CHAPTER - 2 LITERATURE SURVEY AND REVIEW

Transcription

1 CHAPTER - 2 LITERATURE SURVEY AND REVIEW 2.1 INTRODUCTION: In past, some results of theoretical and experimental studies on dynamic loosening of threaded fasteners have been reported. However, a detailed review of these studies is essential especially with reference to soft foot condition. Hence, a literature survey was carried out in this area. Various technical papers articles published in the Journals of professional societies, papers published in Proceedings of the National/ International conferences from various organisations such as, ASME, JSME, I.Mech.E., Elsevier etc. are referred for this purpose. Also, the Handbooks in the field of Machine Design, Reliability, Maintenance, Statistics etc. and books related to Joints, Fasteners, Design of Experiments etc., have been studied. A number of websites such as etc. available on Internet have also been visited. Along with this search for literature, some industries have been visited to get the first hand information about the industrial practices to minimize the effects of soft foot condition [Refer Appendix I]. Also discussions with some resource persons have been carried out on the problems concerned with design and dynamic loosening of threaded fasteners. The literature review has been carried out mainly on following aspects, (i) (ii) (iii) (iv) (v) (vi) Bolted Joint Design Historical Perspectives Threaded Fastener Preload Tightening Tools and Techniques for Threaded Fasteners Dynamic Loosening of Threaded Fasteners Thread Locking Tools and Techniques Soft Foot Development in Machine Structures and its Effects on Dynamic Loosening of Threaded Fasteners 2.2 BOLTED JOINT DESIGN HISTORICAL PERSPECTIVES: From the dawn of humankind (in fact, maybe even before), the ability to join similar or dissimilar materials has been central to the creation of useful tools, the manufacture of products, and the erection of structures. Joining was undoubtedly one of the first manufacturing technology. With the passage of time, the need for and benefits of joining have not abated; they have grown. More diverse materials were fabricated into more 8

2 sophisticated components, and these components were joined in more diverse and effective ways to produce more sophisticated assemblies. Today, from a Wheatstone resistance bridge to the Whitestone suspension bridge, from missiles to MEMS, joining is a critically important consideration in both design and manufacture [4]. Most machines and structures are assemblies of simpler components. Fastening and joining technology used to assemble the components is a key feature of practically all modern machines. Threaded fasteners are probably the best choice to apply a desired clamp load to assemble a joint, at a low cost, with the option to disassemble if and when necessary. Furthermore, the simplicity of its mechanism of developing and maintaining the desired clamp force made it very popular and it has become one of the most accepted engineering products. There is no record of the invention of the first threaded fastener, although the concept of the screw thread was known by Archimedes (d. 212 B.C. ) who used it in a water lift device. There is little doubt that the inventor of the first nut to complement a screw had no idea of the revolution he began. Today, literally billions of threaded fasteners are used throughout industry, in a wide variety of sizes and shapes. It has been said that one automobile manufacturer a few years ago used 11, 500 different fasteners in his product [5, 6]. A reliable joint must be both strong and stable in the operating environment. It should prevent slip, separation, vibration, misalignment, and wear of parts. Joints and bolts must deflect elastically, but they must also be rigid enough to provide structural integrity. Rigid joints result from specifying enough bolts, adequate bolt diameters, strong enough bolt and joint materials, and sturdy joint members [7]. Following brief review gives the information about some of the references concerned with the issues of bolted joint design. In the past, assemblies were routinely overdesigned, using more and larger fasteners than necessary, to provide up to 10 times actual load-bearing requirements. As weight and cost reduction become competitive necessities safety factors have been reduced, and fasteners are being subjected to higher stresses and loads. There has been a concerted effort within the International Standards Organization (ISO) to develop procedures for the design and evaluation of fastener product standards on a sound technical basis. Alexander [8] has described methods accepted by Technical Committee- 2 of ISO for the strength design of mechanical fasteners employing the ISO R68 thread profile. 9

3 Kulak et al. [9, 10] have provided a state-of-the-art summary of the experimental and theoretical studies undertaken to provide an understanding of the behavior and strength of riveted and bolted structural joints. Design recommendations are provided for both allowable stress design and load factor design. In both cases, major consideration is given to the fundamental behavior of the joint and its ultimate capacity. They have also contributed to create a primer for the structural engineer with the necessary information to select suitable high-strength bolts, specify the methods of their installation and inspection, and to design connections that use this type of fastener. Barrett [11] has written a manual for design engineers to enable them to choose appropriate fasteners for their designs. Subject matter includes fastener material selection, platings, lubricants, corrosion, locking methods, washers, inserts, thread types and classes, fatigue loading, and fastener torque. A section on design criteria covers the derivation of torque formulas, loads on a fastener group, combining simultaneous shear and tension loads, pullout load for tapped holes, grip length, head styles, and fastener strengths. The second half of this manual presents general guidelines and selection criteria for rivets and lockbolts. Chaddock [12] has also provided similar information along with the list of relevant British Standards for threaded fasteners. Brown et al. [13] have prepared a technical report for SANDIA Laboratory, which provides general guidance for the design and analysis of bolted joint connections. An overview of the current methods used to analyze bolted joint connections is given. Several methods for the design and analysis of bolted joint connections are presented. Guidance is provided for general bolted joint design, computation of preload uncertainty and preload loss, and the calculation of the bolted joint factor of safety. Axial loads, shear loads, thermal loads, and thread tear out are used in factor of safety calculations. Additionally, limited guidance is provided for fatigue considerations. An overview of an associated Mathcad Worksheet containing all bolted joint design formulae has been provided. Proper design, assembly, preload and inspection of bolted connections remains an important activity for operators of any manufacturing or process commercial plants. Nuclear Maintenance Application Centre s Bolted Joint Maintenance and Applications Guide [14] have addressed the general good bolting practices and guidelines for threaded fasteners usage. The Guide is subdivided by major application into pressureretaining joints, mechanical joints, and structural joints. Additional information on 10

