Numerical and Experimental Characterization of a Railroad Switch Machine

machines

Article Numerical and Experimental Characterization of a Railroad Switch Machine

Dario Croccolo, Massimiliano De Agostinis, Stefano Fini * ID , Giorgio Olmi and Francesco Robusto ID Department of Industrial Engineering, University of Bologna, Viale del Risorgimento 2, 40136 Bologna, Italy; [email protected] (D.C.); [email protected] (M.D.A.); [email protected] (G.O.); [email protected] (F.R.) * Correspondence: stefano.ﬁ[email protected]; Tel.: +39-051-2093455

Received: 15 January 2018; Accepted: 12 February 2018; Published: 17 February 2018

Abstract: This contribution deals with the numerical and experimental characterization of the structural behavior of a railroad switch machine. Railroad switch machines must meet a number of safety-related conditions such as, for instance, exhibiting the appropriate resistance against any undesired movements of the points due to the extreme forces exerted by a passing train. This occurrence can produce very high stress on the components, which has to be predicted by designers. In order to assist them in the development of new machines and in defining what the critical components are, FEA models have been built and stresses have been calculated on the internal components of the switch machine. The results have been validated by means of an ad-hoc designed experimental apparatus, now installed at the facilities of the Department of Industrial Engineering of the University of Bologna. This apparatus is particularly novel and original, as no Standards are available that provide recommendations for its design, and no previous studies have dealt with the development of similar rigs. Moreover, it has wide potential applications for lab tests aimed at assessing the safety of railroad switch machines and the fulfilment of the specifications by many railway companies.

Keywords: railroad switch; railway junction; FEA; experimental; points

1. Introduction A railroad switch machine (RSM), turnout or set of points is a mechanical installation enabling railway trains to be guided from one track to another, such as at a railway junction or where a spur or siding branches off. One of the key safety requirements of railroad switches is related to achieving a suitable resistance against any undesired movements of the points, due, for instance, to the extreme forces exerted by a passing train in the case of the needle leaned to the rail (force F in Figure1). Many railway companies assume a force F = 100 kN as standard. This work deals with the development of FEA models aimed at accomplishing the structural design of the RSM under the aforementioned operating load. In order to validate such models, an experimental test bench has been designed and manufactured. This comprises two ad-hoc designed fixtures that allow the accommodation of the test piece on a standard INSTRON 8500 500 kN standing press and the application of forces up to a maximum of F = 300 kN. Issues of novelty arise from the lack of studies both in the scientific and in the technical literature dealing with the development of similar fixture devices. The developed testing rig can be used not only for FEA validation purposes, but also for experimental tests aimed at warranting the safety of the RSM and the accomplishment of design requirements by most railway companies. The originality of the performed non-trivial design task arises also from the lack of specific Standards providing recommendations or reference schemes for the execution of lab tests aimed at assessing the structural response of RSM under high loads.

Machines 2018, 6, 6; doi:10.3390/machines6010006 www.mdpi.com/journal/machines Machines 2018, 6, 6 2 of 9 Machines 2018, 6, x FOR PEER REVIEW 2 of 9 Machines 2018, 6, x FOR PEER REVIEW 2 of 9

FigureFigure 1. 1.Geometry Geometry of of a a railroad railroad switch switch.. Figure 1. Geometry of a railroad switch. 2. Materials and Methods 2. Materials and Methods 2. Materials and Methods The Alstom RSM object of the present investigation is shown in Figure 2, along with some balloonsThe AlstomThe highlighting Alstom RSM RSM object the object k ofey the structural of present the present components investigation investigation of the is shown machine. is shown in Figure in Figure2, along 2, along with with some some balloons highlightingballoons highlighting the key structural the key componentsstructural components of the machine. of the machine.

