House Module Options

The ACS SASSI HOUSE module is a standard finite element program which computes the basic frequency independent global mass and stiffness matrices, for both structure and excavated soil.

Two separate finite element models are constructed, one for the structure and the other for the excavated volume of soil. The SSI models can share the same nodal points at/below the ground surface. The nodes that define the excavated soil volume are called interaction nodes.

The finite element library includes (see GROUP):

The excavated soil zones are modeled using the SOLID, PLANE, and LOVEWAVE element types.

The finite element models of the structure and the excavated soil must be selected in such a way that every interaction node below the ground should lie on a soil layer interface. ACS SASSI HOUSE reads the nodal point input data, nodal types, soil layer properties, and element data for the structural and excavated soil elements, then forms the element mass and stiffness matrices for these elements which are later assembled into corresponding global mass and stiffness matrices. These matrices are stored in compacted blocks in preparation for solution by the active column method later in the ACS SASSI ANALYS module. The results are written to File4. If the skin method is to be used (not recommended) for computation of the impedance matrix, the excavated soil elements are once again assembled, but this time in a different format in order to form the global matrices M12 and K12. The columns of these matrices follow the same order as the degrees of freedom to be used later to form the flexibility matrix F12, thus making it possible to carry out the matrix operation efficiently. The matrices M12 and K12 are full matrices which are stored in blocks and then are written to File4.

The node numbering should be as follows: the nodes at or below the ground surface first, layer by layer starting from the bottom. The following figure shows examples of node numbering. Thus, recommend the bottom-up node numbering as a standard convention for ACS SASSI code. Using the bottom-up node numbering scheme any SSI analysis options would be used correctly.

For element numbering in ACS SASSI, there is no special restriction in the element numbering for structures with no embedment. The element numbering has to be in continuous sequence. However, for structures with embedment, the element numbering of the solid elements used to define the excavation volume has to be based on a top-bottom element numbering scheme, from ground surface to baserock, so that the far-field soil layers associated with the “layers” of solid elements are assigned in an increasing order.

As an extra caution when building a SSI model, please always check your solid elements soil layer and material assignments in the HOUSE output file before the SSI analysis run.

The following options allow you to specify the analysis options for ACS SASSI HOUSE module:

• Operation Mode - Select the operation mode from Solution and Data Check.
• Dimension of Analysis - Select the dimension of analysis from 1D, 2D, and 3D (1D not available)

Flexible Volume Methods Select either the Flexible Volume methods with the option of the Flexible Volume, Direct Method or Skin Method, or the Flexible Interface with the option of Direct Method (equivalent to subtraction). For surface foundation these methods are identical. For embedded or buried structures, we recommend the Flexible Interface method (or subtraction method). The Flexible Interface method uses only the interaction nodes that are on the lateral boundary of the excavated volume that translated in large run time and file size reduction.

Thus, for the direct flexible volume method all the nodes that are in the excavated soil volume needs to be defined as interaction nodes (see INT instruction - section ). For the direct flexible interface method only the nodes that are on the lateral surface of the excavated volume have to be defined as interaction nodes (see INT instruction - section ). Both the flexible volume and interface methods involve inversion of a full complex symmetrical matrix as big as 3 x total number of interaction nodes.

For the skin method, all interaction nodes are divided into three different types, namely: interface (nodes by boundary), intermediate (nodes connected directly to interface nodes), and internal (remaining nodes) (see INT instruction - section ). This method involves inversion of full complex symmetrical matrix only as big as 3 x total number of interface nodes, and therefore is considerably faster.

It should also be noted that all the nodes in the superstructure (above the ground surface) are not connected to the soil and therefore are not interaction nodes.

• Acceleration of Gravity - Type the acceleration of gravity. The value is the same as set in the Analysis Options - SITE dialog box.
• Ground Elevation - Type the ground elevation. This value is also used to determine whether an element belongs to the superstructure or to the excavated soil (see ETYPE).
• Input Data - Click this button to enter data for the pile interface (ACS SASSI PINT module is not available).
• Soil Motion - Select the soil motion type from Coherent and Incoherent.

The seismic motion incoherence option applies only to three-dimensional SSI models with no axis of symmetry, i.e. only for full models (not applicable for 2D models or half-models). It is required that the interaction node numbering to start from the bottom layer at baserock up to the ground surface.

The incoherent motion analysis includes different coherency model options: 1) Luco-Wong model (Luco and Wong, 1986), that is a physics-based, anisotropic parametric model with different coherence parameters for X, Y and Z motion components (see input below), and 2) Abrahamson models, 1993, 2005 (for surface foundations) and 2006 (for embedded foundations) (Abrahamson, 1993, 2005, 2006), that are empirical-based models developed based on many earthquake records on different soil conditions, with no input parameters. Luco-Wong model can be applied with or without wave passage option checked. Abrahamson are applied only with wave passage selection checked.