4 procurement and fastener receipt inspection is also included. This document is useful to plant engineering and maintenance personnel responsible for procedures, assembly, inspection, and troubleshooting the various types of bolted connections used in manufacturing and process plant applications. Yakushev [15] has provided the information on the effect of manufacturing technology and basic thread parameters on the strength parameters together with practical recommendations for increasing the fatigue strength of threaded connections. One can get the state-of-the-art information about the threaded fasteners, the current trends of research and new developments in fastening technology from the latest manufacturers catalogues and handbooks [16-21]. 2.3 THREADED FASTENER PRELOAD: Threaded fasteners are commonly used in assemblies due to the advantages such as the ability to develop a clamping force, and the ease of disassembly for maintenance and repair. Clamping force in a bolt is commonly developed by turning the engaged nut such that it moves against a clamped component and causes an axial elongation in the bolt. The resulting clamping force is a function of the joint stiffness and the bolt axial elongation. In general, joints must be preloaded enough so that external loads do not cause conditions known as gapping and tension release. Gapping is the separation of bolted parts caused by a force large enough to pull the joint apart. Tension release results when an external force compresses the joint members, causing the bolt looseness. Either condition can cause premature fatigue failure, part misalignment (possibly reducing preload), joint instability. Setting the preload to avoid gapping and tension release requires an understanding of how the joint behaves during preloading and under applied stress. In simple terms, a preload produces stresses within parts that help strengthen the joint [22]. Bickford [23] from his study has suggested four statistically significant factors having impact on the preload achieved in a bolt when torque is applied to it. (i) Lubricity of the fastener s contact surfaces (ii) Condition of the parts (surface finish, thread damage, presence of rust etc.) (iii) Part fit (male-to-female threads and fastener-to-hole) (iv) Tools and procedures used to assemble the joint He also felt that the preload produced in a bolt by the applied torque is not necessarily the final goal of the assembler. What needs to be known is the clamping force created between joint members. 11

5 Haviland [24] has explained the preload generation process as, when a screw or bolt is tighten by applying torque to the head. A clockwise torque makes the bolt-to-nut distance shorter. If a resistance is met (such as clamping a flange), the bolt will continue to rotate until a balance is obtained between the torque applied to the head and the sum of bolt tension and friction. According to him the equilibrium relationship is given as, T = K d F i (2.1) Where, T = Torque (N.m), d = Nominal diameter of bolt-in. (m), F i = Preload or clamp load (N), K = an empirical constant, which takes into account friction and the variable diameter under the head and in the threads where friction is acting. Values of K can be determined experimentally. The variation in friction and, therefore, K is wide since it is the result of extremely high pressure between surfaces that may be rough, smooth, oxidized, chemically treated, and/or lubricated. K varies between 0.11 and 0.17 or ± 20 percent. In majority cases, the bolts are tightened by means of torque wrench and then the preload is dependent upon the coefficient of friction between the screw threads of the nut and bolt. Lambert [25] has considered the effect of normally distributed variations in the coefficient of friction for developing Torque-Tension relationship, which given as, ( ) ( ) α = semi-angle of thread form r = mean radius of screw thread r m = mean radius of nut seating (or bearing radius) t = Number of threads per unit length μ t = Coefficient of friction between screw threads μ b = Coefficient of friction between mating surface and bolt under-head This shows that T/F i is approximately linear with μ t and μ b. The coefficient of friction mainly depends upon a number of factors such as; (i) The method of manufacture and surface finish of the threads. (ii) The degree of lubrication and nature of the lubricant. (iii) The number of times the bolt has been previously tightened. Mean value of coefficient of friction = 0.15 and Practical range 0.1 to 0.2 i.e. ± 33 % are observed during the study. 12