Figure 2. 3d model of the Alstom RSM: (1) body; (2) lower plate; (3) pin; (4) hammer; (5) switching FigurerodFigure 2.; (3d6) 2cam. model 3d; model(7) ofdetection the of the Alstom rodAlstom; (8 RSM:) armRSM:. ( 1 )(1 body;) body (2; )(2 lower) lower plate; plate; ( 3(3)) pin; pin; ((44)) hammer;hammer; ( 5)) switching switching rod; (6) cam;rod;( (76)) detectioncam; (7) detection rod; (8) rod arm.; (8) arm. Due to confidentiality-related issues, the working principles of the machine cannot be described in detail.Due Theto confidentiality analysis was limited-related to issues, the verification the working of theprinciples mechanism of the against machine unwanted cannot be movements described ofinDue detail.the to points confidentiality-related The caused analysis by was a passing limited issues,train to the since verification the the working system of principlesthe is equipped mechanism of with the against machinetwo inte unwantedrlocking cannot movements devices. be described In in detail.fact,of Thethe once points analysis the caused full was stroke by limited a haspassing been to thetrain travelled, verification since the and system the of thepoints is mechanismequipped are in the with againstopen two (or inte unwanted closed)rlocking position, devices. movements the In of the pointsswitchingfact, once caused therod full by(5) aisstroke passingsecured has to train been the sincebodytravelled, (1) the by system and means the is ofpoints equipped a hammer are in with (4);the atopen two the interlockingsame(or closed) time, theposition, devices. detection the In fact, oncerodswitching the (7) full is stroke secured rod (5) has isto secured beenthe lower travelled, to platethe body (2) and by (1) themeans by pointsmeans of a areslider,of a inhammer thenot openrepresented (4); (orat the closed) samein the position, time,picture. the Thereforedetection the switching rod (5)therod is locking(7) secured is secured devices to theto comethe body lo intower (1) effectplate by means (2)preventing by means of a hammerany of amovement slider, (4); not at ofrepresented the the same rods, time,when in the thean picture. external detection Therefore force rod is (7) is securedappliedthe tolocking the along lower devices z-axis plate come to (2)the into bypoints, meanseffect and preventing of thereby a slider, to any notthe movement arms represented (8). of the in rods, the picture. when an Therefore external force the lockingis appliedAccording along z -toaxis the to requirements the points, and set thereby by railway to the companies, arms (8). the RSM should be validated under devices come into effect preventing any movement of the rods, when an external force is applied along the actionAccording of a force to the F =requirements 100 kN. The loadset by application railway companies, rate surely theaffect RSMs the should response be validatedof the structure. under z-axis to the points, and thereby to the arms (8). Thethe action testing of force a force of FF == 100100 kNkN. is The set loadby the application railway company rate surely in order affect tso the account response for dynamicof the structure. effects. InTheAccording order testing to forceattain to the of an requirementsF adequate= 100 kN isstiffness set set by bytheof railwaytherailway test fixture,company companies, it hasin order been the t RSModimensioned account should for dynamicfor be validateda maximum effects. under the actionloadIn order of of 3 00to a forcekNattain. The F an = overall 100adequate kN. dimensions Thestiffness load ofof application thethe test test piece fixture, rate are itsurely900 has × 300been affects × dimensioned210 themm response; therefore for a, of maximumthe the fixture structure. The testingwasload conceivedof 3 force00 kN of. inThe F two = overall 100 separate kN dimensions is parts, set by a the lowerof the railway andtest piecean company upper are 900grip, in× 300so order as × 210to toachieve mm account; therefore a certain for dynamic, the flexibility fixture effects. In orderduringwas toconceived attainmounting anin two adequateand separateunmounting stiffness parts, operations a oflower the and test on an the fixture, upper standing grip, it has press. so been as toIn dimensionedachieve order not a certain to transmit for flexibility a maximum any loadunwantedduring of 300 kN.mounting bending The overall and moment unmounting dimensions at the arms, operations of the test on fixture piece the standing arewas 900 shap ×press.ed300 as ×shownIn 210order mm;in not Figure therefore,to transmit 3. While the anythe fixture was conceivedlowerunwanted grip bendingis in atwo simple separatemoment C-shaped at parts, the interface arms, a lower the between test and fixturean the upper actuator was shap grip, threaded so asas andshown to the achieve in arms, Figure athe certain 3. upper While flexibilitygrip the duringhaslower mounting to grip retain is a the andsimple whole unmounting C- shaped RSM by interface operationsmeans between of four on M20the the actuator standing 8.8 class thread bolts. press. and The In the orderbolted arms, not jointthe to upper is transmit doubly grip any overlapped:has to retain this the provision whole RSM allows by doubling means of the four frictional M20 8.8 surfaces class bolts.and hence The boltedthe transmissible joint is doubly load unwanted bending moment at the arms, the test fixture was shaped as shown in Figure3. While the foroverlapped: a given bolt this size provision and class al lows[1,2]. doublingExcept for the a few frictional details, surfaces the fixture and has hence to be the arc transmissible welded, therefore load lower grip is a simple C-shaped interface between the actuator thread and the arms, the upper grip has afor structural a given bolt steel size S275JR and class according [1,2]. Except to [3] forhas a beenfew details, chosen the for fixture its construct has to ion.be arc All welded, the welds therefore were to retainstaticallya structural the whole dimensioned steel RSM S275JR by according according means ofto toStandard four [3] M20has ENbeen 8.8 1993 class chosen-1- bolts.8 [4] for. In its The order construct bolted to assession. joint theAll is doublystressesthe welds and overlapped: were the this provisiondeformationsstatically dimensioned allows on doublingthe fixture according the under frictional to Standard maximum surfaces EN design 1993 and-1 - load8 hence [4] . (FIn the=order 300 transmissible to kN assess), some the load FEAstresses for have and a given been the bolt size andperformeddeformations class [ 1by,2 ].means on Except the of fixture the for commercial a few under details, maximum code the A fixturensys design Workbench. has load to be (F arc = 3 welded,00 kN), sometherefore FEA a have structural been steel S275JRperformed according by tomeans [3] has of the been commercial chosen for code its A construction.nsys Workbench. All the welds were statically dimensioned according to Standard EN 1993-1-8 [4]. In order to assess the stresses and the deformations on the fixture under maximum design load (F = 300 kN), some FEA have been performed by means of the commercial code Ansys Workbench. Machines 2018, 6, 6 3 of 9 Machines 2018, 6, x FOR PEER REVIEW 3 of 9

