Scribd Logo
Upload

Welcome to Scribd!

  • 63 views

    Lecture1 - Robust Design

    Uploaded by

    ds_srinivas
    This document provides an introduction to a university course on robust and probabilistic design. It discusses key topics that will be covered over the course including introduction to robust design concepts, repetition of statistical concepts, introduction to probabilistic design methods, course layout and planned activities. It also outlines preparations for the first group exercise on statistical analysis and introduces the first individual assignment on probabilistic thread fasteners.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Lecture1 - Robust Design

    Uploaded by

    ds_srinivas
    63 views31 pages
    This document provides an introduction to a university course on robust and probabilistic design. It discusses key topics that will be covered over the course including introduction to robust design concepts, repetition of statistical concepts, introduction to probabilistic design methods, course layout and planned activities. It also outlines preparations for the first group exercise on statistical analysis and introduces the first individual assignment on probabilistic thread fasteners.

    Original Description:

    Robust design

    Original Title

    Lecture1_Robust design

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document provides an introduction to a university course on robust and probabilistic design. It discusses key topics that will be covered over the course including introduction to robust design concepts, repetition of statistical concepts, introduction to probabilistic design methods, course layout and planned activities. It also outlines preparations for the first group exercise on statistical analysis and introduces the first individual assignment on probabilistic thread fasteners.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    63 views31 pages

    Lecture1 - Robust Design

    Uploaded by

    ds_srinivas
    This document provides an introduction to a university course on robust and probabilistic design. It discusses key topics that will be covered over the course including introduction to robust design concepts, repetition of statistical concepts, introduction to probabilistic design methods, course layout and planned activities. It also outlines preparations for the first group exercise on statistical analysis and introduces the first individual assignment on probabilistic thread fasteners.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 31

    Robust (and probabilistic) design (MF2024)

    Lecture 1

    Introduction to robust and probabilistic design

    [email protected] http://www.kth.se/itm/inst/mmk/edu/inst_kurser/md/MF2024/
    1

    MF2024 lecture layout


    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction Preparations for the first exercise

    MF2024 lecture layout


    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction Preparations for the first exercise

    - not traditional optimization


    Traditional optimization is usually directed at maximizing (or minimizing) some performance parameter in a deterministic sense.
    Robust design is targeted at making the performance insensitive to variation. Pure robustification is directed at minimizing the variation of a performance parameter under the assumption that the designed nominal level of performance is satisfactory.

    Robust design

    Performance parameter

    Design parameter
    4

    - common sources of variation


    Variation in material properties: density, yield strength, modulus of elasticity, homogeneity, contamination, Variation due to manufacturing process variability: dimensions, heat treatment, residual stresses, Variation due to degradation with time: wear, corrosion, embrittlement, Variation due to environment: temperature, humidity, dust, supply voltage, electromagnetic fields, vibration, Variation in the mode of operation: Who uses the product & How

    Robust design

    Robust design
    - what it is
    A robust product is one that is insensitive to variation. A robust design is one that is insensitive to material property and manufacturing tolerance variation to the extent that a very high proportion of the product is manufactured within specification and without the need for extensive inspection or adjustment. A corollary of this is that a robust design permits the use of the least costly materials and manufacturing processes without compromising quality and performance. A robust product is one that performs and continues to perform to the satisfaction of the customer despite the possible drifts (variation) in its design parameters expected over the normal product life.

    MF2024 lecture layout

    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction Preparations for the first exercise

    Some statistics
    Arithmetic mean value where xi is a single value

    Standard deviation of a population

    Maximum-likelyhood (uncorrected) estimate of the standard deviation

    The sample standard deviation (corrected estimate)

    A minimum of 20 samples are needed for a sufficient set


    8

    Real parameters have (probability) distributions


    Uniform Normal

    Poisson

    Gamma

    Chi-square

    Weibull

    Exponential

    Pareto

    Normal distribution
    The probability density function (the Gaussian function):

    f x; ,

    1 e 2

    x 2

    2 2

    Parameters: expected (mean) value standard deviation >0 The cumulative distribution function (cdf):
    x

    F x; ,

    f x; ,

    du

    10

    Exponential distribution
    Probability density function (pdf);

    f x;
    Parameter: rate >0 Mean value: Median value:

    e 0

    ,x 0 ,x 0

    -1

    ln 2
    x

    Cumulative distribution function (cdf):

    F x;

    1 e

    11

    (2 parameter) Weibull distribution


    Probability density function (pdf)

    f x; k ,

    k x

    k 1

    , x 0; Parameters: >0 scale shape k>1


    Cumulative distribution function (cdf)

    F x; k ,

    1 e

    12

    The Weibull straight line


    Cdf:

