Unified First-Principles Ship Structural Design Based On The Maestro Methodology
Unified First-Principles Ship Structural Design Based On The Maestro Methodology
Unified First-Principles Ship Structural Design Based On The Maestro Methodology
Tobin R. McNatt
Proteus Engineering, 345 Pier One Road, Stevensville, MD 21666, USA
INTRODUCTION
The demands of the shipping industry for more efficient, higher speed, lighter weight, and
lower cost ships are strongly linked to the issue of ship structural design. The economic
success and safety of shipping rely heavily on intelligent structural design that optimizes
the use of new materials, improved fabrication procedures, and efficient life-cycle
maintenance to address the current and future trade requirements. Environmental issues
have never been more distinctly involved in the criteria for successful and safe ship
operation. All of these demands place increasing emphasis on the structural design
process. This paper summarizes the value of unifying ship structural design through
combining rapid structural modeling, global as well as local finite element analysis,
comprehensive failure and limit state analysis, and the use of mathematical optimization to
meet these increasing demands. The unification and continued development and extension
of this methodology based on the MAESTRO computer program is presented:
implementation of improved structural limit state evaluations; automation of detailed finite
element meshing and analysis; implementation of composite structural modeling, analysis,
and failure evaluation; improved support for fatigue design; connectivity to preliminary
design tools for hull form and other naval architectural design tools, such as GODDESS;
integration with 3D CAD-based ship product models; and, integration with cost estimating
and production planning resources. These developments represent significant progress
toward unifying first-principles structural design with the total ship design and
construction planning process. Plans are presented for implementing Multidisciplinary
Design Optimization (MDO) within these integrated ship design technologies. For
structurally intensive ships, such as heavy cargo vessels, high-speed ferries, and naval
combatants, these developments offer improved designs, and lower construction and life-
cycle maintenance costs. Examples of designs and applications are presented to illustrate
all of these developments.
McNatt 2
Basic Capabilities
The basic capabilities of the MAESTRO structural design system are summarized below.
These capabilities and the software system are illustrated in Figures 1 and 2.
4 EVALUATION
(A) FORMULATE CONSTRAINTS (B) EVALUATE ADEQUACY
γ 1γ 2γ 3Q ≤ Q L
CONSTRAINTS SATISFIED?
OBJECTIVE ACHIEVED?
YES
MAESTRO
Software Modules
Natural frequency calculations Detailed Stress Analysis Underwater shock analysis via
in-air and in-water interface to UNDEX code
Rapid Structural Modeling - The MAESTRO Modeler is an interactive graphics tool that
enables the rapid creation (typically in days, rather than weeks or months) of a full ship
finite element model. The same model is used in failure analysis, evaluation of structural
adequacy, and structural optimization. Figure 3 presents a representative full ship
MAESTRO structural model of a high speed ferry.
Finite Element Analysis - MAESTRO’s FEA, normally completed for the entire structure,
determines the stresses (symbolized by Q - “Load Effects” in Figure 1) in all structural
members and for all load cases. MAESTRO offers flexible, ship-oriented and highly
automated specification of loads.
Failure Analysis - MAESTRO’s failure analysis begins with the calculation of stresses (QL
in Step 3 of Figure 1 that would cause failure of each member, for all possible failure
modes (yielding, buckling, plastic hinge, etc.) and other design limits (e.g., deflections).
Table 1 lists the 25 failure modes treated at the individual structural member level.
Additional failure modes or limit states are computed at the overall structural level,
typically using the full hull cross section.
Translation to Other Analysis Formats - Other types of analysis are possible because the
MAESTRO Modeler can automatically create files for other programs, such as NASTRAN,
VAST, and/or dynamics programs as noted in Figure 2.
McNatt 5
Z X
PANELS GIRDER
Collapse Collapse
stiffener flexure torsional buckling
combined buckling plastic buckling, flange
membrane yield plastic buckling, plate
stiffener buckling Yield
Stiffener yield compression, plate & flange
compression, plate & flange tension, plate & flange
tension, plate & flange
Plate unserviceability FRAME
yield Collapse, plastic hinge
transverse bending Yield
longitudinal bending compression, plate & flange
local buckling tension, plate & flange
allowable permanent set
pressure loads
concentrated loads
little more than a basic requirement. The SHIPSTRUCT philosophy and its relationship
with GODDESS is shown within the box outline in Figure 5. Feasibility design involves
trying to optimize the ship design by subjecting the possible variants to the following
calculations and assessments:
It is during this feasibility design stage and the subsequent detailed design stage that the
MAESTRO suite of programs can be utilized as a powerful tool in the design of both
commercial and military ships. To this end, the UK’s Defence Evaluation and Research
Agency (DERA) has been incorporating the MAESTRO system into an integrated structural
design package, shown in Figure 6, for the structural design of ships. To fully integrate the
MAESTRO suite of programs into the UK MOD’s ship design suite of programs,
considerable work has been carried out to develop the linkages between
GODDESS/SHIPSTRUCT and the MAESTRO Modeler code, to enable geometric and
structural data to be transferred.
