Brief summary of engineering doctoral thesis: Analysis of influencing parameters and basics of determining resistance factors of drilled shafts used in bridge substructures in Ho Chi Minh city

Nội dung Text: Brief summary of engineering doctoral thesis: Analysis of influencing parameters and basics of determining resistance factors of drilled shafts used in bridge substructures in Ho Chi Minh city 1 MINISTER OF EDUCATION AND TRAINING UNIVERSITY OF TRANSPORT AND COMMUNICATIONS NGO CHAU PHUONG ANALYSIS OF INFLUENCING PARAMETERS AND BASICS OF DETERMINING RESISTANCE FACTORS OF DRILLED SHAFTS USED IN BRIDGE SUBSTRUCTURES IN HOCHI MINH CITY MAJOR: BRIDGE AND TUNNEL ENGINEERING CODE: 62.58.02.05.03 BRIEF SUMMARY OF ENGINEERING DOCTORAL THESIS Hanoi20142 This thesis completed at: Faculty of Civil Engineering University of Transport and Communications SUPERVISORS: 1. Assoc.Prof. Dr. Tran Duc Nhiem 2. Assoc.Prof.Dr. Nguyen Ngoc Long Reviewer 1: Prof. Dr. Nguyen Nhu Khai, National University of Civil Engineering Vietnam. Reviewer 2: Prof. Dr. Nguyen Dong Anh, Institute of Mechanics Vietnam. Reviewer 3: Dr. Do Huu Thang, Institute of Transport Science and Technology Vietnam. Thesis is defended in front of the UniversityGraded Committee of thesis evaluation according to Decision 1359QĐĐHGTVT, on date 17th June 2014 signed by the Rector of University of Transport and Communications on date………………….. 2014. Readers can find this thesis at: Vietnam National Library Library of the University of Transport and Communications

MINISTER OF EDUCATION AND TRAINING

UNIVERSITY OF TRANSPORT AND COMMUNICATIONS

NGO CHAU PHUONG

ANALYSIS OF INFLUENCING PARAMETERS AND BASICS

OF DETERMINING RESISTANCE FACTORS OF DRILLED SHAFTS USED IN BRIDGE SUBSTRUCTURES IN HO-CHI-

This thesis completed at: Faculty of Civil Engineering

SUPERVISORS:

1 Assoc.Prof Dr Tran Duc Nhiem

2 Assoc.Prof.Dr Nguyen Ngoc Long

National University of Civil Engineering Vietnam

Institute of Mechanics Vietnam

Reviewer 3: Dr Do Huu Thang,

Institute of Transport Science and Technology Vietnam Thesis is defended in front of the University-Graded Committee of

Communications on date……… 2014

Readers can find this thesis at:

- Vietnam National Library

- Library of the University of Transport and Communications

INTRODUCTION

By applying statistics and probability and reliability theory to engineering foundations, the thesis proposes a pattern to determine resistance factors of drilled shafts used in bridge substructures based on statistics characteristics of the ratio between real measured values and

estimated values for Resistance (R) and load effect (Q) Then, by analyzing

statistics characteristics of the capacity based on 24 results of static axial compressive load test of drilled shafts that were constructed by the wet method (bentonite) in cohesive and non-cohesive composite soils in Ho-Chi-Minh city, the thesis captures the determination of resistance factors for four different calculation methods of pile resistance design based on soil base strength condition

Literature Review:

Drilled shafts construction technology was first used in America (1890),

in over the word (1950) and in Vietnam (1990) However, the calculation theory has been developed more slowly One of the trends in the world is to research new problems of applying statistics and probability theories and reliability theory to correct resistance factors based on statistics characteristics of the ratio between real measured values and estimated

values for Resistance (R) and load effect (Q) extracted from a reasonable

number of construction projects Research results are gradually applied to update and to implement for some clauses in standards or specifications and design instructions of developed countries such as in Europe, Japan…and America

Over two more recent decades in Vietnam, accompanied with the development of infrastructure in great scale built in soft soil bases or in urban areas, drilled shafts foundations becomes one of the best solutions The foundations are also widely applied in Ho-Chi-Minh city However, there is not any backgrounds to determine resistance factors of drilled shafts based on analyzing statistics characteristics and reliability analysis using current advanced theories

