You are on page 1of 8

Proceedings of the ASME 2018 37th International Conference on Ocean, Offshore and Arctic Engineering

OMAE2018
June 17-22, 2018, Madrid, Spain

OMAE2018-78090

LARGE SCALE EARTHQUAKE ANALYSIS OF INTEGRATED SUBSEA FACILITIES

Majid Hesar
Subsea 7
Sutton, Surrey, UK

ABSTRACT respectively. These facilities may typically consist of rigid


Subsea facilities located in earthquake-prone regions of the pipelines terminating in sliding PLET structures on skirted or
world can be subjected to severe excitation and have to be hybrid mudmat foundations, connected via rigid spools or
designed in a two-tier manner against both ELE and ALE level jumpers to manifolds on suction pile foundations or X-mas
earthquakes. The return period of these earthquake levels can trees on well conductors cantilevering a few meters above
be 100-300 years and greater than 1000 years, respectively. seabed. Several Flexible flowlines may be hanging off these
These facilities may typically consist of rigid pipelines, sliding structures via goosenecks.
PLET structures on skirted or hybrid mudmat foundations,
connected via rigid spools or jumpers to manifolds on suction DESIGN PHILOSOPHY FOR SUBSEA FACAILITIES
pile foundations, with flexible flowlines hanging off The design document used for seismic design of subsea
goosenecks. facilities is ISO 19901-2, Ref. [1], also known as API RP 2EQ.
In recent practice Subsea 7 have developed a novel ISO 19901-2 is one of a series of standards for offshore
methodology for modelling and simulating the seismic structures. The full series consists of the following international
response of such complex clusters in which the so called standards:
“system effects” are intractable and dominate the response of  ISO 19900 General requirements for offshore
certain critical components. Chief amongst the latter are rigid structures
spools and jumpers that span large distances between PLEM,  ISO 19901 Specific requirements for offshore
PLET and X-Tree structures without touching the seabed. structures
In the nonlinear implicit direct integration dynamic FE  ISO 19902 Fixed steel offshore structures
analyses in Abaqus each of the cluster models in a project is  ISO 19903 Fixed concrete offshore structures
subjected to time history accelerations of seven representative  ISO 19904 Floating offshore structures: Monohulls,
earthquakes, as per ISO 19901-2 requirements. Hysteretic semisubmersibles and spars
damping characteristics of soil-structure interfaces are  ISO 19905 Site-specific assessment of mobile
modelled with special kinematically hardening elements, offshore units: Jack-ups
calibrated to the site-specific seabed geotechnical properties.  ISO 19906 Arctic offshore structures
Specially developed post processing scripts are used to
automatically extract the vital information from the large At present there is no specific code for the design of
amount of data produced and perform the unity checks of subsea structures, although some clients have their own
various components to their respective codes of practice. detailed internal specifications. In current earthquake design
codes and standards a two-tier design approach is adopted. The
INTRODUCTION system must satisfy two distinct Earthquake level criteria,
Subsea assets located in earthquake prone regions of the namely:
world can be subjected to severe excitation. According to the
 ELE: Extreme Level Earthquake (probability of
ISO 19901-2 standard, Ref. [1] these assets need to be designed
occurrence, every 100-300 years)
in a two-tier manner against both ELE and ALE level
 ALE: Abnormal Level Earthquake (probability of
earthquakes. The return period of these earthquake levels can
occurrence > every 1000 years)
be typically 100-300 years and greater than 1000 years,

1 Copyright © 2018 by ASME


SITE RESPONSE ANALYSIS
According to the ISO 19901-2 code for each of these the Once the accelerations arriving beneath the site within the
performance criteria is stated as follows: stiff soil or rock are developed the actual soft soil response at
mudline is determined by relatively simple 1D or 2D models
 ELE: The structure should withstand the seismic event consisting of a series of springs, masses and dashpots, whose
with little or no damage and remain functioning. constants are tuned to represent the soil stratigraphy of the
 ALE: The structure should endure the event without actual site, see Fig. 1. Using such models the acceleration time
suffering complete loss of integrity or major histories at other elevations below the mudline can also be
environmental impact, but members may undergo determined. These are necessary to simulate the input of
plastic deformation. ground motions onto, for example, deep foundations, well
conductors or skirts. An example of the set of orthogonal
Two alternative procedures for seismic design are stated in acceleration time histories at mudline is shown in Fig. 2.
the ISO 19901-2 code, as follows:
 A Simplified procedure: where seismic considerations
are unlikely to govern design of the structure. In these
cases it is proposed that generic seismic maps for
offshore regions of the world can be used.
 A Detailed procedure: where seismic considerations
have a significant impact on design of the structure.
These cases require a site-specific seismic hazard
study to be performed.

