PILE DRIVING ANALYSIS STATE OF THE ART by Lee Leon Lowery, Jr. Associate Research Engineer T. J. Hirsch Research Engineer Thomas C. Edwards Assistant Research Engineer Harry M. Coyle Associate Research Engineer Charles H. Samson, Jr. Research Engineer Research Report 33-13 (Final) Research Study No. 25-62-33 Piling Behavior Sponsored by ‘The Texas Highway Department in cooperation with the U.S. Department of Transportation, Federal Highway Administration Bureau of Public Roads January 1969 TEXAS TRANSPORTATION INSTITUTE Texas A&M University College Station, TexasForeword ‘The information contained herein was developed on the Research Study 25.62.33 entitled "Piling Behavior” which isa cooperative reearch endeavor sponsored joint bythe Texas Highway. Department and the U. S. Department. of Transportation, Federal Highway Administration, Bureau of Public Roads, and also by the authors as evidenced by the number of publications during the past seven years of inense uly: and. research, "The broad’ objective of the project was to filly dovelop the ompuler solution of-the wave equation and is use for pile driving analysis to dtermine ‘values for Whe significant parameters involved to enable engineers. to prodict driving siesses in piling during driving, and Yo estimate the static soi resi fee to penetration on piling atthe time of driving from driving resistance record. The opinions, findings. and conclusions expressed in this report are those of the authors and not necessarily those of the Bureau of Public Roads.Acknowledgments Since this report is intended to summarize the research efforts and experience zained by the authors over a seven-year period, it is impossible to mention all of the Persons, companies, and agencies without whose cooperation and support no “state of the art” in the analysis of piling hy the wave equation would exist ‘The greatest debt of gratitude is due Mr. E. A. L, Smith, Chief Mechanical Engi neer of Raymond International (now retired), who not only fist proposed the methed af analysis but also maintained continuing interest throughout the work and cone tributed significantly to the accomplishments of the Feenrch. His advice and guid ace, based on his extensive field experience, and hie intimate knowledge of te wave tvquation have proven invaluable throughout all phases of the research ‘The authors gratefully acknowledge the assistance and support of Farland C. Bundy and Wayne Henneberger of the Bridge Division of the Texas Highway De. partment, who worked closely with the authors in accomplishing several of the proj: ects, A debt of gratitude is due the Bass Bros. Concrete Company of Victoria, Ross Anglin and Son, General Contractors, and the California Company of New Orleans for their unselfish cooperation with all phases of the field work. It was indeed fortunate to have these foresighted and pri the various jobs. ‘They willingly invested considerable amounts of theit own time and effort to the accomplishment of the research projects, Sincere thanks and our personal appreciation are extended tothe Texas High Department and the U, 8. Department of Transportation, Federal Highway Adiniai traion, Bureau of Public Roads, whose sponsorship made all the research and studies leading to this report a reality. Several graduate students and research assistants contributed significantly to the accomplishment of this work. ‘They were Tom P. Aithart, Gary N Reeves, Paul C'Ghan, Can F: Raba, Gary Gibson, I H, Sulaiman, james I Finley, M1 and Johin Miler. iTABLE OF CONTENTS. LIST OF FIGURES - LIST OF TABLES CHAPTER 1 INTRODUCTION T_ PILE DRIVING ANALYSIS. 21 General 22. Smith’s Numerical Solution of the : 23 Critical Time Interval 24 Effect of Gravity.. Hl PILE. DRIVING HAMMERS _ 3.1 Energy Output of Impact Hammers 3.2 Determination of Hammer Enerzy Output. 3.3. Significance of Driving Accessories... 3.4 Explosive Pressure in Diesel Hammers 3.5. Effect of Ram Elasticity - IV CAPBLOCKS AND CUSHION BLOCKS ...... 4.1 Method Used to Determine Capblock and Cushion 4.2, Iealized Load-Deformation Properties 4.3. Coefficient of Restitution V_ STRESS WAVES IN PILING 1. Comparison with Laboratory Experiments 5.2 Significance of Material Damping in the Pile VI SOIL PROPERTIES... 6.1 General — 62. Equations to Describe Soil Behavior 63 Soil Parameters to Describe Dynamic Soil Resistance During Pile Driving 64 Laboratory Tests on Sands 65. Static Soil Resistance After Pile Driving (Time Effect) - 6.6 Field Test in Clay... - VII USE OF THE WAVE EQUATION TO PREDICT PILE LOAD BEARING CAPACITY AT TIME OF DRIVING... 21 7.1 Introduction — 7.2 Wave Equation Method 73 Comparison of Predictions with Field Tests VIII PREDICTION OF DRIVING STRESSES... B1_ Introduction — 8.2 Comparison of Smith’s Numerical Solution with the Classical Solutio 83 Correlation of Smith's Solution with Field Measurements 84 Effect of Hammer Type and Simulation Method 85. Effect of Soil Resistance 86 Effects of Cushion Stiffness, Coefficient of Restitution, and B 87 Fundamental Driving Stress Considerati 88 Summary of Fundamental Driving Stress Considerations — IX USE OF THE WAVE EQUATION FOR PARAMETER STUDIES. 9.1 Introduction 9.2. Significant Parameters 9.3 Examples of Parameter Studies X SUMMARY AND CONCLUSIONS... Properties .—..APPENDIX A. DEVELOPMENT OF EQUATIONS FOR IMPACT STRESS! IN A LONG, SLENDER, ELASTIC PILE... AL Introd A2_ One Dimensior A3_ Boundary Conditions Ad Solving the 'B ferential Equation AS. Maximum Compressive Stress at the Head of the Pile AG Length of the Stress Wave APPENDIX B. ee : WAVE EQUATION COMPUTER PROGRAM UTILIZATION MA BI B2 Idealization of Hammers. B3. Ram Kinetic Energies Bi Method of Including Coefficient of Restitution in Capblod BS Idealization of Piles — B6 Explanation of Data Input Sheets B7_ Comments on Data Input APPENDIX C - 08/360 FORTRAN IV PROGRAM STATEMENTS. LIST OF REFERENCES. ushion Springs SSASSSHSSSSSSHSSRSS LIST OF FIGURES Figure 21 Method of Representing Pile for Purpose of Calculation (After Smith) 22 Load Deformation Relationships for Internal Springs 23 Load Deformation Characteristics Assumed for Soil Spring M 3.1 Typical Force vs Time Curve for a Diesel Hammer 3.2 Steam Hamme 33 Diesel Hammer 4.1 Stress Strain Curve for # Cushion Block - 4.2 Dynamic and Static Stress Strain Curves for a Fir Cushion . 43° Cushion Test Stand - 44 45 Dynamic Stress Strain Curve for a Micarta Cushion ~ 4.6 Stress vs Strain for Garlock Ashestos Cushion ‘Theoretical vs Experimental Solution ‘Theoretical vs Experimental Sol Theoretical vs Experimental Solution — ‘Theoretical vs Experimental Solution Comparison of Experimental and Theoretical Solutions for Stresee Model Used by Smith to Describe Soil Resistance on Pile. Load Deformation Characteristics of Soil. Load Deformation Properties of Ottawa Sand Determined by Triaxial Tests (Specimens Nominally 3 in. in Diameter by 6.5 in. High) Increase in Strength vs Rate of Loading—Outawa Sand 3” vs “V" for Ottawa Sand. 7 “Set up” or Recovery of Strength After Driving in Cohesive Soil (Ater Pore Pressure Measurements in Clay Stratun—50 fe Depth Uhimate Driving Resistance vs Blows per Inch for an Example Problem Comparison of Wave Equation Predicted Soil Resistance to Soil Resistance Determined by Load Test for Piles Driven in Sands (Data from Table 7.1) 7.3 Comparison of Wave Equation Predicted Soil Resistance to Soil Page sistance Determined 74 Comparison of Wave by Load Tests for16 al a2 83. a4 as 86 a7 8B 91 92 93 on 12 13 al 82 Ba Be Summary of Piles Tested to Failure in Sands (From Reference 72) Wave Equation Ultimate Resistance vs Test Lond Failure (From Reference Moximam ‘Tensile Steess Along a Pile Maximum Compressive Stress long a Pile Stress in Pile Head vs Time for Test Pile Stress at Midength of Pile v# Time for Test Pile Iealized Stress Wave Produesd When Ram Strikes Cushion at Reflection of Stress Wave at Point of « Long Pile - Reflection of Suess Wave at Point of a Short Pile Effect of Ratio of Siress Wave Length on Maximum Tensile Stress for Pile with Point Free Effect of Cushion Stiffness. Ram Weight, and Driving Energy on Stresses Effect of Cushion Stiffness, Ram Weiaht, and Driving Energy on Permanent S Computer Analysis of Pile Hammer Effectiveness in Overcoming Soil Resistance, Ru, When Driving a Pile Under Varying Conditions Evaluation of Vulean 020 Hammer (60,000 ftlb) for Dri Capacity of 2000 Tons Blows/In. vs RU (Total) for Arkaneas Load Test Pile, Long Slender Elastic Pile Ram, Cushion, Pil Case 1—Ram, Capblock. andl Pile Cap. Case Ram, Caphlock, Pile Cap, and Cushion Case I11.—Ram, Anvil, Capblock, and Pile Cap Defini Pile Idealization Example Problem Normal Output (Option 15=1) for Effect of Varying Cushion Stiffness Summary Output for RU (Total) vs Blows In. (Optio Normal Output for Single RU (Totals (Option 11 Detailed Output for Single RU (Total) (Option 1 n of Coefficient of Restitution ob, 1 for Prob. ) for Prob. LIST OF TABLES Summary of Hammer Properties in Operating Characteristics Effect of Cushion Stiffness on ENTHIRU. Effect of Cushion Stiffness on FMAX _ _ Effect of Cushion Stiffness on LIMSET a - Effect of Remoring Load Cell on ENTHRU. LIMSET, and Permanent Set of Pile... Effect of Coefficient of Restitution on ENTHRU (Maximum Point Displacement) Effect of Coefficient of Restitution on ENTHRU. Effect of Breaking Ram Into Segments When Ram Strikes a Cushion or Capblock — Effect of Breaking Ram Into Segments When Ram Strikes « Steel Anvil ical Secant Moduli of Elasticity (E) and Coefficients of Restitution (e} af Various Pile Cushioning Materials = Errors Caused by Assuming J(pointy = 0.1 and J Vide) = Jipoint) for Sand-Supported Piles Only ee Errors Caused by Assuming J(point) = 0.3 and J" (side) Loin) for Clay (for Clay-Supported Pites Only) - - a Errors Caused by Aesuming a Combined Jipoin!) = 01 for Sand and Jtpoint) for Clay Using Equation 7.1. (for Piles Suppwnted by Both Sand and Clay) Variation of Driving Stress With Ram Weight and. Veloc Variation of Driving Stess With Ram Weight nnd Ram Bnergs..- Drop Hammers and Steam Haniets..—_.- 7 Diesel HammersPile Driving Analysis-State of the Art CHAPTER 1 Introduction The tremendous increase in the use of piles in both Jandbased and offshore foundation structures and the ap: pearance of new pile driving methods have created great engineering interest in finding more reliable methods for the analysis and design of piles. Ever since Isaac published his paper. “Reinforced Concrete Pile Formula,” in 193127" it has been recognized that the behavior of piling during driving does not follow the simple Newtonian impact as assumed hy many simplified pile driving formulas, but rather is governed by the one Gimensional wave equation. Unfortunately, an exact mathematical solution to the wave equation ‘was not possible for most practical pile driving problems In 1950, E. A. L. Smith? developed « tractable solution to the wave equation which could be used to solve extremely complex pile driving problems. The solution was based on a discrete element idealization of the actual hammer-pile-soil system coupled with the use of a high speed digital computer. In a paper published in 1960, * he dealt exclusively with the application of wave theory to the investigation of the dynamic behavior of piling during driving. From that time to the present the authors have engaged in research dealing: with wave equation analysis. The major objectives of these studies were as follows: 1, To develop a computer program based upon a procedure developed by Smith to provide the engineer with a mathematical tool with which to investigate the behavior of a pile during driving. 2. To conduct field tests to obtain experimental data with which to correlate the theoretical solution. 3, To make an orderly theoretical computer invest ‘Numerical superscripts refer to corresponding items in the References, gation of the influence of various parameters on the Jehavior of piles during driving and to present the re. sults in the form of charts, diagrams or tables for direct application by office design engineers, 4. To present recommendations concerning good driving practices which would prevent cracking “and. spalling of prestressed concrete piles during driving. 5. To determine the dynamic load-teformation properties of various pile cushion materials. which had tacitly been assumed linear. 6. To determine tke dynamic load-leformation Properties of soils required by the wave equation analysis. 7. To generalize Smith's original method of analy- sis and to develop the full potential of the solution by using: the most recent and accurate parameter values determined experimentally 8. To illustrate the significance of the parameters involved, such as the stiffness and coefficient of restitu: tion of the cushion, ram velocity, matetial damping. in the pile, etc., and to determine the quantitative effect ‘of these parameters where possible 9. To study and if possible evaluate the actual energy output for various pile driving hammers, the magnitudes of which were subject to much disagree- 10. To develop the computer solution for the wave equation so that it may be used to estimate the resistance to penetration of piling at the time of driving from the driving records, 11. To develop a comprehensive users manual for the final computer program to enable its use by others. CHAPTER IT Pile Driving Analysis 2.1 General Th mi iperenine oof ile funtion the appearance’ of neve pile driving. techniques. have tauned preat Interest among engineers in finding nore feliahle'methods of pile anahysie and design. As noted by Dunham?" "A plle driving formula ian attempt to evaluate the resistance of pile to the dynamic fares applied wpon it during the ding and to estimate. from the the statical Longtail load that the pile can su port safely as a part of the permanent substrucure™” In 1851, Sanders (Army Corps of Engineers) pro: hosed the first dy namie pile driving Formula by equating the total enery of the ram at the instant of impact to the work done in foreing down the pile, that is, the product, of the weight of the ram and the stroke was assumed equal to the product of the ultimate soil resistance. by the distance through which the pile moved. Sanders applied a safely factor of 8 to this ultimate soil resist. ance to determine an assumed safe load capacity for the pile. Since that time, a multitude of formulas have been proposed, some of which are semirational, others beingstrictly empirical. Many of the formulas proposed at- Temp! to account for various impact losses through the cushion, eapblock, pile, and soil. When restricted to @ particular soil pile, and driv. ing condition for which correlation factors were derived, dynamic formulas are often able to predict ultimate Dearing capacities which agree with observed test loads, However, since several hundred pile driving formulas hhave ‘been proposed there is usually the problem of choosing an. appropriate or suitable one2*. Also dis iressing is the fact that in many cases no dynamic for- ula yields acceptable results; for example, long. heavy piles generally. show much greate Toads than predicted by pile driving equation is has become Increasingly. siznificant since prestressed concrete piles 172 ft long and 54 in. in diameter have been successfully driven? and more and more large diameter steel piles several hundred feet long are being used in offshore platforms, Numerous field tests have shown that the tse of pile driving formulas may well lead to a foun- dation design ranging from wasteful to dangerous? + Driving stresses are also of major importance in the design of piles, yet comprescive stresses are com. monly determined simply by dividing the ultimate driv. ing resistance by the cross-sectional area of the pile.??2* Furthermore. conventional pile driving analyses are un- able to calculate tensile stresses, which are of the utmost importance in the driving of precast or prestressed con: crete piles This method of stress analysis completely overlooks the true nature of the problem and computed stresses almost never agree with experimental. val- ues Tensile failures of piles have been noted fon numerous occasions?"2""="" and the absence of reliable method of stress analysis has proven to be @ serious problem, Although most engineers today realize that pile driving formalas have serious limitations and cannot be depended upon to give accurate results, they are still used for lack of an adequate substitute. For further discussion of pile formulas in general, the reader is referred to the work of Chellis?