Mechanism and Machine Theory 125 (2018) 52–79

Contents lists available at ScienceDirect

Mechanism and Machine Theory

journal homepage: www.elsevier.com/locate/mechmachtheory

Creativity in design–Science to engineering model

J.S. Rao
Chief Editor Journal of Vibration Engineering and Technologies, Chief Science Officer (Consulting) Altair Engineering, Bangalore, India

a r t i c l e i n f o a b s t r a c t

Article history: This paper is written in honor of my long term friend and mentor Professor Bernard Roth
Received 25 July 2017 who spent over six decades in Academic contributions devoted to Kinematics, Dynamics,
Accepted 29 August 2017
Control, and design of computer controlled mechanical devices. He is continuing actively
Available online 19 September 2017
in his work from the famous school Hasso Plattner Institute of Design cofounded at Stan-
ford (popularly known as d.school). As an Academic Director he is active in Creativity in
Keywords:
Design Design. This paper explores the idea of Creativity directly from Science of 17th–18th cen-
Creativity turies in place of Engineering evolved by Professor Timoshenko at Stanford in early 20th
Science to engineering century at the time of rapid expansion of rotating machinery from de Laval in 1882. An
Fusion reactor example of design that involves extreme temperature ranges, magneto hydrodynamics and
Liquid Lithium Ceramic Breeder optimization in a fusion reactor to produce tritium fuel will be described.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Engineering is defined as “The creative application of scientific principles to design or develop structures, machines,
apparatus, or manufacturing processes, or works utilizing them singly or in combination; or to construct or operate the
same with full cognizance of their design; or to forecast their behavior under specific operating conditions; all as respects
an intended function, economics of operation and safety to life and property.”, see Cooperrider [3]. Design definition begins
with creative application of scientific principles.
The Paleolithic, or Old Stone Age, was the longest that began about 2 million years ago, when stone tools were first used
by humanoid creatures, and ended with the close of the last ice age about 13,0 0 0 BCE, see Rao [14]. Around that time man
found that a section of the tree trunk was able to move under the force of gravity because it was round. If the branches
and twigs of the trunk were removed, the speed of the rolling log improved, see Fig. 1. This is probably the beginning
of creativity in Design. Rao [14] also talked about Hero’s steam turbine, water turbines and wind turbines amongst others
much before science revolution. The scientific thought led to atmospheric or vacuum engines and James Watt made the first
real application of Black’s latent heat discovery to usher the industrial revolution, see Ogg [11]. The work of James Watt, see
Rao [12] is thus the application of science to engineering which led to the birth of industrial revolution.
Science Revolution began in 17th century and defined Science of Solids, Fluids, Electromagnetism …, but was too com-
plex in pre-digital era involving several coupled partial differential equations of the state variables. The reciprocating steam
engine of 18th century ushered the industrial revolution but did not warrant any design principles as we know in 20th
century. It is the first practical rotating machine of de Laval and the dynamo of Alva Edison in 1882 that changed the face
of the world, see General Electric [6] and Rao [14]. This also required a major revision in the way in which design is taught
in our engineering schools.

E-mail address: tvii2k@yahoo.com

http://dx.doi.org/10.1016/j.mechmachtheory.2017.08.023
0094-114X/© 2017 Elsevier Ltd. All rights reserved.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 53

Fig. 1. Man’s earliest creativity achieved to pull and push under log rollers.

Professor Stephen Timoshenko wrote the first version of his famous Strength of materials textbook during 1907 to 1911 at
Kyiv Polytechnic Institute in Ukraine. He moved to US in 1922 and joined Westinghouse, an exciting place to be during the
1920s. This was a period of industrial growth and expansion providing challenges in Mechanics where Timoshenko became
the most influential person. The five years that Timoshenko spent at Westinghouse have been called its “Golden Era of Me-
chanics”. His book on Strength of Materials was written in English in two volumes and published in 1930. Subsequently he
abridged these two volumes into one and published Elements of Strength of Materials primarily addressing undergraduate
students in American colleges in 1935. Since then many books appeared in this subject matter.
From Westinghouse Electric Corporation in 1927, he became a faculty professor in the University of Michigan where he
created the first bachelor’s and doctoral programs in engineering mechanics. From 1936 onward he was a professor at Stan-
ford University. He was the leading point to transform engineering education around the world by bringing Approximate
Engineering Approach in Design. This enabled advanced design in all fields of engineering activity. Different disciplines in en-
gineering were borne from Science. They are actually domains in applications of Mechanical, Electrical, Aeronautical, Marine
etc. besides Static Structural engineering of Civil Engineering. However, they all emanated from just Science of 17th and 18th
centuries.
In absence of computational means, Timoshenko’s approach became widespread throughout the world; the basic equa-
tions of scientific revolution got simplified. Navier-Cauchy equations are given in 1822 involving 15 state variables, a set of
coupled partial differential equations, see Rao [19]. Similar is the case with Euler equations given first in 1755 and Navier–
Stokes equations involving viscous effects; they are seven coupled partial differential equations with seven state variables,
see Rao [20]. Therefore the need for approximations and Professor Timoshenko pioneered in this engineering effort. Factors
of safety, Stress concentrations, testing during design all became norm of the day. Until the end of 20th century these time
consuming and expensive methods prevailed in academia as well as industry. Things began to change in this approach ever
since digital computers became affordable and common.
The invention of digital computer using valves in Philadelphia (ENIAC) during the II World War has changed the scenario
gradually; this was accelerated by Transistors and subsequently Integrated Circuits in 1960’s and it is all on a sudden a
boom for number crunching. The high performance computing that followed gradually replaced engineering approximations
through Strength of Materials by Elasticity and Lagrangian approach.
Initially the computers are main frames, each research institute having one central facility with punching machines for
cards for input to the computer, e.g., IBM 360. Research laboratories wrote out computer programs in Fortran and Algol
languages. The output from the computer was also in the form of cards and printouts from these cards were taken and pro-
cessed by hand to understand the results. Even the expensive supercomputers of late 1980’s are small in memory and speed
when compared with computer clusters today. The industry was also following this cumbersome and expensive procedure.
Academic institutions by necessity had individual researchers having programs exclusively with them.
Things have changed with high performance computing. It is now “Science to Engineering” or “Simulation Based Engi-
neering Science SBES” through several commercially available codes to do the drudgery arising out of Finite Element Meth-
ods or Finite Volume Methods. This has revolutionized the outlook for designs; what used to be months of design time by
several skilled engineers, it is now few seconds of numerical effort.

