CSSE374 Course Intro

1
CSSE374 Course Intro
Right Architecture in Tunisia Its always good
to ask the client what they want. It might not
look like what youd expected.

• Steve Chenoweth
• Day 1, Dec 1, 2008

2
Today

• Whats in the course why
• Arch design - How do you know if you succeed?
• Lets start with some bridges
• And some basic principles
• And an intro to the book

3
Whats in the course why

• Lets look at course web site (on AFS)
• Policies
• Schedule, with links to everything that will
  exist
• Ideas for projects from past classes
• Discuss how the projects will work
• Teams of 2 How formed?
• By 6 AM Thurs On Angel Wiki Survey, identify a
  team-mate.
• If youd like to be the one team of 3, let me
  know.
• Acting as clients, architects, and implementers
• On 3 different projects

4
Lessons from five bridgesWhat are Success and
Failure?

• Split into five groups in class
• Each group looks at one bridge, and list answers
  to these questions
• In what way(s) was this bridge a success? (or
  might it be a success)
• In what way(s) was this bridge a failure? (or
  might it be a failure)
• How could better requirements (versus design)
  have made a difference?
• How do you imagine the lessons here could apply
  to software development?
• This Should Take About 15 Minutes
• 5 Minutes to read
• 5 Minutes - Groups discuss their bridge
• 5 Minutes A quick class readout from all the
  groups

5
The Britannia Tubular BridgeMenai Strait, UK
1850-1970

• The Britannia Tubular Bridge was a civil
  engineering project in the late 1840s, and opened
  in 1850, to complete a section of the British
  railway system. The bridge is 1500 feet long, in
  four sections There are two approaches of about
  250 feet each, and two center spans of 500 feet.
  These four spans were built of rectangular,
  wrought-iron tubes, connected together to form a
  single, 1500-foot-long tube through which the
  trains passed.
• Because of intense competition in the railroad
  business, completion time was of extreme
  importance to the railroad company. Therefore
  the project was fast-tracked, with construction
  underway before all details of the design were
  worked out. Most visibly we have the tall
  towersat the time construction started, it was
  not known for sure whether the tubes themselves
  would be sufficient to carry the trains, or
  whether additional support would be needed. The
  towers were intended to support suspension chains
  if these were needed. Calculations during
  construction showed they werent, so the towers
  remain superfluously tall.
• Total cost of the Britannia Bridge was 600,000,
  very high in 1840s money. About 75 of this was
  spent on the construction and raising of the
  tubes, which weighed some 1,500 tons each (7,000
  pounds per foot of railway track). The bridge
  was used from its opening until 1970, when a fire
  (in a wooden roof built to protect the iron tubes
  from rain) damaged it beyond repair. The
  folklore is that the fire was set accidentally by
  kids using torches to look for bats in the
  tunnel.
• The Britannia Bridge was a celebrated engineering
  feat at the time however, only a few other
  tubular railway bridges were built.
  Truss-and-girder bridges requiring on the order
  of 5,000 pounds of iron per foot of track were
  under construction by 1860, and suspension
  bridges requiring less than 3,000 pounds/foot
  were in use by 1870. In addition, passengers
  found the experience of passing through the tubes
  (where temperatures could reach 110 degrees on a
  summer day) in trains pulled by coal-fired
  locomotives to be quite unpleasant. The
  Britannia Bridge may be viewed as the ancestor of
  the many box-girder bridges built since the
  1940s, although the tubular girders of these
  bridges are welded rather than riveted, traffic
  rides on top of the tube rather than inside it.
• Source Henry Petroski, Design Paradigms

A painting of the Britannia Bridge, showing how
the tubes form a 1500-foot iron tunnel, 100 feet
above the water.
6
The Niagara Gorge Suspension BridgeNew York,
USA 1855-1897