Machines 2018, 6, x FOR PEER REVIEW 3 of 9

Figure 3. Loading scheme: (1) test piece (switch machine); (2) lower grip; (3) upper grip. Figure 3. Loading scheme: (1) test piece (switch machine); (2) lower grip; (3) upper grip. FigureFigure 4a shows 3. Loading the boundary scheme: (conditions1) test piece applied (switch machine)to the model:; (2) lower the uppergrip; (3 grip) upper has grip been. fixed at theFigure upper4a end shows and the loaded boundary by two conditionsequal forces applied Fz = 150 to kN, the one model: at each the arm. upper The gripmodel has has been been ﬁxed at themeshed upperFigure with end 4a SOLID187 and shows loaded the Tetrahedral boundary by two equalconditionsand Hexahedral forces applied Fz elements, = 150to the kN, model: Figure one the at4beach .upper The arm.material grip Thehas is been modela structural fixed has at been meshedsteel,the withupper whereas SOLID187 end the and bonded loaded Tetrahedral contacts by two are andequal managed Hexahedral forces by Fz means = elements,150 ofkN, the one pure Figure at penaltyeach4b. arm. The contact The material algorithm,model is has a structural withbeen steel,themeshed whereas normal with stif the SOLID187fness bonded factor Tetrahedral contacts set to FKN are an = d0.01, managed Hexahedral following by elements, meansthe lines of Figuresuggested the pure4b. Theby penalty [5,6] material. The contact isanalysis a structural algorithm, and steel, whereas the bonded contacts are managed by means of the pure penalty contact algorithm, with withmodel the normal parameters stiffness are summarized factor set to in FKN Table = 1 0.01,. following the lines suggested by [5,6]. The analysis the normal stiffness factor set to FKN = 0.01, following the lines suggested by [5,6]. The analysis and and model parameters are summarized in Table1. model parameters are summarized in Table 1.