2. Stanford and other initiatives

With changing times Stanford University Mechanical Engineering adopted curriculum that expects Undergraduates to
demonstrate the following amongst other things through their projects:

• Application of knowledge of mathematics, science, and engineering.

• Design and conduct experiments, as well as to analyze and interpret data.
• Design a system, component, or process to meet desired needs.
• Ability to function on multidisciplinary teams.
54 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

• Identify, formulate and solve engineering problems.

• Usage of the techniques, skills, and modern engineering tools.
• Design of thermal and mechanical systems.
Accordingly, Stanford uses projects extensively and offer project-based courses at all levels, see, Faste and Roth [5]. They
identify the projects to come from three sources:
1. Industry Sponsors.
2. Instructor generates the project.
3. Students generate the projects.
Industry sponsors have a tendency to state the problem in ways that define a favored solution. Usually the teacher
himself is the source as in second category. The third category where students generate the projects tend to create the most
enthusiasm on the part of the student. Philosophically self-generated projects are attractive. If one wants the students to
become leaders, they have to learn to generate worthwhile assignments.
Typically category 1 industry projects are carried out in a few industry houses in countries like India, usually called out-
sourcing industries. They begin with a baseline provided by them. One needs CAE tools. This is discussed as a project from
Ford by Faste and Roth [5]. The second and third categories are usually carried out in Industrial Design Centers, not as suc-
cessfully as they seem to be independent of design practices otherwise from basic sciences. Indian Institute of Technology,
Bombay, started a program in 1979–83 under UNDP support to improve the quality of industrial design, see Industrial De-
sign Center [8]. The main finding of this project in India can be stated as: To commit to new responsibilities by broadening
the educational program and work towards influencing the design thinking at a national and international level.
Sheppard and Jenison [27] discussed examples of Freshman Design Education, these design courses are explicitly inte-
grated with freshman level mathematics and engineering mechanics courses. Problem Solving Approach was adopted by
many US schools, e.g., Ciocanel and Elahinia [2]. Machine design is a required course at junior year to learn essential skills
for senior design projects. There is a great need for comprehensive and integrated software due to its complicate nature of
the course materials. These tools empower students to solve more challenging open-ended and/or integrated design prob-
lems, and to conduct design projects for a more rewarding experience in machine design, see Kyu-Jung and Rezaei [10].
Modern computer simulations make stress analysis easy. As they continue to replace classical mathematical methods
of analysis, these software programs require users to have a solid understanding of the fundamental principles on which
they are based. Bower [1] in his book describes the physical and mathematical laws that govern deformation and internal
forces in solids. He introduced use of the finite element analysis to model deformation and failure, and describes theory and
implementation of FEM including a chapter on analyzing failure in materials and structures.
US were concerned with the engineering instruction and the need to introduce SBES. A high level panel was set up as a
Presidential initiative with six US government departments, National Science Foundation (NSF), Department of Energy (DOE),
National Aeronautics and Space Administration (NASA), National Institutes of Health (NIH), National Institute of Standards
and Technology (NIST) and Department of Defense (DOD), see WTEC–Panel Report [29]. This report says on p.79, that “The
use of simulation and computational models is pervasive in almost every engineering discipline and at every stage in the
life cycle of an engineered system. Typical examples of the use of simulation in engineering include manufacturing process
modeling; continuum models of bulk transformation processes; structural analysis; finite element models of deformation
and failure modes; computational fluid dynamics for turbulence modeling; multi-physics models for engineered products;
system dynamics models for kinematics and vibration analysis; modeling and analysis of civil infrastructures; network mod-
els for communication and transportation systems; enterprise and supply chain models; and simulation and gaming models
for training, situation assessment, and education.”
Essentially, there are only basic sciences; solid mechanics, fluid mechanics, thermodynamics, heat transfer and elec-
tromagnetism as given in 17th and 18th centuries; they are all expressed as coupled partial differential equations. These
equations can all be simulated using numerical methods. This simulation has become possible with the advent of digital
computers for different domains of engineering as illustrated in Fig. 2.
The design essentially consists of two parts, the first one on the concept and creativity in making the outline and then the
second part on the application of science to make an analytical design. In the analytical design part Professor Timoshenko
made highly innovative contribution to the world of education by evolving Approximate Engineering Approach in the absence
of number crunching ability from the digital computers. Soon after de Laval’s nozzle and Edison dynamo, the needs of
exploding design were met using this approach together with testing and usage of factors such as stress concentration factor,
factor of safety etc. Thanks to the invention of digital computers the time consuming and testing intensive approximate
engineering design approach, the mainstay of 20th century industrial practice, is gradually replaced by a direct Science to
Engineering approach through Simulation and High Performance Computing (HPC). Industry was quick to adapt this approach
to cut the design time and costs.
Some practical design aspects such as what bolt should be used in a submarine to space craft are not a part of this
discussion. Also domain and manufacturing aspects amongst others are also excluded. They are not science driven; only
analytical design that depends on Science to Engineering Approach is considered.
The educational institutions have also gradually adapted to the new approach, though there are still semblances of Ap-
proximate Engineering Approach prevalent, especially in developing countries. There is a need to adapt Science to Engineer-
ing approach from the very beginning of instruction to the young students and relegate Approximate Engineering approach
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 55

