Modeling and Simulation for Material Selection and Mechanical Design Part 17 potx

with rolled thread after heat treatment depends significantly on the preload level (reason: strain hardening and residual stresses from thread rolling are not compensated by a new grain structure from heat treatment, so nonlinear profiles from loading stresses and residual stresses are superposed). The test principle for determining axial fatigue load s aspermt2 is defined in ISO 3800 [30] or more detailed in DIN 969 [10] for threaded fastening ele- ments. Normally, the screw shank is the location of fatigue failure, but the clamped part or nut thread component can also end in fatigue failure, e.g., thin sheet metals as clamped part and a screw head with locking teeth. If the location of fatigue failure is at screw head fillet, significant bending of screw is probable (see also Fig. 54). Minimizing fatigue problems can be realized by reducing the screw stressing (e.g., larger screw size, lower additional force for screw in fastening system), proper screw section design (see Fig. 2, e.g., sufficient radius at head-to-shank-fillet, perpendicularity of screw axis and head support, smooth transition of each discontinuity at screw shank, such as different diameters of screw, running out of thread to unthreaded shank), no overlap- ping of stress concentrations (e.g., chamfer at clamped part under head or at Figure 53 Typical cross-section of screw M18 Â 1.5-12.9 failed in fatigue, result of testing procedure according ISO 3800, stress amplitude s a ¼80 MPa, mean stress 555 MPa, symmetric axial force without bending moment. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. first nut thread flank, avoiding corrosive pittings at thread flanks or at head-shank-transition). The most established actions to increase the fatigue limit of the screw itself are discussed in Fig. 41. Figure 53 contains a cross-section of a screw M18 Â 1.5 which has failed in fatigue. A plane fracture zone can be seen at outer regions of the cross-section (area of crack initiation and crack propagation) and an unplane fracture area in the center of the cross-section (residual area of rapid failure under preload). Figure 54 in contrast to Fig. 53, gives an impression of a fatigue failure with significant bending moment under external loading. This result was obtained with a transversal vibrational test (see also Fig. 76). Now, the area of crack initiation and crack propagation with ‘‘cycle lines’’ is clearly differ- ent from the residual fracture area. The size of this second part of cross-sec- tion gives information whether the acting preload at the event of failure was high or not (fatigue failures often are induced by wrong tightening or loss of preload caused by relaxation or self-loosening). F. Aspects of Quality Management The overall objective for quality aspects of a threaded fastening system is to guarantee sufficient preload. This preload normally is not specified directly. Due to this reason, a large number of details must fit together, which have to be realized by different responsibilities. Figure 55 demon- strates seven main groups of authorities which have to give their contri- bution to quality of the fastening system. The drawn boxes make clear that every authority has its objective, its risk and takes actions due to different criteria. Besides this, it is important for clarifying failures that each authority in most cases belongs to different business units or com- panies, so various communication interfaces exist, which have to work without deficit or error. So, from the point of organization, a ‘‘fastening manager’’ is recommended. A few more quality aspects are:  The calculation of threaded fastening systems in any case is an approximation because numerous details have to be estimated like real screw loading in the system, local fatigue strength of screw, material inhomogeneities, real external loading spectrum or others. By these uncertainties, the compatibility of a designed fastening system with experience from former solutions is valuable, and critical calculations have to be verified by experimental testing. But proper designing due to guidelines of this chapter avoids a lot of failure situations and reduces testing expense. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. every screw failure during operation, the assembly process has to be ana- lyzed too. Always, three general reasons for failures during operation have to be considered: 1. wrong initial preload (tightening process, poor design), 2. wrong residual preload (relaxation by creeping of materials, gaskets), and 3. overloading mechanically, thermal or reactive (too high operating load with plastification or sliding, too high temperature with creeping or decreasing of strength, too strong environment with significant corrosion). G. Cost Accounting of Fastening Systems Always cost accounting of a fastening system has to be done due to life cycle of the component system (product) because only this life cycle cost can be compared to the customer value. Then, all boundary conditions of Figs. 1 and 32 have to be included and evaluated monetarily—the fastening element is only one contribution to this life cycle cost. Figure 