> Isaacs is thought to have first pointed out the occurrence of wave action in piling during driving, He proposed a solution to the wave equation assuming that the point of the pile was fixed and that side resistance ‘was absent. These assumptions were so restrictive that the solution was probably never used in practice, Cum- ast in_an earlier writing noted that althouzh the pile driving formulas were based on numerous erroneous assumptions and that only the wave equation could be expected to yield accurate results for all driving con ditions, he also pointed out that such solutions invelved long ‘and complicated! mathematical expressions 50 that thei use for practical problems woud involve Jaborous, numerical caleulations.”” In. fact, with. the advent of a multitule of different type driving hammers and driving conditions. an exact solution to the wave equation was not known, 2.2 Smith's Numer Wave Equation In 1950, Smith®# proposed @ more realistic solution to the problem of longitudinal impact. This solution is I Solution of the + | stroxe i lee ‘) eee Figure 2.1. Method of representing pile jor purpose of analysis (after. Smith). based on dividing the distributed mass of the pile into a number of concentrated weights Wi11 through Wp), which are connected by weizhtless springs K(1) through Kip—1), with the addition of soil resistance acting on the masses, as illustrated in Figure 2.1(b). Time is also divided into small increments. Smith's proposed solution involved the idealization of the actual continuous pile shown in Figure 2.1(a), as a series of weights and springs as shown in Figure 2A (bi. For the idealized system he set up a series of ‘equations of motion in the form of finite difference tions which were easily. solved using high-speed computers. Smith extended his original method of analy- sis to include various nonlineal parameters such as elasto- plastic soil resistance including velocity damping and others. Figure 21 illustrates the idealization of the pile system ‘sujgesied hy SmPh, Tn general. the system is considered to be composed of (see Figure 21(a)}: 1, Aram, to which an initial velocity is imparted by the pile drivers 2. A capblock (cushioning material) 3. A pile caps 4. A cushion block (cushioning material) ; 5. A piles and 6. The supporting medium, or soil eesIn Figure 2.11b) are shown the idealizations for the various components of the actual pile, The ram, cap: lock. pile cap, cushion block. and pile are pictured as appropriate diserete weights and springs. The frictional soil resistance on the side of the pile is represented by a series of side springs; the point resistance is accounted ‘a single spring at the point of the pile. The char- istics of these various components will be discussed in greater detail later in this report. Actual situations may deviate from that illustrated in Figure 2.1. For example, a cushion block may not be used or an anvil may he placed between the ram and capblock. However, such cases are readily accommo: dated, Internal Springs. ‘The ratn. capblock. pile cap. and cushion block may" in general le considered to. consiet, of “internal” springs, although in the representation of, Figure 2.1(b! the ram and the pile cap are assumed Figid (a reasonable assumption’ formany” practical cases) Figures 2.2(a) and 2.2(b) suggest different possi bilities for representing. the load-deformation character. ics of the internal springs. In Figure 2.2(a), the rial is considered to experience no internal damping jgure 2.211) the material is assumed to have inter. In nal damping according to tke linear relationship shown. External Springs. The resistance to dynamic load ing afforded bythe sil in shear along the outer surface of tele an ving a the pinto the pe tstzemely complex. Figure 23 shows the Ioad-dtorma 1 Figure 2.2. Load-dejormation relationships for internal springs. 1 Figure 23. Load-deformation characteristics assumed for soil spring m. tion characteristics assumed for the soil in Smith's cedure, exclusive of damping effects. ‘The path OABC- DEFG ‘represents loading and unloading in side friction For the point, only compressive loading max take place and the loading and unloading path would be along OABCF. It is seen that the characteristics of Figure 2.3 are defined essentially by the quantities “Q” and “Ru.” “Q" is termed the soil quake and represents the maxi- mum deformation which may occur elastically. “Ru” is the ultimate ground resistance, ot the load al which the soil spring behaves purely: plastically A loaddeformation diagram of the type in Figure 25 may be established separately for each spring, Thus, K'fm) “equals Rum) divided’ by Qim), where’ K'tm} is the spring constant (Guring elastic deformation) for external spring in. Basic Equations. Equations (23) through (2.7) wore developed by Smith. Dim.) = D(mt—1) + 12M Vma—1) (2.3) Cimt) = Dima) — Dom+1,) (24) Fim) = C(m.t) Kim) (25) Rima) = (Dima) ~ Dima) Kem) 1 + Jim) Vimt—i)) (2.6) Vemma) = Vimi—1) + [Flm=14) — Fim) = Rn. gt (23) wore (= functional designation; rm = element number: = number of time interval At = size of ime interval (see); Clomt) = compression of internal spring min time interval t (in.);Dim.) = displacement of element m in time interval 1 Gin.) D'im,t) = plastic displacement of external soil eprin m in time interval U tints Fim.) = force in internal spring m in time interval Ib); = acceleration due to gravity (ft/sec); J{m) = damping constant of soil at element m (see/ft); K(m) = spring constant associated with internal spring m (Ib/‘in.) K’(m) = spring constant associated with external soil spring m tUbyin.t; Rim.) = force exerted by external spring m on element m in time interval t (Ub); Vint) = velocity of clement m in time interval & (ft/sec); and Wom) = weight of element m (Ib). ‘This notation differs slightly from that used by Smith. Also, Smith restricts tke soil damping constant J to to values, one for the point of the pile in bearing, land one for the side of the pile in fri While the present knowledge of damping behavior of soils perhaps does not justify: greater refinement, it is reasonable 10 use this notation as a function of m for the sake of generality. The use of a spring constant Kim) implies a load- deformation behavior of the sort shown in Figure 2.2(a). For this situation, Ktun) is the slope of the straight line. Smith develops special relationships to account for in ternal damping in the eapblock and the eushion block. He obtains instead of Equation 2.5 the following equa = | Km) ClA) a 28) Clm,t)maz = temporary maximum value of Cum.t). With reference to Figure 2.1, Equation (2.8) would he applicable inthe calculation’ of the forces in internal spr = 1 aml m = 2. ‘The load-deformation relationship characterized by Equation (2.1) is illustvated by the path OABCDEO in Figure 221b). For a pile cap or a cushion block, no tensile forces ean exist: cose sequently, only this part of the diagtam applies. Inter mittent unloadingloading is typified by the path ABC. fetablished by control of the: quantity Clint) nay in Equation (2.8), ‘The slopes of lines AB, BC. and DE depend upon the coofficient of restitution e(m) ‘The computations proceed as fellows: 1, The initial velocity of the ram is determined from the properties of the pile driver. Other time. dependent quantities are initialized at zero or to satisfy static equilibrium conditions, 2. Displacements Dim,1) are ealeulated by Eq tion (2.3). It is to be noted that V(1,0) is the initial velocity of the ram, 3. Compressions C(m,1) are calculated by Equa: tion (2.4) 4, Internal spring forces Fim,1) are calculated by Equation (2.5) oF Equation (2.8) as appropriate 5. External spring forces R(m,l) are ealeulated by Equation (26) 6. Velocities V(m.1) are calculated by Equation (27) 7. The cycle is repeated for successive time inter- vale In Equation (2.6), the plastic deformation D' (ma) for a given extemal sping follows Figure. (28) and may he determined hy special routines For example, ‘en Bima) is less thant Qn), D(a). tetos when Dm) iv gteater than Qin) long tng AD see Figure 23), Ditmnt) is equal to Dlont) Qin Smith notes that Equation (2.6) produces no damp- ing when Dim,t) — D'(m,t) becomes zero. He sug- fests an alternate equation to be used afler Dim,t) first ecomes equal to Qn): Rom) = (Dom) — D'omay] R'm) + Jim) K'(m) Qt) Vims—2) (29) Care must he used to satisfy conditions at the head and point of the pile. Consider Equation (2.5). When D, where p is number of the last element of the Kip) must be set equal to zero since there is no Pip) Force (see Figure 1.1). Beneath the point of the pile, the soil spring must be prevented from exerting tension on the pile point. In applying Equation (2.7) to the ram (m = 1), one should set F(0,t) equal to zero, For the idealization of Figure 21, it is apparent that the spring associated with K(2) represents both the cushion block and the top element of the pile. Its spring ale may he obtained by the following equatie (210) a KO) one A more complete discussion of digits] computer programming details and recommenced values for varie ous physical quantities are given in 0 Appendices, From the point af view of hasic mechanics. the wave eauation solution is a method of analysis well founded physically and mathematically 2.3 Critical Time Interval The accuracy of the diserete-element solution is also related to the size of the time increment At. Heising.?!™ in his discussion of the equation of motion for free Jongitudinal vibrations in a continuous elastic bar, points‘out that the discrote-clement solution is an exael solution of the partial differential equation wher aL a VEp where AL is the segment length. Smith ° draws @ simi- lar conclusion and has expressed the critical time inter- val as follows: 1, “= wor VE (211) (any If a time increment larger than that given by Equa: tion 2.11 is used, the diserete-element solution will di verge and no valid results can be obtained. As pointed out by Smith, in this case the numerical calculation of the discrete-clement stress wave does not progress as rapidly as the actual stress wave. Consequently, the value of At given by Equation (2.11) ie called the Mert cal” value. Helsing? has also pointed out that when is used in a discretelement solution, a less accurate solution is obtained for the continuous bar. As At be- comes progressively smaller, the solution approaches the actual behavior of the discrete-element system (segment lengths equal to AL) used to simulate the pile ‘This in general leads to a less accurate solution for the longitudinal vibrations of a slender continuous bar. If, however, the discretelement system were divided into a large number of sezments, the behavior of this simulated pile would be essentially the same us that of the slender continuous bar irrespective of how small At becomes, provided This means that if the pile is divided into only a few segments, the accuracy of the solution will he more sensi- tive to the choice of At than if it is divided into many segments. For practical problems, a choice of At equal wut one-half the “eritical” value appears suitable lastic springs and materials of different densi- ties and elastic moduli are usually involved, 24 Effect of Gravily ‘The procedure as originally presented! by Smith id not account for the static weight of the pile. In other words. at t = 0 all springs, both internal and exiernal, exert zer0 force. Slated symbolically, F(m0) = Rimo) = 0 If the effect of pravity is to be included. these forces ‘must be given initial values to produce equilibrium of the system. Strictly speaking. these initial values should be thove in effect as a result of the previous blow. How= exer, not only would it be arskward to “keep books” on the pile throughout the drising #0 as 10 identify the initiat conditions for successive blows. hut it is highly questionable that this refinement is justified in light of other uncert which exist, relatively simple scheme has been developed as @ means of gelling the gravity effect into the compu: tations. ‘Smith suggests thet the external (soil) springs be assumed to resist the static weight of the system accord: ing to the relationship Rim.0) = [Ratm) /Ru(total) ] (Witotal) ] (2.12) where Witotal) = total static weight resisted by soil Ub); and Ru(total) = total ultimate ground resistance (Ib). ‘The quantity Witotal) is found by » Wi total) = Wib) + Fle) + ¥ Wom (23) m=2 where Wb) = weight of body of hammer, excluding. ram (Ib}; and F(c) = force exerted by compressed gases, as under the ram of a diesel hammer (Ib). ‘The internal forces which initially exist in the pile may now be obtaine Fi10) = Wiby + Fle) (24) and in general, Fim) = Fim—1,0) + Wom) — Romo) (2.15) In the absence of compressed gases and hammer weight resting on the pile system, the righthand side of Equa tion (2.11) is zero, he amount tht each intersl spring am oom pressed may now be expresse: (m0) = F(m,0)/K(m) (2.16) By working upward from the point, one finds displace. ments from D(p,0) = Rip.0)/K'tp) (217) D(m,0) = D(m+1,0) + Clm.0) (mp) (2.18) For the inclusion of gravity. Equation (2.7) should be moilified as follows Voma) = Vomit) + [Fim—1,) — Flam.) = Rim + Wem} fA (29) In order that the initial conditions of the external springs he compatible with the assumed initial forces H(m.0) and initial displacements Dim.0), plastic dis: placements D'tm.0) should be set equal to Dim,0) — Rm) /K' Gn).CHAPTER III Pile Driving Hammers 3.1 Energy Output of Impact Hammer One of the most significant parameters involved in pile driving is the energy output of the hammer. This energy output must be known or assumed before the wave equation or dynamie formula can be applied. AL though most manufacturers of pile driving equipment furnish maximum energy ratings for their hammers, these are usually downgraded hy foundation experts for various reasons. A number of conditions such as poor hammer condition, lack of lubrication, and wear’ are Known to seriously reduce energy output of « hammer. Jn addition, the energy output of many hammers ean be controlled by regulating. the steam pressure or quan- tity of diesel fuel supplied to the hammer. Therefore, a method was needed to determine a simple and uniform method which would accurately predict the energy output of a variety of hammers in general use. ‘Towards this purpose, the information generated by the Michigan Erate Highway” Commniscion in 1963. and. presented in their paper entitled “A Performance. Investigation of Pile Driving Hammers and Piles” by the Ofice of Test ing and Research, was used. These data were analyzed by'the wave equation to determine the pile driver eneray which would have been required to produce the reported behavior.’ # 3.2 Determination of Hammer Energy Output Diesel Hammers. At present the manufacturers of diesel hammers arrive at the energy delivered per blow by two different methods. One manufacturer feels that “Since the amount of (diesel) fuel injected per blow is constant, the compression pressure is constant, and the temperature constant, the energy delivered to the piling is algo constant.”"" The energy output per blow is thus computed as the kinetic eneray of the falling’ ram plus tte explosive energy found hy thermodynamics. Other manufacturers simply ive the enerzy output per blow as the product of the weight of the rampiston Wy and the length of the stroke hor the equivalent stroke in the case of closedend diesel hammers. The energy ratings given by these two. methods Aitfer considerably since the tam stroke h varies greatly thereby eausing much controversy as to which if either, method is correct and what energy oulput should be used Jn dynamic pile analysis In conventional single acting steam hammers the steam’ pressure or eneray is used to raise the rain for tach Blow The magnitude of the steam force is too Small to force the pile downward. and. consequently i teorks only on the ram to restore ite potential enele}, We xh, for the next bow. ‘In a diese! hammer, on ihe ather hand, the diesel expive pressure used to raise the ram is, for a short time at least, relatively large (see Figure 3.1) While this explosive foree works on the ram to restore ils potential energy Wy x h, the initially large explosive presire also does some useful work on the pile given by: R= f Fd (a1) where F = the explosive foree, and 4s = the infinitesimal distance through which the force acts. Since the total energy output is the sum of the Kinetic ‘energy at impact plus the work done by the Ewa = Ey + Ey (32) where Eyea = the total energy output per blow, Ey = the kinetic energy of the ram at the nstant of impact, and E, = the diesel explosive energy which does useful work on the pile 1k has been noted that after the ram passes the exhaust ports, the energy required to compress the alr. fuel misture is nearly identical to that uined by the remaining fall (4) of the ram. Therefore, the velocity of the ram at the exhaust porls js essentially the same a at impact, snd the kinetic energy at impact ean be closely approximated by B= We (hd) where Wy = the ram weight h = the total o ved stroke of the ram, and d = the distance the ram moves after closing the exhaust ports and impacts with the anvil. | _ sinensis foe ee TIME (SEC x 10°*) Figure 3.1 r Typical force vs time curve for a diesel AThe total amount of explosive energy: Eman is dlependent upon the arnount of diesel fuel injected com: pression pressure, and temperatures and therefore, may ary somewhat. Unfortunately, the wave equation must be used in cach case to determine the exact magnitude of Ey since it not only depends on the hammer characteristics. but also on the characteristics of the anvil. helmet, cushion, pile, and soil resistance. However, values of Ey deter Inined by the wave equation for teveral typical pile prob. lems indicates tha i is usually stall i proportion to the {otal explosive energy output per blox., and furtherinore, that itis on the same order of magnitude as We. % Thus, Equation (3.1) can be simplified by assuming: E= Wa Xd (34) Substituting Equations (3.3) and (3.4) into Equation 13.1) gives: Evwat = Ex + E, so that: Wath dit Wad (35) oat = Wy 3.6) The results given by this equation were compared with experimental values and the average efficiency was found to be 100% Steam Hammers. Using the same equation for com: parison with experimental values indicated an efficiency ating of 60% for the single-acting steam hammers, and 87° for the double-acting hammer, based! on an energy output given by Ewut = Wa eo) In order to determine an equivalent ram stroke for the double-acting hammers, the internal steam pressure above the ram which is forcing it down must he taken into consideration, The manufacturers of such haminers state that the maximum steam pressure or force shoul not exceed the weight of the housing or easing, or the housing may he lifted off the ile. Thus the maximum downward force on the ram is limited to the tolal weight of the ram and housing, Since these forces both act on the ram as it falls through the actual ram stroke h, they add Kinelic energy to the ram, which is given by: Ewan = Wah + Fh (3.8) where Wy = the ram weight, Fr = a steam force not exceeding the weight of the hammer housing, and h = the observed or actual’ ram stroke Since the actual seam pressure is not always applied at the rated maximum. the actual steam force ean. he expressed as: Ge m= (ho ) Wn 69) where Wi = the hammer housing weight, P = the operating pressure, and Prat = the maximum sated steam pressure. ‘The total energy output is then given by Exar = Wah + ( (3.10) This can be reduced in terms of Equation (3.7) by using an equivalent stroke hy which will give the same energy output as Equation (3.10). ‘Thus: Evan = Wa by (an) Setting Equations (3.10) and (3.11) equal yields Veh, = We he id h mhewens (Lowy ) ‘or solving for the equivalent stroke: _ » Wu Conclusions. The preceding discussion has shown that itis possible to determine reasonable values of hame ner energy output simply by taking the product of the ram weight and Sts observed or equvelat stoke, and applying an efficiency factor. This method of energy rating can be applied to all types of impact pile drivers with reasonable accuracy A brief summary of thie simple procedure for ar: riving at hammer energies and initial ram velocities as follows: Open End Diesel Hammers E = Wrh (e) Va = Vie th=d) te) where Wy, = ram weight Vix = initial ram velocity hh = observed total stroke of ram d= Distance from anvil to exhaust ports = efficiency of open end diesel hammers, approximately 100% when energy. is computed by’ this method, Closed End Diesel Hammers Wr hy fel Vu = V2e thd ley where Wy = ram weight Vu = initial ram velocity h, = equivalent stroke derived from bounce chamber pressure page = distance from any = efficieney of elosed end diesel hammers, approsialely 1005) when energy is cramputed by this method. to eshaust ports Double-Acting Stean Hammers E = Wake (e) v “Note: “For the Link Belt Hammers, this energy can be wad directly from the manufacturer's chart using bounce chamber pressure wage.= ram welt he = equivalent ram stroke Wu Pr We ‘actual or physical ram stroke operat ug steam pressure ‘maximum steam pressure recommended by manufacturer Wu = weight of hammer housing e efficiency of double-acting steam ham- mers, “approximately 8546" by this tnethod Single-Aeting Steam Hammers E = Wah ie) Va = Vigh (ey where Wy = ram weight ho = ram stroke © = efficiency of single-acting. steam ham: mers, normally recommended around 75% to 85°. In a study of the Micki gan data, a figure of 607% was found, The writers feel the 60% figure is un- usually low and would not recommend it asa typical value. character- 3. A summary of the properties and operat istics of various hammers is given il 3.3 Significance of Driving Accessories In 1965 the Michigan State Highway Cormmission completed an extensive tescarch program designed. to obtain a better understanding. of the complex problem of pile driving. ‘Though number of specilie objectives were given, one was of primary importance, As noted by Housel? "Hammer energy actuslly delivered to the pile, as compared with the manufacturer's rated energy, a the foal pot of 4 major, portion, of hie inva gation of piledriving hammers” Tn ether words, th hoped to determine the energy delivered to the pile and to compare these values with the manulacture’s ratings. The energy transmitied to the pile was termed “ENTHRU” by the investigators and’ was determined by the summation ENTHRU = 3Fas Where Fs the average force onthe top ofthe ile ding a short interval of time, was measured hy 1 specially designed load cell, and AS, the incremental movement of the head of the pile daring this time interval, was found using sisplacement transducers, andor reduced from accelerometer data. Tt should be pointed out that ENTHRU is not the total energy output of the hammer blo, but only a measure of that portion of the energy delivered below the load-ell assembly Many variables influence the value of ENTHRU, ‘As was noted in the Michigan report: “Haminer type and operating conditions; pile type, mass, rigidity, and length; and the type and condition of cap blocks were all factors that affect ENTHRU. but when, how, and how much could not be ascertained with any degree of TABLE 3. SUMMARY OF HAMMER PROPERTIES AND OPERATING CHARACTERISTICS Hanmer Hanmer Maximum Ram Hammer Anil Maxed Bated Maxima Gap Nam’ “ype” “Rated” Werght Mowsing weeucit mun or (ft) Steam Baplocive eek scturer Energy “aby Wee “aig Bathe Pros Presure Normally at a Tent sore ti)" Specified stroke fon i Varn 5900 S000 4700 300 us, to xRonn 1330 mec tesoo © "som ‘ean foc bison See 10G__Ehoo sooo 1888 : Tin oki 18,000 3.867 nee 5 earta dias : TE tome 30000 5070 17950803 28,00 16,0002, wo 800092 46300 aylon ok azo i Dean © 22400 2800 m5 800108 08.000 DEAO 92,000 4,000 TARO 8.00 az 158,000 nylon disk © wat fi : Danae Bio a6 a0 ss ones ee Soe German, oak paw 4380 uy 819 tas rws00 1S a5 akcertainty.” However, the wave equation can account for each of these factors so that their effects can he dlotermined. ‘The maximum displacement of the head of the pile was also reported and was designated LIMSET. Oseillo. eraphie records of force vs time measured in the load cell were also reported. Since force was measured only atthe load cell, the single maximum observed values for each cate was called FMAX. equation can be used to determine (among tities) the displacement Dim.) of mass “m” ”. as well as the force Fim.) acting on any ‘mass “mat time “.” Thus the equation for ENTHRU al any point in the system can he determined by simply letting the computer calculate the equation previously mentioned: THRU = 2FAS or in terms of the wave equation: = [ist 5] X [Dim+1,) — Dim+14—-1] ENTHRU(m) where ENTHRU (mi = the work done on any mase (m + 1p. m = the mass number, and ‘ © time interval number. ENTHRU js greatly influenced hy several parame. ters, especially the type, condition, and coefficient of restitution of the cushion, and the weight of extra driv- ing aps. Tt has been shown that the coefficient of restitu mm alone can change ENTHRU by 20%, simply. by changing ¢ from 0.2 to 0.6. Nor is this variation in unlikely since cushion condition varied from new to “hadly burnt” and “chips added, ‘The wave equation was therefore used to. analyz0 certain Michigan’ problems to deterinine the influence fof cushion stiffness, e, additional driving cap weights, Ariving resistance encountered, etc. Table 3.5 shows how ENTHRU andl SET increases when the load cell assembly is removed from Michigan piles. EFFECT OF CUSHION STIFFN RANSMITTED TO THE PILE (ENT on RU) a 55 Mes ae TABLE 38. EFFECT OF CUSHION STIFFNESS ON’. THE MAXIMUM FORCE MEASURBD AT THE LOAD LL (MAX) FMAX (kip) Tan ‘ushion Stiffness (kip/in. Vadly Ryn —<_Custion Stifnees (Np7ing Uvises) kip) “110802700 97,000 a2) SSS 8 D lattes bata. : x 14 sl 18 ‘30198 811569 2 50 205 ar Veo — 350 ast 1i60 30-364 522 1a 16 % 75 Boe saa 150288 Bee 1/858 From Table 3.2, it can be seen that ENTHRU does not always increase with increasing cushion stiffness, and furthermore, the maximum increase in ENTHRU. noted here is relatively small—only about 10%, When different cushions are used, the eoefficient of restitution will probably change. Since the coefficient of restitution of the cushion ‘may affect ENTHRU, a number of cases were solved with “e” ranging from 0.2 10 0.6. As shown in Tables 3.6 and 3.7, an increase in “e" from 0.2 to 0.6 normally inereases ENTHRU from 18 to 20%, while increasing the permanent set from 6 to 11%. Thus, for the case shotsn, the coefficient of restitution of the cushion has 2 greater influence on rate of penetration and ENTHRU than does ite stiffness, This same effect was noted in the other solutions, and the cases shown in Tables 3.6 and 3.7 are typical of the resulls found in other cases ‘As can be seen from Table 3.3, any increase in cushion stiffness also increases the driving stress. Thus, according to the wave equation, increasing the cushion stiffness to increase the rate of penetration (for example by not replacing the cushion until it has been beaten to 4 fraction of its original height or by omitting. the eush- jon entirely) is both inefficient and poor practice be. cause of the high stresses induced in the pile. It would he beter to use a cushion having a high coefficient of restitution and a low cushion stiffness in order to in crease ENTHRU and to limit the driving stress. Unfortunately. the tremendous variety of driving accessories precludes general conclusions to be drawn TABLE 2.4, EFFECT OF CUSHION STIFFNESS ON THE MAXIMUM, DISPLACEMENT OF THE HEAD OF THE PILE (LIMSET) ENTHRU (kin fo LIMSET (inp Tam Cishion Sifines Gy YE Cashion Suffners p/n Vanets RUT Guthlon Stifitees Ceipyingd = AE RuT (felsec) kip) * 3401080 Ufe/ace) (kin) O_O 30 9 aoa 040g he i 8 i f EY ae io a0 03 a a a ” a : e a i ua S 6 i 3 n i fi iaTABLE EFFECT OF REMOVING LOAD CELL ON ENTHRU, LU AND PERMANENT SET OF PILE ENTHRU LINSET PERMANENT SET iy ft in) (in Ram With Withowt Withowt