A simulation study of models for combinations of random loads.

72

Transcript of A simulation study of models for combinations of random loads.

Page 1: A simulation study of models for combinations of random loads.
Page 2: A simulation study of models for combinations of random loads.

££ff -

IF0R'a

Page 3: A simulation study of models for combinations of random loads.
Page 4: A simulation study of models for combinations of random loads.
Page 5: A simulation study of models for combinations of random loads.

NAVAL POSTGRADUATE SCHOOL

Monterey, California

THESISA SIMULATION STUDY OF MODELS

FORCOMBINATIONS OF RANDOM LOADS

byIloh, Jang Kab

September 1984

Thesis Advisor: P. A. Jacobs

Approved for public release; distribution unlimited

T222990

Page 6: A simulation study of models for combinations of random loads.
Page 7: A simulation study of models for combinations of random loads.

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE fWhen Data Entered)

REPORT DOCUMENTATION PAGE READ INSTRUCTIONSBEFORE COMPLETING FORM

1. REPORT NUMBER 2. GOVT ACCESSION NO. 3 RECIPIENT'S CATALOG NUMBER

4. TITLE (and Subtitle)

A Simulation Study of Models for

Combinations of Random Loads

5. TYPE OF REPORT ft PERIOD COVERED

Master's Thesis;September 1984

6. PERFORMING ORG. REPORT NUMBER

7. AuTHORfij

Nph, Jang Kab

B. CONTRACT OR GRANT NUMBERS.)

9. PERFORMING ORGANIZATION NAME AND ADDRESS

Naval Postgraduate SchoolMonterey, California 93943

10. PROGRAM ELEMENT, PROJECT, TASKAREA ft WORK UNIT NUMBERS

II. CONTROLLING OFFICE NAME AND ADDRESS

Naval Postgraduate School

Monterey, California 93943

12. REPORT DATE

September 198413. NUMBER OF PAGES

60t«. MONITORING AGENCY NAME & ADDRESSf// different from Controlling Office) 15. SECURITY CLASS, (of thie report)

Unclassified15a. DECLASSIFICATION/ DOWNGRADING

SCHEDULE

16- DISTRIBUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT '0/ (he abstract entered In Block 20, It dltterent trom Report)

18. SUPPLEMENTARY N'OT^S

r

i '9 KEY WORDS (Continue on reverse aide if necessary and Identity by block number)

Constant LoadShock LoadFirst-passage Time

'20. ABSTRACT (Continue on reverse aide If necessary and Identify by block number)

This thesis describes a model for a combination of random loads actingupon a physical structure, such as a building or ship. The various loadsrepresented in a model might be winds, tidal effects, or even earthquakes orsnow loading. Asumptotic results are given for the first-passage time for theload combination process to exceed a given stress level exceeding structuralstrength. The accuracy of using the asymptotic results to approximate thefirst-passage time, or time to structural failure, distribution is assessedby simul ation .

DD , jam 73 1473 EDITION OF 1 NOV 65 IS OBSOLETE]

S N 0102- L c - 014- 6601UNCLASSIFIED

SECURITY CLASSIFICATION OF THIS P*GE (When Data Entered)

Page 8: A simulation study of models for combinations of random loads.

ABSTRACT

This thesis describes a model for a combination of

random leads acting upon a physical structure, such as a

building or ship. The various loads represented in the

model might be winds, tidal effects, or even earthquakes or

snow loading. Asymptotic results are given for the first-

passage time for the load combination process to exceed a

given stress level exceeding structural strength. The accu-

racy of using the asymptotic results to approximate the

first-passage time, cr time to structural failure, distribu-

tion is assessed by simulation.

Page 9: A simulation study of models for combinations of random loads.

Approved for public release; distribution unlimited

A Simulation Study of Modelsfor

Combinations of Random Loads

by

Noh, Jang KabMajor, Korea Air Force

B.S., Korea Air Force Academy, 1974

Submitted in partial fulfillment of therequirements for the degree of

MASTER OF SCIENCE IN OPERATION RESEARCH

from the

NAVAI POSTGRADUATE SCHOOLSeptember 198U

Page 10: A simulation study of models for combinations of random loads.

NAV< ^ /

MONTEREY. CAI ^J ' *•'

TABLE OF CONTENTS

I. DESCRIPTION OF A STOCHASTIC MODEL FOR

COMBINATIONS OF RANDOM LOADS 8

II. ASYMPTOTIC RESULTS FOR THE DISTRIBUTION OF THE

FIRST-PASSAGE TIME T x11

A. THE DISTRIBUTION OF Tx FOR LARGE X 11

B. THE TAIL OF THE DISTRIBUTION OF T x FOR

FINITE X 12

III. COMPUTER SIMULATION 16

IV. ANALYSIS OF TKF SIMULATION RESULTS 21

A. THE DISTRIBUTION OF Tx FOR INCREASING X ... 21

1. Constant and Shock Load Magnitudes

Have Identical Exponential

Distr itutions 21

2. Constant and Shock Loal Magnitudes

Have Different Exponential

Distr itutions 23

3. Arrival Rate of Shock Loads Depends

on the Constant Load Magnitude 24

4. Load Magnitudes Have a Pareto

Distribution 25

3. ^HE EXPONENTIAL TAIL OF THE DISTRIBUTION

CF Tx FOR FINITE X 26

C. CONCLUSIONS 27

V. SIMULATION ALGORITHM " 45

A. VARIABLE DEFINITION 45

B. ALGORITHM 45

Page 11: A simulation study of models for combinations of random loads.

APPENDIX A: COMPUTER PROGRAM 47

LIST Cr REFERENCES 59

INITIAL DISTRIBUTION IIST 60

Page 12: A simulation study of models for combinations of random loads.

LIST OF TABLES

A.I

A.

2

A.

3

B.1

E.2

B.3

C.1

C.2

C.3

D. 1

E.2

D.3

1.1

2.2

F. 1

F.2

G.1

Identical Exponential

Identical Exponential

Identical Exponential

Different Exponential

Different Exponential

Different Exponential

case A-1

case A-2

case A-3

case B-1

case B-2

28

29

30

31

32

33

34

case B-3

Varying Arrival Rate ( case C-1)

Varying Arrival Rate ( case C-2 ) 35

Varying Arrival Rate ( case C-3 ) 36

Pareto Distribution ( case D-1) 37

Pareto Distribution ( case D-2 ) 38

Pareto Distribution ( case D-3) 39

^ (x) and r(x) ( case E)

40

Quantiles ( case E) 41

%{x) and r*(x) ( case F)

42

Quantiles ( case F) 43

Confidence Interval in case E 44

Page 13: A simulation study of models for combinations of random loads.

LIST OF FIGURES

1. 1

3. 1

The Load Combination Process 10

Generating T^ 17

Page 14: A simulation study of models for combinations of random loads.

I. DESCRIPTION OF A STOCHASTIC MODEL

FOR COMBINATIONS OF RANDOM LOADS

Many physical structures are threatened by combinations

of loads of varying magnitudes from various sources. For

instance, bridges, piers, dams, and buildings can experience

loads frcm wind, snow, ice, tides, earthquakes, and so

forth. In many instances the total load or stress experi-

enced by a structure varies in time in an apparently random

fashion. Certain components of the loads vary ratter

slowly; others occur more nearly as impulses, such as those

associated with winds or earthquakes. The problem is to

design a structure to withstand the superposition of random

loads- from many sources with at least an approximately

understood probability.

The purpose of this thesis is to describe and investi-

gate certain simple but somewhat realistic probabilistic

loal models for use in design, and perhaps safety assessment

of structures. In this thesis we confine attention to the

superposition of just two load types; shock loads, and

constant loads. For example, winds gusts, flash floois, and

earthquakes have varying magnitudes and have relatively

short durations in comparison to the times between their

occurrences. These will be modelled as instantaneous shock

loads. Cn the other hand, snow, ice, or water accumulation,

or even the presence of slowly moving vehicles present loads

that remain nearly constant in time, occasionally changing

to new levels. These will be modeled as constant loads that

change infrequently. Throughout this investigation it will

be assumed that the effective stress exerted by several

types of loads acting simultaneously can be expressed as a

linear combination of the component loads, the load

components being treated as stochastic processes.

8

Page 15: A simulation study of models for combinations of random loads.

The times between changes in magnitude of the constant

load process are independently and identically distrituted

with distribution function H. The successive magnitudes of

the constant load are independently and identically distrib-

uted with distribution function F. Given the constant load

process, the shock load process is a compound Poisson

process. The successive shock load magnitudes are indepen-

dently and identically distributed with distribution func-

tion G. The conditional rate of arrival of shocks, given

that the magnitude cf the constant load process is x, is

Mix) .

let X (t) be the magnitude of constant load process at

time t; and Y(t) be the magnitude of shock load process at

time t; Z (t) =X (t) +Y ( t) is the superposition of the two loads

at time t, and M(t)=SUP Z (s) , the maximum load combination

in [C,t], see figure 1.1.

Let Tx={inf t>0: Z(t)>x } ie, the first-passage time

for the load combination to exceed a stress level x. Tx

will represent the time to failure of a structure whose

strength is x, and is subjected to a stress history

[Z(t) ; t>0} .

Gaver and Jacobs (1981) studied the above model for the

case in which the distribution H is exponential and the

shock load rate m is a constant independent of the shock

load process. Related load combination models that have

been studied include Peir and Cornell (1973) [Ref. 1] Wen

(1977) [Ref. 2] Pearce and tfen (1983) [Ref. 3].

Asymptotic results which appear in Gaver and Jacobs

(1934). and Jacobs (1984) will be described in chapter 2 for

the distribution of Tx as x-* e» and for the tail of the

distribution of Tx for finite x.

In chapter 3, a simulation program for the load

combination model will be described.

Page 16: A simulation study of models for combinations of random loads.

i

X(t )

i

1>

<

> •

•—•— ~w

Y(t)i

.

1 1 1 fcw

Z (t)i

t t

<ll t—L •

1

V y

. , ,.

.

Figure 1.1 The Load Combination Process.

In chapter 4, a simulation study of the use of the

asymptotic results of chapter 2 to approximate the

distribution of Tx will be described.

10

Page 17: A simulation study of models for combinations of random loads.

II. ASYMPTOTIC RESULTS FOE THE DISTRIBUTION OF

THE FIRST- PASSAGE TIME Tx

A. THE DISTRIBDTION CF TK FOR LARGE X

This section summarizes some of the results found in

Gaver and Jacobs (1981) [Ref. 4] for the model in which the

distribution H is exponential (x) and the shock load arrival

rate m is constant independent of the constant load process.

First, we consider the distribution function of the

maximum load combination to occur during [0,t], M (t) .

H(t)=P£M(t)<X}x

=P (M (t) <x, T ( >t} +P [M (t) <X,T, <t} (2-1)

where Ti is the arrival time of the first shock.

A renewal argument yields;

P{M(t)<x,T, >t}= e~MJ*e*Pt-Aii &*'$)) pcty)

? {M (t)< x ,T, <t} = J* *t-**dMJ* ext[-AAvGOL-w) HxCt-VJ fcAf)

where G (x) =1-G (x) .

Next take Laplace transforms with respect to t of (2-1)

.

£<*) = /.* e-^HA±) it

= Mx (g )+ Am x (£ )H x (f )

(2-2)

where

-\(S ) = /X[S^ iMGrtt-W1 F(dy) (2-3)

Hence

It seems to be difficult tc invert H ^ (£ ) for any inter-

esting choice of the distributions F (x) and G (x) . However,

useful information can still be gleaned from eguation (2-4).

First, from the defirition of T% ,

E {M (t) <x}=P (T x >t) (2-5)

Thus, H^) = /V 5 * p(Tx >t) dt (2-6)

let % ->o, in eguation (2-6) to find that

1 1

Page 18: A simulation study of models for combinations of random loads.

m(x)=E(Tx )=Hx(0)= Mxto)

(2-7)/ -*Mx(o)

The next result is the limiting property for the first

time the load combination process exceeds a given level x.

is exponential in the sense

____(2- 8)

The limiting distribution of Tx

that

liff E{m (x)H

Tx >t}=€-t

A *^ Ac

where x =inf {t ; F*G (t) = 1}

The prccf of this result for a more general model that

includes that of chapter 2 can be found in Jacobs (1984)

[Ref. 5].

B. THE TAIL OF THE DISTRIBUTION OF Tx FOR FINITE X

In this section we will summarize results for the asymp-

totic behavior of P (Tx >t) for large t and fixed x as t-»a>

for the model in which the distribution H is exponential and

the shock load arrival rate is constant. More details and

results for a more general model that includes that of

chapter 2 can be found in Jacobs (1984)

.

Frcm the equations (2-1) and (2-5)

p(Tx >t)= e-x*£ Ptctyj e-*** (*"W*

+ *}?e-Xi JLs£ F ave-M * i*'*>s pcTx>t-j; (2-9)

let

l (ds)= // fUi) ^-O+^oiOs ds (2-10)

* (t) = h* f«i) ^V»i {<- e<*+«™W) (2-11)

1 <" >

=IT ?<**) tt^T) (2" 12)

The renewal equation (2-9) gives;

P(Tx >t)=g(t)+ Jo€lUs) H x (t-s) (2-13)

where g (t) = j* f^cdy.) e"^ +>u ^ Cx "^3 * (2-14)

Following the development in Feller (1971;" page 376)

[Ref. 6 ] it will be assumed that there exists a number %such that

i= J; exw * UH) (2-15)

12

Page 19: A simulation study of models for combinations of random loads.

This roct is unique and since the distribution of I is

defective, is positive.

Let

L*(ds)= e*WS L(ds) (2-16)

g*(t)= e%(-x)i gCt; (2-17)

P* (Tx >t) = t%(xH

H K (t) (2-18)

Then

P*(Tx >t)=g*(t) + J*F*{TK >(t-s)}I*(ds) (2-19)

It is assumed that the assumption for the key renewal

theorem holds. Application of the key renewal theorem

yields

lira cX(xH ? (Tx >t) = -4 f g* (t) dt (2-20)

Rewriting the equation (2-15), the equation determining

% is

=f

xFcdy; — = (2-22)

and

= />C^; —j-j^j-jj^ (2-23)

The key renewal theorem implies;

Lim e*^*?(Tx >t) = —(2- 24)^ ro

V*^fc> 7 £ -,

Examples: l> + MaCx-¥J -^fIt appears to be difficult to solve for X analytically

in general. We will discuss the numerical computation- of 4£

for two examples. In both examples, A-M^l and the

distribution of the shock and constant load magnitudes F and

G are of the form;

F (x) = e-a*

G (x)= e-b*

13

Page 20: A simulation study of models for combinations of random loads.

In example 1, a=b = 1 and in example 2, a/b=2.

To compute the "k for F (x) = e"°*

let

f (*) = L*PC«l«

,G(x) = e-tx

A

where A = xx = 1.

(2-25)

Fcr model 1,

f (x) =T773gT{('^)C/-e-)

')-e-x[M'-3c+e-^-^((/^;€-*4e'x ))}

—(2-26)

forX<1 + e"x andx*1.

For X=1, f (x) has a singular point, so by using the

L' HOPITAI RUL3 [ Ref . 7]

lim f(x) = Lim g'(x)/h'(x)

where g &) ^ s the numerator and h(x) is the denominator

cf the function f(X).

For % =1

,

f (1)=-£ (e*-e'x)

(2-27)

For model 2,

+ e~itx

[^(/-^f-^)-A((/'3c)€-^4e- t>x

)]} (2-28)

for % <1+e"kx and ft *1.

For^=1 f (X) has also singular point, so by using the same

method as used in model 1;

forX =1

(2-29)

Differentiation of the equation (2-25) gives

(2-30)

-,-ixZcuation (2-30) is greater than for 0<K1+e'"aRd^*1. So

the function f (x) is a monotone increasing function in the

interval 0<X<l+e~i>* . Using this result, we compute the K

from the equation (2-22) by applying the bisection method

[Ref. 8].

Using the computed x we compute the constant |f* such that;

is the constant of eguation (2-24).

14

Page 21: A simulation study of models for combinations of random loads.

r cr case a=b=1

,

4- _L^ 4=T7W{[ C

'- x+e"X) -Ze

"X^ (/^^"X) -T^W]

For case a/b=2,

e-* (z-*.) -3}— (2-31)

15

Page 22: A simulation study of models for combinations of random loads.

III. COMPUTER SIMULATION

This section describes the computer simulation model.

We denote the constant load arrival rate by x, the shock

load arrival rate by^ f the distribution of constant load

magnitude by F (x) , and the distribution of shock load magni-

tude by G(x). The simulation model will be used to study

the quality of the approximations of the distribution of Tx

by the asymptotic results described in chapter 2.

The following describes the simulation algorithm. The

stress levels are x, <x 2 <x3<- x -n

-

1, Set T = 0, N o=0.

2, Set 7=0, generate a constant load magnitude, C, and

duration interval Tc. Find the largest index i>N

such that X; <C.

Put N, equal to that index and

set

lx , =T for N <i<N, ,

set N =H, .

3, Generate a shock load magnitude, S, and a time until

the shock occurs, Ts.

4, If the shock interval does not exceed the excess life

of the constant load duration,Tc, find the largest

index i>N such that x;<C+S.

Let N, be that index.

Set Tx; =T+V+Ts for N <i<N,.

Set N =N, i V=V+Ts, Tc=Tc-Ts. Go to step 3.

5, If the shock interval exceeds the excess life time

of the constant load duration,

set T=T+Tc, go to step 2.

16

Page 23: A simulation study of models for combinations of random loads.

One realization cf the simulation is completed when one

T* is commuted for each level x;

, 1<i<n. Each simulation

consisted of 5000 replications. Further details are

(

z (t)

x4

x3

x2

xl

Tx3

- -Tx4 - - -

11 1

- . Tx2

I

,

Txl-I

I I

-

wt

Figure 3.1 Generating T,

included in chapter 5.

The Key subprograms are following. Other programs will

be explained by reading from the output results.

EAKDC ; generate the Tx .

CC.M ?UT ; compute the P(T*=0), true P(T*=0), sample mean,

true mean, standard deviation, coefficient of

variance.

Sample probability

P(Tx =0)= JMm^i °f T*=o is obtained fromr vx * ; Sample. j,'3€

the sample.

True protability

r(Tx =0)=1-P(Tx>0)=1-P(M(0)<x)=1- J*F(dy)=F (t) is

obtained from the equation (2-5).

The sample mean

x =xrL Tx: and

17

Page 24: A simulation study of models for combinations of random loads.

the true mean E(TX ) is obtained in special

cases from the equation (2-7)

.

See details in Gaver and Jacobs (1981).

Sample standard deviation

^ = T _, 1. ("fo;~"*)i s obtained from the sample,

i-

1

and

Sample coefficient of variance

CCv/J = = is obtained from the sample

mean and sample standard deviation.

CCMP ; compute the sample guantiles, and the quantiles

cf an exponential distribution with mean E(TxJ.

A sample quantile is

&. =(o<^sanple size) th order statistic of sample and

fd. - -E(T*)jfr*>( /- •<) is the quantile of exponential

distribution with mean E (Tx ) .

CCM ; compute the quantile of the sample conditional

distribution given Tx>0 and the exponential distrib-

ution with mean E (Tx ) /P (T y >0) .

The sample quantile of the conditional distribut-

ion

£o(C=[ et x (number of positive Ty ) ]th order statistic

is obtained for sample cf positive Tx .

The conditional quantile of the exponential dist-

ribution with mean E (Tx ) /P (T*>0) is

SK ; compute the 'K such that;

*={X(x): t=!?FUv x+^^.^ 3 b Y applying the

bisection method.

INT ; compute the constant J"* such that ;

A

by substituting the X found in SK.

18

Page 25: A simulation study of models for combinations of random loads.

The following cases were simulated.

distribution case

a , exponential 1,

(a=b=1) 2,

3,

b , exponential If

(a = 2,b=1) 2,

3,

c, exponential 1,

with varying 2,

arrival rate 3,

d,Jareto case 1,

2,

3

A -U(c)

1

2

1

1/c

1121

1-e"

ii

1-e-2X

1-/+ X

1-BfV+*P1-

G(X)

1-e-X

1-er*-x

1-e'x

/+x

1-e'x

1-e'x

1--4^x

These different settings will be referred to as case (A-1),

(A-2), (J-3), (3-1), (B-2), (B-3), (C-1), (C-2), (C-3),

(D-1), (E-2) , (D-3) respectively throughout this paper. For

each above case, we compare the asymptotic result (2-8) and

the experimental results.

The next setting is to examine the exponential approxi-

mation of the tail of the distribution of T* for finite x.

The cases of identical exponential distributions,

F(x)=G"(x)= e' y and different exponential distributions,

F (x) =e'AX G(x)=e~ fcx where a/b = 2

were considered.

The following conditions are considered.

e, A =M =1 f (x) =G (x) =e-*

f f A=m=1 F(x)=e' 2x ,G(x)=e~*

19

Page 26: A simulation study of models for combinations of random loads.

These different settings will be referred to as case (E) and

case (F) respectively. For each case, compute the dC , and

constant tf"

x and compare the guantiles of the asymptotic

results (2-24) to the simulated data for each level of x.

20

Page 27: A simulation study of models for combinations of random loads.

IV. ANALYSIS OF THE SIMULATION RESULTS

As indicated in chapter 2 the exact distribution of Tx

is difficult to obtain in general. In this chapter simula-

tion will be used to study the accuracy of the approxima-

tions of the distribution of T* by the exponential

distributions suggested by the asymptotic results of chapter

A. THE DISTRIBUTION CF T* FOE INCREASING X

The limiting result (2-8) indicated that the distribu-

tion of T* approaches the exponential as level x increases.

This result suggests that the distribution of T* can be

approximated by an exponential distribution at least for

large x. In this section this exponential approximation will

be studied via a simulation.

1 • Constant an d Shock Load Magni t udes Have Identical

Exponential Distributions

In the first collection of three models to be

considered F (x) =G (x) = e~*. The times between constant load

changes are exponential with parameter A and the shock load

arrival rate ax is a constant independent of the constant

load process.

For these cases it is possible to derive an analyt-

ical expression for E (T* ) .

In the Tables, P(Ty =0) refers to the simulated

sample probability that T* =0; X-3AE gives the simulated

21

Page 28: A simulation study of models for combinations of random loads.

sample irear. of T x ; ST-DEV is the simulated standard

deviation of Tx ; VAF (X-B) is the variance of the sample

mean; COEF-V is the sample standard deviation divided by the

sample mean. T P(7 X =0) is the theoretical probability that

T* = 0, ?(x) ; T x-bar is the theoretical mean of T* computed

from the equation (4-1) .

The simulated coefficient of variation in all three

cases decreases as the level x increases becoming quite

close to the theoretical exponential value of 1 when x=5.

To compare the simulated distribution of T x to the

approximating exponential distribution, quantiles from the

simulated data were computed and compared to the corre-

sponding guantiles cf an exponential distribution having

theoretical mean (4-1). The simulated quantiles were

computed as described in chapter 3. The exponential

o(-guantile is given ty

gol=-a.lr. (W)

where a is the mean of the exponential.

For each level of x, the first row in the tables

(A . 1 ) , (A. 2), (A. 3) for the quantiles of the distribution of

T x gives the quantiles of the simulated data and the second

row gives the corresponding exponential quantiles for an

exponential distribution having theoretical mean (4-1) .

Cne way the distribution of Tx differs from an expo-

nential is that it has an atom at 0. In particular,

P (T*=0) =F (x) =e~ x

for the case considered here.

Guantiles were used to coopare the simulated condi-

tional distribution of Ty given T* >0 to an exponential

distribution having mean ?, (T* ) /P (Tx >0) . The simulated guan-

tiles for the conditional distribution were computed as

described in chapter 3. These quantiles were compared to

those of an exponential distribution having theoretical mean

F(Tx) /P (Tx >0) .

22

Page 29: A simulation study of models for combinations of random loads.

In the tables (A.1), (A. 2) , and (A. 3) for the guan-

tiles of the conditional distribution for each level x, the

first row shows the simulated guantile and the second row

the corresponding approximating exponential guantile.

The guantiles of the simulated conditional distribu-

tion are much closer to their exponential approximation than

the guantiles for the unconditional distribution to their

exponential approximation. However the guantiles for the

unconditional distribution get closer to those of their

approximating exponential as x increases.

