain and Linear-Time Domain). Finally, under Analysis Tag, the
user can see the identifiers, which are IDs that are included in the analysis results to help users
identify the kind of soil model analysis that DEEPSOIL performed (See Table 3.2 for Soil Models
Descriptions).
Figure 3.1. Step 1/5: Choose type of analysis.

Note: Before creating a new profile, or opening an existing profile, it is recommended to verify
the Default Working Directory from the menu Options (Figure 3.2). If a different directory is
preferred, press the Change button to bring up a folder browser and select the preferred directory.
At the same window, the user can choose the Default Units (English or Metric), the Default
Language, the Multi Core, the Graph Colors and the choice to allow DEEPSOIL to collect
anonymous data to improve user experience.
Figure 3.2. Choose the Default Working Directory.
Under Analysis Method the following options are available

• Linear
• Equivalent Linear
• Nonlinear
Depending on the Analysis Method choice of the user, different Solution Type, Default Soil Model,
Default Hysteretic Re/Unloading Formulation choices may be available. The available
combinations for each Analysis Method are presented in tabular format in Table 3.1.

Table 3.1. Available analysis option in DEEPSOIL 7
Analysis Method Soil Model Hysteretic Pore Pressure
Re/Un-loading Options
Formulation
Linear Frequency
and Time - - -
domain
Equivalent Frequency ● General Quadratic/ ● Non-Masing
-
Linear domain Hyperbolic Model ● Masing
Nonlinear Time (GQ/H) ● Generate Excess
domain ● Pressure-Dependent Porewater Pressure
Modified Kondner ● Enable Dissipation
Zelasko (MKZ) ● Make top of
● Yee et al. (2013) Profile Permeable
● Discrete Points ● Make Bottom of
(Equivalent Linear) Profile Perm.
● User-Defined (UMAT)

Table 3.2. IDs of different Soil Models in DEEPSOIL – Soil Model Descriptions
ID Model Description
DS-FL0 Frequency Domain Linear
DS-EL0 Frequency Domain Equivalent Linear - Discrete Points
DS-EL1 Frequency Domain Equivalent Linear - MKZ with Masing Rules
DS-EL2 Frequency Domain Equivalent Linear - MKZ with Non-Masing Rules
DS-EL3 Frequency Domain Equivalent Linear - GQ/H with Masing Rules
DS-EL4 Frequency Domain Equivalent Linear - GQ/H with Non-Masing Rules
DS-TL0 Time Domain Linear
DS-NL1 Time Domain Nonlinear - MKZ with Masing Rules
DS-NL2 Time Domain Nonlinear - MKZ with Non-Masing Rules
DS-NL3 Time Domain Nonlinear - GQ/H with Masing Rules
DS-NL4 Time Domain Nonlinear - GQ/H with Masing Rules
-PWP0 Porewater pressure generation without dissipation
-PWP1 Porewater pressure generation and dissipation - permeable halfspace
-PWP2 Porewater pressure generation and dissipation - impermeable halfspace
Note: To review the input parameters, you can select the Input Summary menu at any moment.
The Input Summary window (Figure 3.3) may be viewed any time after completing step 1. Note:
tabs will only appear after the corresponding parameters have been inputted.
Figure 3.3. Input Summary window.

3.1.1 Linear Analysis
A Linear Analysis (Lin) model can be solved in the following two ways:
-Frequency Domain
-Time Domain
In both linear site response analyses, the maximum soil stiffness and a constant damping ratio are
considered throughout the entire time history.
3.1.2 Equivalent Linear Analysis
The Equivalent Linear (EL) model employs an iterative procedure in the selection of the shear
modulus and damping ratio soil properties as pioneered in program SHAKE. These properties can
be defined by discrete points or by defining the soil parameters that define the backbone curve of
one of the nonlinear models.
The option of defining the soil curves using discrete points is only applicable for the Equivalent
Linear analysis. For this option, the G/Gmax and damping ratio (%) are defined as functions of
shear strain (%).
3.1.3 Deconvolution via Frequency Domain Analysis
This approach is the same as the frequency-domain equivalent linear analysis approaches except
that the input motion can be applied at the ground surface or anywhere else in the soil column.
The corresponding rock motion is then computed and provided to the user.
Deconvolution requires definition of a soil profile. The following properties need to be defined
for each layer:
• Thickness
• Shear Wave Velocity (𝑉𝑠) or Initial Shear Modulus (𝐺𝑚𝑎𝑥)
• Unit Weight
• Damping Ratio (%)
To perform the deconvolution,

1. Open or create an frequency domain profile (Linear or Equivalent Linear Analysis)
2. Enter the requested information into the table on Step 2a, as shown in Figure 3.4.

3. Additional layers may be added using the Add Layer button. Unwanted layers may
similarly be removed using the Remove Layer button.
4. Select the bottom layer and check the box labeled Deconvolution near the bottom of the
window
5. Specify the point of application of the ground motion by selecting the appropriate layer in
the drop-down list.
6. Use the circular buttons to select the type of ground motions to be generated as output.
7. Click Next to advance to Step 3 and select the locations for output and the motion(s) to be
deconvolved.
8. Click Next to advance to Step 5 and set the frequency-domain parameters.
9. Click Analyze.
Figure 3.4. Deconvolution analysis parameters.
The output from a deconvolution analysis is a set of DEEPSOIL-formatted motions. Regardless

of the output selection, there will be a file named “Deconvolved - [motion name].txt” that is the
motion at the top of rock (bottom of profile). Additional files will be produced for each layer
output requested and will be named “Deconvolved - [motion name] - layer [#].txt”. These files
can be used directly in DEEPSOIL.

Note: Deconvolution cannot be performed in the time domain analysis. Finding the motion at the
bottom of the soil profile given the motion at the ground surface is an inverse problem in nonlinear
analysis that is complex to solve and is not amenable to a simple deconvolution computation.
3.1.4 Non-Linear Analysis
Non-linear (NL) analysis solves the equations of motions in time domain using the Newmark β
method (implicit) or the Heun’s Method (explicit). Several soil models are available for users to
select from, summarized above and in Table 3.1. The analysis can be performed with or without
porewater pressure generation.
The user has the option of obtaining the site response results using the equivalent linear method
automatically whenever nonlinear site response analysis is conducted. It is highly recommended
that EL results are always examined whenever a NL analysis is conducted. This can be done by
checking the box labeled Equivalent Linear – Frequency Domain as it is shown in Figure 3.5.
Figure 3.5. Complementary Equivalent Linear-Frequency Domain analysis.
3.2 Soil Profile Definition: Step 2 of 5

The Soil Profile Definition window (Figure 3.6) consists of a visual display of the soil profile (Soil
Profile Plot), the Soil Profile Metrics section and either one of the two tabs: (a) Layer Properties
and (b)., Advanced Table View.

The user can define the properties of each layer of the soil profile (Thickness, Shear Wave Velocity
(VS) or Initial Shear Modulus (Gmax) and Unit Weight) using one or both the Layer Properties and
the Advanced Table View tabs.
Additional layers may be added using the Add Layer(s) button. Unwanted layers may similarly be
removed using the Remove Layer(s) button.
If the user selects to generate porewater pressure during the analysis (nonlinear analyses only),
additional parameters must be specified, including the model to be used and their respective
parameters. Each model and the required inputs are discussed in detail in Section 4.
The check box Water Table is used to choose the depth of the water table by clicking the drop-
down menu and selecting the layer that the water table will be above. The Advanced Table View
tab displays every layer beneath the water table by changing the background color to blue. The
location of the water table affects the calculations only when introducing the pressure dependent
soil parameters or performing an effective stress analysis. The location of the water table does not
influence the frequency domain solution.
The Layer Properties (Figure 3.7) tab is divided in five sections: i. the Current Soil Properties, ii.
the Reference Curve iii. the Curve Fitting iv. the Save Materials and v. Other Material Files. In
the right side of the window the plots of G/Gmax, Damping Ratio and Shear Strength vs Shear
Strain are shown. Single Element Test
The Previous Layer and Next Layer buttons on the top right corner of the window can be used to
select the layers. Alternatively, the user can double click on the layer he wants to modify from the
visual display on the left side of the window.

Figure 3.6. Soil Profile Definition – Advanced Table View window.
Figure 3.7. Layer Properties Tab.
The Advanced Table View (Figure 3.6) tab summarizes the input parameters of each layer along
with the generated information from the Layer Properties tab in a tabular format.

The user must specify the typical soil properties of each layer based on the type of analysis that
was selected (Linear, Nonlinear, etc). The input parameters for each soil model are discussed in
Chapter 4.
3.2.1 Single Element Test
Single Element Test option (Figure 3.8) is provided under Layer Properties Tab (Figure 3.7) in
order to test the soil model behavior for given strain path. Soil Model can be changed to any of
available options. Additionally, different damping models and pore water pressure options can be
selected to evaluate the soil hysteresis behavior. Soil backbone curve can be plotted on top of
hysteresis loop. Figure 3.8 shows the hysteresis behavior for soil layer for which MKZ soil model
and Masing type of damping model is adopted.
3.2.2 Maximum Frequency (for Time Domain Analysis only)

Upon completing the definition of the soil and model properties, the user is shown a plot of the
maximum frequency versus depth for each layer (Figure 3.9). A plot of maximum frequencies
(Hz) versus depths of all layers are displayed. The maximum frequency is the highest frequency
that the layer can propagate and is calculated as: fmax = VS/4H, where VS is the shear wave velocity
of the layer, and H is the layer thickness. To increase the fmax, the thickness of the layer should be
decreased. This check is performed solely for time domain analyses. It is recommended that the
layers have the same maximum frequency throughout the soil profile, though this is not required.
For all layers, the maximum frequency should generally be a minimum of 30 Hz.

Figure 3.8 Single Element Test Window
Figure 3.9. Soil Profile Plot.

3.2.3 Implied Strength Profile (Step 2)
Upon completing the definition of the soil and model properties, the user is shown a plot of the
implied strength of the soil profile. The window provides three plots for the user to view: implied
shear strength versus depth, normalized implied shear strength (shear strength divided by effective
vertical stress) versus depth, and implied friction angle versus depth (Figure 3.9). The shear
strength and friction angle are also provided in the table to the right for closer inspection. The
implied shear strength is calculated from the modulus reduction curves entered as part of step 2a.
At each point on the curve, the shear stress is calculated using the following equation:
𝐺
𝜏 = 𝜌𝑉𝑠2 𝛾 (3.1)
𝐺𝑜
where, τ is the shear stress at the given point, ρ is the mass density of the soil, VS is the shear wave
velocity in the given layer, G is the shear modulus at the given point, G0 is the shear modulus at
0% shear strain, γ is the shear strain at the given point.
The maximum value of shear stress for the given layer is then plotted at the depth corresponding
to that layer. Using this maximum value, the implied friction angle is then calculated using the
following equation:
𝜏𝑚𝑎𝑥
𝜙 = 𝑡𝑎𝑛−1 ( ) (3.2)
𝜎′𝑣
Where 𝜙 is the friction angle, 𝜏𝑚𝑎𝑥 is the maximum shear stress as calculated above, and 𝜎′𝑣 is
the effective vertical stress at the mid-depth of each layer.
The user is encouraged to carefully check the provided plots. If the implied strength or friction
angle of particular layer is deemed unreasonable, the user should consider modifying the modulus
reduction curve for the layer to provide a more realistic implied strength or friction angle.
3.2.4 Halfspace Definition (Bedrock)
As part of the Soil Profile Definition, the user must also define the rock / half-space properties of
the bottom of the profile. This can be done through the Layer Properties tab by double clicking on
the last bottom of the Soil Profile Plot at the left side of the window (Figure 3.10).
The user has the option of selecting either an Elastic Half-space or a Rigid Half-space. An
informational display (Information Regarding Rock Properties) explains that an elastic half-space

should be selected if an outcrop motion is being used and a rigid half-space should be selected if
a within motion is being used. If an elastic half-space is being used, the user must supply the shear
wave velocity (or modulus), unit weight, and damping ratio of the half-space. If a rigid half-space
is being used, no input parameters are required.
In general, the shear wave velocity of the bedrock should be greater than that of the overlying soil
profile. It should be noted that the bedrock damping ratio has no effect in time domain analyses
and only a negligible effect in frequency domain analyses regardless of the value specified by the
user.
Bedrock properties may be saved by giving the bedrock a name and pressing the Save Bedrock
button. The new bedrock will appear in the list of saved bedrocks below. To use a saved bedrock,
select the file from the list box and press the Load button.
If the analysis includes porewater pressure generation and dissipation with a permeable half-space,
the user is also given the option to specify the coefficient of consolidation Cv for the halfspace. If
no value is specified, DEEPSOIL will use the coefficient of consolidation Cv of the last layer for
the half-space as well. If the user is conducting a frequency domain analysis, deconvolution can
be performed rather than a forward analysis. Deconvolution is discussed in section 3.1.3.
Figure 3.10. Halfspace Definition – “Bedrock”.