Witk Withont Valecity Lead Lond Lead Load ‘Load Case’? Uisee) Celt ell Cell Celt Cell - 16. Oa 038 0.35 DTP.15, 36 0.67 805 = DLIP-s, — 80.2 from wave equation analyses in all but tke most general of terms, Although the effect of driving accessories is quite variable, it was generally noted that the inclusion of additional elements between the driving hammer and the pile and/or the inclusion of heavier driving accessories Consistently decreased both the energy transmitted to the head of the pile and the permanent set per blow of the hammer. Increasing cushion stiffness will increase com: pressive and tensile stresses induced in a pile during 3.4 Explosive Pressure in Diesel Hammers In order to account for the effect of explosive force diesel hammers, the force between the ram and the ‘anvil is assumed to reach some maximum due to the impact between the ram and anvil, and then decrease However, should this impact force tend to decrease below some specified minimum, it is assumed that the diesel explosive preseure maintains this pecified minimum, force between the ram and anvil for a given time, after TABLE 26, EFFECT OF COEFFICIES T OF RESTITUTION ON MAXIMUM POINT DISPLACEN which the force tapers to zero. As shown in Figure 3, the force between the ram and anvil reaches sotne ma mum due to the steel on steel impact, afterwards the force decreases to the minimum diesel’ explosive force on the anvil. This force is maintained for 10 millisec- onds, thereafter decreasing to zero at 12.5 milliseconds. The ‘properties of this curve, including values of the minimum explosive force and time over which this force acts, were determined from the manufacturer's published literature for the diesel hammers, The effect of explosive preseure ova found to be extremely variable, possibly more so than the effect of the driving accessories, and few conclusions could be drawn. ‘The only consistent effect that could be ob- served was that if the maximum impact force induc by the falling ram was insufficient to produce perma- = nent set, the addition of explosive force had little or no effect on the solution. In other words, unless the par ticular hammer, driving accessories, pile, and sail con: ditions were such that it was possible to get the pile moving. the explosive force, being so much smaller than the maximum impact force, had no effect ite = yen, Maximum Point Displacement in.) ‘ LD. (kip) (ft/see), e = 02 e— 04 c= 08 é BLTP- 100 3 7B 7 aie 296 0 B EH us a8 te __ 8 is i ae s&s 355 ra ote tae 50 7 3 ie a ng oa oo 3 158 M0 te hog x Bs EFFECT OF COEFFICIENT OF RESTITUTION ON ENTHIU 4 Ram THRU Maximum - ¥3 Pile RUT Velocity ENTHNU (hip ft) Change 2c tb fis ties o=aE o> 0d ta 30 B 8 ii is a Ed i 2 8 i 3 HH eS oe.However, the addition of explosive pressure in creased the permanent set of the pile in some cases where the maximum impact force is sufficient to start the pile moving; on the other hand, ils addition was found ineffective in an equal number of circumstances, The explosive forces assumed to, be acting within ious diesel hammers are lined in Table SI. These Forces were determined by. experiment. pessonal corre spondence withthe banner manutacturere and Tom their published Tterature 3.5 Effect of Ram Elastic In 1960, when Smith first proposed his numerical solution to the wave equation for application to pile ‘ing problems. he suggested that since the ram is usually short in length, it can in many eases be repre: sented by a single weight having infinite stiffaess. The example illustrated in Figure 2.1 makes this assumption, since K(1) represents the spring constant of only the capblock, the elasticity of the ram having heen neglected, Smith also noted that if greater accuracy. was desived, the ram could also be divided into a series of weights and springs, as is the pile. As noted in Figures 3.2 and 3.3, there is a signifi- cant difference hetween the steam or drop hammers ancd el hammers, i.e. the steam hammer normally strikes a relatively soft capblock, whereas the diesel hammer in- volves steel on steel impact between the ram and anvil. }.—__ HAMMERBASE CAPBLOCK |e PILE CAP custo \ mS ADAPTER FS _—___ PIPE PILE PIPE PILE CLOSED AT TIP Figure 3.2. Steam hammer — capaLock — PILE CAP —— cusHION —— apaPTer }——— PIPE PILE PIPE PILE CLOSED AT TIP Figure 3.3. Diesel hammer. nine the influence of dividing the ram into a number of segments, several ram lengths ranging from 2 to 10 ft were assumed, driving 100-ft pile having point resistance only. The total weight of the pile being driven varied from 1500 to 10.000 Ib while the ultimate soil resistance ranged from 0 to 10,000.15. The cushion was assumed to have a stiffness of 2,000 kips per in. Table 348 lists the results found for a typical prob- Jem solved in this study, the problem consisting ‘of a 1O.it long ram traveling’ al 20 fps striking a cushion with a stiffness of 2000 kips per in. ‘The pile used was @ 1004t 12H53 steel pile, driven by @ 5,001) ram. e OT OF BREAKING THE RAM INTO RAM STRIKES A CUSHION OR CAPBLOCK Maxi- ‘mu Maxi- Comprese ‘mum Length “alve Point Number of Pile Fore Force in Displace- of Ram — SenmentsinTile "Bile ment Divisions “tip (cp) Gad) 1 m9 fda 1 50 age Bas Btu 1 Bene eas 58 1 a earrieary 2 135308 Bae S ORS 10 135382918080TABLE 49. EFFECT OF BREAKING RAM INTO. MENTS WHEN RAM STRU A STEEL ANVIL Maximum Compressive Length Maximum, Number of Bach Force on Pile Point Anvil Ram, of Tans Ram ‘AL Displace- Weight Length Disisions Segment Center ‘ment ® ft a = in “2000 10 10.0) 0.207 50. 0.159 20 ous i 0.125 ® a 0.188 20 oun i0 0118 6 60 0.155 20 0130 2 10 01109, 1000 10.0 0.160 BO. 0159 20 O51 io. 0353 8 80. bast 20 O14 in 0484 os 0138 6 6 457 0.187 20 se O28 Xo bie o:109 as, 520 oun No pile cap wae included in the solution, the cushion held constant and the ram was divided equally into seg. hreing placed directly between the hammer and the head tf the pile. Since the ram was divided into very short lengths, the pile was also divided into short segments As shown in Table 3.8, the solution is not changed to any significant extent whether the ram is divided into 1, 2, or 10 segments. The time interval was held con: slant in each case. In the case of a diesel hammer, the ram strikes directly on a steel anvil rather than on a cushion, This, ‘makes the choice of a spring rate between the ram and anvil difficult because the impact occurs hetween two steel elements, The most obvious solution is to place a Spring having a rate dictated by the elasticity of the ram and/or anvil. A second possible solution is to break the ram into a series of weights and springs as isthe pile To determine when the ram should be divided, a parameter study was run jn which the ram length varied between 6 and 10 ft, and the anvil weight varied from, 1,000 to 2,000 Ib. Tn each case the ram parameter was rient lengths as noted in Table 3.9. These variables were picked because of their possible influence on the solution, The pile used was gat having a cushion of 2000 Beton tnd ma 12458 point bearing pile ip per in. spring rate placed the anvil and the head of the pile. ‘The soil used were RU = 500 kips, Q = 0.1 ina 0.15 see. ‘These factors were held con’ for all problems listed in Tables 38 and 3.9. ‘The most obvious result shown in Table 3.9 is that when the steel ram impacts diteetly ona steel anvil, dividing the ram into segments has a marked effect on the solution. ‘An unexpected result of the study was that even when the ram was short, breaking it into segments still effected the solution. As seen in Table 3.9, the solutions for forces and displacements for both 6 through 10 ft ram lengths continue to change until a ram segment length of 2 ft was reached for the 2,000-Ib anvil and a segment length of 1 ft for the 1,00-Ib anvil was reached. CHAPTER IV Capblock and Cushions 4.1 Methods Used to Determine Capblock amd Cushion Properties ‘As used here. the word “capblock” sefers 10 the material placed between the pile driving hammer” and the steel helmet. "The term “cushion” refers to the ma- terial placed hetween the steel helmet and pile (usually used only when driving eonerete piles) Although a capblock and cushion serve several purposes. their primary function is to limit impact stresses in both the pile and hamier. In general, it has,boent found that a wood capblock is quite effective in reducing driving stresses, more so (haw a suff capblock material such as Micarta. However, the stiffer Micaria is usually more durable and transmits a greater percentage of the hammer's energy to the pile because ‘of its higher coefficient of restitution, For example, when fouricen different cases of the Michigan studt weve saved by the wave uation, the Micaria assemblies averaged 1") more effiient than capblock assemblies of wood. However, the increased Cushion stiffness in some of these eases inreased the impact stresses to. point shere damaze to the pile oF hammer might result during diving. “The increase in stress as pariculasly important. when concrete_or prestressed concrete pice nee driven. "When diving Eonerete piles, itis also frequen necessary to include cushioning material between the helmet ‘and the head of the pile to ditibute the impact load uniformly over the surface of the pile hesd and prevent spalling. To. apply the wave equation to pile driving, Smith assumed that the cushions stesestain cue” want Series of sraight lines as shown in Figure 4. ALhough this curve was found to be sulfcienty accurate to pre det maximum compresive stresses inthe pil, the of the ares wave often disagreed with that of the actual stress wave, "To eliminate the effecte of sil resistance several tat piles were sunpended horizontally above the round. These tat piles were instrumented with stain ages at several pointe slong the length of the pile, and Especially at the head af the pile. “A cushion wes placed athe head of the pile wich was then hit by a hori Zonlally swinging i, and. duplacements, forces, and accelerations of both the rain and head of the pile were measured, "Thus by knowing the fore t the head of the pile and the lative displacement between the ram aud the heal of the ple. the fore exered inthe eushion STRESS STRAIN Stress-sirain curve for a cushion block, Figure 4.3. 000-—— —________, | = rame s-8 cone STRAIN UN/IN) Figure 4.2. Dynamic and static strese-sivain curves for 4 fir cushion. and the compression in the cushion at all times could bbe calculated. ‘Thus the cushion’s stress-strain diagram could be plotted to determine whether or not it was actually a straight line. Using this metkod, the dynamic stress-strain prop- erties were measured for several types of cushions, 1 was further determined that the siresestrain curves were nat Hnear as war assumed by ‘Sith bat Taller appeared as show in Figure 4. Because it as Cxteemely dificult to determine the dynamic sreseetrain Curve by this method, n cushion West stand wae eon structed as shown in Figuse 43 in any attempt to simplify the procedure, Since i was not known kow much the rigidity of the. pedestal affected the cushion's behavior, several cushions whose stress-strain curve had been previously determined by the first method were checked. These studies indicated that the curves determined by either method were similar and that tke cushion test stand could be used to accurately study the dynamic load- deformation properties of cushioning material,a = ni LL, L___f Figure 4.3. Cushion test stand. ‘Throughout this investigation, a static. stress-strain curve was also determined for each of the cushions. ‘Surprisingly. the static and dynamic stress-strain curves for wood cushions agreed remarkably well. Atypical example of this azreement is shown in Figure 4.2,” The stress-strain curves for a number of other materials commonly used 2s pile cushions and capblocks, namely oak, Micarta, and asbestos are shown by Figures 4.44.6, 4.2 Mealized Load-Deformation Properties ‘The major difficulty encountered in trying to. use the dynamic curves determined for the various cushion materials was that it was extremely difficult to input the information required by the wave equation, Although the initial portion of the curve was nearly parabolic, the top sezment and unloading portion were extremely com: plex. “This prevented the curve from being input in equation form, and required numerous points on the curve to be specified. Fortunately. it was found that the wave equation accurately predicted both the shape and magnitude of the stress wave induced in the pile even if a Tinear Force deformation curve was assumed for the eushion, so long Toad as th i portion was based on the secant mevlul of elasticity for the materiel ‘as opposed to the initial, final, or average modulus of elasticity), and the unload ing portion of the curve was hased on the avtwal dy namic cooflicient of restitution. ‘Typical. secant moduli of elasticity and coeflicient of restitution values for various materials are presented in Table 4.1, TABLE 44. TT! E ® psi Micarta Plastic 450,000, Oak’ (Green) -45,000° Asbestos. Dites Fir Plywood Pine Plywood Gum “Properties of wood with load applied perpendicular to ‘wood grain. 4.3 Coefficient of Restitution Although the cushion is needed to limit the driving stresses in both hammer and pile, ils internal damping reduces the available driving energy transmitted to. the head of the pile. Figure 4.1 illustrates this energy loss, with the input energy being given by the area ABC while the energy output is given by area BCD. This energy loss is commonly termed coefficient of restitution of the cushion “e", in which Area BCD) ‘Area ABD ee esi) stress STRAN UK) Figure 44, Dynamic stressstrain curve for an oak cushionSTRAIN (INK) Figure 45. Dynamic stressstrain curve for « micarta cushion. Once the coefficient of restitution for the material is known, the slope of the unloading curve can be deter mined as’ noted in Figure 4.1, For practical pile driving problems, secant moduli 3000 2500. 