Dissertation (PhD)--University of Stellenbosch, 2007. / ENGLISH ABSTRACT: To keep pace with international trends, the introduction of geotechnical limit state design in
South Africa is inevitable. To pave the way for implementation of limit state pile design in
the country, the study quantifies model uncertainty in the classic static pile design formula
under the Southern African geologic environment. The generated model uncertainty
statistics are used to calibrate partial resistance factors in a reliability-based design
framework.
A series of pile performance predictions by the static formula are compared with measured
performances. To capture the distinct soil types for the geologic region of Southern Africa
as well as the local pile design and construction experience base, pile load tests and
associated geotechnical data from the Southern African geologic environment are used. The
methodology of collecting, compiling, and analyzing the pile load tests to derive the
measured ultimate pile capacities is described. To facilitate the computation of the
theoretical capacities, the site specific geotechnical data in the database are transformed to
the desired engineering soil properties through well established empirical correlations.
For a given pile test case, model uncertainty is presented in terms of a model factor
computed as the ratio of the measured to the theoretical capacity, leading to n realisations of
the model factor. To facilitate further interpretation and generalisation of the model factor
realisation data, statistical analysis is carried out. The statistical analysis comprises of
graphical representation by histograms, outliers detection and correction of erroneous
values, and using the corrected data to compute the sample moments (mean, standard
deviations, skewness and kurtosis) needed in reliability analysis. The analyses demonstrate
that driven piles depict higher variability compared to bored piles irrespective of materials
type. Furthermore, for a given pile installation method (driven or bored) the variability in
non-cohesive materials is higher than that in cohesive materials.
In addition to the above statistics, reliability analysis requires the theoretical probability
distribution for the random variable under consideration. Accordingly it is demonstrated that
the lognormal distribution is the most appropriate theoretical model for the model factor.
Another key basis for reliability theory is the notion of randomness of the basic variables.
To verify that the variation in the model factor is not explainable by deterministic variations
in the database, an investigation of correlation of the model factor with underlying pile
design parameters is carried out. It is shown that such correlation is generally weak. Correlation can have a significant impact on the calculated reliability index if not accounted
for. Accordingly, the effects of the exhibited correlation is investigated through an approach
based on regression theory in which systematic effects of design parameters are taken into
account (generalised model factor). The model factor statistics from the conventional
approach and those from the generalised model factor approach are used to determine
reliability indexes implied by the current design practice. It is demonstrated that no
significant improvement in values of the reliability indexes is gained by taking into account
the effects of the weak correlation.
The model factor statistics derived on the basis of the standard model factor approach are
used to calibrate resistance factors. Four first order reliability methods are employed for the
calibration of resistance factors. These include; the Mean Value First-Order Second
Moment approach, an Approximate Mean Value First-Order Second Moment approach, the
Advanced First-Order Second Moment approach using Excel spreadsheet, and the
Advanced First-Order Second Moment approach (design point method). The resistance
factors from the various calibration methods are presented for the target reliability index
values of 2.0, 2.5, and 3.0. The analyses of the results demonstrate that for a given target
reliability index, the resistance factors from the different methods are comparable.
Furthermore, it is shown that for a given material type, the resistance factors are quite close
irrespective of the pile installation method, suggesting differentiation of partial factors in
terms of materials types only. Finally, resistance factors for use in probabilistic limit state
pile design in South Africa are recommended. / AFRIKAANSE OPSOMMING: Ten einde in pas te bly met internasionale neigings, is dit onafwendbaar dat geotegniese
limietstaat-ontwerp in Suid Afrika ingevoer word. Ter voorbereiding vir die plaaslike
toepassing van limietstaatontwerp op heipale, kwantifiseer hierdie ondersoek onsekerheid
rondom die model vir klassieke statiese heipaalontwerpformules in die Suid Afrikaanse
geologiese omgewing. Die statistiek van modelonsekerheid wat gegenereer is, word
gebruik om parsiële weerstandsfaktore in ’n betoubaarheid-gebasseerde ontwerpraamwerk
te kalibreer.
’n Reeks voorspellings van die gedrag van heipale volgens die statiese formules word
vergelyk met die gemete gedrag. Om die kenmerkende grond-tipes in die geologiese gebied
van Suidelike Afrika sowel as die plaaslike ondervinding met heipaalontwerp en -
konstruksie vas te lê, word heipaaltoetse en die gassosieerde geotegniese data vanuit hierdie
geologiese omgewing gebruik. Die metodiek vir die versameling, saamstelling en analise
van heipaaltoetse om uiterste kapasiteite daarvan te bepaal, word beskryf. Terreinspesifieke
geotegniese data in die databasis word getransformeer na die vereisde ingenieurseienskappe
volgens gevestigde empiriese korrelasies.