Comparing tables (A.1)/ (A. 2) , and (A. 3) , it appears

that increasing the arrival rate of shock or the rate of

change of the constant load magnitude decreases the guan-

tiles. The change of arrival rate of shocks appears to have

the greater effect.

2- _~2H§tant and Shock Load Magnitudes Have Different

l2£oneiltial Distributions

In the next three cases, the distribution of the

shock load magnitudes is exponential with parameter 1 and

the distribution of the constant load magnitudes is exponen-

tial with mean 0.5. The other model assumptions in tables

(B.1), (B.2), (B.3) correspond to those in tables (A.1),

(A. 2) , and (A. 3) .

As before, for each level x, the first row in the

guantile table gives the simulated guantile and the second

row gives the approximating exponential guantile using the

exponential distribution with theoretical mean. Comparing

the guantiles of T* in table (B.1) with (A.1) (respectively

(B.2) with (A. 2) and (E.3) with (A. 3) ) , it is seen that

decreasing the mean constant load magnitudes has increased

the guantiles. Further, it appears that the convergence of

the distribution of 1% to an exponential is faster in the

case of smaller mean constant load magnitude.

23

Page 30: A simulation study of models for combinations of random loads.

The exponential approximation to the conditional

distribution of Tx given Tx >0 has theoretical mean

E(TX )/P(T* >0) . Once again for small levels x,it appears

that the exponential approximation to the conditional

distributior of Tx is better than that for the unconditional

one.

3 • Arrival Rate of Shock Loads Depends on the Con stan t

load Magnit u de

In the next 3 cases, constant load magnitudes and

shock iDad magnitudes are exponential with mean 1. Constant

load magnitudes change according to a Poisson process with

rate 1. Given the constant load magnitude at time t is C,

the probability a shock load will arrive in the time

interval [t,t+h] is M(C) h+o (h) .

In case (C-1) the conditional shock load arrival

rate is ax[c^=C. Comparison with table (A.1) indicates that

conditional shock arrival rate x(c)~' =C has the effect of

decreasing the guantiles of Tx . Further the exponential

approximation to the guantiles of T x is not as good as in

case (A-1) .

The approximating exponential distribution to the

conditional distribution of Tx given T x >0 has a mean of

Z (Tx ) /P (Tx >0) . The guantiles of the exponential approxima-

tion to the guantiles of the conditional distribution appear

to be closer than those for the unconditional distribution.

In case (C-2) shock loads arrive with conditional

arrival rate aj^C)^ =exp (C) . Comparing the table (C.1) and

(C.2) , it is seen that the guantiles of the case ^(C)"* =C are

less than those for the case ^c(C) =exp (C) . This is because

interarrival times tends to be larger for the case

A(C)H =exp (C) . Comparing tables (A.1) and (C.2) indicates

that the exponential approximation to the guantiles of Tx is

not as good as in the case M='\

.

2U

Page 31: A simulation study of models for combinations of random loads.

In case (C-3) the conditional shock arrival rate is

Micf = exp (-C) . Comparing tables (C.2) and (C.3) indicates

that the quantiles fcr the case

yU{C)'1 =exj (-C) are less than those for the case ^(C)"* =exp (C)

ard furthermore the exponential approximation of Tx appears

to be close for the case (C-3) than that of for the case

(C-2) . In all of the cases the simulated mean was used as

the parameter in the exponential approximations.

4 • load Mag,ni tu des Have a Pareto Distribution

In the next three cases, the times between constant

load changes of magnitude are exponential with mean 1. Fhock

loads arrive according to a Poisson process with rate 2. In

case (D-1) the constant load magnitudes have a Pareto

distribution with parameter <*=1 and the shock load magni-

tudes are exponentially distributed with mean 1. Comparison

with table (A. 2) indicates that the Pareto constant load

magnitude has the effect of decreasing the juantiles of Tx .

Further the exponential approximation to the quantiles of Tx

is not good in the Pareto case; (the approximating exponen-

tial has a mean of the simulated sample mean).

The approximating exponential distribution of Tic

given T^ >0 to the conditional distribution has a mean of

E (Tx ) /P (T x >0) . Once again the quantiles of exponential

approximation to the quantiles of the conditional distribu-

tion appear to be closer than those for the unconditional

distr ibution.

In case (P-2) the constant load magnitudes have a

Pareto distribution with parameter o< =2. Comparing tables

(D.1) and (D. 2) it is seen that the quantiles of the simu-

lated distribution cf Tx for the caseoC=2 are larger than

those for the casec(=1. Comparing tables (D.2) and (A. 2)

indicate that the exponential approximation to the quantiles

of Tx is not as good for the Pareto case as for the

exponential case.

Page 32: A simulation study of models for combinations of random loads.

In case (D-3) the constant load magnitudes have

Pareto distribution with parameter o< =1 and the shock load

magnitudes have Pareto distribution with c*=1. In all three

cases the siaulated mean was used in the approximating guan-

tiles.

B. TEE EXPONENTIAL TAIL OF THE DISTRIBUTION OF Tx FOR

FINITE X

In this section simulation will be used to study the

exponential approximation

suggested by the asymptotic result

This is an approximation for P (Tx >t) for fixed finite x; it

should become more accurate as t becomes large.

Two cases are considered. In both cases shock loads

arrive according to a Poisson process with rate 1 and

constant load magnitudes change at the times of arrival of a

Poisson process with rate 1. The shock load magnitudes have

exponential distribution with mean 1. In case (E) , the

constant loal magnitude distribution is exponential with

mean 1; In case (F) it is exponential with mean 0.5.

Description of the computation of % and T* for these two

cases can be found in chapters 2,3. It can be seen from the

computed value of 3((x) and <f(x) appearing in tables, (E.1)

and (F.1) that X (x) is approaching the value 1/E(Tx) (where

F(Ty) is the theoretical mean) as x becomes larger and for, &.

the cases computed X(x)<-—: . Further , r (x) is approachingE(7x)

1 as x becomes larger.

. To study the exponential approximation to the distribu-

tion of I , cuantiles from the simulated data were compute!.

These can be found in tables (Z. 2) and (F. 2) . For each level

x, the first row presents the simulated guantile; the second

row gives the approximating ^-guantile computed by

^=~[-ln£i*(W)}] ( I )

26

Page 33: A simulation study of models for combinations of random loads.

The third row shows the approximating ^ -quar. tile computed by

qo( =-srx )ln(1-oO ( EL )

where the theoretical mean is used for E (Tx )

.

Comparing the quantiles in tables (E. 2) and (F.2) it can

be seen that the approximating quantiles (1) are close to

the simulated ones even for «>{~;.40. The approximating quan-

tiles computed in (jj;) do less well but improve as x becomes

larger. The two approximating quantiles approach one another

as x becomes larger, as expected. Approximating quantile (il)

has the advantage, however , of being easier to compute.

Table (G.1) presents 90 % confidence intervals for

^-quantiles with dL>0.9 for case (E) . These confidence inter-

vals were computed using a large sample approximation, see

Conover (1930; page 111-114) [ Fef . 9].

Comparison with table (E.2) suggests that the approximating

quantiles approximates the simulated ones well for these

values.

C. CCNCIDSIONS

Approximating guantile i. is always better than approxi-

mating quantile £ . Approximation I approximates q^ well for

s(,as small as 0.40. However I is more difficult to compute

than £ .

The performance of approximating quantile 2L can be

improved by chancing it to

y ' PCTx>o);

and using it to approximate the quantiles of the conditional

distribution of Ty qiven Ty>0; for the models considered

P(Tx =0) is easy to compute being equal to F (x) .

27

Page 34: A simulation study of models for combinations of random loads.

Zuj<mUZ—t-Kna—zHCLZUJ

—OO.KOCJ^uj

I

W

u

• rtf

kj: •H•«->

w CH-1 0)

en a< oH cu

XwHrd

U•H4->

fl

0>

13

• QUJ3(-1- UJ

c> «_ onr "5KEUJt- _><«!-•UJ ^lazr > o<a -iQ_«ex U.4 < *-

t «o>Q. •*_<~Q

1 or <>— Ht-aO_> £.£.**-!a «»< 1

Z (-I-**»/»iUO4ZOO

• 03II» u>o^>i^>

4> >—•^ ^s* e©<m fv—•m

t> rrvHHOOOOO

OOOOOO-ifNiO*OOOOOOOOO'Hoooooooooo

>jT^oir»f— On^oooin O* u~\ h") *4f f\jr^» CO'"^ <^

H» OO-HfsJi^lTV^-*

OO^rvn^N^- ,o

<x •©•oormO'O^onjinp-

CD P-VWWT^iMi'lV >r

OO^Nn>fNrr»^

i— O -OPs|f*)CD •*«"> —MOm* -O^INHOOOOOOCl ••••••••••oooooooooo

O <fNOO-f »r -4".4>CD .f »*

!_»*, «n JUy * mm --—(_> £.

f— <Omw—*OOOOO0 ^Q- oooooooooo O

O —•-^rvrgmm -* #* m

•rvcm oofN*«vrgo^omj^>a>-^o*'MO Of1*f*rs|

O co<"^m<"\|inr<»^r»*r*>r«>*J>O c -rHrft^rr)^^o* *rn'T>-nr>^)^ moco-oor"--Hnjjin*o jj-coo• •••••••••••••»•••••

h-m-^ojrgcomi/N-^m-^in-H -* *'*>fN(O r,*ACO O r,J ,,Wl/-'r*«-rOUO'J>" ,'>n.-«^"*m(^ O—4 "U «J lO -A* JJ '.T* H> -O <i"» u"\ ~« CD -* -f" Q —*'J*r<}

MO^H-rOOCMAH'>r»0<M^M(#m)r.cO

o* -rco vo"io 'TOO oo9>Nvaxo -r o */^j

0-*CD -«'M -O —<d>QCO^-*p-r>rvLM/>-«(7t O-HO* mo^«*>»rm<-«coin-^o,»co-0',»mof*>.-«r—

|

o • • • t • • • ••••—•—•—« -4r^rg gf *J •*>*©

•o *mo -«-*Nmf* -*" «"* o>m co f\j rsj co r+ —»o>co 0*"O oOO co o*» r-r- co cxmo coO0<04 <o -r• •*-* rvX307»n'V^^OOO' ,'tf,r*^ -T "*>!*>

c* •• . ...... t

OO —«—« (V—«mmm \r\ co cow rg co co r*. co^ mj

lACO'MOcO>»P*-f^>'0 ON^C*C"*fl}JNK*m (Tvcm^-t *or^^jrM^j-^^4^f»w>r-ir^^^'-*rn

O 00 00-*-*«%|rg^^ 0^>00 0^-t -r^f^^a -^ ^^^4 <M rs* rCf **>

f>» ^^rnfs-CO-*1 ^ -T-t COCO ^»00-^C7'C?* OT^rNj-^• ••••»•••••••*«•»•••O OOOO-*-«rvjrgm)'r»iO'0^0'T»fTt-^'^r\joj

O" OOOOHHH^f\<f\j4?^0-000'OOc?f

o^ <-»oo -t" i*^—•^ *n t/> -o f"»of^^ ^ co cot*q* ^^*counrr>*H>rrnr^tnoc7^»OiAtr\u*\c7v >0

»n o-^<^ -r -ocof>JTOr^^^>ro<N^r»o-^rvtr\^o

O 00 0000*^-* r*Jr\jm<n*ru"*r--cDiMc\!coco

r»-0-o<>i0^0o r,J^> l'^f—ocorgcoort>-^r*'TN-rO'N^'T >r -r*rv^^ v»orn'vm*«vn

-# O*^or«»^r «coOT orsiLHsT corner-VO'W—

f^ rs|f*-o^-*, *oeoo*ineomrg-Oo •©^•^-O

« •»•»•*••••?•*•••••••

^•OO-t^-iCUsrrg^/oncoo^^-Tnicoco-cf

CF .OOOOOOOOOO^HHf^fM^finN^

^ ,h ii>^NfO<>f*-mcomo^^m'Ocoo>0 m r*-r— in ^T •^^^^mjOrsiO^o-fmo

rj OO O^O^-h-* ^TP^co^^^^-rN/iAOC^r^O

O" OO OOOOOOOOO-<-*-<NNm'nin>0

O "^ O M»fTnr>r^ir\0 TO«""int^r»

-- t>0 c^OO-^CJOiO r'K^f>uicut^WNOCo*<JU

CJ 0000000000000 00^-*-^<-Nj(M

>ooooooooooujmou^o*no*noiAO

O* r»rsimc7»'-H^»'T>>#o-^—«»A'^-M-^fnco -o-^coO ON'O^OO'* <>f»»co(y»^)-^m'N-^^>^*coco• •••••• •*••••••••!••«

O? meof -t-f*r»OOi/>*A^rnvf\iA^^f>j-^r*>^-,—.—,—« fN* r^ rr>m^ U-y CO XJrsj ^

p4o »r» inO CO -O i^J rsj ^4^ if\^ *4 eo ^» CO (N »^»

O <J*—•f^'A -* ^V r**n» co -^ aJ p^r* **> -« *•* <D r*- ft)

«_>0-^^'«'0tf»t7»r-gw.rrnr*>u)*^,^r»^f.r>j

o <n -ra» <j co <n*n r*omr«4 unm coon# >r *><

O* <^-^coo*c^-*'inrgrt>o ,OfN4'^»nr».^)C^cof,-Nr

CO fM04-or,»0 {©0'N*'T'NO-0<N*^r'-nj<r',*r>*st.• ^^-4 0^>«n"5t r»r»vnmr*>pgrr>(T>oi>~*inm-Q

(7 •*••••••• ••••••—* -* -^-^rsi rM rf\mtnm CO CO ^rsiO^ CO t> co -*m

-*-h^h p>intni »r -r

O—•Lnr-n^'nOT'O®'-» j-O'-n^^- >^fr- o^O T>T-^^ rN' (>* cOr*'-si-H^)Ou^rsiO-OcOLn

O1 OO'^^'M'Mm^ > >h-N-00^-0 -*"-*• f*l^-4^-4-^rMfNjfn^

ry f*>m •>» r»»n >f ^O -rm^ >r <\» -o «1 *r f—*u « -* «r ««»nw» co to -»«>•-• co -^f*ir*tn •**''• r*m

h, cocO'N(rN*co«r»'^'-*-* ra-*'N'Mr>s*-*-rrnr»>gO

O oo-^'^-^-^'^i'M'*' -^-O O<^0 >f-T*-*-^f,vrg

o* rgom o1* fv o* a» '^ -o r* p* tn -o o*^ o* o> co co

• ••»•»••••»••••••»••CJ 0000-»-*rgrMrnrr»^srr^^oO'0-0'», 4'

oco<MfNj-r C'*»o^na^rn'Of^^^H^^coo ,''0>,«n m-*r*r-oo ,n^rn'Ti*tnr^rr>.^-^f*>fT>^i^»

O oOOO-^'^^^*^'^l rM p*^'^u^incOCOrMrgcOco^-•—4—«^^

^co^«MOf,*pn-^co-»-4>^^h"0*nr-0'^rnr— ui rr> i-r> co **j -r co rg ^n rg rs* rv—4 rgo rn r»- rsi *r*

*C m'T»^ntnPta-r*-*'Hr^p-trv^>f^v»{j»o-^c>co'o• •••••«•••«••#••••••

CT 000000^-<HHNN',»n«^(J'(7,n^

^o 1*1 <m nio %o in ~*inw rg -^^v h»m—«m rsi<m^finr-r-sr u^ tr>'NJLr" >4Conrtujr)U^i^inrnu^r*

wi rgrsi<^roinirtfN.«>^^rvf»-cou-\h- w-4 4*fTiLn-o• ••••••§•••••••<•••••

f^^^tnoOm cotn^ co rg co 0>o <Ntnjco co tn -^m oOiro^l^N-o«coif*tror«car*or"0

t •••••••*••••§#•%•••«CJ OOOOOOOOOO^HMMmnuM/\NN

O ODNO ^^4 0CD>0^f-'Orr»-H'*'*T> ,A r^^^r<4 .-i^rgfjn«'i >tmh»r*-0'-*'0^*4"-ocDt7v coo

• t • • • • t"••••#••%•»•••O OOOOOOOOOO^-4-4HNN<^^«*\-0

^* uo*-*—t-^-^i^rNjrnrnirMnr^cv-Hrva/cDintJO

O OOOOOOOOOOOOOOhh^^Nn

O -* -^ r-j ng

28

Page 35: A simulation study of models for combinations of random loads.

VIOlZUJ— XUJZOUJZOa — i_>u.

—OQ.XOUHUJ

•"•o n#»* rn^m -c a>^~ i >•-

ivD^o-D^ij o Ji^jin' *i»o^^oA'ny<

<N

CMI

U

w c-J Q)

en a*»: oH a.

*w

U•H

Ca?

-»- UJl/> <« QaC ar^uj^uj -3>— •—

I > Oa- a. ** -» *.

a. ^ _j— o»— *-^-oclj

o* «*< i

Z >-•-*:*

* OU3XX

ift -> sfrO^O-O -«-'© tft

gj -O C3* rg o »*- r\J ft -Org

ooooorg g* o^'n

ooooooooooooowoooooo

^O^COO —«f>4> ift<N

uj r* sj-tft ^ r rg.o j-*o-oO f\ ^CT -r -^ ^ .c —«r»rr)

I • • • •

t- 000-^rgrr>g> r-O-0

CO co ^o -o in -o (No g* f--

o

0C tft^tftfgccor-vorggv

00O— -.r W > OOO

oooooooooo

CDO^rg^O'O-O^CD ~

^"•' w-«oooooo ^• -Joooooooooo o

TO-to r*mOOQ* —•-* r*> 4} tft, rgo gj p- _* r*» tj»

CD -O O rs»a>o^ rsj^ coO^ a> a> rn^ rN# r>rg -^Q> rgrggjeootft g>a>rgrgcgg)-*f—cocDrg-H'x>c3

OtftrgO ^tf>r-mrnrgr*(^o«0^ T-«0-«

<J oooo-

U>rnmrf>g> Otftr,»(7*COeDOrg-Hg>tt>r-gco.t, rr)

r-» ^*rg TtftCT'Omift.fnrotft <J ^tftrorocQ -0^>O

O OOOOOO-^wrgrgromtfttftCOeorgrgcDO

0*> ^*-*^O tft*» ift IN U"*e> tftC> CO -^ CD ift CD iA.

gj o-^f">'*x;^»o-^ f*»ao ,</r*orv rT> rncoii>-*-sf

OOOOOO- (t-f-oocT'-r g*

corj*> -o ino—^^cc^io tft oto^o^®rg^OrgrngjvOcO-O'OP'''*' co-^^Oiftrn J"\

tft 0'-*"^ r^*yift's*coojr»">C^O<7*-*'0<© r^ r*0<^

O 00000000H-4HNNff|^^N.K-<0

-^ »TC0iftO'**t7*OlftC0^>r0^--^'O-^OrN;iftC7»fn-^-TtAvn>r*-o <^J4> co^iftgrgrgr^r>-

O OOOOOOOOO^^HNN^ntftHMBU)

^) rgiftrNioOf*g-^*Nco ^ »rsj^«Mift.h-o

1*1 OO0-^O <WW'T*ft'^"00>ft ^? f*>-^tft r*-0 -O

ooooooooooo-

l«1 f*>NO'OOH>ONO ,^N',l'* ^f^-wcO»ft Ift rgmr*«J-0-^OCDr*v/*fsj«r C^ff>'^C7, -Trg OOO-^O' g—•«^igru-\^)eo-^«ia?<7*ajO r, tu>

C/ OOOOOOOOOOOO-^^'H'HfM'^-t^

-— ift OrVC7s OOC7»lft'^'i*>-^OO r,,-^sC

g> O f*- O^- 3> «1 *^ CO <N -f -fm rn > rg

ooooooooooooo-

V tft tP NniNO^'T'tn^'*f*i(^o

ooooooooooooooooo-*^-^

fNO'^f*> ,MCO',ta CO«0(^'ftm'DO-0<0^ift -Oi^l

C3K3inift'g'g'g'T»V'tw»-«-^r^ y.--*- r-iQ

CO rg rrj m coO r\IO^ TV-H -0O tft i ft .^O 0^ rf>

CO tft^-^OTOf-O'-^tft'J^-^'M O'^(DHOOOH

C7 .•<•..••..•....•••OQ^^^4^rgrgmrn 1n tnp^r,^4^ r,p^. gj tn

ift iftin^rg>-o.O*JOC7»O ngO r^rnOCr, >rnrn-*rg

CT OOOO-^-*—t-^r\»M^^«0t>O-t*<N»'N~4_irgrg

OO—*rn<*g-^cor>'*>--u^-ft-^cO'<> ^rgrggjift

^- tftiftcoco f>J-^'s-^*ift<ftr-.t-*u-Mft -r ^or*o»rn

C7 0000—*-^—«-<'NfM'T>fT>tftiftcoco'vrgcoO

r- j- j- ,o .Q a*O •©-«'*• tft COOO ft -"Nj-O-Orgrrt

>*i •*** O OV^COfT>'^tr V»COCU-^n^^» 4» CT''-»X\'^

t •••#•••«••••••••«••

rgrg c*1* ^TJ^a*^ o-^ o<^»0^**- co^otn -Otf\ n>f|-T t^O«/'7t-T-J 1^|.^-^'>j(-.cor»i<r>-^-^• ••••••••%•••••§•#•%

Q* rg co C7»O -jrg -^ gj co c> rn rgco ift -^O co co ift

gr rgojmro-r»ft r*'s"OC3tntf>nm4> ti»g'g>rg(g

0>«^ <N»0^"N|>r^rgrO<>J^-Ocoa)0(Mrnr^gjntg^grtncD-^—««^cu-* ^^JtOwi^OT

m -^-^rv*-grnn^-tftf>»r*o-^^> og^tft-*r^T>^-

CJJ O000OO00OO'*'4'*-'f>"VWTn^

f»vO fr"^ ,O r^-O r^"^^ -o^*^ft<t)'^^M•^^*t7'^gift mrnCT'^r— aJCT^-OOtftV* -T'NCOrgixiftO'^rg •-•-^-^-^rg'^r^g'u-t ocot«">"tC7kUvotit-0

O OOOOOOOOOOOO^^HN'Vrt)*^

—•gjiftrgrgorg^^rg.^O-O -^O^ 1^ CO—«-*rr>-^O-t'tft-^-^-wrg^p*. g> O^rgrn-O rg^- >t (>

rg -*,-*-4-*'\|rgrni-n-l' > -0^OO^n^ ttft

r-*O rn r>* co r»> «^ -* co rgmv to -O r^* *^ tft'j* tft -*;

o oooooooooooooooo^^^

o o o o o o o o o o o o o o

O——*rgrwn-,(-i ^. g> tft

0-4<-*(Nrgrr>m^«rm •g *-4 rg rg r*i f*l

29

Page 36: A simulation study of models for combinations of random loads.

