3y ago
2.35 MB
70 Pages
Last View : 3d ago
Last Download : 4m ago
Upload by : Nora Drum

REPORT TO JOINT INDUSTRYULSLEA PROJECT SPONSORSLOADING AND CAPACITY CHARACTERISTICS OFPILE FOUNDATIONSCorrelation of Calculation Results with U LS LEAbyZhaohui JinandProfessor Robert BeaMarine Technology &Management GroupDepartment of Civil & Environmental EngineeringlJNIVERSITY OF CALIFORNIA AT BERKELEYJanuary,1998I

TABLE OF CONTENTSPartPage No.1. Introduction . .12. Approach and Description of Analysis Models .43. Lateral Pile Response . 254. Axial Pile Response .455. Summary and Conclusion.60References . 63II

Introduction1.1BackgroundIn the field of Offshore Engineering, foundation is always an essential element forfixed platforms. In the Gulf of Mexico, most fixed platforms are pile-supported. Theresponse of the pile foundations of these platforms to the external loading, both dynamicand static, are a classical subject for offshore engineers. With the extension of offshoreoperations into deeper water and more hostile environments, concerns with implicationsof foundation design on overall platform costs, and the need to incorporate more realisticfoundation response characterizations into the requalification analyses of existingstructures have brought a recent focus on the dynamic response of marine foundations.The generalengineering guidelines for treatment of such problems are still underdevelopment and subject to update. The needs for high quality experimental andanalytical research are highlighted. Pile foundations are the key components for thewhole drilling and production systems, which determine the safety, serviceability,durability and compatibility of such systems. It is desired that the pile foundations aredesigned in a safe and economic manner.An study was performed by the Marine Technology and Management Group atthe University of California at Berkeley in 1997, parallel to the joint industry ULSLEAproject. This study has an emphasis on the dynamic response of a single pile, especiallyfor the ultimate state.The pile's response and capacities are calculated by differentcomputer models. As the output of this effort, the calculation results by these differentmethods are correlated with those obtained by ULSLEA.1

1.2Objectives of this ReportA lot of detailed research has been done with respect to pile response in this Thework performed include: Extending SPASM computer code used in previous study (Jin and Bea,August, 1997) tothe non-linear pile behavior after first yielding occurs,obtaining an estimation of the ultimate capacity of a single pile to lateralloading. Developing an analysis model which is capable of handling boththenonlinear piles and nonlinear-hysteretic soil resistance, with proper supportingcomputer codes, such as Drain3D. With the developed model, performing analysis of the pile's behaviors beyondthe elastic range up to final collapse. Using the analysis model ULSLEA, predict the ultimate capacity of a singlepile by a simplified approach. Correlating the pile capacity predictions obtained by the different analysismodels above to estimate the bias and uncertainties from the predictionmethods.The purpose of this report is to summarize the results of the detailed complex pileresponse analysis and correlate them with the results from the ULSLEA program.Excellent agreements are obtained between the detailed complex analysis tools andULSLEA programs. By comparing the calculation results from different methods, thisreport demonstrated the validity of the simplified ULSLEA. These results provide theinformation on the bias of the ultimate capacities that ULSLEA can develop comparedwith the API guideline and the detailed numerical simulation method such as Drain3Dmodel.2

1.3Organization of the ReportThis report is divided into 5 parts. Chapter 1 is the background and objectives ofthe report. Chapter 2 gives a review of the current research in this field; describes thebasic assumption and approach in this study; details the building-up process of theanalysis models in this study, with an emphasis on the new DRAIN3D numerical model.Chapter 3 is about the lateral response of a single pile, including the first yieldingcapacity, ultimate capacity. The analysis addresses pile-soil systems behaviors understatic loading, cyclic loading and fast loading. The lateral responses were predicted bySPASM, DRAIN3D, ULSLEA3.0 and ULSLEA phase IV. Chapter 4 summarizes theanalysis of the axial response of a single pile. The basic approach is the same as that forlateral response. The behaviors under static, fast and cyclic loading are studied. Thecalculating code is DRAIN3D and UlSLEA3.0. A calculation is also performedaccording to the API RP2A guideline.Chapter 5 is the conclusions andrecommendations.Fig. 1.1 Typical pile supported platform3

