Scribd
Upload
262 views228 pages

Performance of Jacking Pipes

A research about the Pipe Jacking Methodology and possible uses and limitations by Kevin J Ripley

Uploaded by

mohafrah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
262 views228 pages

Performance of Jacking Pipes

A research about the Pipe Jacking Methodology and possible uses and limitations by Kevin J Ripley

Uploaded by

mohafrah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 228
The Performance of Jacked Pipes by Kevin John Ripley A Thesis submitted for the Degice uf Douius of Philusuphy at the University of Oxford, Magdalen College Hilary Lerm, 1989. ‘Tho Performance of Jacked Pipes. Kevin J. Ripley Magdalen College, University of Oxford. A Thesis submitted for the Degree of Doctor of Philosophy. Hilary Term, 1989, ABSTRACT Pipejacking.is a tunnel construction technique which is increasing in popularity, but fundamental research is necessary to fully understand the extent of its possible uses and limitations. This dissertation reports on laboratory research into the performance of reinforced concrete pipes, assessment of pipe jolnts and the use of Joint packing materials. “The research hat addressed specific probleme which the tunnelling fraternity have raised. Model pipes have been constructed at scales of 1:6 and 1: 10.5 using reinforced microconcrete and they have been tested in either a sand filled chamber or between supporting yokes. Current British Standard tests have been used as a control on the quality of pipe manufacture. Data have been recorded of changes in soil pressures, pipe geometry and strains induced in the pipes. The tests have investigated defonuation of pipes, deflection angles Letweon vouscuulive pipes, dist tribution of stress concentrations and the effects of the use of joint packing materials on allowable Jacking loads and induced stress magnitndes in the pipes. Arreview of current pipejacking practice is presented and recommendations for the control and ‘supervision of pipejectiag, ‘operations are made. The conclusions include recommendations tor fieldwork monitoring and implications of this stage of the research to industry. Recommenda- tions arc made for maximum installation jacking loads for any given deflection angle between pipes. The prediction of friction angles at the pipe soil interface have been assessed at different sail stress levels and new recommendations are made. The effects of cyclic loading on the pipejacking system and transfer from jacking load to ground loading once the pipes are installed are presented. The criteria used in the selection of a recommended joint packing material for ‘use in jacking operations have been included. Failure modes ot pipes are stated and recom- mendations are made for pipe design, installation and monitoring to predict and prevent such failures. This dissertation is the report on the first stage of an overall programme of research which is ow set to progress with monitoring of pipejacking operations on several construction sits. Contents ABSTRACT Contents Acknowledgements Symbols CHAPTER 1 INTRODUCTION CHAPTER 2 BACKGROUND 2.1 Inuuduction: 2.2 Pipejacking 23 The need or research: 2.4 Oxford research 2.5 Jacked pipes and their design 25.1 Joints 2'5.2 Pipes 2.5.3 Manufacture 25.4 Joint packing materials 25.5 Load transfer at joints 2.6 Prediction of jacking forces 2.7 Reinforced microconerete 2.7.1 Model concrete 2.7.2 Model reinforcement 2.713 Toim sealing rings. CHAPTER 3 APPARATUS DESIGN 3.1 Introduction 3.2 Chamber and reaction frame 3.3 Hydhaulics 3.4 Transducers and instrumentation 35 Soil stressing system CHAPTER 4 MODELLING 4.1 Microconcrete 4.1.1 Trial mixes 4.1.2 Characteristic tests 4.1.3 Pipe manufacture 4.1.4 Pipe tests 4.1.5 Manufacturing tolerances 4.2 Reinforcement 4:3 Joint sealing rings CHAPTER & SAND CHAMBER TESTS 5.1 Introduction 5.2 Leighton buzzard 14-25 sand 53 Pipe tests 5.4 Results of the tests ‘3.4.1 Vertical and hortzontal positions 5.4.2 Joint gap 5.4.3 Preseure celle 5.4.4 Electrical resistance strain gauges (ERSG's) 1 Principal strain analysis and sign convention 5.4.42 Strain gauge results 5.4.5 Changes of deflection angle between pipes 5.5 Suumuay pyre SPREE eM PESSRELE BSGRSRRAES PREPPY PEPEpY pry eee eee ee ee RRSSS2aSss PRGA ARGU CHAPTER 6 MISALIGNED PIPE TESTS 6.1 Introduction 6.2 Misalignment apparatus 6.3 Test (6-4 Diagonal axial toading results 6.4.1 Deformations of the pipe 2 fect of pipe yokes 6.4.3 Strain gauge readings 6.4.4 Failure modes 65 Edge loading results 65:1 Deformations of the pipe 6.3.2 Suain gauge readings 65.3 Failure modes 6.6 Summary of misalignment tects CHAPTER 7 JOINT PACKING MATERIAL 7.1 Introduction 7.2 Test programme 73 Results 731 Stress/strain relationship 73.2 Pffect of cyclic loading 7/3.3 Variation of packing material thickness 73.4 Changes of material densities during saturation and drying 13.9 tect of material conditions 1.3.6 Differences between materials 7(3.7 Strain gauge readings 7328 Tensile strain distribution 7.4 Stress distribution at joints 7/5 Profiled packing material CHAPTER § MOVING PIPES 8.1 Outline 8.2 Test procedure 8.3 Results Resistance to jacking Boundary pressures 8.4 Summary CHAPTER 9 ANALYSIS AND DISCUSSION 911 Intenduction 9.2 Realignment during chamber tests 9.3 Changes of pipe alignment in practice 9/4 Distnbution of soil stress on the pipe 9.5 Angle of friction between sand and pipe 9.6 Stresscs measured at the chamber boundary during moving pipe tests 9.7 Joint packing materials 9.8 Profiled joint packing materials 9.9 British Standard crushing tests CHAPIEK 10, CUNCLUDING REMARKS REFERENCES ii 61 61 61 66 Oil Gil 615 6-19 Qe RB 627 REE RBB PP PVP PPP BSBSRREBaaHSe PESRSEL EB SB pepreserer E Sgaspeense Q Acknowledgements This research is the outcome of much effort and encouragement from many interested parties. I wish to express my gratitude and thanks for the support and stimulation given to me by all members of the Pipe Jacking Association and Concrete Pipe Association and for their con- siderable financial backing. The project was also financed by the Science and Engineering Research Council. Twould like to thank Professor Peter Wrath for s assistance in instigating the research project at Oxford and Dr. George Milligan for his excellent supervision during the work and his perpetual stimulation of new ideas. I consider he gave me the unique opportunity and freedom to develop the research project and allowed me to choose the direction of its progress. The Soil Mechani Group at Oxford provided an enjoyable atmosphere in which to work and the friendship of ‘members of the group was 4 continuing source of inspiration. My own participation in this work would not have been possible but for the foresight of my employers Delta Civil Engineering Co. Ltd., who gave me the time away from the Company to carry out the work. They have been a continued source of support and loyalty. I should express my gratitude and indebtedness especially to Amos who continues to prompt me with new goals and fully supports my attaining them. Finally, I would like to thank my parents and family for their snpport aver many years. and it «gives me great pleasure to thank Margaret for her encouragement, time and patience during the period while I was writing this dissertation. iii reves Feo Re een f Pmeos & Symbols Initial packing material thickness Packing material compression Contact breadth of packing material Pipe external diameter Radius of pipe end profile Elastic modulus Minimum voids ratio Maximum voids ratio Applied jacking load Horizontal pipe movement Relative density Relative dilatancy index Coefficient of earth pressure at rest Pipe length Rending moment at crown Bending moment at springing point Total soil load on pipe Mean effective soil stress External radius of pipe Anternal radius ot pipe Wall thickness of pipe Vertical pipe movernent Diametrical contact width at pipe joint Radial angle around pipe Angular deflection at pipe joints Unit weight of soil cL ‘Angle of friction betwoon pipe and aoil Major principal strain Minor pri cipal suain Poisson's tativ Maaimuu inducod sucss ‘Total horizontal soil stress Maximum stress at pipe joint Radial stress on outside of pipe Total vertical soil stress Internal angle of friction of the sand Critical state friction angle Direction of principal strain CHAPTER 1 INTRODUCTION Jacked pipes are used in the Civil Engineering industry for the installation of underground pipelines by tunnelling using a trenchless construction technique called pipejacking. The system has been used for many years but its increasing world-wide popularity and use over the last decade has led to the need for a fundamental assessment of its uses and limitations. A programme of research, beginning with the laboratory modelling presented in this dis- sertation, has been instigated. It will lead onto correlation with fieldwork and monitoring of prototype pipeiacks. Problems with pipejacking were reported by Shullock (1982) when a contract encountered ‘unexpected ground conditions which eventually resulted in failure of some of the pipes being