Selection of the appropriate procedure depends on:

 Exposure Level of structure


 Expected Intensity of seismic event
 Characteristics of seismic event

For the integrated subsea facilities which are the subject of Figure 1 Site response analysis model
the present paper, clearly a simplified procedure is out of the
question. Hence a site-specific Probabilistic Seismic Hazard
Assessment (PSHA) has to be performed in advance of the
project detailed design stage in order for the appropriate ground
motion data to be available in time.

PROBABILISTIC SEISMIC HAZARD ASSESSMNET


(PSHA)
Detained discussion of this topic is outside the scope of
this Paper, but the major steps in a PSHA can be summarised as
follows:

 Identify main seismic sources in the area that can


affect the site
 Define their seismic activities using catalogues and
historical records
 Create a seismo-tectonic model
 Use Ground Motion Prediction Equations (attenuation
relationships)
 Develop hazard curves (Probability of exceedance
versus spectral acceleration)
 Construct Uniform Hazard Spectra (UHS)

Figure 2 Set of orthogonal acceleration time histories.

2 Copyright © 2018 by ASME


mercy of the structural modal response of these long slender
Acceleration time histories are typically produced at time pieces. Abaqus software is used throughout, Ref. [2].
intervals of 0.01 or 0.02 seconds. And typical signals from
earthquakes can be 25 to 100 seconds long. SOIL-STRUCTURE INTERACTION MODELLING
Depending on the position of the dominant faulting A central aspect of any earthquake analysis is the
relative to the location of the offshore field directionality of the modelling of interaction between structure and soil. Input
input accelerations may have to be considered. However, the accelerations predicted from seismo-tectonic models, whether
ISO 19901-2 code allows arbitrary orientation of input in the form of spectral or time history accelerations, are free-
accelerations relative to the global axes of the subsea system field accelerations i.e. the accelerations of soil medium at
model. mudline in the absence of any structure or pipeline. As such
In order to account for the random nature of earthquakes these accelerations cannot be imposed directly onto the
the code states that the facility must be analyzed against at least structure. At the interface between the structure and soil body
seven earthquake time history records. However, if the energy is dissipated in the form of hysteretic and viscous
structure passes the first four earthquakes augmented by a damping.
factor of 1.05 then the rest is not necessary. A further important consideration is that it is not always
possible to model the structure supported on a single set of soil
SYSTEM ARCHITECTURE AND ANALYSIS TYPES springs centred on foundation geometric centre, as is often
Types of seismic analysis can be broadly classified into practiced in reported literature. (Note that in this Paper the term
three categories as follows: “spring” is used to refer to any soil-structure interaction
 Quasi-static Analysis, which is suitable for smaller elements, even though they may represent elasto-plastic and
items securely fastened to larger structure e.g. valves hysteretic behavior). This would be fine if the main point of
and instrumentation. concern was away from the foundation elements themselves
 Response Spectrum Analysis, which is typically used and the foundation itself was not the main focus of the analysis.
for ELE seismic design of structures where they are However, if it is desired that the local stresses on foundation
expected to remain within the elastic limit. It uses a elements, such as mudmat base plates and hence the sliding
square root of sum of squares approach to combine rails be obtained from the same global analysis, then the soil
forces from various mode shapes. The scope of this pressures on these elements need to be modelled in a
method for subsea structures is somewhat limited. distributed manner. The more the soil springs are distributed
 Time History Analysis, where a time integration the better will the stress distribution prediction.
scheme is used to solve the dynamic equations of Ideally a large volume of the soil medium itself in which
motion throughout the time steps defining the the integrated facility being considered is situated should also
earthquake signal. be included in the model and the free-field input accelerations
should be applied to the far field boundaries of the soil
Whenever there are individual but interconnected medium. In this way the task of taking the important soil-
structures the behavior of the whole complex becomes difficult structure interactions into account would be automatically
to determine from a knowledge of the response of individual taken care of.
components in isolation. In other words the so called “system It should be remembered that the structural parts of all the
effects” become intractable. In such circumstances the only facilities in the model still need to be present and meshed in
way to determine the true response of all parts of the system is sufficient detail to be post-processed directly and code checked.
to use an integrated model and adopt a time history analysis With such an advanced soil model the large volume of 3D soil
methodology. Experience shows that although the magnitude elements requires a massive amount of computer processing
of the ALE earthquake may be moderate, the architectural form power and is currently prohibitively expensive, even with
of these integrated subsea systems makes some parts of them current computer technology available within typical offshore
vulnerable to sustaining inelastic deformations due to system oil and gas projects.
effects. Therefore, it would be naive to dismiss the seismic Furthermore, interference of the seismic waves reflected
effect of such apparently low level of shaking. off the boundaries back into the soil volume would invalidate
In the present work a novel methodology has been the correct functioning of the simulations. 3D Infinite elements
developed for modelling and simulating the seismic response of with non-reflecting (silent) far field boundaries could be used
such complex subsea clusters in which system effects not only to overcome this problem but also greatly reduce the
potentially dominate the response of certain critical volume of soil required to be present in the model, e.g. see Ref
components. Chief amongst these components are rigid spools [11]. Even so to obtain meaningful results an advanced and
and jumpers that span large distances between PLEM, PLET elaborate cyclic soil constitutive model would need to be
and X-Tree structures without touching the seabed. In employed which would take account of the hysteretic
particular, the hubs connecting the extremities of these spools behaviour of soil, e.g. see Refs [12]. An advanced soil sampling
or jumpers to host structures are the weak points that are at the and laboratory testing plan would need to be programmed very