3.3 Step 3 of 5: Input Motion Selection
The Input Motion Selection allows the user to select specify the input motion(s) to be used in the
analysis. The input motion(s) must be selected from the current input motion library (to which the
user may add additional motions, see section 2.2.8). The motions may be selected by checking the
appropriate checkbox in the second column of the window. All motions can be selected or
deselected by using the Select All button at the bottom of the motion list. Once a motion is selected,
DEEPSOIL will calculate and plot the acceleration, velocity, displacement, Arias intensity and
Housner intensity time histories as well as the Fourier amplitude spectrum and Peak spectral
acceleration (Figure 3.11). If multiple motions are selected, a single motion can be highlighted in
the plots by clicking on it either in the motion list or in its column in the table below the plots.
The user can utilize the dropdown tools menu at the bottom of the window (Figure 3.12) for each
motion in which one of the following options can be chosen: i. Estimate Kappa, ii. Baseline
Correction, iii. Timestep Reduction, iv. Show Tripartite Graph and v. Single Motion View. Finally,
the additional dropdown menu Spectral plots can be used to display: a. a smoothed FAS, b. the
Duhamel Integral, c. the Duhamel Integral with Timestep Correction (zero-padded in frequency
domain) d. the Newmark Beta Method and e. the Newmark Beta Method with Timestep Correction
(zero-padded in frequency domain).
Figure 3.11. Input Motion Selection.

3.4 Step 4 of 5: Viscous/Small-Strain Damping Definition
The Viscous/Small-Strain Damping Definition step appears only for time domain analyses and allows
the user to set the viscous damping formulation and select the optimum modes/frequencies for the
analysis (Figure 3.13). This window is unique to DEEPSOIL and it helps control the introduction
of numerical damping through frequency dependent nature of the viscous damping formulation.
Note that when multiple input motions are selected for an analysis, the viscous damping
formulation and the selected modes/frequencies are the same for all the selected input motions.
The following options must be specified:
• Damping Matrix Type:
o Frequency Independent (recommended)
o Rayleigh Damping
▪ 1 mode/freq.
▪ 2 modes/freq. (Rayleigh)
▪ 4 modes/freq. (Extended Rayleigh)
• Damping Matrix Update:
▪ Yes
▪ No
The user can also press the Plot Damping Curve to generate a plot of the Normalized Damping
Ratio. This option is available only when the Rayleigh Damping option is activated. Also, the user
can choose whether the damping matrix will be recalculated at each step of the analysis or not by
choosing the appropriate circular button in the Damping Matrix Update section. Finally, the user
can plot the Frequency Domain Solution and the Time Domain Solution for his motion of choice
using the corresponding buttons from the Linear Response Evaluation section. For more details on
this stage, please refer to Example 6 in the tutorial.
Viscous damping formulation is used to model small strain damping. The viscous damping
formulation results in frequency dependent damping and can introduce significant artificial
damping. It is therefore important to select an appropriate viscous damping formulation and
corresponding coefficients to reduce the numerical damping (Hashash and Park, 2002; Park and
Hashash, 2004). There are three types of Rayleigh damping formulations in DEEPSOIL, as listed
below. It is, however, recommended that the frequency independent damping formulation be
selected for most analyses.

When ready to proceed, click Next.
Figure 3.12. Dropdown tools menu.
Figure 3.13. Viscous/Small-Strain Damping Definition.

3.4.1 Frequency Independent Damping Formulation
This procedure solves for the eigenvalues and eigenvectors of the damping matrix and requires no
specification of modes or frequencies. This formulation removes many of the limitations of
Rayleigh Damping and does not greatly increase the required analysis time in most situations. A
complete explanation of the damping formulation is presented in Phillips and Hashash, 2009.
3.4.2 Rayleigh Damping formulation types
• Simplified Rayleigh Damping formulation (1 mode/frequency)
Uses one mode/frequency to define viscous damping.
• Full Rayleigh Damping formulation (2 modes/frequencies)
Uses two modes/frequencies to define viscous damping.
• Extended Rayleigh Damping formulation (4 modes)
Uses four modes/frequencies to define viscous damping.
A complete explanation of the extended Rayleigh damping formulation is presented in Park and
Hashash, 2004.
3.4.2.1 Modes/frequencies selection

There are two options available for selecting modes. The first option is choosing the natural modes
(e.g. 1st and 2nd modes). The second option is choosing the frequencies for Rayleigh damping
directly. The resulting Rayleigh damping curve can be displayed by pressing Show Rayleigh
Damping and the curve will be displayed at the right bottom window. Note again that the viscous
damping is frequency dependent. The goal in time domain analysis is to make the viscous damping
as constant as possible at significant frequencies.
3.4.2.2 Verification of the selected modes/frequencies

The time domain solution uses the frequency dependent Rayleigh damping formulation, whereas
actual viscous damping of soils is known to be fairly frequency independent. The frequency
domain solution uses frequency independent viscous damping. The appropriateness of the chosen
modes/frequencies should be therefore verified with the linear frequency domain solution.

Press Graph Lin. Freq. Domain. The results of the linear frequency domain solution (Frequency
ratio vs. Freq. and Response spectrum plots) will be displayed as blue curves. The goal is to choose
the appropriate modes/frequencies that compare well with the linear frequency domain solution.
Enter the desired modes/frequencies as input. Then press the Check with Lin. Time Domain button.
The results (in the same window as frequency domain solution) will be displayed as pink curves.
Choose the modes/frequencies that agree well with the linear frequency domain solution. This is
an iterative procedure and optimum modes/frequencies should be chosen by trial and error.
3.4.2.3 Damping Matrix Update
This option is only applicable for nonlinear solutions. During the excitation, soil stiffness and the
frequencies corresponding to the natural modes of the profile change at each time step. The natural
modes selected are recalculated at each time step to incorporate the change in stiffness and the
damping matrix is recalculated.
This feature is enabled by clicking the Update Matrix option in the Damping Matrix Update
selection. Note that using this feature may significantly increase the time required to complete an
analysis.
3.5 Step 5 of 5: Analysis Control Definition

In this stage of analysis, the user may specify options to be used for either the frequency domain
or time domain analysis as well as define the output settings (Figure 3.14).
Figure 3.14. Analysis Control Definition.

3.5.1 Frequency domain analysis
The options in a frequency domain analysis are:
• Number of Iterations
• Effective Shear Strain Definition
• Complex Shear Modulus Formulation
o Frequency Independent
o Frequency Dependent
o Simplified
3.5.1.1 Number of Iterations

Determines the number of iterations in performing an equivalent linear analysis. Check whether
the solution has converged and the selected iteration number is sufficient by clicking Check
Convergence tab after running the analysis.
3.5.1.2 Effective Shear Strain Definition

When performing an equivalent linear analysis, the effective strain needs to be defined. An
effective shear strain, calculated as a percentage of the maximum strain, is used to obtain new
estimates of shear modulus and damping ratio. The default and recommended value is 0.65 (65%).
The following equation relates this value to earthquake magnitude.
𝑀−1
𝑆𝑆𝑅 = (3.3)
10
3.5.1.3 Complex Shear Modulus

DEEPSOIL allows a choice among three types of complex shear modulus formulae in performing
frequency domain analysis:
• Frequency Independent Complex Shear Modulus (Kramer, 1996):

The frequency independent shear modulus results in frequency independent damping, and
is thus recommended to be used in the analysis. This is the same modulus used in
SHAKE91.

𝐺 ∗ = 𝐺 (1 + 𝑖2𝜉) (3.4)
• Frequency Dependent Complex Shear modulus (Udaka, 1975)

The frequency dependent shear modulus results in frequency dependent damping, and
should thus be used with caution.
𝐺 ∗ = 𝐺 (1 − 2𝜉 2 + 𝑖2√1 − 𝜉 2 (3.5)
• Simplified Complex Shear modulus (Kramer, 1996)

This is a simplified form of frequency independent shear modulus defined as:
𝐺 ∗ = 𝐺 (1 − 𝜉 2 + 𝑖2𝜉) (3.6)
3.5.2 Time domain analysis

For a time domain analysis, the options are:
• Step Control
o Flexible
o Fixed
• Maximum Strain Increment
• Number of Sub-Increments
The accuracy of the time domain solution depends on the time step selected. There are two options
in choosing the time step (Hashash and Park, 2001).
3.5.2.1 Flexible Step

A time increment is subdivided only if computed strains in the soil exceed a specified maximum
strain increment.
The procedure is the same as that for the Fixed Step above, except the Flexible option is chosen.
Type the desired Maximum Strain Increment into the text box. The default and recommended
value is 0.005 (%).
3.5.2.2 Fixed Step

Each time-step is divided into N equal sub-increments throughout the time series.

To choose this option:
• Click the option button labeled Fixed
• DEEPSOIL responds by disabling the text box labeled Maximum Strain Increment and
enabling Number of sub-increments
• Type the desired integer value of sub-increments into the text box
3.5.2.3 Integration Method

There are two available time integration methods:
• the Newmark β method (implicit) and
• the Heun’s Method (explicit).
3.5.2.4 Time-history Interpolation Method

This option is only available when the flexible step is selected. When subdividing a time step,
accelerations must be computed at intermediate points. DEEPSOIL implements two subdivision
strategies: 1) linear time-domain interpolation and 2) zero-padded frequency-domain interpolation.
Linear (time-domain) interpolation is the classical approach in which the change in acceleration is
simply divided into equal increments. This method has been shown to fundamentally alter the
motion by adding energy to the signal at frequencies above the Nyquest frequency of the original
signal. This can potentially add high frequency noise to the output signal.
Zero-padded frequency-domain interpolation is often referred to as “perfect interpolation” because
it allows for increased resolution (reduced time step) without adding energy above the Nyquist
frequency of the original signal. This means that the intermediate points are added to the signal in
a manner that is consistent with the actual behavior of the propagating wave. However, they are
not reported in the output and hence can cause a distortion in the output motion. Results from this
method should always be compared to the linear interpolation results.
3.5.2.5 Output Settings

The users can choose the layer(s) for which the results are presented. This can be done by checking
the appropriate checkbox in the first column of the window. There are four different choices: i.
Surface only, ii. All Layers iii. At Specific Depth and iv. At Specific Layers.

All layers can be selected using the All Layers button. Specific layers can be selected by clicking
at the appropriate boxes from the Want Output column of the table. It needs to be noticed that
requesting time-history output for additional layers will increase the time required for analyses to
complete. Therefore, it is recommended that the user only request time-history output for layers of
interest.
The user is also provided with the choice to generate an output displacement animation, by clicking
the box in the Displacement Animation section. As it is stated in the warning note, generating the
displacement animation will slow down the speed of the analysis.
After all the analysis parameters are completed, the user proceeds by clicking the Analyse button
in the right bottom of the window. An Analysis Running window will appear, showing the progress
(Figure 3.15).
Figure 3.15. Analysis Running.
3.6 Results
After the completion of the analysis, the following output for each selected layer will be directly
exported to a text file “Results - motion.txt” in the working directory specified using the Options
menu.
The Results window (Figure 3.16) consists of a visual display of the Motions and Layers selection
and the following tabs: a. Time History Plots, b. Stress-Strain Plots, c. Spectral Plots, d. Profile

Plots, e. Mobilized Strength, f. Displacement Animation, g. Response Spectra Summary and h.
Check Convergence.
If multiple motions were selected for analysis, the output can be found in the user’s working
directory in a folder named “Batch_Output”. Within this folder, there will be a folder
corresponding to each profile and within this folder there will be the folders of each of the motions,
that contain the results from each motion.
If a single motion was selected for analysis, the results can be found in the user’s working
directory.
3.6.1 Time History Plots tab

In the Time History Plots tab, the user can see the following plots: a. Acceleration, b. Velocity, c.
Displacement and d. Arias Intensity (see Figure 3.16).
3.6.2 Stress-Strain Plots tab

In the Stress-Strain Plots tab, the user can see the following plots: a. Shear Strain, b. Shear Stress
Ratio, c. Excess PWP (if applicable) and d. Strain Stress Ratio (Figure 3.17).
Figure 3.16. Results - Time History Plots.

Figure 3.17. Results – Stress Strain Plots.
3.6.3 Spectral Plots tab

In the Spectral Plots tab, the user can see the following plots: a. Peak Spectral Acceleration, b.
Fourier Amplitude and c. Fourier Amplitude Ratio (Figure 3.18).
Figure 3.18. Results – Spectral Plots.

3.6.4 Profile Plots tab
In the Profile Plots tab, the user can see the following plots: a. PGA, b. PGD, c. Max Strain, d.
Max Stress, e. Max PWP Ratio (if applicable) and f. Effective Vertical Stress (Figure 3.19).
Figure 3.19. Results – Profile Plots.
3.6.5 Mobilized Strength tab

In the Mobilized Strength tab, the user can see the following plots: a. Mobilized Shear Strength, b.
Normalized Shear Strength and c. Mobilized Friction Angle (Figure 3.20).

Figure 3.20. Results – Mobilized Strength.
Figure 3.21. Results – Displacement Animation.

3.6.6 Displacement Animation tab
In the Displacement Animation tab, the user can see an animation of the Displacement profile over
time (Figure 3.21). The user can modify the speed of the animation by adjusting the slide bar Speed
as well as start, stop and pick a time instance to be plotted using the Stop and Finish buttons and
by adjusting the slide bar Time, respectively. If multiple motions are selected, the user should pick
motion of choice for the animation using the dropdown menu at the top of the soil profile visual.
3.6.7 Response Spectra Summary tab

In the Response Spectra Summary tab, the user can see a plot of the Peak Spectral Acceleration
for the motion as well as the selected layer(s) (Figure 3.22). If multiple motions are selected, the
user should pick motion of choice for the animation using the dropdown menu at the top of the
soil profile visual.
Figure 3.22. Results – Response Spectra Summary.
3.6.8 Check Convergence tab

To view the convergence of the solution, click Check Convergence Tab This option enables
checking whether the solution has converged in an equivalent linear analysis. Plots of maximum

strain profiles for each iteration are displayed (Figure 3.23). To view the layersin the plots, check
Show Layers.
Figure 3.23 Check Convergence Tab
3.6.9 Output data file
Output data for each layer is automatically exported to “Results – motion.txt” in the user’s working
directory.
DEEPSOIL provides the user with the option to export the analysis results to a Microsoft Excel®
file or an LS-DYNA® file. This is done by clicking the Export to Excel or the Export to LS-DYNA
buttons respectively, located in the left bottom part of the results window. Note that this feature
requires that Microsoft Excel® or LS-DYNA® is installed on the system. Also, by clicking on the
Show results in folder view the user is directed to the results’ folder.