Vir ’n gegewe heipaaltoets word modelonsekerheid weergegee in terme van ’n modelfaktor
wat bereken word as die verhouding van die gemete tot die teoretiese kapasiteit waaruit n
uitkomstes van die modelfaktor dus gegenereer word. Om verdere interpretasie en
veralgemening van die modelfaktordata te vergemaklik, word ’n statistiese analise daarop
uitgevoer. Die statistiese analise bestaan uit grafiese voorstellings deur middel van
histogramme, uitkenning van uitskieters en verbetering van foutiewe waardes, waarna die
statistiese momente (gemiddeld, standaardafwyking, skeefheid en kurtose) vir gebruik in
betroubaarheidsanalise bereken word. Volgens die analises toon ingedrewe heipale ’n
groter veranderlikheid as geboorde pale, ongeag die grondtipe. Verder is die
veranderlikheid van heipale in kohesielose materiale hoër as in kohesiewe materiale, ongeag
die installasiemetode (ingedrewe of geboor).
Bykomend tot bogemelde statistiek, vereis betroubaarheidsanalise die teoretiese
waarskynlikheidsdistribusie van die ewekansige veranderlike onder beskouing.
Ooreenkomstig word illustreer dat die log-normale verspreiding die mees toepaslike
verspreiding vir die modelfaktor is. ’n Verdere sleutelvereiste vir betroubaarheidsteorie is die mate van ewekansigheid van die basiese veranderlikes. Om te bepaal of die variasie in
die modelfaktor nie deur deterministiese veranderlikes in die databasis verduidelik kan word
nie, word ’n ondersoek na die korrelasie van die modelfaktor met onderliggende heipaalontwerpfaktore
uitgevoer. Sodanige korrelasie is in die algemeen as laag bevind.
Korrelasie kan ’n belangrike invloed op die berekende betroubaarheidsindeks hê indien dit
nie in ag geneem word nie. Dienooreenkomstig word die effek van die getoonde korrelasie
ondersoek met behulp van die metode van regressie-analise waarin sistematiese effekte van
ontwerpparameters in berekening gebring word (veralgemeende modelfaktor). Die
modelfaktorstatistiek wat volg uit die konvensionele benadering en dié van die
veralgemeende benadering word gebruik om betroubaarheidsindekse te bepaal wat deur die
bestaande ontwerppraktyk geïmpliseer word. Die bevinding is dat daar nie ’n
noemenswaardige verbetering in die waardes van die betroubaarheidsindekse is wanneer die
effek van die swak korrelasie in berekening gebring word nie.
Die statistiek van die modelfaktor wat afgelei is volgens die standaardbenadering word
gebruik om die weerstandsfaktore te kalibreer. Vier eerste-orde betroubaarheidsmetodes
word gebruik om die weerstandsfaktore te kalibreer, naamlik die Gemiddelde Waarde
Eerste-Orde Tweede Moment benadering, die Benaderde Gemiddelde Waarde Eerste-Orde
Tweede Moment benadering, die Gevorderde Eerste-Orde Tweede Moment benadering
waarin ’n Excel sigblad gebruik word en die Gevorderde Eerste-Orde Tweede Moment
benadering (die ontwerppuntmetode). Die weerstandsfaktore vanaf die verskillende
kalibrasiemetodes word weergegee vir waardes van 2.0, 2.5 en 3.0 van die teikenbetroubaarheidsindeks.
’n Ontleding van die resultate toon dat vir ’n gegewe teiken
betroubaarheidsindeks die weerstandsfaktore vanaf die verskillende metodes vergelykbaar
is. Verder word getoon dat vir ’n gegewe grondsoort, die weerstandsfaktore vir verskillende
metodes van installasie van die heipaal nie veel verskil nie. Dit wil dus voorkom asof
parsiële faktore in terme van die grondsoort uitgedruk kan word. Ten slotte word
weerstandsfaktore vir gebruik in plastiese limietstaatontwerp van heipale in Suid Afrika
aanbeveel.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/21612 |
| Date | 12 1900 |
| Creators | Dithinde, Mahongo |
| Contributors | De Wet, M., Retief, J. V., Phoon, K.K., Stellenbosch University. Faculty of Engineering. Dept. of Civil Engineering. |
| Publisher | Stellenbosch : Stellenbosch University |
| Source Sets | South African National ETD Portal |
| Language | en_ZA |
| Detected Language | Unknown |
| Type | Thesis |
| Format | xii, 314 leaves : ill. |
| Rights | Stellenbosch University |
Page generated in 0.0021 seconds