• The Niagara Gorge Suspension Bridge, was designed
  by John Roebling (who would later be known for
  the Brooklyn Bridge). The bridge was definitive
  proof that suspension bridges could carry rail
  traffic. Before that time, most bridge
  designers, particularly those in Britain, had
  maintained that suspension bridges were
  unsuitable for rail traffic, despite their
  ability to span longer distances with much less
  material than truss-and-girder designs.
• For railroads, the problem with using suspension
  bridges was that the concentrated load of the
  locomotive(s) would cause the bridge to deflect.
  This created a low spot. If the bridge
  deflected far enough, the low spot would be so
  low that the train couldnt climb back up. (Very
  embarrassing for the designers!) Additionally,
  suspension bridges had gotten something of a
  reputation for being blown down by strong
  windsfor instance, the Menai Straits bridge in
  England (a road bridge near the Britannia Bridge)
  had been severely damaged by wind on several
  occasions.
• Roebling solved the problems of carrying rail
  traffic by stiffening the bridge deck with a deep
  wooden truss, and tying the deck in place with a
  series of cables, both above (to the towers) and
  below (to points on the gorge wall). When the
  Niagara Bridge opened in 1855, it was at 820 feet
  nearly twice the length of the longest single
  railway span built to date, and used barely half
  the material per foot of track, compared to other
  designs. Roebling wrote an engineering report on
  the bridge, systematically examining the forces
  that could cause a suspension bridge to fail, and
  describing in detail the provisions his design
  made for stiffness and stability. The main point
  of this report, of course, was to prove that his
  design idea was the best choice for the job.
• While Roebling understood and dealt with the
  physical forces that could make suspension
  bridges unsuitable for trains, he had not
  anticipated the growth of railroad trains.
  Trains in general, and locomotives in particular,
  grew much heavier in the 40 years after the
  Niagara Gorge bridge was built, leading to its
  replacement by an arch bridge in 1897. By this
  time Roebling had designed and built many longer
  suspension bridges, including the Brooklyn
  Bridge, and suspension was well-established as
  the approach of choice for long bridges in North
  America.
• Source Henry Petroski, Design Paradigms

An illustration of the Niagara Gorge Bridge.
Visible are the diagonal stays above and below
the bridge deck, the deep truss (with trains
running above and carriages below), and the
relatively small train by todays standards.
7
The Tacoma Narrows BridgeWashington State,
USA1940

• The Tacoma Narrows bridge is possibly the most
  famous bridge in the world, having had the great
  misfortune of collapsing in front of the movie
  camera. Millions of physics students have
  watched the bridges twisting as an example of
  harmonic motion. The collapse movie is even
  included as a part of the definition of Bridge
  in the Microsoft Bookshelf CD-ROM package!
• The bridge was opened in 1940 to connect the city
  of Tacoma with the navy yards on the other side
  of the Narrows. It was designed by Leon
  Mosseiff, a highly-respected engineer who had
  worked earlier on the George Washington and
  Bronx-Whitestone bridges in New York and was one
  of the designers of the Golden Gate bridge.
  Mosseiff had developed the modified deflection
  theory, which held that the weight of a bridge
  deck could resist wind pressure without the need
  for additional trussing or stays hence Mosseiff
  favored an aesthetically-pleasing slender deck
  braced only by 8-foot tall plate girders. Some
  engineers had been concerned about the extreme
  length-to-width ratio of the Tacoma Narrows
  bridge, but Mosseiffs calculations and
  reputation overcame any objections.
• From the day it opened, the bridge moved in the
  wind, much more than expected, and gained the
  nickname Galloping Gertie. Often the movement
  took the form of an increasing oscillation, and
  the bridge became the object of study to
  determine why it moved. Therefore, when it
  started moving in a fairly light (40 mph) wind on
  November 7, 1940, the cameras were ready. This
  time the oscillations built until cables snapped
  and a substantial section of the roadway fell.
  Because of ample warning, there were no human
  injuries. The bridge was completely destroyed
  even the towers were bent to the point where they
  had to be torn down and rebuilt.
• Post-collapse analysis, including wind-tunnel
  testing of models, showed that the flexible
  bridge deck was flapping like a flag in the
  windthe flexing of the bridge changed how it
  responded to wind, creating a positive feedback
  loop. Mosseiffs theory had dealt only with
  static wind loading and ignored dynamic effects.
  It was not until the late 1940s that David
  Steinman (a member of the team investigating the
  collapse) felt this problem of aerodynamic
  movement in suspension bridges was solved he and
  applied his findings to the Macinac Bridge in
  Michigan.
• Meanwhile, the Tacoma Narrows Bridge was rebuilt,
  with a more traditional deep truss stiffening its
  roadway. The replacement bridge has stayed put
  in much higher winds and is in use today.
  Indeed, as of this writing, a second, similar
  bridge is planned to be built next to it, to
  handle increased automobile traffic across the
  strait.
• Source Petroski, Design Paradigms Levy
  Salvadori, Why Buildings Fall Down

Galloping Gertie crosses the finish line.
8
The Humber BridgeHumberside County, UK 1981-