Basic Sciences Simulation

Civil Engineering

Solid Mechanics
Mechanical Engineering

Fluid Mechanics

Engineering
Electrical Engineering

Thermodynamics Mathematics
Aerospace Engineering

Heat Transfer
Marine Engineering

Electromagnetism
Chemical Engineering

Fig. 2. Science to Engineering.

I Year

Mathematics History since invention of wheel to present day

Physics Specific domain introduction
Freehand Workshop Practice
Practice
II Year Use software in
addition to hand
Solids Adopt to calculation
Fluids Core
Replace or add digital
Electromagnetics signal acquisition and
Digital Signal Processing analysis

III + IV Years
Energy Conservation Fluid Statics;
Lagrangian Conservation Eliminate handbooks
Mechanics Momentum and Energy Stress concentration,
factor of safety …
Finite Element Finite Volume Methods
bring in SBES for
Methods One D diffusion and
Vibration, Fatigue,
One D structures Convection Life and Optimization

Fig. 3. Science to Engineering Approach in under graduate class.

to elective History classes in the later years. A general Science to Engineering approach is discussed by Rao [18] as given in
Fig. 3.
The outlines of curriculum in each of the four years of undergraduate program are suggested as follows.
I Year: The notion that mathematics, particularly differential equations, is a separate subject from physics may be dis-
pelled. It should be emphasized that simulation or parameterization of physics necessitated the invention of differential
equations and their numerical solutions need matrix algebra, Doherty and Keller [4] and Keller [9]. Basic sciences, viz.,
Solid Mechanics, Fluid Mechanics and Electromagnetism should be used in deriving and treating Differential equations and
explain how Linear Algebra plays a role in solution of state quantities.
Develop an introductory course on history of machinery and how science played a role in engineering; see Rao [13,21]
56 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 4. ITER cryostat.

Fig. 5. Location of TBM in cryostat.

II Year: This is where Basic Science Courses on Solids, Fluids and Electromagnetism may be introduced as a core. In
addition a core course on Digital Signal Processing may be introduced as Analog Measurements have faded out. This practice
may be introduced in key laboratories.
III and IV Years: Design courses are to be developed by dropping the concepts of approximation, stress concentration
factor, factor of safety etc. The peak stresses etc. obtained by FEM or FVM have to be incorporated following industry prac-
tices.
Besides the core courses, one may introduce Kinematics of Machinery following SBES and HPC and Dynamics of Ma-
chinery. Following modern practices these two courses define the periodic time dependent forces. An FFT algorithm can be
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 57

Fig. 6. Schematic of radial buildup of fusion breeding blanket.

Fig. 7. Schematic of concept LLCB-TBM.

used in design classes to obtain the harmonic forces. Assuming no flutter condition the alternating stresses can be obtained
which pave the way for fatigue.
Then a course on Vibrations can be offered which leads to critical speeds where resonance occurs with large dynamic
stresses limited only by damping. High cycle fatigue and low cycle fatigue with fracture mechanics can be a course that can
be given in final year. This will lead to Life Estimation that dictates design of torpedoes to automobiles.
Vibration of rotating structures is a special subject Rotor Dynamics where the excitation mainly comes from the imbal-
ance. All rotors have special characteristics, e.g., misalignment, supports and seals, casing and foundation. This subject also
leads to asset management or condition monitoring, a subject of growing importance in protecting the rotating machinery
in diverse fields of application.
In recent years optimization is becoming an important part of the design, e.g., weight optimization, shape optimization
etc., and no design is considered complete without this optimization. An advanced course can be offered in final year or
58 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Preprocessing Solver Postprocessor

for FEA

CAD Solves Non- Displays

Model homogenous results from
system of nodal
linear deformations
Physics e.g., strains
Mesh equations set
Solids stresses and
up using
Problem Physics in the principal
preprocessor stresses etc. in
Stiffness color format or
Boundary
Matrix … gives nodal
Conditions
values.

Material
Properties

Fig. 8. Single physics codes.

Fig. 9. Multi-Physics Platform Approach.

graduate level. The author of this paper is giving courses to popularize the SBES approach or Science to Engineering ap-
proach. One may refer to the courses offered e.g., in US, China and India respectively, Rao [15–17].
The subject matter is not new, but the reorganization and adaptation allows considerable saving in time for instruction
and bring Science to Engineering culture instead of Approximate Engineering. This will bring our undergraduate students
ready to enter industry and be a part of original modern designs of 21st century without any expensive and time consum-
ing retraining. Two examples can be cited: 1. Topology optimization in obtaining a new 70-seater aircraft wing, Topology
Optimization of Aircraft Wing, see Rao et al. [22] and 2. Flutter of an Aircraft Wing, see Rao et al. [25].
A multi-physics example of design of Liquid Lithium Ceramic Breeder (LLCB) for Tritium Fuel in a Fusion Reactor by a
group of fresh graduates following SBES approach with HPC is next illustrated.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 59

Fig. 10. LLCB TBM assembly installed in one of the half ports.

