Revisiting The Ship Design Spiral 365b2w

A Farewell to the Design Spiral Horst Nowacki, Technische Universität Berlin Invited Note Presented at the Mini-Symposium on Ship Design, Ship Hydrodynamics &Maritime Safety Athens, September 30, 2016 30 September 2016

1

Contents 1. Introduction

2. Early History of the Design Spiral 3. Innovation in Ship Design Methodology

4. Modern Concepts of Ship Design 5. Comparison Sequential vs. Concurrent Engineering 6. Achievements and Contributions 7. Conclusions and Congratulations

References 2

1. Introduction My note will actually combine three farewells: Farewell to the design spiral (from sequential to concurrent design)  Farewell to 20th century design methodology (abandoning single ship viewpoint)  Farewell to Prof. Papanikolaou after more than 30 years of research and teaching at NTUA (1985-2016) 

3

1. Introduction (cont.)

I claim the design spiral is an incomplete, inflexible, misleading and hence obsolete paradigm of the design process. It ought to be replaced by a systems analysis based, concurrent engineering, team work oriented paradigm.

4

2. Early History of the Design Spiral: Evans (1959) J. Harvey Evans: 1958, 1959, 1963, [1] ,[2], [3]. „Optimized Design of Midship Section“. Spiral depicts sequential, cyclical decision process . Purpose: Optimal, low weight midship section structure. Basis: Class rules. At every intersection of a spoke with a spiral a calculation task is performed to update the structure. Result: Low weight, feasible design.

5

2. Early History of the Design Spiral: MarAd (1965) Murphy, Sabat, Taylor, MarAD, 1965: „Low cost ship design.“ Scope: Preliminary design cycles up to reliable cost estimate. Sequential, cyclical approach in rigid sequence. Univariate systematic variation with constraints. Rules based. Result: Single feasible ship with lowest building and fuel cost.

6

2. Early History of the Design Spiral: BSRA Hurst, Buxton, BSRA, 1991/92. Design spiral for preliminary design up to shipyard tender. Elaborate sequence of design steps. Basis of software system BRITSHIPS. Sequential, cyclical approach with increasing complexity. Result: Single feasible ship with economically reliable tender.

7

2. Early History of the Design Spiral: SDC (1980) Ship Design and Construction, Kiss, Taggart (1980) Gale, Lamb (2003) Both editions drawn from Evans. Sequential, cyclical approach. Preliminary design up to contract design. Subdivision into 4 design phases.Result: Single feasible design of low cost. Used in parametric design studies.

8

2. Early History of the Design Spiral: Comparison Design Spirals No.

Author(s)

Date

Criterion

Scope

Basis

Objective

1

J.H. Evans [1], [2], [3]

1959

Weight or Cost

Midship Section Structure

Class. rules

Cheapest or lightest structure

2

MarAd (Murphy et al.) [4]

1965

Least Cost

Prelim. Design

Rules

Least cost ship

3

BSRA (Hurst, Buxton) [5], [6]

1971

RFR, NPV, Tradeoffs

Design up to tender

Rules

Best tenders

4

SDC (Kiss, Gale) [7], [8]

1980

Cost

Prelim. Design

Rules

Economic, feasible design

5

Papanikolaou [9]

2014

Same as in No. 4

From rules to goals

Basis for optimization 9

3. Innovation in Ship Design Methodology Innovative Elements: Systems Engineering (1950) • Computer Aided Design (1965) • Optimization (1965) • Economization (1967) • Parametric Design (1970) • Discretized Analysis Tools (1980) • Concurrent Engineering (1985) • Probabilistic Risk Analysis (2000) • Globalization (2000) • Open Systems Communication (2000) •

10

3. Innovation in Ship Design Methodology (cont cont.) .)

R = Requirements S = Solution Space D = Design variables P = Parameters M = Measure of merit C = Constraints The Systems Approach 11

4. Modern Concepts in Ship Design The Product Model • The Process Model • The Concurrent Engineering Approach •

12

4. Modern Concepts in Ship Design: The Product Model The Product Model contains all the data describing the ship in its design, production and potentially operating phases (Life Cycle Product Model). The Product Model is built up step by step by synthesis actions and is analyzed at each stage by analytical methods. Thus it is gradually completed. The sequence of steps in developing the product model is in principle arbitrary. Whenever the information is incomplete for performing a certain step, it must be substituted by provisional, assumed or approximated data. Thus the product model is open to overlapping concurrent engineering actions. 13

4. Modern Concepts in Ship Design: The Process Model

The Product Model ought be supplemented by a Process Model, which describes all the activities that lead to the completion of the product and potentially in its operation throughout its life. The Process Model is formulated in of the network of activities in the product lifecycle or at least to its delivery. This activity model is expressed by a set of transactions and events.