6 Porter [26] also has similar findings from the formula derived for Torque-Tension relationship using US Military Handbook (MIL-HBK-60, 1992, Appendix B), applicable to threads with 60 degree flank angles. ( ) ( ) Where, T = Torque applied to fastener, T p = Self-locking nut prevailing torque, p = thread pitch (mm), μ t = coefficient of thread friction and d m = nut base mean (friction) diameter It is essential that for a given fastener system, the variation in thread pitch, nut base diameter, and pitch diameter should only make minor contributions to preload variation. The major contributors are the applied torque, and the two coefficients of friction. By comparing the developed tension values for different bolt sizes and material grade from the formula with the bolt ultimate strength, it is seen that the preload varies from approximately 22% to 89% of the rated nut strength The Degree of Tightening The degree of the initial tightening is usually selected according to the requirements of tightness for the joint. Available design practice recommends tightening of bolts until maximum induced nominal stress reaches a value not greater than 0.7 times the yield strength of bolt material. Generally, in designing a joint, a minimum preload of 80% of the minimum guaranteed bolt yield strength is considered sufficient. In some cases, because of elevated temperatures and differing material coefficients of thermal expansion, a lower level may be recommended. The actual preload applied during assembly should be higher than the amount specified. Generally, fasteners of grade 8.8 or higher can take a preload much higher than 80% of the minimum guaranteed yield strength. This is because their actual strength is significantly higher than the guaranteed strength, and also because their tenacity is some 20 times higher than the local yielding in the bolt or nut. Second, new fasteners have built-in stresses that are relieved when the joint is loaded first time. Accordingly, bolted joints should always be retightened after proof testing, or early in the service life. [Webjorn, 27] Recent studies recommend tightening of bolts up to plastic. The main advantages of inducing such a high preload in bolt are the reduction of the amplitude of alternating force in bolt and the uniformity in load distribution between threads in contact. It is to be noticed that when plastic deformations take place the nut cannot be easily assembled after it has been disassembled. Therefore, enough clearance must be provided in the 13

7 original dimensions to provide for the plastic deformation if a bolt is to be preloaded to plastic deformation and subsequently reassembled. Although the bolt fatigue strength decreases as the plastic lengthening progresses, thread rolling after heat-treatment and well finished thread root radius make it possible to provide satisfactory fatigue strength in the plastic region. From the mathematical formulation of the tightening process, a diagram was developed by Motosh [28], which enables the determination of the necessary tightening torque, for a certain bolt, to induce the maximum utilizable preload in bolt upto the plastic range. The charts are constructed with the ranges for the coefficients of friction: μ t up to 0.5 and μ b up to 0.3. Bray and Levi [29] have studied the tightening characteristics of bolt-nut-washer assemblies for torque-tension relationship. The test program included tightening bolts up to the yield point under a planned series of combination of factors and levels as given in table 2.1. A simple bolt-testing machine was built for the tests. A motor-driven wrench with a strain-gage bridge measuring the input torque tightened the nut at low speed while a two-component strain-gage transducer measured bolt torque and tension. Provision was also made for measuring the angle of rotation of the nut and the torsion of the bolt s shank. The tests were planned and analyzed with statistical techniques in order to obtain estimates of the effects of each factor and the more important interactions as well as of the errors to be considered for the tests of significance. A comparison between the torque wrench and the turn- of-nut method of bolt tightening is made in terms of closeness of bolt tension control. From the experimental analysis it is seen that the higher loads can be induced in the bolt by tightening the nut before reaching the yield point with the nuts cadmium plated, Molykote G lubrication and soft nuts cadmium plated or hard nuts zinc plated. The range of axial loads at yield point is from 0.8 to 1.2 the average value. Table 2.1: Factor and Level Matrix for Torque-Tension Relationship Factors Level 0 1 A Bolt surface treatment Cadmium plating Zinc plating B Nut surface treatment Cadmium plating Zinc plating C Nut material nominal yield strength (kg/mm 2 ) D Washer treatment None Case hardening E Lubrication Light motor oil Molykote G F Fit between mating threads Loose Tight 14

8 Croccolo et al. [30] have developed an experimental tool and procedure useful to define accurately the friction coefficients in bolted joints and, therefore, at relating precisely the input bolt torque T to the bolt preloading force F i. The components under investigation were clamped joints made of aluminium alloy used in front motorbike suspensions to connect steering plates and legs or legs and wheel pin (static failures occurred during the tightening phase of such components because of the bending stress produced by preloading forces in the compliant part of the clamp). In order to evaluate accurately the friction coefficients and to gather their effective mathematical expression, the Design of Experiment (DOE) method was applied. At first, a screening analysis was performed: the applied torque T (5 20 Nm), the bolt type (hexagon flange bolts) and the bolt diameters (M6 1 or M8 1.25) are not significant in changing the friction coefficient values. The bolts used are zinc plated and realized with a product grade B while tightening was always performed at room temperature. In light of the screening tests a full factorial plane, characterized by four variables with two levels each, was designed and given in table 2.2. Three replicas were carried out, in order to reduce the influence of noise (experimental error) and any non-investigated factors, a total of = 48 experimental tests were carried out. Table 2.2: Summarized DOE Parameters Variable Low level (0) High level (1) A. Lubrication Unlubricated Lubricated B. Forming process Cast Forged C. Surface finishing Spray-painted Anodized D. Number of tightening First tightening (Unspoiled surfaces) Sixth tightening (Spoiled surfaces) From the experiments carried out it was observed that the friction conditions are significantly affected by surface finishing, lubrication and the number of tightening and loosening, whereas the forming process (forged or cast aluminium alloy) seems to have no significant influence. Nassar and Sun [31] have introduced a novel experimental procedure and test set-up for studying the effect of surface roughness on the torque-tension relationship of threaded fastener (M12 size). Three levels of surface roughness (N4, N5, and N6) are considered for the fastener underhead and the joint surface (as-received versus clean and dry). The study was conducted for two joint materials (steel and aluminium), two fastener classes (8.8 and 10.9), and for coarse and fine threads. The torque-tension data were expressed in terms of the value of the nut factor (k) as well as its scatter. Also the effect of the number of tightening on surface roughness and on the torque- 15