—XUJZ

—00. *OOX.UJ

«>• oo

mmNN^ ©r,»m'-«>oorgin'^rriP^omf«i-<?o> co-^inrgm-j'gooo.o^rgo-J'^'OVingrco

r»eorgcgVrgorg-OOCOin -OO^O-OrO*)**

w \f» -o co^m in i/* <-*mmo -T t*\-^ -«or* r»o

rg -0m <n 0> Is* r*> TO—•m r» CO —I d» rn COO CO <*> co

O CD -0 ^ -Ti rn - g rg o r/io r» CO ?3 -0 ^ —* 0> —• -*" ^

ft ••*••«*•«•••*•••••••3 r^rs*rOrOinincocors|rga>cor-.cDrripT»»nin*g^

—»—(—,.*-" jrgsj- r -o ooo—»-^

SO coo*"O oco rnco ooin in o> g^co g> -©ft> *t

v* v^ooo T,mr,"i/»rr|inr'»oin rg-^<-r**/vor,*rg

4 rg r*> n | inm UWQO

mI

• o »UJDy-t- ai

>/> <-. aocUJ»- _!<«-»UJ «ItrrZ > <J>•* •-•LJ_*«CC a^-ma.m acu>->o. «J"<U

1 «<ii- t-t-ao_> iC2«l-Ja. «< 1

Z t-l-**—

i

V>V>0(J<>:uo

• oan• uu^ko

CO •© r* <7> **r-mO -O C7* co o> cr -JO <oo <o V» -*CDm mC7"^^<^^'OmrgrNjh.r«jgr-<(^ 07*^0*^*</» O'ncT'-f *>* ^J^inconomoo-i-oor-CFNco

in <0 r» gf CO rn f*r* ^-» f*> f*- -Hr» rg rg -T *nr% gj rt>

-ro co co -*o r»mm r* -*> f*> -o <mh» r*>m c?» co>r <Ni <^ r«- o* -f-m gf rsjm -4* a*^ rsj»n r*> gf gr -^ ct»

fn^>cor*mg> g^ r -* c* >r\ —* to -o ct« to in o o^

co t7»O "1 ^H gj frin |rgom r«* r* inO 0> CD —4 <*> rry

m r-fT)g3r-io O'nHm'* gj*eo-«co**or*co>n-*O* inm<gn>m>t nviOO'^'^-^m-r'n'^OJ^O

O '^'^^cgjrnmmnco'DiN^jcoBcocomrNih.gj)-^—i —»-^' -^' g -r -# « v

mfn-OCO-O Tco-^rvp^g^ 4>Hinh <7» »n (7»^*0>v/tcomr—r^oi^co-JFO'V j-^a)i/Mf»«uinO'*l

m^o^mr^rvjtn-t -fm>*-ico\n^a»ON.O'nOrvp^-r gjg3^in-Of*>^00'COOO p^fTlro Oco id r j • g to -o r> to n "T^- -Of7-cr— -gr^OTJ T

rn

w

t fti

< •H4J

w a^n a<! oH Q<

KWH(0

U•H4->

a0)

»o

mjJ^corgdO^O^O> ocO'Ominf'^o^cocon^I rginf*>rr>o>rnor^-^^^U. O^r'^—lOOOOOO

ooooooo-*^-o>oooooooooooooooooooo

o a* •• ^ 4*m (*>»n r*>o00^0 gj o^ COO^ 0> rg-^ rs»rg in ^h -0 co 'nO -0^ n-^r*r- OOin J- -4

ac cor*-r o>>f ro-*fr>f»o< cof*>ocD'«co-«rn>«'rcd o,»coo*-*-^^co^mo>I

r-igJf0i,n>co^rgrNJj(^

OOO^njfr)tno,« >r «^

r>g t_Kn \/* <*>-^ -+jr~ n 1 1(|

tn c>»hm <-i corg ft%—« r»»

O "Orgm co >jrn-g^O

oooooooooo

r» rgrgr*.g3r*>rgjrg-^gf>inr»l g^rg;cocoi**'fc*0>r

,»C7»cOf>jr^g^gjrJNcom»f rno>—*coo>comg30

h» f-«rgjmm-^OC0c0OC7*a>-0rf>rgo^^0^-H-^-f

O OOOOH^^^fl)N-t4>NNHfH^)NO^

rsiomrnmorsjmramg3^or*0't,pr>-^ p,>coVs

-T" r* rg iv o^ u i gr r- rg*w -«r» *d -o —«»n?—t r»> ^r r* com<n rg rgtn in« » ^f fn Jf coO <*!*H

« r\jm rg g> rjroO •©O P"" r» -O rf^ p- g> rg gj«mroOrOrgrgr-inrg^ir* Oh^g^oo^rcoo^-*

tn o-^rgrr>in-o<^»o-0'0 -0<)H-^n-t»na)O'N

CT OOOOOOO-^^-grgrggf g^gagjo^Omm

rgO^-g?OOs000 3>'0'^vn'OrgrT>rr|0^noo-to-4 *rr»-*m<^ ,oo»nrgrg-^cD'*/rH

gf o^HOrgngf -<;r*-^rgco^^HO<?r**^rgorg

CJ OOOOOOOO'

-^ m^-^rOrg^Tnornmr^or-rgtnrginr* r"nrg^>aI '^r-.grr-r*nnn(/>P*iO*rgr

rn 000«-*-**rnrninr-corgr^o-*rgrgcoo>nco

(_» OOOOOOOOOO-

h- o^r^0x ^»fs»co»t, '-«P»-rr>»-*0> ,*'*»-^^r^in m nimtnr*-rr>rgogjO'>r 'gom-oo^-rg

O OOOOOOOOOOO^^^rjfMm^-O-O

>f CD *-trr>C7*ing3^<0>-OincOrT>0^ r,JCO

^ o om rf>r**t,r^u^^4rr>*NJ'0'^comoOOO -<OW -4 r?> rr) ir\r» CO rg prj O^O -*^ -O 0>

f^ ^p^oo O'Om-rr^r^r-g^in'OO'-^o^co-^in

O Oo^-^^^-^rgrgrnrnmmcocorf»f*>o>0>*00-^r>rn

coo -OOP'-rnrgm Orfi o>corrto> gfr^cog> -goU-.-a1 u lAU-v t| U ,^^,i| (ov , 1|t-.^ni U \H|

gjjgjO^CT* -Tr?>»-»-^rgrgoo> 4>- >f rgrnorgrgtn

O 0000-^-4rgrgrr>rr>mg

Of-*inr*oorging3corgcorg»rinorn'00»^r*r-oo g5>4;'V'gr»g5r*uirgcO'rin-^'^r*(/.-r r r>.r-oo o-o >T -rr-r-r- <i gr O—t-*r^-i

o oooo-

©O^trtrnrgcoo g3r«--0rr>QOOOrgrg0x g>ininmrOCT'OOrgrggj g^gf g>o ^mrgrgmr*m rorrfintnr-co^^'^aocococornritintnr-O'grg

• •••••••#•#••••#••••CJ ©oOoO©-*-^*-4^cgrg-$'*g>gDO>o,»tf%m

ro rg r* gf-o—* co *QQ >0 in in C7» C?» r^r- ^i rg rg gjOOOO co V» coO ? r» co O" -T 'OOrgr» -^r* u~t

rgrg gf pT)mu> cov/» mrr) OOrg-gr* cwrgr^ rgrg

O OOOOOOOOHHNNmrng1 ^

CO (*) CO in -^ rn f*>m g^^ »* rnO r0 <N—« -Om^Ocucur»r» -H-^-4ni rg *j v\ gj io—«rgr» r^O V -W

oooooooooo- <rgi>*mm«m

g^ co «-^ rg aj> rf» *tO roin coOmm r-on coo 0>m m>rrgrg<vrriO-^^r*>j- ,*''^cDco-^^<-<rgrnrg <•+**«»grgrt)(T>grmr»r»»-4^-«cor*r*r*«4'-4rgfT)

t #•«•••#••!•*••••••*#CJ OOOOOOOOOOHH^HfM^^tOO

rgtn rg rg gj- co ^ -Om-go in co >fO 0> ^O -r co f"»

^^h-r^mmOOr^Oco^Ocorgorgoco-*--i t-H —4 ^^ rg rg rO rr(m gj coO rr> ro •-»—* rg—« r- o>

O" OOOOOOOOOOOOH^HNffli^*^ -J 3 OOOOOOOOOOOO^H'MN'n^-t't

co cost' >Qrgo-OrgrgrgMJ —«J ^ CO -T co co ~* r*O u »•-» rsj r» -gr rg .-« -^o• ••••t«*toooooooooo

r- H H in Or-CD^inn-o-O-T -g»rgo>r«»

< rg m i^ *o<o«it'UO r* ,'tU','y>ui')Hn *^ 0OO0OO0-«0 ,X |N-T ,», «wysn«iHHJ> t I**ltltt«>t«*O © 0000000000000000-*HfMN

<0>r ^<-trT>rg«*»r*^>rorgrg^tr'CV, coorg

o ooooooooooooooho^-^nn

^rgrgromgr>rmOHHNN(^P)t^in

oooooooomo'no^O^Oi o. HHNNmnf^w

30

Page 37: A simulation study of models for combinations of random loads.

CQ

W-CQ<

I

CQ

0)

w(0

id•H»->

a

aoa.XW

aq;

M0)

«M«H•HQ

ZlU<-.uz-«>-WOI-ZwXUJZOUUZOu.—wa.

OOKIU

> <J

a ^ -^ *.

J *: t -i

<< I

aizououxx

0> ** -*m -» o om ~o o

ooooooo-*-*-*ooooooooo-^

Kio •*O -J- 'St so U"»^ —

et fNjf»-in'>jooc7»<o r'40

© ,©^N.O*'»O-*m^-0^

X ••••••••

- -om^j^oooooo— ^-(OOOOOOOOa • ••••••••oooooooooo

O * eors#rg OO'VrgrNjrsi .-,

^ n-tyooouuo<J ^w •••«••••• ^o. oc^ooooooo o

o-^-^rgtvm^^r -**n

O »^- re, _, u-, <> _«0—*^U">^D ^ (> CMW „* ir» -« C?>

<-4 1-«-^—4ro r*> ^ >f to© rs| r»> rs^ -^ia*-^^pgr\|rr>rn

-* -o (?»o <M ni -«r »no <o ao*n -t ^r i/* -o ina -* o

O -» —•O<M •*• f»-O GO 0* -O -# "Om r»—4 *0 "1 rs|oo^ >ooinfMO',,fnNinrtio*A-fi''i/*^G'noy

CT ••%••• • ••••••••*-•-^ rgrg fn >T *> -O (->O '* f*

00 >* -^ © -^ <» rg ,f f — -d ijTi^r'j^i-n f ^^lN

a » • ••*•*•••

—«-H

-*r\nr>omr\jm^m.o -o-^-^tncow<o^*"Cor>i

f* •fl-OOinocor^-ODC?'CT, maO(7*[J*'n^H^, r*rri^

«Mcgcoo*maO!>m<\ir«»Or\#Tto ^mo-O^J-Orno*rno-^«*<7**-»f"»mco'">jO^-<7'ir\'0'*»0-o

o* cyoo ** ^ "O in rn co tn o» r*-m aoin «-* ^» -^ >oy. -^) O'^ j- — r— j- ^-i/» aj ^> 'mi^i ai ai —* rn r -O

OOOO^^^^^^rgfnm*mr»coao^-J, 'nfncDa5

t «•«*•••#••••••#•••••

m 5*h"O -* *t r\tO -^ ^O ^" t>^^ <"0 •*> **> *

t •»•#%••••••••••••••a oooooo^^-*HNfn4, in<x»onmHH

^O^fneO-^rNj t a0h»^^oojg>rrih*rfK)r»>O

f>J O'-f-^'Ni TinOJCD'*>>'rim'K»(DW rtT*l'*>OT3

>oooooooooo-J • • t1 & — ^rityimm-t + m

-_—_^ <

-*—irgrgrr^rri

4>m <A rt)Q p<» o r*>i^> ni -^^ rf>m^ >orv^ra

-^*n f*> rgu> o»cn;« »rg-^fT(o ao<o C7*0 in</> -^ad

• ••••••#••••••••

o*mO'A-coof,*.ni o -^rnc^ ^aor^cD'0'O r>^»

•Ococoaof^JO^MO ft^^-OCOO ,s•0^ ^r^ojo

O* ao^ N- -* O* rs$r\(o ^*O -* -• "*"^O i>^ -^

^JflOO^NNtO /mN^IOH T -T COO TW

^roo> ^cDf\jOfnmrf>*»tf,

\ ^fpcrn 1 ^- -*f»»

^ ^r*-r»rgrgv^or^r'»rgmof»''''T* rn-r i',

» rvtn

r) m«iin<ncoco^ ^^rr><yr^o rvo r^' -o *-^*v»

rgrg-T sir— r*.^^uo(7*i^Ovuo' ,gO'Tw^'-

rj r\jrjrnrOiALnX!> -t ^^^^-^ r>jW^ rr>OT0

C7 OOOOOOOOOOH-H^NN^^ODS

OOOOOH N N

31

Page 38: A simulation study of models for combinations of random loads.

(Nl

CNI

CQ

Q)

W(0

o

t fd

CQ •H4->

W aJ <D

CQ art oH cu

Xw+jaQ)

U0)

*H«H•HQ

I—. _J

Ztu«*-*

Q. -UU.-oaxOOHUiO •UJ"*

o

- -J« <-*UJ m t <\ £X > -3

u. «* .j—«oI < <

-» ^ < «> _j

& <t< I

2 »-»-**

WOO* OOXI

f*»^oin rsiOin'o-*o> »--nOcoor-com*«.>f'

u. -r-400000^^0

O—• <TO -r r* r\# r. > cry .©

OOOHNO"*ND OOOOOOO-"TOC0OOOOOOOO-*^oooooooooo

O^"" CD *7> O -O <"sj 3*» tf"\ •©

-Omino^'-'O OfVNCOlT""'"--«OC>O rrir'-in

at (^^ v*ou-*m«N'*' * »r*

^ c*cu-«r*'omO'T«yco fT> rn ^» p-

) in r* as f*> *"• *»

1 n^NH ^m ct> mn> t>

X •••#••••#•OO **<Nm in co >r en a?

'""J"*-* MnrwiM"«

I C*rnco<'r> r'"m0'»m~«0 Xi— «(^t-«00OOO0«w rn-^oOOOOOOO0. t • U.OOOOOOOOOO o

LU^ -JO **" -OOrNjcococo -»

X 'fc*n ) ^»MCJ(^CJ ^.

t— n>—OUO«JOOOO <— • ••»•»•#•» _}Q. OOOOOOOOOO O

U) OOOOOOOOOO •

I » t - t • • • • •

rvj \AO >H—»O irw*) rsja>oNi <0 t+ •© **> r»"M «"* -fC?> co.ocdO'0'T'"s>w-«*-i-o O .*v/»t>'*>*reo<J*^O •~*»nf^»^rf>OrMa>0 0-^OOrr>oo*>'s-inO's"

t ••••••••••••••«•••#•3 rs*,-«fT>rr>.0U"*O(7,'iJ"\mf»m-4-#iri -OOO o <;^ HHNN#*OflOO'ON

ino »*0* -h -f r»» rn to r\jo f>* -O »*> c?» >*oN **i

o w>-*»*»"*or»pr).rinrneo'''»-«0'O lo<>s*r»^

-*m-4W ># <*1 n l *>Mlc*V"*m

^ cr> <--<"<J-^0 -O^^a-v-n**- <"vj (7* -'>—«'*- jprn,r

ip> "»o-r r>*Ooo>in'T>in rT*-^'0'nt*"C*n,r -oajo*

0* r*» en-*o-O '"**- -^*n in r\i<*>«© tr^fN-orsja* u"*f»

CJ •••»••••••••••••>••*OO—•-^rn rsl in -t <X> t* rs|r^—•o r*vom •> BD «)

oi •.••#••••••••.«..•.00-^-^*N*r>#'n'ninino>»oo -t ^ r*>fn^h*o-^

>t —* -0 tn^iA^rnrgmrso^o^o^^iAm o^-^Df^f^ooinoc^r^fMinoo-rnof^o

O OO^^H-*NNt^f*^NN^C)rg!M^fn

r* -T -j-ooommmtn-*-* ^r^^ajn^o^H^O OOOOH-^NN'f^ -o^*-•or-^-co^in^o

<t> -*o p* to to* to^*o*• h«^m cd (7*mo **

O i>jrn<J'0'^'-^i^'C^-^-^0-^*" r^-r«1'^*i^>-*

O OOOO^'^'^^'^nt/Mntoton'^'^H-tifl

t> Oh-'N'Mtnoto't o^"nOH*> t^^'*!.'^^^-.s.^ o^^Oro-o aoro^^-to^ CT-O

O OOOOOO-^t^Nn^-O^OO'flOiA-0

rg^rgi**f*-o-Ot> -f -otv»^mraom«n^^4vn,-n tr- rNj V r*<0>»'*>CO-r-*lJ*COrgr'>^-00

o oooooo-4-i^-<NfM>r'rNpT*vD^

-^ O f*> D> —* nl-*<Mm >M T5 -H •©O 'J' <NO <MO

O* OOOOOOOOOO-^NN-f •* OOO^

»or\jtnrsja>-*-Oto-*-H >rsiotnfnrsj^o-4-f^-f*>-OrotO'*^^*rs)^^-^-p»Of-'J"m'Oao^5

o oooooooooo-*-*-*<M'*>'*)mmaocD

O tom-u o«^rr)^nj^.r«»f^in—iQ^y •*"

m oouov>^-•,,<'T^ ,,l«|^'/, t*«> ,' , ^ tooO OOO000OOO0OOOO*HNN«f>f

>ooooooooooiuinomov"»omou"\0X

O *n *A-h -^ -t 'Wrvi rs| c?«OW (O -^ -C •*" rs4rg (n 4>t^» o*> rn<DQ*r,'-flo-^—*-^o-o *T-0(/,r*JTtot7«r^O rr|-r (Oo-nrsi^-«^o^oo r0^">|t^.^'c> -rt •«•••••*••••>••••••••O rsjrs|n>ri>-0'0OO*n^^"»'0 ,-<'Vsnf^or^aD'^

< <» <i <

cotOK-oin,otoOtOur\0°s*rginrrto-^-^rs|»n

V tJO'^'^ntrnr^iry^inr*) - *f»s/* 0-^<0"lf*rn

O rs*r^nirr(inini**a>r^rr>fnc>*iu>(nr*^.n»-^a>r>»^*«^rs|rs*rr|pr1 «w\u»n/*^ *

<0^'0'00co rr)to^rc0s> -^rsjO rr>o'— contT-m ^aj(7n/»-TCD-^t,*rs|coin-TO>n'T»n>m-^r* -r

(/* »nin»n»n-^0'0*nin"i-^v/>-rtn>rr* r cr-«o -«

CJ -»_4rjrg ^ 4- ^> -OOOr- O^N ff> t>0*»vn rsJ-^—•-^•-•"«<>('*•^ +j <; — —

«

O ^Ht(7> OtOiA'09>intNt01irsi-4 s|t>nr*)

C7 ••••••••••••••••••#«-4-4rsj*-«m«nmiAcora->rsjfr)-*<~imrrtrr>rnao ^)

pH -* rsl rsjrn r»>mm co a)

rn-«-» n o ^>'n^rA-»ao^ro^'-H>^o-*>st, K.«0 min-^t>C0C?*rnrri'>J'O'*-O , MO-^tJ«rs*ors|rsJ

coco f«>-4^iAi^-OiriO'^ rO^OtO^'* OJ ••«••••*••• * « t •

OO -«-* <>* rs* rf| romm C*O *r *» *n **s ** *.OQ-^•*^fs*f*irn ^> />

m rnm<r*ocot>-H^^o^'Tiinin'>r<i'?iin r*'r^r» r*»^--H r,s|33r0OO (0'Or-»f0r- ,

*,»O'>JT rr»C0''-s|

O 00-*-^-*^^'*> rP*^^t'r*^*NirsiOOrvrs*-4rsi

•"*r-in«iT'«>to-*',*mrgcor>jinr-in-oo^' T^ 00000-00 0'-«-« f,,-a>'Y>0'^J^O-«-<r^t •••••••••«••»••••••O oo-*'-«-^^'M r^*'*"0'0-^-^'*- r,*cocosnin

rsi-^rs|-T"lQaDrgo's"inrs|rs|(7*in^O(>'00^*

4) sT T^f^-rs*rg^o-«—«0-^ T •* TtTV*"r»"nn

o oooo-4 -*^* rs' rrt',">u"\mcD<om'Ti-*-«JT -f.H^rsjrsirnrn

P-O r^^t0^rsF'M^'vf'>»'0-*, -0-^tO>cOt>^r*OC7'0-f-* rt^ fW^Oinrv;t7*^^r*>^*3)r>.C7, *Hm nirn^i os"tMAu»',ivraj{>'T)f> rsf'^o^^'-4

OOOOOO-^^fMfsj'O^'O OOO-O 0>A O

O -O rn ^*- r* -* rM > »-• CD C7» N* U» O*"* in rsi -H **>

r-p*.fT> 4- v*0'»rs#rs*-or^tDtD| vrv/'CO»nfr|rs|^i ^<# rw^sy o (j —,_<r»%N x> -c -g j* ~r O 1

," / ' -J

cj oooooO-^-^w+r^rsi^^r^r-fsi-^eoa*

COO^^^rs. ^ rr>>—«^— ^n-fsj-^^Nfsji/>fs|^)• W(yuO t©I>'«airstf»>^^istujOL»-H«vr-f|

-^ —t*^

LU CDrsuno—•mt?*fNlOCO*Mr«-0 O ^0>* %T *" r*^rs* <-t-* , sffv,)n^^tM^-0'O*'> o rV"^'-'* r*"'s* ***

UJ • ••••••••••••••••»•-j (j ooooooooooH^fMN-r^oooo

z< coco-^mp-vnCT'Orn-T'^w ^n'vvnt^rgco^o >©

3 -H-^c^C7t C^o^c^ln^*Ln'0's-ln^*^nlnoa^pCJ rs| _iH-4H^n * >r«»h»rvrgoo r,

sl r«J^n'">i-^ T

-i O OOOOOOOOOO^*-•nimmm into od<zo-• myivrsr-f *"«')^J'*'<Usris)-o<ii , lU»'* ,t»i'

CJ t ••••§••••Z O OOOOOOOOOOOOOO^^NN^flau >oooooooooo• ujmomo^ou^omo

1 o-"-*rgr^rnm^-rvn

32

Page 39: A simulation study of models for combinations of random loads.

m

rnI

w(0

t

CQ(0

•H-

W -^ 0>

CQ cs

< oH CI.