(a) (b)

Figure 4. FEA on the fixture upper grip: (a) Boundary conditions; (b) mesh; (c) total deformation ; (c) (d) (d) equivalent von Mises stress. Figure 4. FEA on the fixture upper grip: (a) Boundary conditions; (b) mesh; (c) total deformation; Figure(d) 4. equivalentFEA on von the Mises ﬁxture stress upper . grip: (a) Boundary conditions; (b) mesh; (c) total deformation; (d) equivalent von Mises stress.

Machines 2018, 6, x FOR PEER REVIEW 4 of 9

Machines 2018, 6, 6 4 of 9 Table 1. Analysis and model parameters.

Number of Nodes Table 1. AnalysisElement and Types model parameters.Elastic Modulus Poisson Ratio

[-] [-] E [Mpa] ν [-] Number of Nodes Element Types Elastic Modulus Poisson Ratio

Fixture 80[-],000 SOLID187 (Tetrahedral [-] and Hexahedral) E200 [Mpa],000 ν0.3[-] Fixture 80,000 SOLID187 (Tetrahedral and Hexahedral) 200,000 0.3

As can be appreciated by looking at Figure 4c, the total deformation is Δtot_max = 1.3 mm, As can be appreciated by looking at Figure4c, the total deformation is ∆tot_max = 1.3 mm, whereas the maximum von Mises equivalent stress remains below 190 MPa (see Figure 4d); such a whereas the maximum von Mises equivalent stress remains below 190 MPa (see Figure4d); such a stress level is well below the material yield point SY = 275 MPa. Since the model is linear, a maximum stress level is well below the material yield point SY = 275 MPa. Since the model is linear, a maximum deformation of about Δtot_nom = 0.4 mm can be expected at nominal load, which is deemed deformation of about ∆tot_nom = 0.4 mm can be expected at nominal load, which is deemed acceptable. acceptable. The assembly procedure of the test rig requires quite a number of subsequent operations, The assembly procedure of the test rig requires quite a number of subsequent operations, briefly brieﬂy summarized in Figure5. In particular, Figure5d shows a detailed view of the arms of the summarized in Figure 5. In particular, Figure 5d shows a detailed view of the arms of the machine machine when these are clamped by the lower grip. When the assembly is done, the load cell undergoes when these are clamped by the lower grip. When the assembly is done, the load cell undergoes zero zero calibration and the test can begin. The main goal of the experiment is to provide a validation calibration and the test can begin. The main goal of the experiment is to provide a validation of the of the FEA models of the RSM that will be described in the following. A secondary aim of the FEA models of the RSM that will be described in the following. A secondary aim of the experimentation is to determine how much of the total load is borne by the switching rod and how experimentation is to determine how much of the total load is borne by the switching rod and how much by the detection rod. In order to accomplish this twofold task, three components of the RSM much by the detection rod. In order to accomplish this twofold task, three components of the RSM were instrumented by strain gauges: the two arms and the pin (see Figure6). were instrumented by strain gauges: the two arms and the pin (see Figure 6).

(a) (b) (c)

(d) (e)

Figure 5.5. Test arrangement:arrangement: (a(a)) Placement Placement of of the the lower lower grip, grip, (b ()b pre-mounting) pre-mounting of theof the upper upper grip grip with with the thetest piece;test piece (c) placement; (c) placement of the upperof the grip upper with grip a forklift; with (ad )forklift details; of(d the) details assembly of ;(thee) assembly ﬁnal conﬁguration.; (e) final configuration.

Machines 2018, 6, 6 5 of 9 Machines 2018, 6, x FOR PEER REVIEW 5 of 9 Machines 2018, 6, x FOR PEER REVIEW 5 of 9

Figure 6. (a) Placement of the strain gauges on the arm and (b) on the reworked pin. FigureFigure 6.6.( (aa)) PlacementPlacement ofof thethe strainstrain gaugesgauges onon thethe armarm andand ( b(b)) on on the the reworked reworked pin. pin.