3 Copyright © 2018 by ASME


early into the project schedule in order to help calibrate the obtaining static load-displacement responses for each DOF in
cyclic degradation and damping characteristics of such a turn in a separate geotechnical model, e.g. see Fig. 4. They are
sophisticated soil model. then distributed in a manner that takes account of the
This Paper describes the next best alternative, which is not interaction of loading responses in different DOF direction e.g.
to include any 3D soil elements but to take the soil-structure Feng, et.al., Ref. [5]. Of course attention has to be paid to the
interaction effects into account in a distributed manner, while combination of the set of design static loads.
still applying the free-field accelerations. This approach
drastically reduces the computational effort required and yet
maintains the essential elements of the simulation. The only 3D
soil elements used are within the soil plug region of mudmat
foundations, but these use a simple Mohr-Coulomb soil
constitutive model. Their main purpose is to model the
presence of soil’s mass for dynamic purposes and its stiffness
for the purpose of obtaining the right stress field on either side
of the skirts and base plate.
In this way the problem reduces to one of finding the
appropriate distributed non-linear soil-structure interaction
“springs” which are not only equivalent to the global static
force-displacement response of each structure (e.g. a PLET
foundation), but also carries the appropriate damping
characteristics. In the present work the hysteretic damping
characteristics of soil-structure interfaces are modelled with Figure 4 Static geotechnical model to derive foundation
connector elements with kinematically hardening load-displacement response springs.
characteristics, calibrated to the site-specific geotechnical
properties. An example is schematically illustrated in Fig. 3. The result of this process is to end up with different parts
of the foundation carrying different spring intensities, the
corners having a greater intensity than the middle sections, see
for example, Fig. 5.

Figure 3 Distributed equivalent soil springs


surrounding a mudmat base.

Connector elements are also used for other components in


contact with seabed. For example the soil-structure interaction
elements supporting the X-mas tree well conductor use the
shape of conductor P-y curves reported by Jeanjean et.al., Ref. Figure 5 Distribution of foundation spring intensities.
[3] as the backbone curves, or the pipe-soil interaction (PSI)
connectors use the shape of axial, lateral and vertical PSI Quantitative definition of hysteretic damping is given in
curves, from Safebuck JIP, Ref. [4]. standard literature, e.g. Kramer, Ref. [13].
Radiation damping effects are modelled by use of linear
CONTROL OF SOIL DAMPING dashpots. In any case the total damping has to be limited to
It is important to account for the DSSI effects. Both about 20%, according to Ref. [1].
hysteretic damping at soil-structure interface and radiation
damping due to viscous effects within soil mass have to be ADDED MASS EFFECTS
accounted for. Hysteretic damping is inherently taken care of A specific feature of dynamic simulations for subsea
by virtue of the use of non-linear backbone curves for structures is the fact that everything above mudline is
distributed soil-structure springs. These are produced by first submerged and hence attracts additional forces resulting from