4 Soil Models
A variety of models are available for DEEPSOIL analyses. These models include: a) Equivalent
Linear, b) Hyperbolic (MR, MRD, DC), c) a Non-Masing Hyperbolic model (MRDF), and d)
Porewater Pressure Generation and Dissipation.
4.1 Backbone Curves
4.1.1 Hyperbolic / Pressure-Dependent Hyperbolic (MKZ)
DEEPSOIL incorporates the pressure-dependent hyperbolic model. The modified hyperbolic

model, developed by (Matasovic, 1993), is based on the hyperbolic model by (Konder and Zelasko,
1963), but adds two additional parameters Beta (β) and s that adjust the shape of the backbone
curve:
𝐺0 𝛾
𝜏=
𝛾 𝑠 (4.1)
1 + 𝛽 (𝛾 )
𝑟
where G0 = initial shear modulus,  = shear strength,  = shear strain. Beta, s, and  r are the model
parameters, respectively. There is no coupling between the confining pressure and shear stress.
DEEPSOIL extends the model to allow coupling by making  r confining pressure dependent as
follows (Hashash and Park, 2001):
𝑏
𝜎𝑣′
𝛾𝑟 = 𝑅𝑒𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑆𝑡𝑟𝑎𝑖𝑛 ( ) (4.2)
𝑅𝑒𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑆𝑡𝑟𝑒𝑠𝑠
where v’ is the effective vertical stress. Reference stress is the vertical effective stress at which
 r = Ref. stress. This model is termed as the “pressure-dependent hyperbolic model.”
The pressure-dependent modified hyperbolic model is almost linear at small strains and results in
zero hysteretic damping at small strains. Small strain damping has to be added separately to
simulate actual soil behavior which exhibits damping even at very small strains (Hashash and Park,
2001). The small strain damping is defined as
1 𝑑 (4.3)
𝜉 = 𝑆𝑚𝑎𝑙𝑙 𝑆𝑡𝑟𝑎𝑖𝑛 𝐷𝑎𝑚𝑝𝑖𝑛𝑔 𝑅𝑎𝑡𝑖𝑜 ( ′ )
𝜎𝑣
where d can be set to zero in case a pressure independent small strain damping is desired.
In summary, the parameters to be defined in addition to the layer properties are:

• Reference Strain
• Stress-strain curve parameter, Beta (β)
• Stress-strain curve parameter, s
• Pressure dependent (reference strain) parameter, b
• Reference Stress
• Pressure dependent (damping curve) parameter, d
4.1.2 Generalized Quadratic/Hyperbolic (GQ/H) Model with Shear Strength Control
(Darendeli 2001) study constructs the shear strength - shear strain curves based on the
experimentally obtained data. At small strains the data is collected using resonant column test, and
towards the medium shear strain levels the torsional shear test results are used. The values are
extrapolated at the large strain levels. This extrapolation may underestimate or overestimate the
shear strength at large strains. Therefore, shear strength correction is necessary to account for the
correct shear strength at large strains (Phillips and Hashash 2009). General Quadratic/ Hyperbolic
model proposed by (Groholski et al. 2016) has a curve fitting scheme that automatically corrects
the reference curves (such as Darendeli (2001)) based on the specified shear strength at the large
strains (the parameter τmax in the eq. (4.5)). The curve fitting parameters θ1 through θ5 (eq. (4.5))
are used to preserve the modulus reduction curves obtained from reference studies as much as
possible and modifies the large strain values based on the specified large strain shear strength.
The parameters τmax, and θ1 through θ5 are required to construct the shear strength corrected shear
strength - shear strain curves. Obtaining τmax is straightforward and user only needs to determine
the shear strength of the simulated soil material at large strains. The parameters θ1 through θ5 can
be obtained based on the reference study (Groholski et al. 2016). One easy way to obtain these
parameters is using DEEPSOIL (a 1-D nonlinear site response analysis software, (Hashash et al.
2016)). The user can create the layered domain in DEEPSOIL software and select the available
reference curve. Upon constructing the layered domain, GQ/H curve fitting routine calculates the
shear strength corrected shear strength - shear strain curve and provides the parameters θ1 through
θ5. These values can be directly used in soil hysteretic material without necessity to define any
reference shear strength - shear strain curve. The material model uses the τmax, G0, and θ1 through
θ5 to construct the shear strength - shear strain curve using the following functions:

𝛾
𝜃4 ∗ ( )𝜃5
𝛾𝑟
𝜃𝜏 = 𝜃1 + 𝜃2 ∗ 𝜃 𝛾 (4.4)
𝜃3 + 𝜃4 ∗ (𝛾 )𝜃5
5
𝑟
where, 𝛾𝑟 is the reference strain and is calculated as 𝛾𝑟 = 𝜏𝑚𝑎𝑥 /𝐺0 . Once the θτ is determined, the
shear strength - shear strain curve is constructed as follows:
1 𝛾 𝛾 2 𝛾
√
𝜏 = 𝜏𝑚𝑎𝑥 ∗ [ ∗ {1 + ( ) − {1 + } − 4 ∗ 𝜃𝜏 ∗ }] (4.5)
𝜃𝜏 𝛾𝑟 𝛾𝑟 𝛾𝑟
4.2 Hysteretic (Unload-Reload) Behavior
4.2.1 Masing Rules
When the user wishes to fit a soil curve (i.e. determine the model parameters which most closely
match the defined curves), the following options are available:
MR: Procedure to find the parameters that provide the best fit for the modulus reduction
curve with potentially significant mismatch of the damping curve.
MRD: Procedure to find the parameters that provide the best fit for both the modulus
reduction and damping curve.
DC: Procedure to find the parameters that provide the best fit for the damping curve
with potentially significant mismatch of the backbone curve.
4.2.2 Non-Masing Unload-Reload Rules
The non-Masing model included in DEEPSOIL is the MRDF Pressure-Dependent Hyperbolic

model (Phillips and Hashash, 2009). This model is implemented as a reduction factor which
effectively alters the Masing rules. By introducing the reduction factor, the modulus reduction and
damping curves can be fit simultaneously. The damping behavior is modified as:
𝜉𝑀𝑎𝑠𝑖𝑛𝑔𝐻𝑦𝑠𝑡𝑒𝑟𝑒𝑡𝑖𝑐 = 𝐹(𝛾𝑚𝑎𝑥 ) ∗ 𝜉𝑀𝑎𝑠𝑖𝑛𝑔 (4.6)

where F(γm) is the reduction factor calculated as a function of γm,, the maximum shear strain
experienced by the soil at any given time, and ξMasing is the hysteretic damping calculated using the
Masing rules, based on the modulus reduction curve. Two formulations for F(γm) are implemented
in DEEPSOIL and are discussed in the following sections.
4.2.2.1 MRDF-UIUC
The MRDF Pressure-Dependent Hyperbolic (Phillips and Hashash, 2009) model available in
DEEPSOIL allows the user to introduce a reduction factor into the hyperbolic model. The
reduction factor has the form:
𝐹(𝛾𝑚 ) = 𝑃1 − 𝑃2 (1 − 𝐺(𝛾𝑚 )⁄𝐺0 )𝑃3 (4.7)
where 𝛾𝑚 is the maximum shear strain experienced at any given time, 𝐺(𝛾𝑚 ) is the shear modulus
at 𝛾𝑚 , and P1, P2, and P3 are the fitting parameters.
By setting P1= 1 and P2= 0, the reduction factor is equal to 1 (regardless of the value of P3), and
the model is reduced to the Extended Masing criteria.
4.2.2.2 MRDF-Darendeli
The MRDF Pressure-Dependent Hyperbolic model (Phillips and Hashash, 2009) can also be used
with alternative formulations for the reduction factor. One alternative is the formulation proposed
by Darendeli, 2001. This formulation is an empirically-based modified hyperbolic model to predict
the nonlinear dynamic responses of different soil types. The developed model is implemented as a
reduction factor with the form:
𝐹(𝛾𝑚 ) = 𝑃1 (𝐺(𝛾𝑚 )⁄𝐺0 )𝑃2 (4.8)
where 𝛾𝑚 is the maximum shear strain experienced at any given time, 𝐺(𝛾𝑚 ) is the shear modulus
at 𝛾𝑚 , and P1 and P2 are the fitting parameters.
By setting P1= 1 and P2= 0, the reduction factor is equal to 1, and the model is reduced to the
Extended Masing criteria.

4.2.2.3 Non-Masing Unload-Reload Formulation
The hyperbolic / pressure-dependent hyperbolic unload-reload equation is modified with the
reduction factor, 𝐹(𝛾𝑚 ), as follows:
𝐺0 ((𝛾 − 𝛾𝑟𝑒𝑣 )⁄2) 𝐺0 (𝛾 − 𝛾𝑟𝑒𝑣 ) 𝐺0 (𝛾 − 𝛾𝑟𝑒𝑣 )

𝜏 = 𝐹(𝛾𝑚 ) [2 𝑠− 𝑠 ]+ + 𝜏𝑟𝑒𝑣 (4.9)
(𝛾 − 𝛾𝑟𝑒𝑣 ) 𝛾𝑚 𝛾𝑚 𝑠
1 + 𝛽 ( 2𝛾 ) 1 + 𝛽 (𝛾 ) 1 + 𝛽 (𝛾 )
𝑟 𝑟 𝑟
4.3 Porewater Pressure Generation & Dissipation
The following table summarizes the available excess pore water pressure generation models and
required parameters.
Table 4.1 Available Excess Pore Water Pressure Generation Models and Parameters
PWP Model Input Input Input Input Input Input Input

Soil Type Abbrev.
Model No: 1 2 3 4 5 6 7
Dobry &
Sand S-M/D 1 f p F s γtvp v -
Matasovic
Matasovic
Clay C-M 2 s r A B C D γtvp
& Vucetic
Dr
GMP Cohesioneless GMP 3 α FC(%) - - v -
(%)
Park & Ahn Sand P/A 4 α β Dru=1.0 CSRt - v -
Generalized Any G 5 α β - - - v -
Each model is described in the following sections. The user is referred to the original sources for
additional details.
4.3.1 Dobry/Matasovic Model for Sand
The Matasovic (1992) pore water pressure generation parameters must be determined by a curve-
fitting procedure of cyclic undrained lab-test data. Once the data is obtained, the following
equation, proposed by Matasovic and Vucetic (1993, 1995), can be used to determine the best-fit
parameters to be used in the analysis.

The excess pore water pressure is generated using the following equation:
𝑠
𝑝 ∗ 𝑓 ∗ 𝑁𝑐 ∗ 𝐹 ∗ (𝛾𝑐 − 𝛾𝑡𝑣𝑝 )
𝑢𝑁 = 𝑠 (4.10)
1 + 𝑓 ∗ 𝑁𝑐 ∗ 𝐹 ∗ (𝛾𝑐 − 𝛾𝑡𝑣𝑝 )
Table 4.2 Description of Dobry/Matasovic Model Parameters

VARIABLE DESCRIPTION
uN Normalized excess pore pressure (ru).
Neq Equivalent number of cycles.
γc The current reversal shear strain.
γtvp Threshold shear strain value.
p Curve fitting parameter.
s Curve fitting parameter.
F Curve fitting parameter.
f Dimensionality factor.
v Degradation parameter
4.3.1.1 Remarks:
The uN parameter is defined as the normalized excess pore water pressure ratio (ru = u’ / σv’).
Neq is the equivalent number of cycles calculated for the most recent strain reversal. For uniform
strain cycles, the equivalent number of cycles is the same as the number of loading cycles. For
irregular strain cycles, since the cycle number does not increase uniformly, Neq is calculated at
strain reversals using the uN obtained from the previous step and is then incremented by 0.5 for the
current step.
γtvp is the shear strain value below which reversals will not generate excess pore water pressure.
f is used to account for loading in multiple dimensions. f = 1 is used for 1D motion. f = 2 is used
for 2D motion. Note that assigning a value of f = 2 does not double the excess pore water pressure,
because f is included in both the numerator and denominator of the equation.

F, s, and p are the curve fitting parameters and can be obtained from laboratory tests.
The degradation parameter, v, is discussed in further detail in section 4.3.6
4.3.1.2 Suggested Values:

Carlton (2014) presents empirical correlations for the curve fitting parameters F and s for sands.
The best data fit is shown in Figure 4.1 and have the following functional forms:
(−1.55)
𝐹 = 3810 ∗ 𝑉𝑠 (4.11)
𝑠 = (𝐹𝐶 + 1)0.1252 (4.12)
where Vs is the shear wave velocity in m/s and FC is the percentage of fines content. The fit is
produced using the data from Table 4.3.
Table 4.3 shows that the values of p range within +-7.1% of 1 for different types and relative
densities of sands. For practical purposes, p = 1 is often assumed in the absence of laboratory data.