Fig. 11. LLCB TBM assembly.

3. Design aspects of an LLCB following SBES approach with HPC

ITER in Cadarache, France was engaged in the design of 500 MW fusion reactors. With our experience in SBES I made
a presentation to the Director Engineering and his team and was pleasantly surprised at their invitation to provide design
support a Tritium Breeding Module. Back in Bangalore, there was a clear opposition to take up a project in so-called nuclear
physics area. The management also was advising against such a project. I tried to convince the bright young well trained
team to take up the challenge. I reminded them several times in that fortnight that the design is mechanical in nature the
loads are from nucleonic and thermal in nature. I had some very young and bright fresh graduates from IITs, yet a fusion
reactor scared all (I should say including me as well). Pleasantly one day the leader of the team peeped into my office and
said they are willing to take up the challenge. SBES won and so the project was on.
Briefly, International Thermonuclear Experimental Reactor (ITER) Cryostat, Fig. 4, is 24 m high and 24 m in dia weighing
around 23,350 tons. It has a central solenoid, toroidal and poloidal field coils forming the superconducting magnetic field
for the plasma contained in the tokomak maintained in a molecular level vacuum. The plasma of volume 840 m3 is at
100 million degrees centigrade temperature or more that moves in a toroidal path under the influence of the magnetic field.
The vacuum vessel is made of 9 sectors comprising of a blanket with 440 modules. The cryostat is the most complex multi-
physics system. Two isotopes of hydrogen, Deuterium and Tritium form the plasma under high pressure and temperature.
Whereas Deuterium is stable and available in sea water, Tritium is produced by the reactor itself. Producing Tritium in the
Power Plant is the most important aspect in design. Tritium is bred in the reactor in Tritium Breeding Modules (TBM) and
60 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 12. TBM support system.

Fig. 13. LLCB TBM dimensions.

this project is designing a TBM. Fig. 5 shows the location of TBM in the cryostat and Fig. 6 shows a schematic of radial
buildup of fusion breeding blanket.
Lead-Lithium Cooled Ceramic Breeder (LLCB) blanket module breeds Tritium and extracts heat from the fusion reactor.
This module is heated by the 14 MeV neutrons and the surface heat flux from the fusion Plasma and is kept in front of
the Plasma. In the LLCB Tritium breeding module the coolant is Lead-Lithium (LL) flowing around the ceramic breeder zone.
The Plasma facing surface of the module is the First Wall (FW) made of Ferrite Martensitic Steel (FMS) cooled separately
by high pressure Helium gas. The flowing LL experiences very high magnetic field in the reactor environment (∼ 4 Tesla) in
transverse direction and develops high MHD pressure drops due to induced current interaction with the magnetic field. This
modifies the flow velocity profiles and turbulence characteristics, and hence the heat transfer and MHD pressure drop in the
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 61

Fig. 14. Section view at TBM mid-plane without stiffeners.

Fig. 15. TBM FW assembly with internal He channel details.

62 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 16. Header CFD domain.

Fig. 17. Baseline–velocity contours- m/s.

flow channels and manifolds. The engineering design of the LLCB Tritium breeding module should be an optimized design
in various aspects such as, mechanical, thermal, structural, thermal hydraulics, thermo mechanics, magneto hydrodynamic
pressure drops and electromagnetic loads. The TBM system includes several associated equipment, such as primary and
secondary coolant circuits and components, Tritium management components, liquid breeder loop for liquid breeder TBM’s,
instrumentation packaging and control system, safety detection systems and valves.
Since it faces plasma (FW), TBM has to be cooled such that its walls, particularly facing Plasma can withstand this
heat. He is circulated to provide the necessary cooling in TBMs. Lithium Titanate is used to breed Tritium in the breeders.
Beryllium is used in Helium Cooled Solid Breeder (HCSB-TBM) to provide for extra neutrons. LLCBs are dual cooled (DCLL),
He and Pb-Li are used; the Pb-Li providing additional neutrons. Figs. 6 and 7 give Schematics of radial buildup of fusion
breeding blanket and concept of LLCB-TBM.
J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 63

Fig. 18. Optimization results–velocity magnitude - m/s.

Fig. 19. Inner back plate header CFD model.

Uniform Distribution achieved in all the

channels variation 10%

Fig. 20. Inner back plate header velocity magnitude - m/s.

64 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Table 1
Liquid TBMs developed in the various countries.

Country Name of TBM Liquid metal used

US DCLL – Dual Coolant Lead Lithium PbLi

EU HCLL – Helium Cooled Lithium Lead PbLi
Korea HCML – Helium Cooled Molten Lithium Li
India LLCB - Lead-Lithium Cooled Ceramic Breeder PbLi
China DFLL – Dual Functional Lithium Lead PbLi

He IN
Circuit 2

Inner Back Plate

Fig. 21. Design of inlet header.