Therefore, studying of the backgrounds to determine the resistance factors based on reliability analysis is new and attracts international and domestic researchers It is the reason why I selected this topic to research

Thesis title: “Analysis of influencing parameters and bases to determine

resistance factors of drilled shafts used in bridge substructures in Minh City”

Ho-Chi-Objectives: Determine resistance factors according to soil base strength

which is equivalent to methods presented in specifications being applied

Structure: Drilled shafts used in bridge substructures

Research Scope: Predictive resistance and real resistance obtained from

the results of static axial compressive load tests for drilled shafts in Chi-Minh City casted in cohesive and non-conhesive composite soils (sand, sandy, clay, clay mud, ) by wet method; to determine general resistance factors according to soil base strength condition for four different methods

Ho-of pile resistance design: 1) Russian Method specified in TCXDVN 98; 2) Japanese Method (JRA 2002 SHB -Part IV); 3) Reese&O'Neill (1988) and 4) O'Neill&Reese (1999) Problems relating to load statistics characteristics, general resistance factors for various types of soil, local areas, and type of structure as well as pile shalf and tip resistance factors are not performed in this thesis and they are recommented for future studies

205-Scientific and practical meaning of the topic:

- Apply advanced theories of statistics analysis and reliability to propose a pattern to determine resistance factors of drilled shafts based on statistics data of the ratio between real measured values and estimated

values for Resistance (R) and load effect (Q)

- The thesis has analyzed and determined statistics characteristics of the ratio between real measured values and estimated values; to determine resistance coefficients for the four methods from 24 static load test results

of drilled shafts constructed in cohesive and discrete composite soils in

Ho-Chi-Minh city subjected to static axial compression and other applicable

data

- Research results of the thesis can be used as reference documents in design bearing capacity of drilled shafts used for bridge substructures constructed in Ho-Chi-Minh city or similarly geological areas

Chapter 1 GENERAL 1.1 Drilled shafts and its application in infrastructure construction

1.1.1 Definitions, structural characteristics and technology

Drilled shafts of bridge substructures (Drilled Shafts) : are a part of piers and abutments; they are constructed by raw concrete casting in pre-bored holes with or without steel case inside The piles are subjected to loads transferred from foundation foots and then transfers the loads into surrounding soil base

Wet drilled shafts construction method (wet method): to drill holes and

to cast the piles in water or in bore mud and a temporary tube wall segment

is put in the boring hole top Applicable for cohesive, discrete and high groundwater level areas

Drilled shafts cross-section maybe cylindrically constant throughout the pile length, this pile type is called simple one; or cylindrical-shaped but widened at bore hole tip area

1.1.2 The utilization of the pile in Vietnam and in the world

Through analysis, the author recognizes the need to use the drilled shafts is growing both in Vietnam and in the world Almost foundation solutions for traffic, civil and industries from medium to large scale in Vietnam are using drilled shafts foundations

1.1.3 Current status and characteristics of drilled shafts used in HCM City Through analysis, the drilled shafts foundation for constructions here is also applied a lot in recent years Most of the piles are constructed by the wet method (in bentonite) through the mixture soil layers with combined cohesive and discrete soils, these layers can be weak, average or good in load bearing 1.1.4 Some structural characteristics, drilled shafts technology in Vietnam Due to the characteristics of the technology, the complexity of geology; experience level of the participants in the management, design and construction limits and especially the system of processes, standards are still in the process of integration and are not complete and existing many problems Therefore, the quality of the drilled shafts or pile resistance depends very much

on the aforementioned elements

1.2 Design drilled shafts based on reliability according to Load and Resistance Factor Design method (LRFD)

The design method LRFD is based on reliability,as the load effects with their

particular factors (Qtk) shall not exceed the resistance with their particular factors (Rtk)

Through analysis of the historical development of the design philosophies and design standards such as Allowable Stress Design (ASD), Limited States Design or Load Factor Design (LSD; LFD), Reliability- Based Design (RBD) and the method with reliability factors separately or Load and Resistance