used for construction by pipejacking. The Pipe Jacking Association (PJA) and Construction Tndusuies Reseach and Lifuuuation Association (CTRIA) lad alicady undertaken a review of pipejacking which was published hy Craig (1983) This ontlined some. specific tapies for investigation and research which were taken up by the PTA with the commencement of this research programme. The principal topics for investigation were outlined as: friction loads in different ground conditions; behaviour of pipe joints; cyclic loading effects; reduction of, frictional resistance and prediction of installation forces. Research began with investigations {into load capacities and distribution of loads in model pipes. Introduction 1-1 ‘The Inhoratory work reported here has concentrated an the: first three points Prediction and reduction of friction forces will become more clear as a result of the overall programme. ‘The use of scale models and testing in the laboratory has enabled conditions during tests to be controlled and has led to reproduction of the pipe failure modes occasionally experienced on pipejacking contracts. This has enabled assessment of the reasons for failures and con- cclustons as to how they can be prevented, ‘The report of the laboratory research begins with Chapter 2 which gives a background to the pipejacking method of construction and its uses. It continues to explain the reasons for is hoped might be achieved and a review of previous microconcrete ‘modelling techniques. A section in the Chapter presents information about current practice the research, what it in pipe design and manufacture and literature about load transfer at joints between pipes. ‘Chapter 3 continues to introduce the design uf the apparatus used in the laboratory and how tests were controlled and monitored, ‘The techniques of modelling investigated for manufacture of jacking pipes are presented in Chapter 4. Methods of controlling standards and quality in model preparation are reviewed and reasons for the selection of the microconcrete mix and reinforcement are stated. (Chapters 5 to 8 deal in turn with four different test series. Chapter S presents tests and results carried out on model pipes in a sand filled stress controlled chamber, Results are included assessing how pipes move and interact with the sand when axial load is applied to them. As aresult of some of the findings in this test series, apparatus was designed to maintain pipes {in misaligned positions while axial load was applied. Results from this test series are pres- ented in Chapter 6 which reports on the results of tests modelling the two most common modes of load application. Diagonal luading and edge loading of the pipes were used in the tests. Introduction 1-2 Toint packing materials and their changing characteristics when subjected to cyclic loading are the subject of Chapter 7. Predictions of axial load capacities of prototype pipes can begin tobe made as aresult ofthese tests. Recommendations are made for the most suitable packing ‘material to be used in pipejacking. Chapter 8 assesses interaction at the pipe/soil interface and how the friction angle at the interface affects predictions of jacking load. ‘The results from all the test sorics are reviewed and discussed in Chaptor 9. Distribution of soil stresses on the pipe are analysed. Changes of pipe alignments during misaligned tests are related to changes in the orientation of pipes during prototype pipejacking. Axial load capacities of pipes are assessed and figures tabulated for capacities at various angular deflections. The use of profiled joint packing materials is reviewed and results of the angle of friction between the pipe and sand are discussed. British Standard tests on pipes are presented aud results ave assessed for improved control aud specification of the tests. ‘Theconclusions from the research are drawn togetherin Chapter 10 where recommendations for the future fieldwork monitoring in the project are made. Control and management of pipejacking operations are assessed and how, if at all, the pipejackers’ and designers’ philosophies might be encouraged tochange as aresult of the continuing research programme, Introduction 1-3 CHAPTER 2 BACKGROUND 2.1 Introduction Pipejacking is a tunnel construction technique used to install pipes into a near horizontally excavated hole. It is being used more frequently in preference to trenching or segmental tunnelling; pipejacks have increased their United Kingdom market share from 10% to over 30% during the last five years. The advent of microtunnelling and its increasing use will increase this figure. Pipefacking 1s used to avold interference with surface waffic, services and the environment. Increasing pressure is being brought upon authorities to use the techniqueas itis ‘minimum disruption’ method and popularin urban areas forinfrastructure renewal. Pipejacking techniques are used to provide pipelines, conduits or access for sewers, gas and water mains, electric and telephone cables, sewer relining, subways and land drainage. Installation of pipcs by jacking roquises large forces to be provided by hydraulic powerpacks and rams. A thrust ring distribntes these forces around the circumference of the pipe being jacked. Jacking equipment is accommodated in a thrust pit, shown in Plate 2.1, with a thrust wall to provide a reaction against which jacking forces can be applied, as depicted in Figure 21. Background 2-1 q : 5 2 a Plate 2.1 A view of a typical pipejacking thrust pit. Excavation is normally carried out within a shield in front of the pipeline being jacked. The ground is excavated using pneumatic tools, hydraulic shovels or boring machines, the soil being transported along the pipeline to the surface and pipes advanced as the groundis mined. Corrections to the alignment of the pipes are made using hydraulic rams in the shield in conjunction with frequent surveying to fixed reference points Background 2-3 2.2 Pipejacking ‘The use of pipejacking has hecame enmmon place during the last two decades The earliest recorded use of the pipejacking method was in America about 1910, Richardson and Mayo (1941). The basic principles of the technique have been presented in detail by the American Pipe Jacking Association (1960), Richardson (1970), Hough (1974), Drennon (1979) and. Clark and Thompson (1983). The British Pipe Jacking Association publish notes giving guidance on design and practice (Pipe Jacking Assoclatlon 1981 and 1986). The increasing use of modem technology and 2 competitive tendering market have led to many innovative methods being introduced in recent years. Advances in technology have been particularly significant in Japan and West Germany, where pipes from 200mm to 4000mm diameter are jacked in almost any ground conditions, to alignments which include horizontal and vertical curves, over ever increasing distances, Tho introduction of mechanised cxeavation has increased jacking distances. In Great Britain, Wallis (1982) reports an 180mm diameter pipejack. 460m long in London Clay: Byles (1983) reports an increase in length to 560m in water bearing sands and gravels; Winfield (1986) records 690m of 1950mm diameter pipes jacked through sandstone and siltstone, whilst in 1988, 1500m of 330mm external diameter pipes were jacked in mixed ground conditions around a number of curves, Dumbleton (1988). A point is reached in most con- struction comacts when it becomes more economic voexcavate a thrust shaft and commence anew pipejack, rather than be delayed by increasing muck haulage distances and casts, but improvements in techniques are constantly increasing this distance. New methods of construction and new materials are constantly being developed and implemented. Richardson and Scruby (1981) reported on the development of the Uni-Tunnel system in which pipes are jacked forward by intlatable bladders positioned between suc- cessive pipes. A now jointing profile is reported by Cole (1986) used for pipejacking in Background 2-4 Greenwich and Newman (1986) reports on the use of vitrified clay pipecin Newcastle White et al. (1988) presented details of the use of new joint packing materials, and microtunnel technology using unreinforced concrete pipes in a joint venture project based in Yorkshire. ‘A new British Standard for concrete pipejacking pipes in Great Britain is to be published shortly as BS 9911: part 4 (1986 dratt). Concrete jacking pipes are generally manufactured by contrifugal apinning or vertical strength greater than SON/mm’, Pipes are normally reinforced with spiral and longitudinal steel spot welded to form internal and external cages. The specific method of manufacture 1S using concrete with a 28 day characteristic crushing and material characteristics are dependent upon the manufacturer. Materials used for manufacturing jacking pipes include:- 1 Unreinforced canerete Reinforced concrete, Vitrified clay. Glass reinforced plastic. Asbestos cement, aunrurn Steel and ductile tron. 7 Concrete and recin laminatec or composites. Manufacturers of concrete pipes design their pipes to comply with BS 8110: 1985 and BS ‘5911: part 4 (1986 drafi). The pipes must withstand loadings imposed under the various tests of BS 5911. They are designed to higher standards and strengths if ground conditions or Clients’ specifications impose more onerous design requirements. Background 2-5 2.3 The need for rescarch Until 1980, any research carried out into pipejacking was commissioned by individual ‘companies to meet their specific requirements. No attempt was made to collate this research and in the course of time their documentation has been mislaid. The Pipe Jacking Association. (PIA) and Construction Industries Research and Information Association (CIRIA) have been, instrumental in establishing a research programme with full support from all interested parties. CIKIA undertook a review ot pipejacking and published a Technical Note 112, Craig (1983), as a state of the art review, which recommonded that the following areas required research:- 1 Friction loads in different ground conditions. Characteristics of joints and joint packing materials. Effects of cyclic loading on pipes. 