4 Copyright © 2018 by ASME


added mass effects. The added masses depend on the shape and stability of such an explicit simulation are orders of magnitude
size of structural elements and vary depending on the direction smaller than that needed by an implicit scheme, although in the
of acceleration of the member in question. The appropriate explicit scheme the formation of global system stiffness matrix
characteristics of every part of the model have to be calculated is avoided. In the present simulations an implicit dynamic
and data prepared for Abaqus to implement during the analysis. scheme is used with the minimum of time increments required
These calculations are generally based on the guidelines in while maintaining analysis accuracy, see for example Bathe,
DNV, Ref. [6]. Ref. [7].
The input acceleration time histories are typically defined
INTERFACING WITH THIRD PARTY STRUCTURES at time intervals of 0.01 to 0.02 seconds. And typical
Typical offshore projects involve multiple contractors, earthquake signatures can range from 25 to over 100 seconds.
some of which supply substantial specialist or proprietary parts In order to ensure that accuracy is maintained and numerical
of the system e.g. manifolds, x-mas tress, etc. often the main convergence problems are avoided, it is found necessary to use
contractor responsible for seismic design of the whole system time increments about an order of magnitude smaller than
does not have full details of these third party structures, or the these.
design of the component is not developed sufficiently in time to
be included in detail in the global system model. ORGANISATION OF MODELS
It is therefore, necessary to build global models with these In a typical subsea project there may well be several
structures as intelligent “black boxes”, meaning that the black clusters of facilities each to be modelled separately. The
box has the overall geometry and positions of the interface organisation of model building and distribution of the work
points with other parts of the system e.g. spool, jumper or in order to meet the tight deadlines is a challenge within a fast
flexible gooseneck connection hubs. It also carries the vital track project. The overall duration to complete the modelling,
dynamic characteristics of the structure it represents, such as analysis and reporting for each model depends on the
the six DOF foundation hysteretic springs, structural masses at complexity of the model and is typically assigned to a single
the CoG, directional added masses, etc. engineer.
It is common practice for the owner of the black boxes to The strategy adopted in the present work is to have one
carry out their own seismic design simulations. This requires engineer responsible for creating and housekeeping of the basic
the interface loads to be passed from the main global model to components common to all models, as well as components
the black box model. Close collaboration in a timely manner is unique to each model. Once the structural engineers have
essential for this interfacing to be successful. effectively finalised their design, their CAD drawings are
converted into a finite element model by use of standard
conversion software and imported into Abaqus CAE. All these
CHOICE OF DYNAMIC ANALYSIS SCHEME components are kept within a single master CAE file like a
Although in structural dynamic work it is customary to opt repository for all the Parts.
for frequency domain analyses on the grounds of reduced By combining instances of these parts the different
processing cost, the presence of non-linearities in the model geometric models are all created by the house keeping
preclude this approach. Soils under cyclic loading conditions engineer, meshed and handed over to individual engineers
always exhibit highly nonlinear and hysteretic behavior due to responsible for each model. In this way uniformity amongst
their particulate nature. Other sources of nonlinearity include models is assured and great time saving is achieved. The
viscous dashpots representing radiation damping in soil individual model engineers then complete their own model and
medium and friction present in certain parts of the model. input of all the manual data such as added mass, pressure,
These nonlinearities rule out the class of modal superposition temperature, density, damping, etc.
methods and call for a direct integration time domain analysis This conveyor belt approach and maintenance of
approach, where the equations of motion are integrated in their uniformity across the models is not only beneficial from an
time domain directly without any modification and the passage efficiency point of view, but is a necessity for the purpose of
full acceleration time history through the system for each unifying the post-processing and reporting effort.
earthquake is simulated by incrementally marching through
time.
The choice of Implicit versus Explicit dynamic analysis is AN EXAMPLE SYSTEM MODEL
also dictated by the model characteristics i.e. by the density, A typical integrated subsea facility system model is shown
stiffness and smallest element size. The presence of such in Fig. 6. The facility consists of a manifold in the centre of the
largely contrasting stiffnesses between steel and soil-structure cluster which is a third party “black box”. This is connected to
interaction elements and 3D soil elements within the soil plug three structures by means of rigid spools which do not touch
means that the steel material would dominate the calculation of the seabed, a 16” PLET, a 10” PLET and a X-mas Tree
maximum stable time increment for an explicit dynamic Structure. Each of the PLETs is made up of a skirted
simulation. The time increments required for numerical foundation unit locked onto the seabed by full penetration of its