8 3 STRESS IN PSI 8 3 1000; 500: 006 = one ok 0” alg STRAIN IN IN. PER IN. Figure 46. Stress vs strain for garlack asbestos cushion of elasticity values for well consolidated cushions should be used. ‘Table 4.1 shows typical secant moduli of well consolidated wood cushions. ‘Table 4.1 also lists the coefficient of restitution for the materials which should bbe used when analyzing the problem by the wave equa tion, CHAPTER V Stress Waves in Piling 5.1 Comparison with Laboratory Experiments p x As noted in the preved 1 section, several test piles were instrumented pended horizontally above the ground. This example pile was a steel pile, 85 ft in length with a cross-sectional area of 21.46 eq, in, ‘The cushion was oak, 7 in. thick, The ram had a weight of 2128 Ib and a velocity of 13.98 fps. The cushion wae clamped to the head of the pile and then struck by a horizontally swinging ram. The pile was instrumented with strain gages at six points along the pile, and dis- placements and accelerations of both the rain and head of the pile were also measured, In order to utilize Smith's solution to the wave equation, the following information is required: 1. The 2. The actual dynamic stress-strain curve for the cushion, normally ial velocity and weizht of the ram, 3. The area and length of the pile, and 4. ‘The density and modulus of elasticity of the pi Since the stress in curve for the cushion was un- known, the numerical solution was rewritten such that ii was'not needed. This was possible since the pile was instrumented with a strain page approximately 1 ft from the head of the pile which recorded the actual stressFigure 5.1. Theoretical vs experimental solution, Strain 25 ft from pile head. induced in pile by the ram and cushion, The force measured at the head of the pile was then piaced directly at the head of the pile and the wave equation was used to compute stresses and displacements at all of the gage points along the pile. Figures 5.1 and 5.2 present typi- ‘eal comparisons between. the experimental: resulls and Wave equation solutions at two points on the pile, and illustrate the degree of accuracy obiained by tse of the wave equation. It must be emphasized that this excellent correlation between experimental and theoretical results was in ef fect obtained by using the actual dynamic load-deforma- tion curve for that particular case. However, as men tioned earlier, the stress-strain curve for the cushion is normally assumed to he linear as shown in Figure 4.1. ‘To determine how much the use of the linear stress strain curve will affect the solution, the previous case was rerun using the straight [ine strees-strain curves. As noted in Figures 5.3 and 5.4, the solutions for the linear and nonlinear cushion assumptions agreed favorably. The use of the straight line assumplion is reasonable since it gives fsirly accurate results for both maximum Figure 5.3. Theoretical vs experimental solution for strains 25 ft from the pile head. tensile and compressive stresses, Furthermore, it pre- dicts the shape of the stress wate reasonably well 5.2 Significance of Material Damping in the Pile Other parameters were often varied in an attempt to obtain more accurate results, one of which was. the material damping capacity of the pile material. How: ever, most suspended pile cases studied stronzly indicated that damping would be negligible hecauee of the extreme: ly low rate of decay of the stiess wave in the pile. The only pile in which damping was thought to be signifi cant was a lightweight concrete ple with a statie modulus of elasticity of 3.96 x 10" and a “sonic” modulus of elasticity of 4.63 x 10° psi. This problem was chosen since E, was relatively larger than E. indicating the pos sibility of rather high damping. It can be seen in Figure 5.5 that the magnitude of the experimental results di- Iminishes slightly after four cycles. The magnitud> of tical solution with damping neglected would not. Figure 5.5 compares the experimental and theo- retical solutions for stresses when Smith's proposed method of damping is included. In this ease, the ex- Figure 5.2. Theoretical vs experimental solution, Strain 52 fe from pile head. Figure 54. Theoretical vs experimental solution for strains 52 Jt jrom the pile head.Gitepeeegsg Figure 5.5. Comparison of experimental and theoretical solutions for stresses 25 jt from the pile head. perimental and theoretical solutions are in excellent’ Agreement, both in wave shape and rate’ of decays Although it is extremely interesting to be able to predict the dynamic behavior of pil Fae), most practically” the. primary wu sresea induced “in the the fist or second pass of the sires wave slong the pile, Daring this time, the effects of damping are eatremely small even for the lightweight aggregate ple and are apparently of no practical importance,“ Whethce this conclusion vill be aeeurate for timber or other pies having mach higher doping caplet eter se Pacity could affect the results earlier in the soluiog aad thus be of significance. le which occur It should be emphasized that the above conclusions are valid only for normal pile driving conditions. If the wave must be studied for an extended period of time, damping in the pile may be significant and should be accounted for. CHAPTER VI Soil Properties 6.1 General A limited amount of work has heen done on soil properties and their effects on the wave equation solution of the piling behavior problem. A total of three re search reports concerning soil properties have been pub lished! by the Texas Transportation Institute during the “Piling Behavior” study. Research Reports 33-7 and 337A" 62 give the resulls of a series of laboratory dynamic (impact) and static tests conducted on satu: tated sands. Research Report 33-0" gives the. resulls of a field test on a full scale instrumented pile in clay A brief summary of the results of these tests are given in this chapter. 6.2 Equations to Deseribe Soil Behavior Examination of Equation (6.1) shows that Smith's equation describes a type of Kelvin theological model as shown in Figure 6.1. Rimt) = (Dima) — D'tm.v1K'(m) {1 + Im) Vom 10} (1) The soil spring behaves elastically until the deformation Démat)_ equals Q and then it yields plastically with @ load-eformation property as shown in Figure 62¢a) The dashpot J develops a resisting force proportional to the velocity of loailing V. as modified the t Kelvin model slightly’ as sho (62) equation will produce a dynamic loaddeformation bee havior shown by path OABCDEF in Figure 6.21b). Hf lerms_ in Equation (6.11 are examined, it can be seen that Smith’s dashpot force is given by [Dim.t) = D'en.ay] Keim) (Jam) Vom) The dimensions of J are seo/ft and it is assumed to be independent of the total soil resistance or size of the pile. It is also assumed to be constant for a given soil under tive conditions as is the static shear strength of the Soil from which Ru on a pile segment is determined, Ru is defined as the maximum soil resistance on a pile segment, Smith notes that Equation (6.1) produves no damp- ing when Diin.t) —— D'(m.t) heeomes zero. He suze gests an alternate equation to be used after Dim,t) first hecomes equal to Qin): Romi) = [Dima — D'Gm)) K'tm) + J(m) Rut) V(m,t—4) (62) Care must be used to satisfy conditions at the point of the pile, Consider Equation (6.1) when m =p, where p is the number of the last element of the pile Kip) is used as the point soil spring and Jip) as the point soil damping constant. Also at the point of the vn ar TEA Bue} 22 6 oe gmt as uae aen volar RESItaNC (aus “Tone Q = 0. in, and I(point) = 0.15, I1side) = 0.03. Soil resistance was assumed to be 50% at the point and 50% Jriction distributed uniformly over the embedded length below a depth of 10 ft. Hammer efficiency assumed to be 80%. sf rte somos ces 0 Figure 753. Summary of piles tested to failure in sands. TABLE 7.3. ERRORS CAUSED BY ASSUMING A COMBINED J(point) = 0.1 FOR SAND AND J(point) = 0. FOR CLAY USING EQ. 7.1 (For Piles Supported by Both Sand and lay) ‘Tes ance at ‘Soil % Error ae Average °% Brror “Indicates piles which exceeded the testing equipment’s capacity, and could not be load-tated to failure **Rye for these piles wore actual lond test measurements corrected to account for soll “setup.”NOTE: TEST FAILURE LOADS ARE THOSE EVALUATED BY EBASCO'S ENGINEERS 0%, $5 20% 150 100) 50 WAVE EQUATION ULTIMATE RESISTANCE (RU- TONS ° 60 120 180240 TEST LOAD FAILURE - TONS Figure 7.6. Wave equation. ultimate resistance vs test load failure (after Rej. 7.2, data jrom Fig. 7-5) (sands). the damping constants J(point) = 0.3 and J'(side) = Spoink) /8 gave the best correlation, The accuracy of the correlation is shown in Figure 7.3 to be appro mately =+50%. If more than one soil was involved the damping constant used was a weighted average calculated from Hipoint) = Z (RK, x I(point ] (7a) where R, = the ratio of the amount of resistance of each type of soil “i”, to the total soil resistance, both determined after setup hhas ceased, and Upoint) J'(side) ‘able 7.3 shows the damping constant that, was calculated from Equation 7.1 using Jipoint) = 0.3 for clay and J(point) = 0.1 for sand. The accuracy of the correlation, as shown in Figure 7.4 wae approximately 225%. Mosley?# has found a similar correlation with 12 piles driven in sand. Figure 7.5 is a summary of the piles tested. Figure 7.6 shows that all resistances on these piles fall within 20% of that predicted by the wave equation. CHAPTER VIII Prediction of Driving Stresses 8.1 Introduction Tn Appendix. A. the exact solution for the stress weave introduced into a long slender elas pile is de ved ubing the clasical one-dimensional wave equation ‘The solution of this equation depends upon eran it fs assumed that the pile ty prismatic With lateral dimensions small comparison to Hts length (Poisons effects can be negleel that the ile and cushion material are linearly elastic, and the Yam has nfnile rigidity (assumed to hea rigid body). The equation ‘which poverns the stress amplitude in_ most practical cass, shows that the mogoltae of the stress Induced at the head of the pile by the falling. ram, i Sir proportional fo th weleity of the raat in pact. The equation further shows al the siffnesses of the cushion and pile alee have a significant elfect on the magnitude of the sires generated. The soll resistance tir the side and at the point of the pile will also affect, the magnitude of the strestes in the pile Chapter IF discusses Smith's numerical solution of the onedimensional wave equation, This. particular lechnigue fy sling he wate eution is ach es for application to problems which ean have. inlatic cuchions and material properties as well ag rol on the Se and point of the pile. Chapter V diceuses the eneration of stress waves in piling, the signifieance of Materia damping, in the. pile and’ the effects of pile dimensions on driveabiliy This chapter demonstrates. the validity of Smith's numerical solution by comparing its results with the theoretical solution of Appendix A and with field date, 8.2 Comparison of Smith's Numerical Solution with the Classical Solution For the purpose of correlation, consider a concrete pile, square In cross-section, with an area of 489 in# and 90 ft long. ‘The modulus of elasticity of the pile material is assumed to be 5 x 10* psi. The pile is cone sidered to be free at the top with the bottom end fixed rigidly against movement. No side resistance ie present. £ roe 3 oy i == 3 sa STANCE FROM WEAD OF PILE IN FEET 2 Figure 8.1. Maxinum tensile stress along the pile.MaXMUM COMPRESSIVE STRESS «4 P51 % STANCE FROM HEAD OF PILE\IN FEET Figure 8.2. Maximum compressive stress along the pile The following information is also applicable to the cor- relation Weight of the ram 11,500 Ib, 1445 fps 3,930,000 Ib in., Velocity of the ram Cushion block Coefficient of restitution of the cushion hlock = 1.00 fess Solutions have been obtained for the exact solution of the one-dimensional “wave equation and. for Smith's swumerical method using 10 sezments. Previous studies"! had shown that segment lengths of L/10. would. yield aceurate results, Figures 8.1 and 82 show comparisons of the maximum tensile stress and: maximum compres: sive stress, respectively. versus position along. the lenzth of the pile. Note the time interval used ime differene- ing interval used in the numerical selution) for solutions shown is varied from 1/1410 seconds (this is the eritical time differencing interval) to. 1/20,000 seconds, " Note that when the differencing interval became very stnal, 1/5000 seconds. the accuracy of the olution was it improved. Note also. that the numerical soltion js very close to the exact solution. Other comparisons have been made for the stresses at olher points inthe pile and for other combinations of the end boundary conditions “Heising*? and Smith*? have shown that the diserete-clement numerical solution is an exact soli- tion of the onedimensional wave equation when AL Av VE/p where, AU = critical time differencing interval, segment length, modulus of elasticity, and p = mass density of the pile material This time interval is the “critical” time intereal For practical problems, a choice of At = onedhalf the “eritieal value,” appears suitable since inelastic springs, materials of different densities, and elastic moduli are usually involved, 8.3 Correlations of Smith's Solution with Field Measurements In previous reports*** the writers have shown several correlations of the wave equation with stresses measured in piles during the time of driving in the field. Typical examples of these correlations are shown in Fig. ures 83 and 84. The significant conclusions draven from these tests are as follows 1, The maximum compressive stresses occurred at the head of the pile. 2, Maximum tensile stresses were found to occur near the midpoint of the piles. 3. The computed compressive stresses and dis: placements agree very well with the measured data, 4. The computed tensile stresses appeared high but in view of the unknown dynamic properties of the soil, concrete, and cushioning materials involved in the problem, the’ quantitative comparisons shown were con- sidered good, 84 Effect of Hammer Type and Simulation Method It has been shown"? (see Chapter IIL) that the ram. of a pile hammer can be idealized as a riaid body pro- vided it strikes on a capblock or cushion. If the ram strikes directly on steel, as in te case of the diesel ham- mers, the accuracy of the solution for stresses is im- proved by breaking the ram into segments, Figure 8.3, Stress in pile head vs time for test pile.Figure 84. Stress at mid-length of pile vs time for test pile. For diesel hammers, the explosive force used to raise the hammer for the nest blow does work on the pile and should be included In all hammer simulations, all. parts which are in the force transmission chain should be included. The housing and other parts which do not serve to transmit, the driving energy may be neglected, Refer to Appendix B, Tables B.1_ and B2, for recommended values for use in the simulation, 85 Effect of Soil Resistance If soil borings are ‘lable, the distribution of the soil resistance on the pile should be estimated from sol Shear strength date. ‘In_ general, piles in unilonn co hesive soils will have the il’ resistance distributed uniformly in side frietion with about 10 t0. 20°. potat resistance." Gohesionless soils can generally be-simue lated with a triangular friction distibution with, about 40%. in side friction and 60"f “of the total resistance at the point. "The actual distributions. used will Course, depend on the properties of the sols ple tepe, ete and shoul be studjed for cach’ ease important to note. however, that the soil distibution will affect the magnitude of the driving. stresses. Thi is particularly true for the reflected tensile stresses. In most investigations for driving stresses itis best to Van) the distribution over the expected range and choose the ‘most conservative result, fellciedtenile stresses ar highest wien the soil existance acting nt the pile point is'small. 