4 Bffecis uf lubricams in reducing friction, 5 Development of a site investigation test to predict friction forces. Kirkland (1982) had previously reported on the requirements for pipejacking research when he stated theneed forresearchers to establish more scientific facts to supporta sound concept. ollowing publication of the echnical Note 112, a second phase of CIKIA's Work involves the collcetion and analysis of pipejacking data. This relics on contractors answering ques tionnaires about their jacking contracts. No details of the survey have been published. CIRIA have published a Technical Note 127, Watson (1987), which is a survey of trenchless construction for underground services. The Note recommends future work in full scale field trials, monitoring by an independent body and recording of detailed statistics relating par- Gculatly we method of installadon, materials used and vusts. Many ofthe uther tecluiiques Background 2-6 ced in trenchless construction have been adopted from the principles of pipejacking. The findings of pipejacking research are often equally applicable to the other trenchless tech- niques. Following the recommendations of CIRIA and enthusiasm of the PJA in promoting and supporting research,a project commenced at Uxtord University dunng 1486 ands the subject of thio thesis. Milligan (1986), reports on current research programmes being undertaken. Failures of clayware jacking pipes have been reported (Winey 1987) as occurring on a Scottish contract, Large numbers of pipes were cracking, prompting research intojoint details forclayware pipes. More recently in Germany, pipes were seen to crack due to excess jacking load and poor control of alignment (Winney 1988a and Winney 1988b). It should be noted that curt « production teclniques du not permit reinforcement uf elay pipes. Clay wae pipes ate only used on pipejacks less than 600mm in diameter. Itis understood that allowable jacking stresses for clay pipes are less than for concrete pipes and economic considerations most often rule out their use. 2.4 Oxford research Pipejacking research at Oxford University was commenced by the author with support from the Science and Engineering Research Council (SERC) for equipment and supervision and with the PJA and Conerete Pipe Accociation (CPA) providing further finance for staff, together with invaluable guidance, advice and industrial involvement. There was limited knowledge about the behaviour of jacked pipes during installation and disagreement on the explanations of some failures in the field. Research proposals were broadly based to allow a specific direction to be chosen once a basic understanding of the problems had been achieved. The initial research programme presented in this thesis is based on laboratory testing of scalc models. Background 2-7 ‘The overall objectives of the research are to improve understanding of the loads applied to {jacked pipes, both during and after installation and to establish a fieldwork programme. It is imended to make recommendations of the most suitable and economic data collection for the exercise of monitoring performance of full-scale pipejacking operations. ‘Model test results have a limited application to the prototype behaviour due to the ditticulties ‘of modelling precisely many aspects of the problem such as self weight, stress level, stress path and dilation behaviour. Theoretical calculations can be carried out before results of ‘model tests are applied to prototype predictions. Tests can later be carried out on prototype pipejacks to correlate to model test results, but with a large saving in cost. Bassett (1979), outlines the advantages of small scale physical models for studying soil Juanics probleurs:~ 1 Testing models is relatively cheap, easily modified and quick to conduct. Soil properties can be chosen within known limits. Simple controlled variations can be introduced one by one as required. Internal movernents of soil or rupture lines can be observed. Homogeneity of soil provides easier analysis. auweoun ects are repeatable to provide data for experimental and statistical scatter. ‘The scale model testing programme at Oxford has been used to investigate the following points: 1 The distribution of concrete strains during pipejacking and how these change when soil loading uansfers tw the pipe ay itis installed in its final position. 2. Parameters relating the allowahle jacking Inad to the soil type and angular misalign- ‘ments between consecutive pipes. Background 2-8 3. Thedistribntianaf stress concentrations hetween misaligned pipes, thense ofa packing ‘material to reduce these concentrations and to find a suitable joint profile to minimise stresses and retain pipeline watertightness and integrity. 4. The possibility of using a suitable jacking strength test to supplement the existing British Standard tests and to test the compliance of model pipes with BS 5911: part 4 (1986 draft). S The effect of coil strece level on the behaviour of the pipes during installation, in particular changes in shape and orientation of the pipes with different applied jacking loads. 6 The most suitable and economic measurements to be made to monitor performance of full-scale pipe jacking operation, 7. The various failure modes occurring in jacked pipes and how pipes can be designed, installed and monitored to predict anu preveut such failutes. 2.5 Jacked pipes and their design ‘The primary loads imposed on a pipejack are:- 1 Installation jacking load. 2. Face vesistance during jacking, 4. Ground load, both surcharge and transient 4 Loads from footings and buildings. 5. Fluid pressure, internal and external. Background 2-9 British manufacturers’ design calculations assess compliance of the pipes’ strength to withstand loadings imposed on the installed pipe. Allowance for installation loads receives scant attention, possibly due to the uncertainty surrounding loads, stresses and their dis- tribution throughout the jacking cycle. ‘The British Standard 5911: part4 (1986 dratt) details the testrequirements tor jacked concrete pipes. The drafting committee were aware at the time of writing that research was required to establish a suitable test to assess a pipe’s ability to sustain installation end loads. The forward to the Standard states "The joint face strength testis included in the absence of a suitable jacking strength test. Research is presently being undertaken in order to devise such atest and this may be incorporated in a future revision." The British Standard Institute is looking towards the results ofthis research programme to provide guidance on amore suitable 2.5.1 Joints ‘The functional requirements of a joint on a jacked pipe are adapted from Clarke (1968) as:- 1 Itshould be constructed to the standards laid down in BS 5911: part 4 (1986 draft). 2 It should be designed to permit angular and axial movement large enough to tolerate maximum displacement likely to occur, withour damage or loss of waterighmness. 3 It chould be decigned to withstand the force applied during installation without detrimental damage. 4 It should remain efficient throughout its working life ic. use materials that do not change with time or within the exposed environments. Background 2-10 5 It should be cimple to make and dismantle in the limited space of the jacking thrust pit. 6 It should remain effective under vibration. 