Approach and Descriptionof the Analysis Models2.1Current ResearchThe performance (load-deformation) characteristics of pile foundations is aclassical research subject in the design and construction of offshore oil drilling platforms.Realistic modeling of pile foundation is crucial to the validity of the results of static anddynamic structural analyses of offshore platforms. Furthermore, the comprehension ofboth the static and dynamic response of a single pile to the external loading is the comerstone for all the analyses in this field. The state-of-art design technique and theoryconcerning pile foundation are still under development, though there have been extensiveresearch effort on this topic. In the past design and construction practice, too often havefoundation failures been predicted to be the dominant failure mode of platforms, seldomhave the observed failure modes included failure of foundation elements. This factindicates that the traditional methods of predicting the ultimate capacity of pilefoundations are in general conservatively biased.The state-of-practice design criteria of pile foundation have a static or pseudo static approach. This kind of calculation method is widely used. To keep the analysistractable in interpreting of complicated pile-soil interaction phenomenon, this kind ofmethod has several simplifying assumptions: The capacities of offshore piles can be calculated using methods based primarily upontests of relatively short onshore piles that were loaded slowly to failure, i.e. thevalidity of the method are usually correlated to the static loading test;4

Pile capacity reductions due to the degradation of soil resistance by cyclically appliedloading do not need to be considered explicitly; and Pile capacity increases that can occur in clay soils during rapid loading also do notneed to be considered explicitly.This static pile-capacity method has been used to determine the pile foundationconfiguration for the pile foundations of the more than 6,000 offshore platforms that arenow located on the world's continental shelves. These foundations have had a remarkablygood record of reliability. This has proved the validity of the current design criteria. TheAmerican Petroleum Institute (API) has developed such guidelines for evaluation of thecapacity of the pile foundations (API RP2A, 20th edition 1993). These guidelines addressa wide scope of topics such as operating and environmental loading; determination ofstatic capacity; influences on capacity, stiffness; applications of discrete element andcontinuum analytical models; use of in situ and laboratory soil test and prototype pile load tests in soil characterizations; evaluation of load, resistance, and deformationcharacterizations at serviceability and ultimate limit states; and interpretations andapplications of results. These guidelines represent the culmination of a 20-yeardevelopment effort of worldwide research regarding pile foundation performance.However, as stated in the principal assumption of the static-capacity calculationmethod, two important factors, which affects the in-situ performance of pile foundationsare not addressed: the loss of strength and stiffness of the pile-soil system due to cyclicloading, which is obvious in the wave loading during a hurricane; and the increase ofstrength and stiffness due to the high loading rate effects, which is typical in anearthquake. Bea (1984) summarized the trends that have been observed for piles testedfor these two effects. Load rate effects can result in effective increases in pile strengthand stiffness on the order of 20% to 80% or more for loading rates consistent with waveaction. Assuming the trend continues without degradation, the expected increases forearthquake loading rates would be much higher, shall be to the order of 2-3. Cyclic5

loading tends to result in progressive deterioration of pile foundation strength andstiffness.Hysteresis curves generated for piles will tend to exhibit pinching andsoftening for repeated cycles. Tests have shown that the soil support for the pile in the topstratums will suffer a drastic reduction due to cyclic loading from wave force. How toreflect these two effects in the prediction of the real in-situ pile foundation response isstill under investigation. For the sake of conservation, current state-of-the-practice inoffshore engineering tends to recognize cyclic degradation in determining response. Thisis achieved by implicitly incorporating these effects into the static capacity analysis, orby including such negative effects in the safety index in pile foundation design.Meanwhile, the beneficial loading rate effects are not taken into account. It is obviousthat further research is needed in this area to better define the interaction between thesetwo phenomena.Assessing the structural integrity of an offshore platform reqmres balancebetweenconsiderations of capacity and economic. In the case of foundations, thisrequirement translates into the need of better understanding of their performance andmore realistic modeling of their behavior so that the foundations are not designed withunnecessary reserve capacity. This need has led to focus on the study of dynamicresponse of pile foundations. This analysis incorporates the two major factors notaddressed by static method. It also involves other important factors affecting the realdynamic response of the pile foundations. This effort has resulted in numerous valuableinformation in guiding the design and construction.Bea (1984) published a key note paper on the dynamic response of the pilefoundation. It provides a summary of the basic approaches to the investigation of thisproblem. The main concerns in the prediction of the pile foundation behaviors are asfollows:6

Dynamic response depends primarily on external loading patterns and theinherent structure properties; Environmental loadings are dynamic. Loadings on platforms are developedfrom the motionless ocean and earth crust. It is crucial that they are wellunderstood; Non-linearity is a key concern in the analysis: in presence of soil, which arehighly nonlinear, the pile foundation exhibits complicated coupling actionbetween the soil and the steel piles;2.2 High strain rates increase strength and stiffness; Cyclic strains decrease strength and stiffness; Cyclic loading leads to accumulated displacements; Damping developed from pile foundation is important;Uncertainties in the Pile response predictionIn practice, designers of offshore platforms and pile foundations deal withnumerous uncertainties, including imperfect knowledge of the frontier such as: the loadsto which the superstructure is subjected; the behavior of the superstructure under thoseloads; and how the founding soils respond when those loads are transmitted to them viathe foundation piles.In the frontiers mentioned above, Tang(l 988, 1990), Bea(l 983),Folse(l 989),Ruiz( 1984, 1986), Yegian and Hadley( 1979) , Olson( 1984), Kullhawy( 1984 ), Briaud andTucker( 1986) have identified various factors affecting offshore pile capacity prediction.These studies suggested estimation of the calculation bias the uncertainty statisticsproperties associated with these factors. The conclusions concerning the majorcomponent of uncertainties involved in the prediction ofsummarized as follows:7pile capacitycan be