The TBM assembly consists of FW, top (covering) plate, bottom (covering) plate, ceramic breeder zones, inlet and outlet
channels for Pb-Li flow, Helium purge lines, back plate, supports for TBM, piping and manifolds. The FW along with top
plate, bottom plate and back plate form a closed enclosure. The TBM is inserted in a water-cooled steel frame, 20 mm-thick,
which provides a standardized interface with the ITER basic structure, including thermal insulation of the basic machine.
TBM is mechanically connected to the back wall of the frame through flexible supports. It is supported at the back side
with dowel pins and support keys. The ambient condition for TBM excluding the wall facing plasma is a vacuum of the
order of 10−8 torr.
Major functions of LLCB are: 1. Transmutation reaction between Lithium in blanket and neutrons from Plasma that gener-
ates Tritium and heat, 2. Lead acting as a neutron multiplier increases the number of neutrons and hence the higher Tritium
breeding and 3. Heat extraction from the test blanket by Lead, Lithium and Helium. Lead Lithium continuously circulates
and acts as a coolant for the ceramic breeder, Helium acts as a coolant for the FW, top plate, bottom plate and back plate.
The FW of TBM receives heat flux from Plasma. It has two Helium circuits (Purge Helium circuit and Coolant Helium circuit)
and a Lead Lithium circuit. The Purge Helium circuit is for extracting Tritium and Coolant Helium circuit is for cooling FW,
top and bottom plate and back plate. The Lead Lithium circuit acts as a coolant for ceramic breeder and neutron multiplier
for higher Tritium breeding. Wong et al. [28] gave an overview of liquid metal TBM concepts and programs in progress in
different countries as given in Table 1.
A team of 18 engineers envisaged through discussions and workshops the design of LLCB to be 1. CFD analysis of the
baseline Test Blanket Module (TBM) to find out the temperature, pressure and velocity distribution, 2. Optimize the flow
path if required to meet the design criteria, 3. MHD analysis to estimate the pressure drop due to induced current inter-
action with magnetic field, 4. FEA analysis of flow optimized design, and 5. Optimize the TBM design to meet required
structural design criteria. This is following Stanford practice of creativity in developing the concepts through workshops
and consultations with ITER India included. Single Physics applications work on the principle given in Fig. 8 and Platform
approach is adopted with multi-physics concepts, as outlined in Fig. 9.
The Dual Coolant, Pb-17Li (DCLL) TBM is designed using low activation Ferritic steel as structural material and He as
a coolant for the first wall and structure. The TBM is designed to occupy one poloidal half section of the equatorial test
port designated for testing blanket modules. It is mounted inside a water cooled frame designed to hold two different test
modules.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 65

Outlets

He IN Circuit 2- 0.8 kg/s

Fig. 22. Inlet Header CFD model.

Fig. 23. Optimization of inlet header.

A baseline TBM model is first developed and used for CFD and thermo-mechanical analysis and optimization to fix flow
paths, Electro Magnetic analysis for estimating EMag loads and structural design. Fig. 10 shows the TBM baseline model as it
is inserted into the frame occupying the test port. The TBM is a rectangular structure with a faceted FW designed to match
the surface of ITER shielding blanket. Its design provides flow channels for the Pb-17Li to flow poloidally at a slow speed
while the He coolant is flowing throughout the TBM structure. The TBM assembly is shown in Fig. 11.
The TBM is designed to use a center key support system to allow for supporting and centering it inside the frame.
Peripheral (corner) support studs are to retain the TBM and stabilize it during operation, and allow for movement and
expansion without restrain. The support key will be inserted into matching slot in the shielding block located behind the
TBM, and the four positioning pins will be inserted into the shielding block and used to set the TBM into the proper position
and provide the radial support needed at the top and bottom during operation as shown in Fig 12.
As the TBM will be mounted inside water cooled support frame the overall TBM size is 1660 mm tall (Poloidal) by
480 mm wide (Toroidal) as shown in Fig. 13. The total radial depth of TBM is 534 mm. The TBM is designed to use Helium
66 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 24. Inlet header and convergent divergent main header CFD model.

Fig. 25. Results–inlet header and convergent divergent main header.

as the primary coolant and Pb-17Li as the self-cooled breeding material. It is designed to accommodate the two flows
internally and maintain a total separation between them. Also it is designed to withstand the maximum He pressures in
case of an internal leak from the He into the Pb-17Li chambers. Fig. 14 shows a cross-section of the TBM at the equatorial
plane.
The FW assembly is designed to withstand the heat flux from the Plasma chamber and to maintain the TBM structure
temperature below the allowable limits. It is a U shaped structure made of FMS material and designed with internal He
coolant channel. The coolant channels are designed to allow multiple passes across the FW in order to maximize heat
removal.
There are a total of 64 coolant channels, 20 mm wide by 20 mm depth in the TBM FW assembly. There are two circuits 1
and 2, both have 32 channels each and are interconnected by a series of manifolds in the back plates. The He flow between
the two circuits is always in a counter flow arrangement to achieve a uniform temperature distribution across the FW
surface. The two He circuits flowing through the FW channels are separated from each other and only mixed in the outlet
manifold prior to entering into the outlet pipe. Circuit 1 of the He flow channels have openings at the edge face of the FW
and these openings will match up with the outer back plate manifolds where the four passes are routed through. Circuit 2
on the other hand has the channel openings on the inner face of the FW as shown in Fig. 15. The FW structure is composed
of a 4 mm thick FMS plate facing the plasma. This plate will have a 2 mm Beryllium layer on top of the FMS FW.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 67

Fig. 26. Inlet header fixed main header configurations CFD model 1.

Fig. 27. Inlet header and fixed main header model 1 result.

Outlet

He IN Circuit 2- 0.8kg/s

Fig. 28. Inner back plate header–CFD model 2.

68 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 29. Results inner back plate header–velocity magnitude m/s.

Fig. 30. Results inner back plate header channels–velocity magnitude m/s.