Figure 4.1 a) Carlton (2014), best fit correlating Vs (m/sec) to parameter F of Dobry pore water pressure
model for sands. b) Carlton (2014), best fit correlating FC (%) to parameter s of Dobry pore water pressure
model for sands

Table 4.3: Material Parameters for Low Plasticity Silts and Sands for the Matasovic and Vucetic (1993)
pore pressure generation model (From Carlton, 2014)
Pore Water Pressure Model Parameters
k
Material Reference γtv
(ft/sec) v f p F s
(%)
Warrenton, Oregon Silt
recovered from 130 to 248 ft
b.g.s; 73%<fines<99%;
Dickenson not
32.9%<water content<37.3%; 1 1 1 0.493 1.761 0.06
(2008) reported
86.3<γdry< 88.9 pcf;882<
Vs<1086 fps; OCR = 1.0; PI =
10, LL = 37
Stillaguamish River Silt,
Washington; recovered from
Anderson et not not
30 to 95 ft b.g.s; 60%< fines< 2 1.05 0.3 1.5 0.02
al. (2010) reported reported
90%; 600ft<Vs<900 ft/s; PI=8-
10; LL=31-32
Bangding Sand (BS); poorly-

graded commercially avaliable Dobry et al.
5.5x10-4 1 1 1 10.9 1 0.017
sane; Dγ＝40%；Dγ＝0.19； (1985)
Cc＝0.9；Cu＝1.4；γd,min =
90pcf;γd,max=106pcf
Wildlife Site Sand A(WSA);

Vucetic and
void ratio 0.84 to 0.85; 37% 9.8x10-4 1 2 1.04 2.6 1.7 0.02
Dobry (1988)
fines; N≈5;Vs≈350ft/s
Wildlife Site Sand B(WSB);
void ratio 0.74 to 0.76; 25% Vucetic and
6.6x10-3 1 2 1.04 2.6 1.7 0.02
fines; N≈6 to 13;Vs ≈ 450 to Dobry (1988)
500 ft/s
Heber Road Site Sand PB; void
Vucetic and
ratio 0.7; 15% fines; Vs ≈ 500 1.4x10-4 1 2 1.05 1.706 1.09 0.024
Dobry (1989)
to 600 ft/s
Heber Road Site Sand PB; void
Vucetic and
ratio 0.7; 22% fines; Vs ≈ 400 3.9x10-5 1 1 1.071 1.333 1.08 0.022
Dobry (1990)
to 466 ft/s

Santa Monica Beach
Sand(SMB); clean uniform
Matasovic
beach sand similar to Monterey 3.3x10-1 3.8 1 1 0.73 1 0.02
(1993)
No. 0; void ratio = 0.56; zero
fines; dense; Vs ≈ 867 ft/s
Owi Island Sand at depths
Thilakaratne
from 6 to 14 m b.g.s.; silty fine
and 6.6x10-3 1 2 1.005 3 1.8 0.025
sand placed as hydraulic fill;
Vucetic(1987)
18% <fines<35%
Owi Island Sand at depths of 6 Thilakaratne
m; placed as hydraulic fill; and 9.8x10-4 1 2 0.95 2.5 1.6 0.015
50% <fines<85% Vucetic(1988)
Mei et al. (2015) developed correlation for the curve fitting parameter F using 123 cyclic shear test
results compiled from literature. Two soil index properties, relative density (Dr) and uniformity
coefficient (Cu) are used in the correlation and it is applicable to sub-angular to sub-rounded clean
sands.
Figure 4.2 Proposed correlation to estimate curve-fitting parameter F (Mei et al. 2015)

4.3.2 Matasovic and Vucetic Model for Clays
Matasovic and Vucetic (1995) propose the following equation for the excess pore water pressure
generation for clays:
𝑟 𝑟 𝑟
𝑢𝑁 = 𝐴𝑁𝑐 −3𝑠(𝛾𝑐−𝛾𝑡𝑣𝑝 ) + 𝐵𝑁𝑐 −2𝑠(𝛾𝑐 −𝛾𝑡𝑣𝑝 ) + 𝐶𝑁𝑐 −𝑠(𝛾𝑐−𝛾𝑡𝑣𝑝 ) + 𝐷 (4.13)
Table 4.4 Description of Matasovic and Vucetic Model Parameters

uN Normalized excess pore pressure (ru)
Neq Equivalent number of cycles
γc The most recent reversal shear strain.
γtvp Threshold shear strain value.
r Curve fitting parameter.
s Curve fitting parameter.
A Curve fitting coefficients
B Curve fitting coefficients
C Curve fitting coefficients
D Curve fitting coefficients
4.3.2.1 Remarks:
The uN parameter is the same as in normalized excess pore water pressure ratio (ru = u’ / σv’)
Neq is the equivalent number of cycles calculated for the most recent strain reversal. For uniform
strain cycles, the equivalent number of cycles is the same as the number of loading cycles. For
irregular strain cycles, since the cycle number does not increase uniformly, and Neq is calculated
using the uN obtained from previous step and is then incremented by 0.5 for the current step.
γtvp is the threshold shear strain value below which reversals will not generate excess pore water
pressure.

Carlton (2014) presents empirical correlations for the curve fitting parameters s, r, A, B, C, and D.
Table 4.5 and Figure 4.3 (solid black lines) are used in the correlations. The parameter t in the
figure corresponds to s*(γc - γtvp)r. The empirical correlations take the following functional forms:
𝑠 = 1.6374 𝑥 𝑃𝐼 −0.802 𝑥 𝑂𝐶𝑅 −0.417 (4.14)
𝑟 = 0.7911 𝑥 𝑃𝐼 −0.113 𝑥 𝑂𝐶𝑅−0.147 (4.15)
7.6451 𝑓𝑜𝑟 𝑂𝐶𝑅 < 1.1 (4.16)

𝐴={
15.641 𝑥 𝑂𝐶𝑅−0.242 𝑓𝑜𝑟 𝑂𝐶𝑅 ≥ 1.1
−14.714 𝑓𝑜𝑟 𝑂𝐶𝑅 < 1.1 (4.17)
𝐵={
−33.691 𝑥 𝑂𝐶𝑅 −0.33 𝑓𝑜𝑟 𝑂𝐶𝑅 ≥ 1.1
6.38 𝑓𝑜𝑟 𝑂𝐶𝑅 < 1.1 (4.18)
𝐶={
21.45 𝑥 𝑂𝐶𝑅 −0.468 𝑓𝑜𝑟 𝑂𝐶𝑅 ≥ 1.1
0.6922 𝑓𝑜𝑟 𝑂𝐶𝑅 < 1.1 (4.19)
𝐷={
−3.4708 𝑥 𝑂𝐶𝑅 −0.857 𝑓𝑜𝑟 𝑂𝐶𝑅 ≥ 1.1
where OCR is the over consolidation ratio, and PI is the plasticity index.
Figure 4.3 Comparison of the curves given by Matasovic (1993) and Vucetic (1992) (solid black lines)
for t, for different values of PI and OCR and the correlations presented (dotted red lines). (Carlton, 2014)
Table 4.5 Material parameters for the Matasovic and Vucetic (1995) clay pore pressure generation model
(From Carlton, 2014)

Pore Water Pressure and Degradation Model
γtvp
Material Reference Parameter
(%)
s r A B C D
Marine
Matasovic
Clay
and Vucetic 0.1 0.075 0.495 7.6451 -14.7174 6.3800 0.6922
(OCR =
(1995)
1.0)
Marine Matasovic
Clay and Vucetic 0.1 0.064 0.520 14.62 -30.5124 18.4265 -2.5343
(OCR = (1995)
1.4)
Marine Matasovic
Clay and Vucetic 0.1 0.054 0.480 12.95 -26.3287 15.3736 -1.9944
(OCR = (1995)
2.0)
Marine Matasovic
Clay and Vucetic 0.1 0.042 0.423 11.263 -21.4595 11.2404 -1.0443
(OCR = (1995)
4.0)
4.3.3 GMP (Green, Mitcher and Polito) Model for Cohesionless Soil
The GMP model (Green et al. 2000) is an energy-based pore pressure generation model. The
excess pore pressure is calculated as follows:
𝑊𝑠
𝑟𝑢 = 𝛼 √ (4.20)
𝑃𝐸𝐶
Table 4.6 Description of GMP Model Parameters

ru Normalized excess pore pressure.
Ws Normalized dissipated energy per unit volume of soil.
PEC Pseudo energy capacity.
α Scale factor.
4.3.3.1 Remarks:
The dissipated energy, 𝑊𝑠 , is calculated as the area beneath the current stress-strain path and has
the following functional form:
𝑛
1
𝑊𝑠 = ′ ∑(𝜏𝑖+1 + 𝜏𝑖 ) ∗ (𝛾𝑖+1 − 𝛾𝑖 ) (4.21)
2𝜎0
𝑖=1
In DEEPSOIL, a scale factor “α” is introduced to allow for scaling of the generated excess pore
water pressure to match laboratory or field data.
The GMP model is a special case of the Berrill and Davis model (Berrill and Davis, 1985) that has
the form ru = α x Wsβ. In GMP model, α and β values are replaced by (1/PEC)0.5 and 0.5
respectively.
The degradation parameter is as described by Matasovic (1993) and uses the same functional form
as defined in the Matasovic model for sands (see section 4.3.6).

The determination of the 𝑃𝐸𝐶 calibration parameter can be conducted either via graphical
procedure or by use of an empirical relationship. The graphical procedure is described in detail by
Green et al. (2000). However, this causes an interruption in the analysis as it requires the
construction of the graphical procedure outside of the site response analysis software.
Polito et al. (2008) derived an empirical relationship between 𝑃𝐸𝐶, relative density (𝐷𝑟 ), and fines
content (FC) from a large database of laboratory data on non-plastic silt-sand mixtures, ranging
from clean sands to pure silts. The use of this empirical relationship allows the use of the GMP
model directly in the nonlinear site response analysis software by removing the need to find the
value of 𝑃𝐸𝐶 through graphical procedures. The empirical relationship is defined as:

𝐹𝐶 < 35%: 𝑒 0.0139𝐷𝑟 − 1.021
ln(𝑃𝐸𝐶) = { (4.22)
𝐹𝐶 ≥ 35%: − 0.597 ∗ 𝐹𝐶 0.312 + 𝑒 0.0139𝐷𝑟 − 1.021
4.3.4 Generalized Energy-based PWP Generation Model
This model allows for a user-defined excess pore water pressure generation model based on the
framework adopted from Berrill and Davis (1985) and Green et al. (2000). The model is energy-
based and the excess pore water pressure is calculated as follows:
β
𝑟𝑢 = α ∗ 𝑊𝑠 (4.23)
The model is a generalized form of GMP model, and uses the same general functional form
presented in the Berrill and Davis (1985) formulation. α and β are curve fitting parameters and can
be extracted from laboratory tests. Ws is the dissipated energy and is calculated using the
formulation defined in the GMP model.
Table 4.7 Description of Generalized Model Parameters

α Curve fitting coefficient
β Curve fitting parameter
Ws Normalized dissipated energy per unit volume of soil
4.3.5 Park and Ahn Model for Sand
The Park and Ahn (2013) model is a stress-based excess pore water pressure generation model that
uses the concept of a damage parameter to account for the accumulation of stress. The excess pore
water pressure is calculated as follows:
1
( )
2 𝐷 2𝛽
(4.24)
𝑟𝑢 = arcsin ( )
𝜋 𝐷𝑟𝑢=1.0

where the damage parameter D at each time step can be calculated as:
𝐷𝑖+1 = 𝐷𝑖 + Δ𝐷 (4.25)
Δ𝐷 = 2(𝐶𝑆𝑅𝑖+1 − 𝐶𝑆𝑅𝑡 )𝛼 (4.26)
In an incremental form, the generation model becomes:

1 1
2 𝐷𝑖+1 2𝛽 2 𝐷𝑖 2𝛽
𝑑𝑟𝑢 = (𝑟𝑢 )𝑖+1 − (𝑟𝑢 )𝑖 = 𝑎𝑟𝑐𝑠𝑖𝑛 [( ) ] − 𝑎𝑟𝑐𝑠𝑖𝑛 [( ) ] (4.27)
𝜋 𝐷𝑟𝑢=1.0 𝜋 𝐷𝑟𝑢=1.0
Table 4.8 Description of Park and Ahn Model Parameters

ru Normalized excess pore pressure.
D Damage parameter.
Dru = 1.0 Damage parameter at the initiation of liquefaction
CSRt Threshold shear stress ratio value.
α A calibration parameter
β Empirical constant

4.3.5.1 Remarks:
CSRt is the threshold shear stress ratio below which reversals will not generate excess pore water
pressure.