Soil properties uncertainties; Load parameters uncertainties; Prediction model error;Uncertainties in the soil properties are a major contribution to the overall systemuncertainties. There are several very important soil parameters needed to define the p-yand t-z curves for pile foundation analysis, such as undrained shear strength, unit weight,friction angle, and shear modulus. As an inherent character of soil mechanics, theseparameters subject to large variation due to natural inhomogeneous properties of in-situsoils, and distribution during lab or in-situ test, not to mention the system variationderived from various test methods. The major uncertainty sources in soil parameters arelisted as follows: Non-standard sampling or test methodsTheir effects are not completely avoidable on the determination of the soilproperties, even though very high quality test are performed. Spatial variation of soil propertiesThis variation is due to the randomness associated with the natural depositionprocess, which is the inherent variability with the macro geological structure.Consequently, soil properties do vary along the length of a pile and across thesite. Insufficient number of soil samplesThis leads to error in the interpretation of the soil properties based on widelyscattered locations in field, thus affect the averaged soil properties input to theanalytical prediction model. Systematic error of soil propertiesSometimes, despite the availability of a large amount of measured data, theestimationof the soil properties could still be subject to significant error.8

The reason 1s simply that all the measurements made could have beenconsistently too high or too low due to common sample disturbance,calibration error of instrument, or other factors.Load uncertainties could also be very large. An offshore platform subjects toenvironmental load induced by waves, wind and possibly earthquakes, which all havelarge inherent probability of variation. For example, the annual maximum wave heightfluctuates considerably between years. This inherent variability is further magnified bythe uncertain dynamic transfer function relating the wave characteristics to the inducedloading at the pile head. The main loadings transmitted onto a pile head take the form ofaxial load, lateral load and possibly bending and torsion moments. However, the structurebehavior at connection between jackets and piles is extremely difficult to predict. Theforms and values of actual loading and boundary restraint on the pile have a large rangeof variation. Moreover, the patterns of cyclic degradation and high loading rate effects inthe dynamic analysis are subject to insufficient understanding, thus involve largeuncertainties in the pile capacity prediction.Each pile capacity prediction method has some simplifying assumptions. Thisuncertainty is a system error which vary among different prediction methods.Experiments and field tests indicate that even if the soil parameters in the input toprediction model could be accurately determined and if the applied loads are carefullycontrolled, discrepancy would still prevail between the predicted and measure pileresponses. Besides this, the numerical and discretization procedure in the currentprediction model to solve the beam-column equation could also impose additionaluncertainties. Furthermore, most present prediction models are correlated to the load testresults to verify their validity. Thus the discrepancy between in site pile capacity duringoperation and those measured in load test program impose another uncertainties on thepile foundation analysis. The pile capacity measured at a load test does not necessarilyhave the same capacity of a similar pile during a storm. For instance, load tests are9

generally performed within 100 days of pile installation, whereas the maximum loadapplied to a pile during a structure's lifetime may occur years after installation. For mostnormally consolidated clay where soil strength around a pile generally increases withtime after pile installation, the capacity measured during load tests could significantlyunderestimate the actual pile capacity due to this reconsolidation effect. Generally thecapacity measured during the load tests with relatively slower loading rate underestimatethe actual pile capacity. Other factors can also be identified that would cause adiscrepancy between the load test capacity and the actual capacity during operation, suchas soil reconsolidation, pile compressibility, jacking error during load test, etc., thusincrease the uncertainties in the prediction model. As an example, Table 2.1 summarizesthe biases for axial pile capacities.Table 2.1 Bias in axial pile capacity in normally consolidated claysComponentReference ConditionActual ConditionMedian BiasBoringDrill mud, heave compensationSea water, driH collars1.2-1.3SamplingPush large diameterWire-line small diameter1.5-2.0TestingRemolded, reconsolidated,Unconfined compression1.5-2.0Upper boundLower bound1.5-2.0StaticStorm wave1.5-2.0earthquake2.0-2.5Limit equilibrium1.2-1.310 days1.5-1.8direct simple ear finite element, t-z, qz degradingage10 yearsThis report summanzes the results and conclusions of the research effort onidentifying the uncertainties in the prediction models by the means of comparingpredicted capacities obtained from several methods.IO