Outlet

He IN Circuit 1- 0.8kg/s

Fig. 31. Outer back plate header–CFD model 2.

 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 69

Fig. 32. Outer back plate header results–velocity magnitude m/s.

Fig. 33. First wall channels results–velocity magnitude m/s.

4. Magneto-Hydrodynamics-flow

Magneto-hydrodynamics is the study about the interaction of electric currents in the liquid metal with magnetic field,
Rao and Sankar [24]. Electric currents are induced in the liquid metal moving in the presence of an externally applied trans-
verse magnetic field. These currents will form a close loop either through the walls, if they are conducting or through the
thin boundary layers, known as Hartmann layers if the walls are electrically insulated. The magnitude of these currents,
and the corresponding MHD pressure, is inversely proportional to the resistance of the electric circuit. Because the Hart-
mann layers have very large electrical resistance, much larger than the wall resistance, the electrical currents and the MHD
pressure in insulating wall ducts are significantly reduced.
MHD pressure drop is studied by a commercial code and verified against the theory, Rao and Sankar [23]. The Hartmann
layer thickness decreases with increasing Hartmann number. The velocity profile obtained is compared with the profile
predicted by Hunt [7]. The Hartmann layer thickness follows same pattern as predicted analytically. Here CFD calculations
70 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 34. LLCB TBM assembly.

Fig. 35. TBM support system.

were performed to determine the helium gas flow distribution in the FW channels before optimizing the design of the
headers for the FW helium flow circuits.
Cooling of the FW is achieved with two counter-flowing helium circuits. Each circuit consists of eight channels making
four passes of the FW see Fig. 16. Helium transitions from one pass to the next through headers. The configuration of the
headers determines the uniformity of channel flow distribution from one pass to the next.
The three-dimensional CFD model consists of a single header with eight inlet channels and eight outlet channels. The
mass flow rate is defined for each channel inlet and the resulting steady-state outlet rates are calculated. Turbulence
is accounted for by utilizing the k-epsilon turbulence model. Helium density is taken to be 5.63 kg/m3 and viscosity as
3.5 × 10−5 kg/m-s. The results obtained for the baseline model are shown in Fig. 17. Maximum velocity observed is 322 m/s
as shown and it is high; further the flow is not uniform in outlet channels. Therefore, we have to optimize header configu-
ration of the baseline.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 71

20

10

R2 480

30(2mm beryllium thk)

20

20 506

R2

28 28

64 hole Helium flow Circuit

Circuit-1: 32 holes
Circuit-2: 32 holes

All dimensions in mm

Fig. 36. TBM FW assembly with internal He channel details.

Optimization of complex problems has become economical in the recent years because of advances in the design as
well as availability of commercial codes. In earlier practices, dedicated codes were required to be developed to achieve a
specific optimization problem. The optimization here is carried out by Adaptive Response Surface Method (ARSM). Using
Multi-Physics Platform approach, Fig. 9, with mesh morphing techniques, Velocity of each channel to be 45 m/s is taken as
objective function. The results obtained in Fig. 18 show outlet channel uniform flow distribution, with maximum velocity
limited to 104 m/s. The variation in the outlet channel flows is limited to 11%.
The optimized header model used in inner back plate helium flow assembly for CFD analysis is shown in Fig. 19. The
flow results with the optimized model are given in Fig. 20. Uniform distribution was achieved in all the channels with a
maximum variation of only 10%.
The Helium inlet pipe passes through the outer back plate and connects to the inner back plate in the lower right corner
and then it is distributed through the first manifold as shown in Fig. 21. From the inlet manifold the flow goes to 8 He inlet
channels. Then the He flows through the FW in four passes and is then collected at the last manifold. Manifold inlet design
is required to get uniform distribution in first 8 helium channels.
Inlet header CFD model is shown in Fig 22. The results showed middle channels mass flow is high compared to side
channels of header and different configurations are evaluated to make the distribution in the inlet uniform. Therefore ob-
structions are used in inlet header to divide the jet velocity and get uniform distribution in all the channels as shown in
Fig 23.
Both the inlet and main headers are fitted in inner back plate and full CFD model is shown in Fig 24. The results obtained
in Fig. 25 show that the flow is non-uniform in all the channels due to small areas in convergent divergent regions that
restrict the flow from inlet header.
Considering the TBM design constraints, different header configurations, incorporating both fixed and variable cross-
sections, were evaluated. Of the two header configurations, analysis results indicate that a 45 mm × 80 mm fixed-area header
configuration, model 1, Fig. 26 achieves the most uniform flow distribution, with outlet channel flows varying by as much
as 14%. Therefore, in order to meet a minimum channel flow velocity of 45 m/s, some channels experience velocities as high
as 53 m/s. Fig. 27 shows the flow distribution and the average pressure loss for the headers is only 0.2 MPa.
Considering the TBM design constraints, previous model CFD analysis shows that minimum thickness required for inner
back plate is high (60 mm) and based on this parameter, Model 2 concept is taken for analysis as shown in Fig. 28. The
results obtained in Figs. 29 and 30 shows that the flow is uniform in all the channels as well as this flow arrangement gives
the necessary cooling to the inner back plate.
The average pressure loss for the headers is 0.187 MPa only as against 0.2 MPa compared to previous model and the
pressure drop is reduced by 6%. Fig. 31 shows Model 2 for outer back plate. A similar arrangement is used to perform flow
72 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 37. Breeder assembly.

Fig. 38. Back plate assembly.

 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 73

Fig. 39. Section view of the TBM in the poloidal direction showing the Pb-17Li flow circuit.