McNatt 9
CAD Systems
GODDESS (ANVIL)
Structural Load
SHIPSTRUCT Prediction Finite Element
Modelers
FEMGV/PATRAN/HYPERMESH
NASTRAN Structural
Analysis Codes
Further work has been carried out on the MAESTRO Analysis code to implement the UK
Warship Design Limit States, detailed in NES 154/SSCP23 (Dow [2,3]), covering
buckling, yielding and failure of structural components. This work essentially provides an
alternative set of limit state evaluations to the 25 failure modes used by MAESTRO to
assess the individual structural members. Making use of these failure modes, structural
optimization can be carried out using MAESTRO to develop a ship structural design based
on the MOD’s warship design criteria. A major application of this software in the UK has
been to investigate the effect of degradation of material properties on the adequacy of large
composite ship hulls. To enable this type of analysis to be carried out efficiently, a set of
failure criteria for composite structural components are being developed by DERA. These
will be implemented in MAESTRO as another alternative set of limit states. Some work in
modifying existing failure criteria has already been carried out, which enabled hull life
assessment of existing composite warships to be accomplished.
Further development work has been carried out on the MAESTRO code to incorporate the
DERA Ultimate Strength Procedure (Smith and Dow [4]) into the MAESTRO software as
an additional Adequacy Parameter to be considered when developing the ship structural
design. These developments represent a significant step towards the integration of the
MAESTRO software package into the MOD design procedure for UK warships.
The MAESTRO system, refer to Figure 2, includes a Detailed Stress Analysis (DSA)
module which facilitates fine meshing local portions of the global structural model. The
DSA Modeler also supports adding structural details and/or modification of global
structure to reflect local details. Boundary conditions are automatically applied using
either a top down or a superelement approach. The top down approach applies
displacements from the global analysis as boundary conditions for the local finite element
analysis. The superelement approach condenses the fine mesh finite element model to the
master nodes of the global model, and conducts the global analysis incorporating the
structural stiffness of the local fine mesh model. An example of a DSA model is provided
in Figure 4, presented previously.
Composite structures pose unique requirements with respect to failure or limit state
evaluation. MAESTRO uses the Tsai-Wu interaction formula for first ply failure and last
ply failure of orthotropic panels, as in paragraph 1B202 of Part 3, Chapter 4 of the HSLC
Rules. The formula uses the strength values Xt, Xc, Yt, Yc, and S which are part of the
input data. These must always be verified by material testing. In the finite element
analysis, the stresses are calculated in every ply of every panel. A check is then made of
each ply for first ply failure and last ply failure by substituting the stresses into the Tsai-
Wu formula and using the factors of safety R given in the Rules. Core shear stresses and
maximum deflections are checked against the maximum permissible values given in the
Rules. The program also checks for panel buckling. For sandwich panels, this includes
local buckling of the skins. These failure capabilities are currently under development.
Fatigue analysis typically requires using a ship motions and loads program (usually strip
theory) to calculate the wave-induced pressures on the ship, the displacements and
rotations at the C.G. of the ship, and the vertical and horizontal bending moments at
selected positions along the ship length. The analysis then uses a large (usually three cargo
hold) finite element model, and involves all three types of cyclic loads:
• external pressures due to waves and ship motions
• internal pressures due to the accelerations of cargo and ballast
• vertical and horizontal bending moments at the ends of the model
The number of load cases is very large, even with a linear, spectral-based unit wave
approach. The number of load combinations is roughly:
• 18 ship-to-wave headings
• 25 wave frequencies
• 5 ship speeds
• 2 or 4 loading conditions (2 for a tanker; 4 for a bulk carrier)
This gives a total of either 4500 or 9000 load cases. Three principal features of MAESTRO
provide major benefits in making this fatigue analysis process more efficient.
McNatt 13
Rapid Global Structural Modeling - The MAESTRO Modeler allows the construction of
a three hold finite element model in from five to seven days, depending on the geometry
(e.g. corrugated bulkheads take longer because of the intersections at the upper and lower
stools). It is the same Global Modeler that is used to build whole ship finite element
models for strength and vibration analysis. Thus in many cases the global fatigue model
will already be available (a MAESTRO model is inherently modular, such that a three cargo
hold model can be extracted from a larger model in a few minutes).
Rapid Local Structural Modeling and Analysis - The MAESTRO Detailed Stress
Analysis (DSA) Modeler performs fine mesh modeling and stress analysis of any portion
of structure, using a library of 24 element types and powerful interactive graphics meshing
tools. Templates allow the rapid modeling of complex details such as corrugated plating,
cutouts and brackets. The DSA Modeler has a smooth interface with the Global Modeler.