9 tension relationship was investigated. The surface roughness is measured before tightening, and after each loosening using a WYKO optical profiling system. This study would aid design engineers in the process of developing more reliable torque specifications for the tightening of critical bolted joints, by taking into consideration the effect of the surface roughness and condition. Groper [32] has developed a preload measuring device based on the principle of stretch-control method. The device utilizes elastic properties of steels to measure the preload in fasteners. The device a 'measuring fastener,' has an axial hole which carries a non-loaded pin. The pin is attached, at some convenient location to the body of the fastener. It is presented for observation at the fastener's head so that elongations are observed as a change in the relative position of fastener's head and the free end of the pin. The load induced in the fastener readily measured at the time the fastener is assembled by applying a dial indicator. The calibration of the measuring fasteners shows that the elongation δ is proportional to the preload, but slightly larger than the elongation given by equation δ = F il/ae. Some 'mushrooming' effect of the fastener's head contributes to an increasing number of deflections. From calibrations it is found that the preload (F i) is proportional to the elongation δ, ( ) Where, F i = Bolt Preload, A = Bolt cross-section area, L = Bolt Effective Length, E = Modulus of Elasticity of Bolt Material, δ = Bolt Stretch and C = factor larger than unity which represents the 'mushrooming' effect of the head. For usual hex- headed fasteners it is found that 1.05 < C < TIGHTENING TOOLS AND TECHNIQUES FOR THREADED FASTENERS: Bolt tightening tools and procedures are the significant factors playing vital role in development of appropriate amount of bolt preload. Extensive investigations carried out on torque-tension relationships prove that under most uncontrolled situations using torque as a measure of tension can lead to a error as large as ± 50%. Even under controlled conditions torque on its own is not a reliable measure of tension. On the other hand, the reliable preload measuring systems are cumbersome and expensive. A comparison of various methods available for achieving preload in terms of their reliability and relative cost is given in Figure 2.1 [33]. 16

10 2.4.1 Methods of Bolt Tightening There are two methods of tightening screws, the elastic region tightening method and the plastic region tightening method. Generally, because of easy control, the elastic region tightening method (the torque method) has been used. On the other hand, from the standpoint of the axial tension necessary for tightening, the plastic region tightening method, which is little influenced by friction and allows the maximum bolt capacity to be used, is superior to the torque method. Also, it has been clarified that there are no problems related to addition of external force after tightening [34, 35]. However, the plastic region tightening method has not come into wide use since the reliability of control method and the method of checking quality after tightening are not adequate for introduction into a mass production line. Figure 2.1: Comparison of Common Tightening Methods [33] In the torque controlled method, the axial load in the bolt varies largely due to high variations in the tightening torque and coefficient of friction, and consequently the tightening operation must be controlled to prevent the maximum axial force from exceeding the bolt yielding point. This causes a large variation in axial load of the bolted joints, giving rise to excessively high or low axial loads, thus posing a problem in controlling the bolted joint quality. In the case of the plastic region tightening method, the frictional effect is largely eliminated, and this dependence on the bolt yielding point results in lower variations in bolt load. The plastic region tightening is a method to preload the bolt into beyond its yield point. Two techniques are available for bringing the bolt load accurately into the plastic region. One is called the gradient- controlled 17

11 method in which the ratio of change in torque moment (ΔT) to change in torque angle (Δθ), (i.e., the torque-angle ratio, ΔT/Δθ) is monitored during bolting. Another is the angle-controlled method utilizing the nut snug torque T snug. The tests conducted by Yamashita et al. [36] indicate that the plastic region tightening method makes it possible to raise the minimum bolt load by 50% and reduce the bolt load variation to about one half as compared with the torque controlled method. Since torque controlled tightening is highly influenced by friction, the snug torque should be set as low as possible in order to minimize the influence of friction. There is a high probability that the recommended snug torques will ensure that the gaps in the investigated joint are closed and that there will not be any rupture or plastic elongation of the screw before applying the desired tightening angle. Toth [37] has proposed a method of utilizing a screw over its yield point even under simple tightening conditions using Monte-Carlo simulations. The simulations consider the probabilistic nature of different variables necessary to calculate snug torques required for screwed joints. The proposed torque and angle controlled tightening technique can be used in assembly plants as well as in small workshops. The technique predicts permanent elongation, maximum tightening angle, final torque and preload after tightening. Further, a recommendation on the minimum and maximum snug torques prior to application of the angle over the yield point of a screw has also been presented. ( ) ( ) (2.5) ( ( ) ) ( ) ( ) Where, δ = bolt elongation, k j = spring constant for joint, k b = spring constant for bolt, dm = bolt nominal diameter and d b = bearinf pressure diameter. Toth [38] also performed similar analysis using Taylor s series expansions, and he observed almost identical results compared with Monte Carlo simulations. Monte-Carlo simulation is preferable if a computer is used for the calculations, while the Taylor s series expansions can be used if the calculations are to be done without the help of a computer. Junker and Wallace [39] have proposed a methodology of determination of fastener dimensions as a function of the tightening method is based on the concept of tightening factor (α A), which is defined as the ratio of the highest to the lowest assembled preload, i.e. F imax / F imin. Based on the experiments conducted they have determined the value of 18