7 Itshould be capable of withstanding the service loading conditions. ‘Traditionally jacked pipe joints in the United Kingdom have been an in wall spigot and socket as shown in Figure 2.2a. Disquict in the industry about the performance of the in wall joint and its ability to transmit longitudinal loads has led to the introduction of new joint details. The most often used of the new joints is shown in Figure 2.2b; the steel collar joint. ‘The main reason for its introduction was the belief that jacking loads would be better ‘transmitted through the centre of the jacked pipe wall rather than its edges. An increase of available end area would improve the pipe’s jacking load carrying capacity. Limited design or research was carried out co ascertain the parameters for configuration and performance of the steel collar joint detail. The CIRIA Technical Notes 112 and 127 give many more examples of joint profiles and arrangements used both in the United Kingdom and around the world. The more common of these details are reproduced in Figure 2.3. Background 2-11 a STFFI COLI AR oS a RUBBER RING N\A JOINT PACKING MATERIAL, INTERNAL DIAMETER, a) LEZLEY EXTERNAL Cy DIAMETER Spee \ \ iS = N a RUBBER SEAL JOINT PACKING MATERIAL INTERNAL? piaweTER b) STEEL COLLAR JOINT Figure 2.2 Types of pipe joint common in Britain, Background 2-12 q EXTERNAL DIAMETER ‘SEALING RING INTERNAL pret aia JOINT FOR GRP JACKING PIPE SOFT RUBBER RING eaten STEEL COLLAR 7 ff Ui / fo PROFILED SEAL INTERNAL, PACKING RING DIAMETER, GERMAN CLAYWARE JACKING PIPE JOINT. PAC! DOWEL JOINT Figure 2.3 Other common pipe joint details, Background 2-13, 2.5.2 Pipes Jacking pipes are designed to withstand the Class H proof load of BS S911. Pipes are also designed to CP 110: 1972 and BS 8110: 1985 to withstand ground loading conditions only. No specific calculations are carried out to assess a pipe’s axial loading capacity. For the purposes of design it is usual to make the assumption that in ideal jacking conditions the pipes are subjected to an axial load evenly distributed around the curcumterence ot the pipe. A view along a pipejack is shown in Plate 2.2. The worst case is assumed to be the pipe nearest the thrust ring which will be subjected to the maximum jacking force from the rams, Plate 2.2 General view along a pipejack during construction. Background 2-14 Tnatypical design example for an 1800mm intemal diameter pipe, ajackingload of 10,000kN produces an average direct compression at a pipe’s spigot face of 11N/mm and an average direct compression inthe full pipe section of 9N/mm?, A manufacturer's qualifying statement ‘might be along the lines that “excessive departure from uniform jacking conditions will reduce the acceptable load. Equally some increase in jacking load may be practicable under ‘uniform conditions. Expenence rather than theoretical calculations wall be the ultimate guide." No data are quoted, and it is one of the aims of the pipcjacking research programme to enable manufacturers and contractors to offer clients and design engineers a theoretically based method for axial load design of pipes for jacking. British pipe manufacturers specify their pipes as capable of sustaining a uniformly distributed end stress of between 10 and 15N/mm* (dependent on manufacturer and pipe diameter). ‘These figuics appcar vonscivative if loads are applied tw che full cmd area of the pipe, but allow for end tolerances and jacking misalignment i an athitrary fashion 2.5.3 Manufacture Until recently the majority of pipes used in the United Kingdom were manufactured with reinforced concrete by centrifugal spinning. During spinning, concrete is poured into pipe ‘mouldsin two operations. The centrifugal forces drive out entrained air, butalso cause some particle segregation. Two distinct layers can be seen with a layer of cement paste separating them, when a completed pipe is cut. Pipesare now commonly cast vertically with concrete poured between inner and outer moulds from one end because this method of construction is much safer and cleaner, and requires Jess skilled personnel for the operation. The moulds have vibrators attached to them which are operated in sequence as the moulds are filled. Some segregation occurs due to the dif- Background 2-15 ference in vibration times from bottom to top of the mould. The top end of the pipe, when cast, has to be hand finished using a trowel which results in a less smooth end profile to the pipe than could have been achieved with the use of end moulds. 2.5.4 Joint packing materials Compressible joint packing material is recommended by all jacking pipe manufacturers. However, not all ae prepared to specity the type ot material or advise in any detail on its ‘use, Thie research project has been investigating various packing materials, to what extent bine misalignment can be accepted with each type. and where the packing should be posi- tioned. ‘The purpose of the packing materials is to provide a compressible medium between the two adjacent concrete faces. This aids distribution of stress concentrations over larger areas and avoids c« trations of sucss induced by uncveness un the pipo’s cuds. It should be nuted that misalignment hetween consecutive pipes can result from a lack of pipe. end equarenese even though they comply with manufacturing tolerances. Itis therefore important tobe aware of the end squareness of pipes. 2.5.5 Load transfer at joints ‘The design of joints and distribution of load across their faces has not been investigated in any detail. The Concrete Pipe Association of Australia (1983) published a theury based on material properties, their etrese/etrain relationship and compressive strengths (Figure 2.4). ‘Since commencement of research at Oxford, Hornung et al. (1987) presented a guide for calculation and design of jacking pipes according to German codes and specifications. It presents the authors’ considerations for loading, calculation and dimensioning criteria. The Background 2-16 guide details information considered on a particular contract and presents the design con- siderations for pipe deflections and a suitable packing material. A summary of this guide’s recommendations on packing material and stress distribution at the pipe joint is presented in Figure 2.5 and relates them to jacking pipes around curves. No allowance has been made for pipe elasticity and only the material properties of the timber packing and orientation of the plpe are considered. Clarke (1968) wrote a manual about buried pipelines which noted some points that are relevant to pipejacking. Clarke wrote that radial compression of a joint sealing ring induces circumferential stress in the pipe’s socket and collar. The intensity of the stress is dependent fon the hardness of the rubber and the degree of its compression. The load is additional to the stresses induced by internal and external service loads and jacking loads imposed during installation. Tests on the jointing rubber sing show a rapid increase in socket Ivading ay compression of the jointing ring is increased The compression needs to he sufficient ta ‘maintain a watertight seal during testing and service life, whilst allowing for stress relaxation of the rubber. Background 2-17 ‘The analysis presented by the Concrete Pipe Association of Australia is summarised as follows:- Retnttt Total deformation Aa =Aa +AL ¥ 2. fF , oe % 5 = Joim Packer Concrete 4 buto=" where t= wall thickness and = wall thickness at joint Permissible joint deflection: B— tan oe x from sables (Paaaiget c vt Vv T PACKER ory | ; TEE fw ; Figure 2.4 Australian joint stress distribution. Background 2-18 “TIMBER ELASTIC MOOULUS E, re zueae, a0, Figure 2.5 German joint stress distribution (Homung et al. 1987). 2.6 Prediction of jacking forces Pipejacking contractors use empirical methods to predict jacking forces and allow for pipe/soil friction forces in different ground conditions. Typical values of friction loads are listed in Table 2.1 below, Craig (1983). Background 2-19 Ground Type External Load (kN/m*) Rock 23 Boulder Clay 5-18 rm Clay 3-20 Wet Sand 10-15 Silt 5-20 Dry Loose Sand 25-45 Fill 45 Table 2.1 Typical values of Friction Loads on Pipes Prediction of jacking forces is ohtained hy multiplying fignres fram Table. 2.1 by the total extemal surface area of the pipes being installed. Jacking forces can be reduced, Durden (1982), by the use of injected lubricants, intermediate jacking stations (shown in Plate 2.3), jacking n twodirections and making underground connections and using mechanised jacking ‘methods which speed up construction operations. Analytical predictions of jacking forces are detailed by Auld (1982) based on standard coil properties and theories and are discussed later in this Section. Auld’s predictions of jacking Toads are based on ground pressures, the angle of internal friction of the surrounding soil and an "experienced guess", The following points are listed as influencing jacking loads:- 1 Resistance at excavation face. Amount of ovexcut during excavation, Variatinne in grand canditions Recommencement of jacking after a weekend stoppage. Steps at joint. awa oe Joint deformation. Background 2-20 Plate 2.3 An intermediate jacking station installed between pipes. 7 Misalignment of jacking pipe. 8 Jacking around curves. 