engi -8700 civil design projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture...

19
COURSE ENGI -8700 Confederation Bridge CIVIL Design Project INSTRUCTOR Dr. Steve Bruneau EN.4013Ph 8642119 [email protected] TOPIC Lecture Monte Carlo simulation Case Study for REFERENCES. Case Study for cost and load estimating FILES WEB http://www.engr.mun.c a/~sbruneau/teaching/ 8700project http://www.confederationbridge.com/en/about_the_bridge/bridge_faqs.php

Upload: others

Post on 02-Mar-2020

22 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Confederation Bridge

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

REFERENCES.

Case Study for cost and load estimating

FILES

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

http://www.confederationbridge.com/en/about_the_bridge/bridge_faqs.php

Page 2: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forConfed Bridge

REFERENCES.

Case Study for cost and load estimating

Confed Bridge

Case Study

S Bruneau

FILES

S.Bruneau

Records

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 3: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Class Study in Monte Carlo Simulation for Probabilistic Load Estimating . . .

Case Study: Confederation Bridge, PEI, NB

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

Ice Loading on Pier Diameter

TOPIC

Lecture

Monte Carlo simulation

Case Study forThickness

REFERENCES.

Case Study for cost and load estimating

Simplified Assumptions for Illustrative Purposes

FILES

A 100 year design load for pier stability required by developer. Models suggest:

1) Pier Force = Lesser of Ice Crushing or Driving Force

2) Ice Crush Resist = Ice Thickness * constantsWEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

2) Ice Crush Resist Ice Thickness constants

3) Driving Force = Floe Diameter * Wind * constants

4) Ice thickness is independent of floe size is independent of wind

Page 4: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

start

Input data

Monte Carlo Simulation for Bridge Ice Loads

CIVIL

Design Project

Ice Thick, Floe diameter, Wind

Simulation: random number for each p d f to get a simulated value for each

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

p.d.f. to get a simulated value for each parameter (thick, diameter, wind)

C t f iRepeat for No of

Repeat for No of floes per year

TOPIC

Lecture

Monte Carlo simulation

Case Study for

Compute force on pier

Sort forces into f

years of Simulation

REFERENCES.

Case Study for cost and load estimating

frequency intervals track

years also

FILES

Prob(x) = Freq(x)/ No. simulations

Cumulative(x) = sum of prob up to Freq(x)

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Tabulate data

Plot Data

End

Page 5: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Simplified Assumptions for Illustrative Purposes

F(pier) = lesser of F(thick), F(diameter, wind)Th th f t t llCIVIL

Design Project

diameter

wind thickness

These three factors are naturally random independent variables that affect loading every time a collision occurs – perhaps millions f ti th lif f th

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected] We must research and judge what the best representative probability

distributions are that will provide the best long term estimates for each In this

diameter of times over the life of the structure, therefore. . .

TOPIC

Lecture

Monte Carlo simulation

Case Study for

distributions are that will provide the best long term estimates for each. In this case study lets assume that a triangular probability distribution is sufficient.

The Triangular Distribution is typically used in cases where the relationship between variables is known but data is scarce (possibly because of the

REFERENCES.

Case Study for cost and load estimating

between variables is known but data is scarce (possibly because of the high cost of collection). It is based on a knowledge of the minimum and maximum and an "inspired guess" as to the modal value. the triangular distribution is a continuous probability distribution with lower limit a, mode c and upper limit b

FILES

c and upper limit b

2/(b-a)

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project a bc

x

(height of triangle)

Page 6: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Simplified Assumptions for Illustrative Purposes

Therefore, we define the triangular probability density functions for our random variables with estimates for a, b and c. For example:

CIVIL

Design Project

random variables with estimates for a, b and c. For example:

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for Now, what do we know: For ice Thickness

REFERENCES.

Case Study for cost and load estimating

1.01. The area under a probability density function is

equal to 1.0 (triangular included)

2. The random variable (thickness etc) always has a

For ice Thickness

FILES 0 20.5

value somewhere between a and b

3. For each collision in a monte carlo simulation we need a value for the random variable that is between a and b and weighted towards c in

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

between a and b and weighted towards c in accordance with the probability function.

We get this random value each time using the following …

Page 7: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Simplified Assumptions for Illustrative Purposes

Density function for CIVIL

Design Project

yIce Thickness

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

Probability of Thickness (l) = P

REFERENCES.