8.6 Effects of Cushion Stiffness, Coefficient of Restitution, and Pile Material Damping Tt has been shown*® (see Chapter IV) that the actual load deformation curve for a cushion is not a straight line, but is parabolic. However, a straight line which has a slope given by the secant modulus will give reasonably accurate results. The cushion’s dynamic coefficient of restitution was found to agree with com monly recommended values. It has also been shown that the effect of internal damping in the concrete and piles will usually have negligible effect on the ing, stresses. 8.7 Fundamental Driving Stress Considerations ‘The purpose of this discussion is to briefly describe and discuss the phenomena of impact stresses during driving. Compressive Stresses, High compressive stress at the head of the pile can he caused by the following: 1, Insufficient cushioning material between the pile Griving ram and the pile will result in a very high com pressive stress on impact. 2. When a pile is struck by a ram at a very high Nelacity. or from a very high drop, a sree wave of high magnitude is produced. This stress is directly. propor. tional to the ram velocity. If the pile is idealized as a long elastic rod, with an elastic cushion on top an equation for the compres: sive stress can be developed (see Appendix A). The approximate equations for the maximum compressive stress at the pile head are as follows Notations used art © max = maximum compressive stress at_ pile Eead (psi. W = 1am weight (tb), v ram impact velocity (in, Via, h ram free fall (in.), « | K cushion stiffness (Ub/in.), Ac Ee 6 = cross-sectional area of cushion (in2), modulus of elasticity of cushion (psi), te = initial uncompressed thickness of, cushion (in), t = time (see), A cross-sectional area of pile (in2), E = modulus of elasticity of pile (psi), 1, = length of pile (in.), unit weight of pile (lb/in2),Cael oncp ~KV A VP where t is found from the expressi m =v tan (CV Case, on =p Case II, n>p KV" 7 max A VaFS pe aimh (tv (8.3) where t is found from the expression tanh (t Vi Equations (8.1), (8.2), or (8.3) can be used to determine the maximum compressive stress at the pile head. For most practical pile problems » will be less than p and Equation (tk1) will be used. However, this is not always the case, For a given pile these equations can be used to determine a desirable combination of ram weight W, ram velocity V, and cushion stiffness K so as not to’ exceed a given allowable compressive stress at the pile head, To illustrate the ‘of the equations consider the following situation. Given: Concrete Pile L, = 65 fe A = 200 in y= 0.0968 Ib/ind (150 1/4, E = 5.00 x 10" psi Green oak cushion, grain horizontal 200 in 45,000 psi (for properties of wood sce Chaps ter IV! te = 30 in Ko = AE = 30 x 10° Ihyin Steel ram W = 5000 th ho = 46 in V = V2gh = 167 in./sec B= 306 in,/scc? Calculations: K = a n= RVR = ree Ke 1 > = f= awe Since n

p) The third and final case is where n is greater than For this condition, the solution of the homogeneous ferential equation, given by Equation (A.14), assumes the form Ve =e" [A et VEO Apel V Va = et [Ay sinh t Vi + Ay cosh t Vi — FI] ‘The general solution then becomes Wt¥ ve (aan Applying the boundary conditions required by Equation AMY yields =A 4 Me oma, + ME =oMg A= aE Substituting the required boundary condition given by’, Equation AZ then gives eel = Sne™ (Ay sinh t Vin P+ As cosh fe __ nM AE 1 Rewriting Equation A.24 gives KVe ett Vo. Nir Substituting Equation A.25 into Equation A.6 gives OF e* sinh CVn — pF (A.26) [»- ©* (cosh t VF +m sinh tVOF=P | 55) where n= KV Ewi = \/ke pe VN Equation A.26 gives the stress at the head of the pile as a function of time in the ease where n is greater than p, AS Maximum Compressive Stress at the Mead of the Pile ‘To compute the maximum compressive stress at the pile head, Equations A.19, A.23, and A.26 ave required. Numerical studies of these equations have shown that if the last term in each equation is omitted, litle accuracy is lost, and the expression becomes relatively. we. Since it is necessary to know the time, t, at the maximum stress oceurs, Equations A.19, A’23, .26 will be differentiated with respect to time and set equal 10 0, This in turn will allow the maximumstress to be found. ‘The following sotations are again the ram impact velucity ( cushion stiffness (Ub per = time (seconds) the cross-sectional area of the pile (in.*) ‘modulus of elasticity of the pile (psi) unit weight of the pile (Jb per in.%) wcceleration due to gravity (886 in. per cross-sectional area of the cushion (in) the modulus of elasticity of the cushion (psi) cushion thickness (in.) In order to further simplify the solutions, the fol- lowing approximate equations for the maximum com- pressive stress are presented: Case 1 (where n is less than p) AVP oe where tis given by the equation tan (UV Case 2 (where n is equal to p) = — [AY _ wie (a2 % (max) Fe where the value of t was given by Co where t is found from the expression tanh t Vr] Equations AZT. A.28 ond A.29 cay be ase to asco the miami utes A he hed Of the ple" moa! practic! ple geaklone Gale To than p and Euton Ac wi atl he sel Aiough W's nol alway te ease For a given pile these equations can he used to determine the proper combination of ram weight, W. ram velocity, V, and the required cushion stiffness, Ky in order to prevent excessive stresses at the head of the In most cases, there is some mivimum amount of < energy which must he available to drive the pile. ‘For examp'e, the maximum energy output avail- able to a drop hammer is givea by its kinetic energy at the instant of impact. Therefore, ye 5 KE=W should be equal to oF greater than the energy required. Tt would” appear that the most efficient way to ineresse hammer energy would be by increasing the ram velocity V. "However, Equations A.27, A28, and A.29 show thet the maximum compressive sve at the head of the pile yi crane gropstionaly with vloiy. On the ober Thand, to inereae driving energy the maximum compres: sive stress at the head: of the pile increases slighty as 'W increases, It is therefore desirable (considering diving stresses) to increase the ram weigh, W, if the le driving situation requires that the driving energy be Increased. Once the ram weight and ite velocity & im pact have been selected, the spring rate of the cushion {K) cai be varied to hold the maximum compressive stress within allowable Timits AG Length of the Stress Wave It is known that the magnitude of the reflected stresses in a pile will be a function of the length of the sitess wave and the length of the pile. The length of this. stress wave is easily found from Equations A.19, A23, and A.26, If the last term is again omitted in each of these equations, little accuracy is Jost and relatively. simple expressions are oblained for the siress at the head of the pile, Omitting the last term in Equation A.19 yields _ _nkv AVpae Equation A.30 is seen to equal 0 at tin again at em sin t VF (A230) et equals 0 and VP 1us, the second of these equations gives the duration of the impulse stress. Noting Equation A.1 2, the stress wave velocity, ¢, is found to he ac ‘The length of the stress wa then obtained fom L=a forn

p the stress wave is infinitely long, L.=APPENDIX B Wave Equation Computer Program Utilization Manual BI Introduction This appendix describes the utilization of the com. puter program for the application of the one-dimensional wave equation to the investigation of a pile during driving. ‘The, program can be used to obtain the following information for one blow of the pile driver's ram for any specified soil resistance: 1. Stresses the pile 2 Displacement of the pile (penetration per blow). 3, Static load capacity of the pile for specified soil resistance and distribution, ‘This ca: pacity is the static resistance at the time of driving and does not reflect soil set-up due to consolidation, ‘The program is valuable in that system parameters nored before ‘in pile driving formulas) can be and their effects nvetzaed, Tt makes po neering evaluation of driving equipment and pile type, rather than relying only upon experience and judgement, In order to simulate a given system, the following information is essential: 1. Pile driver. a) energy and efficiency of hammer, 1b) weight and dimensions of ram, ©) weight and dimensions of anvil (if induded), 4) dimensions and mechanical properties of eapblocks, " } weight and’ dimensions of pile cap helmet, £) and dimensions and mechanical. prop- erties of cushion, 2 Dimensions ‘gh and mechanical prop 3. Soil medium, a) embedment of pile, b} distribution of frictional soil resistance over the embedded length of the pile expressed as a percentage of the tolal static soil resistance, €) Point soil resistance expressed as a per- ‘centage of the total static soil resistance, 4) _ublimate elastic displacement for the soil ‘on the side and point of pile, ©) and the damping constant for the soit on the side and point of the pile, It should he recognized that the solution obtained with the program represents the results for one blow of the hammer at the specified soil emhedment and soil resistance. ‘The techniques for idealization can be categorized in three groups: 1, the hammer and driving accessories, 2 the pile, and 3. the soil, B2 Mdealization of Hammers The program is formulated to handle drop ham- mers, single, double, and differential acting steam ham- mers and diesel hammers that operate on the head of the pile. The techniques presented in this section are general in scope and are presented. for. illustration, Appendix B gives the idealizations and pertinent infor. mation for the most common hammers Figures BL through B3 describe the idealization for the following cases: 1, Case 1 —Ram, capblock, pile cap, and pile (Figure B1). 2. Case I —Ram, capblock, pile eap, cushion, and pile ‘(Figure 82) Case IlL—Ram, anvil, capblock, pile cap, and pile (Figure B3). Soro «y re ern, xc) MOD = wetghe of eam, 08) ray = ADE, an + MOLES ittnve of ee capbocky (A/i8) NG) + erone seettonal ran of the capoek, (a? Fa) = nodatun of LU) + ehlokees of the capbioc, ie) Figure B.1. Case Iran, capblock, and pile cap.a ——— erwe.nw ———— Sr rnen, not Figure B.2 Case I—ram, capblock, pile cap, and eushion. B3 Ram Kinetic Energies The kinetic energy of the ram for specific hammer types can be calculated as follows: 1. Drop hammers and single acting steam hammers: : En = W() (h) (er) (Bay °° eg = ram Kinetic energy, (fb) W(1) = ram weight, (Ib) h = ram stroke, (11) hammer mechanical efficieney (usually between 0.75 and 0.85 for most single acting hammers). 2. Differential and doubleeting. steam hammers: = Pau , Wh) En =h [: a wi Jeo e (B2) where hh = actual ram stroke, (ft) Pecsar = actual steam pressure, (psid Pract = manufacturers rated steam pressure, (psi) W(h) = hammer housing weight, (Ib) W(1) = ram weight, (Ib) er = efficiency is approximately 85% for these hammers. 3. Diesel hammers: here BE WA) Oe 6) + 0) (B3) th, = actual ram stroke for open-end hammers, and the effective stroke (includes effect of ounce chamber pressure) for. closed-end hammers, (ft). The energy Ey for the closed-end Link Belt hammers, can be read directly from the manufacturer's chart using bounce chamber pressure), ex = elficieney of diesel hammers is approximate: ly 100% © = distance from hottomdead-center of anvil to exhaust ports, (ft). Work done on the pile by the diesel explosive force is automatically accounted for by using an explosive pres- sure (see Sample Problem and Table 2). Caleulations for idealization W(1) = weight of ram, (Ib) Ka) = AGED, teas of te aback (ib/in.) Whee A(1) = cross sectional area of the capblock, (in?) E(1) = modulus of elasticity of the eapblock (psi) L(1) = thickness of the eapblock, (in) Note: See Table 4.1 for capblock properties. Calculations for idealization W(L) = Weight of ram, Ib) ——— [—— rine re erat, 0) en = AED ag gesttnas ot the son an CUD = mete of elvesetty of cam acetals (Ge) LC) = tenth of tans ln Figure B3. Case Il—rar, anvil, capblock, and pile cap.K(1) = Stiffness of the Capblock, (Ib/in.) K(2)¢ = Sffness of cushion, (Ib/in.) K(2)p = Stiffness of pile spring, (Ib/in.) " Ke Kip chine KQ) = RSE Rp combined stiffness of K(2)o and K(2), in series Note: See Table 4.1 for capblock and cushion properties. a Ke = where A = cross-sectional area of cushion, in E = secant modulus of elasticity of cushion ma- terial, psi L = thickness of cushion, in. Calculations for idealization ‘W(1) = Weight of ram, (Ib) K() = AGL EO he ston of the sam (Ib/in.) wer A(1) = ram cross sectional area, (in.) naw win—et WAMMER cusHion C81 eee PILE K (2), —+2 Pie Pie w(s1—+| DROP HAMMERS SINGLE ACTING STEAM HAMMERS L(1) = length of ram, (in.) This calculation assumes that the pile cap is rigid, In the hammer idealization, note that the parts com- posing the pile driver are physically separated, ie, the Tam is capable of transmitting compressive force to the anvil but not tension. ‘The same is true of the interface between the anvil and pile cap, and the pile cap and the head of the pile. ‘The program cont provisions for eliminating the capability of transmitting tensile forces between adjacent segments. ‘The me- chanies of this provision are more fully explained in the following sectio ‘Tables B1 and B2 list the information needed for the simulation of the most common types of pile driving hammers. BA Methods of Including Coefficient of Restitution in Capblock and Cushion Springs In the case where K(1) is a capblock (Cases 1, 11, and III), and K(2) is a cushion (Case IT), it is desirable to include the energy loss due the coefficient of restitu. tion of the particular material. j+—— HOUSING w, Ss capaLocK (1) =Ag (fp —me car wie = sion xii, a aM Wil) DOUBLE AND DIFFERENTIAL ACTING ‘STEAM HAMMERS a) @) TABLE 81 DROP HAMMERS AND STEAM HAMMERS wamuen [rere] Tan | Yast | Nes at.tos| tame [tia | STROM] cote] SF wat ss | « | 3000 | — = 2 2 300 | | 0.80 wxrss [a | 8000 [= = 22 }2 3] a 325 |= [0.80 vuUcAN + a 3000 | 1000 = ge |e EF ais 300 =| 0.80 vac 2 [a [000 [woe f= 2 *§ | 2 8) BP poe wucan 300] | 3000 | 1000 Se [te i toe | 20 | os vurcan soc | 6 | s0co | 1000 a é 8 t29_| 120 | 068 wuscan aoc |e | e000 | 2000 | essa ]$ § |$ &| 2 138 | 120 | 00s vutean taoc| @ | 14000 = 13986 2 g 20 | 140 | 008 REPRESENTATIVE VALUES FOR PILE NORMALLY UEEO IN” MIGHWAY CONSTRUCTIONigure B4 the evefficient of restitution is defined r m1 . iarase = V Energy input (BS) Tin Case II it is necessary to combine springs K(2)e and K(21q to determine the equivalent spring K(2)- Tn this instance it js also neceseary to determine the coeffi- cient of restitution of the combined springs. The stiff. 