KwJ->

a0)

MQ»

iH4-1

•HQ

*«OHZ

Oiuzd

OUKUJCJ »w —'•* ow

O t

• o I

*-- UJ

a orzruj^UJ o—

_

»— _J< <3 —Ui «X(£ZI > o

^ cro :>

u. ^ _;—

a

a. << |

z. r-»- **

• ouii

mO-n^- O > O —»<"sjrs*

g0^in i"rt<"NJ'«"ttJ>—<Og>

-r-^oooo^ooo

OOOOOiNjininOiN

oaoooooo<-»-«

^ino^'^^'rcoco'ngj - j _.-~ '

w ^ -ri g> LP rr,

Ct rg rg n-, <j —•—»--, T -U O^ m-*^ <"*<>« l n"> tf«-tff«1

co rsj;ocoON->J''— rsjinp-

I -TOO-0-H*Na3P»-*ico

< *-»gp o\/**»*y»—•—*ijw»

— ^ti^hOOOOOO— ^hOOOOOOOOa- •• u.OOOOOOOOOO O <N

aji>oooO"'5 -O ~h-* rn r—o <0 *n r» g^co

r\* •-» rg CD -^O rg ny -O f\i rt> CO CMTk r*>O 17» P*O

m^mo •Oo,»co-*~*o>J, -oo's- -torn—»*? -j*

intor*-^r^in>r<7'r>i>roa3 0u^h.rv-r i^ -*o

gf njf*- <v -^ rr>OO O WSO r^ -« iT» f*-lA ^OO -4 J3

V*-*«4iO i»i«—•^^0»4/i*n/*n|^r-^grw/ni gp !» g*

r- ^m^m in inm wrm -g> o* rn q* rn n» rno co -*-** •••••••*•••••••••#

<j 00-^<-«'N r^-^^ r"'s**^ rN'^'0<^'*> ir^ tog

^m^th oa^-mcd^-^^O O-orn-^0,»p»pncojj^h'-g'^sfajfn^CHOH^ >rnjp»p*ot>

—«p*m'MCD'0'4>'-**nnmr\i-^Q>»w*^-«p*-

^•-g-eincDgjr**— grm-rmco-iO'O-OP-'gr-gr

3 OOOOW^'

O ^^TCDgjrsjgjc^^rNj M«rrri-<jgjgj»*^oO H-

i— pn-^oOOOt>OQL> ^* t •••••»•• • 3Q. OOOOOOOOOO o

*mro >r »f* <© <B ni -» rv ro -0 3)

c» oot>ot>o-

IflfMHN 00>K^CDlA<iO^CM)O OkACO

-^ir>^rN4fnoa>fno*-^*Atnp*' g? -rr^O'N0> r*-g^r«-«-ir<-i/>(©o-Orgfs»f,»tA-4yr^or»3>o^

-r —«o ~u i/i a* o w^j^o^vjj'jy-o-in r ^ a

-<-«n-im^0-<O-M)OO>^-<-' g*g>t^

rr> rnm g^ %rm a>—•p* r^ r*> co <n f> •*> r><ooc o>OrNJOu"\CM"vjg^m<N g3mrr»^r»r*cDO'7'f,*Wl

HW^WOOT -rnir^uif*lv/*»J*-r « -©m*nrn

g3 4>(7««0'OfNJ rD^*'00,»u^r\jff«rsjgfrnfT>-0>»,

gj OiOOiO-g^ONO'MHO •Tf^HtO^Oooooa^mocp'-^cr'rnrn-^mrga) ^ ^^^

r» <y*o of-OOOOu"tu-\r*o p,J<D prir»rNjrnr,-m

^g3cc^>rncOii^^»g3r*-u>n^ir»Lr»P*"ir» o^-r»r^jfgj--og3(ti^'^' -grorMOcor*-r*o—»-

O) ornr»rno-T'Ti-^insrr-rn^^-^ -g> «z>'nij*rvj

oo^fn^o^^jor-inu/ut-^sr ^ly^oogj o -^ -* v*O n )mm -© •-• ' ^ *» —« (/» t/» in <^gr*o

m<DP*gf-g>m-*,'inm-rinp»-rg^NO« gj^-^ajmrnf>jf*>«>-^in',*^"'ixjp*o,»^'UO -o^o-t v

•gf ninio-g3 0«-i<30Cflo-^tr'-^fT>'g/'r*-^ OLT"i> gr

O OOOOhh#-*-*rNg*^<"nm

f* O N

rng3»^rs*a0tr>'r>O,« Of>0*AinrgyO^-OCOO

rsioinp^mtDm-Oh-fNirn-c-t -to-^-^rgco -^

tr(7"4*m'^ r>*'Jrmrgcotr"^t>'tC7s mnigf»>#-o

mcocor-{7»p*JPf>'*'nj p," r^'MOf>^0'7' l^-g>

tn^f- -Op^-Oco-^-^incD-O o•4> co-J^* rmoo*-»_#rs|r^gr 4" COCO ^n rr>-^ r'J^ Ot^OfNJO 4* CD

O OOOO OOOO-^HN^r^fomoOO^'O -J 3 00000000HHNN(n«)in^OO O O

«1 ^ g;u^'«f^-^f, i^a)wi^iN'ronn/>«gf «-«i^i v*-/ag^^inc»r*(g»T- yr^r»y^gn

3 OOOOOOOOOO^^-^^rgfNgfNf p-p»

n»c_>o-^mo Tf^ornwcogjcotr-to^o^i-r

^-< oo»-«^r>jr\jmrn-<j ^oOO'^"'N"CO<Aiinr»coVa oooooooooo'

<rgr\jrrtrn-4*>4*tnO - -" N fN 0-*^rNjr^rrtm>f

33

Page 40: A simulation study of models for combinations of random loads.

w

I

U

w

U

CD

ft)

>•HuU

a•Ht>i

M(0

5Z ~WO zt-na, lij

~x ZOuJ O

oaiiuo • —•

O •

.2

• ORUJ^»— h- LU

*/> <^ Qa orzuj^UJ tJXh-

X. > o^ ~io-j<a. avii^ acO:»u. *-»—o

I or<*

a. «*< i

z »-»-^^

# OUII

<u<o<N "W" uj ou -r **• aj

I f».©C7-.OOOco*»<\|,*U» NO^N-*-»0000UJ • •••••««••

o^fvmr\jcDo*-o>vnQOOO--rVN -O^csjOOOOOOO—«r*K>oooooooooooooooooooo

mc*<"vjC3-4CD-*0*-*0

C>rsjr\j<-O'*t*©.jrsj«0Q

• •••••••••oOo-*m m*r\ (7» r*>-h

-4<M

u><iv"f i u * -jj >y j-

m

• •*»••••«

in c* *« n>—4 ao rv*> —•"

oooooooooo

O CCNiO>0 ONOO(DD

a. oooooooooo

X 0*H^fMrVft>fn-^N**lA

O* »A 'OO« r^O CT»« -4^ rsj -^o» co r*. j> «> f>» ft>

CJ OON-^J/ftV^iA-4^^tf»K^-4(>^-«o-t

C* Nn^'^iinnr»r.in'^'nrsOfv^f1 »-«'r>'4 a• • ••«••••••••••••••••

CJ cj)0-*^rr>fN#rv«ru*»*i*e'n'T>—•mnnn^NO4*4 rv V« K*>M *< » CO <0

CDr»r>*i> -sT O -T O^r^l^fn^^OJ-f -O -£> r»- C7>

m omr<».r>©r»rgrnn»co-^.oo*'0-<-*T*«P4 j-r*

t **••••••«•••••#•«••or ooHONH*n«^Ho«ooiAOQNi—»—• —»-^« ^'-g » -r <4#^J

mc0C0O^>V>'t-7O'4m4)N9>O^(V)^KN{> U*t*rnf'"trn-^*,*C*»-«<*>-4*^U,»(7'iT) r*>-*'$' CO -T"

CjT • •••*••••••'•••••••••0000-4-**>fM(ft^ONrt>fNJOO-40^^

—*—i r>*^-» rr» pr>^ ^

ty» to ooo r*\ **m t*\ -*»»• oxfT* co^- —4n^ -* n*o •*

t -*-^irnn-^OOCMnn-)f— T C*'"'- 0^*'*>*0'>i-<cj ••••*••••••*•*

O1 OOOOOO^H(>jMt >*n.r».-«wc*coo*aoM»4HH^^

u*>mCDrVCON*'0-4-0*0-ON»',r»<0 ^f'-OojvO'OOniQi^Ki^w v^'^i'H/Oni^V*'^ CO

cj oooooo—^NM't-ro'OoO'O'Oiftj'l ift ««JT>|

(>»noN ^co'N--« rMrMO',^'*'ru'no',>^>-4r- LU i f> JO r_j u —. — , .j (_> _« a % (— CO—• « ' f- <_, i_.

• •»*;••*•*»»••••#•, •- »

CJ 0000000--*^(*)^^^Nf*fMMiDaj

t •••••##•»#•••!••?••or oooooooo-<HNNftimm»no^^^

^ nO*m •Of*'OCJO*M^^ ^^r*>-Ornf^m-0

ooooooooo^-*-Ho**N'n'*'-©'Ooo

-* rntDtnr*>rnniaoo-^-^'Ooo-^«ON-ff'0

< •••••••fl»«*t««t««»*i<J OOOO^J^OOOOOO-^-^-^r^fti •*••* <J^"

IA --CDO-^O-^^O^-OfSCMO'^Oi'OU^CD

CJ OOOOOOOOOOOO^^HiNromiftinK

u. -^ ^- -> »*- r— rn oo^o-^No^torgoO r*J 0000-«0(Mrg^-irv*,~ajr>j'^c* >j't>0<A

O OOOOOOOOOOOOOH-HHfMN-f*UJ-J

*- t7> wi y* oo-^-^oajfnr^^aotD—4 -r^^^ u> n

i v r^ptii* -rmrvr^aj«j^-4j—iv«r«« -^ C?oaot->t->0-^O f'^^^n>rr(irMrta3i/» *r io-^3 ••tit ••••••O CF OOOOOOOOOOOOOOOOOH^N

8>oooooooooouj»nou^OW^O»nOino_J • • • • • t • t • •

1 OHHNN(tjm*^iO

0*> CO-^rsjfv.QDcocOrvaJO^u*"* on-i, o^t7>fO -^O's*f>4-^a3r^^ >f-T<N'r>-rr-C0Ofl0 r^(X>'A

^ ^n ^» <noo <N -^<n eo ^nr»» -^ <t» »n ^mr» as r«i

CD «>*-o<;nT*Jl/*u» -4Ji*cO'-"0-^"»",* »y*'-* *" Hi

• •••#•••••••»••••••••

*N <M <"N# rs<-4O ^T f>* CO sO ^^O *!^ °4»"^HI CO^0> r» J0P*>ro?*r\ ^-^'UC0 , '4'4J'^f>>Ot/t n»U>i#>-4

& Oo**'^ rN* ,M ,*Nr n*'©*"4'Or,»'Of"-'in.OOrgrNj.

tn C> p*0> CD^ i/> ^ Ht) -4 -0 O CD >TOm GO^ C7* *H

CjT ••••••••»•••••••••••OOH-#r>i-4fommmcoeon*>r\JOC*-*p**«.r-

pa|a4CV*4f*Kn^Hf

co OmO(7«tfMn^ *to \sy rry c\4—*mv +*-***•mv• *ni-nco'^"', if'>w^'^» ow,^^^^* h,^'<)

Ql •••••»••••••••••••••OOOO -*-* <*s| fMm fn iAin O^ co> »*) <M -^ **) f>

p^>4mnjrm<i

-<-0-o.eQi>oiAflO>*io-f ^ o*nm*Min-^

r» n->m-00^-4t7» (J>^-«0^'» h,00',)s^^

r-—•—*-4rs|fM

^gJ^f^om^nP>"^«'>m^-^O•»'-Hm«^<0•0aDC^tf \ O ^*» <J* •** —•» >• uji * <• i«-\ uji« itvx-r n ) aju>r»

r» »Nr>ju^inci*O'0<?^»p-<*^'N-O',*-r-yir\w-^^

o ooooo^^-^o**^^>r-o-ooooom^

corOO*>r*irs|rn'MO>O rV»^-*Am'0»0*A *^0r»o<-r'inrvri-C3*^-t^aji/'U-\ •r^-f'j"O^'0^

o oooooo^H-4Nm'n»ftm'*^fsiN^^

mrr»inroo^inc*20-^inin t ^^^^^rsioo'^-*fN in<o -tO iO f* -o <? r* a>o b o-^-^^n'^-*s- com -^-^fv^u-nr>aoo^ sfinf^^ <ja>coo'V>*^ rO

• ••••••••••••••••••••Or OOOOOOOO^HNrxrt^m'OT'^^ >f

*T rn r*> rtl "O -* -4 ^^^* -^ -^ *> -O rg aj -4 -OO CO 05

t ••••••••••••••••••••CJ 00000000-^-^-4-*<^ r>**»''*,,0>000

n*lOO -0 ^4—* -O c0—«O 0> CO -«^t> ->UMAOO

• ••^••••••••••••••••t

u. o^o^pJO^H^org-^^inn^oO^u^O—<oaoLJ m »* >c;tP'<»'r--*, »HOmmn»r^i/>0'* y*n»(7*ri.O

UJ • •••••••••••••••••••—i cj ooooooooooo^^^^NmcitfMr

z< ^o-toocoff-Hi/if-^iiin-o-•^coco'"*o3 <NJ>r--^rn5)f>J-40>O>0^**si^C0 rr>C0^4O'^>CJ rj 000-4H-^N'r»'r|tAO^ONin(>00-*0

-J O OOOOOOOOOOOO^—«hn^^ >t

o

o • ••••••••••••••••••••z O OOOOOOOOOOOOOOOO^^^N

> o o o o oui m o <n o «*>

O ^ H rg M n

34

Page 41: A simulation study of models for combinations of random loads.

I

U0)

w(0

u

• 0)

U 4J

w en

•< (d

H t»

•HMU

c•H>i

cd

>

v»o<Z«--)CL Z UJ

-tx 2OUJQO~OUXcj • w

>-_ UJ^ <— oa oezaoUJ OIKr- j«a——

!

UJ ^X^^x :> oa. a. < ^X.^. QtO.»a. -*_>-- tj

»- t-*-GCO—

»

-: *- i -j

Q. *< |

• oaxx

•V \S* -TO"" f I ' "* u » iw sT

| CD o ,J"» o^OCD**'*-^

ooorg -oeo-^m © •*

O OOOO-^mm OCT*

oooooooooo

O~wg rn in o* © o rgo-grgg>»—

OC -^0<~g O-O^-^g—^O

a toooooooooo

O cooog* o-o-orv-r © •-«

I— >«r'Vg-«OOOOt»>0 ^*- • • t Ja. oooooooooo o

X 0-*-*<MrNJ.rr*mNr g/«-\

OTg-^otn-orncooo-go^-m«*(m»h

3 fN*-^tf>rno*r*-ir>fnr^rnf>iprr|«>0'^<OfT»co>—•—•rg rg -* g> h- gj <"^—•O CO -•O

CD ni-v/^/»^» WtU*'*'UO*>*-•-* r""l/» rt !''*'-»V»g>O* gfo*in'TlLf»gjf*V«gfc rgo*'"-«">co-^gjgf r^cO'g

3 »>*-«grn»r»gJ/r'V^pgor*gr.Hr»o-»''^ingrco

O •*m rg -• gr gr h. o* ij»oorn r* groo —<cor**r\ t/»i*»rncD odo-oo g}-*r»r*^*m''» gjotrom^^^U^ODO^ *0*-^rwf»»rn>rO<~^ ,J0r>»

• ••*••••••••••#•••••3 ^O^^iAinoO^gJinf*- O-0-**»' rsiW0> Of*.

• -*r-gr l^r^^ 4H/*mcO-^^g'C3 -Of^mg^oO(J »««••#••••••# •

—*o*N -* ** rn »*» >0N -4 ~*o in g> >om ^ -^r* cm-^ -H rg r\j rn rn in >f\ C7* v* imr\

inr\jco»".r\j^-T-*>_^oo-^^m -*m'D f''-

CD •OrvfOcOmrO^r* ©-HO>©C0O'"»'rlC^ t*»CD T• inin"»<*r>c7, '*"0*> rr* rr

»rvo'r»'C -f^oor'N

rj «•••*••*•••.•••••.••

t»*CTk -0in fM |^O'0in'©OOCDC0^>0 *tr^rr>rf\

^* m^r^-^^'^>rr>-^0-^'^-^mvnin'nrMrNifrvm

<y 00-^-^*>J'M^-<*'s*'^-rMfsiOO'n rf>irMno-^^^ -^ r\j oj n^ rnin ino (7»

000*®0^c/'Nrof**ina)'^-«Nint>N^«

3 00-4«-»rs|r^mf*>^) 0OO f**f^OO'^t^r*O

**• f«ioo in compg -^ r*i <rom -o -o in r-twr»

3 0000-4*-*fNirg^^-r^cDrsJrO'-* ,wm-OroO

fsi(\»oo^'NJOMnr*o /»coiA-»- -ornocom*n Orgrnino-^OO rr**nmo T^(o^ino^',»

3 oooo —*•-•-*w^^ino^Om-OOf-rom

^O -O O f» h*OO <M CO—* «N (7» r» inf^ -* ^* tt>-*^)r»>r'*'0 0'»«,i'Tic>'Orrt<ricoin-r-rfnN

< o-^o ^mcofr»mm*nrTf-^^nmn^^

^3 tk»r>tnmQO r^'*--o>o-Ojpeocor^-^m

rn O^^ornrg or-o^nco^o-^o'N'r ^jrn^n^in

r*> eorgcoO'^inmcoi^i^^-ococ?, 's--0'Ni-^m C7* '*c^conlr-^cDc^'^^n»nf,|*m^moc7•<^ ooo f^O''»* wo-*-rtJ u >^>'^ ,*' ,^' r* ~rw<-* OOOOOOOO^HNl^n'TO^^H'OCO

pg rg o^mfs)N^o>0H'v®oot>O»^ C7» f»-.co K-Oin-0^oOOt^'* cOC7'rr>

<-g ooo^O"i^'Of*Hino«^n ^^ -J* co -r r*

3 OOOOOOOOOHH-4Nf^iiAK83(M^

gr •-• (^ i/»cOsfrgoo'-r", itr'C7*in',nn

3 000000000000^^-*N(n>liA«

>ooooooooooujmomomomomo1 o-^-^rgrg(Ting>gr»n

0> ocQC7»,f'^m-Tco'*»r— mr—OO •T<DO ,»n»r>j

• ••••#••••••••*••••

t» n»ft>*nm»**«or«»rnfT>rgcD«»—loo-u^^-ro•^-^rg<>io»PT>'Vtncpi^»r»i«»r-»^j

~h ^-«.g«^

in ^noo -rr*O"! o* co <0 m-^<DO C7» ro co co rgm -o -o W'O-^'—i**- *^i^<j»o -oo-u^coocoor"

3 f^^^g^ -O O00 h""0cor,w'*»inmn^rgop««0

CT"*>in-^-4co -* "M —«C7» <n^•r»pgi**rgr»-^O <nOx t/> rv< rgo fs-n *r rn co -O **»OO rslO rs< *T rg o

OfNmco-^ijrMnginr* oOin^OV^^^trr^m

-^-Hcoco(7*c7"s«a3mco-o^»o—*o-^ ,nco'0'*g

g>O h" r**C7»'*>'Omco-HCOCDO-*0 O^<tm-0u> t^» * rr> xvo i* cj n i v/* -v r» gr *n ^* ./* *^» gr ni «v

r* CT'o^'Oino-^NO'^'OoajnooHiA'^o

3 OO-^-^rgrggf^g) OO-^^COCT^OCOCO^O

^(*»0'V'>iH-ro>no^^taDN03*'AtO

r*h.mff>rr,cro p^-*'—•t*f*-0-ococ>o * OO»n »nmtj»y»m^*g*4> cDC7*C7»rou»inrrio-^tr-0-^

3 oooo-*~*cgrMmmtn-oo4>Og)'"-N»r,-mg>—i—f—«rjrM ^-g>

mo fM'vcovro-*'co^rNjorap*r*o^in^i^>C7tgr fvcj tr rio tr co <-• co -^oo co pgv in Vs »^f*

g/ gr g>f^gj^-«-^p»N«om -rr*^g^*'^•lnv»li^r>Jo,,

3 OOOOw

lAr^-tdfO^O-O g>rgjo>fMr»fV—tf»»U%—<i^lfl# OLfsu^cotf, '"*g'-Tt/*«-<a»cu'^v rn<;srg-co«>''g• rn rnrMgf gfcr»r*rg r>*C7'OOr"g(j»g* B^ion^N

u oooooo-*-^-^«xmm-ru-»*»wiggr-^m

u, ogj-^om^o^m^^rT'nf'-gjgjcy'gJ'vcrmgj(_» m gr^cr*CT>rn | gooc7»'>gO''*~*cDinoo<',,71,T»

rg r^rgiirt>vouou>-y«< 4»0"»rvon u »-^'^

ui • » • • • t • • • • • • f •

< Orn ,nrg4v csiN-0,'inC7'0,'romrg'^ ^rgh- om3 ocoooco'vt^r*<.g3mcomoo©OGOox Og>3 rg ^a—*-^^ j- ^•i-»r*rg--sj'^o—* t XJinco^O'^

-J 3 OOOOOOOOHH-^rgnn-rmMOf*)^

CD— u>mrnco^ J-> -H -J U-i'^'M O 1' to* -T '-o *" CO

wu;g,

'f i^'V*'Oi/,v, (/"4;'^U'*ii/, v*'»''*U

oooooooooooo^ ^rgrgmg^tn**

O O O O QO u > O *n o

^g rg rg m ro

35

Page 42: A simulation study of models for combinations of random loads.

5LU —

Zt3*-i/>O^ZU1Q. * UJ—X ZOujOO

m

• UJUJ• O 1

uj 3t-t- UJ

l/l <-• aa.UJh- -i<~—*UJ «UJ->-Z

^p-,^ z > <J^ — LJ-11BL B«t^ J£

ro ^ au>i

Qp ^-J-"U1 a'

<

u 1- •-•-xuJ Z^ «l-JQ. <«» 1

qj Zw ^•cwu(0

u• OUII• o^j^^o

• CJ

u 4->

Id

w OS

yACQ HwC Id

H >•HHM«c

Oa•H>iUid

t>

> «Mtr'm-rO'"V'TOO^

Cj w— —•— -* —«—t-*—-«

—. —.t,u-\—mo-« -o^n

0000-«<M ^**rsj-*OOOOOOOO-*^oooooooaoooooooooooo

| ••»••••••

*- ^nN-«ooooooOOOOOOOOOO o

O cocOeocorst^o-O-OCO —

«

^ pi uv f*. r* 'j* r« »o 'V* »

*• » • • oa. oooooooooo CJ

*rN*r>*nK> >r >r*n

r>* n- co .o*- »n*n eo -$•m *n •* #—•*o—»-^ -O fN*

O *M >f" OCO .TCOmO* T>—«C*U*C7***r'>r^O 0-*>fO rn*^mrr\-4C0.rC>:9CM*'*'0in0X-4OOC0 rr>C*

?> co P» •*"o **\ »rt -o i*l f— *n cotn •* 0* •*• <*> *»

>T -Ofrirvor»^rvoO>-HCO JO-^fnO'v»*"Otnm •H*rtr*inv7»'>*r»10 n*0*Otn<V\nin-*''-c*-*rTl^ fnf^i->«^-Htr\r«»(jX4ntN*"»r^c0J0>-^O^O'^ r^

Hino^^NN^sN(rNinift«foooo-4o o^--*c?»a>.*'^c*in ,n,*>r •r»n',»^fv**>H'*

^ »••••••#•%••#•••-•*•OO -••^'N* -**>^• 'T -O -O 0"C0 rs#r># h- J^> r^ (>4

i Ki li IfNiCM

i>_. O-* <0 COC^^OO ^0^*00*0^ -f^f^O"-CQ NOWlNiAH^dJHVJ'^^OOO'^'^COOt ** ^O^'O'T '(^ON"^ <» •T'*>fVO -O <*»T>V»»^

CJ #•«««••*••••••••••«OOO 0*-« —#<V r\»mm ** >* *0 -O O* co rs>—<m t f\

^•mcoCD^O-O'OO ^r\J-^T"si^»—«>0'r^cOC>m co >r >y^co-oiNiO oco**^ ^-o—-^ ot^O'W

•^ or* ^m -o^tcn-^** tf» ^nion corg -<of* ^rnmo^O^ninn^nf*! o O "> *^ *> *o r*

t ••••••••••••••t**ttor ooooo-*^^Nfsirnmt^'Ooa)©-«H

co-^mcornr*-^^} -r^o^^ ^i>(u^-*n

•V o^"iTflNQH of*fnmn'A r»oo*n'^^^

CT OOOOOO-H-^-^-^rsjnimr^vr^ -O O ^ d

* ••••••••••••••••••••CF 00000000-<-«-«HN'M'n'n^iA^^

r\j^ jjrniTifT> jjnjco(7»o pr»covrvmv»\no (o'f O^O r^<\»^r*T •t»r»c?'0-», 'OC7TorTnr*rN.cP»

v/ oooooooooooou-*-*^ 1^""*'

rsj o*»nrn^-injj^^mp*cm/> co ***> o*^-in r* ^-^ *r ^r^r^n>ro«fV'^0'n(t>-^',torg O00-^O r^*^'T> rV*'>-T',*',-'™*-^m vj-^-t T>t •••••••••••••••••«•#

CJ OOOOOOOOOOOOCK^'^H'^iMN

V3 O QOm-^COiMOmrf>cO r^ rr»<M-«^CO

o ooooooooooooooohh^-n

>© \r\ v r**rm*f •oor*o ,N'tT*'UM^ai*X ut CO ^irMtj VV*CM"*WWII >*^" -«W

o oooooooooooooooooooh

^oooooooooouj»now^oinowio>no

O* r>i-^r^»n-^^»^h»^ ©ojin rOfVr^fTioo,"^V*o* o^o^-t^^cooo-^^w^fn-^o^^No m'^4'•mm>t^•-o^^cr»^n^^>c>co^»^n^»^r -Oin

^cof^mco^o^iMO o ,K*^m^ f,">'0-^'<*co

l^ -* ^''l-^ QW'T aj^r^^r i^r»ir» co r«v^.^r^ ^

m r* .^ c^ co in *» -ro 'T^ '"* o* o» *»•**>")oo >nin -*'(i»3>f*>thrr»o f,»'^co-^<t>« ^-^«\ JJO*n-0(^ cocom f-|ivo T'**''*'« ,«Jcr'O r, t',,»m«v^T• ••••••••tt*« ••••••••

CJ *^-< CM rs»*nf*>m• -O -O O* CO fNir^t^ -OrrirNjOO

3 ••»••••*•»••••#••••--»-«-4_*rN|rj'r>rr,m ^. jj O^T'^'-^r.mN

CO r^-Hij'^HtnrnOOCT'r^C^ *>^»^» OH>-^©t(tl• 00°*'Tl fs» ''•**' * t^NOllUHN J'CP'O'Oa • #••••••

-* -^-^—* -^-* rs* r>i *nrn^ -r j> -OO CO <N| -4O in

rsltjrr^O^^^OinTO »OO i0^'*^'f»-f'flm f* r>-o ^ <n > O ^*^O /iNas^ ^ 0<~N*rg^O-n

«0(*>COOOO'MCO^>NO-T-^0',1JCO'*s|^'inC^ lN

r* r»f- va»*NJ'*>'s-'s*inin»n ^^**^«r\^»^c7»a>

fT)O prl <)COO>0*CDf*rM»VN-in<n ^rNirr>fT>-^r>4

O tfMnr-fSBt^cfw^xocMn-omvjcocjJjaocDO

O 00O0OOH^H^(MN(T)^^m^0i»0

n(>n -^*notn»nm^moocom-^co r^<x)co

»n ^^^inv-O'O'N* cooinf^»*n»^tf»-^inco

O OOOoOOOH-4-^HNNN'<>^'ttf>'0 -O

cD-t, r*-ocor\;corrta>rjc^^^^t)in*f'coc*>^cD>^pg rg^ «-« r* tn n> jjo <** *" «> -*> in r» co fVO r* fV

^ n»n-)fT>^^ix\'4>r*-crO'N-Tf,"0-»''^»incor-o

f<> -O—«0-^\mn'*V-^!BO^ c'* u^ t'^'s"C>-OrOCO

• m r^«Mr>*(v'*>m'rin»n'fc»coO-^'rin'7»'r>'0-^in

lj »n co4»^*r»i4j-^^r^.^O'-''m JJinrv^r.^'^-if*>

UJ • tat*«»*t««**«t*-J CJ OOOOOOOOOOOOOhhhh'MNN

2< rNj^r<trsf^-*^<Orr)'Ococo^m^incor<*r>jcoin3 ^^JDCOO^, rnfriOO-^^C>*C7, Cv—I 'X* J&m 0*

-J CJ ooooooooooooooo-*-—<^N

Qr— -*/<»r*<*-n/*-^LJ>u »-^r>ju>U'»* , MO'*i, i«;"in|*-• ^^ oouoo-^^-^H^^''ii>jsr"iinT^'«cjU> • •••••••••••f>«>»*««»Z CJ ooooooooooooooooooo-*

O ^ -. rg rg m rn T

36

Page 43: A simulation study of models for combinations of random loads.

o h-

*^ OUJ

O I mo

O-~0.UJO w

UJQ •

QC •-««rg

0> ^Q^O^^inun^-OO^rnrgrjirjiQyirorngr^ggj-g0> -0 o-f f^^oin^o-t gjg^^g?* ©<>h-co-*

**-m g> ur\ f*V* fM—«OO rJr-r- CO ~*r»—• rgm CO

tr* *n i^O*"go><j*<>gO'0<~grg<)<^»iVU*0 ,^<JCjcoco• »••••••••••«•••%••••

*n -rocom o* -o ct* cd to ft) in g> co g?o o>O **• *^IT> ajtfi ornii">rg-^rgtAr»rnu\mrg gfrg~*cO g**n

O OO-*0fN*^rnrgj-fnin grgjunoo^OO <©»*<>

•"•o-^un-ocorgo-^ogj-rg^gv g>-«Qr> t^mO r^^^o^OO -O t'^O-^^O^N'OW 0'^<"g

CI rgrAirnrN*rnrnmiAr»r*r>c»^ornrn-Ot/\cDco

w coco g> gr rn rngjm'~go uJ'«Jinrnr^rn grr^gv^in

to *"*OO 0***n rn •-*O»H *jm r»m^ co ^ r- r* y r»»nrnvoMn*ninmr»g> oa>'^HV4 -o^oco

-*—«—• •-• rgr* rt> r»> g> gf g> un r» r*- <© COOO rg —«

I

Q

u

co

% a. -* •«. a.