Fig. 40. Section view of the lower Pb-17Li manifolds.

analysis of outer back plate as shown in Figs. 32 and 33. This analysis shows the Main Header Thickness to be 30 mm inner
and outer back plate thickness required is 40 mm.
There are several other analyses to be done before fixing the final design: FW Conjugate Heat Transfer Analysis, Lead
Lithium Flow Pressure drop, Lead Lithium Flow and Heat Transfer, Conjugate heat transfer analysis of Lead Lithium Flow,
Top and bottom plate Thermo-Hydraulics, 3D - LLCB Full Thermo-Hydraulic Analysis, LLCB Purge gas Manifold Design and
CFD Porous Media Analysis, Full TBM Baseline Thermo Mechanical Structural Analysis, Creep Damage Limit Evaluation, Buck-
ling Analysis, Seismic Analysis before arriving at final design. The team was able to accomplish this by using SBES with
HPC. If one adopts the training at the undergraduate level and keep it through engineering design practice as outlined in
Fig. 3 coupled with discussions and workshops starting with the concepts generated through creativity original designs can
be achieved in the least time.
Just the final design of the LLCB achieved by a team following the above is given next.
74 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 41. 3D section view of the Pb-17Li pipes.

Fig. 42. 3D view of the LLCB showing the breeder plates with Pb-17Li outlet flow.

The TBM assembly is shown in Fig. 34. The TBM module is designed to occupy one poloidal half of the equatorial test
port. The support system is shown in Fig. 35. The TBM will be mounted inside water cooled support frame. Based on this
frame restriction the overall TBM size is 1660 mm tall (Poloidal) by 480 mm wide (Toroidal) is retained same as in baseline
model as shown in Fig. 13. The total radial depth of TBM is 534 mm.
The baseline model dimensions of the FW structure are retained with a 4 mm thick FMS plate facing the plasma. This
plate will have a 2 mm Beryllium layer on top of the FMS FW, see Fig. 36. The five ceramic breeders in the TBM are as
shown in the assembly Fig. 37.
Fig. 38. shows the overall back plate assembly view of the two back plates.
Figs. 39–42 show the details of the Pb-17Li flow circuit
The FW He flow for Circuit 1 in detail is shown in Fig. 43. The flow in Circuit 2 follows a similar path, only it starts
at the opposite end of the FW as shown in Fig. 44. Fig. 45 shows the top plate internal flow channels with the top cover
removed for clarity. The He flow circuit into the bottom plate is same as top plate circuit, see Fig. 46.
The purge gas flow channels for breeders 2 and 4 in the top plate and for breeders 1, 3 and 5 in bottom plate are shown
in Figs. 47 and 48 respectively.
Two exploded views of the LLCB are provided in Fig. 49. This is the complete design of the LLCB designed entirely in
India [26], a new technology allowing Sun to be brought on to this earth and provide pollution free energy.
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 75

Fig. 43. He flow details in the outer back plate for FW He coolant in Circuit 1.

Fig. 44. He flow details in the inner back plate for FW He coolant in Circuit 2.

5. Closing remarks

Stanford University led by Professor Timoshenko paved the way for engineering design with the advent of steam turbine
and electric power plants. Professor Roth from the same University brought the modern practices and changes in the focus
of design instruction with the advent of digital computers.
I visited Cadarache to explore the possibility of contributing to ITER Fusion Reactor project. Surprisingly a warm welcome
was given by the Director of the design group. This is mainly because of our young engineers’ contribution to challenging
projects such as Kaveri Engine, High Speed Cryogenic Pump and providing successful designs; this is besides the Indian
talent in OEM’s projects.
76 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 45. Top plate flow details.

Fig. 46. Bottom plate flow details.

How did an average group of young engineers from a developing country India make this contribution? They brought
1680 science to 1980 high performance computing to provide simulation based engineering science in place of century old
approximate engineering of no-computer era. That did the trick. This paper described such an approach and changes needed
in implementing the way engineering undergraduate classes can be oriented using Science to Engineering Approach.
Though there was confidence in handling advanced engineering projects, natural resistance came in addressing what
appears to be an Atomic Physics project; even the management was not in favor. The young minds after a while accepted
the challenge as a result they produced LLCB for the first time in India and world. The major contributors in this project are:
Vasanthkumar Mahadevappa, M. Saravana, R. Ashok, Hari Sankar, China babu, G. Selva, H. Liya, K. Raj, G. Raju, M. Mahesh,
R. Deepak, R. Jayaram, Deiva, S. Suresh, Partha Dey, R. Rejin and A. Nagesh. The list is otherwise exhaustive and I am very
thankful to those young talented engineers who worked with me in IITs Kharagpur and Delhi and in Bangalore, Vijayawada
and Coimbatore for the last 57 years; actually they were teaching me continuously and updating me. I am very thankful
 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 77

Fig. 47. Top Purge gas flow assembly for breeder 2 and 4.

Fig. 48. Bottom Purge gas flow assembly for breeder 1, 3 and 5.

to my teachers from 1950’s to have inculcated in me a thought process leading to several highly satisfactory engineering
works.
Late Dr. Mahalingam, a veteran scholar and visionary, discovered me in 2005 and 2010 and his friendship resulted in me
joining his vision to India in 2012. He has been always encouraging me to advance technology and I remain deeply grateful
for his advice.
78 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79

Fig. 49. Full LLCB exploded views.