8 hess of the apring in the restitution phase is the slope of the line DB in Figure Bi “ Kon = (B6) Since, _ ae Energy output = Area BCD = Fy (Ae — An) /2 Ore) Energy input = Area ABC Figure BA. Definition of coefficient of restitution. ge = Fnlde = Kan Fy (9 Koo ‘ ‘The combined restitution stiffness of K(2)q and K(2),, can be determined from, “ : (for restitution a8 phase DB in Kon = 3 Bn Kp Figure Ba) Ye—~ wo nores ron TABLE 82 ‘for ctu atone . = seca AMD {roy ory "rom a0 KY row 2) where g = carevocx «420 + overege volves g SS ee wor SF crsnon ciaie 1 me ec 4 |-—— mut wie TABLE B.2- DIESEL HAMMERS wo] war] wor) KOR] KGL fwaxre] ¢ ] en POONA] re wamer | 8) | tar | to) fas efector] Me | MO [Ment | cen fereail OEE | wat_ve-20 [2000 [00] gl a2 soo*[ 002 | >| aes00 | 100 mer pesse | -se0e [re] 8838[ ser | ose sorter] [38000 [00 ict oe-40 | +000-[ aso] * =| v2.0 eoo*] 116] 5 8 [iseo00| 100 cewmae 05 oo [azo] 111 [es] 6 a|_ [Teoosfoas_| + 5 [lasx00| 100 pecwas_p.a [2750 | —e16 sist we} y | “8 [aoc roe] 2 y [esr00 [100 Devwac_o 22] s880-[ 1576 aarp ase] 2 | 8° Ceooet ioe] = 3 [ser00- [1001 Devuac_pea | 9500 | 2081 wez{ oes] 3 |“ [oops] ¥ 5 [200000 | 100 tunic oT veo [720 sas | 155) 3 [Ceestoee| 3 8 [000 [00 siz 087 wees [10s Tar[ 080] fg [se000 | 100 245] 40007 veo | 185) ase"[ vas) & [T0000] 190 zo | sore | 79) woes | 186 s20[ 0s | _§ [e000 | 100from Equation (A-7), 2 ee Lene Kar > Kae * Kaye (2)? = Kl peat K(2y Ride Ky + e(2),® Ke] Kee K(2)y KO) Khe + ROp (2) (B.8) V rate ey, BER + KCC BS Mdealization of Piles The idealization of the pile is handled by breaking the continuous pile into discrete segments, Each ses- ment is represented by its weight and_ spring. repre: senting the total segment stiffness. In Figure B5, the weight representing the segment is assumed to be con: cenirated at the end of the sezment away from the point ‘of impact. This places the spring on top of the weight whose stiffness it represents, ie, K(2) is associated with Wi) Piles should he broken into segments not to exceed approximately 10 feet in lengths, but into not Tess than five segments. The stiffness of each pile segment spring is calculated from Atm) E¢m)) Key Bleccn A sine (3) Ke) Z ‘pyar 4 x) sac (692 0— stack (71 0—s suace (8): 0— EALIZED re BS. Pile idealization. where Kon1) = spring stiffiess for segment m, (bfin.) A(m) = cross sectional area of segment m, (in) E(m) = modulus of elasticity of the mat of segment m, (psi) L(m) = length of segment m, (in.) ‘The weight of each pile segment is calculated by hee Wit) = bm) Lom) nit weight of pile material, (Ib/in.) If the pile is tapered, the average value of Am) should be used. : The program has provisions for handling cases where the physical construction of the pile prohibits the transmission of tensile stresses or is capable of trang mitting tensile stresses only after a specified movement of a mechanical joint (joint slack or looseness). ‘These conditions occur with certain types of pile splices. ‘The program provides for this eventuality’ by entering the following 1) Ifa joint (a joint is defined as the interface Detween two segments) can tvansmit tension, the slack or looseness is entered as SLACK (m) = 0. (Refer to Figure BS) 2) If a joint is completely loose, no tension ean be transinited.and SLACK” (im) should be made avery large number, ie SLACK (mn) 1000.0. 3) If a joint is capable of moving 1.25 in. before transmitting tension, SLACK (mn! = 125, the physical value of the slack or Tooseness in 4 joint is entered in iuches ‘The SLACK (m) values are always associated spring Kim). In Figure BS, if tension ean he trans: mitted across the interface hetween segments 3 and 4, the slack value would be associated with spring K(3), ie, SLACK (3) = 0. ‘The interfaces between the various parts composing, the pile driver (ram, eapblock, pile eap, its.) which ean- not transmit tension are also handled by setting the SLACK values equal to 1000, B6 Explanation of Data Input Sheets Data for the Pile Driving Analysis program is en- tered on two sheets. Page 1 contains data pertaining to the physical parameters of a particular pile. Page 2 is used to vary the soil, pile driver. or cushion charac. teristics for the pile deseribed on page 1. Examples of the data sheeis follow the explanation, Page 1 Case No, = Any combination of up to six alpha- botie or mimerieal characters used for fdentitying’ information. ‘These char uctors will identify all’ problems as sociated with the pile data entered on Sheets 1 and 2, No. of Probe, = Total number of problems listed on page 2.A/DELITA P = SLACK (1) SLACK (2), SLACK (3) Option 2 Option 2 = Option 8 Option 4 This space way: be Yet blank jn-most cages as the program calculates. the critical tine interval fron the para clere of the system. ‘The value cal- culated. is ‘AT = 219.608 VK/W) one, desives to tse a spe- ifie DELTA T, it may. be entered. ‘The problem will then compare the entered vale With the critical value faleulated by the above formula and tise the larger ‘of the two. ‘This is done so that the user cannot inadver= ently" enter @ value too small and Renee Introduce’ nstabiity” into the numerical process Total number of weights including ram of hammer, follower, and helmet, ‘This indicates a specified looseness Detween WI) and” W(2). in inches. This js the amount of movement Te: quired: before K(1) will take tension. If there is complete tensile. freedom of Kup, then enter SLACK. (1). = Too0. Léave’ blank if option 3 fs “2%, fee notes on Slack (1). ‘This J an option for the manual en- try of the cross secHional ares of each Segment. (a) Enter I" and all AREAS will automatically’ be 'set™ equal to 1.00. In this case, draw a horizontal line through all AR Ale portion of a {do ot ente (b) “Enter "2" if the cross sectional fea of euch segment is to be entered manually in the AREA rows, dn this care enter ARBAS (1) to'(P) inclu ‘This isan, option for the manual entry of soil resistances. (a) Enter “2” if the soil resistances (expressed as a percentage of the totai soll resistance) are to be entered manually in the RU rovs. The RU Values are entered from (1) to (Pt 1) inclusive, Note that (Pt) ie the point resistance and all others are fide! resistances. ‘The tatal of all RU percentages entered must total 100% Cb)" “Enter "1" if the aoit resistances fre not listed in the IUU rows but are indicated under Option 12 on page 2 ‘This is an option for manual entry of the SLACK values, (a), Enter “1” if SLACK values from SLACK (4) to SLACK (Pi) are all 000 (indicating K(4) to KP —"t) ean take tension). Im this ease only SLACK" (1)"t0 SLACK ("ate ea fered. in Taw 1. Draw a horizontal line through ail’ SLACK ‘ows in the lower partion of page 1. In this eaxe do not enter any" values in the Slack (5) "Enter “2 if SLACK values are 42 he entered manuaity. In this eas SLACK (i) SLACK (3) in row T ray be Heft blank ‘This is an option on the rontine used to'simulate the material behavior af springs KU), K(2), and (3). (a) Enter “1 for use of Smith's routine 3 and 4 () Enter "2" for use of ‘Tesas AGM routinet Its suggested thst Option 4 blank in whieh ense fis automatically set equal to 2 IPRINT = This Ison option on the amount of data printed out when the Tong form output is used (Option 15 = 2). It Option 15 = 2, IERINT is the print Interval expressed. as the number of me intervals, As an example, If print out Je required every 10th time Imuereal, 10 "would. be IPRINT. If Option 15 is| leave IPRINT blank = NSEG 1 is the mass number of the first pile sexment. If NSHG 1 is left blank NSEC 1 = 2 will be used by the program. ‘Tho total weight of each segment, in pounds, is centered in the rows marked W(2), Wi3},.... W(24). The weights, W's, are entered for 2 to P inclusive, Note that W(1) is not entered as it will be included on page 2. ‘The spring stiffness of each sexment, in Ib/in. is entered in the rows marked K(1), K(2), ~.-. Ki24). ‘The stiffnesses, K's, are entered from 1 19 PL incl. sive. Spring K(P) is the soil spring at the pile tip and is ealeulated by the program from the soil data entered ‘on Page 2. If Option 1 = 2, the averaze area of each segment must be entered in the rows marked A). A‘2) J. The units of A should be consistent with’ the Units desired in the output. The hasie force unit ‘of the output is the pound. The arcas, A's, are entered from 1 to P inclusive. A(P — 1) and AGP) in most instances will be the same, Areas of sezments of th hammer are usually entered as ACL) = LO. ete., since stress values obtained for these srzments are nol usually of concern. If Option 1 = 1, the area row should be marked through with a solid horizontal line indicating. no data cards are to be included. If Option 2 = 2 the side soil resistance on each, segment, expressed as a percentage of the total soil resist ance, is entered in the rows marked RU11, RU(2b RU(241. The soil resistances. RU's, are entered from 1 to P + 1 inclusive. ‘The vaiue of RU (P + 1) is the pile tip resistance. Mark out all rows when Option 2=1. If Option 3 = 2 the physical slack or looseness, expressed in inches, is entered in each row. marked LACK (11, SLACK (21, .... SLACK — (24) LACK’s are entered from 1to P’— 1 inclusive, If here is no slack, enter 0.0: if there is complete loose- ness. enter 10000, SLACK (P) is automatically. set ‘equal to 1000.0 since the point soil spring cannot take tension. If Option 3 = 1, mark out all rows. Note that the forms have 21 spaces for Wis, K's, A's, RU’s, and SLACK's, The prota is capable of handling @ pile with a maximum of 149 sezments. Ad. ditional canis may he added to each parameter as needed, Page 2 wa) = The weight of the pile driver's ram in' pounds. Ne ‘The number of the epring: for which K(NC) "ie helng’ varied Keno) ‘The spring constant of the spring bec ing varied in Ibs/in, ‘Only one spring can take on variable values per eas Option 4 may be left.EFF = The efficiency of the pile hammer. JSIDE = Deming comtant for the soil on the". ENERGY = toate fe, of the falling ram cee lated by" Equation Bf FEXP = The eel ee forse, in pounds) hie “set on the vil of ERES (1) = The. covfficent of restitation of fevaplsive fre Sears ERES (2) = The Cae of restitution of FEXP blank. spring K(2) Option 11. = This option provides for single or ERES (3) = The coefficient of restitution of multiple calculations. spring ROS) 2), Baler, 0" of pape eae, RU (TOTAL) = This space should be, used only when date ae eA VS BLOW! Option 12 = 2, In this ign suitable values of Eh ea a Gori pete HUchOTALS ace ‘11 = 1, leave this entry blank. a: Pee “ susan a ae lounge, AE TORRL) “unde the entered on page 2. Urs eetcaae’ "SY Orn tm te ct tn Mo = the bi (a) Brier 0" for wpitorm dts tion of aide tection from segment Moto ®. — a on Pa (0) Enter 2° for a triangular Q Po = coe e wie gk at the point, Nor- {ehatign of ‘ie fetion ttm ser CSIP BASH Nerant eM SGot RS Sg oF $0 Fae Oe SiuePe ta pase J PONT = Dapping constant fr the silat te Option 18. = Ths, soon rorln for commuter BCE DRIVING BNET TSS Bron Fy map TEXAS ABM UNIVERSITY ze3| i jose: —__| ore Tea #88, § TTPO eC . TAT pce LECCE eC] PER Se TOT MTP TIT PILE TOC ETTTE ETCTTT CTETTEE POUT TOTP ELC {TITEL mcr ToL vee PLCC AU eC Pc bo PT Hy PST TTS CECT CECE CCT OC TOT OEY ma Hg TT ag eh He wR faba PTET) UT TRE TPT PT) er ear ae PTO OT OO FE oP HeeATAGU On VGA URN PLANO DOT GRPPOTAEGTCFATTAGI[evATOMURT OAD VRS nGEfOanta cet ugar cg igre eo oo me FE Pe PERT TTT EA | JRC PE TET EERE Et TT} FLEET TPT LT ET) ae eee eee de PREG 2 SETTLE TTP TET ETT BRUTE Ee Tees ae rd eee ee ETT WE ano AREAS TOP CL, Hes AND SLAGKA 1 TO PNG AVE | TO Pet WELerated 6 INO) for RUCTOTAL) VS BLOW/ or RU B7 Comments on Data Input (0) ‘Biter ™1 far emputer plot of On page 2 of the input form, provisions are made ea If no plot is desired, leave for varying, ae ioe ‘of any soning. K(1) through pie K(P — 11, in the hammer or pile idealization. This is ee tte ie nee, ceeteses ecard by Ternerinat ie amebertol tthe spring to be changed in the NC column and then the stiffness of nc spring KING) jn the KINC) ‘column. As soon an this 2) Enter (2° i te forces of gravity Grobler is completed, the spring silfaesses, K (NC), wil {ions siernate in efece excludes be eset automatically to the value on paze 1 of the Gicig of hei rom ti aces input forms er fe in horizontal postion or for The program is capable of handling pile ideal an extreme batter. tions with a maximum of 149 segments. There is no Option 15 = This option provides for versatility limit on the number of problems that can be run for : Inthe: outpu format, nee Go) enter“ for a normal data Printout “7 ei for extra, detail in Semple eee an ao Printouts This alfernate gives ber ‘onsider the pile shown in Figure B-6. Slee at the pring Inveteal, Specie Pile: 16 in. square prestressed concrete pile, 26 ft as UPTINE, gh Bane 1 "in length. The modulis of the concrete je 7.62 © 108 {gp bnter <8" for short output This psi and its unit weight ie 154 Ib/It The pile is ax Maren DEOW IN VE ABCHDEREY, sumed to be embedded for ils full length. Option 13,29 should be sed ony Pile hammer: Hypothetical diesel hammer with when Option 11 = 2. 4850 Ib ram with an input ram kinetic energy of 39,800 SPECIAL NOTE. Whete anything listed for Prob- lh, The explosive fovee produced by he dive) Tucl Jem Vis to be repeated for Problem 2 3, ele. draw an is 198.700 Ih. "The stiffness of the os arrow down through the last problem to indicate repel; X10 Ib/in. ‘The anvil is assumed PILE DRIVING ANALYSIS TEXAS ABM UNIVERSITY RU (TOTAL) Kove lnesfenesfenes| oa la. wo [8 owt | lere.| enencv sie por] sive J a { | NOTE: iF oPtiow niet, AUCTOTAL) Nor REQUIREDrwaael lesa ( acruan ue we oeauizeo Pe Figure B.6. Sample problem. In order to illustrate the utilization of the input data sheets and explain the oulput data sheets, four prob lems ate considered. Problem 1 and Problem 2 are concemed with the driving effects produced by two different cushions. The object of these two eases is to determine the dynamic: static resistance curves (RUITOTAL) VS BLOWS’TN,) for one blow of the hammer. In Problem 1, the cushion is assumed to have a cross sectional area equal to that of the pile. is 644 in, thick and has a modulus of elas: lieity of 1.0. 