1 ac<

a. <*-* |

# uun

<w o<r y-TUUuvMQ

g* sy r\| r* rgf— tf> (j* rg ^-4 gr grm gr rg ,-#—i nO c©

m

mn m_j r*» u |—• » >* r» cj i/> lt* n »ct» r^ —• n >h rrj grm00-*0-^»-«"g—«rnrgrnrngr g>g?mf»g3cor»

—. ><-- t g*rgm-«O-«iAO'noi)C0g?'©rrvM

rgW^ * unCT* CO 4* rj^r<- in rnrgo*or-»r«» 4* uvj

f* -^r\j-T ^-fs-'N-—«Oin>rt-HO^in ^ rsj^ CD ©u-\

-* f<> «0 »*- CD <f\ >? -G •* t*VO ^O <^ IT* -*W CO -^-^

r- 0-^^rniT*-0 0**Cr''nrr)(s.fs.njrjcor*r^n^o*, Os

• #••••••*••«•••»#•••

0> P*O^»-0C7**0rgoO-riT»C0'NC0MT^-O',*i',-r^

CP ••••••••«••••••••«

3 0000-*^-*^^f\rvHNrslpnrn4'^^-4'iAiA

0000-*-«-<-i- ^rNjrNjrgrNjmfnvf-*'-*^

w-3

mH

•HUPw•HQ

OOOOOO^HNrOoooooooooooooooooooooooooooooo

(7» ** r*- co 5D -^ rg ynfVMO^-^o^-cofNico^*

0-^-^ |^J pn *rin'^coo-^ frnrN'Ocr»'-«^, w^GOO

in ^--4^ a>o>ocoir>co>o-<—«oOQiA

0-*^MM-^rr>*^invn*^-^-OW"^fn<J otr» V«(N

ooooooooooo--o-h-«-«^-*-*<-<j

o*1 rg <j rg ^^rg rg -y r>O u^m rg rsj t> r*m rn sf- <-*

O , ^ r>-ao'+)a>f>-r-c^—»mauo -*> -T^nM^n

corj»coO'>»^mrngvf*rgO"00"*»o rriin>t'^*o^^rsjgv^c^r^-^-^-orgin-^ino-^^m-ygr

»H r*»rng- imiAN^oonn -o© vo^^r^r*

CJ OOOOOOOO-M^-^^^^NNNNN

o

Mid

r* -^gr co <^» rgo* o^ g? w*

r>^m> -^in o* rn a? **s tr

r» -^*r\ *$ -^ co—| —< CD **

<-»r«mif—.*_#g»-co"ir»"g

^OO'n^OrNiOCD'^ XX gji-*wrn»' - !_* j <_>—• -^ -H> ooO^ncoinrgocD-O•~ -o^g* 'n^r>jrgrg_^-^Ck •••••••!•• U.OOOOOOOOOO o

r» o i-<fx>r* g>o^—*morf>^^orjTgo*g3r» -o ijo'^rninc* gr -rgr rfca^co-^fT^"

o ooooooooooooooo^o^w—

Ooo-^OHO^O^Hin^ona) g*i/» o-^

OOOOOOOOOC»C5 0<_»000000'-«

© *• * <H eoo—ifNitn-Omrg-J-»n gr ca g> f^ -- ^^romo"!^-©^'^ OOOOO-^O^OniO^OU^—•^rgr^nii^O OOOOOOOOOOOOOOOOOOOO

g>o^rmg>g>rgm *r-0'"grr\ t*. h i*- > rg -^ ,~*a"\—*—'n

rsi OOOOOHO-'O'MO^O ^OiAO O—*r^

C7 OOOOOOOOOOOOOOOOOOOO

gjmm g»- —( gjm f> o> g? rgin r» o rg—imm O" 4*rgrnorg-^rrl r*'r»inGO"itf» o^-i'ntO'»'nr*rgrgrgnmt g»gpinmr*r-(j»(/*—•r^grg*' gjr*v*o

000000000000-4-4HH^h^N

rn fC^r-o^fOcorgomirirjo^f*- ^norgrvj'n«»-oofgajocD-^' ,Jg*g'i/»a*grcofng'-^-t;-^•-«•*< rg rg rgrnm gf ir\m O <3 P*» cp 0"Q -^NH "T

oooooooo<->oooooo——|*^-^^^

gj rg com g> in g> r* ir> rnOom in^ - ^r^ h-

p

m iWl gjcorgg,o ro-^g>r-*^.^ao'om rvp«a3 g^

rg ^^-H—^r-^rgr-^mrn^g^u^m <u-v»,*Go^"r«o-^

<J ooooooooooowoooooo^*^

cDmONrno<n r^OrnOirtm-4(7,Ko®0-trg O"*—«—•—*-4rgfgf^>rn •» -}• tm^o-0N l)<D

O OOOOOOOOOOOOOOOOOOOO

g»oooN(DOfNta) —

«

r*c«jniujii»<>rfCJujr-« <.

» • Joooooooooo o

<;mg^rg««»m'*"m»^g3—- .

"

', w \ uj — y l u- -r (.i^ j-

-^ ooooooyoo-«o-»oHoi>io^unO OOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOO

oooooooooo••••#•«*

o o o o

,-h —t rj rg rr, m g-

37

Page 44: A simulation study of models for combinations of random loads.

z ~—u >-

O • i/>0Q. —Q.

O J'Q.aio ~-

'JJO •

• Q <UJO*-y- uj

to <». ote *5£21U— _/<« —UJ 'lazX •> o4 «o-^<at Ot «* * 1

i^ < ao>a.

1 act<N i* -— ULU

_j Z.e*»_l1 a " < 1

Q Z •->-**«/iw_iiJ<:zu3

0)

(0

• UUII• ijOt/>i/i

U

fN wt c

O. o .> fN-J/lHLT«>«k/MJ*O r-'-"*

1 -* CM*1 -f—O« -O lTVf»

•H u. .0<N r'J-«-«OOOOOW 4->

UJU

hJ 3 uca -Q«fl •HH M

0-*'rt'0r03>a>o®mID OOOO'^'V'AHff'flO

M 1 OOOOOOOhjHX oooooooooo

•H 5oooooooooo

Qrr>—i •-•m -rv tr*f» <o o

+J» -*^Ofnn(MninO'Ti

\n ror*-r\jr*-ao.o*»f\im^0) o ^^rsj^m^rn t^QM 1

00-*-«r><r'1«">^0> <«>

«J «/> ~4

Oi

X • ••••••#••

w ^rg^HOOOOOOa • *oooooooooo

ut•^ -J

O r^o^JOO-O-OrN/rgrg *«. «y ~>— -r *r "lo*"*1^ p—

P- *f^-*-*O000O0 <** • • • • 3CL OOOOOOOOOO o

O* Of^-^^-O-*'OO<T'O'*-4'rMJ-»rfMnN0-^'^rna* *>*^>*r^*^mo-«^t,

's*'*>3><o-*ir*^ (N'y>-*ot •••••••••!••••«•••••

07 r>J-^rfV>i»A^a>f^^HOf*»nr*>r*%r*>f>iu>rt^rt>^

f*» Co «^—I -*rt>0*A *> 0* (*)0* <0 Ni**>0* —« *A (X> fVj

• ••»*•••*••*•••§*•••

rg >r o*a t>a? <vrg ino a> -o -ro*^—• •« *• -o *o»A COCP»rgcor»4n<jxcg.© CNO^C^lJO^*^

NO -00*00 •©'*>«>Ox o*r<».^.*.ocg*»i50*"<>«o gr -*> ^m oco^^o^-T^n # o^*-u"\f*)<?>>r• m-ro'>oiO'0-r<^ f,*-oir> tO'Ti^'n-^oo ,D

af ••%• •• •

OO-*0 -*—• «^*^*f% •*»»tma (O CD -*-^ vn *noO

m -^-^-or>go^tr»'*icD'^^rO'r»0'OOiA rr>r»t»nP* <# 4>coa0 ^rr^^^—irg^jco*— •^^'*»r» fvOf>5»

CT oooo^-»NNnm*t',»07i (>nfnh.r*

CT OOOO'^-W (mim-*- tO'CaxD^HintA

« -tr^^in^y*n*OHOHtfl* ?o<i^i3

»nor>J-^'* (I>*A-0'^-^-^<X)CT'C0tJ0d>a>rvor*-tn ON"iT>0'Oi/soino'N rr»N'*' |0<Dmtn'fl^

fNj >» ^o •* * ooo e* ^ *» -^ ao -^a> rgm fsimlT\ r—O f*O "1CO rsrf 00 OO *T\ H> -O sT r» f~gO **} "H

g> O*^«^^'*>if>^^»O-«'«>r*rntnf^*i»<»(0rgvt ••••••••••••••••••••O OOOOOOOO^H-^HN^omt -t « O

c» oooooooooo*->^*^-^«**>*''»pn'*''e

W> ^0*N-fOa)NO'V^tnN 0^'0m co n^coo * «^cir\o*-v» g'o*^ ooaorg*>* OOO-^'^'^'^-T •* ^Jf*(7"^*'f*»0',lf,»'V-• ••••••••••••••«t««t*

(j OOOOOOOOOOOOHHHf^r^Nmn

P«* nNO<l't>t OOTMaoO'^JOi'^ -to»MOO c* ^rgin-ro-< r^'0>/>rgu-\^t^w^

<N OOO-^Or^-^^n^Au^^tO-^rgi^N-O ^ CO

C7 OOOGOOOOOOOOO-«-t^^N^^

~-« fT) »r f»icou»'4>rsi',,»o>y aO't —««^ tOf •*/ O *"X fti'Of ninxn 'T'^'^^/

«-< ooooo^o^oN-^fi^in rr» -oi/»o>rn

O" 0000000000000000000-^

a» r^^^^in 0'<>'^oac»o*rvtnco ,N'*,*iAooiAa> *• •* -o ooo tooN-o^o io**i4>m**)<0<+th

ff> afi> -OO> «!•* <NOW1m rf> ">m f*»h" O 0* ^«rn

Lf *^-^-^*HO*-v>auo'J,, t/»'T> rvr^fn* toi*'N• ••••••«••••••••••••O ^ojniyi-f *«'OU 17"r -4*OOV*oo^i>«vrfl'N

r>4 -T 0* O* °**rv -^ ^3 r«» t* r>« -o cgrn O*in 0*AO ^/>

• *••#•••••••••••••••

p* -^o^ f*noo <-<o -^o rn*> </* p* -o*^o o*>

9 •••••••••• ••

^^p* h* -^mO "I r>JO *Nm rf> k. rs#mo >»" 00 -^ 4*

CO n» o^aJO-H-JO'AOOrnHltirv^N'Hinoi'H

O t •#*•••••OO—•—»—»—i^irsjvT ^irnnaoao—«-^u-iu"v-*—

«

*A sT>tOOO fTl-*'*>-<^0,"X>>r 0D»> >"0«>*'*lp*. f^r^^—. o o r tmmoo^^cr'O a *<*><**

(j oo^H-^HfMN'tmTiiAKNff'O'nf'KO'o

AiO^JJO^AOt ^H*ff)n»«ONin^OtP^ «OCM>-f^OOOOT ^fnTl^^-OP^TOO

O 0000-^-iNNff>'n>rt'0«a)JDH^ifiO

^^tr^O^^O O-o-frgooofM^rgo^-^rao

O gr ^^-r»oO'JMAn»rtinv>fT>r*Cjor-. o^CTW»rg

O 0000-^-^«-«'^P^<N'rt'f>^4> O-OOOCD-Hrg

W3 4>*\imO»'M rV*— 01>0**»aDO •ON-H'nO oP*« r*. in tA COrNJO^OfMVAO-t ^O^iA-rO** ©m rr)fn»A*r\r*co*H'^',*r*ir»tjn«»©O^Or>»r-0(g

o oooooo»4HH-4NN'r>m4iA'0'0 (y, c*

N*lAf»P— r^^>\AO**r *0 0*P*-Om^0^*0^*00P*-r*OO r*o»AcoiALriAf»-^o Trg©r«-r-^r\j

•C Airg^ jtfiy'Uijrg-jr-w^^ © r* 4/ 0* *» OJ

O OOOOOOOO^H^H^fVfi^'rt«J4

^Ps|©^0,*r*)f*>-*\Ain''-OU"»0*'fN|P*oxO J"*• i>troOa>v*'\J^-rgoo ^^^t_»oJO(/'«^^-u»'4j* m r-»^fN*rgT>>fvA©«00>» r>J rr>coaoiAiA>r ^>Q^

u ooooooooo^>^-*-^-««',* r>"T» fn'C*'

IL -^lAP*0*'0*'-^C0-^4> OO^-'*''rt0DiA<sJ0> ^0©CJ lA lAtAfvrsirg^g-o^'^iauiAr^r^int/'OC^-T-T

CM -^•-*f^rMnipn4*u>^0p».0*O ,', *At/»OP»r-»r*T0Ui t •»»*«tt«**«tt**»t«l|-i O OOOOOOOOOOO-^^HHiMNNnmp"

z< 0*0 ©coo^A©o*»^" oooNN^m-^^O'^3 i-Hf\i»^-fs-trv©iAco^^O'O r^™*X'^ r^rnr«.roaoO rg •^-*-^"*rgrgro rrtiAiA^-CO-^—<iA OO-^00'>

-J O OOOOOOOOOOOO-^M-^MrwNN

-^ nr*. —« ^ -ijLp.—t T,fv'^—*" ci uj o *<• tJu-^-O'-wp— uimww^N >o <*»«*> •«•*" «a*t>ju\o<»« , ' ,* r*tJ

O • •••••§#••#•••••#••Z O 00000000000000000-^-4-^

X o—^rgrMmrngr-riA

o 1T> o• t

IM INJ n

o o o o

I O -c h rj N

38

Page 45: A simulation study of models for combinations of random loads.

5^

QW-a<

nI

Q0)

w

u

3

W•HQO

0>

U«d

-

t/*-*-;wi O»-l-4tOO i^»-^

~*(J-o— a. •

%-*—r-o—uuoa to«r^»-a. uj

. S• a.

* Ui* O I

UJD •

h-k- IU*/» «— oae 0CZUJ3aj o»->-

_»« « —

i

Ui ««.a^A > Jm. -*t_»-j^=L CI •»•*! ».^ ocu->U. «_<~l_»

I ar«»•— --etaJ -I ^. «5 -1

Cfc «< 1

s *->-«:i— "l^>UO•^iU J

* our x• uu^o

3 —•0<"VJ-*fWV* r*><N|rf>f*>,*rt> lft,#-trnA'©t/\r»-0

rt1U"\m *VO ^"Hl * "^ "** W* CT» eg 0»1> r^

C3 OO-^O—rg— eg <%j ,«><•*mfSK^r^^r^ -*-*

m fgo t*J^-*cr*-«r-'^o<"rvr'"^-x-*cv'>j,rO''*i*>» —•rg r'"»r0^U*>C0'-»O><"V-*-* ,

"r*0U-» ,30*»OO

o »vr^ rsi •>mo its -o-o—<o<7»-g^.o -o-r **-—

rgrgrgrnUT"r> O -T '"•^0=0 00^0 fNO^I^O

ty oooooooo"<0' oooooo-

•^ o — —-- -j -\> -n ,* «^r\to oooooooooot oooooooooox oooooooooo2 oooooooooo>

^ <^ <"g j"t)'*»'^*-otnuj r^O-J'ls—• -*n *> **> ff* .ft

I • •

>- OOOOO—"---1/1

a^ -— ,_ — vijt-^ -^ ^j

ooooooo—<H

* OOr^* O «MO -D r*

*0 0*^m^<wo T3 -O

OOOOOOOOOO

1^ C7- -« -fl O* «1 <"g—*O »* ^* k- CI "• «V #O -^* -^

^» 0-^ r>*'r»^^0 0<oa3OJ»'N-* rr)i"r>invnr**,o

O OOOOOOOOOO—O-^-*—^-^-^^-^ o oooooo-

'n^t>o^ N*<D-o-*-oO'*'^*'7, '0>r(ri-o ^o

ooooooooooooooo- o oooooooooo-

ct oooooooooooooooooooo

r* %ro^^tocoo*r\*rrsjfvj^>.o-o-o-r'^—

<

C7 oooooooooooooooooooo

*_^ oooouoouoooouououooo

m ^ <*> O 0* <C03'7»fM(>^— O'N* « rg <? v r~>« i^-o—**^ -W-O

00O00—^0'-*0'—O^O""*— rr>—«'', >'"*'*'

oooooooooooooooooooo

m >f iTs -* lT> m -OiTt^i-nO —<~gm -O CT- ^J irt 3J ^O-t^iJ*-^ooooaoo-,o-*o^ONO nJONO ,,i

oooooooooooooooooooo

C7 oooooooooooooooo——I—•—

l

oooooooooooooooooooo

o ooaooooooo*-*c3<-»ooo<->oou

UJ • ••§ ttatvtt**!-i o oooooooooooooooooooo

OO r^ rv^r^l/^<t)'Tl-^-^^/^•^3D•^-—«o -r^'^^-

o oooooooooooooooooooo

** •••• ^a. oooooooooo o

— ooooocjooooooo—o—o—o—3 oooooooooooooooooooo

ooo o<->ooo -*—-*—•-"— -^ -*-*—-* -*

oooooooooooooooooooo

o o o O '

OH-4NNm(Ti>f O -n ^ rg (NJ m rr>

39

Page 46: A simulation study of models for combinations of random loads.

o <o •^ Vo <a o •co »H "» o

• •

1*1 o o -i

o ~+ 09 \t\

o ""» fM UJOJ <M <N O• • • *

fNJ O o ^

o o » <N<-> <•» <*l *~»

o m i»* mp o n in>J> rr) pr» _• • • •

«M O O -*

o <r» h» 1TN

o IA O mN m *• ~-*

O *• •-• f*l

o -< o> 71

OinO^O*JJ^N-<>ino>-< o -* -I* -H> ^-.MnJUO-nlOfM^-fOnCO • • • •

I -OO ""• <M ~t —< -M -T J^ rr> 4> —1O 7> *» IN* O O «H

*"* O ...... ft. ...... ON.-0 33U* |.|i>|.-*»*.«*«-^^»^—ip4»4i^p4H ui u^ v/^ n

[J Ol > ifl N• • •

„ - -< o o -•

0000—•-— rvj rvj-r^^ ojo^h©OOOOOOOOOOO—«-<-N/tN| OOOOOOOOOOOOOOOO o

t- U x <_>oooooo<-"->o«_»i_>ooo -o *n r» n• — • • • •• ....

rj -w* QC OOOOOOOOOOOOOOO -* O O iH

:>

H _* O *•. -o <u

J M *"* (^iTi 0"^OfNjxl">-OOl>r\j m>f\J -o o coL^ ," > O^O' fnrirMO B'M"10N>»«1 ->/ -0 0"> r"(

CM jjp- tu noi^f'iJ'i'iOrg/io^iAS.o ....,3; Iq l_J n^o«i-iO"100-,rV»'0'\'\ «* O O "«

*"* _, t-

(0 • •

• »•••*•••*«••. • o c

— —- —- ur»-i«f-»™<*>"tJJ'Ov«(AJ O 'V **• >>*

OQ.O 1oxo X«UJ »-^»— —IQ.O XO UJ. w a

-< a.

• atII » -< 3_/ iu a:

# < t- 1—« > <— a oc.

a J —

«» Of «* -Jf— ** A. ^ •

1 > a xo_j->—<<l t—

1 1 cc r ~-t——— X a.