• From its completion in 1981 until the opening of
  the Akashi-Kaikyo Bridge in Japan, the Humber
  Bridge was the longest suspension bridge in the
  world. It also holds the distinction of being
  the first major suspension bridge to use concrete
  towers. Its deck, rather than having a deep
  truss to resist wind movements, is in the form of
  an inverted wing, so that the bridge actually
  becomes heavier and more stable in wind.
  However, the bridge also boasts one of the most
  rapidly increasing debts of any public project,
  with no real long-term hope of ever paying off
  the loans taken to build it.
• The idea of building a road bridge across the
  Humber Estuary first came up in 1959, as a
  project to assist local development by increasing
  mobility and promoting industrial growth. Actual
  design and construction began in 1971, with an
  estimated cost of 23 million and a planned
  opening in 1976. As usual, the financial
  viability of the bridge relied on a certain,
  perhaps optimistic, timetable for the project.
• The construction project ran into serious
  problems in the foundations for the south support
  tower. The geology was riskythe foundations
  went into Kimmeridge clay, which is highly
  sensitive to water content if it gets too dry,
  it crumbles if it gets too wet, it turns to mud.
  Since the south tower is located in the river,
  this is a tough problem. It got worse when the
  half-sunk caisson encountered artesian water.
  The caisson had to be modified in place, and the
  remaining excavation and sinking was done
  underwater by divers. As a result the
  foundations for the south tower were not
  completed until 1976the original planned opening
  date for the bridge.
• Meanwhile, there was a steel strike, inflation,
  recession, and soaring interest rates for the
  project to contend with. These all took their
  toll in time and money, so that when the bridge
  finally opened in 1981 (five years late) it had
  cost 91 million, with another 54 million in
  interest already accrued, for a total opening-day
  debt of 145 million.
• Once the bridge was open, the planned traffic
  flow failed to materialize. Industrial
  development in the region was stagnant, and in
  the words of one local commentator, nobody in
  particular wants to travel between Grimsby and
  Hull. Traffic on the bridge has never been
  anywhere near enough to cover interest, so the
  debt continues to increase. If current trends
  continue, the debt will rise to the billions and
  then trillions of pounds by 2043, the date the
  bridge would be, under its original plans, paid
  for.
• The engineers involved in building the bridge
  apparently suspected the traffic foreccasts were
  unlikely to be achieved. However, once
  construction had been started, they had to
  consider the benefit of local jobs in doing that,
  and also the reaction theyd get from political
  and community leaders who had sponsored the
  bridge.
• Sources Bignell Fortune, Understanding
  Systems Failures Hawkes, Structures

The Humber Bridge during a typical rush hour.
Whats missing in this picture?
9
Strait of Messina Bridge Project (Once due to
start construction in 2004, then cancelled by new
Italian government in 2006. In the end, who
knows?)
would have to be huge to cause trouble, as the
bridge could face without damage a seismic action
corresponding to 7.1 magnitude in Richter scale
(severer than the earthquake that destroyed
Messina in 1908). Construction had been slated to
start in 2004, then 2006 with completion in 2012.
The only obstacle left was funding. The bridge
was expected to cost five billion dollars. The
bridge would be 60 m (196 feet wide) and have 12
lanes for traffic and two lanes in the middle for
trains. This would allow 140,000 vehicles and 200
trains per day. This would cut down transit times
of up to 12 hours (via ferry) down to
minutes. Ever since Giuseppe Garibaldi landed in
Sicily in 1860, completing the unification of the
nation, Italians have considered building a
bridge over the Strait of Messina, so this bridge
is tied to long-held political aspirations.
Sicily and the neighboring mainland area of
Calabria are among the poorest regions of Italy.
Analysts say the bridge would boost the economy
there, attracting tourism. However, critics
pointed to the environmental impacts of the
bridge and its construction, and also to the fact
that many other, less glamorous infrastructure
projects were desperately needed. And there was
a crack in the bridges role in national unity,
in that opponents suggest the construction
project was a banquet for organised crime. In
2006 two ministers of the newly elected
government of Romano Prodi stated their
opposition to the project when taking up office.
In August 2006, the project was announced as
"under review" for budgetary reasons. Citing
concerns that the project was too expensive, was
likely to enrich criminal gangs, and might not be
earthquake-proof after all, the project was
terminated in October 2006, over protests from
southern Italian legislators. Provided by
Stretto di Messina S.p.A. and news sources.
This project if completed would be one of the
Landmark Bridges of the 21st century, the longest
span suspension bridge ever built (between
towers). The Strait of Messina divides the
island of Sicily from Calabria in southern Italy,
and is 2 miles (3km) wide. The overall length is
not a big problem, per se, but water depth, wind,
and earthquakes all must be considered. After 50
years of study, the bridge looked possible if
maybe not economical. To avoid the problem of
the deep water, the solution was to design the
longest suspension bridge ever. It would have a
3300 m (2 mi) main span and 180 m (590 ft) side
spans (overall length 3.7 km(2.5 mi)). The main
piers would be founded in 120 m (400 ft) of
water. There was to be a new patented lighter
deck design which dealt with aerodynamic and
seismic problems. The wind would be no problem as
the aerodynamic features of the bridge would
allow it to withstand 216 km/hr (134 mi/hr).
Earthquakes
10
Some basic principles

• Architecture / design is about a project
• Know who defines success failure
• Results doing right things right
• Understanding the problem and the solution
• Knowing how to get from one to the other

11
Architecture / design is about a project

• Whats a project?
• A sequence of unique, complex, and connected
  activities having one goal or purpose and that
  must be completed by a specified time, within
  budget, and according to a specification.
• -- From Effective Project Management
  Traditional, Adaptive, Extreme, by Robert K.
  Wysocki, p. 4. (This years book for CSSE 372).
• For architecture, the meaning here is
  specification requirements. For lower-level
  design it could mean more. Why?