10° pai. In Problem 2 the cushion area and properties are the same as ia Problem 1, but the thickness is 3% in. In Problem 1 and 2 the’ soil side friction is assumed to have a triangular distribution with 10% point resistance. The soil constants are: (a) Q= Y = 010 in, (b) J = 05 see. fe) ¥ = 0005 sec. /t Problems 3 and 4 illustrate the use of program to investigate the penetration of a pile to 200 tons of static soil resistance produced hy one hlow of the hammer. In Problem 3 the soil resistance is distvibuled uniformly along the side with 10% at the point. ‘The cushion is the same as in Case 1. In Problem 4 the soil has a tri angular distribution along the side with 10°% soil resi. ance (same as Problem'2).. The cushion is the same as in Problem 2. Problem 4 will aleo illustrate the use ff the output option (OPTION 15). The following calculations llustrate the computa, tions for the hammer and pile idealization, (a) Pile: The pile is broken into eight equal length segments of 39 in. The spring sliffness for cach sepment is, where AGB), = 254 in? E(3), = 7.32 X 108 psi (3), = 39 in, therefore "4 (254) (7.32_x 108) i Kaj, = SR ASA = 51.0 X 10° bin (b) Cushion: Spring K(3) in Figure B6 (b) represents the combined stiffness of the cushion and first pile segment. In Problem 1 and 3 ABI EB)e Lidl K(3)e where A()q = 254 int Eile 00 X 10° psi 6.25 in, x 108) ah pgp = 405 % 10° Ib/in, ‘The combined stiffness of K(3)q and Ki3)y is KiSie X Ki3), 405) 151.0) (10%) KO) = Rigi FRG, OS F510) 0) K(31 = 226 x 10° Ib/in The coefficient of restitution for the combined springs is assumed to be 0.50. For Problem 2 and 4 similar calculation yields K(3) = 31.3 X 10° Ib/in, The output data sheets are completed as follows: Page 1 (Same for all 4 probleme) No. of Problems on ‘page 2 A/DELTA T = 040, since the program will ealeulate the corvect value. 4, there sre 4 problems to be solved P I, there are 11 weights (3 for the hhaimmer and 8 for the pile) SLACK'S = all set equal to 1000 since there is com= Blete looseness between the ram, anvly fapbiock, pile’ cap, cushion, sid” pile head, OPTION 1 = 2 alt areas are entered manually in AREA ‘ows, OPTION 2 = 1, since OPTION 12 is used to describe the toil distribution OPTION 3 = 1, all vile segments are connected, Hence SLAGK'(4) 40 SLACK (iO) 0 OPTION 4 = lefk blank since it is desired to use the AGM rostine IPRINT 10, in Problem 4, OPTION 15 = 2, itis desired to print output every 10 itera Hons NSBGI 4, the first pile segment, see Figure Bo). ws entor the weight of each element in Ih Note that Wir) is blank since it will bbe entered on page 2xs. = enter all spring stiffnosses for the pile FEXP = 158,700, tb. the diesel explosive force. ° stem ronaidered to be Pasi, Ae he OPTION 11 =} for programe generated RU(TOTAL) titinesses to these ‘alues after euch ierieueeee red patient eae OPTION 12 = 2. for triangular side soil resistance a's enter all eress sectional areas of pile OPTION 13 = leave blank Segments only. computer plotted Page 2—Problem 1 curve is not de OPTION 14 = 1, to indicate gravity. wa) 4860 Thy the ram weight, OPTION 15. = 1, for normal data output Ne 4, the cihion spring number, see Fig- Page 2, Problem 2 _ ie Both) Only, the value of KG), is ehange, BGI" = gpupamy, tbe ate ot te com Bic) 2 Sone EFF 1.00, diesel hammers are considered to ae ‘be 100% efficient. Page 2, Problem 3 ENERGY = 20,800, the input energy for this par- ‘Te valve of K(d) and the OPTIONS are changed ticlat hammes Blow: =3 = io, cottivient of restitution of Kio) BRES(L) = OU enfient of vettaton of steel AN aan ERES(2) = 0.80, coefficient of restitution of cap- ‘RU(TOTAL) Sele Ib for a 200 ton total static soil Block Snatertal resistance. ERES(2) = 0.0, cotficient of restitution of com- «OPTION 11 = 2 for single, calculation using RU BRESC) = Fined ceenon end Hea atom fc Croan} "2" sooo. RUUTORAL) = leave blank, since OPTION, OPTION 12 = 2 for uniform aide soll resistance dis- Sit ME PREeY!S IE rag 2, Praam & at In this problem the cushion and the options are chan POINT. 10%. NC =3 ae Mo ithe first pile segment with side soil RIN) = Resistance KG) 31,800,000 urownt = bin, @ OPTION 12 = 2, for triangolar side soil resistance UsiDe 010, @ distbuton: aor = oan OPTION 15. = 2 for outnt at interval expressed JSIDE, ‘3 a a ITPRINT" on page Bite SAVING ANALYSTS TEXAS AAM UNIVERSITY Ha el RU ST BY Get esi voll el (TIT Wlrleoet | Wael | [isle | [ei COOL was we 7 ca OF “ee | Buleb oleh | Blioe\dobe| Tey baba ay aes [ates ae prises : os 4 . Ea eStats eae et eee [ee ane He Oa ete ag eee ~ Bae Leen Lee Pelee ERE eee TL ee eee Te i eae | eee eee Slee ag [ sue at | aoe | ake aa eee NOTES: one oF MORE PROBLEMS WuST BE LISTED ON PAGE @ 7 : = WE AND AREAS 170 Pick ws AND SLAGHY 1 TO P-1 MeLy FUE ITO PFI EL (Pot 1S EAU UNDER POINT OF PILEPILE DRIVING ANALYSIS: TEXAS ABM UNIVERSITY lenesjenesferes] aU crovaui was) KIN i : f i313 zksooee) hislolda; lerr.| enenoy | [es [a ‘Bhisldole lesa pe A-Aasie loans @ay67 aTTwo Lolo Layo Pwr side a i : 1 ola a ll i FexP oro hr NOTE: iF opriow wire, RU(ToTaL) NOT REQUIRED ‘The output for the four sample problems are shown ‘igures BT through Bl. Figure Bz is the output for one point on the RU(TOTAL) VS BLOWS/IN. curve generated for Problem 1. ‘The block of data on. the upper part of the figure is a printout of the input in Fi Ke data. ‘The RU(TOTAL} value of 1.090,962.1 is the total static soil resistance for which this problem was tun. This value was generated by the program and is ‘only one point of 10 used to develop the data for the total RU(TOTAL) VS BLOWS/INCH curve shown in Figure B.?. Normal output (option 1: 1) for Prob, 1.900) RUCTOTAL) IN TONS. 80 oso 885 BLOWS PER INCH Figure B.&. Effect of varying cushion stiffness. Figure BS. The second block of data shows the maxi- mum compressive and tensile stresses andl the maxinnutn displacement of each segment. The column labeled TIMEN jis the time interval at which the maximum compressive stress (MAN C STRESS) occurred, i... the maximum compressive stress of 7432 psi occurred in segment 5 at time interval 11 (11/9413.9 sec.). Similar data is printed for each point on the RU(TOTAL) VS * BLOWS/IN. shown in Figure BB. Figure B9 shows the summary of the data for the RU(TOTAL) VS BLOWS/IN. for Problems 1 and 2 Data of this type can be used to construct curves like that shown in Figure B8. ‘These curves can be used to compare the effects of cushion stiffness (the cushion stiffness, K(3)¢, in Problem 2 was twice that in Problem 1). "Note the stiffer cushion (Problem 2) produces the rmost efficient driving since for a specified resistance the penetration per blow is larger (BLOWS/IN. is smaller). Figure B-10 is a typical output when RU(TOTAL) is specified, ‘The maximum penetration of the point of the pile under one blow of the hammer is 0.473011 in. listed under DMAX(M), and the permanent set is 0.473011-0.100000 (the ground quake Q) or 0.373011 in, Note that the input data is listed’as well as the maximum stresses and the displacement of each segment. Figure B-11 is a sampling of the output when data is desired at some specified interval (OPTION 15 = 2, IPRINT = 1). The input information is listed in the first block of data. ‘The next txo blocks show the stress: ces at time interval N= 0 and N= I, -The data is defined as follows: iM) = ieplacement of cach mase_polnt, cay the compression in each spring, (in.), STRESS(Bf) = stress in cach segment, (psi), Fay force in each spring, (Ib) Ra force in each soil spring, (Th), wont) weight of each sezment, (1b), van velocity of each seament, (Eps), DPRIME(M) = elastic displacement of soil, (in Figure BY. Summary ousput for RU (total) vs blows/in, (option I1=1) for Prob. 1 and 2.Figure B10. Normal ouput for single RU (total) (option 11 = 2) for Prob. 3. KPRIME(QH = soil spring stiffness, (Ib/in), Time interval N = 0 is for the pile under the FMAXC(M) = maximum, compressive force'in seg influence of gravily alone. The particular output listed man (i) am in Figure Bell shows that the point of the pile of Prob- TCM) = Maximum “tensile force in. segment, i ’ pil ee et) * = Jem 4 would penetrate 0.002355 in, under gravity slone. Figure B11. Detailed output Jor single RU (total) (optionAPPENDIX C OS-360 Fortran IV Program Statements The listing that follows is known as an NREF liste ing, Each statement is numbered, for reference. consecu- tively from the first to the last statement. The variables and program statement numbers are indexed by their reference number. This listing facilitates finding each variable in the program and makes the logic much easier to follove A flow diagram of the program logic is included for reference.weree ence LiStimeREFERENCES 1.1 Isaacs, D. V., “Reinforced Concrete Pile Formula,” Inst. Aust. Eng. J., Vol. 12, 1981 12 Smith, E. A. La “Pile Driving Impact” Proceed: ings, Industriel Computation Seminar, ‘September, 1956, International Business Machines Corp., New York, N.Y., 1951, p. dd 13 Sith, E. A, Le “Pile Driving Anaiysis by the fave’ Equation,” Proceedings, ASCE, August, 1960. ee " 21 Isaacs, D. V., “Reinforced Conerete Pile Formula,” Inst. Aust, Eng. J., Vol. 12, 1931. 22 Smith, E. A. L., “Pile Driving Impact,” Proceed: ings, Industrial Computation Seminar, September, 1950, International Business Machines Corp., New York, N. Y., 1951, p. 44. 23° Smith, E, A. Ly, “Pile Driving Analysis by the Wave’ Equation,” Proceedings, ASCE, August, 1960. 24 Dunham, €. W., Foundations of Structures, Mee Graw-Hill Book’ Company, New York, 1963. 25 Chellis, R. D., Pile Foundations, McGraw-Hill Book Co., New Yerk, 1951. 26 Fowler, J. W., Project,” Civil Chellis, RD. Pile Foundations, McGraw-Hill Book Co., New’ York, 1951. 28 Leonards, G. A. Foundation Engineering, Me- Graw-Hill Book Co, New York, 1962. 29 Janes, R. L.. “Driving Stresses in Steel Beating Piles." Dissertation at Illinois Institute of Tech: nology, June, 1953. thesapeake Bay Bridge-Tunnel ‘neering, December, 1963. 210 Cummings, A. E., “Dynamic Pile Driving For rmulas,” Journal of the Boston Society of Civil Engineers, January, 1940. 21 Gardner, S. V. and Holt, D., “Some Experiences with Prestressed Concrete Piles.” Proceedings of the Institute of Civil Engineers, London, Vol. 18, January, 1961. 212 Glanville, W. H., Grime, G., Fox, E.1N., and Davies, W. W., “An Investigation ‘of the Stresses in Rein: forced’ Concrete Piles During Driving.” British Bldg. Research Board Technical Paper No. 20, DSLR, 1938. 213 Helsing, W. P., Discussion of “Impact -and Lonei- tudinal "Wave Transmission” by E, A. L, Smith Transactions, ASME. August, 1955, p. 963. AL Explosion Adds Driving Fare te Mins Nessie) Sarat eeatak Ie December, 1960, 2 Housel, W. S, “Pile Load Capacity: Estimates and Test Res Foundations’ Division, AS September, 1965, is" Journal of the Soil Mechanics and Proc. Paper 4433, 33 61 62 63 64 65 66 67 al a2 83 ad Lowery, L. L., Hirsch, T_ J. and Samson, C. H., “Bile Driving’ Analyae—Simulation of Hammers Cushions, Piles and Soils,” Research Report 33-9, ‘Texas Transportation Institute, August, 1967. Chan, P. C., Hirsch, T. J. and Coyle. H, M., “A Laboratory Study of Dynamic Load-Deformation and Damping Properties of Sands Concerned with a Pile-Soil System,” Texas Transportation Institute, Research Report No. 33-7, June, 1967. Reeves, GN, Coyle, H. M. and Hirsch, T. J. “Investigation” of Sands Subjected to Dynamic Loading.” Texas Transportation Institute, Research Report’ No. 33-7A, December, 1967. Airhart, T. P., Hirsch, T. J. and Coyle, H, Ma, “PileSoil System Response in Clay as a Function of Excess Pore Water Pressure and Other Soil Properties,” Texas Transportation Institute, Re- search Report No, 33-8, September, 1967. Chellis, R. D., “Pile Foundations,” McGraw: Book Co, New York, 1951. Analysis by the * Transactions, ASCE, Paper No. 3306, Vol. 127, Part 1, 1962. Forehand, P, W. and Reese, J. Ls. “Pile Driving Analysis Using the Wave Equation,” Master of Science in Engineering Thesis, Princeton Univer- sity, 1963. Mansur, C. I. “Pile Driving and Loading Test, Presented at ASCE Convention, New York, Octo- ber 22, 1964. Lowery, Le ny Jr Edwards, L, C., and Hirech T. 1, “Use of the Wave Fquation to Predict $0 Resistance on a Pile During Driving.” Texas Transportation Institute, Research Report 33-10, August, 1968. Moseley, Ez T., “Test Piles in Sand at Helena, Arkansas. Wave Equation Analysis,” Foundation Facts, Vol. 3, No. 2, Raymond International, Con- exete Pile: Divison, New York, 1967. Samson, C. H., Hirsch, T. J. and Lowery, L, Ly “Computer Study of The Dynamic Behavior of Piling,” Journal of the Structural Division, Pro- ceilings, ASCE, Paper No. 2608, ST4, August, 1963, p. 413, Heising, W. P., Discussion of “Impact and Longi- tudinal Wave “Transmission” by E. A. 1. Smith, Transactions, ASME, August, 1955, p. 963. Smith, BALL Ws a 102, Part ly pe U4 “Pile Driving A: ysis by the VEEL, Vol. LZ, Hirsch. T. J., “A Report on Stresses in Long Pre- stressed Concrete Piles During Driving,” Research Report No. 27, Texas Transportation Institute, September, 1963.85 86 91 92 98 4 Hirsch, . J., “A Report on Field Tests of Pre- streseed Concrete Piles Duving Driving,” Progress Report, Project No, 25.6233, Texas Transports: tion Institute, August, 196: Lowery, L. L., Hirsch, T, J, and Samson, C. H., “Pile Driving’ Analysis—Simulation of Hammers, Cushions, Piles and Soil:” Research, Report Ne. 339, Texas Transportation Institute, August, 1967, Hirsch, T. J., “A Report on Computer Study of Variables which Affeet the Behavior of Concrete Piles: During Driving,” Progress leport, Project No. 25.6233, Texas Transportation Institute, August, 1968. Hirsch, TJ, and Samson, C. H. “Driving Prace "Research Re- 3-3, Texas Transportation Institute, April, McClelland, B., Fockt, J., and Emrich, W., “Prob: Jems in Design and Installation of Heavily Lo ed Pipe Piles,” Presented to ASCE Specialty (i ference on Civil Engineering in the Oceans, San Francisco, September, 1967. Lowery, L. La Hirsch, T. J. and Edwards, T. C. “Use of the Wave Equation to Predict Soil Resi ance on a Pile During Driving,” Research Report No, 88-10, Texas Transportation Institute, April, 1967. Lowery, L. L, Hirsch, T, J. and Samson, C. H., “Pile Driving Analysis—Simulation of Hammers, Bi. B2, BB. Bs, Cushions, Piles and Soils,” Research Report No. « 33-9, Texas Transportation Institute, August, 1967. Smith, E. A, Ly “Pile Driving Analysis by the Wave Equation, Journal of the Soil Mechanics and Foundations “Division, Proceedings of the American Society of Civil Engineers, Proc. Paper 2574, SM4, August, 1960, pp. 35-61 Hirsch, T. J., and Edwards, T. C., “Impact Loade Deformation” Properties of’ Pile Cushioning Ma. ils,” Research Report, 33-4, Project 25-62-33, Piling Behavior, Texas ‘Transportation Institute, Texas AGM University, College Station, Texas May, 1966, p. 12. Smith, E. A. Ly “Pile Driving Analysis by the Wave’ Equation,” Journal of the Soi! Mechanics and Foundations Division, Proceedings of the ‘American Society of Civil Engineers, Proc. Paper 2574, SM4, August, 1960, p. 47. Samson, C. H., Hirsch, T. J.. and Lowery, L, Ly “Computer Stily of Dynamic Behavior of Piling,* Journal of the Structural Division, Proceedings of ithe American Society of Civil Engineers, Proceed- ings Paper 3608, ST4, August, 1963, p. 419. Smith, E. A. L,, “Pile Driving Analysis by the Wave Equation,” Journal of the Soil Mechanics and Foundations Division, Proceedings of the ‘American Society of Civil Engineers, Proc. Paper 2574, SM4, August, 1960, p. 44.