• Coherence Parameter X Direction (Horizontal Component) - Type the coherence parameter value for the Luco-Wong model. This parameter value lies usually between 0.10 and 0.30. Higher values can be used to determine upper-bounds of the incoherence motion effects on SSI response. The incoherent motion field can be isotropic or anisotropic for horizontal components.
• Coherence Parameter Y Direction (Horizontal Component) - Type the coherence parameter for the other horizontal motion component.
• Coherence Parameter Z Direction (Vertical Components) - Type the coherence parameter for vertical motion component.
• Mean Soil S-Wave Velocity - Type the mean soil shear wave velocity. This parameter may be computed by statistical averaging for an equivalent uniform soil deposit.
• Number of Embedment Layers - Type the number of embedment layers defined by the set of interaction nodes. If flexible interface method (direct) is used then the internal nodes should be not defined as interaction nodes. If flexible volume is used then the interaction nodes are all the nodes that define the excavation volume, except some nodes that can be associated with some internal structures that vibrate independent from the excavation soil (not connected with the soil-foundation interface nodes).
• Time Step of Seismic Motion - Type the time step of control motion. The value is the same as set in the Analysis Options - SITE dialog box.
• Nr. of Fourier Components - Type the number of values to be used in the Fourier transform. The value is the same as set in the Analysis Options - SITE dialog box.
• Frequency Set Number - Type the number of the frequency set. The value is the same as set in the Analysis Options - SITE dialog box.
• Print Coherence Functions - Reset this check box to disable printing. Set this option to enable the printing of the computed coherence matrix versus the computed coherence matrix at all interaction nodes. This option is provided to check the numerical accuracy of the random field decomposition, but printed output is extremely large. The printed output will be saved in FILE16.
• Wave Passage - Select this button for activating the wave passage option. This selection is required for Abrahamson models.
• Apparent Velocity for Line D - Type the apparent velocity for line D (if wave passage option is enabled).
• Angle of Line D with X Axis - Type the angle of horizontal line D with the x axis (if wave passage option is enabled).
• Unlagged Incoherency Model - Three incoherence models can be selected: Type 1 for the 1986 Luco-Wong model; Type 2 for the 1993 Abrahamson model for all soil types; Type 3 for the 2005 Abrahamson model for surface foundations on all sites; and Type 4 for the 2006 Abrahamson model for embedded foundations for all sites and Type 5 for 2007 Abrahamson model for hard-rock sites.

Motion Incoherency Simulation

This option is used for simulating the seismic motion incoherency field at the interaction nodes. The user has the options to use either Deterministic (Mean) Input assuming zero phase angles between different motion spatial wavelength components, or Stochastic Input assuming random phase angles in the interval [-180, 180] degrees. If the user selects a pair of a random, arbitrary SEED numbers for the Horizontal and the Vertical components, and a 180 degree angle for the Random Phase angle for different wavelength components, then, a stochastic motion incoherency field is simulated and further used in SSI analysis. If the user selects zero values for the two SEED numbers and the Phase Angle, then a deterministic (median) motion incoherency field is generated and used in SSI analysis.

Because of complex Fourier representation of motion spatial variation, the computed incoherency transfer functions (ITF) are not highly smoothed curves, but more oscillating curves (they represent a convolution between the random spatial variation Fourier amplitude that is different at each frequency and the structural transfer functions). If an user wants to compare these computed ITF functions with those obtained using classical linear random vibration theory based on smoothed PSD inputs, then, the computed incoherent ITF needs to be smoothed (the unsmoothed PSD is called periodogram, and it has uncorrelated neighbor frequency components; typical PSD estimate is based on spectral windowing techniques that introduce a level of statistical correlation between neighbor frequency components). Alternatively, and better form a technical point of view, the user can use a number of stochastic simulations, and then apply statistical averaging to the computed ITF realizations (with or without additional smoothing). To perform quickly several statistical SSI analyses, the ANALYS restart option for New Environment should be used. This ANALYS restart option reduces the SSI run time by a factor of 4 or 5.

• Use Multiple Excitation - Select this button for activating the multiple excitation option.
• Input Motion Number - Select the number of the active input motion. All following data refers to this motion.
• First Foundation Node - Type the number of the first interaction node for the selected isolated foundation.
• Last Foundation Node - Type the number of the last foundation node for the selected foundation. These foundation nodes have to be defined in a sequential node order with unit increment. Thus, it is required that the interaction nodes to be defined in a sequential order for each foundation, so that ranges of node numbers can be defined for each isolated foundation, i.e. do not skip node numbers from one foundation to another foundation.
• X Coord. of Control Point - The X-coordinate is used to define the input motion location application.
• Y Coord. of Control Point - The Y-coordinate is used to define the input motion location application.
• Z Coord. of Control Point - The Z-coordinate is used to define the input motion location application.
• Spectral Amplification Ratios - Type the spectral amplification ratios. This factors define the ratios (at frequency solution points) between the Fourier amplitude of the motion with respect to the input motion. The number of ratios must be equal to the number of frequencies. Use blank, tab, ‘,’, or ‘;’ as separators.

Nonlinear SSI Analysis Input

If the nonlinear SSI analysis is used, then the user needs to click on the Nonlinear SSI Input Data to define the input for the initial soil properties for the near-field soil element groups. By clicking the Nonlinear SSI Input Data a new input file is opened for editing.

This file has extension .pin. The user needs to input in a free-format the following information (see example shown below in the figure):

• 1^st line: Number of nonlinear soil element groups, effective strain factor, number of soil material curves defined in SOIL (soil constitutive model);
• 2^nd line: Number of the nonlinear soil element group, number of materials (could be equal with the number of layers or not) in the group and number of solid elements in the group
• 3^rd line and after define a loop over the number of soil materials, with each line including: Initial shear modulus reduction factor (1.00 indicates same shear modulus as in free-field), initial damping ratio factor (1.00 indicates the same damping as in free-field), soil material curve order number. The block of lines after 1st line, needs to be input for all nonlinear soil element groups. The figure below is shown an example with a single group of nonlinear soil elements, an effective strain factor of 0.60 and 2 soil material curves. The order number of the nonlinear soil group is 2, the number of soil materials in the group is 5, and total number of elements in the group is 180.