Factor Design (LRFD), the author found that design calculations of drilled shafts foundation according LRFD method is an advanced method and have being trusted and applied by many countries in the world

1.3 Analysis of literature to determine drilled shafts resistance factors based on reliability used for bridge substructures in the world

1.4 Analysis of literature showing the LRFD application and determining resistance factors for bridge design in Vietnam

1.5 Current challenging problems

Some current problems in bridge design standards 22TCN272-05 and AASHTO LRFD 2012 (2007) are shown in Table 1.1

Table 1.1 List of current problems in the standards 22TCN272-05 and

AASHTO LRFD 2012 (2007)

Method for determination

of resistance in cohesive and

discrete soil

05 methods existed from before 1988

01 method O'Neill&Reese (1999) Resistance factors are not

specified for:

Sandy soil, cohesive and discrete soil Cohesive and discrete soil Officially Applied Year 2005 2007

Determination of ultimate

resistance under static load

Many method based

on

TCXDVN269-2002

5% pile diameter or merged settlement pile Recomendations when

resistance factors used

The resistance factors are not the standard values for all states of America and of course not accurate for other countries, including Vietnam Some shortcomings of the related scientific studies:

- The study of resistance factors correction of deep foundation of Paikowsky et al (2004): Didn’t mention the resistance factors of method O'Neill & Reese (1999), just mention method Reese & O'Neill (1988) for sandy and clay mixture soil conditions on the basis of 44 load test results of drilled shafts in Florida

- Liang (2009): Proposed resistance factors for method O'Neill &

Reese (1999), but only suggested for sandy and clay conditions in the U.S

- Murad et al (2013): Proposed resistance factors for method O'Neill

& Reese (1999) for mixed cohesive and discrete soil condition in Louisiana

& Mississippi on the basis of 34 pile load test results, but there were 26 values extrapolated to static load test results due to not try to break the pile

- There is no study regarding the research objectives of this thesis in Vietnam

From the above-mentioned problems, the author proposes the targets, content and research methodology of the thesis as decribed in items 1.6 and 1.7

1.6 Targets of the topic

Quantitative study of factors affecting the estimated resistance results of the four methods compared with actual field resistance of drilled shafts under the ground conditions in the area of HCMC This means that the author has determined the statistics characteristics of the ratio of the real measured resistance and the expected one (resistance bias factor, λ R);

To research the basis of determining the resistance factors and to propose the resistance factors for drilled shafts foundations of bridge substructures in HCMC area for the four methods

1.7 Content and Research Methodology

To research the basis of determining the resistance factors for drilled shafts using probability and statistics theory and advanced reliability theory Specifically, the survey collected from 24 results of static pile load tests in HCM City, the author conducted a study to identify typical statistics

of the ratio of the measured and estimated resistances (Resistance bias factor, λ R); From that way, the authod determined the resistance factors for

the four methods on the basis of reliability analysis

DRILLED SHAFTS BASED ON RELIABILITY THEORY

According to AASHTO LRFD, drilled shalfs axial resistance factors according to soil base strength condition are factors determined based on the statistical characteristics of the nominal resistance, mainly calculated from the variability of characteristic parameters of the ground around the pile, the pile size, level of expertise (professional) of human - device participating in the implementation phase of the project and the uncertainty

of prediction method for nominal resistance; but also related to the statistical characteristics of load effects through the identification process

2.1 Method to analyze the statistical characteristics

2.1.1 Determination of minimum size of samples

Sample size is estimated by: (2.1)

In which: σ and zα/2 , z θ: common standard deviation and standard deviation with error probabilities α, θ from the normal distribution; ɛ: allowable

error; C: is a constant related to error probability Type I and Type II

2 /2

For example, to determine the sample size for the thesis: With some prediction methods of drilled shafts resistance that accept averaged

estimated error of about 50% (=1/FS, FS=2: safety factor) with reliable

interval of 0,95 (i.e., α=0,05) and θ = 0,2 Previous studies indicate

standard deviations of the resistance bias factor from 0,27 to 0,74 Thus, the

effect factor is: ES = 0,5/0,74 = 0,456 and C=7,85 By applying the formula

(2.1) to estimate the required sample size for the study:

To compare with recommendation of Murad (2013), the number of test piles for the study area at least is ≥ 20 piles Thus, with 24 results of static axial compressive load tests for drilled shafts in Ho Chi Minh City area can

be considered reliable enough for analysis in order to meet the research objectives of the thesis

2.1.2 Testing method of suitable probability distribution for the random

bias factor

Through analysis, the Shapiro-Wilk method or the Pearson chi-square (when the sample size is less than 50) is recommended with the following principles: the empirical distribution consists with assumed theoretical

distribution (standard or logarithmic, ) when the match probability (P) is

greater than 0.05

2.1.3 Correction method for statistical characteristics of random bias factor

For foundation structures, the laws of probability distributions of random bias factor often match or nearly match the normal standard distribution or standard logarithm

Through research, the authod

proposes two correction methods of

statistical characteristics for

logarithmic distribution form

according to the the principle (Allen,

2005): Based on the graph of the

cumulative probability function to

examine the conformity with one of

the two cases, 1) consistent with the

entire collection data (FTAD method

-fit to All data) or 2) only consistent

with the area of small values at

distribution tail (BFTT-Best method

fit to tail) (Figure 2.1)

Figure 2.1 Cumulative probability density function of resistance bias

2.2 Reliability Analysis Method

When analyzing the reliability, the incident probability is the condition that the limited state has been reached The adjustment factors are selected

to ensure that incident probability of each limited state is very small and

acceptable The probability density functions of load effects (Q) and resistance (R) with the assumption of two independent normally distributed

variables (Figure 2.2) Safety range or the safety factor is the difference

between R and Q, the quantitative quantity for the safety is reliability or safety probability, Ps:

In which: Φ(.): normalized distribution functions; β: index of reliability

Index of reliability is determined based on averaged number and standard deviation as follows:

-R Q G

µ µµ

If R and Q follows the normal logarithm distribution, safety range, G, is

determined as follows: (Figure 2.3):

2.3 Methods to determine pile body resistances

The thesis has researched four methods to determine the pile body resistance: Method in accordance with the safety factor of the design philosophy of allowable stress (ASD); first-order secondary moment

method (FOSM); First-order reliability method (FOSM); Monte Carlo method (MCS) After analyzing the advantages and disadvantages of these four method, the author proposes to select Monte Carlo analysis method to determine the resistance factorss

Safety range, G, is applied to determine resistance factors as R and Q

follow the normal logarithm distribution:

L

Q Q

Q Q

2 To select statistical parameters of design load effect (Q) and load

factors: the representive is static load bias factor (λ D) and live load effect

bias factor (λ L) complied with the standard AASHTO LRFD

3 To analyze the statistical characteristics of resistance (R): the

representive is resistance bias factor, λ R, which is the ratio of measured ultimate resistance (Rtd) and predicted nominal resistance (Rdt):

a To determine the measured ultimate resistance Rtd from results of pile

static load tests according to soil base condition, this is the trial load value

at a settlement of 5% of pile diameter or merged settlement pile (AASHTO LRFD 2012, TCVN 9393-2012);

b To predict the nominal resistance (Rdt) based on calculation theory;

c To determine the resistance bias factor, λ R =R td /R dt;

d To analize, calculate the statistical parameters (μ, σ) and to verify the

form of distribution density function (standard, logarithm, ) suitable for λ R;

4 To analyze and to determine the resistance factors of drilled shafts (ϕ) on the basis of analyzing reliability follwing Monte Carlo method with the target reliability index satisfied, β t;

5 To recommend to correct the resistance factors for calculation method

The above procedure is shown in Figure 2.4

Figure 2.4.Analysis model to determine pile resistance factors on the

basis of ensuring the target reliability index

Results obtained in Chapter 2

- Recommend to use relative random resistance bias factor (λ R) with a

minimum sample size of 20 to analyze statistical characteristics When choosing a probability distribution function (cumulative), it is needed to consider between 2 cumulative distribution functions which fit to the entire real values (FTAD) and cumulative distribution function calibrated in accordance with the actual value area at the tail of distribution (BFTT)