12 Torque Measurement Twist-of-nut Turn-of-nut Bolt Heating Hydraulic Tensioning Strain Gauge Micrometer Measurement Built-in Strain Gauge Load Indicating Washer tightening factor for torque-control as 1.7 and 1 for angle-control and yield-control. The procedure begins by assuming that the design loads are known. It then relates the maximum preload F imax that can be given by a fastener to the minimum preload, F imin that is required and can be obtained with any particular tightening method: F imax = α A F imin. The next step is to establish the screw diameter that has a load capacity, F sp (proof strength), equal to or greater than the required maximum preload, F imax. Generally the load capacity of the screw, F sp for torque-controlled tightening is that which gives combined stresses that are 90 percent of the yield point of the screw material. From the calculation it is observed that the tightening factor, which exhibits the scatter in the mean preload plays vital role in the determination of bolt size based on a particular tightening method. Kerley [40] in his report has presented a matrix (refer table 2.3) giving guidelines for selecting an appropriate bolt tightening method. Table 2.3: Matrix for Selecting the Right Tension Control Method [40] Requires special components Requires special tools Low cost Sensitive to Misalignment Reusable Needs loosening to check Special access requirement Minimum size limitation Single load value Requires stress calculation Can be used in confined space Requires extra bolt length Can give remote indication Sensitive to variation in thread length A pocket Guide for Tightening Technique [41] prepared by Atlas Copco, demonstrates the application of power tools for the bolted joint assembly and the influence of tool selection on the quality of the joint. 19

13 2.5 DYNAMIC LOOSENING OF THREADED FASTENERS: It has been accepted that fasteners turn loose when subjected to dynamic loads in the form of shock, vibration or cyclic thermal loading. This reduces the clamping force and leads to joint failure. Such failures result in higher maintenance expense, costly downtime in machines, and can be catastrophic in safety critical applications such as boiler feed pumps, lubricant or coolant circulating pumps, rolling mills etc. Perhaps the incomplete understanding of the loosening process is responsible for the lack of formal quantitative design guidelines to avoid loosening in spite of the widespread nature of the problem. General guidelines followed for minimizing the problem include using bolts with large length to diameter ratio, using high preloads to avoid slip between joints, reorienting the joint such that the fastener is subjected to axial loading instead of shear loading, and adding features to minimize slip at the joint. Solutions such as reorienting the joints or redesigning the joint are not always feasible, and effectiveness of the other guidelines is generally difficult to evaluate. As a result, the most widespread solution is the use of various commercially available "anti- loosening" products. However, studies on effectiveness of such features have pointed to limitations; such as add to the cost, weight as well as assembly and disassembly time [42]. Pai and Hess [43] have shown that the primary cause for loosening is a result of slip between the mating fastener surfaces, which is explained by a block on an incline system shown as shown in figure 2.2. Figure 2.2: Block on incline system [43] In such a system, when sufficient force is applied in the transverse direction, the block not only moves in the transverse direction, but also downwards. In a threaded fastener, the lead of the thread provides the direction for decreased potential energy, while the external loads provide the additional loads required to cause slip. Loosening by definition requires slip between the fastener surfaces, more specifically; it requires slip in the circumferential direction in the loosening sense. Sakai [44, 45, 46] has carried out a theoretical analysis on the bolt loosening mechanism on the transversely loaded joints whose clamped parts slip relatively, and some 20

14 experiments were also carried out. It was observed that the necessary condition tor the bolt to rotate loosely with the clamped parts slipping relatively is that the friction coefficient (μ t) be smaller than As the self-supporting condition of the screw thread is μ t > 0.017, a bolt with the clamped parts slipping relatively can rotate loosely, even though the self-supporting condition is satisfied. The friction coefficient of fasteners (μ b) with a relative slip occurring on the bearing surface was measured to be It has been verified experimentally that the friction coefficient under relative slip can become small enough for the bolt to rotate. For the friction coefficient to satisfy the above-mentioned bolt loosening condition, it is necessary for a relative slip to occur on the bearing surface. Based on this fact, the critical slippage is defined as the minimum slip between the clamped parts needed for the relative slip to occur on the bearing surface. Junker [47] has developed vibration machine (later it becomes milestone in the analysis of vibration loosening) which generates transverse forces and displacements in preloaded joints as well as combinations of transverse and axial forces. Displacement, transverse forces, preload, and angle of rotation can also be measured. Additionally, the fatigue strength of bolted sheet metal joints can be evaluated, which is a means of rating the damage that a locking element causes to the surface of the clamped parts. Pai and Hess [48] have further developed the understanding of loosening under shear loading originally studied by Junker. For slip to occur, the friction force between two contact surfaces must be overcome by the resultant force tangent to the interface. For slip, it follows that loosening of a threaded fastener will occur if the following two conditions are satisfied at all the fastener contact surfaces: (i) at least one of the forces acting at the contact surfaces acts in the loosening direction (i.e., there is a loosening moment acting on the fastener) - The loosening moment required for loosening is developed due to the circumferential component of the reaction force around the thread helix. The preload, the applied shear force, and the bending moment contribute to the reaction force at the threads that develops the loosening moment. In addition, torsional energy stored during tightening can contribute to initial stages of loosening, and (ii) the resultant of all forces acting tangent to the contact surface overcomes the friction force - Factors that contribute to overcoming the friction force are the shear force, the bending moment, and the elastic deformations of the contacting bodies. In addition to loosening caused by complete slip as revealed by 21