2.3Analysis models in this studyThe prevailing analytical models in use at present time are the discrete Winklerfoundation model of beam-column basedon non-linear soil support. Numerousresearchers have studied the pile-soil capacity problems using this model(Matlock, 1978;Kagawa, 1986; PMB, 1988; Bea, 1992; Wang, 1996; Lok and Pestana 1997). This modelis superior to the finite element model since its predictionfits the measured pilebehaviors better. The Winkler foundation model is illustrated in Fig. 2.1. This study alsotakes the discrete element method as basic approach. There are two major predictionmodels in this study: the SPASM lateral response prediction model, and the moreversatile analytical model developed by DRAIN3D structure analysis software package.The purpose of the these models is to investigate the ultimate behavior of the pile-soilsystem. ULSLEA is a simplified prediction model of the ultimate capacities of pilefoundation.SOILCOLUMNFig 2.1 Winkler pile foundation model11

2.3.1SPASM modelThere is a detailed description of the SPASM model used in this study in theearlier report of the first phase research (Jin and Bea, August, 1997). The basicmechanism of the model is shown in Fig. 2.2. This model has performed an excellentprediction of the lateral response of a single pile with different loading patterns and pilehead rigidities, while the steel pile is still in the elastic range.Rotational restraint ReDepth ofreducedresistance zoneY!YyXrResidual Resistanoe Pr depends on the depth from mud lineSprings and dashpot tosimulate the near-field soilresponseIFree-field soil/columnPile station:lumped elastioityY/YvFig. 2.2 Illustration of SPASM analysis model in this study12

One thing shall be noted is that this model assumes an elastic pile in soils. Thevalidity of the model is doubtable after the first yielding occurs in the pile. To keep aclear total picture of the pile response, this study assumes the steel pile exhibits anelastic-perfectly-plastic behavior. The second-order strain-hardening phenomenon isneglected. This means the stiffness of the pile will reduce to zero after yielding takesplace. So there will be a significant loss in the stiffness

American Petroleum Institute (API) has developed such guidelines for evaluation of the capacity of the pile foundations (API RP2A, 20th edition 1993). These guidelines address a wide scope of topics such as operating and environmental loading; determination of static capacity; influences on capacity, stiffness; applications of discrete element and continuum analytical models; use of in situ .

Related Documents:

loading stations for loading up to 1000TPH feature the popular and rugged MD30-EV, MV30-EV Series loading spouts and the MA30-EV Series which includes a filter module, air purging system and clean air fan. From a single stand alone loading station with a sepa-rate loading spout and dust collector to an integral Vaculoader, spout positioner

Dimensions: Main loading door is 8' wide x 10' high (2.4m x 3m). Loading door opens into loading area. This loading area then opens into a hallway, followed by the backstage shop. From main loading door to stage is 60' (18.2m) straight on except for a slight 15 angle between the loading area and backstage shop.

We offer three types of loading spouts to suit your needs: 1. Tanker loading spout (type NZU) 2. Truck/Wagon/Open pile loading spout (type NZO) 3. Combined loading spout (type NZK) HENNLICH ENGINEERING is the leading manufacturer of telescopic loading spouts - we have been producing them since 1996. We create customized solutions for your .

Summary Loading changes antenna characteristics - Feed Impedance, Gain and Angle of Radiation No reduction in the requirements for - height above earth - ground/radial systems With expedient application of the loading the efficiency of the system can be kept high - Apply 'end' loading first - Apply inductive loading in the order of ---

Transport Stock Chaser Personnel Carrier Burden Carrier BSC BUV BBC 4-wheel Cushion Walkie Reach StackerWalkie Reach Stacker 5,500lb capacity 4,500lb capacity 6,500lb capacity 6,500lb capacity 10,000lb towing capacity 4,400 - 5,500lb capacity 600lb capacity 1,000lb capacity . MAY 2014

Vortex Loading Spouts enable the fast and dust-free loading of many types of dry bulk solids such as powders, pellets, and granules. Standard volume loading spouts consist of internal material stacking cones that contain and direct the material flow using gravity and negative pressure. A flexible outer sleeve encompasses the stacking cones to .

of-the-art loading facilities to load and unload cryogenic liquids from ships ranging from all sizes. JLA developed the CryoTec Marine Loading Arm, to provide a safe and reliable solution for ship loading- and unloading facilities worldwide. The CryoTec is ideal for loading or unloadi

“BROTH LOADING” Afternoon and evening until 6–7 P.M. Banish those cravings and load up on fiber! Enjoy more broth and Broth-Loading Soups! Up to 48 ounces of bone broth can be consumed during the broth loading phase, including 8–16 ounces of Broth-Loading Soup. What is a Broth-Loading Soup?