The pin and arms had been previously reworked in order to accommodate the sensors (Vishay TheThe pin and and arms arms had had been been previously previously reworked reworked in order in order to accommodate to accommodate the sensors the sensors(Vishay Precision Group J2A-XXS047K-350); in particular, the pin required both milling and boring (VishayPrecision Precision Group Group J2A-XXS047K J2A-XXS047K-350);-350); in particular in particular,, the the pin pin required required both both milling milling and boring boring operations in order to achieve a plane surface for the application of the strain gauge, as well as a operationsoperations in in order order to to achieve achieve a plane a plane surface surface for the for application the application of the of strain the gauge,strain asgauge well, asas awell passage as a passage for the cables. The arms were instrumented by means of two strain gauges each; the strain forpassage the cables. for the The cables. arms The were arms instrumented were instrumented by means by of means two strain of two gauges strain each; gauges the each strain; the gauges strain gauges were connected in a half-bridge fashion to the Wheatstone circuit. The pin was instrumented weregauges connected were connected in a half-bridge in a half fashion-bridge to fashion the Wheatstone to the Wheatstone circuit. The circuit. pin was The instrumented pin was instrumented by means by means of a single strain gauge; a dummy gauge, which served as a temperature drift compensator, ofby a means single strainof a single gauge; strain a dummy gauge;gauge, a dummy which gauge, served which as a served temperature as a temperature drift compensator, drift compensator, was glued was glued to an identical, unloaded pin placed in the testing room. The adhesive used for the towas an identical,glued to anunloaded identical, pin unloaded placed in thepin testing placed room. in the The testing adhesive room. used The for adhesive the installation used for was the installation was the M-BOND 200 by Vishay Precision Group. All the sensors were installed by a theinstallation M-BOND was 200 the by VishayM-BOND Precision 200 by Group.Vishay AllPrecision the sensors Group. were All installedthe sensors by awere certiﬁed installed operator, by a certified operator, following the guidelines suggested by the Standards [7–9]. Data acquisition was followingcertified operator, the guidelines following suggested the guidelines by the Standards suggested [7– by9]. the Data Standards acquisition [7– was9]. D managedata acquisition by means was managed by means of the NI 9237 sampling card plugged into a NI cDAQ-9184 carrier. The FEA ofmanaged the NI 9237 by means sampling of the card NI plugged 9237 sampling into a NI card cDAQ-9184 plugged carrier. into a TheNI cDAQ FEA model-9184 carrier. of the RSM The wasFEA model of the RSM was developed by means of the Ansys code V.17. Due to the complexity of the developedmodel of the by meansRSM was of the developed Ansys code by V.17.means Due of tothe the Ansys complexity code V. of17. the Due assembly, to the submodelingcomplexity of was the assembly, submodeling was leveraged, by considering half a model at a time, as if the machine were leveraged,assembly, submodeling by considering was half leveraged, a model by at considering a time, as if half the a machine model at were a time, cut as along if the its machine mid-plane, were cut along its mid-plane, normal to the x-axis. In this way, it was possible to find a satisfactory balance normalcut along to theits midx-axis.-plane, In this normal way, to it the was x- possibleaxis. In this to ﬁnd way, a satisfactoryit was possible balance to find between a satisfactory accuracy balance and between accuracy and computational cost. Both models were meshed with tetrahedral elements computationalbetween accuracy cost. and Both computational models were meshed cost. Both with models tetrahedral were elementsmeshed withSOLID187, tetrahedral switching elements rod SOLID187, switching rod side (Figure 7), and detection rod side (Figure 8). The analysis and model sideSOLID187, (Figure 7switching), and detection rod side rod (Figure side (Figure 7), and8). detection The analysis rod and side model (Figure parameters 8). The analysis are summarized and model parameters are summarized in Table 2. inparameters Table2. are summarized in Table 2.

Figure 7. Boundary conditions for the half model comprising the switching rod. FigureFigure 7.7.Boundary Boundary conditionsconditions forfor thethe halfhalfmodel model comprisingcomprising thethe switching switching rod. rod.

Machines 2018, 6, 6 6 of 9 Machines 2018, 6, x FOR PEER REVIEW 6 of 9 Machines 2018, 6, x FOR PEER REVIEW 6 of 9

Figure 8. Boundary conditions for the half model comprising the detection rod. FigureFigure 8.8. BoundaryBoundary conditions conditions for for the the half half model model comprising comprising the the detection detection rod rod..