    F x; k ,

    1 e

    Logarithmic transform:

    1 ln 1 F x

    1 1 F x; k , k x

    Once again:

    1 ln ln 1 F x
    Parameter substitution

    k ln x

    k ln

    The Weibull distribution as a straight line:

    k X

    Other distributions can also be transformed into a linear form


    13

    Matlab probability plot


    nc=[3 3 5 7 7 7 11 11 12 18] probplot(normal,nc)
    Probability
    Probability plot for Normal distribution 0.95 0.9 Probability plot for Exponential distribution 0.95 0.9

    probplot(weibulll,nc)

    0.5 0.25 0.1 0.05 0 5 10 Data 15 20

    Probability

    probplot(exponential,nc)

    0.75

    0.75 0.5 0.25 0.1 0.01

    10 Data

    15

    20

    Probability plot for Weibull distribution

    F(x)

    x0 too large
    Probability

    0.999 0.995 0.99 0.95 0.9 0.75 0.5 0.25 0.1 0.05 0 10 10 Data
    1

    x0 too small x

    3-parameter Weibull: add parameter x0

    10

    14

    MF2024 course layout


    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction Preparations for the first exercise

    15

    Reliability block scheme


    - probabilistic design
    p(s)

    S S Strength Traditional factor of safety (uncertainty): S F S L Stress L and strength S are of a probabilistic nature! S L S M Probabilistic factor of safety: 2 2

    L Stress

    Strength /Stress

    L
    16

    Reliability block scheme


    - probabilistic design

    p(s)

    S
    The surface represents failures

    Strength/ stress

    Design criteria:

    What is the acceptable failure probability?


    17

    Probabilistic design
    the philosophy
    All design parameters are treated as variables and the basic result from the analysis is a probability of failure.

    The effect of realistic variability of the design input parameters can thus be obtained.

    The probabilistic analysis methodology is capable of producing:


    Safety (risk) quantification Design variable sensitivity analysis Cost/weight reduction scenarios Optimum inspection intervals Probabilistic structural analysis: The structural analysis methodology in conjunction with statistical characterization of applied loading, geometry, material behavior, and expected environment.
    18

    Robust and probabilistic design


    Probabilistic design is a mathematically based

    engineering design methodology for producing high quality mass-produced products. quality.

    Robust design is a methodology for optimizing this

    19

    MF2024 Robust and probabilistic design


    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction
    Homepage Course layout Examination

    Preparations for the first exercise

    20

    MF2024 - HT2011
    Twelve lectures: 12x2 hours Four group exercises: 4x2 hours Two individual assignments A written exam 3 credits 3 credits

    Group exercises (and assignments): 4x2 hours 1 - Statistical analysis 2 - Anova 3 - Design of experiments 4 - Human-machine interaction Two individual assignments: 1 - Probabilistic thread fastener 2 - Robust flight
    Final grading: based on individual assignment & written exam
    21

    Course homepage

    22

    MF2024 HT11 - Course PM

    23

    MF2024 HT11 Planned activities

    24

    (Exercises and) assignments

    25

    MF2024 Robust and probabilistic design


    Introduction to Robust design Repetition of some Engineering statistics concepts Introduction to Probabilistic design Course introduction Preparations for the first exercise

    26

    Paper clip # of cycles to fatigue (n:n)

    Test condition:
    n paper clips n individuals Each individual is testing 1 unique paper clip

    Test procedure:
    Bend 90 degrees and back Count # of cycles until the clip breaks Write down the number of cycles in the test protocol
    27

    Paper clip # of cycles to fatigue (n:1)

    Test condition:
    n paper clips 1 individual Each individual is testing 1 unique paper clip

    Test procedure:
    Bend 90 degrees and back Count # of cycles until the clip breaks Write down the number of cycles in the test protocol
    28

    Snap fastener snap force (n:n)


    Snap fastener Scale Force N

    Test condition:

    864,276 g

    n fasteners n individuals Each individual is testing 1 unique snap fastener

    Test procedure:
    Place the fastener on the measuring device Measure the maximum force (force at the snap click point) when you close the snap fastener Write down the normal load N in the test protocol
    29

    Snap fastener snap force (n:1)


    Snap fastener Scale Force N

    Test condition:

    864,276 g

    n fasteners 1 individual The individual is testing n unique snap fasteners

    Test procedure (for each clip):


    Place the fastener on the measuring device Measure the maximum force (force at the snap click point) when you close the snap fastener Write down the normal load N in the test protocol
    30

    Whats next
    Thursday, October 27, 10-12 (Glader) Exercise 1: Statistical analysis Monday, October 31, 10-12 (M35) - Probabilistic design & introduction to individual Assignment 1

    31

    You might also like