9 Injection of a lubricant into the overbreak void. 10 The use of interjack stations. Haslem (1986) and O'Reilly and Rogers (1987) have presented papers proposing new approaches to predictions of jacking forces based on the analysis of field data from selected pipejacks. Haslem’s analysis was for pipejacks in London Clay and O'Reilly's for pipejacks Background 2-21 in sandstone, Both authore presented methods for dealing with misalignment of pipes. More work is required for predictions in different and variable ground conditions: a series of analyses applicable to each type of ground condition would be an ideal outcome, ‘The paper presented by Haslem studied the behaviour of an clastic cylinder resting in a cylindrical void in an clastic continuum. Comparisons made between the predicted jacking forces and those capericuced in he ficld wee found w give an under esti wate of jaching forces. Adjustments were male to allow for misalignment af pipes hut were fonind to be negligible. Hastem concluded that the differences he found between predicted jacking forces and those encountered on the studied pipejacks may have been from another cause. It is interesting to compare these remarks with the values of the angle of friction between the pipe and soil discussed in Section 9.5. (O'Reilly and Rogers used the came techniquee to compare predicted and field measurements and find a similar low estimate for pipeiacks through cohesive materials. However, pre~ dictions for pipejacks through rock produced comparable results if the site conditions were uniform. Laboratory tests were carried out to assess the contact area between pipes and clay and the results from the experimental data confirmed a need to account for plastic behaviour of clay and the ettect of time. ‘The commonly used empirical approach predicts jacking forces with reasonable accnracy. Errors encountered using this approach are most likely due to variation in soil properties or large misalignments of the pipejack. The presentations of Haslem and O'Reilly only allow for small misalignments of the pipejacks and one consistent ground condition. ‘The analysis of each author 1s relevant to the ground conditions and contractor of the specific contract. Variatione between the jacking loads experionced by the different contractore can Background 2-22 he due to the method of working, labour employed, amount of supervision, jacking arrangement, pressure of lubricant injection and excavation method (type of machine orhand. excavation). Auld (1982) presented theories for calculation of pressure distribution due to cohesionless ‘ground loading onto a jacked pipe. Auld stated that he was not satisfied with the pressure distribution obtained from the analysis. ‘The approach is ambiguous because no allowance is made for changing depth below ground tevel (see Section 9.4). ‘The author of this thesis has taken a simplistic approach at this stage in assuming no change of vertical or horizontal stress with the changing depth below ground level (Figure 2.6). Using Mobr’s circle for stress in the soil, the radial stress at any point on the pipe is given by (a, +6) (6,- 6) wiQ’2t ry tyne where 6, = total stress on pipe 6, = total horizontal soil stress ©, = total vertical soil stress a= vertical angle to position of strcas calculation The total load on the pipe can be obtained by integrating. = ©,+0y) (6, Pee 2f[ea22 + where R = pipe external radius. o Dos 2a Pays =TR(0,46,) Eq 2.2 Background 2-23 ‘The frictional resistance which needs to be overcame: along the pipeline being pushed into the ground is F =Pygtand Bq 23 where 6 is the angle of friction between the pipe and soil. 5 is generally taken as 0,7 of the angle of internal friction, 6, of the soil. thtt 0. v Figure 2.6 Proposed stress distribution on a pipe. More detailed analysis is presented in Chapter 9 which allows for soul stress variations with depth, Background 2-24 2.7 Reinforced micraconcrete The use of reinforced microconcrete for scale model testing of reinforced concrete structures hhas been frequently used and developed, The principles of the use of microconerete have been reviewed and the more influential characteristics investigated to enable selection of a suitable design mix and reinforcement material to model the jacking pipes. Concsets ishighly inclasti iu compression and wcusiou, This vreates problems iu avempuing, to model n minforeed concrete structure to fail at the modelled capacity and in the correct ‘mode. Thereinforcement alsopresents amajor difficulty; its strengthand bond characteristics must be given careful attention for correct modelling, ‘The failure criteria for model concrete should be identical to that for the prototype concrete, However, concrete does not have a well detined failure criterion. Therefore, model concrete should have geometrically cimilar otreca/strain curves for uniaxial comprescion and tension. and the model and prototype strains at failure should be the same (Sabnis et al. 1983). The current research has placed emphasis on reproduction of failure mode and correlation between model and prototype British Standard tests. 2.7.1 Model concrete ‘A well graded aggregate is recommended to produce concrete with minimum voids, maxi- ‘mum strength and minimum volume change during curing. The maximum aggregate size is determined by geometric scaling. Background 2-25 ‘The engineering properties of concrete are influenced by: 1 Water|cement ratio. 2 Aggregate/cement ratio. 3 Aggregate dimensions and properties. 4 Cement ype. 5 History of moisture available and temperature during curing. 6 Testage. 7 ‘Type of stress caused by loading. 8 Duration and rate of loading. ‘The influence of aggregate size on deformation of concrete was investigated by the Cement and Concrete Association, Hughes and Chapman (1966). It was concluded that as the size ‘of aggregate increased so compressive strength and elastic modulus in compression decreased. Tensile strength and elastic modulus in tension decreased with increased aggregate size and roundness. ‘More recent investigations have been conducted into the effects of size on microconcrete behaviour, Sabnis and Mirza (1979), the modelling ot stress/strain relationships ot structural concrete, Noor and Wijayasri (1982), and the comparative flexural behaviour of model and prototype beams. Evans and Clarke (1981). The conclusions reached in each of the studies have influenced the tests and choice of microconcrete and model reinforcement used for this research. Each of the investigators used Rapid Hardening Portland Cement (RHPC) tocnable results to be obtained in the laboratory more quickly. RHPC produces concrete with a seven day compressive strength comparable to the twenty eight day strength if Ordinary Portland Cement (OPC) were wed, Previous investigawors have used a variety of different aggregates in concrete mixes. The results of their tests gave some indication as to how sensitive mixes are to the influence of aggregate properties, mix proportions and test procedures upon the Background 2-26 compressive strengthe obtained, Figure 2.7. Maintenance of standarls and procedures throughout a test series is essential to the manufacture of concrete with consistent quality and to enable valid comparisons to be made. 2.7.2. Model reinforcement ‘The properties requiring investigation during the choice of a reinforcement to mode! the behaviour of provorype relnforcements are summarised as follows:- 1 Yield and ultimate strengths are to be duplicated. 2 Bond at the steelfconcrete interface is to be similar. 3 Duetility of reinforcement is to be the same as in the prototype. ‘The author's review encountered many types ot bar and wire having been investigated tor use as a model reinforecment. These included:- 1 Round steel wire and bar. 2 Square steel wire. 3 Custom deformed wire. 4° Threaded rod. 3 Commercially deformed wire, It was apparent that one of the modelling problems would be to reproduce a similar number and pattern of cracks to the prototype. The conclusions as to the most suitable wire to use pointed to a black annealed stee! with deformations somewhere between those produced by knurled wheels and those of a threaded bar (White and Clark (1980), Noor and Khalid (1980) and Evans and Clarke (1981). Background 2-27 ‘MIX MIX PROPORTIONS AGGREGATE COMPRESSIVE REFERENCE REFERENCE W/C A/C TYPE STRENGTH RATIO RATIO (N/mm) A 08 3.25 2mm DOWN 21 WHITE (1983) B 07 3.6 6mm DOWN 2B SARNIS & MIRZA (1979) c 0.54 24 3mm DOWN 27 HUGHES & CHAPMAN: (1966) D 07 3.5 10mm DOWN 34 TSUI & MIRZA (1969) B06 «= 4 BdmmDOWN 35.“ BVANS & CLARK (1978) F 0.44 28 2mm DOWN 4 NOOR & WUAYASRI (1982) G 0.63 5.6 10mm DOWN 42 WHITE & CLARK. (1980) H 0.515 275 236mm 4B WALDRON, PINKNEY DOWN, & PERRY (1980) T 0s 38 ‘Imm DOWN: 0 WHITE (1983) J 04 25 10mm DOWN 14 WHITE & CLARK 1980) °° ‘COMPARISON OF MICROCONCRETE MIXES: ua i e § os : z f iF. 5 os : . 