Table 2. Analysis and model parameters. TableTable 2. 2. AAnalysisnalysis and and model model parameters parameters.. Steel Components Cast Iron Components Steel Components Cast Iron Components Number of Steel Components Cast Iron Components NumberNumber of of Element Types Elastic Poison Elastic Poisson Nodes ElementElement Types Types ElasticElastic PoisonPoison ElasticElastic PoisPoissonson NodesNodes Modulus Ratio Modulus Ratio ModulusModulus RatioRatio ModulusModulus RatioRatio [-] [-] E [Mpa] ν [-] E [Mpa] ν [-] [-] [-][-] [-]E [Mpa] E [Mpa] ν [-ν] [-]E [Mpa] E [Mpa] ν ν[-][-] RSMRSM switchingswitching SOLID187SOLID187 (Tetrahedral (Tetrahedral RSM switching 360360,000,000 SOLID187 (Tetrahedral 200200,000,000 0.3 0.3 169 169,000,000 0.2750.275 rodrod sideside 360,000 andand Hexahedral) Hexahedral) 200,000 0.3 169,000 0.275 rod side and Hexahedral) RSM detection SOLID187 (Tetrahedral RSM detection 485,000 SOLID187 (Tetrahedral 200,000 0.3 169,000 0.275 RSM roddetection side 485,000 SOLID187and Hexahedral)(Tetrahedral 200,000 0.3 169,000 0.275 rod side 485,000 and Hexahedral) 200,000 0.3 169,000 0.275 rod side and Hexahedral)

InIn the the case case of of the the switching switching rod,rod, the the stressesstresses on on thethe pinpin andand thosethose onon thethe hammerhammer werewere sampled,sampled, In the case of the switching rod, the stresses on the pin and those on the hammer were sampled, andand subsequentlysubsequently comparedcompared with the the experimental experimental outcomes. outcomes. In Inthe the case case of ofthe the detection detection rod, rod, the and subsequently compared with the experimental outcomes. In the case of the detection rod, the thestresses stresses on the on pins the pinsthat connect that connect the lower the lowerplate to plate the body to the were body examined, were examined, as a function as a functionof the actual of stresses on the pins that connect the lower plate to the body were examined, as a function of the actual thebolt actual preload bolt of preload the joint. of the joint. bolt preload of the joint. 3. Results and Discussion 3. Results and Discussion 3. Results and Discussion The results from a tensile test carried out on the ad-hoc developed test bench are shown in The results from a tensile test carried out on the ad-hoc developed test bench are shown in FigureThe9 . results from a tensile test carried out on the ad-hoc developed test bench are shown in Figure 9. Figure 9.

(a) (b) (a) (b) Figure 9. (a) View of the test bench and (b) plot of the results in terms of stresses on the pin and on Figure 9. (a) View of the test bench and (b) plot of the results in terms of stresses on the pin and on Figurethe arms 9. ( anda) View force of at the the test load bench cell. and (b) plot of the results in terms of stresses on the pin and on the the arms and force at the load cell. arms and force at the load cell. Figure 9 reports the data relevant to a test run until a maximum force of F = 160 kN. At the peak Figure 9 reports the data relevant to a test run until a maximum force of F = 160 kN. At the peak load, one of the pins connecting the lower plate with the body failed. The first outcome of the load, one of the pins connecting the lower plate with the body failed. The first outcome of the

Machines 2018, 6, 6 7 of 9

MachinesFigure 2018,9 6 ,reports x FOR PEER the REVIEW data relevant to a test run until a maximum force of F = 160 kN. At7 the of 9 peak load, one of the pins connecting the lower plate with the body failed. The ﬁrst outcome of the experiment is the knowledge of the force distributiondistribution on the two arms: the great majority of the total force (82%) reaches the body by passing through the chain of components named named the the switching switching rod, rod, the pin and the hammer. The remaining part (18%) passes through the detection rod, the slider and the lower plate, eventually reaching the body. Running each of the FEA models by applying the appropriate fraction of the total load to the arm under investigation, it was possible to validate the numerical results. For example, looking at Figure 1010aa,, it is possibl possiblee to observe the equivalent stresses calculated according to the von Mises criterion on the half machine comprising the the switching switching rod. rod. Figure 1010bb reportsreports the theσ YσY bending bending stresses stresses on theon the pin pin that that supports supports the hammer, the hammer, when when this sub-system this sub- issystem loaded is withloaded 82% with the 82% total the load total F = load 160 kN.F = 160 kN.