56 proposes a funda- mental approach to total cost accounting, which takes into account the main types of cost related to a fastening system. Optionsforcostoptimizingarethreadrolling(Fig.7),standardmate- rials(Fig.10),coarsetolerancesforgeometry(Table2).Butonemust always remember that the guaranteeing of reliable function has to be of higher priority than the cost for a technical system. Otherwise, the product has no customer value and therefore no market. IV. EXAMPLES OF DESIGN A. Fastening with Optimized Initial Preload Traditionally, screws are tightened with torque control (see Fig. 19). The tightening torque T tot is specified for the conditions with lowest friction. This is shown on the left side of Fig. 57 (as a supplement to Fig. 51) for a steel screw 8.8, tensile strength of 800 MPa. The first case A considers a low friction situation with coefficients of m t ¼m h ¼0.08. This screw at 20 N m tightening torque is stressed up to 0.9 Â R p0.2 (rhomb marking) and produces a preload of almost 20 kN. Because of high screw stressing, the torque specification cannot be increased over 20 N m if yielding or breaking of screw has to be avoided in any case. For the same screw in large series assembly lines, the frictional situa- tion can change to m t ¼m h ¼0.16 (curve B). Then, for the same tightening Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. defined minimum plastification before fracture, therefore, the manufacturing process must be optimized. In principle, using yield point controlled tighten- ing, the screw cannot be overloaded. Using angular controlled tightening, the screw cannot be overloaded, if the snug torque is low enough (e.g. 10 N m þ908 in Fig. 57). Figure 26 explains why also a screw tightened with angular control can be loaded additionally. As a side-effect, the diagrams in Fig. 57 make it clear that torque-con- trolled tightening (and a steel screw for aluminum components) is an ‘‘old- fashioned’’ and not very optimized solution from the viewpoint of engineer- ing threaded fastening systems. The diagrams confirm that a torque value mentions ‘‘nothing’’ about the preload acting in the joint. B. Fastening with Small Thread Engagement For every component design, the necessary minimum thread engagement needs space and therefore generates component weight. A minimization of thread engagement is required. Figure 33 contains the basic mechanics and suggests the use of relatively high-strength nut thread material or the use of relatively low-strength screw material. Figure 58 Measured low minimum thread engagement for Aluform TM screws M6 (From Ref. 15.) Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. Figure 58 gives a practical verification of the calculation from Fig. 33 for a pull-out-test with RIBE-Aluform # screws [60] M6 (R m > 400 MPa) engaged to an aluminum nut thread plate (R m ¼300 MPa) with certain lengths of thread engagement t e . The bar diagram shows the maximum pull-out-forces F zmax in the event of failure. The upper level of %9kN belongs to a tensile screw breaking in the screw shank. The lower level belongs to nut thread stripping. The transition begins exactly at the point 0.7 Â d which is also pre- dicted by Fig. 33 (do not forget to consider chamfer of 1 Â P in Fig. 33). So, indeed Aluform # screws have a reliable behavior against stripping also for low-strength nut thread components and low thread engagements t e , which would never be fulfilled with a steel screw 8.8 or higher. C. Fastening of High-strength Components High-strength components provide the chance for small threaded fastening systems because contact pressure at screw head as well as thread flanks can be in the range between minimum of R p0.2 and R m of the materials in con- tact. Besides this, hard surfaces are almost unaffected by roughness chan- Figure 59 Screw head design with small head diameter for tightening on hard surface. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. ging with adhesive or abrasive wear meachanisms, so that the frictional situation is constant over a wide range of preload. Figure 59 demonstrates a small-diameter screw head design for tigh- tening on hardened component surface in the range of R m ¼1400 MPa (for high-strength screw materials see Table 9). Such a screw of dimension M11 Â 1.5 and a screw tensile strength of 1150 MPa produces an initial pre- load of about 60 kN. This leads to a mean contact pressure of 950 MPa (compare also diagram in Fig. 39 and explanations related). In addition, for such design, the lubrication of the screw is of significant importance. These screws offer the possibility for small space flanges and in consequence for a compressed design of component. In contrast to light metal com‘ ponents, these high-strength materials possess significant mass density, but can be used for an extreme compact design. The same aspects are valid for thread engagement but at least t e ¼0.8 Â d should be realized to avoid stripping of screw thread flanks (compare asymptotic behavior of Fig. 33 for high nut thread strength R mnut ). If fastening high-strength components, precise support geometries and small contact roughness is required, then peak contact pressure is avoided, which can be the