3C^1-JQ. 1<t 1

?H»- -i ii >-»'"U'J LU

*-<. J_» -J• cj r 1 1

• UUVIl/l X

OOOOOOHHrfiNlMlMI'l* ~*

so r« -r -r~ O s»- -^

•* ~Q <^ 00nrg^ni3)^000 >T

O

rNJ ^* -^i7* ** . . #

I'oKunnu^v^flM'lD'^'xair. ^ o ^^ >>*

H)OCOtT"-— '0»J^ir>00'IO ,J' ^o

to o j-i *"*i n r>jn —« —•—'o o o o o

I....... ....... • O u'i CO

OOOOOOOOOOOOOOO •*••x . . .

h* O -* •*

i °— OOOOOOOOOOOOOOO

ui-or\jtMO uJ |^"STH/s uwv-*'\i r, »>0

OOOOOOOOOOOOOOO

OOOOOOOOOOOOOOO v/ UJ -» — X^K

OOOO —* ^H •-* -^ ^-» rsj rg rg rsj rsj rn

40

Page 47: A simulation study of models for combinations of random loads.

w

o»«**rt *of» <orn»» Of*o» o>m ^n <oo»n m-om »n^fn inm« mo** -**»<o -m^N m-oo -o*r\-*O f**tf%» o^fx «©o'*> «uoo* o<o rJ ^ojrr) o- _• <x> <v-*o -#rjf>j ornr*» r»mo "^ohj m-^—* mN*4 >rco<oo -o o -/ nno oo-o 'O'*^ «n*r-^ mr*>o ^^o o-c*> oor» a».©i*> ©mf*i o*»m <d^n -•m oooO —I.^O r>#f>*-* "H*V-< f*»mf* *-*-«) .»r*w-v,r -O-0«"> f-f**.<) «9»N 009* mrs|-^ tf>*-rn N»r»^ —OO ">^,|N

<r*»r># <vmr>t +*-o *>»*»o oorsj a> ^ w-»oo mo^O ^O"* <rout o—itnf*>*«^ mmr~**m <*> j- -.« ^m""** itfOIWl f«|/t^ nif*-" i^om UO*J 00 1** ^»ooo eve*** o<x>4 *^'^ "*oo* •©•"••hj ^r*r*r 1*11*1*1 m^r**! -oom •-*<7«(^ » o -©

O p^» «o »« *o «^*^ m»f^ w*»i>* *-*-<-n •nm-c -o-<7*> *f *i o»cof"> oot

*^*-^<o ^*%« <erM«n 'O'*o o*r* o.^m «-rw <yiv»4 w^* moo> *e#n»n r-»o»*> o»^ worn ^inp*^ rx—»-^ m-f o mrmm-O fMO*-* f"»r>4iTt +rw 0"00 «JOO -©-*<">* m-OU> ONm r»o*>< -^,^,^- wr v *j rf>*f*\O pwai f>»*>*« r»*»»0 i>*"^*M r—»-"0 m <*><</ O'O'M r»»»«^ */r»o «U*»0 OO^n -C*« *T-*»n -**-^»n otnO

o ooo ' »o -*-*•* «^o< * wmn fnmoi *><#•»*» <#**-# mmm oo-o »«>»» oo« —-*o w^m«N gw»^

o^m >oo**> "^^tg o^m o»n^ *»-<-si-o ***'»* o-ors* Of*vM ^w\«> Nf»* omm mo^ f>«o«> -o^**O f^O-*" Wh—N lAw^-• -O^r-O O"- O -Q sT***—• OO-*' *rw»«A OO-O •-•OO |>»*^tf\ mO<D ^^rsio **»m»A ID-O®* "i-^>« f-^-»»#> ——*uj m*>*^ v^l^w-v *" -^O 00*W» • »«*•-« rvr-v^U ——•<) UOO ^r*«"» «-.«->-• —o« 0«*»*T9 • • * ••• • • • •«• *•••« ••« » • • • * f • • ••• ••• •

OOO OOO -* *0 «** —•-*—«NMN mr*c>4 mtr\t+* * *rn tntrt^ *-©*n p*f»^j »cd« OOO* r>M-t«4

O 'SA^ 0*-#i>| r>4*0+ (**^-*«%rsr^ #mK ^O**^ "*00O r^^^f^ rs»m<N nO><^ ^m-O *^0-n h»o*»n9 **» (M^>\TS -^O'* Nt>»^ OOO* -T-^O ^rsir* (^>00 TO'OC^ ^%r»'» r^rstvn o—'*T -- * -O «0>A ''•ajr^• OO"* f^rM^ -0<0<f> 0*<O(D <M**»iO ir*»r»^ ooi*» ^H*^«N ©coc * -*fX «**^^> 7>o»» <D(x> o oj^^» p^ «qO • •• ••• •• •• •••

OOO OOO OOO OOO **-«i «i t m -*-»-^ cvtnmM Olf>*oj mmm <*+ r*\ ^-r^- muMA ^<o<o »»oo

^ <x>fNj^ ^>ryf*t Ok«^ -^©O-t-^-^ ^*o-< vtoo N.o*rt -+\f\r*\ f+*v\>{ »vm ^ f»>^o is,»w\in r\j<nNm *f r*-^o *tn<^ -<®o r*m«» \nu%-^ oo*^ a>oo ^-^*^ ^o'X ^>oo cr*a>« »-*a>aj oo-m oo-J^ OO^ . «.*<->•> J- j- * ^-^f*. c^ r» •jv rs, rv*-Nj OtTttn ^CT»d) •"T-*'^ 0*CT*W •^-^^ »^. <Q t^OCD V«9(B <ao^

O OOO OOO OOO OOO OOO i. i i « i i i jp< r^v^J fs#*>|r>j mm'n ^«r *• ** *>MMA <»^-0

f<» O^O —»-0« *\yOV* C^tDO ^*rvj^ 0-0t7» (DO CD <T*m<* f*>®-* 0**^0 0*>*"» fVt^l» -O >O IHaj^(1) "» <o«">4J -*r»»« -*— 'j vtOMiyoMt ( »-«<^ ^/ # <- um«-^ ^»>o«» *'»/«i —a>«C -ch-nt m^itn r^f^^j

pj »a h. ooh oo*n m«^-r iom-o '•'•»ooo mm'n <><?<» ooo v\-** oco> ^o»ntA rsirvr>i ooo ooo* "* 9 OOO OOO OOO OOO Ooo — i— •* w+m+*4 m*m+m4 rv>«f>j r\»ojN tt>estr* mmm *r * s ir\miA ^-o-o

n» ^ mmiO «^-o»*l »•«» >^rr»r^ r*>irt(N

O O ^-rr* u^-r « v^i-mou.o OT>-^OOO OO^ OOf*! rv*^* -^ -«• -*j o o x> <S*<x>0

w(JCQ tO ooo ooo ooo ooo ooo ooo oo-

•H^ «r\ ^ m^»-— oeoO cv^»>^ -*oo ^^O isttnrg p*-^-^ ^«n a>a><"*j n^o nan p^*--o•• m -r o«mt\ sTP'^ocO cdit»o f— ir»^ ou-o ^>^-^ o-TO ^ro-^ ^O-T r^-oo orvnA

^J *l OOO OO-^ 00*x OO*^ r^^*' ^•^o io\r»f^» p*i^»o OO—• «>#r** - »ntr»r» irtgo »*>r<>«# a>r*-o mm-*"

C3 3 OOO OOO OOO OOO OOO OOO OOO OOO **0-H **—»-^ ' ' « -^^<M fMfs»r>i rsirsjca fnrnm

-^j * -r fsi a> m <Ofofr>»* -ro-r w^rn^ tf\-*»^ »>mo o^^r *v^-»n no« -*om OO'*V^" «n -* 9 »n O -c aof»»-r rsri^^i m»*nr «r—*»n *ro^ mmio ir\oo «i/,w« a? -r -r «>o<C

ooo oo—* OO-h oo^ 00"» ^c mr*m ^r * <; -v-o x» woo oor^ 'T>r^«» <o«p* oo*4 rnf*>«n

a ooo ooo ooo ooo oooooo ooo ooo ooo oo-«

A O * O *^ r*\ rs4 r*\ir\*r m^O 0\*® -Om -O N*< «^^ O^-O ->>#'^> lf%f\*«fn 9 <"X «3 ** o "^ ^^ rN* <j aa *(^i^ f^mr^ ^•%/"^" mmwt motn */*•**/* «»iy^*

•^ OOO OOO O^J-^ OO-^ OOr* oO"> -^o^ *V—»-r MNfl -#-*r* ^>m» WHO OON r^f^'T nmNl» ooo oou ooo <-»oo oooooo ooo o^>o oo*j o^o ooo <-»o^-* -*-*•* -*-*-• -*-*—

*«* O * e-4 -o misa>8) «\-*^ ors/rsj rsi<x>4> (Nnrt^o *\.^»«» f*r»m fMfsjr- *ir»-r

tf\ fn ^ o •* 0^_m-*»^ ^*-*-on rM^ao moo moo OO-*" o>**>«^ NOD tDr—rr\rx OOO poo OO*^ OO** OO'^OO'M OO*^ ^^om *>*-** mr^«r» ^^r^ viinm h-*^»o vo»*N «-»-*•<

X O OOO OOO OOO OOO OOO OOO OOO OOO OOO ooo ooo OOO oo-* OO-* •*—-

•+ o o* f*> mm \r\ •tt fr» w -^o»Ht' «>^\-^ ^rrt^to (Do** m»-© o—-o fsif^^ii rsj »A ^» »< mo O O O-^^" o-^»Ti O*^^ 0-*"\A (NfOCO O-^*^ r»-fT>-^

S r<i OOO OOO OOO OCH OO-* OO-< OON OO^O -^O'*) ****-*+ rnrMin mm-o in>r^* h*^)0 «<o-^

O OOO OOO ooo ooo oooooo OOO OOO OOO OOO OOO OO© OOO ooo o©-*111

w4

^ mm t^ *% *\ ^* *T -m >r r» —• (^»'<-*•^nlr•«t»0'^'C^' ,*

< m4 OOO OOO OOO OOO OOOOOO OO-* OO-* OO— OO*^* OO'X 00"1 —O^ *+*m)<r rwxrt

o o ooo ooo ooo ooo oooooo OOO OOO OOO OOO ooo ooo OOO ooo ooo

41

Page 48: A simulation study of models for combinations of random loads.

o CM en mo O o> oot

<* o•o

en O o ~4

o rM en n»o -4 ^* oo3 •-• -^ o• t • t

'> o o »H

O o CO -ft

<J "1 ••1 •-«

-o —

a

^ o• t • •

CM o o -*

O <o <o r»

? o o f4—4 -* o

o <r> •* o*O' O O rtN H N Ot t • I

<N O O >H

O -* O cm

O «* m en

>0OCMf—c©Oc— CMeo-JcnmN.cn-H OCMcMO>0>j'.^co«0-*Ocn"0-fmr»cn'©eM • • • •

I <7«a}.o<ooo(^inu"\com>», fMr,"»cj» cm O o ^HJCM0cncM-*a-(O0000OOO0Q ........ ...ft ft. .^, , N _ _a ,* -ft —I —a -a -a -a -H .- —a _ ^

in

10

U-,

w «^-J X

< *£Eh

T3a«3

"x

*

•«» CM «> V uy C) m >M r» r* »*• U» r» w «sr mj O o o H CM en in r» -H o -r in •r TM O

1 o o o o O O o O —< -* CM en m CO <NX o o o o o O o o o O O o o o —

<

a.

<

CJ cj o <_» o CJ o o o o O a o O CJ

o o o o O O o o o o o o o o O>

•-. r\J —

1

CO o •0 m tn CO IA 0< ~4 e'- CM CO <M> ^ en * m >! CO •-4 o CM p» m en O Cf CDUJ m •*• •o m C7> CM —

-

o •o m CM m -0 *t -OO •0 o •T •* —t r- r- CO CM 'U it- r- •r p. r-

1 -r ft^ o CM in 00 CM r» en O ch o «n o r>-

o •o o -ru v/* o TCO CM en o

ft t • ft

~* O o H

o CO CO C*O in n» in

o en en o

o m o <o —a * •Ve- -* o r» r« ** IJ>

in .* —1 <7* O en 0o o o* m m a-4 aCM o eo vi CM CO i>"

•4 O O m4

O il\ O -T

o m *~ co-t -r ^ o• • • »

^ O O <H

»/»OOO'^»H-HeMeMcn«ar>*'inr»0»O O H N rg-H O C| t7> cm

^ in m h• • t •tVICQ^NON«»o>rr-or>--ocooo-isin^iijoj

I—ir(-^co^a:rgT-OOC^mm--o OOCMcnxrj-r-oc7>cn-o~ft'4(eMOcov>e^cc"- o<n<oco• >•••••••••••»• o -0 c~ ^4

_, -. -oooO'^«-«cMeMen>»- ^-m»-cx>0 «»••5 S M ° ° "•

1/7 l/»

toi/) r»w»coO'^-^ ,Mcnin'^mo>co-or~ o^^m>v>-a"^ Nr-<r— coa-«^;>8rnii/iV'»^o<Mr<»>*' cjo-^oi14 u. <X —

*rl

l l-'" UJ \J UJ (>J 'M nT" -+ ** -r «» ^ ^/ • • t ••v.^'w (DiM»r 1oo>cn'0'-»'OeMOv/, i'> eMo,>r» • o o »hi »hOQ.O I #•» *>oxo XOOOO^rHeMfMfn«r«rme<-coo• uj > -» —i o m ni o—•—-m -» o r- en oo xT x -o cr> -r -r

O uj pj n) ? o -f M -«i->0 cm cm co en cm O co ~» • • • .• w j n pi' ^^ v» pi r* co n» ni in cm cm u» p* ^ ^f o o »h a-t

cMO cXOC7>^^moOOr-cocMcomcncM /*o r~>roocnrj.-o*reM-«-Hoooo. lU0-*"encM-^OOOO0OOOO0 »or-mo

II n -* 3. >.l — O en vj -0

J U) o;000000000000000 X'teM'MCO#<l— p (_,...• >< ^.O-^CM^« It CX ttl

C£(J — OOOOOOOOOOOCJOOO >,«j x *J —' oe~'< ,^e>-^r»ujcJCji/, -»

, 0'-<''^ -^ociowr-qis n yuivr»^»-ii>ft, '4^ujur^uj»>*aj^>ari>* Oh^'-uO< >0 X<U-T, OOflCUin«J"CM'^'HOCJOO CMUlf."*cj_i_j—i«» i— -o>rene>j^cjoooooosjejo • • » •

I I OCX —•»•••• ..•••«•• -o^-Ten(--t-ct 0.000000000000000 —JZ1L*X-1 Xa <! <j | — —~ululUO LLIOOOOOOtPOOOOOOOO ^'uj — — Xi-CUO JtM^T'OcoOIM^-iVcooeM^T-WcOO -IXUJ —OOXX I ............... ft I «• >, 4ftOU^m XOOOOft-tft^ft^r^a-aJrJtMCMlMCMin •X^<£

v-<>o

42

Page 49: A simulation study of models for combinations of random loads.

0)

fN w• to

&H

w —JCQ w< 0)

H H•HP=id

Sd

-** O 0**1 IT»©*«> h»dD -C **-0.^ O-O O^* OO"* *t>f»>0 WOO ^^^ -*0> o -Q a>—

*

»— -O O'*o ^no (^ c+rf> W"*-»«-« *. w>m fN^ r*^>r- *coo rw-r» r*mm o-o *> w>o»f>< ff\r»v>r "#-*o o -rao o**oc* wr^cfc -* <~*o «*mr*j v -on wkdo <«•>*'• ****>« «©»n.-^ -»*o *»-*»••> «o*<n n o -r od"-**i as—r^n o*>*nt • • • ••• » • • • • • ••• * • • «•* • • • • » » ••• * • • i • • ••• ••• •••

CT r*r*0 «* «nc>| -*^-<*> VMH# O-OO «»N 00-> NrsfN K\in«> •&«» **r*r4 »—**•*. rwyi f*-*0 -*> O—-* -*_—

4

. 4. 1 M fc I r«ofr>l WNrf* «W»y*> -r -T * o-r -*»

w>f»o* -oon *«o o*v* *>*>—« ***f4 o*»«© nvno «oa»o f^fn^ 90'*' •-•?«> ^-^n *-o^ ooo9 ^r— r-W i> ^^^ ^jci rry^-in Vlhn •»»f\0 **U»l* VV«V tt»«f« ntr^Hf V" 1*"'! r > -O J -miv • »*F^» ot>^V* p— oo «r> ^r^. v>o<^- o««» nhh '•*—t<u •c*».^i v«r*f*i «cy> ovo **«n*^ t^-^em no**i **g*-4> —»f*»a»t • • ••* • • « ••* • t ••• t • « ••• ••• ••• • • • ••• • • •»• •••

fc» l' "O •>* r*^ t^nn» o-^-tM nmn »—r»o «ow ooo r#ci>f>t: w>*n Oc*f mmm ws« wr<e (*vr4<4« i^ mm~+>* wmm+m4 »* *-^ c*f>ioi n>r^i !»>••>#•> **•«

rv-*^* *><Mi *4«o ^-«>—

I

i<oq ooo Of*w* "rwj'n r-aj-n f-«%*i r*ontf oo-*> «-\o—

«

o-nn r**>*-%•ft 00*>*) 8D — t-* ONO avrvtf^ Wifl ^f«wn *0*» tr\-^ri| mcoas OOO r»wm o** m no11* n#^ro ^-r^

a "'O -*-*-* r^r^s nvn** <*«*> »\«\»n OOO *&«n. Q»0^ »>*r>*-« * * -» H>t^»* i i I O-O-Q PMH

r%**-< w> 9^. -*^.^ -^r^aj r^^o «r<v« * -*^ »-<T>«) N*0 <0-«0 rsK?r>4 V>9fn mVMO k^^ *-o ^^ «^ao ^r «-—

i

o—«o r-oo ^^m>* i-yr>KO ooo »*-*/» w-r's ^•^^#*>l r>r»«© f-oN -*t^ o*ct> r+r*)W\• i*"-*^^ »n f^o t^o-o w>«i<n 'v^^o ' * «qp o—«(^ *^^*o irv-ow^ 0*^0* ««-«>o« vr*r« -uro-o <9«nm ittW* «9 • • • ••« ••• • • « • ••• ••• I • • • « ••• • • • • t • • • • ••• •••

000 -*-*-4 *•*• 1 r^r^r^ mfn« -r*^» «vn> >o-o-o m » » ^»» > 1 < "Vn^ 4^4 000 ^^^^•OD 4> o~4 m^m <T>~0m «W .. I <t.0(X> OOfl ^*^fM lVo4 0*»*t «%^f^ O >*M OOtNlO «0> DON

00 i>#om o—m noN o^o«oo o—•—

<

o—-^ c^-^i>j c>-* -o wcmo ir>f*--r w»—* a»o9 «>**»^» m^-co ^»—

m

• (*>r^rn Q> ^^, rv^-^ >0 Jj-O rv^*—

«

f-»07^» -* tT« * <~+rryrj rvVN «* 00<B 0*%«n ^("•'O r>*»*>-^ r-^ -r r^

7 • • * • • • • « ••• ••• •«• • • • • • * ••• ••• • • t ••• . • ••• • • •OOO 000 * ii 1 "*^«* r«>#r<4 **r««>J f*lmm +•-+•* nvWN -C-o-O w^-»w 9>9>0^ -^.^^ ^ # ^ «n^*p«

r^-^—

(

Oovy rs*^»^» c^^ « --^oo <r»*\-^ «-.-*.-< »«< O*^ tn-^—

1

aomo »MO*S -^vsr^ «n >r « 40>tl*> o*«oct» «^ p*ml rrv^- ^j o— -o f*-r*.«^ t^»c»-r p-t7»«^ f»-^»» rN*^r>4 o^nn m«"icci m^4 r>#<y*M^ r>»>-^ o#a><#— _-_*^j ^> ^j^j OOO *-4>*n O0HO09 f^*n»n 000 O^--? «*>« tnw>v^ a»a»r^ *n*>*rij mqO rvnr^ aooxo

O 000 o 00 -*-^o »< * » « <>*tn^^ *<*is#^# i*wnm <<#•# m*r\«n o-o-o «>«>« qqq nnn -#-#

**^**i rg «^H noo mo-« p**0» ^»*>#fn oo>« m#—ao c^^-*^ ^-^*-^ <t--o -r.^r» o*<oO »o*n#-^ 0>of«4« *« 1 'j^ r*«| wiLf < UtA4l «M^"» r*«>fM »»p^-*i r*o«C 'T*wn| -*— n# m -r aj «»#»r%r» o^^i »-•—^ a?—-^* OO*^ ^* ***f\ aoaoaa r-*—•-- r-rri OOO <r»r>*^ 1 «*4.

4

C*o»<?» o«V OO«0 f^*-^^ »•*«*«• ^**^» <o aso«J

J 000 o 00 000 ^ » * ! 1 1 1 ivrvm »n**v«># «v»>m mm«> ^#>-#'4> nmm p»^-f» ««»<o opo «vn*4

f^ <o w^n -**r«c* oow 0-^»< Oo** o«>-* ^h^p» v\.r<*>4 mr»jo* ^«?oj OOH f*w\^ «v*<n o—«*HV* «\ mo Oi^"1 «-*mo otoo -*—*•+ -+•**) aj^jr^ r^ncr- m^ v f*cr>«o -o-^<» <»**^ r*«*>r«« frt9«4

'O OO^ *N t>#^ irw>.o *»o>»^ • *n| mnm o*o»(^ wvr* v»t^t^ o-cc 4> -4> '*' •4' -o -o -o * *0 ooaooo

O OOO OOO OOO OOO »»«« "N. IM M+*+m4 *w<*>4 r-4*>gr*4 «vnm ^ «r»w\»l -O-O-O ^"«n»« 0*<?*»«

-fl *» ^^ -»-0^- »*«M*5 r<>09 W<yO J5«^-0 fno»o ^o-*" o**« ^0 o-*-< O^-fl ^^^v0^ <?»o»>j-r »• «^o *Ood f*«noo r«ioo oh»n no** <J^-*^) «-*o-0 >o-— r«- r-*»^ o ,»*'*> ir\o^< od-*0 <*\*»r>(

• t t • • • « ••• •»• • • • • t *•• ••* • * • ••• ••• •*• • • • ••• • •

O OOO O OO OOO OOO OOO « 11 I. 1 < fMofrsJ r+r+r* fTvn»n *•*#"# ««m <0«-0 !*-*»»»

N 41 ^e>'«o t*-+m w«o «*%om m»^^* -ot?»*n f»moo *<>«% -om« 0<o<# r>*o^ <o-*«>i o<vo»O r^ r»— in ftfvo f'V- O-rO o«ao«9 (»OV\ v*<x>-Q 0<O^ o <i ^ O^* -r NflO «*—^O -OX

O^^-^ O ON •*—«f*t P»n« •>*>•<? f*r»OB OVO *vr^rn ^»"r»0 C^<WO <T*^ OOO •OO'^ f«« *-«?•<• ••• • • • ( 4 I «•* It* • • « •#• ••• * • * «•• •«« • • • ••• ••• •«•O OOO O OO OOO OOO OOO OOO p<6» I »* «p^ p<i 1^ p*-«f>| »>*)r«#*N| f»Ks#m fnfn«*> •C'* »»«o«n

« 1— 0* A**>*-* •^oo -o»sr>J >m» -*-t>^* ^-^rt Q«to o<N»*n p**»*Q •*«»•<) »^'y**> moo^~ *> r5"*f «mm —^- -*j r+*Mj •+*r\U* v^pk o-o-'-' aumt* ivv"f *«ri^ h*<*«J* 0"i^ »-»«»«.^

000 o o-^ oon -^»*m »->r^^ -* ^- -o oo'* 9Q9> -*o-« f>*n^ oo** ooh mmm r^-«-* ^*qd4d

ooq ^ 00 t>ao 0^0 000 c^oo 0*00 000 w *>^ i^ *^* * ******* ***fH**i f^»*»*n fn*n»»i

• o h* p4 mnm o<4>f* omo m-#N 0»mM ^e^ o*^»^ o*'* «tvn#»j> «fN*o a»v\*n 0—0•rt O NO h-^co »epi* r>Nao oh.H 0-4>0 mfV"! l/»*»tft «i^o i^m»^ •rr*»«* t^^*r»t rv*»aoN OOO O ^I DON OO'I N-«^ "•'n-C *^0 OO 1"" <D<a<7> OO"^ fnf*>* oo*» 000 *-*^*» 000

K O OOO p OO OOO OOO OOO OOO OOO OOO OOO I > I »« f+ *+*+-* r*rsir*i fMr>*N «^»»>n

k» o o »w**n irv*^- —"0«r* ov\»4 noo ^»oo-«* or^*v^ -*mo *> «ooo ^9^ o-*oIC -t cr« >r. <~j h(d9> mo** r>rs|p* BO ot> ^-^rs* — <x> <?• «4> rgo O^m 0Dir>»^ i*^>0 *r*aoO N OOO O OO OO^ OON "*ON f>*Nm nm^ ^^iA ©On a>f^»a) OOO m*vrn tninO 000* m**>r>

• • • • ••• ft)* ••• ••• •*• •*• • • 4 ««* • • * ••• • • • *•• • • • ••*O OOO O OO OOO OOO OOO OOO OOO OOO OOO OOO ' * < *-»-»-* Hp*-i « rsttvrM

v* f^ — * m> tj ^ h«# n«^ m^*^ rv«j»r^ vimio mm« o-*" 1** O-rt^ <*>«c 1^ « r* O ni m 1* v»«nN ^«if- «»-.v »c*«x *n.-* r-mw iw-u ^wn i>»vr«V -^ OOO O OO OOO OO^ OO-* OO-* OON P*-^m fNirvrn r>trr\*> + r «-> irt*r> O #-».<>*-• tWT* —»0-

<^ O OOO O OO 000 000 000 000 OOO ooo 000 000 000 000 000 000 **-»*

• >oo 0000000000000• UJN^r oWON-C^ooON* o«0

43

Page 50: A simulation study of models for combinations of random loads.