14

4. Modern Concepts in Ship Design Design:: Concurrent Engineering

On the other hand parallel work on related design phases requires close control of mutual dependencies, careful coordination and good team work. Such close control causes extra efforts and cost. Thus labor cost may not diminish.

In Concurrent Engineering the activities of several team may deal with several overlapping phases simultaneously and need not await the completion of the preceding phases. See Figure. Whenever information from other phases is missing, it may be substituted by provisional assumptions and approximations. This saves time toward earlier completion. Earlier product delivery is a competitive advantage. 15

5. Comparison of Sequential vs. Concurrent Engineering Trends in Ship Design Methodology (ca. 1970-2016) Subject

Approach ca. 1970

Modern approach

Paradigm

Design spiral paradigm

Systems paradigm

Sequence

Consecutive steps

Overlapping design phases

Design process

Sequential engineering

Concurrent engineering

Responsibilities

Single persons, one after another

Several persons in design team

Regulatory basis

Rule based design

New IMO approach

Design logic

Rule based design

Goal based design (IMO)

16

5. Comparison of Sequential vs. Concurrent Engineering (cont (cont.) .) Trends in Ship Design Methodology (ca. 1970-2016) Subject

Approach ca. 1970

Modern approach

Decision basis

Semi-empirical

Rational

Uncertainty Treatment

Deterministic model

Probabilistic model

Treatment of Parameters

Several univariate es

Multivariate continuum

Solution search

Trial and error

Optimization with constraints

Criteria

Single objective function and multiple constraints

Multiple measures of merit and constraints

Scope of goals

Favorable tender by builder

Owner’s life cycle cost

17

6. Achievements and Contributions Ship design in the past three or four decades has experienced many new achievements. Prof. Papanikolaou and his team at NTUA have made significant contributions to most of them. To mention only a few top areas of innovation: In design methodology:  Economic efficiency (quantification)  Safety and risk assessment (probabilistic models)  Rationality (measures of merit, cause and effect)  Optimality (unified format, multimodality, MCO)  Versatility (greater scope and depth, many new ship types)  ing for environmental safety  Risk Based Ship Design 18

6. Achievements and Contributions (cont.) In computer technology:  Integration (neutral interfaces)  Open communication (data sharing and exchange)  Longevity (neutral archiving)  Man-machine interaction (simulation, visualization)  Process management (modeling of time dependence, queueing models, process optimization)  Life cycle services (Product Lifecycle Management)  Probabilistic modeling of risks and uncertainties  Simulation and queueing theory 19

Literature 1.

J. Harvey Evans: “A Structural Analysis and Design Integration with Application to the Midship Section Characteristics of Transversely Framed Ships”, Trans. SNAME, vol. 66, 1958. 2. J. Harvey Evans: ”Basic Design Concepts“, J. American Society of Naval Engineers, 1959. 3. J. Harvey Evans, D. Khoushy: “Optimimal Design of Midship Section”, Trans. SNAME, 1963. 4. R. Murphy, D.J. Sabat, R.J. Taylor: “Least Cost Ship Characteristics by Computer Techniques”, Marine Technology, SNAME, April 1965, 5. R. Hurst: “Advanced Techniques in Ship Design and Construction”, Lloyd’s List, December 1971. 6. Ian L. Buxton: “Engineering Economics and Ship Design”, British Ship Research Association, Wallsend, 2nd ed., 1976. 7. Ronald K. Kiss: “Mission Analysis and Basic Design”, 1st ch. in Robert Taggart, ed., “Ship Design and Construction”, SNAME, New York, 1980. 8. Peter A. Gale: “The Ship Design Process”, Ch. 5 in Thomas Lamb, ed.: “Ship Design and Construction”, SNAME, Jersey City, NJ, 2003. 9. Apostolos D. Papanikolaou: “Ship Design”, Springer Science + Business Media, Dordrecht Heidelberg New York London, 2014. 20

7. Conclusions and Congratulations Ship design in the course of five decades of computer use has become more rational, more systematic and analytical, more transparent in its justifications. The quality and safety of the product ship have benefitted from this. Ships have remained a foundation of global trade and of worldwide standards of living. Much work lies ahead of us. The past record of marine design and Computer Aided Ship Design encourages us to look confidently ahead at future challenges to be faced. 21

7. Congratulations Today we are together here to express our gratitude and appreciation to Prof. Papanikolaou, also on behalf of his colleagues, team mates and students, for his personal contributions and his strong commitment to ship design! Congratulations and thank you, Apostolos!

Computer Aided Flower (Bézier)

22

23