origin of crack propagation and fatigue failure of the component. Threaded fastenings with components made of high-strength materials provide the possibility for meeting small space requirements (low thread engagement, small head diameter, small screwing boss diameter at clamped part). If this is combined with overelastic tightening and reliable lubrication, then also lightweight fastening is possible. D. Fastening of Components Made of Brittle Materials ‘‘Brittle’’ means that a material has a low ductility before fracture, which leads to a sudden rupture without plastified deformation in the case of ten- sile testing or overloading a component. As a guide, the fracture toughness from tensile test is for brittle materials smaller than 3–5%). For such a component (e.g., made of magnesium, titanium with hexagon crystal struc- ture or cast iron with high carbon content or ceramic materials), not only is the mean stressing important, but also all local stress peaks have to be minimized. Therefore, thread engagement of a screw in brittle materials should be increased by at least þ20% due to Fig. 33 because of the inhomogeneous stress distribution up to the event of fracture, which is not compensated by local plastification of the nut thread flanks. Brittle materials of low strength (cast iron, magnesium) tend to pro- duce increased adhesive-abrasive wear in the screw head contact zone Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. during tightening (Fig. 60). This results in increased roughness, particles and undefined contact conditions, so the head frictional torque is increased and—if the screw is tightened by torque control—the preload is reduced significantly. To avoid this, the lubrication of screws for brittle materials should be enhanced. Use of thread rolling screws in brittle materials is critical. Figure 61 presents an example from thread rolling with a high-performance thread rolling screw M8 (induction hardened forming point) in high-strength duc- tile gray iron GGG 50. The result is that particles of nut thread material are produced in an unacceptable amount. They lead to poor nut thread quality as well as screw thread damage and therefore to insufficient process capabil- ity for series production (see torquing diagram in Fig. 61 with temporary breakdown of torque curve). E. Fastening of Light Metal Components Light metals, such as aluminum and magnesium, are characterized mechani- cally by low strength and high thermal expansion coefficient. Especially in Figure 60 Adhesive–abrasive wear of head support area after angular controlled tightening of screw M14 Â2-11.9, specification 150 N m þ908 þ908 þ908, preload app. 90 kN, surface gray cast iron GG25. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. tightening or angular controlled tightening (Fig. 26) and a screw with threaded screw shank (Fig. 3). 4. If operating at elevated temperatures, a special adaptation of thermal fit for minimization of thermal stress increase is necessary (often screws made of aluminum are an effective alternative compared to steel screws, Fig. 29). 5. For aluminum components, thread rolling screws are widely used (Fig. 7, Fig. 62). 6. For corrosion stability, use enhanced corrosion protection for steel screws or use aluminum screws (Fig. 63). Figure 64 contains the fastening of a magnesium component with thread rolling screw made of aluminum in two columns: the left side refers to five repetitions of screwing with same screw into the same nut thread hole. The right side refers to the situation where the same screw is screwed into a new pilot hole without nut thread for each repetition. The images of the screwing bosses confirm a high quality of the pro- duced nut thread in magnesium for both columns of Fig. 64. The diagram is very detailed due to the formation of positive torque and prevailing torque (negative). All values are proposed for 1 to 5 screwing operation. Figure 62 Trilobular stud M6 Â 32 for thread rolling in aluminum component, dry lubrication, maximum thread forming torque 3 N m, tightening torque for stud 10 þ0.5 N m, thread engagement 11 mm, casted pilot hole with diameter 5.4–5.6 mm. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. Figure 65 Fastening of aluminum metal foam with screw–sleeve-combination, see also Fig. 44. Figure 66 Thermal decrease of preload by relaxation of threaded fastening systems with magnesium components at elevated temperatures; measured data. (From Ref. 16.) Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. Every case, from 1 to 7, shows the initial preload for steel screw (light bar) and aluminum screw (dark bar) before and after thermal exposure. For case 1 (screw M10), this means 32 kN resp. 19 kN before thermal exposure and 5 kN resp. 9 kN after exposure. The other cases confirm similar behavior. The most extensive preload relaxation occurs for fastening systems with the widely used magnesium alloy AZ91 (high-strength alloy with 9 wt.% aluminum and 1 wt.