Case Study for cost and load estimating

Thickness at Probability P = (l)

FILESProbability of Thickness (n) = P

Thickness at Probability P = (n)

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

So, a random number generator is used to give a value for P between 0 and 1, thus a value of l or n is computed with the formulas above, giving. . .

Page 8: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

A value for thickness, diameter and wind. We then use the deterministic formulas that we know about for Ice Cruching Force and Driving forces to get a force on the pier , ie. . .

CIVIL

Design Project

Monte Carlo Simulation

F(pier) = lesser of F(thick), F(diameter, wind)

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

So we end up with this after many, many collisions are modeled in the simulation

REFERENCES.

Case Study for cost and load estimating

simulation

urre

nce

FILES

babi

lity

of O

ccu

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Force in MN

Pro

0 5 10 15 20 25 30

Page 9: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Simplified Assumptions for Illustrative Purposes

So from this figure we can get the probability of exceedance for e

CIVIL

Design Project

p yany risk level over the period of the simulation.

y of

Occ

urre

nce

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

F i MN

Pro

babi

lity

So, if you want a design load value that will not be exceeded

th i 10 0000 5 10 15 20 25 30

TOPIC

Lecture

Monte Carlo simulation

Case Study for

Force in MN

eede

nce

more than once in a 10,000 years99.99%

REFERENCES.

Case Study for cost and load estimating

obab

ility

of E

xce

FILESPr

oForce in MN

Here’s your number!

24 MN0 5 10 15 20 25 30

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 10: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Fixed Link and the Confederation Bridge

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

REFERENCES.

Case Study for cost and load estimating

FILES

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 11: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Getting Ice Floe Thicknesses

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

REFERENCES.

Case Study for cost and load estimating

FILES

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 12: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

CIVIL

Design Project

Testing Ridge Keel Stength

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forConfed Bridge

REFERENCES.

Case Study for cost and load estimating

Confed Bridge

Case Study

S Bruneau

FILES

S.Bruneau

Records

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 13: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Primary Document for

CIVIL

Design Project

Document for Bridge Design

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forDesign Case

REFERENCES.

Case Study for cost and load estimating

Design Case Studies

Professor’s Sources,

FILES

Professor s Sources, Reports, Commercial Investigations, WWW

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 14: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Primary Document for

CIVIL

Design Project

Document for Bridge Design

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forDesign Case

REFERENCES.

Case Study for cost and load estimating

Design Case Studies

Professor’s Sources,

FILES

Professor s Sources, Reports, Commercial Investigations, WWW

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 15: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Probabilistic Estimating:

CIVIL

Design Project

Estimating: Cost example

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forDesign Case

REFERENCES.

Case Study for cost and load estimating

Design Case Studies

Professor’s Sources,

FILES

Professor s Sources, Reports, Commercial Investigations, WWW

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 16: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Probabilistic Estimating:

CIVIL

Design Project

Estimating: Cost example

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forDesign Case

REFERENCES.

Case Study for cost and load estimating

Design Case Studies

Professor’s Sources,

FILES

Professor s Sources, Reports, Commercial Investigations, WWW

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 17: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Primary Document for

CIVIL

Design Project

Document for Bridge Design

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study forDesign Case

REFERENCES.

Case Study for cost and load estimating

Design Case Studies

Professor’s Sources,

FILES

Professor s Sources, Reports, Commercial Investigations, WWW

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 18: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Primary Document for

CIVIL

Design ProjectBridge Design

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

We were right!

REFERENCES.

Case Study for cost and load estimating

FILES

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project

Page 19: ENGI -8700 CIVIL Design Projectsbruneau/teaching/8700project/archive/classof2011/8700 lecture 5.pdfENGI -8700 Simplified Assumptions for Illustrative Purposes Therefore, we define

COURSEENGI -8700

Class example using excel Filename: monte carlo simulation xnew.xlsx

CIVIL

Design Project

INSTRUCTORDr. Steve Bruneau EN.4013Ph 8642119 [email protected]

TOPIC

Lecture

Monte Carlo simulation

Case Study for

REFERENCES.

Case Study for cost and load estimating

FILES

WEB

http://www.engr.mun.ca/~sbruneau/teaching/8700project