12
Who defines success and failure of the
project? -- Key project players

• The dilemma
• Many key players
• Youd like to satisfy them all, but their needs
  and wants are In conflict
• Your business needs to make money now and in the
  future

Customer Reps
End Users
The Impacted
Developers
Related Proj.
Testers
Process Owners
Some filter about what to do
The System
Like project managers. Perhaps spoken for by
regulators, etc. Often via your marketing
people.
13
Who defines success and failure of the
project? -- Key project players

• A solution Identify your client
• The client controls the money that pays for the
  system
• This sets the priorities on the needs/wants of
  all stakeholders
• The architect/designers job is to assist the
  client in making the decisions that will lead to
  a successful system. Why?

Customer Reps
End Users
The Impacted
Developers
Related Proj.
Testers
Process Owners
Client (Vision )
The System
Like project managers. Perhaps spoken for by
regulators, etc. Often via your marketing
people.
14
Doing the right thing right

• Successful architecture addresses two questions
• Does the system as built accurately implement
  what the specification described? (Does it do the
  thing right?)
• Does the specification accurately describe what
  the important people wanted? (Does it do the
  right thing?)

15
Doing the right thing right

• And we can measure the answers to both of these
• Does it do the thing right?
• Associated with Quality and with verification
  of the project (black box and white box testing
  vs specs).
• Does it do the right thing?
• Associated with requirements, validation of the
  project and black-box acceptance testing in the
  project

16
Doing the right thing right

• What was missing from each of the bridges?
• Does the definition of success and failure
  depend on where we draw the boundary around the
  project?

17
Problems and solutions

• Architects must know high-level problems and
  solutions
• Problem A difference between what a person sees
  and would like to see
• Solution A plan for bringing reality as seen
  and reality as desired into agreement
• In lower-level designs, we merge the high level
  solution ideas with lower-level requirements,
  extending that design.
• Which isnt always smooth

18
Problems and solutions

• Architecture especially involves both problems
  and solutions
• To the architect, a concise problem statement or
  needs statement is crucial It drives
  high-level design decisions
• A few paragraphs to a couple pages, defining the
  center and the boundary of the product scope, and
  leaving room for innovative solutions
• This is a big reason why we insist on problem
  statements

19
How do you get to the solution?
In an ideal world (and often in the clients
head)
Perfect Understanding of Problem
Perfect Blueprint for Solution
...in a single bound!!
In this world, some number of rough sketches
are needed
First Tries at Solution
Better Idea of Problem
Better Ideas of Solution
Good Enough Idea of Problem
Good Enough Blueprint for Solution

(First Architecture)
(First Problem Statement)
(Final Detailed Design)
(Final Requirements)
Note that each step may involve multiple rough
sketch solutions
And of course at the same time the client is
cycling, and getting a better better idea of
what they really wanted
20
What do designers do?

• Half technical
• Half social
• Like explaining what the design means
• Like talking people into doing things
• Architects have to talk to outsiders, too

Ok, here we are. What are we supposed to do?
21
Larmans text

• Widely used in OO Design courses and OO software
  engg courses
• Youll need a copy
• Tonight read Ch 1-3 the intro, including case
  studies
• Tomorrow well also discuss Inception
• Heres whats in the book

22
Larmans text, cntd

• Part I Introduction
• Part 2 Inception
• Part 3 Elaboration iteration 1 Basics
• Part 4 Elaboration iteration 2 More patterns
• Part 5 Elaboration iteration 3 Intermediate
  topics
• Part 6 Special topics

23
Larmans text, cntd

• Part I Introduction
• Explains the iterative approach hell take
• Introduces his two case studies
• Part 2 Inception
• How to start an interative/agile project
• The big picture
• Review of use cases other requirements

24
Larmans text, cntd

• Part I Introduction
• Part 2 Inception
• Part 3 Elaboration iteration 1 Basics
• Part 4 Elaboration iteration 2 More patterns
• Part 5 Elaboration iteration 3 Intermediate
  topics
• Part 6 Special topics