- Recommend to use Monte Carlo method to analysis the reliability as

a basis for determining pile resistance factors and to use the first-order reliability method (FORM) for validation

- Propose a procedure and a pattern to determine pile resistance factors

as shown in item 2.4

Chapter 3 ANALYZING THE PARAMETERS INFLUENCING TO RESISTANCE FACTORS OF DRILLED SHAFTS USED IN BRIDGE SUBSTRUCTURES IN HO CHI MINH CITY

 Define the failure condtion of drilled shafts

piles based on soil base (AASHTO LRFD, 5% pile

diameter of merged)

Determine limit state based on soil base for drilled shafts piles (strength, service states)

Strength state function: g(R,Q)=ϕR – (γDQD+γLQL)

 Determine statistical characteristics for 2 random variables (R: resistance, Q: load effect): Representive of R is resistance bias factor, λ R=Rtd/Rdt Representive of Q is load effect bias factor, (λ D, λL)

Determine λ R, is the ratio of measure ultimate

resistance, R td and predicted nominal resistance, R dt deadload and live load effect bias factor ( Apply the statistical characteristics to λ D,

λL) according to AASHTO LRFD

 Analysis and calculate the statistical

characteristics (μ, σ, V) and verify distribution

density function (standard, loga…) suitable for λ R

Determine reliability index, β and

incident probability, Pf

Select target reliability index βt

(refered to AASHTO LRFD: βt=3,0)

Carlo (MCS) method or fisrt-order reliability method

 Evaluate the reliability index

The parameters that influence the results of determining of pile resistance factors described in Figure 3.1

Figure 3.1 Parameters influencing to determinging of resistance factors (φ)

3.1 Uncertainty factors and statistical characteristics of load effect

In Vietnam, there is no research conditions to determine the rules of distribution of load effects, the author proposes to apply the statistical characteristics and other factors regulated by the AASHTO LRFD design as:γL=1,75, λL=1,15, VL = 0,18; γD = 1,25, λD=1,08, VD = 0,13, QD /Q L =3

where: λD and λL are deadload and live load effect bias factor VD and VL are variation coefficients of dead load and live load; the ratio QD /Q L is of dead

load and live load

3.2 Uncertainties affecting to drilled shafts resistance

The uncertainties affecting the predicted pile resistance should be analyzed to determine the resistance factors for methods to ensure required reliability and they are divided into four main groups: 1) The diversity, the unusual geological structure; 2) The error of measurement (measuring, surveying, testing of characteristic parameters of the material, structure or soil base); 3) The model error and 4) Quality of project administration and construction experience (According to Phoon and Kulhawy (1999), Paikowsky (2004))

To describe the general characteristics of these uncertainties, relative random resistance bias factor (λ R) as outlined in Chapter 2 can be used

3.3 Analyzing selection of methods to predict drilled shafts resistance

On the basis of several popular methods of pile resistance prediction in Vietnam and overseas, the author selected four methods according to soil base condition as mentioned in the research scope

Real geological

layer profile Model of (MH) soil base Model MH applied for design CKN Result in (φ)

Target reliability index (βt)

The formula to determine the unit resistance at the pile tip and pile shaft according to the two standards are briefly introduced in Table 3.1 and Table 3.2.

3.4 Selection of method to determine actual measured ultimate resistance of drilled shafts

Table 3.1 Summary of formula to determine nominal unit resistance of drilled

shafts according to 22TCN 272-05 and AASHTO LRFD 2012

Unit shaft

resistance, q s

1 Cohesive soil (clay, soil with clay dust content higher 50%)

v s

0,59 *

q = N p σ  σ ,

with N60 >50

Table 3.2.Summary of formula to determine nominal unit resistance of drilled

shafts according to TCXDVN 205-98 and JRA 2002-Part IV

Russian method in TCXDVN 205-98

(brief SNIP-205)

JRA 2002-Part IV (brief SHBP4-JRA02) Unit shaft

Refered to table A2, for medium tight

sand has grain components: coarse,

fine, dust If tight state used, then q s