% zinc; relatively stable corrosive behavior; low creeping resistance at temperatures above 1208C). Cases 3 and 5 confirm where the steel screw leads to almost no preload after thermal exposure— here aluminum screws are the only solution for reliable fastening systems. In general, for these situations, an aluminum screw always gives a higher residual preload than a steel screw. F. Fastening of Components with Thread Rolling Screws What is the difference between thread rolling screws and screws for existing nut thread in practice? The fundamental principle of thread rolling screws is Figure 67 Functional behavior of thread rolling screw compared to a screw for existing nut thread used for the same application, see also Ref. 63. Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved. [...]... rivet produces a high level clamping force In general, riveting systems always have material deformation either of clamped part and= or of rivet, so high-strength rivets have to be engineered for each particular application (size and number of rivets, geometry tolerances, and setting parameters) For more information regarding riveting systems, see Ref [59] For standardization of blind rivets, see Ref... tightening and operating) Linear thermal expansion coefficient of screw material Linear thermal expansion coefficient of clamped part material Force Values Force in general Preload acting in screw shank Change of preload (e.g by thermal expansion or relaxation) Stable preload after tightening and sort time relaxation Separating preload for given eccentrical loading with axial force Fax Maximum axial force for. .. d0 Nominal stress area for tensile loading of screw thread; As ¼ [0.5(d2 þ d3)]2p=4 for metric thread system Contact area at head support between bearing surfaces Nominal circle area from thread size Substituted area relevant for resilience of clamped part Moment of Inertia and Polar Section Modulus Moment of inertia for bending of clamped part and screw shank together, used for calculating of load... preload Fp0, before component separating occurs Additional axial force of screw under external axial force Fax Tangential force related to the screw axis Axial force related to the screw axis Necessary preload for safe working of bolted joint Permitted axial stripping force of nut thread at temperature t1 Permitted axial stripping force of nut thread at temperature t2 Permitted axial stripping force of screw... behavior of tightening torque Ttot and preload Fp compared between metric screw for existing nut thread and thread rolling screw of same nominal diameter M8 Part (a) refers to a metric screw in machined nut thread—there is no forming torque for the first eight revolutions until the screw head is in contact with the clamped part surface Then, the preload is generated for further screw turning To obtain... Components More and more sheet metal designs are used for automated production of components with large lot sizes These components should also be fastened automated This can be done in two ways: (1) use of thread rolling screws for sheet metals, and (2) use of staking elements for generating a high-duty thread at thin walled components (for principle see Fig 43) Thread rolling screws for sheet metals... which describes the distance between bending axis of clamped part and axis of external axial force Fax Length transversal to screw axis which describes the distance between bending axis of clamped part and axis of through hole in clamped part Largest length of a screw drive in a plane transversal to screw axis (e.g., width across corners for hexagon drive) Continued Copyright 2004 by Marcel Dekker,... transportation-, and aerospace-industry) A blind rivet requires a minimum of surface preparation before setting (low roughness requirements) The residual stem is important for high shear strength and high fatigue strength of the joint Figure 71 shows an example for a bulbing blind rivet (in contrast to blind rivets with expanding head) which is applied to aluminum sheet metals (sheet thicknesses 2 and 3 mm)... torque of 28 N m is necessary for case (a) A similar situation states case (b), but there exists a forming torque during the first 12 revolutions of %10 N m This forming torque is reduced to %3 N m, if the forming point of the screw is turned outside of the nut thread Therefore, to obtain a preload of 15 kN, the tightening torque has to be increased to 31 N m For case (c), the forming torque in the situation... aspects of total cost saving for designing threaded fastening systems are given in Table 15 In no case, can all aspects be realized for the same application Therefore, the design engineer has the responsibility to set priorities More information due to cost-optimization of fastening systems can be found in Ref [66] Figure 75 Thread rolling with same screw geometry in different materials Copyright 2004 . engi- neered for each particular application (size and number of rivets, geometry tolerances, and setting parameters). For more information regarding rivet- ing systems, see Ref. [59]. For standardization. initial preload for steel screw (light bar) and aluminum screw (dark bar) before and after thermal exposure. For case 1 (screw M10), this means 32 kN resp. 19 kN before thermal exposure and 5 kN resp area relevant for resilience of clamped part Moment of Inertia and Polar Section Modulus I full mm 4 Moment of inertia for bending of clamped part and screw shank together, used for calculating