α is a calibration parameter that can be calculated using the following formulation with the CSR-
N curve obtained from laboratory tests:
𝑀−1
1
𝛼𝑎𝑣𝑒 = ∗ ∑ ( log(𝑁𝑖 /𝑁𝑖+1 )/(log(𝐶𝑆𝑅𝑖+1 − 𝐶𝑆𝑅𝑡 ) − log(𝐶𝑆𝑅𝑖 − 𝐶𝑆𝑅𝑡 )) (4.28)
𝑀
𝑖=1
where M is the total number of data points in CSR-N curve.
β is an empirical constant and for clean sands, a value of 0.7 is suggested.
Dru=1.0 is the value of damage parameter, D, at initiation of liquefaction and can be calculated from
CSR-N curves that are obtained from laboratory tests using the following formula:
𝐷𝑟𝑢=1.0 = 4𝑁(𝐶𝑆𝑅 − 𝐶𝑆𝑅𝑡 )𝛼 (4.29)
4.3.6 Porewater Pressure Degradation Parameters
Matasovic (1993) represents the degradation of the shear strength and shear stiffness of the soil
within the MKZ model by inclusion of two degradation indices. These degradation parameters
have also been implemented (and have similar effects) within the GQ/H model. The degradation
parameters are defined as:
𝛿𝐺 = √1 − 𝑢∗ (4.30)
𝛿𝜏 = 1 − (𝑢∗ )𝑣 (4.31)
Where 𝛿𝐺 is the shear modulus degradation function, 𝛿𝜏 is the shear stress degradation function, 𝑢∗
is the excess porewater pressure normalized by initial effective overburden stress, and 𝑣 is a curve-

fitting parameter to better model the degradation of shear strength with excess pore pressure
generation.
These degradation parameter formulations are implemented for all soil models except the
Matasovic and Vucetic model for clays. The degradation parameters for the Matasovic and Vucetic
model for clays are defined by Matasovic (1995) as:
𝛿𝐺 = 𝛿𝜏 = 𝑁 −1 (4.32)
Where 𝛿𝐺 is the shear modulus degradation function, 𝛿𝜏 is the shear stress degradation function,
and 𝑁 is the number of equivalent cycles.
4.3.7 Porewater Pressure Dissipation
The pore water pressure dissipation model is based on Terzaghi 1-D consolidation theory:
𝜕𝑢 𝜕 2𝑢
= 𝐶𝑣 ( 2 ) (4.33)
𝜕𝑡 𝜕𝑧
where Cv is the coefficient of consolidation.
Dissipation of the excess pore water pressure is assumed to occur in the vertical direction only.
Porewater pressure generation and dissipation occur simultaneously during ground shaking.

5 Randomization of Site Profile Properties
This section is under development.

6 Database Output Structure
DEEPSOIL versions until V7.0 are designed to provide site response analysis results as text file(s).
In the case of single analysis, one text file output is generated as including results of each selected
layer in separate tables of (i) acceleration, strain and stress ratio (shear stress/effective vertical
stress) time histories, (ii) spectral accelerations at 113 oscillator periods from 0.01 to 10.0s, (iii)
Fourier amplitude spectrum (FAS) and Fourier amplitude ratio computed as FASsurface/FASinput,
and (iv) calculated PGA, minimum and maximum displacements at the top of each layer, and max.
strain, max. stress ratio and effective vertical stress at the mid-depth of each layer. The results of
a batch analysis, in which one site profile can be exposed to several input motions, are stored in
subfolders with motion names, and a unique text file is created for each table given in one text file
for every single analysis.
After the introduction of randomization of site profile properties, as VS and dynamic curve
randomization in DEEPSOIL V7.0, users have the ability to run large numbers of analyses (such
as parametric studies) through the interface, and this necessitates handling large amount of output
with complicated structure. Thus, SQLite database structure has been introduced to store the
analysis results in DEEPSOIL V7.0. The transition from using text files for the analysis results to
database files happened for mainly two reasons: (i) databases can handle querying and indexing of
more sophisticated output structure, and (ii) significant reduction of output size (1.5 ≈ 2.0 times)
can be achieved.
The next section gives further details on the database output structure for DEEPSOIL V7.0.
6.1 Database Structure for Analyses Output
Figure 6.1 shows the database structure for DEEPSOIL V7.0 output. It is composed of mainly 6
components as:
(i) Fourier Amplitude Spectra (FAS):

FAS data includes the frequency array (Frequency), FAS of input motion
(Input_Motion_FAS), FAS of computed motions at selected layers (Layer#_FAS), and
ratio of FAS of computed motions at selected layers to that of input motion.

(ii) DEEPSOIL Input:
Input file created by the interface is provided as BLOB data in database output.
(iii) Profile Data:
Profile Data is composed of (a) total PGA (PGA_Total), relative PGA -which is defined
as the difference between the computed PGA and the PGA of bedrock motion
(PGA_Relative)-, minimum and maximum relative displacement (Min_DISP_Relative
& Max_DISP_Relative) at top depth (Depth_Layer_Top) of each layer, (b) initial
effective stress (Initial_Effective_Stress), maximum strain (Max_Strain), and
maximum stress ratio (Max_Stress_Ratio), which is the ratio of shear stress and
effective vertical stress at mid-depth (Depth_Layer_Mid) of each layer.
(iv) Response Spectra (RS):
Response spectrum of input motion (Input_Motion_RS) and calculated RS at selected
layers (Layer#_RS) are provided at 113 periods.
(v) Time Histories:
Calculated acceleration (Layer#_Accel), velocity (Layer#_Vel), displacement
(Layer#_Disp), Arias intensity (Layer#_Arias), strain (Layer#_Strain) and stress
(Layer#_Stress) time history are provided.
(vi) Velocity and Displacement Time Histories:
Velocity time history as relative (Layer#_Vel_Relative) and total (Layer#_Vel_Total),
and displacement time history as relative (Layer#_Disp_Relative) and total
(Layer#_Disp_Total) are provided under this table.

Output Structure
Fourier Amplitude DEEPSOIL Profile Response Time Vel&Disp

Spectra (FAS) Input Data Spectra (RS) Histories Time H.
• Frequency • Period • Time

• Input_Motion_FAS • Input_Motion_RS • Layer#_
• Layer#_FAS • Layer#_RS Vel_Tot
• Layer#_FAS_Ratio al
• Layer#_
Vel_Rel
• Depth_Layer_Top • Time
• PGA_Total • Layer#_Accel
• PGV_Relative • Layer#_Vel
• Min_DISP_Relative • Layer#_Disp
• Max_DISP_Relative • Layer#_Arias
• Depth_Layer_Mid • Layer#_Strain
• Initial_Effective_Stress • Layer#_Stress
• Max_Strain
• Max_Stress_Ratio
Figure 6.1 DEEPSOIL V7.0 Output Structure

7 Examples and Tutorials
The tutorial is intended to help users get familiar with DEEPSOIL. Six examples are prepared to
guide the users through the various features of DEEPSOIL. It is recommended that the examples
are followed in the order they appear. The example soil profiles and strain paths are stored in the
“Examples” folder under the default DEEPSOIL working directory. The motions for use with
example profiles are included under the default DEEPSOIL motion directory.
7.1 Example 1: Undamped Linear Analysis with Resonance
Two analyses included in Example 1 consist of linear analysis with frequency domain
(Example_1A) and time domain solutions (Example_1B), respectively, and selected input motions
with analyses properties are presented in Table 7.1.
Soil
Unit Weight= 20 kN/m3
Vs=500 m/s
20 m
Damping (%) = 0.0
Rigid Half space


Figure 7.1 Soil profile of Example 1
Table 7.1 Analyses properties and input motions for Example 1

Name Analysis Method Solution Type Analysis Tag Input Motion
Example_1A Linear Frequency Domain DS-FL0 ChiChi.txt
Example_1B Linear Time Domain DS-TL0 ChiChi.txt

The flow of analysis through DEEPSOIL V7.0 interface of these two examples are illustrated
below:
• The analysis definition for linear frequency-domain analysis (Example 1A) is given in
Figure 7.2. The solution type in Figure 7.2 should be changed to time-domain for time-
domain linear analysis (Example 1B).
• Figure 7.3 show Soil Profile Definition window to input basic soil properties (thickness,
unit weight, shear wave velocity, effective vertical stress) and soil model properties (small-
strain damping) for soil layer of interest
• Definition of bedrock as Rigid Half-space is presented in Figure 7.4
• After the definition of basic soil properties and small-strain damping for each soil layer,
profile plots as shear wave velocity, maximum frequency (fmax), small-strain damping can
be viewed as in Figure 7.5 along with implied shear strength, normalized implied shear
strength and implied friction angle.
• ChiChi motion is selected in Input Motion Selection window, where time-history
(Acceleration, velocity, displacement, Arias and Housner Intensity) and spectral plots (5%
damped spectral acceleration and Fourier Amplitude Spectrum (FAS) of selected motion
can be viewed, for two analyses in Figure 7.6.
• Viscous/Small-Strain Damping definition for time-domain analysis is presented in Figure
7.7. The recommended options as frequency-independent damping and no update on
damping matrix for Example 1B are selected.
• In Analysis Control Definition window, frequency-independent complex shear modulus
reduction is recommended for Example 1A (Figure 7.8). Flexible time-step control along
with maximum strain increment of 0.005 and time history interpolation method as zero-
padded in frequency-domain are adopted for Example 1B (Figure 7.9).
• Figure 7.10 and Figure 7.11 illustrate the computed time-history (acceleration, velocity,
displacement and arias intensity), and spectral plots (5% damped spectral acceleration,
FAS and FAS ratio) for Example 1A. Export output as Excel file and related Excel output
file are illustrated in Figure 7.12 and Figure 7.13, respectively. Calculated time-histories
and spectral plots for Example 1B are given in Figure 7.14 and Figure 7.15.

Figure 7.2 Step 1: Analysis Type Definition for Example 1A
Figure 7.3 Step 2: Soil Profile Definition for Example 1A and 1B

Figure 7.4 Step 2: Bedrock Profile Definition in Example 1A and 1B
Figure 7.5 Step 2: Soil profile plot for Example 1A and 1B

Figure 7.6 Step 3: Input motion selection window for Example 1A and 1B
Figure 7.7 Step 4: Viscous/Small-Strain Damping Definition for Example 1B

Figure 7.8 Step 5: Analysis Control Definition Window for Example 1A
Figure 7.9 Step 5: Analysis Control Definition for Example 1B

Figure 7.10 Results: Acceleration, Velocity, Displacement and Arias Intensity Time-Histories for Example
1A
Figure 7.11 Results: 5% Damped Spectral Acceleration, Fourier Amplitude Spectrum (FAS) and Fourier
Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 1A

Figure 7.12 Results: Exporting Analysis Results using Export to Excel Option
Figure 7.13 Results: Excel Output for Analysis Results for Example 1A

1B
Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 1B

7.2 Example 2: Undamped Linear Analysis with Elastic Bedrock
with analyses properties are presented in Table 7.2.
Soil
20 m
Vs=500 m/s
Damping (%) = 0.0
Elastic Half space

The flow of analysis for Example 2 is quite similar to that described in Example 1. Only difference
between Example 1 and 2 is bedrock definition, and Elastic Half-space is defined for this case
with:
• Shear wave velocity: 760 m/s

• Unit weight: 25.0 kN/m3
• Damping Ratio: 2.0 %

Figure 7.18 and Figure 7.19 illustrate the computed time-history (acceleration, velocity,
displacement and arias intensity), and spectral plots (5% damped spectral acceleration, FAS and
FAS ratio) for Example 2A. Calculated time-histories and spectral plots for Example 2B are given
in Figure 7.20 and Figure 7.21.

Figure 7.17 Step 2: Bedrock Profile Definition for Example 2A and 2B
Figure 7.18 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for Example 2A

Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 2A
Figure 7.20 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for Example 2B

Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Example 2B

7.3 Example 3: Damped Linear Analysis with Elastic Bedrock
with analyses properties are presented in Table 7.3. Computed time-histories (Acceleration,
Velocity, Displacement, and Arias Intensity) for Example 3A and Example 3B are presented in
Figure 7.23 and Figure 7.25. Response spectrum at first layer of site profile for Example 3A and
3B along with that of input motion are given in Figure 7.24 and Figure 7.26.
Soil
20 m
Vs=500 m/s
Damping (%) = 5.0
Elastic Half space


Figure 7.23 Result: Computed Time-Histories (Acceleration, velocity, displacement, and Arias Intensity)
for Layer 1 for Example 3A
Figure 7.24 Result: Computed 5% Damped Spectral Acceleration for Layer 1 along with that for Input
Motion for Example_3A

for Layer 1 for Example 3B
Motion for Example_3B

7.4 Example 4: Equivalent Linear Analysis with Discrete Points
One example included in Example 4 (Figure 7.27) consists of linear analysis with frequency
domain, and selected input motions with analyses properties are presented in Table 7.4. Computed
time-histories (Acceleration, Velocity, Displacement, and Arias Intensity) and response spectrum
at first layer of site profile along with that of input motion are presented in Figure 7.28 and Figure
7.29.
Soil
20 m Unit Weight= 20 kN/m3
Vs=500 m/s
Elastic Half space

Name Analysis Method Analysis Tag Input Motion
Example_4 Equivalent Linear DS-EL0 ChiChi.txt

for Layer 1 for Example 4
Motion for Example_3A

7.5 Example 5: Nonlinear Analysis (MKZ Soil Model with Masing Re/Unloading Behavior)
Three examples included in Example 5 consist of nonlinear site response analyses with different
types of fitting procedure as DC (Damping Curve Only), MR (Modulus Reduction Only), and
MRD (Modulus Reduction and Damping Curve) adopted for nonlinear dynamic curves (Table
7.5), and MKZ model with Masing type of re/unloading formation is used for dynamic curves
defined for each layer. Soil column is given in Figure 7.30. Selected input motions with analyses
properties are presented in Table 7.5. 5% Damped Spectral Acceleration and Fourier Amplitude
Spectrum computed at Ground Surface for Nonlinear and Equivalent-Linear Analyses using DC,
MR and MRD fitting procedures along with those for Input Motion are presented in Figure 7.31