0)

wId

u

c

>1-1

0)

ua<D

T3•HM-l

aou

>erID

Ui

oI

o ••••«•• • ..«•..,-• «4 »* ** tM n|

_j m h ^ «r m <o -o o r~ <n (7> -* m o» <oO m in w» <•> "» .* -« rv) in T 'JA u\ -^ ui (_>

a-oj-~oo<3 4, -o rn rr'if,'iA—•'-•la*

r»iv;\/'v<vuPii-*<Ohnuu>-^u*<or^_i-^eorno>r>'T>'HfvrjcorO>n5jr»io<_>•••# »••••#••••

—• —« —t —• cnj

O—'•-«rNjrnfn-*u-\-or~^Orgini-.o-* •-! -* —« fNJ

O^rNmO-*O-0u-te0u"*a>»-«e0C7>-or~Oco<NjrgoO>'NJ<D-*'m>»'-«>03r»r^<orNco~r.-<coco<7>mr»r»-*rM^••••••••••••••lp4 «4 >H

a>aor~j-o-C>r- r-.u-imr~(/<rgr\<0J4)h«hnivoi4<i s;ior^r-(rn\nO • • t *••••«o—'—•iNiojrnr^^-j-i -o r» o» ^ m mH •< »4

<MrNPiN-r>gr»^v<uj*>«-<uUH»iNvjr»|cr^-'-gr-rsjo..-r| <^>r->oa3 rJ-j--r •ow»o t_> ~« »-i 'n <nj pi pi -T ui <o «* w <— r-\ m

'tf> «-« O *-• >T fNf""Nj- v »* rs< cn m >« mPI W N -O O >•> O •<> «*l P4 "• "I r» * *3* •oo©«*•-<^M^gmm•»• in 4) r< d o in

r» m >o M W O ^.^0 «*»>»• •© vn n» <\j r>J m0<*'-HO>», co<\IOO.e^-*mmrx«0_».... t4JOO»*-*-*<MPJmNv *• U> •© cO 0> i-J

ir\H*«-«OP*'"*>*>«u>«itiir»inrN'f>iP'tP) uu ui -y v/- J- i_> u\ ui —< v UI <^ P| UJ

if* PI r» »-• u> v* *ro«~\rv»HU*-«u*>^oC7OO^<^^rsir0r0uftn<Qf«.«o<^

>oooooooooooooooujr\j>r«oioor\j>r -ocooin-t-owo

X

H4

Page 51: A simulation study of models for combinations of random loads.

V. SIMULATION ALGORITHM

A. VARIABLE DEFINITION

Tx; ={inf t>0 ; Z(t)>Xi }

where Z (t) =superposition of constant load and shock load

Nrow=maximum number of level X; ; X(

<X i<X i< <x*aw

Ncol=maximum number of repetition of generating T*

Lc=constant load magnitude

Ls=shcck load magnitude

Tc=constant load interval

Ts=shcck load interval

Ipc=ccunting the number of repetition

L=counting the maximum level of X;

E. ALGORITHM

Input parameters ( XrMt a* b, seed numbers )

Initialize the stress level X;; X,<X i <X 3< < x*u,*

ipc=0

Repeat

set T o =0

M=0

TTs =

Repeat

generate constant load;Lc

generate constant interval;Tc

7ind L such that L= {sup i;Lc>x; }

Set Tx; = to for 1<i<L

M=L

Ci=Tc

A, generate the shock load;Ls

generate the shock interval;Ts

B, If Ts>Ci, then

45

Page 52: A simulation study of models for combinations of random loads.

TO=TO+Tc

Ts=Ts-Ci

TTs=0

generate constant load;Lc

generate constant interval;Tc

Find I such that L= {sup i; Lc>Xi}

If L>M, then

Set T*;=T for M<i<L

H=L

Else continue

Ci=Tc

Go to E

Else, Find I such that L= {sup i;Ls + Lc>Xi }

TTs=lls+Ts

Set Tx; =Te+TTs for M<i<L

Ci=Ci-Ts

Go to A

Until (L=Nrow)

Ipc=Ipc+1

Until (lpc=Ncol)

End Algorithm

46

Page 53: A simulation study of models for combinations of random loads.

APPENDIX A

COMPUTER PROGRAM

*****»*** -^

* rg* <* 0: -»

* X « NJ"

* 1— —1 <# < a:* X O t—

1

w« Z f— _J* -J h-4 X z <* LU 1— h- LU >—

»

# > < z >-* LU CC z a z* _l LU a LU* z 1—

<

X z* oO LU H- LU# 00 O 3 a* LU CQ • *

2* CX LL -H ac* H- O a: LU

* 00 X 1— 1- .»

* 00 00 LU a:* OO -JZ •—

)

z 2; LU

* a LUQ Q a < i—« LU >-« O —

> cC LU

* LU LUh- -1 II H- < S* -J< >-iXO—"LUQ. <# X O •"SI- II hh a:# LU CO—

I- XLU<LULU<* »• 00 _) LL ^Q_ OCT 01—0.* * O LUQ.LU LLO LU Z><« # < a:u_o_j O LUcX-Jt— CtLU•* * O i-orrx >o <—

«

Q•* _l oO 1— >'U l-H- LL>Z_IO# ll—iqc:q_) X z~< >0'-"0<l-# 2: z olozluzx \- <_i»— aco*-*# a O Oh-h->x i—

—•0:0:2: —<z* <—

^

*-H o:<z—iq:v-ll 00CQ_ILU< QiO* h- t— QCLUJI^ZLU z< —<CQ QQJ<# *^ < LUCQ OHU. '<iocqcq3:q<<s:«• z Z icyQQ<zozau30<a•« —

<

t—

«

2:z><<2:~ << a:cazo-iOQ« u. cojDzao ZLU^LLQlO _J <<* LU SLU2! JJQQ<S a:o t-oo* Q o> 2: << LU 1.l lu a. ?• h- 2? _i -J» t_i LU 2T "5 1— V- C3O "J" LUO LU _J -<^<r* LU J -3 3: 2: 2: -1—1 _i :o a- lu>-<>-*;*:# —J LL 2!i-i<< LUO-LU-IZ ZJZH-OOOO* CD Ooo~ XI—H-^^DSD<< a:3i^ZOQ•» <I i>o x< 00 00 ( _> q:< —1> 00 t— aax x* •—1 LU lu «q. 2: <C^ (_>>-/ >—C)<t 0000000# QC 2:1x2. UOII !> 11 11 II

•* <5 —•1— II II OOOOOO 11 11 II II II II II

»• > H^2_| • 11 oca a:* II OO II II II II XX Za:a.O.-t<NmrO# X II a:0-J'-<—I—" (— 1— i^LUa 1— CL<<< <* * (-XZZOU^^Haiu)Q<h«ci:a:a:cil# #* ** ** #

* o* o

* zLUQ3

* LL.

« o* —

^

* o*:* ^HOO* — »* X-«* *o* -JO« X-t* •—

•* 0:0:* -a.

» u. •* »—

* LLO _»•M- »rH O* _^. o•» oox z•» * —ILU 3C

* * 00 » o*«.—. o:* * -to z* * 00•* * »r-H LU* -1— _i•» ox •-•

# 2: •+- LL•« < —I .—1» o: o »o »« o 0^00* O 0<"^CL« cC inn?——

1

a- .» .rO -t rnmo—

'

# z <M< »^-i

•M- —< —a:-* * -^rsjrn-t# <i x «02: <<i<<mr-p-p«

« <^o:oooo^>^ >«>^»mr,-n)i>-m•»• sfo:—-lu t • • »oooor«-Lr\t^-rn »0 O*«•«• »o:o—*rsif-tr-i t • t •fNj(M(Nj(\io • •

# « _J^-tQ.LU 11 11 11 (t -^.—t .—1.—• H II II H IIO^O

# * <<(— f—^fMrnxT II II II II .-KNcn^r—l II II II

•fr « LUa:i—^<<<<—iNKl-tXXXXLUJOQ# * o: •"-! 0:0: Ota: LLLLLLLL -imciwQX -J H-

*

OOOOOOOOOOOOOOOOOOOOOOOOOOOO

U7

Page 54: A simulation study of models for combinations of random loads.

LU *H- *

• < * #II Q£ < *_J a: *

> 3*o •» *

•—I > •» < »aC 1—4 -I CC -^ «.

% a:-* a k >t *QC* o <M < *<— 2 < CC #

1

Q-3C

or•

1— tJLU O r-H < *Z o~ CC < OC *< r- z DC »•» •i •»

\— 00 - > •> H U *oo • OsJ 1—

a X o *z * • l_> 00 X •b #a xo •» > m o *o —IrH _J X z #* •-LL o GO H LU •ft

•» "-«» » •> X oo Q *X •»• vt > -—^ •* •» *t-\ - II X •» XcO ^ * •^•» —o M CO h-Qg -»»o0 * CM^ CL< • CC XQ. X » # UL•» XX ro a. < t-x # i9 LU X m X - 1— h- * 1—)

* *-_J *-4 Q h-X »»»— * LL

# > 1 «k o oO 1— h- *t •> * *>

•^ <M o » H— <x « —1 «•»

(MO X o> CO X •—

CCh- * U "2

< •o >—

i

X 1— x-t »LU * -•—

.

O«frf- ox •» a •» UJh- <(NJ * # _I^H QC<< f-iOO —A I— -1 ujq:<_i * O— -* -^ z0£O LU- X O _l * aca * CO O m

•h •» —

«

a o 0_jr\j -o * CM » •» k —1 r-t

rot- • X • o z -»UO<QZ « X— H i-H H O<0 II r-A o —

»

m -IZOQCZ » # l-tr-4 •» •• •» r-t

oca o » Q. -» Z£ o -z •LU2 •»• •« »^» r-« ^-( —I-z <-N. •—i i-« o 03C ^00 * * r-*< •> te a

c\i>-t SL - k + a:->ZOIO -DC * # X < CD H< >• o _J Zo! tfOoi^Z «• * •—•cm > •» X ITcc*--_J—

i

-J o aizoczo » * w u. —

t

CM — -jo aH»2IZ •> o XhO -2 • OCX * H • X X <OQC

rH OhOS X zo h-l-OCX -*:ZI- * oo 00-^ CM HH •—

Q -Z -»

<- i^zi/ia 1— wO •b »ZH Xoo » «k » 2: ZLLX ««• —

.

2T-. -» ^or* •>oo<Tooq' _J •» »CM xro » 1.1— •

o^^ $< n *—

1

Z-H^CM MO^ it?

LU X^J-O -axx »x ex 0-1 • CLa.auu LLO » —nsoooo •»Q —LLLUO H-O-t-— X*- «r> * O OHX# X* -H—t Xi—i »azo.^ + a " •(- o<-3-<xa * O LU ••XLU-".'jj-» LU^-»>-» •

OpHX--a— it _) rflO Q- •'H-O.Q.CL —3: * a Z—1—

<

OO'-^OO'-^Z 2X tuC3 •»<—• O" *"

<

x; >—«UJ ^u.Q.aiiiixa£i<i-u * a. HO . ^-Ql 1 aooo^« •••» •» II —O < hODUI-OUU^ZU * CO i— ot:_i<_jcD3 1— 4*LU3 •

i—ILT\wh-CM MO -».xz C£CM—

.

ZOhWOOol hhQ * Z3 Z3<MJJ_J II -III h-Q o# OOJ•-H || || UJ< •x »o »—

«

II —

1

U.LL >—• • «• 00 O* 0<—*<-iLU^ O—ILUQ^OII -C—II— 3IO--<0^—

>

_Jh- -J *—

(

(-JXJJJJJJQ. * a:_iiuuouootuj CC<h-Z—OOOl-.c£i-| -r-K XX2 J^H Z_I|-_IJJJJJQQ * CD<.h- djuj^ 11 u.

a.ccotx'roLL.'^LLO D <LL 0<LU<<<<<<!-

2

« » Z5LU2! O0i»-«5tfHHhZZ 5LL ». » —Q O o •» KJ^J OOOOOOOO LU •« * 00 Qtt Hi 00««•« •« * «• •J+

O o o # •»• «-«

r*- o o o # rH—

1

l-l CM c* tt

o oouoo O

4 8

Page 55: A simulation study of models for combinations of random loads.

00a

1 H- Q1

m 001 _j to

1o •4-

1 o LL1 z to

1

1

1

1

1

1

1

1

1

1

1

to

oaczrH

o-Joto

LLto

<NLL.

rH

LLto

ai—

"*1 X Ova- 3

-J 1I—

1

ox oo 1 K i—1>-( oo

1ro —^ » zz 1 X xrn •k

»1 H »X i

2 1•» •^H oo 1 <N l-t •> a:

£X 1 X —r\| zz 1 l-H QX— m

to I» *»-tC\j o

a 1•—

1

— »LL 1—H- 1 X r-H—H •• o » 0>»

1 H _X-< o -•X oX i-t 1

•i 0"-«LL c* 3TH-00t— O l-l 1 o •»••» o «o•«_J c* rH 1 Q. -»_!>-» a QiOOOuo rH 1

rH oou hOZH -»-•-» -»

0.0 o 1» oo - r-t »hQ OO O

—Z h- O 1 O OZJ o i—i »S—O Qs b ••

to to h- 1 -J in »o OOmh O •-• >H—

»

-"i O 1 to *:s * -^ h- H-tO-»Q » *o —o o O X LAOCVJ -* _J "O— (— -4 pH*H •a; ^, o h- r\J acTX ^H QOO---I * »uz "^—

»

-* X •k «->Z_ + 02 + 00 Oa -J » O'-t as k- X X » * -J—x«-a—IO - •h- «—l |o + h-O W (— O^-t o ——<ro;o— ro sj-

<rH O-J II a: LU o »_|X UO<LUZOJX Xa LU •»mo "^s I Q _J »l—

1

Z_IOQi iZQh -(

o a:o •o zu 1- Z oo<- —

»

—2O C?— QC—rO —>r—

,

Oq.lT\2 «CLC? < •CLr- •XmK LU •zzu.zu.o 1— l-H M«— o >-«LU LU C£-".t-»H-«00 o> • LLOO «C» »CL* Q--»-*~m to •• <—> • -j » • ^ 1— -tO z -tLUh- —< LU 1— X ~»X -»

a»-< ZXUJ ii•—••—< o^ < LULL QCO o • _I_J_J'-H lOUJHLUrH

LU Ltl h-a i-h w—» LU-HO + z CC z -4-UJO o II o •JJ-O «00>-00»-<-t »C_J zra: • XU-^ + o CC LU •-rn-H- O • t—q._j<<li_q.o o a

1^1- 1—1 LU -ru or— *£>-* u"-o _> *£. (— LI ILU_IOOrHOO>HOLJi_>'-"-'—I—I II —1 II

*:— Z h-o^cl .—4 HI—

1

t—l—< • 1—

Q

LU ^ »<!-—1 11 ii II >-»w— «-.w_JI-._J_J

II LLO OC IDLU_J>-H>-«rH H II II o *-tlUZ !J 0(MLUZ<'-1 QOQlLU. LLLL<0O<o0^"-•CJ3^3 D|-<-M ZOOO II H q£uj c£u_ q:>-.o^ 00 _J (_>>—< i-t >-i-^o o

II t—

Q

aZLUU. II o a_ioaou CO to

~<iuz. CD-* ac*-iZC3 o <—»l

_l a*-*-»q:uj Z5 LOLL

00 to

#rH o »—

1

i—

«

Oo 1—

»

f—

t

O rH1—

1

1—

t

1—

1

ooou ooo

49

Page 56: A simulation study of models for combinations of random loads.

r\l

O

o

Oo

o:(Nl

O

3o

ict:

~i O•—i oO* +-J o0~«l—O— II

oo

00*—

t

Qo.

o

O-iQ£00Z I

•I—I— 00

_)— -JO._l -=X H •—

«

—Q^ »UJ

0>—•(— U.I—<X>—n—iOO+ 00—1 fHt— t— "-H—J II O_i+ t_i z: ii — h-h-oOOO—IO QOU.WII 000<0 O —MHO-J II OO OOOO'-' OO^U.

X

ooh-•—too

• o<—

OH- OOII II IIQQhoo H- O0

OO.O II

H-OoO_J~t<oOO

oo

3oa:

c*-»x0:21—o(NO »o

qzq.H" »!-•

i-h »H-O 1—•*-•

o »->o— _J »o—OO—O »_I-S-IXwQX<LUZ0>•Qa: »0

"t-ZOO(NJJ>-it— UJ• U.00 »0

-J •-"•—UJJH <—00—0u_<a.<u_o o«-«

II II

OO

• +o o o

OO II O •0©0-• »h" »0 • • ''oowo II OO II

II *t•—

• II —I II II oO t-io hho _J —1 a.

0O0OOOOO 00 •-»

<7*

o OO O —

I

oc ~* o oa>

o

1

>*—000

1

1

1

1

1

1

l

1

«or\J-H

00—Q «X00 ShZ:> Olux•> 00 • •»

c0 »v^-»x o<o

l :> ooq:o •»

l > .—

1

Il cq »r<"»w a>1 ct co<ac a.—l a xcca. a:1

m. > »h- LU<1 O »c\J|— oco1 00— CQ<t » CC 1

|Xo£o£— H-X

i coi— a. -0 »j«

1 XoO *-^0 OC1

« ».Q<t-t -»<l

_jac"ooa£— o>OQ. »-CQ 11

00 OH- co cO a: XZHXSQ.Z H—

1

1—

1

•» » *.| ,. <^ ft. —>h- 3c0— - »_J CLLU00 Oa:o<o » O1—

ocaosu^ 1

H» Z<OI— •» » >H-< *. •.m »o-j UJ00H XX .su » -i00 h-h-iT»h- —

«

1

•+-CM •—• > XLU X ••—1—O—

1

- cC-I —XX »oo -<.Q. t— 1— 1— (NJOO XC0 3S

S Olu -Siaz rH | O< Q. p»-^00 » •* X CC00 2I-4-0 -ir>3C-->» Z

0<O^MQON.I-«» •>

UJ O0C—<"3—C£0 » >*<-•

X — »xz:>ox 11

h- UJrrix^-h, •h—1 > 1—

1

Z< SoOQCvO—LU •-•CC-^OO »LU>-H O(— l__ «* »—»0Uj<Ici£^»O

z O OfN—l>TOUJ|— i<l 1 rH

QC Q. O< <x 00 H- >-•a. cu u.

O 2: oictruji—Hzo:a 1 luoH-Q co -a: •—h»2u.xqqwz J—» HX - OC£LU 00< 00H- *•

QC 1— h- X-«. •> >f ** * «

a>sO

o> 000

50

Page 57: A simulation study of models for combinations of random loads.

-»1 * -^

1 j -^ -•<z 1 •—

^-QC* 1 *" X 1

r\l 1 X » —«^ .» ^ si- a.* <xZ 1 # CL •»• CC\U# 1 X — •»

m 1 « LU » Q.rn -»-^

1 # x< -•1- 1

-"» ro -*ujq: z<

u.«•

<\|-»* CM

l/)U0

1 -1 <1 II c£.

1 CD +1 II —1<-»-»

1 -J—X1<•"-•

1

rg | + X11 o-< 1

CO »< —

»

<—— -«

ro-» x-Km •«

^ 1 hxa I-O+ CL*->s 1-4 -J 1 z 1 x Z •'H #1 -» O a 1 LU—LU -» LUC\J< O-»—

1

>* 1 za# 1— co Z + QC •« 1 z 1 ox— *2: O—— tNJ

z

O w1 Q.U4-.-.^ a. -»— •—

f^w 1- K 1 x*^,^^ x-»o 0.*h —

«

< I Lumx<< LU>-iO Xz< O -J a—

-

l <* 0:2: «-_j LUin #0 O o _jCQCQ 1 jtumsH 1—x< *• <\l_l —

»

LLora: 1 < +<rci^- z# 1 <-•

(M >* 1 u. O z SQ.Q. 1 0—ia£<-H- LU^--» <* * # r\j# • «-• —>> I

—<< | QC-* cc<* a:* * * *<t O H» -^—*<^

1 l-Ci:-.—— LUCt'— w—

-^ —»-^ ^ —» * — • < CQH-I--* 1 Z—— 1 (M LL# X "^-a _l ^CQ CD *m Z>x O O c^ctror-* UJS-»-»* u-o-it

1—• ••>r •*» o •X X H >^-^ LUO _J a.ara»»-i Q—»-«h-H».

_»<-• -1^ 1 1 1—— »•-» —i - •© u. l

0000— Z>-iro—*>—' iH Q(M< «~l

o— z O—

»

zzzz>r 0_l-»-$" V. 0* # X < <XX< ^.QC «.

ox — UX^ ^ WM—#Z _l "^ •mm* i-uLuct: 1 -fro: uua.— •»

ZH- ^ ZH *C\ "vf 1— 1— 1—

»

«- ^ z. —t »Q 1—

1

f-4st-r<M 2:0;+—.— OCX* z»+ < • 11 -•</: •—II/) <K<<Q-» m —•-•»— r-» 1-4CQ— OvO<a: <—tCLr-lX. <LU-» X^s: ^-^.^.1— — 1— OOOOoirg r*Xw + —a:* • • rjjQ-UJ <X<— + -H *.

113 -J CO II *TX + X + _i_l -1 _!— •«II HXO t-t t— Q-CQi-HrH 1 H- ^~»GCUJci:— —o< —

-^oolullx ^ »t— <NI|— ^LU LLU.ILU. | # J II H-O>-tLu< 11 q: 1 4"-» II CtX— 1 N.HH x »a: ^O ll^^sll HH«—' 1/5 —003 ^# -J-O x^» II DO-—Q.CQC0Q.--«t———t-KT)^- -HS <.o2zr-»ooo«- II II II II Z—lf\l(NJsl- J-^500 Ot— Lt_ >-»Z-J»-«'—C*:c£X*-< OC_5«—<X O 1 — QiOO")-h^5« • »OX(M(M^-t -»ni/)LOi/)T-'— 000—

t

-.-^LL-— II Q.G.LU-— LJJ O* Of— — f\l "S.

il c\joo»— uo~-OO^t-l'lWWWhU- II II 00 II II OvJ- h- 11 coco 11 11 11 ixiQjH>»a a* »•

2: Z II X II II 001— 1- ZZ II (N>t^QcQ II ZcooicCrH^aro-i—z<< + x z^# •30 OCOh- (NJ-J-O O II 02:2:—i/)X~iQ oa:aQ.jDQ.h < 11 q:^huj <<m >-i

WQ OXLUOOOOQ o^oooooo 11 >>>-«o uaoooh 1— i-u i— 11 11 >- a:<5-«—>r —-• <.CQ II -» II ct-» 11

3 ^?x 2:2:2:a— hi— 1—

O0*

xz-»s«—X— h-

# •«-f\jX

»• +*

O -«oO 00<NI rO 4-sf OOOOOOOO

51

Page 58: A simulation study of models for combinations of random loads.

oo

a IA —•» o*

-^ •

o a* ao c*-• 2*—« •» *^ CO >*\OO o * »o>

2 * Gr •-*. «k * arO Lf\ *

> -»o oQ X-M £.