5 Copyright © 2018 by ASME


skirts and entrapment of a soil plug. The mudmat base plate is Specially adapted Python scripts have been written to open
fully in contact with the soil and carries two sliding rails. The the Abaqus output data base, extract the appropriate data and
frame which carries the heavy superstructure (64te) sits on the perform the code checking of different parts of the system
rails on four sliding shoes. model against their relevant codes of practice. As an example,
the following time history plots are produced:
1. Total reaction forces on PLET foundation springs (i.e.
PLET weight gravity check)
2. PLET foundation mudmat embedment
3. PLET sliding frame relative axial displacements
4. Vertical and lateral reaction forces on sliding rail from
sliding shoes
5. Anchor flange force (PLET to pipeline connection)
6. Spool connector hub reaction forces
7. Piping support reaction forces
8. Manifold suction pile displacements and rotations
9. Manifold CoG displacements and accelerations
Figure 6 Perspective view of example system model
10. X-mas Tree CoG displacements and accelerations
The primary function of the sliding mechanism is to
For certain parts of the model contour plots of key
accommodate the expansions and contractions of the pipeline
variables are more constructive. For example, for steel PLET
connected to the superstructure via the anchor flange during
structures a contour plot of the envelope of plastic strain
normal operation. In order to limit the movements of the sliding
throughout the time history analysis is a very useful visual
top part stoppers are provided at +1.5m (Expansion) and -0.5m
check for ELE analyses. If at any part of the model at any
(Contraction) limits relative to the locked as-installed position
instance of time during the earthquake stresses have exceeded
on the rails. The spool ends are connected to structures by
yield, these can be visually seen at a glance and from the plot
means of special joints called CVC hubs (Cameron Vertical
legend. The following Code Checks can be performed:
Connector), and during installation a special running tool is
1. API code check, e.g. CHS members design check as
used to align and force the spool end into place before locking
per API RP-2A WSD, Ref. [8]
the CVC hubs.
2. AISC code check, e.g. SHS, RHS and I-section
During operation the pipeline expands due to pressure and
members check as per AISC 360-05
temperature rise and depending on the effective axial force
3. DNV-LCC code check, e.g. Pipelines and spool
generated it may overcome the frictional resistance between the
straight linepipe load controlled condition check
shoes and the rails and the sliding frame may move to a new
(LCC) as per DNVGL ST-F101, Ref. [9]
equilibrium position.
4. ASD code check, e.g. Spool induction bend allowable
A series of static analysis steps have to be performed prior
stress design check as per DNVGL ST-F101, Ref. [9]
to the start of earthquake dynamic step in order to establish the
5. ASME code check, e.g. Piping stress check as per
nominal operating conditions within the model. These include
ASME B31.8, Ref. [10]
activation of gravity, raising and joining of flexible pipe ends to
6. PLET mudmat Von Mises stress check.
goosenecks in order for the correct catenary shape to develop,
application of pressure and temperature, etc.
TYPICAL OUTPUT
POST PROCESSING
Please note that the Figs in this section are provided for
As discussed above, earthquake signals can be more than
illustration only, therefore lack of clarity of actual numerical
100 seconds long and defined with time increments of 0.01
values, which helps protect project confidentiality, should not
seconds. With the Abaqus analysis time increments at 0.001
be of concern.
second, this means that up to 100,000 analysis increments are
In Fig. 7 the contours of accumulated plastic strain
performed. With such a large number of analysis frames on
throughout an ELE earthquake analysis is presented. It can be
such large models the output databases can be unmanageably
seen at a glance that no plastic strains have developed, giving
large, hence the output request has to be carefully tailored and
re-assurance that the stresses within the whole have remained
minimized, while at the same time maintaining the essential
within the yield limit. However, this is not a guarantee that the
data required for full extraction of essential output. Given the
system passes the ELE, as for example, the CVC hub bending
fact that at least eight runs are required for each model
or torsional capacities may have been exceeded and need to be
(minimum of 4 earthquakes each at ELE and ALE), the post-
further examined.
processing task has to be automated.