Name Analysis Method Analysis Tag Fitting Type Input Motion
Example_5A Nonlinear DS-NL1 DC ChiChi.txt
Example_5B Nonlinear DS-NL1 MR ChiChi.txt
Example_5C Nonlinear DS-NL1 MRD ChiChi.txt

Soil 1
4m Unit Weight= 20 kN/m3
Vs=250 m/s
Soil 2
3
4m Unit Weight= 20 kN/m
Vs=250 m/s
Soil 3
3
Vs=250 m/s
Soil 4
4m 3
Unit Weight= 20 kN/m
Vs=250 m/s
Soil 5
4m 3
Vs=250 m/s
Elastic Half space



Figure 7.31 5% Damped Spectral Acceleration and Fourier Amplitude Spectrum computed at Ground
Surface for Nonlinear and Equivalent-Linear Analyses using DC, MR and MRD fitting procedures along
with those for Input Motion

7.6 Example 6: Nonlinear Analysis (GQ/H Soil Model with Non-Masing Re/Unloading Behavior)
Example 6 consists of nonlinear site response analysis along with supplementary equivalent-linear
analysis, and GQ/H model with Non-Masing type of re/unloading formulation is used for soil
dynamic curves. Selected input motion together with analyses properties are given in Table 7.6.
Soil 1
4m Unit Weight= 20 kN/m3
Vs=500 m/s
Soil 2
3
Vs=500 m/s
Soil 3
3
Vs=500 m/s
Soil 4
4m 3
Vs=500 m/s
Soil 5
4m 3
Vs=500 m/s
Elastic Half space



Example_6 Nonlinear Time Domain DS-NL2 ChiChi.txt
The flow of analysis for Example 6 can be illustrated as:
• In the Analysis Type Definition window, nonlinear analysis method is selected along with
General Quadratic/Hyperbolic Model (GQ/H) soil model and Non-Masing Re/Unloading
formulation. Additionally, equivalent-linear analysis is performed using the option under
Complementary Analyses. (Figure 7.33)
• The basic soil properties for each layer is defined as given in Figure 7.34 (Only layer 1 is
illustrated in this example). The use of GQ/H model necessitates to input the shear strength
of soil layer to represent the large-strain behavior of soil. The target shear strength of the
nonlinear shear modulus reduction (G/Gmax) is calculated using the Mohr-Coulomb
equation as:
𝜏𝑡𝑎𝑟𝑔𝑒𝑡,𝑖𝑚𝑝𝑙𝑖𝑒𝑑 = 𝑐𝑉𝑠 + 𝜎𝑣′ ∗ tan(𝜙) (7.1)

where, 𝜎𝑣′ is the effective stress at the mid-depth of interested soil layer, 𝜙 is the friction
angle of soil layer (300 for this example), and 𝑐𝑉𝑠 is the judgement-based shear strength
developed at 0.1% shear strain for a linear elastic material with 80% of the maximum shear
modulus derived from the shear wave velocity (VS) of the soil layer under consideration.
(Harmon et al. (2017)) as (Figure 7.35),
𝑐𝑉𝑠 = 𝜌 ∗ 𝑉𝑆2 ∗ 0.8 ∗ 0.1% (7.2)

where ρ is the soil density.
• Reference curve (Figure 7.36) is selected as Darendeli (2001) with parameters of OCR and
PI (%) as 1.0 and 0.0. The coefficient of earth pressure at rest (K0) required for calculation
of reference dynamic curves (Darendeli (2001) model is selected for this example) is
calculated using the (Jaky, 1948) eq. as,

𝐾0 = [1 − sin(𝜙)] ∗ 𝑂𝐶𝑅 sin(𝜙) (7.3)
where OCR is over-consolidation ratio.
• The modulus reduction and damping curve fitting (MRDF) with UIUC Reduction Factor
is used to capture the Non-Masing behavior (Figure 7.37), and GQ/H Model is fitted for a
shear strain range up to 0.05% considering the Modulus Reduction Curve under the
condition that the shear stresses reach 95% of the target shear strength at shear strain of
10% (Figure 7.38). The fitted modulus reduction and damping curve together with fitting
ranges for Layer 1 are illustrated in Figure 7.39.
• Chi-Chi Motion is used, and Frequency-Independent viscous damping along with no
damping matrix update are selected as in Example 1 and 2.
• Parameters for nonlinear and equivalent-linear analysis are assigned as (Figure 7.40):
o For frequency-domain analysis:
▪ Number of iterations: 15
▪ Effective Shear Strain Ratio (SSR): 0.65
▪ Complex Shear Modulus Formulation: Frequency-Independent
o For nonlinear (time-domain) analysis:
▪ Step Control: Flexible
▪ Maximum Strain Increment: 0.005 %
▪ Time History Interpolation Method: Linear in time domain
• Computed time-histories (Acceleration, Velocity, Displacement, and Arias Intensity) from
nonlinear (blue line) and equivalent-linear analyses (red line) are given in Figure 7.41.
Response spectrum at first layer of site profile from nonlinear analysis and equivalent-
linear analysis along with that of input motion are given in Figure 7.42 and Figure 7.43.

Figure 7.33 Step 1: Analysis Type Definition for Example 6.
Figure 7.34 Step 2: Soil Profile Definition for Example 6

Figure 7.35 Step 2: Shear Strength Input for Soil Layer 1
Figure 7.36 Step 2: Definition of Darendeli (2001) Dynamic Curve

Figure 7.37 Step 2: Application of “MRDF with UIUC Reduction Factor” Option
Figure 7.38 Step 2: Fitting of GQ/H Model

Figure 7.39 Step 2: GQ/H Model Fit and Model Parameters for Layer 1
Figure 7.40 Step 5: Analysis Control Definition for Example 6

for Layer 1 via Nonlinear Analysis (blue line) and Equivalent-Linear Analysis (red line) methods
Motion via Nonlinear Analysis Method

Motion via Equivalent-Linear Analysis Method

7.7 Example 7: Nonlinear Analysis / Multi-Layer, Elastic Rock, Pore Water Pressure
Generation and Dissipation
This example is composed of nonlinear analysis as an effective stress analysis with generation and
dissipation of pore water pressure. The steps are mainly identical to those taken in Example 6
except the definition of parameters for pore water pressure model. The flow of analysis through
DEEPSOIL V7.0 interface is illustrated below:
• The nonlinear analysis is selected for site profile in which soils are modelled using GQ/H
model with Non-Masing Re/Unloading formulation. “Generate Excess Porewater
Pressure” option is enabled to enable dissipation and to make both top and bottom of
profile as permeable boundaries. (Figure 7.44)
• Figure 7.45 shows the propertes of first layer, and shear strength is calculated using the eq.
(7.1) in Example 6. Friction angle for each sand layer is assumed as 300. Darendeli (2001)
is selected as reference curve along with K0 calculation by Jaky’s equation in Example 6.
• In Advanced Table View (Figure 7.46), Sand-Vucetic Dobry model is selected with the
paramaters give in Table 7.7.
• Half-space definition is given in Figure 7.47 with VS = 5000 ft/s, unit weight of 160.0 pcf,
and damping ratio of 2.0 %. The resultant soil profile properties can be viewed in Figure
7.48.
• “Kobe” motion is selected for this example, and time-histories (Acceleration, Velocity,
Displacement, Arias and Housner Intensity), FAS, and 5% damped spectral acceleration
for Kobe motion can be viewed in Figure 7.49.
• After analysis, (i) profile plots, (ii) Acceleration, Velocity, Displacement, and Arias
Intensity time histories, (iii) stress-strain plots for layer 1 and layer 3, and (iv) 5% damped
spectral acceleration, Fourier Amplitude Spectrum and Ratio for Layer 1 and Layer 3 are
presented in Figure 7.50, Figure 7.51, Figure 7.52 and Figure 7.53, respectively. Layer 1
and 3 are distinguished by blue and red lines, respetively. Maximum strain of
approximately 0.4% in site profile occurs Layer 3 due to development of significant level
of pore water pressure. (Figure 7.52)

Table 7.7 Sand-Vucetic Dobry PWP model parameters for Sand layer
Max ru Cv (ft2/sec) Cv–exponent f p s γ ν
0.95 0.1 0.0 1.0 1.0 1.0 0.02 3.80

Figure 7.44 Step 1: Analysis Type Definition for Example 7
Figure 7.45 Step 2: Soil Profile in Nonlinear Analysis with PWP Generation and Dissipation

Vucetic/Dobry Model
Figure 7.47 Step 2: Halfspace Definition

Figure 7.48 Step 2: Soil Profile Plot

Figure 7.50 Results: Profile Plots after analyses
7 for Layer 1 (blue line) and Layer 3 (red line).

Figure 7.52 Results: Stress-Strain plots for Layer 1 (blue line) and Layer 3 (red line)
Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Layer 1 (blue line) and Layer
3 (red line)

7.8 Example 8: Nonlinear Analysis / Multi-Layer, Elastic Rock, Pore Pressure Generation
and Dissipation:
Example 8 is identical to Example 7 except that GMP PWP model is adopted for sand layers. The
details of analysis are presented as:
• The definition of parameters for GMP PWP model for each layer is given in Table 7.8, and
also can be viewed in Figure 7.54.
• After analysis, (i) profile plots, (ii) Acceleration, Velocity, Displacement, and Arias
Intensity time histories, (iii) stress-strain plots for layer 1 and layer 3, and (iv) 5% damped
spectral acceleration, Fourier Amplitude Spectrum and Ratio for Layer 1 and Layer 3 are
presented in Figure 7.55, Figure 7.56, Figure 7.57 and Figure 7.58, respectively. Layer 1
and 3 are distinguished by blue and red lines, respetively. Results are similar to Example 7
except that much greater maximum strain of approximately 5.0% in site profile occurs
Layer 3 due to significant level of pore water pressure, which reaches ru = 0.95 (Figure
7.57).
Table 7.8 GMP PWP model parameters for Sand layer

Max ru Cv (ft2/sec) Cv–exponent α Dr (%) FC (%) ν
0.95 0.1 0.0 2.0 95 15 3.80

Figure 7.54 Step 2: Definition of Parameters for PWP Generation and Dissipation Model of Sand-GMP
Figure 7.55 Results: Profile Plots

Figure 7.57 Results: Stress-Strain plots for Layer 1 (blue line) and Layer 3 (red line)

3 (red line)

7.9 Example 9: Equivalent Linear Frequency Domain Analysis / Multi-Layer, Elastic
Rock, Bay Mud Profile
Example 9 is similar to Example 4 (equivalent-linear analysis with discrete points) but includes
31 layers as given in Figure 7.59. This is a typical profile near San Francisco Bay, and is included
to illustrate the capabilities of DEEPSOIL for more realistic profiles. The reference condition is
defined as elastic halfspace with shear wave velocity of 1500 m/s, unit weight of 22 kN/m3 and
damping ratio of 0.0%, and the resultant profile plots can be viewed in Figure 7.61. Analyses
results as (i) computed time-histories (accelertion, velocity, displacement and Arias intensity), and
(ii) 5% damped spectral acceleration, FAS and Fourier Amplitude Ratio at first layer (Fill) are
presented in Figure 7.62 and Figure 7.63, respectively.

Figure 7.59 Step 2: Soil Profile Definition
Figure 7.60 Step 2: Bedrock Profile Definition in Example 9

Figure 7.61 Step 2: Soil profile plot
Figure 7.62 Results: Computed Time-Histories (Acceleration, velocity, displacement, and Arias Intensity)
for Layer 1

Figure 7.63 Results: 5% Damped Spectral Acceleration for Layer 1 (blue line) and for Input Motion (black
line)

7.10 Example 10: Nonlinear Analysis / Multi-Layer, Rigid Rock, Treasure Island Profile
Example 10 consists of nonlinear analysis along with compementary equivalent-linear analysis,

identical to Example 6 but includes 53 layers and located on rigid bedrock. The flow of analysis
through DEEPSOIL V7.0 intefrace is illustrated below:
• Nonlinear analysis using MKZ soil model with Masing Re/Unloading formulation along
with complementary equivalent-linear analysis is selected as given in Figure 7.64.
• The soil profile including 53 layers and MKZ model parameters to create soil backbone
curve for Layer 1 (as representative) can be viewed in Figure 7.65
• The reference condition is determined as rigid bedrock in Figure 7.66, and soil profile plots
can be seen in Figure 7.67
• Kobe motion is selected as input motion, and motion properties can be viewed in Figure
7.68
• Two modes of Rayleigh Damping along with update on Damping Matrix is selected as
given in Figure 7.69
• Parameters for nonlinear and equivalent-linear analysis are assigned as:
▪ Time History Interpolation Method: Zero-padded in frequency domain
• Computed time-histories as acceleration, velocity, displacement and arias intensity from
nonlinear (blue line) and equivalent-linear analysis (red line) at the top layer of soil profile
are given in Figure 7.70
• 5% damped spectral acceleration response spectra from nonlinear and equivalent-linear
analysis at top layer of the soil profile (blue line) along with that for input motion (black
line) are presented in Figure 7.71 and Figure 7.72, respectively.