O (-— X•i xx on i—m

CO •>i— on f- • aoX xi— or z CJ • <^> h- •o » U. O ^> i_-» •. ^ O -JQ •om— <\i w

00 XO^OZ o.•l h-^On •• ujon i <^ -• <«. <•

r LU«-h- »— -JfNJ o in ia m ot- x »oor\ •-• »on • cm r» o> o^•k _|l— inin—4 i-or- r-l t • t •

CO Oujr-^^- 2 • *- O O O OX O OX'^ < or O 1 1 1 1

m 2— -H-Z o O O O o oa:— »o^x * a —»_J • • • •

-^—

*

a.>» JCOfNI •-! •^< i-4 i-H ^-4 -»

mm i— •» QOOh (\J —» _)-«• •* «^ <^ «^• t ••in cCint— u_~5 • _J —-» -»-»-,-»-^ o o o o

O"^ f-ti—t^Hr-4i-< O O O Ooo CQ • ijj 2 * •* • • #ON —

a

** cxo _j »LTMnO> '—

1

* • .-»o 2— U.LLU.U.U, —; —1 —J —

t

## Q.-4 •—

<

X(\|(NIO » 3-* —2 •*x ##«*# < < < <CO CO -U. i— 1 OCJ-J n» h-—

»

-)h- -»mLnm co o-*•*-»#-»«'—« -•XX -»r*- 2 »X »HO •• <I-Z C7* <—h- _KMr«^cM7«cMn-»Ln-»iA-»co-»oT>>:> »—• < xh-~» *o m • a<aa 1 XI O • • • • •f\J

-jr*-->0^-5C7^~7O>_»w —(\j O — •.LOO^Z —4 0^ -IO- 3—11- II OOOOOOZ-2—2w^w2»•« x • Or 3-»--40> * X • 0> LL-IX O II II

-» z-"-—'*-— *x -X »X -X »LHlTi — •4- a.o*-0'3 t-tO'vO • *a.-' a£_j -*_iLU—X XX X X ->f- -5 1— ->l- ->»— "3

^^ Q\Um LU 2IOO* »Q-* « «c?o^^h-«- u_ 2 _J«M 13 t— 1—1 1—• •—1 l-H 1—1M |— —• |— ~-1— «-»h-—CDSsO^OcOO » X 0--4t— coa:r\j'»x Of • —%—

«

»>t—rjr2<U-U_U.li-LLX 1 X 1 X 1 X 1 XXh • • xsy I— o*— ».02^ » —40*-« n ^n(3t-l-lO>-W«MMh. || - || |— || J_ || |—II II <-4r-4 -X »'-H X~»^J x> || || w-» II -4 i— -j ii ii ii ii ii ii on ii m ii in ii oo ii—— 1 Q'O- LU LU ini— 0£>O'-4LU •—t—»c^>-^UJ-) 2u.iAinincoCMnr\iLnr,-Lnc>coo>o>i—

(

<—i CO LO 00 —* h" t— Z<T<-t ••LU-—-_i ao -•II-'3 o O 11 r\ir*-aN 0>C7, r\lor*-O0N Q'0,*CJC>—XX II UU<LU O «-h^ wco'jjuj^i o>—t^»2cyzma o-'Zzzzzai-en-ai-oHOXX II II >h- *--3 o. h- -jcjonujt—«ixm •oaw^—

<

o u.h^_l30«QCZZ 3C 3< ••an— >-<z:- »c_»f-« H ^^

hhoooo:omc< O OLU-* zaro: o- 23:u_h-0 o ooa:Lnin>-«;sox- § OOZf-Q CO i-IO^ LL"-1 •< UQ .

OUJZ 3 «-CJ Xoacui 1

1

1

1

00 CLl— >s.lA»s • •

c*o1

1 «N —i vt

oo •H O or-^ ^ i-Hooo

52

Page 59: A simulation study of models for combinations of random loads.

mor I

s- m

3 -»

-o a3-« in

— 33 h-

3 3•> I—

m —

oI

oI

O f\J I— I—• «— • «.

-H 3 -*C0

O -» ro—

3

O r-4 »Oh_J -» — cot—< -> -«. 03CML **T\-*«• -»w-» a>mao^ -a^—o^m—•

-5XCT>maOO— O0>3-»3>-*'^3 •

-5-1 •VCM^OOOh'J * »—"t— co

— XLU333 31— I—I— II — C0<\I3 »u.X II II II II II II II II II — XO> »H— ">tI— -*-»-»-»—-»-.-»—.0-»3m-»0>.-it— ^-trgro^LnNOfN~ooa^'-ico ••u.^«-» •

I-»»••»•.» »<om -03X

II -i -j ->-»-»-}-> -j -j-j -OX •»>— sO0„—%<—— oO^vO ••«* (— o^-"— "Qi »<-<> O •"-iCJ -»<1hh^>Lu ||—»Z>QOXXXXXXXXXX— ——^h- X00>2>0-LU O^ULO^—I II —Z II r\j-»_»z-i

3^1— I— I— I— r— I— r— I— h-LU—•(— LUC0< LU LUh-h- CT* — I— -J COZ II '-iZO^'ZOw-OJ-mU.t— xxxxxxxxxxh-c^<iH--»2:uj t— z < occ lu<i i o— -> ii it —ia.arzi— ~j—»—3k- m

QC3QCixi—ozz cl >— #-» ••o^cc." iu<ifiaao uxa«uu.^ »o:£ fcU-t-ioc s: :D_jmooi-ua:o -& oozzq oxq o— LL. in h-3 O 0<'-<CM— JZU.X* _J<-lao r» zt—Q o oclu-oz ih • u.— 3 ou-12: cQa£3f— •-« »-x ii a3 (— OC^LU 13 << >vin» LL.O

# # •«• # # # *

m i <n ^ (\i

03 OO r-l O O

— ("0

3 —•• 3

in >—

CM •i

3 ma*•» (NIO— arai

1

I

1

I

i

1

1

1

i

I

I

I

1

I

a. oz

-» <a »oao ao—t-tO O

1 -3 ZXrO< <Xr* ••

a. m mx<<7» a*I— »3 Z00— h- <coo •• m

• ••«• m

3X

a «o3 •

LU 3-1

h-

ZX oi^-im f*- <(T>H- Q.wO>-«Z

<lX'JO<D «co33 •

LL •t- <X^M * 3O xi— » »m

k »inorsi-I< ,_^

LU H-»Mnz Z —

»

-J •>03—< Om -J•-t XOHH » >-i(\i w1— 1—^<X-» i— »m a.Z uu— »xm mON << »xin »<-« a • -^

i

Z> —it— r»-—

w

2 3 -* o3 0LU30 3 a -> •

o »<oz o <a« r-t

_

I

Z-» "^< •^•^ CO •»•

< •oin^» •• r\j <NJZ 0£ oZ 3IOMXJ * • •^Q.3 Q. oa QhOX * #3f- •^ «z< < -»_J>—» a;— h- »™J • LL •«-»•—.

1 —

<

i- 2)«H * 3— * f-r-X O -l#i—

«

•• . fcLUi—

1

n» ••• <<-. • —»—

»

Q x-»in •• •» •> * -J OOU. i—

t

3-»Z h-OfMLLJ X - — -I-J'-' *- z-> -*

a -03 -O ^Hr-« f> CO OLLLL. + *v <I— roo xo^u •> t 0^ QC-t^** ^ -»p* »x a

—in »a>z V.3>0 • saa »^-«—»a. -5 *->x za. —»3 —-^ »3 o<z i—i ••—>z —^— |

LU xmini— 3f\j 3 -^.a:— 1 11 O— II X II X— H-X OM-KOHX J7XJ —i II CL— LUI— _lh- II LU<

00 O' OOU

53

Page 60: A simulation study of models for combinations of random loads.

m m m CO 0>og r*- o> & 0^• • t » •

o o o o oio i

o 1o o 1o• t • • •

.—

I

•—

(

1—

1

1—

(

^-*

o o o o CDo o o o a

-^o* —

«

vi-O v*"—3 —

3< a< H-»C0 < o>

-»;j\ •• a*roc? —» 3—

<

m Har «_» <«im 3 »

»0> 1- 00

imr < of\l< •• 33 m »—<-» <N •«»0> 3 •—»<M K- mt\J3 < o>—< ft 33 - —ft 1—<-- r\j <•CO n* •»

—,_»-»3 *»

-*3rOh— o>—< •< V*3 »C0 3<mu.-» 1—••r-^rH <——*3>—•*•

'

•TO-» o~»< "3 —» •

o>mooa>>- »c\j(— coco

U.LLU.U.U.

mmmcoo*rsjh-O^OCT _^j t i i _j• • • • • <t «3 <t <j ^COooo -»&—«.-»•«.-*#. -»* -»

«—w———m-»m-»m-»co-*0"*— ">*

MMMMMZ-)2-5Z-)ZTZ-5-)Xs.CM^(?aOOK^Nin^ 3>tLLLLU.U.UL <»-<-»<—<*—<>»-hOOOO|- 1— I— <0*-»U.V> t——*

.-_.-«_,_ »x »x »x »x •>xxuj<«<<<< ii a^3 *o» < •»

+ + + + +• -sx-jx-^x-jx-jx ii ii ii ii ii ii ii ii n-»o^<xc*~• *xiH-^HHH« I — I — I

~- I ~- I —,—.—»—»—,—»——.—.O •» •—H »s0in>Oo. a. o_ q. q. x «-•x— x—x— x— i-H <m n ) <rmo r*» ao cr .-• vo -- «-"0 «— r>- >—ZZZZZh II h- II h- II I- II H II »•-»---•• »^ok«-OOH-ii ii ii ii ii ii in ii m ii m ii co ii a^-5~>-5->~5~>-)~^-3-5uj—<lui— h-<mmmrocMr\r\|ir\r— mc^aoaMj^cr*—-*——«—>——•——»^-h-c?2:h- <x <xujc\ir-u>tr>cr>(\ii-*i'-(j(^ua>3U^i-#i— h-i— f— I— *— H-»— h-t— «-i«iac.»-t * *<z.:dzzzz^zot-uhOKUh-oh-xxxxxxxxxxo: uq:-»'*cjzq:««<«««««XXXXXXXXXXi-»U..IinNlIHD

m —— t—t—

o

— 33 ZLUZ3 h-H- OQCUJ< << o

1

1

1

1

1

1

1

1 ct» •1

-, ^ -j

1 ino z •—1

1 h»Q <1 O arr-< 00 i—1 Z I——O1 LU — «-J(^ z

a om<^3 » zm or>« »i— m >-^

z x. >-4Ur*<l U" QCo <S) «— »LL » Z <—

<

m xm »co » ah- X hr\l£M3> m T.3 h- 1— CJCM3<\ir«- OCO f- »^3l-<Z OI—

«

k -•ooh— <a£ * N_

DC X O »oo •> »-in

— 1— om »co-*f\i LU00 LU hmo>o>oz —11—

<

* -00*00 * •—

<

Q -J Xh-31— —•-• H-O I— »h-<—m Z

-J U turn » ••i_)<-t <h z. •M0MftO>« r>< •> —30^ *3 3H ^ »oaiiAZ

0-* •+-o,

> »u. or Oinco<3_jo z mi—io> *>>«. *

•» •»—3min-j #LU X in3H-r"-r^ » *-J h- (\ii—uooa«1—

«

» —•00 »t—i— •1- X X 00<o0_| *z ** _-.tr •• »o m< 3 •>m3in— Xz> Q. -»-<|— IAJOZ r-H

3 2! O— -30 » »O 03CO|--^3 "v

LU O ^-»i— ov <r^-o-»"vX a^ — *3 *Orarc\i »H- *X-* ^ZZHX

lus}- mmmuj •>—<

LU 2< -r r-i a^^a a: v0 >—1— -iQi*"—3—

'»UJ--1—

_> 1- jui-u-»(j;uj^a. ;^r\j< »00|— OLUh-S:2: 0<LU-» »<Ol— i-»0C

O a:o:a:mm •—<za:oCO » r-<a^-4-——'^U-30 «-3<^000 o_h-o;<^>

I» >•>««

•»• «•I ***#I

O O^rO I<Ni

0"» 00 —'

000

5M

Page 61: A simulation study of models for combinations of random loads.

• »o> -^ ^m(33 • -5 -» -«• «^ -» -»w

O — — -»

O -> -3

UJ o oQ Z Z

HOQO # LU UJ• »tr -^ -» a a<33 • -•-»-» * #

o _i—_i — -.-»w_n- -»in mma>—q. incsi r-r-

> i a. i . —<n • _ • •

lu oio in t i mo I

_jLnm »o • <m I o I*- i ©mI «0 *-4 »rH »o «in to »oX3 • —..-t—^ q • fHf>- O **-*

3 U>—-O I-tw I ^twO

» OOOH o——o o—•— t—

i

^<JjO hOO r-iooo«» -» _l#<<h- wjjQ -—_l_)0

>» »(js -.O 3-5-»-»0 r-l^r-|^^ 0## — 0##03 • '-•O Z——*-»{J> U.U.U.U.U. _J-*~»0 —»-»-» » -J— -» -J~»-» _»-»-*.—- 3 —Z »X-5->-^ ##### <-*.-* • <^.-»

O

<—. <-^— <—-»3I——* -JH-—-»o -»inininaoO-»# "J-JO -#T5 • --# T"V*« #->->—# "S "5 «»

<h-z c^—»t— :*:^ « _jc\ir*-crocMn-^«-.w • in-^—— i—oin-»——ao—»——o*-»—«-^

# o-?o »>x i t/>ooo o » • • • •cNj-j^Ciit—ONn^^o • o-j^:£o->:*£:^o ->*:*:-5# _lO— 2r-<l— H^vN, »0 0000002T—uOoOO »Z—1/)00 iOZ-o0i/>:z —»oOo02T—»o0o0—•«. moo u__ix o ii n —ooi— z————' 'Xw »o «>x>»>%in n »xv.>^ »xn.>s »xs\x

• • #ll,i-i q:_j-»_j . lOouj-xxxxxihoom n -51—ooNtfni-oo->i-oo")Hoo->330>-t# LL Z _)"—•-»<-• • || Z>h- 1—!•—••—•<—«»—•—— J— • • C\|in«— I— • iCfh-—h • •«-»l— • •<—H • «C0

^ » • -™._ —^ —»3 I I -^-^Z<I0_LLU.lLU_X I r*HO(\JX I —H^-H-»3X I .-I^X I i-ti-tX I »-l^HC0

x -^f)H 11 i-ioari—»-jjNtqi-i«<^ h-i»-it~ 11 1 1 1—31— 11 1 1 ooi— »— (11 1 1— n 1 11— 11 1 1 »1— 11 11 ——* 11 <-*

11 11 «-.— t—-J 11 11 11 11 11 11 in——'tot— " in-"--w 11 in—— 11 ao—— 11 cr<-—

o

-i— clk-ilu-} -»-»33zll in in in coo in r\i 11 h —ooinr- 11 no. ino n 11 coo 11 11 00 11 11 —u. in •—1 11 — z> o _j_ii— h-o 11 r\jr«-ooo(\j3intno. NQiinin*-i a*3imno3coaoo3000JumNH^zrj^iTiQ «-wa)^oHzzzzzaK'M(N>H orh-r^r^. 31— o*tj>ah-cr>o>(JH o^o^t—

(\J lOU-UZ^o 1_*<J— LL, (J<~# <-»<-» U<J C_»(_# UCJ>-<(_ »cj-4 |_—i 1—h— u. i—h- I—I— I— I— I— I— I— f-cc003 Z <</!•-« <00 <i>0 <o0 <</) <002-I- O OOO - O OQX

- U\m

—» —

.

-53-> -> •*» •.

** ^ >-)-•

Z-*z z UJ-UJ UJ 03u # »# # •**.^

-^~» -,, -«, -•Ornin in 00 00 OO—o>o OCT O »3

•• • • •** » t -» »o ••

mo 1 COO 1 OO 1 U"»

o> 1 1 O IOIM

• • •0 • •O «3O »«H •—

1

O ••-1 *1 r-(— 1

-^-«I^—

.

o«— 0*-^ O—•— 0J• O'J) •OLD •JO-nOQ HOQ -H003— -I-J — -I-J >-_i_J »o<< o<< o<<-»a## o# # o##-i

# * #

^ LA

o r-- in-* o

55

Page 62: A simulation study of models for combinations of random loads.

-J- in-^ *^Of Of\— \—oO m» u* ~.QS

-» O 4-0m t— ~a>.— 00 CBh-C* «» h» -

h- ao •aoc/> O —O•» 0/ roo»

0* m H- —I—

V* tM 00 O •a a 1 f-mm t— LH O

'JO 00 0* mc?cr to a (N^Of —

.

1—to <M 00 t— -*

\T\ —

•» •ON0> a> -^ -»—a -»l— (MOT

Wt rOoO -_. ^t—-^ * » 3 —ON0O-» 1— I—*^-— LL*H oo-•» ».CQ »Os*-— *.ro—.—.ro•» fHOT •^ •HQI *

-» •4- aoco*-t—cocorgoo —U.O -u.* »-^» O-4-l—m^t(JLOO h-<-»—»I*-iH•.xo i-O •.^Qf

in •*-4«•« *> XCT*H-Xr--x «»>omo » »--oUH<)»-r*--— sO—

•» «-»—ooh-— r-t——

«

h-UJt— H-<UJ'-<r*- <i \— oo ( o 2: h-m 2: tuw ac>-< • •»!i>->|—CCO(jac(X-»-»i_ia: *ozz*02inNu.i-»u.«a:

—. LL -——

vO »-=>»o OCJ — Zh-

oooin

ao—zo•ooma:—

a 00 1/)

COGO

o

Qa» uujzi- oq:lu

OO

—00 2i-oa »a:mz2'-io •> ~o II

UJwQi »zt<2:-«o—•< 11 >r1— cc0-^"LUi—«0

c£_jluOcq<i— coZ>LU2t00 ct:—

O

a

Qw

o -*

>-> z —

I—» COO— «*o II

< <z1——•*:

11 a: »o u«-«

-»LUOO — LU

~-2T0O <ZCQi—1 O 1—

<

H-_Jir\ 1—

u<oOOQ

u_>zaza:

^i—QOOiULIo

<<

UJX

LUI—

oo

-4-

<c«;

*-

<

oLUXo

< »

•> »

rsjcc

»_<-51-

O »

000

X^-5— *:o<oocr: 3a -z a<— a:

000—1

r-i » 1—

t

UJ-H ||

zx >-<

—1 cc1— >tlu oo#o oO—JLUOr-):<»- 11

COLUZOO

ood 7

1 <— cc

QC

^°iI —a:•~a: *

X + LULU_1Z

-^ +1—1—+-OO—O*-* • • II 3E

II X II II UJOO H-lctZ II

—iCCh- 1—h- O.

o

oOQOOo m-»or-lh-O—O H-OOO

• tooI «o—Q.I-00—_J^ •

00 • •ftlCQi(i|— »LU<OOHZ0£II LU »oi—h-^:xo • 11 11

uu i-<u--ja

XLUU.U.O—' •—••-»

in

o 00inn 000

56

Page 63: A simulation study of models for combinations of random loads.

Q.XLU

+

ZSLUZt—

I

O

m^- o< —cc —

<o » t-

in —i «t(NXLU

I— Lumo^»h--JZ -»•— II UJZ

lj— >— (_ilu^ -» ZJ ex.

O II II h-l-SI^ZZ}-J a: -tct— i-ih-Q(-h-ooro^Hiuz

O-SLLOO^CtLUo

LL

LU

X

3a.z:

ao

« -»» or—

*

cx<•* aC

<t l<—oca.i x

a:*

<I

- OS_5 OLULU -»ZZ OhH • I

-2-IO2TLU • •

ozar^Oh-LU—-

»

O LU •*

I—* >——"-»U-l— II ex

Z<u_XttOLUi-tO-tU_Q£ 0<

Qd

I

H-

ai a.xo x

lu «3:uj+ r-tLU#

#~»ILU

^JOZ<LU •(—<X2-H I

I I

—-o<-|OiO.OONX •-J"-'LU—l<~»

— * #oluqc:qc<Z# * I

h->t^—I <K1—Occcccc• I ##wQ. •<!—'XfNia:*LU I I

in* <—«-•OQ-Q.o »xx

wOOLULUI I+ +«•»•««

cC#(M

aCI

a.

X

cc

#

<cC

a.xLU

<cc+

<a:

mfM

CXZa—cc

II h-U03IUJZOOCXLUo

I oI o

XH-

Z3coi—

<

cct—00t—

t

Q

<t—CO

ooLUII-

LUK»

a

i->f-<

< »

-<r^tcC

< »a:—•oooZ-4LUwQOZ3LU 3OC3 »a: « -4-

z~» <-O CC^O Q3oOf-t »o•~—-»coa:X^OtQZ—oOO * "iI— *r-l<00

•-•OX « »ZLUr-it— »LU—

I

Z—<<<-» IIMX »</) •-«

(— Ct QC-J^txaJLuci(_># cc# (jog

aC—i ••-JLU'-^

cd<qc:<h--jluqcujzo

cc<-»

LU00

aaujQ_iII CD

"1 II

cc<.Quo

*

QOQ•a •

I <N—^. I I

QQ-CQ•XCQ

(NJIUV.

•^«.QrgO •

* twcOrO'-CQ + *

W-.QQI ^*—co>to*<•<Q

cm a:*

^l a

CM-* t

*O-r0#X I

<LU—<oa— I x-*«—Olu<-QQ— 00* « |

CD | -%,H-»* o—'—>tO<fttc0< *#—CQq:<m*<-'Q--.000

i«•<

i a>--»00O—

I

Q.OJI QU

X#Q • I

LU-4-Q—4-»

-»+ q:-i^-<

^.00 I —.'J 00X I—N.CUO II

<—oa-o#_J —LUQXO^Q *-i

_J »LU •<— *—

l_) || || —CJ-» ZII <C0 || I CQ-»LUCQ<CDi-"«' —.Q

CL>}-—X<<LU0i<oas**«

x

Xh-

OO

X_Dh-ZI—M<h-Z

o

om

O-tOOQOrMOO

rOLn

<j>u><^> uuu

57

Page 64: A simulation study of models for combinations of random loads.

I

X

w JC "5>»LLI —->o^o\

—»^» ^^Qi'N.QiN,.3C>H O *Z »2 •o aim »r0 »r0CC ^ • _i *,_i ,

z r*- ii r- ii r*-

» » i-4LL <-iLL'—tLL^-» ii in «in »LnII X i_i-».-i-»,-i

•» »nl | 1M H »

x X—O>—

t

w t-f-z:-,— * ^—V-^-LUXx» OOX<LUQ—

m>< vu- oo» r*-»(NI—< (Nl ••<%) «*\j »

\^J »* •• 4J •—' 4J —» -4J —

«

LU< HJJ<m<LU<o:

•—iQ^LL,'—"QC^-hQI—(QCh-Q

3U>-HJtUL3LL3EU.a:UJ

Lf\

(Nl (Nl (Nl (Nl

58

Page 65: A simulation study of models for combinations of random loads.

LIST OF EEFERENCES

1. J.C. Peir and C.A. Cornell; Spatial and TemporalVariatilitv of live Loads, Jour. Structural Div. ASCE99 ST5 (1973) pp 903-922.

2. Y.K- »Jen; Statistical Combination of Extreme Loads,Jour Structural Div, ASCE, 103, ST5, (1977) pp1^7-9-TU?3T

3. H.T. Pearce and Y.K. Wen; A Method for The Combinationof Stochastic Time Varying load's Effect's STructUraTR"eseafcE S~er ies No 5(77, DTLT=^HT3=B3- 2 1 , Ci vi 1Engineering Studies, University of Illinois,Urruna, Illinois, June 1983.

H. D.P. Gaver and P. A. Jacobs On Combinations of RandomLoads, SIAM Jour Applied Hath vol 40, No. 3, June 1981.

5. P. A. Jacobs Eirst-Passage Times for Combination ofRandom Loads, Ib preparation.

6. W. "Feller An Introduction to Probability Theory andIts ^plications, volume 2, second

-ecTTtTon, J~o~En tfilev

anl 5~cn~Tnc, TTew York 1971.

i. TJ . Eudin Principles of Mathematical Analysis , Thirdedition, McGraw-Hill, New York", T9"7F.

8. M.S. Pazaraa and CM. Shetty Nonlinear Programming,John Wiley and Son Inc,New York T9"7"9~.

9. F.J. Conover Practical NonDarametric Statistics, JohnT7iley and Son Inc, New York"~T98~0".

59

Page 66: A simulation study of models for combinations of random loads.

INITIAL DISTRIBUTION LIST

No. Copies

1. Library, Code 0142 2Naval Postgraduate SchoolMonterey, Califonia,93943

2. Professor P. A. Jacobs, Code 55JC 2Department of Operation ResearchNaval Postgraduate SchoolMonterey, California, 93943

3. MA J Noh, Jang Kat 3Department or Operation ResearchCenter of Air ForceSeoul Korea

4

.

Defense Technical Information Center 2

Cameron StationAlexandria, Virginia 22314

60

Page 67: A simulation study of models for combinations of random loads.
Page 68: A simulation study of models for combinations of random loads.

1 3 3

Page 69: A simulation study of models for combinations of random loads.
Page 70: A simulation study of models for combinations of random loads.
Page 71: A simulation study of models for combinations of random loads.

101*75

Thesis

c.l

Thesis

c.l

NohA simulation study of

models for combinationsof random loads.

21d*75

NohA simulation study of

models for combinationsof random loads.

Page 72: A simulation study of models for combinations of random loads.