15 Junker, the analysis presented shows that the loosening in fasteners can also result from the accumulation of localized slip in the form of elastic deformation. Daadbin and Chow [49] have proposed a theoretical model to study the loosening caused by the impact loads and investigate the effect of parameters which intensify its cause. The parameters which intensify its cause are, duration of the applied force, initial preload, lead angle and friction Self-Loosening of Bolted Joints Zadoks and Yu [50] have examined the self-loosening behavior of a transversely loaded bolted connection. They used Hertz contact theory to determine the relationship between contact deformation and contact force between two elastic bodies (mass and bolt shank) when impacts occur. A two degree of freedom model of a bolt and a clamped mass was established from which the time histories of the relative displacement and velocity between the mass and the bolt could be obtained. Additionally models of the variations of the friction torques between the mass and the bolt head and between the male and female thread forms as functions of the relative velocities have been presented. The self-loosening process of a bolted joint consists of two distinct stages. The early stage of self-loosening is due to the cyclic plastic deformation of the materials. The second stage of self-loosening is characterized by the backing off of the nut. Jiang et al. [51, 52, 53] have focused on an experimental investigation of the second stage of selfloosening. Over one hundred bolted joints with M specification bolts and nuts were experimentally tested using a specially designed testing apparatus. The experiments mimicked two plates jointed by a bolt and a nut and were subjected to cyclic transverse shear loading. During an experiment, the relative displacement between the two clamped plates was a controlling parameter. For a given preload, the relationship between, relative displacement between the two clamped plates, the amplitude of the relative displacement between the two clamped plates, and the number of loading cycles to loosening followed a pattern similar to a fatigue curve. A larger initial clamping force will results in a higher self-loosening resistance. However, a larger initial clamping force may lead to fatigue failure of the bolt. Zhang et al. [54] have conducted an experimental investigation to study the effects of clamped length and loading direction on the self-loosening behavior of bolted joints by using specially designed fixtures. The experiments mimicked two plates jointed by an M property class 10.9 bolt and a nut. The joints were subjected to cyclic 22

16 external loading. A constant preload of 25 kn was used for all the experiments conducted. During an experiment, the relative displacement between the two clamped plates, δ, was a controlling parameter. The reduction in clamping force, the applied transverse load, and the nut rotation were measured cycle by cycle. The relationship between, Δδ/2, the amplitude of the relative displacement between the two clamped plates, and, N L, the number of loading cycles to loosening is referred to as self-loosening curve and was obtained for different clamped lengths and applied load directions. Similar to a fatigue curve, an endurance limit has been identified from the self-loosening curve. It is seen that increasing the clamped length can enhance the self-loosening endurance limits in terms of the controlled relative displacement of the two clamped plates. However, the load carrying capability was not influenced significantly due to the thickness of the clamped plates. Nassar and Matin [52, 53, 54] have determined the amount of clamp load loss due to a fully reversed cyclic service load is for a bolted assembly in which the fastener and the joint were both tightened initially beyond their respective proportional limits. The cyclic reversed load acts in a direction parallel to the bolt axis. During the first half of each cycle, the cyclic load acts as tensile separating force that increases the fastener tension further into the nonlinear range; it simultaneously reduces the joint clamping force. Thus, after the first one half of the cycle, the clamp load is reduced from its initial value due to the plastic elongation of the fastener. During the second half cycle, the cyclic load compresses the joint further into the plastic range; simultaneously, it reduces the fastener tension. Due to the permanent set in the compressed joint, the clamp load is decreased further at the end of the second half cycle of the service load. Variables taken for this study were the joint-to-fastener stiffness ratio, the ratio of the initial fastener tension to its elastic limit, and the ratio of the external force to its maximum tensile value that would trigger joint separation. From newly developed formula based on the study, it is clear that the amount of clamp load loss is significantly affected by the stiffness ratio of the joint and the fastener, the amplitude of the external force, the level of fastener preload, and by the rate of strain hardening of the fastener and joint materials. Nassar and Housari [58, 59] have studied the effect of the hole clearance and thread fit on the self-loosening of threaded fasteners that are subjected to transverse cyclic loads. Tests were conducted on testing machine that is a modified version of Junker machine. The bolt used in this study is hex head ½ -13 grade 5 bolts; with a fixed thread fit class 23