 J.S. Rao / Mechanism and Machine Theory 125 (2018) 52–79 79

References

[1] A.F. Bower, Applied Mechanics of Solids, CRC Press, 2009 ISBN 9781439802472.
[2] C. Ciocanel, M. Elahinia, Teaching engineering laboratories based on a problem solving approach, in: Proceedings of the 2008 American Society for
Engineering Education Pacific Southwest Annual Conference, March 27-28, 2008, Northern Arizona University, Flagstaff, AZ, Radisson Woodlands Hotel,
2008.
[3] B. Cooperrider, The importance of divergent thinking in engineering design, in: Pacific Southwest Annual Conference, Continuous Improvement in
Engineering Education, March 27-28, 2008, Proceedings of the 2008 American Society for Engineering Education, Northern Arizona University, Flagstaff,
AZ, Radisson Woodlands Hotel, 2008.
[4] R.E. Doherty, E.G. Keller, Mathematics of Modern Engineering, vol. 1, General Electric Co., John Wiley, 1936 Copyright.
[5] R. Faste, B. Roth, The design projects and contests – the rules of the game, J. Rob. Mechatron. 10 (1) (1997) 1.
[6] General Electric, A Century of Progress, The General Electric Story 1876-1978, Hall of History, Schenectady, New York, 1981.
[7] J.C.R. Hunt, Magneto hydrodynamic flow in rectangular ducts, J. Fluid Mech. 21 (1965) 577.
[8] Industrial Design Center, IIT, Bombay, 1985, UNDP/IND/77//013, FMR/SC/OPS/85/204, Paris.
[9] E.G. Keller, Mathematics of Modern Engineering, 2, General Electric Co., John Wiley, 1942 Copyright.
[10] K. Kyu-Jung, A. Rezaei, Efficient use of computational tools in machine design, in: Proceedings of the 2008, Northern Arizona University, Flagstaff, AZ,
Radisson Woodlands Hotel, March 27-28, 2008, American Society for Engineering Education Pacific Southwest Annual Conference, 2008.
[11] David Ogg, Europe of the Ancient Regime: 1715-1783, Harper & Row, 1965.
[12] J.S. Rao, Watt - two hundred years after his retirement, in: 10th World Congress on the Theory of Machines and Mechanisms, vol. 1, Oulu, Finland,
1999, p. 63.
[13] J.S. Rao, in: Ceccarelli Marco (Ed.), Kinematics of Machinery Through HyperWorks, History of Mechanism and Machine Science, vol. 18, Springer
Publication, 2011 ISBN 978-94-007-1155-6.
[14] J.S. Rao, in: Ceccarelli Marco (Ed.), History of Rotating Machinery Dynamics, History of Mechanism and Machine Science, vol. 20, Springer Publication,
2011 ISBN 978-94-007-1164-8.
[15] J.S. Rao, Weight and shape optimization course august 3, 2013, ASME International Design Engineering Technical Conferences (IDETC) and Computers
and Information in Engineering Conference (CIE), Portland, OR, USA, 2013.
[16] J.S. Rao, Vibrations, life, optimization of rotating machinery course, August 28 to 30, 2013, A Course to China Engineers in Rotating Machinery, 2013
Translating instructor Mr. Ji Yuan Ye, 2013.
[17] J.S. Rao, Stress, Flow and Heat, Vibration, Life, Optimization and Fusion, 2-6 February 2015, Kumaraguru College of Technology, Coimbatore, 2015.
[18] J.S. Rao, Science to engineering approach in engineering education, 14th World Congress in Mechanism and Machine Science, 25-30 October, Taipei,
Taiwan, 2015.
[19] J.S. Rao, Simulation Based Engineering in Solid Mechanics, Springer Publication, 2017 ISBN: 978-3-319-47613-1.
[20] J.S. Rao, Simulation Based Engineering in Fluid Flow Design, Springer Publication, 2017 ISBN: 978-3-319-46381-0.
[21] J.S. Rao, Man and Quest for Energy, Krishtel EMaging, Chennai, 2017.
[22] J.S. Rao, S. Kiran, J.V. Kamesh, M.A. Padmanabhan, S. Chandra, Topology optimization of aircraft wing, J. Aerosp. Sci. Technol. 61 (3) (2009) 402.
[23] J.S. Rao, H. Sankar, Numerical simulation of MHD effects on convective heat transfer characteristics of flow of liquid metal in annular tube, J. Fus. Eng.
Des. 86 (2-3) (2011) 183.
[24] J.S. Rao, H. Sankar, Magneto hydro-dynamics and heat transfer in liquid metal flows, Heat Transfer chapter 4, Intech Publishers, 2011 ISBN
978-953-307-317-0.
[25] J.S. Rao, R.K. Tharikaa, P. Shiva Kumar, Flutter of aircraft wing, ICVE 2015, September, Shanghai, China, 18-20, 2015.
[26] J.S. Rao, et al., Design Description Document for the Dual Coolant Pb 17li (Dcll) Test Blanket Module, Institute of Plasma Research, India, 2008 Report
to the ITER test blanket working group (TBWG), (2008).
[27] S. Sheppard, R. Jenison, Examples of freshman design education, Int. J. Eng. Educat. 13 (4) (1997) 248.
[28] C.P.C. Wong, J.F. Salavy, Y. Kim, I. Kirillov, E.R. Kumar, N.B. Morley, S. Tanaka, Y.C. Wu, Overview of liquid metal TBM concepts and programs, J. Fus.
Eng. Des. 83 (2008) 850.
[29] WTEC – Panel Report, 2009. http://www.wtec.org/sbes/SBES-GlobalFinalReport.pdf.