Figure 7.64 Step 1: Analysis Type Definition
Figure 7.65 Step 2: Soil Profile Definition (Layer 1)

Figure 7.66 Step 2: Bedrock Profile Definition
Figure 7.67 Step 2: Soil profile plot

Figure 7.68 Step 3: Properties of Kobe Motion
Figure 7.69 Step 4: Viscous/Small-Strain Damping Definition

from Nonlinear (blue line) and Equivalent-Linear Analysis (red line) for Layer 1
Figure 7.71 Results: 5% Damped Spectral Acceleration for Layer 1 via Nonlinear Analysis


7.11 Example 11: Nonlinear Analysis / Multi-Layer, Elastic Rock, MRDF
Example 11 is composed of 80-layer profile located on elastic rock, and is generated to further
illustrate the the capabilities of DEEPSOIL for more realistic profiles as well as the MRDF curve
parameters. The flow of analysis through DEEPSOIL V7.0 intefrace is illustrated below:
• Nonlinear analysis using MKZ soil model with Non-Masing Re/Unloading Formulation
along with complementary equivalent-linear analysis is selected as given in Figure 7.73.
• The soil profile including 80 layers and MKZ model parameters to create soil backbone
curve for Layer 1 (as representative) can be viewed in Figure 7.74.
• The reference condition is determined as elastic halfspace with shear wave velocity of 2492
ft/s, unit weight of 137 pcf. and no damping in Figure 7.75, and soil profile plots can be
seen in Figure 7.76.
• Kobe motion is selected as input motion, and selection of Rayleigh Damping with 2
modes/freqs along with no update in damping matrix is illustrated in Figure 7.77.
▪ Time History Interpolation Method: Zero-padded in frequency domain
• Computed time-histories as acceleration, velocity, displacement and arias intensity from
nonlinear (blue line) and equivalent-linear analysis (red line) at the top layer of soil profile
are given in Figure 7.78
• 5% damped spectral acceleration response spectra from nonlinear and equivalent-linear
analysis at top layer of the soil profile (blue line) along with that for input motion (black
line) are presented in Figure 7.79 and Figure 7.80, respectively.

Figure 7.74 Step 2: Soil Profile Definition (Layer 1)

Figure 7.75 Step 2: Halfspace Definition
Figure 7.76 Step 2: Soil profile plots

Figure 7.77 Step 4: Viscous/Small-Strain Damping Definition
from Nonlinear (blue line) and Equivalent-Linear Analysis (red line) for Layer 1

Figure 7.79 Results: 5% Damped Spectral Acceleration for Layer 1 via Nonlinear Analysis

7.12 Example 12: Nonlinear Analysis with Auto-Profile Generation Option:
This example is generated to illustrate the new feature in DEEPSOIL V7.0 as auto-profile
generation, which allows to (i) divide the soil layers considering the maximum frequency (fmax =
VS/4H), and (ii) automatically fit the selected soil model to reference curve for each layer. The site
profile in this example is one of the 70650 site profiles used in Harmon et al. (2017), and is
composed of 20 clay layers with constant unit weight of 19 kN/m3, PI = 15%, OCR = 1.5, and
friction angle (𝜙) of 250 along the profile. The shear wave velocity monotonically increases with
depth and, shear strength required for fitting GQ/H Model at each layer is calculated as explained
in Example 6.
The flow of analysis through DEEPSOIL V7.0 interface of this example is illustrated below:
• The analysis type definition for nonlinear analysis with auto-profile generation is illustrated
in Figure 7.81. The only difference from Example 6 is to turn Automatic Profile Generation
“on”.
• In the Automatic-Profile generation page (Figure 7.82), maximum frequency (fmax = VS/4H)
value is provided as 50.0 Hz to specify the maximum layer thickness, and randomization
option for soil thickness, velocity and dynamic curve is turned off at this stage.
• Mean soil profile definition is illustrated in Figure 7.83 for Layer 1. The required basic soil
parameters as soil thickness, unit weight, shear wave velocity (as constant for each layer),
and shear strength (as constant for each layer) needed for fitting GQ/H Model have been
provided. These values for each soil layer can be viewed through Advanced Table View
(Figure 7.84) Mean reference curve is selected as Darendeli (2001), and K0 parameter is
calculated as 0.685 using the same eq. in Example 6.
• Bedrock is defined as elastic half-space with VS of 3000 m/s for CENA (Hashash et al.,
2014), unit weight of 22 kN/m3, and damping ratio of 0.5%.
• The subdivided profile and GQ/H Model fit performed automatically for each subdivided
layer is illustrated in Figure 7.86 (only Layer 1 is shown as representative). Additionally,
Soil profile plots are presented in Figure 7.87

• Input motion for nonlinear and complementary equivalent-linear analyses is selected as
ChiChi (Step 3), and Frequency-independent viscous-damping along with “No update in
damping matrix” is used (Step 4).
• Parameters for nonlinear and equivalent-linear analysis are assigned as (Figure 7.88):
• Computed time-histories (acceleration, velocity, displacement and Arias Intensity) from
nonlinear (blue line) and equivalent-linear (red line) analyses is presented in Figure 7.89.
5% damped spectral acceleration at the top of Layer along with that of input motion from
nonlinear and equivalent-linear analysis are given in Figure 7.90 and Figure 7.91.

Figure 7.82 Step 1: Inputs to Automatic Profile Generation

Figure 7.83 Step 2: Mean Soil Profile Definition (for Layer 1)
Figure 7.84 Step 2: Advanced Table View

Figure 7.85 Step 2: Bedrock definition as Elastic Half-space
Figure 7.86 Step 2: Subdivision of each Layer Thickness and GQ/H Model Fit to Layer 1

Figure 7.87 Step 2: Soil Profile Plots
Figure 7.88 Step 5: Parameters for Nonlinear and Equivalent-Linear Analyses

Figure 7.89 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for Nonlinear
Analysis (blue line) and Equivalent-Linear Analysis (red line)
Figure 7.90 Results: 5% Damped Spectral Acceleration for Layer 1 (blue line) along with for Input Motion
(black line) from Nonlinear Analysis

(black line) from Equivalent-Linear Analysis

7.13 Example 13: Nonlinear Analysis with Randomized Soil Profile:
Example 13 is created to illustrate the new features in DEEPSOIL V7.0 as Thickness, VS and
Dynamic Curve randomizations. The mean site profile used as input to site realizations is identical
to that in Example 12. The theory behind the site realization will explained in Chapter 6 in detail.
• The analysis type definition for Example 13 is identical to that for Example 12 (Figure
7.81).
• The maximum frequency (fmax) is specified as 50.0 Hz, and parameters for thickness, VS
and dynamic curve randomization are given below (Figure 7.92):
o Thickness Randomization: (Default parameters given in Toro (1995))
▪ Minimum Thickness: 0.03 m
▪ Maximum Thickness: 152.40
▪ C1: 10.86
▪ C2: 0.89
▪ C3: 1.98
o Velocity Randomization:
▪ Input Parameter Method: User Defined
▪ σ: 0.2
▪ ρ0 (Rho0): 0.0
▪ Δ (Delta): 1.0
▪ ρ200 (Rho200): 1.0
▪ b: 0.0
▪ h0: 0.0
o Dynamic Curve Randomization:
▪ ρ1: -1.0
▪ ρ2: 1.0
▪ σ Randomization Bound: 1.50
o Number of Random Profiles: 20
It is of importance to note that:

o ρ1 represents the correlation between G/Gmax and dynamic curve for each layer.
Perfectly negatively correlated G/Gmax (ρ1 = -1.0) and damping curves are assumed
in this example.
o ρ2 is inter-layer correlation for dynamic curves. Dynamic curves are perfectly
positively correlated (ρ2 = 1.0) for this example
o σ randomization bound is assigned as 1.5.
• After realization, layer properties of each realization can be viewed as given in Figure 7.93
(Layer 1 of 1st profile realization is illustrated as representative, and each realization can
be viewed using pop-up menu in red circle). GQ/H Model fitted to selected reference curve,
which is Darendeli (2001) as an only option for dynamic curve randomization, for each
layer can be monitored along with curve bounds (orange lines in Figure 7.93). Additionally,
options for exporting the thickness and VS of each layer for randomized VS profiles along
with mean soil profile are provided as (i) Excel file, (ii) SQLite database, and (iii) Enhanced
Metafile (EMF) of VS plot. Figure 7.95 shows the VS profiles of 20 site profile realizations
along with (i) mean input profile, (ii) logarithmic mean of 20 randomized VS profile, and
95% confidence interval (CI) on logarithmic mean.
• Figure 7.96 illustrates the soil profile plots for 1st realization. The purple and blue lines
represent plots for mean profile and selected realization, respectively (1st realization as
illustrated by red circle is shown in Figure 7.96).
• Input motion for nonlinear and complementary equivalent-linear analyses is selected as
ChiChi (Step 3), and Frequency-independent viscous-damping along with “No update in
damping matrix” is used (Step 4).

• Computed time-histories (acceleration, velocity, displacement and Arias Intensity) from
nonlinear (blue line) and equivalent-linear (red line) analyses is presented in Figure 7.97.
5% damped spectral acceleration at the top of Layer along with that of input motion from
nonlinear and equivalent-linear analysis are given in Figure 7.98 and Figure 7.99.

Figure 7.92 Step 1: Specification of Maximum Frequency and Definition of Parameters for Thickness, VS
and Dynamic Curve Randomization
Figure 7.93 Step 2: Properties of Layer 1 for first Realization after Randomization

Figure 7.94 Step 2: Export Options for Site Profile Realizations
Figure 7.95 Randomized VS Profiles along with their logarithmic mean and 95% Confidence Interval on
Logarithmic Mean

Figure 7.96 Step 2: Soil Profile Plot for Profile Realizations
Figure 7.97 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for Nonlinear
Analysis

Amplitude Ratio (FAS at ground surface divided by FAS at input motion) for Nonlinear Analysis
(black line) from Equivalent-Linear Analysis

7.14 Example 14: Nonlinear Analysis by Soil Profile with various Soil Models at Different Layers:
Example 14 is designed to illustrate the new feature in DEEPSOIL as allowing users to adopt
various soil models to different soil layers. This site profile is extracted as shear-beam from 2-D
finite element model of centrifuge model for Concrete-face rockfill dams with depth of 352.2 m,
which is composed of rock-fill with 202.2 m depth (from top to 42th layer (included) and
underlying bedrock layers. The properties of each layer can be detailed as:
• For Rockfill Layer:

o Friction angle is selected constant value of 450 throughout depth.
o Stress-strain behavior of rockfill material is assumed to follow Non-Masing
Re/Unloading Formulation, and backbone formulation is computed using GQ/H
soil model fitted to Darendeli (2001) reference curve.
o “MRDF with UIUC Formulation” is selected to capture the Non-Masing behavior
of Rockfill material.
o K0 and shear strength value required for Darendeli (2001) curve and GQ/H model
are calculated as given in Example 6, respectively.
• For Bedrock Layer:
o It is assumed to behave linearly under earthquake excitation.
o VS and Dmin values are assumed to be constant as 1500 m/s and 0.1%, respectively.
• Nonlinear analysis without any complementary analyses is selected, and GQ/H soil model
with Non-Masing Re/Unloading formulation is used for the analysis.
• Figure 7.100 shows the assignment of properties for Rockfill material at Step 2.
• After properties of all Rockfill layers are assigned, the Soil Models from Layer 43 to 58
should be converted to “Linear” to represent the bedrock behavior (red rectangular in
Figure 7.101) along with removal of “Reduction Factor Formulation” in Figure 7.102
• Bedrock layer properties are assigned as given in Figure 7.103.
• Half-space is defined as “Rigid Halfspace” (Figure 7.104), and soil profile plots are
presented in Figure 7.105
• Parameters for nonlinear is assigned as:
o Step Control: Flexible

o Maximum Strain Increment: 0.005 %
o Time History Interpolation Method: Linear in time domain
• Stress-strain plots for top of rockfill (Layer 1 with red line) and bedrock material (Layer
43 with blue line) are presented in Figure 7.106
• Computed time-histories (Acceleration, Velocity, Displacement and Arias Intensity) at top
of rockfill (Layer 1 with red line) and bedrock material (Layer 43 with blue line) are
presented in Figure 7.107
• 5% damped spectral acceleration and FAS at top of rockfill (Layer 1 with blue line) and
bedrock material (Layer 43 with yellow line) are presented in Figure 7.108.