17 of 2A. Three hole clearance values were used; namely, 3%, 6%, and 10% of the bolt nominal diameter. Three different nut thread fits for the ½ -13 nut size were investigated; namely, 1B, 2B, and 3B. Hence, the three test combinations of bolt/nut thread fit were 1B-2A, 2B-2A, and 3B-3A. The bolts and nuts were ultrasonically cleaned and dried, then lubricated with SAE 5W-30 engine oil. Each bolt is tightened initially to 27 kn, and the upper plate is then subjected to a cyclic displacement of 1.25 mm. Experimental results show that the hole clearance has a significant effect on the loosening rate per cycle assuming that the cyclic displacement amplitude is large enough to cause the bolt head to slide and fully consume the hole clearance. A larger hole clearance causes loosening at higher rate. Gambrell [57] has carried out series of tests on both coarse and fine thread bolts. For the tests he has used those variables of a threaded fastener, such as initial preload, lubrication, frequency of loading and number of cycles of load applications and dynamic-static ratio (DSR) which is the quotient of the repeated load divided by the initial preload. From the study it is observed that the severe loosening occurs during the first 3,000 cycles. As the number of cycles increases from 3,000 to 5,000, the rate of loosening steadily decreases. Within the range of frequencies tested, frequency has no effect on loosening. Sanclemente and Hess [61] have experimentally investigated mechanical loosening in bolted joints due to cyclic transverse loads. In this work a balanced number of factors are studied. They correspond to basic factors necessarily defined during the fastener selection process and include preload, fastener material elastic modulus, nominal diameter, thread pitch, hole fit and lubrication. Sixty-four tests have been performed as part of a 2 6 nested-factorial design (refer table 2.4) in which the nominal diameter is the nesting factor of preload, thread pitch and hole fit. The loss of preload is indicative of fastener loosening and is measured throughout the experiment. The initial preload and the residual preload (remaining preload after the test) define the response variable; ( ) ( ) Where Y is the dimensionless response variable, which is simply a measure or indication of preload loss. It can take a value between 0 and 1. Zero indicates no loosening and one indicates complete loosening. The best performance (Y = 0.01) was obtained with the treatment combination defined by the following conditions: aluminum fastener, 12.7 mm nominal diameter, UNF threads, high preload, tight fit hole 24

18 and lubrication. Both the experimental and the mathematical model results show that tighter thread fit between the bolt and the nut threads reduces the loosening rate. Table 2.4: Parametric Study Test Matrix [61] Factor Description Level Low High A Fastener material elastic modulus Aluminum Stainless steel B Nominal diameter 6.4 mm (1/4 ) 12.7 mm (1/2 ) C (B) Thread pitch UNC UNF D (B) Preload 32 % Yield 64 % Yield E (B) Hole fit Loose Tight F Lubrication Dry SAE 30 oil Alkelani and Nassar [62, 63, 64] have experimentally determined the clamp load loss due to elastic interaction and gasket creep relaxation in bolted joints. Studied parameters include the gasket material (styrene butadiene rubber and flexible graphite) and thickness (1.5, 3, 4.5 and 6 mm for styrene butadiene and 1.5 and 3 mm for flexible graphite), bolt spacing (varied by using a different number of bolts; three, five, and seven), tightening sequence (sequential, star and simultaneous tightening), fastener grip length, and level of the fastener preload. The joint is composed of two steel flanges and a gasket made of styrene butadiene rubber or flexible graphite. The flanges are fastened together using M Class 10.9 fasteners. Force washers are used to monitor bolt tensions in real time. Bolt tightening at room temperature is accomplished by using either an electric digital torque wrench with various control options and by using a production-size multiple spindle fastening system that is capable of simultaneous tightening of all fasteners. From this study it was seen that the elastic interaction between the various fasteners in gasketed flanged joints is significantly influenced by the gasket material and thickness, bolt spacing, and by the tightening strategy, whereas the amount of elastic interaction in hard non-gasketed joints is insignificant. The largest percentage of preload loss is in bolts that are tightened first, and the smallest percentage preload loss is in bolts that are tightened last. For the same torque value, using simultaneous tightening produces a higher and more uniform clamp load in the joint, and it significantly reduces the amount of preload drop off due to the combined effect of elastic interaction and gasket creep relaxation. The fastener grip length has an insignificant effect on clamp load loss. This study provides an insight into enhancing the reliability of gasketed bolted joint by recommending the use of simultaneous tightening of all bolts when possible, in order to achieve a uniform and more stable clamp load and reduce the effect of elastic interaction. 25