It automatically begins with and uses all relevant information from the Global Modeler
(geometry, master nodes, etc.). Likewise it automatically uses the boundary node
displacements from all the unit loads (wave pressures, cargo/ballast pressures and bending
moments) and, if relevant, any cyclic pressures that act directly on the local model. This
smooth interface avoids the laborious and error-prone process of transferring data from a
global model to a separate local model, and it thus improves both the modeling speed and
the integrity of the local model. In a MAESTRO fatigue analysis, the stresses from the
local model are actually Stress Influence Coefficients, and the next point explains how they
allow a much more rapid and efficient fatigue analysis than would otherwise be possible.
Automated Unit Load Analysis Method - This method dramatically reduces the number
of load cases, with no loss of accuracy. For example, for a bulk carrier the number of load
cases is reduced from 9000 to about 1200. The result is that MAESTRO's global analysis
takes only five hours on a standard 133 MHz PC (one time only, regardless of the number
of structural details to be analyzed) and the local analysis takes about eight hours per
structural detail, assuming about 30,000 degrees of freedom per detail. The results from
these MAESTRO analyses (the Stress Influence Coefficients) are the unit wave stresses,
which are needed in a first principles fatigue analysis. The MAESTRO Stress Influence
Coefficients are used to obtain these stresses. At this point, ship motions analyses, such as
strip theory results, must be interfaced with the MAESTRO results. This requires that the
pressures from the strip theory analysis must be allocated or mapped to the MAESTRO
patches. Based on this mapping, the actual stress responses at the structural details can be
generated and used to complete the fatigue analysis.
relationships for the particular structures and their associated fabrication processes.
Structural data in ESTI-MATE can be further transferred to the production planning
modules of the PERCEPTION integrated shipbuilding production management system.
PERCEPTION provides build strategy and work package planning and scheduling,
materials purchasing and inventory management, and labor planning, as well as real-time
feedback against schedules and budgets during the construction process.
DESIGN EXAMPLE
Currently, the UK Ministry of Defence is carrying out design studies on a novel, advanced
trimaran hull for use as a frigate. This design has a long slender main hull fitted with
smaller outriggers to provide stability for the platform. The principal motivator for a
slender hull form for a large warship is reduced resistance, other advantages of the
trimaran hull form are increased deck area and internal volume provided by the wide cross-
deck structure. The three hulled design introduces a range of unknowns for the structural
designer. The main areas of structural risk are:
A concept design of a trimaran warship was produced using CONDES and GODDESS. A
preliminary structural design was then produced using SHIPSTRUCT, and a MAESTRO
McNatt 15
model, shown in Figure 7, was developed. MAESTRO was then used to assess the
adequacy of the preliminary structural design and to develop improved structural
capability. The MAESTRO model of a bulkhead showing the course mesh modeling using
additional beams and panels is shown in Figure 8. Subsequent to the MAESTRO model
being created, analysis was carried out on deck, shell and bulkhead structures to investigate
structural effectiveness, load paths and fatigue stresses using the MAESTRO DSA Solver in
conjunction with NASTRAN. Figure 9 shows a fine mesh model of the bulkhead created
using MAESTRO DSA. This approach enabled fairly detailed design studies to be carried
out at a very early stage in the design/assessment process.
CONCLUSION
This paper summarizes the increasing role that ship structural design is playing in the
overall ship design process, with particular emphasis on defining structure earlier in the
process than has been traditional. Structural design is also incorporating the use of limit
states or failure modes in a more comprehensive way by automating these calculations,
which in the past have required time consuming manual modeling and calculations. The
MAESTRO structural design methodology employs this first-principles approach to unify
rapid structural modeling, finite element analysis at both global and local levels, limit state
evaluation, and structural optimization. Specific aspects of structural design continue to
receive development attention, including structural fatigue analysis and composite
structure design. The unified approach offered by the MAESTRO system is shown to be an
effective design tool with which to implement design criteria for a specific class of ships or
for a specific safety or regulatory/design approval authority’s requirements. In general, the
unified approach to ship structural design improves the quality and efficiency of the design
and the integration of structural design with the overall ship design process.
McNatt 17
REFERENCES
1• Hughes, O. F., “Ship Structural Design,” Society of Naval Architects and Marine
Engineers, 1988.
2• Dow, R., NES 154, “Design Standards for Surface Ship Structures,” Issue 1,
December 1989.
3• Dow, R., SSCP23, “Design of Surface Ship Structures,” Volumes 1 & 2, December
1989.
4• Smith, C. S. and R.S. Dow, “Ultimate Strength of a Ship’s Hull Under Biaxial
Bending,” ARE Report, AMTE(S) TR86204, December 1986.
5• Hughes, O. F., “Two First Principles Structural Designs of a Fast Ferry - All-
Aluminum and All-Composite,” Proceedings of FAST 97 Conference, Sydney,
Australia, June 1997.