0 = Compressive sTaENoTerue?) ‘ [AGGREGATE/CENENT RATIO Figure 2.7 Comparison of mix proportions and compressive strengths. Background 2-28 2.73 Joint sealing rings. Pipes with in wall jaints have always ised a cineular section rolling muibher ring to provide water tightness and integrity of joints. The development of other joint profiles has led to changes of the cross section of the rubber ring and to its material properties. The most commonly used joint sealing ring for steel collar joints is shown in Figure 2.2b, It is a rectangular section, located in a groove on the pipe spigot and uses two fins to locate and. seal in the steel collar, The first fin is lower than the second fin and positions tne jointing pipes into their correct central alignment, The steel collar then meete the eecond fin which provides watertightness and sealing as it is compressed and bent. Background 2-29 CHAPTER 3 APPARATUS DESIGN 3.1 Intraduetion Details of the design for the testing apparatus, instrumentation and systems incorporated in the testrig are presented in this Chapter-Design calculations are not included but were carried out with reference to British Standard 5500 (1985), Donel (1976), Williams and Aalami (1979) and Structural Steelwork Properties and Safe Load Tables (1978). The equipment was designed to enable modelling of pipejacking through a stress controlled soil chamber. Later modifications allowed for testing of pipes in free ait and cyclic load testing of juint packing materials. Pipes were mannfactured ta climensions snfficiently large to allow neces for measurement of deflections, deformations and concrete strains. 3.2 Chamber and reaction frame ‘The aim of the research was toimprove understanding of the factors controlling the behaviour of jacked pipes during installation and after completion of jacking operations. The apparatus ‘was designed ro enable various end loads to be applled 10 static and moving pipes. Verdcal stress (due to overburden), can be applied and varied independently of horizontal stress. Apparatus Design 3-1 ‘The outside diameter of all model pipes was standardised at 200mm. Madel pipes were constructed with wall thickness to internal diameter ratios of 1:6 and 1:12. These represent ‘true scale models of 900mm and 1800mm internal diameter prototype pipes at scales of 1:6 and 1:10.5 respectively. Further details of pipe manufacture and material selection are ‘resented in the nextChapter. The minimum internal diameter of model pipes was 150mm, ‘which allowed adequate space for mounting transducers to record pipe movements and. ‘concrete strains, ‘The soil testing chamber is cubical with internal dimensions of 600mm and is shown in Plate 3.1. This allows forthe pipes to be surrounded by sand equal in thickness to one pipe diameter, ‘Transducers were mounted in the chamber wall to monitor changes in sand stress at the boundary of the chamber, as it was evident from previous research that boundary conditions wore guing Ww influence the behavivur of the pipes. The chamber was sufficicully lung w enable two pipe eections to be completely retained wit in the chamber. The. common, between the two pipes was approximately central within the chamber during static testing and this was the position where most monitoring was carried out. ‘The test chamber clamped into a compressive testing framework which both supported the chamber and provided a reaction frame, The general arrangement can be seen in Figure 3.1. Jacking forces were transmitted to model pipes by hydraulio jacks, Apparatus Design 3-2 Plate 3.1. A general view of the test apparatus. ‘The test chamber was filled with Leighton Buzzard 14-25 sand. To do this, the chamber was rotated ninety degrees trom its test orientation and the top removed. Model pipes were set ‘vertically in their test attitude before placing sand around them, thus enabling a uniformly dense sample to he obtained. ‘The sand pouring technique was chosen after discussions withresearch colleagues at Oxford University whohave worked withthe same sand poured into varioustestchambers. Reference ‘was made to Bieganousky and Marcuson (1976) and to seminar notes of a Calibré ‘Chamber Conference at Southampron University, Last (1983). The sand was poured through a steel plate perforated with 6mm holes on 290mm grid. Relow the steel plate were. two 4mum sieve grids to act as diffusers of the jets of sand. The sieves were maintained at least ‘550mm above the sand surface as shown in Figure 3.2 and adense sand sample was produced. Apparatus Design 3-3 Apparatus Design 3-4 ‘Tests on the pouring technique indicated a sample with a relative density of 85%. It was necessary to include a 2mm stainless steel wire sieve mesh above the perforated plate. This enabled larger pieces of microconcrete contaminating the sand to be removed. < 2mm SIEVE | at PERFORATED PLATE. 4mm SIEVE 34 Z| 2 z — E ie Cc —_ = hana (ALY, +200) ‘otal horizontal displacement between ends of the pipe u AL, = (ALH, ~ALH,) ‘Total vertical displacement between ends of the pipe ALV,= Th (ALV, ALY) ‘Total diagonal length between extreme ends of the pipes D =\(ALH,) + (ALV;) + (440)?) Deflection angle between consecutive pipes 24 pt4p? B =cos"| 2 spiel) Eq 5.1 2D,D; Analysis and visual inspection of the early deflected pipe tests indicated that sand was entering the gap between pipe ends, preventing full realignment of the pipes. This resulted in axial load transfer at positions away from the theoretical contact point and crushing of sand grains. Some spalling of concrete (shown in Plate 5.2) at the areas of sand penetration was visually evident. Whilet this would be true modelling of the situation at full size in granular materials, it was considered necessary to mit sand ingress into the pipe joint gap. This was satisfactorily achieved With the use of strips of masking tape. Sand Chamber Tests 5-12 Plate 5.2 Spalling at common plain end of pipes (Test 14). Results of chan, ¢ in pipe alignment are presented in graphic form by plotting either angular deflection or location of point of contact against applied axial load. Graphs from three tests are presented in Figure 5.5. The tests had the same initial deflection angle of 0.5° horizontally and boundary stress conditions of 5OkPa vertical and SOkPa horizontal and the data were taken from the results of Tests 20, 25 and 29. The only intended difference between the tests was the type of pipe being tested; Test 20 was on a thin walled, plain ended pipe; Test 25 was on a thick walled, in wall juinted pipe and Test 29 was uu a sicel cullaved, thin walled pips. all pipes weie identically reinforced The plot in Figure 5.52 shows how all the pipes attempt to align themselves but as alignment is cccurring so Figure 5.5b shows that the point of contact between the pipes is changing. Test 29 maintains almost the same point of contact during the test but does not fully align itself, possibly Sand Chamber Tests 5-13 due to the ingress of sand into the joint or the ends of one of the pipes not being square. The probability of non square ends is also shown due to the larger than intended initial deflection angle on this test. 09 FOR ALL TESTS: INTENDED INTIAL DEFLECTION - 05° VERTICAL BOUNDARY STRESS 50 KPa HORIZONTAL BOUNDARY STRESS 5OkPa wl erennoe eee : reer nese saan 8 a " pees ron Tears 2rd 25 WERE Tan i : eer eel & oe | mate be je a i z | i ree wee ° * ‘0 150 20 LOAD ga) Figure .5a Alignment of pipes. Pipes used in Test 20 do nor fully align but the position of contact changes from the left hand side to the top of the pipes. The pipes from Test 25 align themselves and the position of contact between the pipes changes from the left side of the pipes to the right side quite suddenly at about TOKN of applied axial load. One problem noted throughout this test series was that pipes were aligning themselves or reaching ‘ point when no further change of alignment was occurring, but the pipes had not visibly tailed. The rescarch was aimed at asscssing jacking loads of misaligned pipes and investigating the different failure modes. This is not possible ifthe pipes are always moving to an ‘aligned! position. Sand Chamber Tests 5-14 g 8 LOCATON OF POINT OF CONTACT (Gageees) a 100 POINTOF 0 Joowract Ao ‘VIEWED FROM END OF LOAD ° ° 50 100 180 20 LOAD 0) Figure 5.5b Location of point of contact between pipes. ‘Throughout this test series it proved difficult to compare tests. Even when initial conditions were identical, pipes moved differontly during the tost and it is therefore important to asscss the ues dimensional orientation changes of consecutive pipes before attempting to interpret test results. 5.4.2 Joint gap ‘The previous assessment of pipe deflection was measured relative to the model pipes’ internal surfaces. This method of analysis takes no account of manufacturing tolerances and the possibility of pipe ends not being square relative to the internal surface. To obtain an assessment of pipe end squareness, three LVDT’s were mounted spanning the common joint between the two pipes. ‘The LVDT"s had their zero datum readings taken in free air with one pipe placed on top of the other. Changes in the measured joint gap at the start of testing and during loading would enable analysis to be made of the contact point relative tothe pipe end squareness. This can be compared. to the deflections obtained from the alignment LVDT’s. Sand Chamber Tests 5-15 ‘Measurements of the changes af joint gap hetween the pipes for Tests 20, 25 and 20 are presented in Figure 5.6 and can be compared to the deflection data shown in Figure 5.5. Joint gaps recorded during Test 20 all tended towards a reading of -0.2mm compared with the joint gap ‘zero' reading taken when no axial load was applied. The gap was uniform around the pipes’ circumference and the point of contact was recorded at the top of the pipes. Before the application of axial load the pipes had a joint gap difference of 0.9mm which agreed with tho previously calculated angle of deflection of 0.52° between pipes. Joint gap values recorded during Test 25 indicated that the gap on the right of the pipe became much less than that on the left, This can be compared to the data recorded of the points of contact which also moved from the left to the right of the pipe. However, data recorded during Test 29 show the joint gap close to the bottom ot the pipe tobe the only position at which the gap is changing by asignificantamount. ‘During this teat the point of contact remained on the left side of the pipe. There appears 19 be 10 logical explanation for the changes in the joint gaps recorded. Sand Chamber Tests 5-16 Test 29 . Figure 5.6 Variation of joint gaps. Sand Chamber Tests 5-17 5.4.3 Pressure cells ‘The pressure cells were clamped in the chamber wall with their measuring diaphragme flush with the inside surface of the wall. In this test series they were mounted in a straight line vertically below the centre line of the pipes and at 100mm centres. Using results from Test 25 the pressure distribution shown in Figure 5.7 is observed. 