19 Izumi et al. [65] have evaluated the relation between preload and tightening torque under the elastic tightening process and the load distribution of thread by threedimensional FEM using ANSYS 7.0. As a standard model, they employed a threaded fastener involving an M16 screw (diameter and pitch of thread, 16 and 2 mm, respectively, diameter and height of bolt head, 24 and 10 mm, respectively), with 10- pitch internal thread and 5-pitch external thread and a cylindrical clamped component whose inner radius, outer radius, and thickness are 18, 50, and 35 mm, respectively. The FEM model is shown in figure 2.3. Figure 2.3: Finite Element Model for Tightening Process [65] The grade and position of crossover is 6H/6g. The detailed shape of the bolt head and curvature of the bottom of thread are not taken into account. With regard to the nut, the hexagonal corner used to apply the tightening torque is not modeled. The contact element pair TARGE170 and CONTA174, which realizes surface surface contact between three-dimensional objects and can handle the Coulomb friction, is used. The total number of nodes and elements are 23,946 and 12,821, respectively. The tightening analysis is performed by applying the constrained circumferential displacement on the side surface of the nut, followed by removing it. The side surface of the bolt head is fixed and the contact of the bolt-head bearing surface is completely stuck in order to avoid rigid rotation. The Young modulus and Poisson ratio of all materials (bolt, nut, and clamped component) are 205 GPa and 0.3. The friction coefficient of the contact surface is set at All analysis was restricted to low preloads (elastic range) to avoid the influence of yielding. The external force normal to the axial direction is modeled by applying the constrained displacement with 0.3 mm amplitude to the edge surface of the movable plate. Based on a comparison with classical material-mechanics based theories and experimental results, it was observed that the loosening due to shear loading is 26

20 initiated when complete thread slip is achieved prior to bolt-head slip, which conventional theory has considered the initiation point Issues of Non-conformance The mechanical performance of threaded components is a function of component material properties, thread geometry, and the environment in which the components are subjected. Other characteristics such as coatings, assembly technique, and thread lubricant will also influence mechanical performance of a threaded component in an assembly. Thread geometry and dimensional conformance of threaded components significantly influences the mechanical performance of threaded fasteners. From the experiments conducted by Dong and Hess [66, 67] resulted in failure time data from a vibration induced loosening apparatus for fastener combinations within dimensional conformance as well as for fastener combinations with nonconforming pitch, minor and major diameters. Data from the tests show a significantly degraded resistance to vibration for the fastener combinations with undersized pitch and major bolt diameters or oversized pitch and minor nut diameters, compared to fastener combinations within conformance. Leon et al. [68] have investigated the effect of thread dimensional conformance of fasteners on yield and tensile strength. A total 13 fastener combinations as mentioned in table 2.5 for 5 samples each i.e. 65 test specimens were tested. Bolt pitch diameter conformance (mm) Table 2.5: Fastener Combinations [68] Bolt major diameter conformance (mm) Nut pitch diameter conformance (mm) Nut major diameter conformance (mm) within within within within Undersized Undersized Oversized Oversized Undersized Undersized Oversized Oversized Two modes of failure were observed from the tests: (1) failure through the bolt cross section, and (2) bolt thread strip. The first mode of failure was observed for 57 of the 65 tests performed and the remaining eight specimens failed in bolt thread strip. In an effort to quantify the relation between the yield strength (Sy) and the four dimensional variables (i.e., bolt pitch diameter, Dp, bolt major diameter, D M, nut pitch diameter, dp, and nut minor diameter, dm), the following multiple linear regression model was computed using the method of least squares: Sy = 18,133,940 Dp 2 + 2,792,535 D 2 M - 3,516,167d 2 P - 1,510,968 dm 2 (2.8) 27

21 2.6 THREAD LOCKING TOOLS AND TECHNIQUES: Once the fasteners are properly tightened, the design problem is to prevent loss of bolt preload. High preload produces high friction forces at the mating threads and reduces joint motion (for example, from vibration) that tends to loosen the joint. However, the inevitable movement of parts, relaxation, changes in temperature, or vibration will eventually work any screw thread loose. Thus, some additional methods for bolt preload retention are required. These methods are the: friction and positive locking methods. The more widely used friction types include deflected beam all-metal locknuts, nonmetallic insert-type locknuts, and self-locking screws with an integral locking feature. Lock-washers are not recommended for high-strength applications because of the possibility of scoring the bearing surfaces, and total loss of lock in the event of any loosening. Also, lock-washers introduce a bending component to the fastener. Positive locking methods include lock wire, cotter pins, and various methods of keying or cementing the screw thread in place. Cementing is used only when there is no need for future disassembly. Tack welding is not recommended because of possible metallurgical damage [69]. Motion, vibration, or shock can cause a momentary loss of pressure between joint members that allows them to slip sideways. A wide variety of locking fasteners resist this loosening. Some fasteners use an accessory part to provide a mechanical lock. Others resist loosening through the use of prevailing torque. Prevailing torque develops while a fastener is being tightened or loosened. It is usually measured while the fastener is in motion under zero axial load. So a free-spinning nut has no prevailing torque. Prevailing torque is provided either by physical features that create interference or by chemical adhesives [70]. Prevailing torque features in fasteners include nonmetallic types such as nylon rings or nylon strips added to the threads as well as metallic types which utilize, for example, out of round or thread distortion. Prevailing torque locking fasteners are widely used because the prevailing torque is quantifiable during assembly which makes this locking feature verifiable. The primary problem with such locking fasteners observed in practice is degradation of prevailing torque with repeated reuse due to plastic deformation or wear. When properly used, thread adhesives generally provide excellent locking, but are not verifiable during assembly since such product requires curing. In addition, proper thread preparation, e.g., cleaning and priming, is critical for performance [71]. 28