6 Copyright © 2018 by ASME


Figure 7 Accumulated Plastic Strain (legend values all zero)

Lateral and Vertical reaction forces in the four sliding


shoes of the 16” PLET are shown in Figs. 8 and 9, respectively.
Figure 10 Typical lateral soil spring acting on the face of
This type of information is passed back to the structural
skirt
designer of the PLET for confirmation and verification.
The python post-processing script extracts the relevant
stresses, etc. and performs the code checking of groups of
different types of elements in the model to their relevant codes
of practice, as listed in the previous section. In order to
visualize this output, the script writes the code check values
back into the Abaqus output database as new user-defined
variables, which can then be contoured, just like any native
Abaqus output variable. Some examples of these code check
contours are given in Figs. 11 to 13.

Figure 8 Lateral reaction forces at 16” PLET’s shoes on


sliding rail system.

Figure 11 API code check – frame tubular.

Figure 9 vertical reaction forces at 16” PLET’s shoes on


sliding rail system.

Fig. 10 (inset) presents the response of a typical lateral soil


spring (highlighted in red) together with its input backbone
curve. The hysteretic response of the soil spring can be seen to
match the input behaviour well.
Figure 12 DNV-LCC code check – spools/Jumpers

7 Copyright © 2018 by ASME


PSHA Probabilistic Seismic Hazard Assessment
PSI Pipe Soil Interaction

ACKNOWLEDGMENTS
The author is grateful to the management of Subsea 7 for
permission to publish this work. Any opinions and views
expressed are those of the author and not of Subsea 7.
Figure 13 Von-mises stress header piping system of PLETs.

This paper illustrates that large-scale simulation of subsea REFERENCES


systems is not only possible but also necessary to be carried out 1. Seismic Design Procedures and Criteria for Offshore
in an integrated manner. Integration of the subsea clusters into Structures, ISO 19901-2: 2004. Also known as API RP
a single model simulates reality better than conventional 2EQ.
approaches and removes the need for major simplifications or 2. Abaqus/Standard 6.13-3, Dassault Systemes Simulia Corp.,
sub-modelling. They capture the “system effects” which would Providence, RI, USA.
otherwise be masked if different units were to be modelled 3. Re-assessment of P-Y curves for soft clays from centrifuge
separately. testing and finite element modeling. OTC-20158-MS.
Realistic modelling of dynamic soil structure interaction Offshore Technology Conference, Houston, USA.
effects (DSSI) is essential for the success of any time history 4. SAFEBUCK JIP - Observations of Axial Pipe-soil
analysis, and the elements used in Abaqus for this purpose Interaction from Testing on Soft Natural Clays. OTC-
prove capable of doing this very well. They allow a relatively 21249-MS. Offshore Technology Conference, Houston,
simple and yet very effective means of negating the need for USA.
inclusion of a large number of 3D soil elements. This would 5. Effect of a surficial crust on mudmat capacity under three-
make it impractical to use within projects when multiple dimensional loading, Feng, X., S. Gouvernac, M. F.
number of earthquakes have to be simulated on each cluster Randolph, R. Wallerand and P. Dimmock, 2015,
model. Géotechnique, Vol. 65, No. 7, pp. 590-603.
Post-processing of the results is a major challenge and a 6. DNV-OS-H204, Offshore Installation Operations, Nov.
vital part of such large-scale simulations. In-house Python 2013.
scripts efficiently sift through the massive output databases and 7. Finite Element Procedures, Klaus-Jurgen Bathe, 1996,
produce colour contour plots of the envelopes of key output Prentice Hall.
parameters, as well as code checking of every member against 8. API RP 2A WSD, American Petroleum Institute, “
its relevant Code of practice. 9. DNVGL-ST-F101
10. ASME B31.8
11. Absorbing boundary conditions for seismic analysis in
ABBREVIATIONS ABAQUS. A.H. Nielsen, ABAQUS Users’ Conference
ALE Abnormal Level Earthquake 2006.
CAD Computer Aided Design 12. An anisotropic, critical state model for soils subject to
CAE Complete Abaqus Environment cyclic loading. Z. Mroz, V.A.Norris, O.C.Zienkiewics.
CoG Centre of Gravity Geotechnique, Vol 31, Issue 4, Dec. 1981. PP. 451-469.
DNV Det Norsk Veritas 13. S.L. Kramer, Geotechnical Earthquake Engineering.
DOF Degree of Freedom Prentice Hall, 1996.
DSSI Dynamic Soil-Structure Interactions
ELE Extreme Level Earthquake
PLEM Pipeline End Manifold
PLET Pipeline End Terminations

8 Copyright © 2018 by ASME

You might also like