35 RESULTS FROM TEST 25 20 mir = ° 5 = pe od § 2 TT 7 T oo 008 oe os 012 LOADED ANCA [EFFECTIVE AREA Figure 7.4 Plot of ultimate bearing stress variations (Williams 1979). Joint Packing 7-9 7.3 Results Examination of the initial eyotic loading test results revealed several points of interest. Tt was clear that packing materials were being permanently deformed (Plates 7.2 and 7.3) and the material properties were totally different after one load application had been performed compared with the properties of the material before testing began. Consideration of the conditions in which the packing material would be used led to all analysis being carried out on results recorded subsequent to the application of the turst load. Plate 7.2. Typical joint packing materials. Some of the initial materials tested were immediately ruled ont of the suheeqnent test prow ‘gramme. 18mm thick planed timber compressed along the direction of its grains rather than Joint Packing 7-10 Plate 7.3. Typical joint packing materials. in the direction of load application which led to dieplacement of one concrete cube relative to the other. The timber was chosen as knot free and was not considered to be a viable proposition for a packing material. Blockboard disintegrated on application of load and was not tested further; hardboard was obtained in 3mm thickness only and was not compressible; 12mm softboard compressed to 4.5mm and thea rescubled dense fibucbuaid. Guuxl oue side shuticring ply wood compressed nevenly dne to many gaps in the ply as it had heen laminated. Further testing and analysis, was carried out on dense fibreboard, plywood and chipboard. Joint Packing 7-11 7.3.1 Stress/strain relationship ‘Data have been processed to give details of the stress and etrain conditions induced in the packing material. Stress is obtained by dividing the load applied to the packing material by the area of material under compression and load is assumed to be distributed uniformly. ‘Strainis the change in thickness ofthe material since it was first loaded divided by its original thickness. Plots similar to those presented in Figure 7.5 were obtained, but these plots do ‘not present the information in a suttable format, DENSE FIBREBOARD Yammy 18pm , so | LOADAND UNLOAD CYCLES N's 1 AND 952 1 : / . Lo LE a LL ° 2 ” sTaAn (4) EAN STRESS (Nine) Figure 7.5. Stesy/strain plot for dense fibreboard. ‘The graphs were re-drawn to plot strain as a change in thickness during the current load cycle divided by thickness of packing material atthe start of the cycle. The new graphs are Joint Packing 7-12 presented in Figure 7.6, but thicker packing materials appear to be strained to a lesser extent than thin materials, yet are more compressible overall and therefore better at distributing stress concentrations. a We PLOTIS FOR APPLICATION OF LOAD (ON THE 362nd CYOLE Puvte ‘STRESS Nima) 8 12 16 2 'STRAIN(%) Figure 7.6 Stressftrain plos for 362 load cycles. Ifthe compressive properties of the packing material are the only consideration, the material with maximum compression is required. This is not evident from the strees/strain plote presented and the data are replotted as compression in millimetres during the current load cycle against induced stress in Figure 7.7. ‘The first ofthese plots shows clearly the greater compressibility of thicker packing materials. ‘The second plot shows the differences for similar thickness packing materials and depicts 18uun douse filrclvatd ay the most coupressibic iatorial. Joint Packing 7-13 (MEAN STRESS (Nite) EAN STRESS (Nin em FIBREBOARO ‘sfmm FIBREBOARD 4 ‘COMPRESSION (mm) ‘COMPRESSION ON 362nd LOAD APPLICATION (mm) Figure 7.7. Stress{compression plots for various packing materials. Joint Packing 7-14 7.3.2 Effect of cyclic loading ‘A graph plotting time. against load is presented in Fignre 7 and indicates the magnitnde. and duration of load applied during a typical joint packing material test. The effect of cyclic loading on a pipejack is important due to the method of pipe installation. Many of the pipes are loaded a large number of times during installation i.e. on a 100 metre long pipejack ‘without an interjack station the average number of load applications to the pipes is 200. If an intermediate Jacking station fs used it's Likely that pipes nearest to it will be subjected to an even larger number of load cycles and the largest magnitudes of load. ‘The tests on the packing materials showed how the material properties changed during the application of a large number of load cycles. A graph showing how the material is compressed and recovers its thickness to a lesser extent with each load application is presented in Figure 7.9. The behaviour was much more noticeable in tests on saturated materials due to drying out of the material during the test; the duration of cach test was fiftecn hours. Joint Packing 7-15 16 es 361 PYCLES € 0 sechs CYCLE = EHeaTS z END. 8 E 5 ane ° 1 00 1000 1500 sa (een) cyeNet 2 9 4 § 8 7 8 el sip sezhsCvCLE REEATS ge 5 i» | | w4 ° ‘TME (Ses) Figure 7.8. Duration of load applications for cyclic loading. Joint Packing 7-16 MEAN STRESS (Winn) MEAN STRESS (Nerf) 8 6 ‘8mm FIBREBOARD WITH MAXIMUM MEAN STRESS OF 15Wnn* COMPAECION (rm) 1 10mm FIBREBOARD WaT MAXIMUM MEAN STRESS OF su Y, oman A 4 a 2 ‘COMPRESSION (rem) Joint Packing 7-17 re 7.9. fect ot cyclic Loading on compression of dense fibreboard. During the test series different i I packing material thicknesses were tested and the dif. ference in their behaviour under compression loads can be seen in Figure 7.7 where more compression is shown to take place in thicker materials. It is not until the geometry of a jacking pipe is considered that the effects of different compressions of packing materials can be fully appreciated. ‘When considering pipes in a straight line with perfeetly aquaze ends relative to cach other, the packing material compression and hence its thickness have no effect on the distribution of stress concentrations because stress is uniform. It is not until pipes are considered whose ends are not square, even and aligned that the requirements of the packing material com- pression need be considered. It could easily be argued that this is the circumstance on all pipejack joints and that ideally square pipes never occur. Materials that are more. compressible when subjected to high etresses are of most henefit in assisting distribution of stress concentrations. The ideal scenario would be to find a material that resulted in equal stresses over the pipes’ end area whatever the magnitude of deflection between pipes or unevenness in pipe end profile. This is clearly impossible to achieve, so a ‘material that has a lot of compressibility in order to distribute stresses over as great an area as possible is required. To give some indication as to how deflected pipes atfect ine com- pressive requirements of the packing material Table 7.1 can be examined. The table detaile variations in pipe joint gaps related to angular deflections and further quotes typical measurements of horizontal line and/or vertical level on a prototype jack that would be associated with the particular deflection angle. When reading the table, the deviations quoted are maxima over the length of one pipe assuming the previous pipe is exactly on the proposed centreline at both ends and that the pipe ends are square to each other. Deviations for linc ut Ievel assutnc dhat the uther has uv deviation and figures for Tine and level assume both have equal deviations from the intended centreline. Joint Packing 7-18 Maximum intemal joint| ‘Equivalent Line and/or Level deviations Angular gap difference (mm) Deflection (mm) Line or Level only ‘Line and Level Degrees) | 150 900 1800] 232 1100 2440] 232 1100 2440 pipe pine pine | pipe pine _ pipe 0} 0.00 0.00 0.00] 0.00 0.00 0.00} 0.00 0.00 0.00 0.025) 0.07 0.39 0.79} 0.10 0.48 1.06] 0.07 034 0.75 0.05) 0.13 0.79 157] 020 096 213) 0.14 068 151 