Figure 7.100 Step 2: Basic Soil Properties and GQ/H Model Fit to Rockfill Layers (Layer 1 as
representative)
Figure 7.101 Step 2: Assignment of Linear Material to Bedrock Layers

Figure 7.102 Step 2: Removal of Reduction Factor Formulation for linear Bedrock layers
Figure 7.103 Step 2: Basic Soil Properties of Bedrock Layers (Layer 45 as representative)

Figure 7.104 Step 2: Rigid half-space definition
Figure 7.105 Step 2: Soil Profile Plots

Figure 7.106 Results: Stress-strain plots for Soil Layer 1 (GQ/H Model with Non-Masing Un/Reloading
Formulation (red line) and Soil Layer 43 (linear elastic soil material (blue line)
Figure 7.107 Results: Acceleration, Velocity, Displacement and Arias Int. Time-Histories for Layer 1 (red
line) and Layer 43 (blue line)

43 (yellow line)

8 References
Amir-Faryar, B., M. Sherif Aggour & Richard H. McCuen (2016). Universal model forms for
predicting the shear modulus and material damping of soils. Geomechanics and Geoengineering.
http://dx.doi.org/10.1080/17486025.2016.1162332
Andrus, R. D., et al. (2003). Guide for Estimating the Dynamic Properties of South Carolina
Soils for Ground Response Analysis, Clemson University.
Berrill, J.B. & Davis, R.O. (1985). Energy Dissipation and Seismic Liquefaction of Sands:
Revised Model. JSSMFE Soils and Foundations 25(2): 106-118.
Carlton, B. (2014). “An Improved Description of the Seismic Response of Sites with
High Plasticity Soils, Organic Clays, and Deep Soft Soil Deposits.” PhD Thesis.
University of California, Berkeley
Chopra, Anil K. (1995) Dynamic of Structures, Theory and applications to Earthquake

Engineering Englewood Cliffs, New Jersey: Prentice Hall.
Clough, Ray W., and Joseph Penzien (1993) Dynamics of structures, New York: McGraw-Hill.
Darendeli, M. B. (2001). Development of a New Family of Normalized Modulus Reduction

and Material Damping Curves, Department of Civil, Architectural and Environmental
Engineering, The University of Texas, Austin, Texas.
Duncan, James M., and Chin-Yung Chang (1970) "Nonlinear analysis of stress and strain in
soils," Journal of the Soil Mechanics and Foundations Division, Vol. 96, No. SM5, pp 1629-1653.
Finn, W. D. L., Lee, K. L, and Martin, G. R. (1977) “An effective stress model for liquefaction.”
Journal of the Geotechnical Engineering Division, ASCE, Vol. 103, No. GT6, 517-533.

Green, R.A., Mitchell, J.K. and Polito, C.P. (2000). "An Energy-Based Pore Pressure
Generation Model for Cohesionless Soils", Proceedings: John Booker Memorial Symposium,
Melbourne, Australia, November 16-17, 2000.
Groholski, D., Hashash, Y., Kim, B., Musgrove, M., Harmon, J., and Stewart, J. (2016).
"Simplified Model for Small-Strain Nonlinearity and Strength in 1D Seismic Site Response
Analysis." J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)GT.1943-5606.0001496, 04016042.
Hardin, B. O. and Drnevich, V. P. (1972) “Shear modulus and damping in soils: Design
equations and curves.” Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 98,
No. SM7, 667-692.
Hashash, Y. M. A., and D. Park (2002) "Viscous damping formulation and high frequency
motion propagation in nonlinear site response analysis," Soil Dynamics and Earthquake
Engineering, Vol. 22, No. 7, pp. 611-624.
Hashash, Youssef M. A., and Duhee Park (2001) "Non-linear one-dimensional seismic ground
motion propagation in the Mississippi embayment," Engineering Geology, Vol. 62, No. 1-3, pp
185-206.
Hashash, Y.M.A., Phillips, C. and Groholski, D. (2010). "Recent advances in non-linear site
response analysis", Fifth International Conference on Recent Advances in Geotechnical
Earthquake Engineering and Soil Dynamics, Paper no. OSP 4.
Hudson, M., Idriss, I.M., and Beikae, M. 1994. (1994) "QUAD4M - A computer program to
evaluate the seismic response of soil structures using finite element procedures and incorporating
a compliant base." Davis, CA: Center for Geotechnical Modeling, Department of Civil and
Environmental Engineering, University of California, Davis, CA.
Idriss, I. M. and Seed, H. B. (1968) “Seismic response of horizontal soil layers.” Journal of the
Soil Mechanics and Foundations Division, ASCE, Vol. 94, No. SM4, pp 1003-1031.

Idriss, I. M. and Sun, J. I. (1992) “User’s Manual for SHAKE91, A Computer Program for
Conducting Equivalent Linear Seismic Response Analyses of Horizontally Layered Soil Deposits”
Kramer, Steven Lawrence (1996) Geotechnical earthquake engineering, Upper Saddle River,
N.J.: Prentice Hall.
Konder, R. L. and Zelasko, J. S. (1963) “A hyperbolic stress-strain formulation of sands.”

Proceedings of the 2nd Pan American Conference on Soil Mechanics and Foundation Engineering,
Sao Paulo, Brasil, 289-324.
Lee, M. K. W. and Finn, W. D. L (1975) “DESRA-1, Program for the dynamic effective stress
response analysis of soil deposits including liquefaction evaluation.” Soil Mechanics Series No.
36, Department of Civil Engineering, University of British Columbia, Vancouver, Canada.
Lee, M. K. W. and Finn, W. D. L (1978) “DESRA-2, Dynamic effective stress response analysis
of soil deposits with energy transmitting boundary including assessment of liquefaction potential.”
Soil Mechanics Series No. 36, Department of Civil Engineering, University of British Columbia,
Vancouver, Canada.
Matasovic, Neven, and M. Vucetic (1993) "Cyclic Characterization of Liquefiable Sands,"

ASCE Journal of Geotechnical and Geoenvironmental Engineering, Vol. 119, No. 11, pp 1805-
1822.
Matasovic, N. (1992) “Seismic response of composite horizontally-layered soil deposits.” Ph.D.

Thesis, University of California, Los Angeles.
Matasovic, Neven, and M. Vucetic (1995) "Generalized Cyclic Degradation-Pore Pressure

Generation Model for Clays," ASCE Journal of Geotechnical and Geoenvironmental Engineering,
Vol. 121, No. 1, pp 33-42.

Mei, X., Olson, S.M., and Hashash Y.M. (2015) “Empirical curve-fitting parameters for a
porewater pressure generation model for use in 1-D effective stress-based site response analysis,”
6th International Conference on Earthquake Geotechnical Engineering, 1-4 November 2015,
Christchurch, New Zealand.
Menq, Farn-Yuh (2003). Dynamic Properties of Sandy and Gravelly Soils, Department of Civil,
Architectural and Environmental Engineering, The University of Texas, Austin, Texas.
Mugan, Ata, and Gregory M. Hulbe (2001) "Frequency-domain analysis of time-integration

methods for semidiscrete finite element equations - part II: Hyperbolic and parabolic-hyperbolic
problems," International Journal for numerical Methods in Engineering, Vol. 51, No. 3, pp 351-
376.
Newmark, Nathan M. (1959) "A method of computation for structural dynamics," Journal of
the Engineering Mechanics Division, Vol. EM 3, pp 67-94.
Park D., J.-K. Ahn. (2013) “Accumulated Stress Based Model for Prediction of Residual Pore
Pressure,” Proceedings of the 18th International Conference on Soil Mechanics and Geotechnical
Engineering, Paris, pp 1567-1570.
Park, D., and Y. M. A. Hashash (2004) "Soil damping formulation in nonlinear time domain
site response analysis," Journal of Earthquake Engineering, Vol. 8, No. 2, pp 249-274.
PEER (2010) "PEER Ground Motion Database Web Application," PEER.
Phillips, Camilo, and Youssef M. A. Hashash (2008) "A new simplified constitutive model to
simultaneously match modulus reduction and damping soil curves for nonlinear site response
analysis," Geotechnical Earthquake Engineering & Soil Dynamics IV (GEESD IV). Sacramento,
California.

Phillips, C. and Hashash, Y. (2009) “Damping formulation for non-linear 1D site response
analyses” Soil Dynamics and Earthquake Engineering, v. 29, pp 1143–1158.
Phillips, C., Kottke, A.R., Hashash, Y.M.A., and Rathje, E.M. (2012) “Significance of ground
motion timestep in one dimensional site response analysis” Soil Dynamics and Earthquake
Engineering, v. 43, pp 202–217.
Richart, F.E., Hall, J.R., and Woods, R.D. (1970). “Vibrations of Soils and Foundations,”
Prentice Hall, Englewood Cliffs, New Jersey, 401 pp.
Roblee, Cliff and Chiou, Brian. (2004) “A proposed geoindex model for design selection of
non-linear properties for site response analyses.” Caltrans Geo-Research Group. Sacramento, CA.
Sanderson, Conrad. (2010) “Armadillo: An Open Source C++ Linear Algebra Library for Fast
Prototyping and Computationally Intensive Experiments.” Technical Report, NICTA.
Sanderson, Conrad and Ryan Curtin. (2016) Armadillo: a template-based C++ library for linear
algebra. Journal of Open Source Software, Vol. 1, pp. 26.
Schnabel, P. B., Lysmer, J. and Seed, H. B. (1972) “SHAKE: A computer program for
earthquake response analysis of horizontally layered sites.” Report No. EERC 72-12, Earthquake
Engineering Research Center, University of California, Berkeley, California.
Seed, H. B. and Idriss, I. M. (1970) “Soil moduli and damping factors for dynamic response
analyses.” Report No. EERC 70-10, Earthquake Engineering Research Center, University of
California, Berkeley, California, 40p.
Udaka, Takekazu (1975) "Analysis of Response of Large Embankments to Traveling Base

Motions," Department of Civil and Environmental Engineering. Berkeley: University of
California, p 346.

Vucetic, M. (1992). “Soil Properties and Seismic Response.” In Proceedings of 10th World
Conference of Earthquake Engineering, July 19 – 24, Madrid, Spain, 1199 – 1204.

9 APPENDICES
This section is under development.

Deepsoil User Manual v7

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Deepsoil User Manual v7

Uploaded by

Copyright:

Available Formats

Nonlinear and Equivalent Linear Seismic Site Response of

One-Dimensional Soil Columns

Department of Civil and Environmental Engineering

DEEPSOIL User Manual V 7.0 Page 1 of 169 February 1, 2018

1 Program Background and Installation .................................................................................. 14

1.1 About the Program ................................................................................................... 14

1.2 A Word of Advise and Caution ............................................................................... 15

1.3 Historical Development ........................................................................................... 15

1.4 Program Installation................................................................................................. 19

2 Program Organization ........................................................................................................... 21

2.1 Profiles Tab.............................................................................................................. 22

2.2 Motions Tab ............................................................................................................. 22

2.2.1 Baseline Correction ................................................................................................. 25

2.2.2 Response Spectra Calculation Methods................................................................... 27

2.2.3 Fourier Amplitude Spectrum Calculation and Smoothing Operation ..................... 29

2.2.4 Arias Intensity.......................................................................................................... 31

2.2.5 Significant Duration................................................................................................. 31

2.2.6 Housner Intensity ..................................................................................................... 31

2.2.7 Estimation of Kappa (κ) .......................................................................................... 32

2.2.8 Adding New Input Motions ..................................................................................... 33

2.3 Analysis Tab ............................................................................................................ 33

3 Analysis Flow ....................................................................................................................... 34

3.1 Analysis Definition: Step 1 of 5 .............................................................................. 34

3.1.1 Linear Analysis ........................................................................................................ 38

3.1.2 Equivalent Linear Analysis ..................................................................................... 38

3.1.3 Deconvolution via Frequency Domain Analysis ..................................................... 38

3.1.4 Non-Linear Analysis ................................................................................................ 40

3.2 Soil Profile Definition: Step 2 of 5 .......................................................................... 40

3.2.2 Maximum Frequency (for Time Domain Analysis only) ........................................ 43

3.2.3 Implied Strength Profile (Step 2) ............................................................................. 45

3.2.4 Halfspace Definition (Bedrock) ............................................................................... 45

3.3 Step 3 of 5: Input Motion Selection......................................................................... 47

3.4 Step 4 of 5: Viscous/Small-Strain Damping Definition .......................................... 48

3.4.1 Frequency Independent Damping Formulation ....................................................... 50

3.4.2 Rayleigh Damping formulation types ...................................................................... 50

3.5 Step 5 of 5: Analysis Control Definition ................................................................. 51

3.5.1 Frequency domain analysis ..................................................................................... 52

3.5.2 Time domain analysis .............................................................................................. 53

3.6 Results ..................................................................................................................... 55

3.6.1 Time History Plots tab ............................................................................................. 56

3.6.2 Stress-Strain Plots tab ............................................................................................. 56

3.6.3 Spectral Plots tab ..................................................................................................... 57

3.6.4 Profile Plots tab ....................................................................................................... 58

3.6.5 Mobilized Strength tab ............................................................................................. 58

3.6.6 Displacement Animation tab .................................................................................... 60

3.6.7 Response Spectra Summary tab ............................................................................... 60

3.6.8 Check Convergence tab ........................................................................................... 60

3.6.9 Output data file ........................................................................................................ 61

4 Soil Models ........................................................................................................................... 62

4.1 Backbone Curves ..................................................................................................... 62

4.1.1 Hyperbolic / Pressure-Dependent Hyperbolic (MKZ) ............................................ 62

DEEPSOIL User Manual V 7.0 Page 3 of 169 February 1, 2018

4.2.1 Masing Rules ........................................................................................................... 64

4.2.2 Non-Masing Unload-Reload Rules.......................................................................... 64

4.3 Porewater Pressure Generation & Dissipation ........................................................ 66

4.3.1 Dobry/Matasovic Model for Sand ........................................................................... 66

4.3.2 Matasovic and Vucetic Model for Clays ................................................................. 72

4.3.4 Generalized Energy-based PWP Generation Model................................................ 76

4.3.5 Park and Ahn Model for Sand ................................................................................. 76

4.3.6 Porewater Pressure Degradation Parameters ........................................................... 78

4.3.7 Porewater Pressure Dissipation ............................................................................... 79

5 Randomization of Site Profile Properties ............................................................................. 80

6 Database Output Structure .................................................................................................... 81

6.1 Database Structure for Analyses Output.................................................................. 81