0.075] 0.20 1.18 236] 030 144 3.19} 0.21 102 226 Ol} 0.26 197 344] O40 L¥Z 4.26] 0.29 1.36 3.0L 015] 039 236 471/ 0.61 2988 6.39] 0.43 201 1.52 0.2| 052 3.14 628] 081 384 852] 057 272 6.02 03) 0.79 471 9.43] 121 5.76 12.78] 086 407 9.04 04] 105 6.28 1257 7.68 17.04] 115 543 12.05 05] 131 7.85 15.71 9.60 21.30] 143 6.79 15.05 075] 196 11.78 23.56] 3.04 1440 31.94] 215 10.18 2258 1) 2.62 15.71 9142] 405 1920 42.59] 286 1958 30.11 195] 327 196% 3997] $06 2400 5323] 45R 1697 37.65 15] 393 2356 47.12) 607 2880 6387| 4.29 2036 45.18 Table 7.1. Comparison between angular deflections between pipes and typical line and level surveys uu pipejacks. Joint Packing 7-19 7.3.4 Changes of material densities during saturation and drying ‘Whilst packing material was being saturated totally immersed in water. regular measurements ‘were being taken of the changes in thickness and weight occurring; the data are presented in Figure 73. ‘The difference between the materials’ properties when immersed in water can be seen by comparing German chipboard 10 chipboard. The chipboard readily absorbed warer and. became more dense than water whereas the German chipboard appeared to be more resilient to water absorption and never attained a greater density than water. Dense fibreboard and chipboardretained some oftheir increased thickness once they were removed from the water and dried. ‘The changes ot densities ot other materials tell between that ot dense fibreboard andGerman chipboard. Ordinary chipboard became frail whon saturated and began to disintegrate if it was not carefully handled. 7.3.5 Effect of material con¢ ions The test results were greatly influenced by the history of atmospheric conditions which the packing material had experienced. It was important to test the material in the various conditions to simulate possible treatment and conditions the material might be subjected to ‘on a constuction site, Itis possible that the packing material would be left in the open air, samrated in a pipejack with water mmnning in the invert or saturated by ground water on the external face of an in wall jointed pipe. The differences in the compressive properties of the material experienced during the tests are presented in Figure 7.10 where it can be seen how saturated materials are much more Joint Packing 7-20 camprescihle than materials which have not been treated! Packing material which hac heen saturated and then dried at room temperature returns to the same thickness when compressed as material that had not been saturated but has more thickness recovery when unloaded. ‘These behaviours were found to be the case for all materials tested and clearly demonstrate that the saturation of the packing material is not detrimental to the compressive behaviour or imegrity. If r were practically possible, saturation of packing matertal would be a distinct advantage in distributing jacking loads over larger areas. “ih VOLE He's 1 and 098 ANC PLOTTED FOR 12mm ORIGINAL THICKNESS DENSE FIBRESOARD fo ASSUPPLIED {edry SATURATED and ROOM DRIED fust SATURATED MEAN APPUED STRESS N/mm? ACTUAL MATERIAL THICKNESS (mm) Figure 7.10 Ettect ot matenal condition on compression characteristics. 7.3.6 Nifferences hetween materials ‘The materials disregarded at the early stage of this test series were presented in Section 73; they were generally degrading when compressed. Comments in this section will be about dense fibreboard, chipboard and plywood. Joint Packing 7-21 Analysis of the resnlte hae heen carried ant an data from the last load cycle applied and the compression of the packing material has been analysed to provide data on jacking load capacity Thesedata have been based on assumed maximum allowable stresses and have been calculated for packing material ona 1200mmexternal diameter and 900mm internal diameter pipe. The author assumes that the full end area of the pipe is used for load transfer and are resentedinFigure 7.11 and Figure 7.12 by comparing stess distributions for various angular deflections between pipes and relating these to total jacking loads. Itean be s 3 that as joint deflections increase so load is rapidly concentrated onto a small area of the end of the pipe and hence results in reductions in jacking load capacity. Distribution of stresses on the end of a pipe is the same for a given distance from the edge of the pipe imespective of pipe diameter, but increased end area results in larger jacking loads being acceptable. Consideration of joint profiles and retated packing matcrial is prosented in Chapter 9, It takes into account the pipes’ elasticity as this will increase the end area available for load transfer. At this stage it can be commented that dense fibreboard gives greatest capacity for trans- mitting jacking forces between pipes and for distributing concentrations of stress. Saturated packing material ls beneficial bur perhaps not practicable. In general, water enhances the ability of a packing material to dietribute load and the material is beneficial towards dis- tribution of loads on pipe ends which are not square. Joint Packing 7-22 JACKING LOAD GAPAGITY OF PACKING MATERIALS ON MISALIGNED 900mm INTERNAL DIAMETER PIPES A WITH MAXIMUM ALLOWABLE STRESS OF 15 N/mnit FRtaweT 0.1 Degrees 1.0 Degrees Distribution of Load at stated Angular Deflections, Figure 7.11. Jacking load capacity of a 900mm internal diameter pipe. Joint Packing 7-23 “suonsoyap reynfue snowea 10} uonnqinNsIp sang Z]"z aang (ww) x uomIsog or soo | og (see00) WILIWVIT RESIN 3g] VINO 00°90 = voseiieq seynduy sadiq uaamiaq suoNoayaq seinBuy snowen 10} spuz adi uO UONNqUIsIg SsalIs ;eoIdA | 6 sseuig Joint Packing 7-24 Effects of packing material properties on deflected pipe joints should be related to deflections ‘measuredrelative tothe pipe end squareness. This gives the best indication of the magnitude of joint gap the packing material must bridge in order to transfer load. Squareness of the pipe end is thus a very important factor in considering a pipe’s jacking capacity. It should bbe noted that pipes complying to end squareness tolerances of the British Standard could ‘have ayoint gap of up to 8mm ona YW0mm internal diameter pipe. his equates toadeflection of 0.5 degrees even before pipes arc deflected relative to cach other. The British Standard also allows for pipe squareness to be 4mm across the wall thickness as shown in Figure 7.13. DIMENSIONS ARE APPLICABLE TOA 900mm INTERNAL DIAMETER PIPE: j__————————— 4mm SQUARENESS —{ ToLeRaNce Zay} ky -} arm TOLERANCE dan ‘rym SQUARENESS. FRE INCLINED IN J] TOLERANCE EITHER DIRECTION, REA PICTORIAL REPRESENTATION OF BRITISH STANDARD SQUARENESS. TOLERANCES AND HOW IT IS POSSIBLE FOR THEM TO MAGNIFY DEFLECTION ANGLES BETWEEN CONSECUTIVE PIPES. Figure 7.13 Conflict between British Standard squareness tolerances. Joint Packing 7-25 Materials chosen to best suit these compressive requirements have been listed in Table 7.2 with the most compressible, dense fibreboards, appearing at the top of the table. The table is calculated with all materials allowed the same maximum stress and lists materials in order of their ability to transmit axial load with a deflection of 0.2 degrees at the joint. Almost the same order would appear in the table independent of stress level or deflection angle. Joint Packing 7-26 MATERIAL | _LOAD CAPACITY OF PACKING MATERIAL (Tonnes) "ANGULAR DEFLECTION (Degrees) 03 05 0.7 0 0.025 0.05 OL 02 75 FBI1850 2572 1738 1189 689 «374245129 CB1830 2572 1613 1082 588 «310 196 «993335 FBISWET | 766 685 567 426 269 «187 119 78 56 FB1250 2572 1523938 $23, 266 162, 79 AQ 28 CBISWET | 766 629 516 384 239-166 106849 CVBWET 766 636 «522,303.56 177,ss2—s—s7S SH FBIDWET | 766 627 479 317 177 1233 72 44 29 PLY1850__ | 2572 1371 820 451-222-132, «63.33 22. PLYI8WET 766 «603462307 169 18 69 41027 FBI2W 766 595 447 _—«295_—*163_—sd113;_—S S398, PLY IZWEL Joo 38H 444 LYS 102 Az o> jy FBI2WET3 | 766 577 439 291 161 111 66 39 26 FBI2WET 766 «575437 290 160 11 65 39 26 CBI2WET | 766 577 434 285 157 10963385 PLY650 2572 974 «547280114 isis 10 CvBcHIP | “766 515 359 216 116 «76 «6412315 FBI815 766 545 «(380-214 «118. S77 40 CVBDRY 766 S515) 337 = 208 13 15 40 23 4 FB650 2572 912 «506-257 «108-28 CVBCHIP2 766 S11 356 200 113 75 AL 2 ou cBIaDRY | 766 494 328 187 99 Gf 32 17 Li Dew 76 «474 «9 18397 SB FBOWET 166 486 «317 «18095 BLT CBIRIS 766 «482305 170 on 55 27 14 9 PLYI8DRY 76 «415-263 143 2 44 22 u a PLY1815 76 421-258 «14271 #2. CBIZDRY I 409 B13)’ YT FBI2DRY | 766 428 247 138 66 38 18 9 6 FBI2I5 766 407 «234130 sia“ (ssa cBI2I5 766388221122, 5B 338— 5S BS PLYGWET 1766-381 215 119 ‘56 32 15 8 3 FBODRY 176 ae 92 es 180 pent 5 es 0 mee 2 ees LO eee Se) R615 66 36 «1753 PLyi2pRY | 766 307 170 89 39 2 0 5 3 SOFTWET | 766 272 14 76 32 18 8&8 4 2 PLYODRY | 766 238 132 64 24 13 6 3 2 SOFTDRY 766 = 229 124 61 24 13, 6 i 2 PLYGIS 76 29 127 GL NOTE + All materials treed to maxi- mum stress level of 15 N/mm’ unless yareRIAL TYPE po ‘marked with "*". Those marked with the fo cvs rarcuent APPLIED STRESS asterisk were tested to 50 N/mm. ratcueed (eum) All data are calculated for a 900mm inter- pry

You might also like