(a) (b)

Figure 10. (a) von Mises stress plot on the half machine—switching rod side, and (b) bending σY Figure 10. (a) von Mises stress plot on the half machine—switching rod side, and (b) bending σY stresses on the pin. stresses on the pin.

As can be appreciated from Figure 10b, the numerical peak of the bending stress on the pin (σY_FEAAs can = 477 be appreciatedMPa) is very from close Figure to that 10b, measured the numerical during peak the of experimental the bending test stress on on the the same pin (componentσY_FEA =477 (σY_EXP MPa) is = very 450 closeMPa, to see that Figure measured 9). The during error, the calculated experimental according test on the to sameEquation component (1), is (acceptable.σY_EXP = 450 MPa, see Figure9). The error, calculated according to Equation (1), is acceptable.

σY_FEA − σY_EXP e% = Y _ FEA Y _ EXP · 100 = 6% (1) e%  σY_EXP 100  6% (1) Y _ EXP Once the FEA model has been validated, it can be used for carrying out some comparisons considering,Once the for FEA example, model the has joint been between validated, the it lower can be plate used and for the carrying body. Suchout some joints comparisons comprise a patternconsidering, of eight for M8 example, 8.8. screws, the joint working between in parallel the lower with aplate couple and of parallelthe body. pins Such of d joint = 6 mms comprise diameter, a manufacturedpattern of eight according M8 8.8. to screws, Standard working [10]. It can in parallel be assumed with that a couple this joint of must parall withstandel pins of the d shearing= 6 mm loaddiameter, transmitted manufactured by the slider according to the to body Standard via the [10] lower. It plate.can be These assumed pins that are coupledthis joint with must interference withstand (H7/m6).the shearing Since load the transmitted screws are tightenedby the slider under to preloadthe body control via the upon lower assembly, plate. These and some pins uncertaintiesare coupled with regardinterference to the (H7/m6). friction coefﬁcientsSince the screws cannot are be tightened avoided under [11,12], preload the load control borne upon by the assembly, parallel pinsand maysome vary uncertainties depending with on the regard effective to the preload friction of thecoefficients screws and cannot on the be friction avoided coefﬁcient [11,12], the at the load interface borne betweenby the parallel the body pins and may the vary plate. depending In order toon estimate the effective such variation,preload of some the screws parametric and on analyses the friction were run,coefficient for example at the byinterface imposing between different the preloadbody and levels the plate. on the In screws. order to The estimate screw preloadsuch variation, was assigned some viaparametric the preload analyses tool availablewere run, in for the example Ansys Workbenchby imposing environment. different preload Figure levels 11 reports on the ascrews. plot of The the amountscrew preload of shearing was assigned force borne via bythe the preload switching tool rodavailable side pin in (T’swi)the Ansys and Workbench by the detection environment. rod side pinFigure (T’det) 11 reports as functions a plot ofof thethe actualamount screw of shearing preload force Fv. Each borne of by the the dashed switching lines rod represents side pin the (T’swi) force actingand by on the the detection relevant rod parallel side pin,pin whereas(T’det) as the functions solid lines of representthe actual the screw fraction preload of load Fv. borneEach byof the genericdashed lines pin. represents the force acting on the relevant parallel pin, whereas the solid lines represent the fraction of load borne by the generic pin.

Machines 2018, 6, 6 8 of 9

Machines 2018, 6, x FOR PEER REVIEW 8 of 9

FigureFigure 11. Shearing 11. Shearing force force on the on switching/detectionthe switching/detection rod rod sideside parallel pin pin versus versus screw screw preload. preload.

ItFigure can be11 . seenShearing that force the moston the loadedswitching/detection pin is that rodon theside detection parallel pin rod versus side screw (closer preload. to the slider), Itregardless can be seen of the that screw the preload. most loaded Nonetheless, pin is the that magnitude on the detection of the load rod borne side by (closer the pins to decreases the slider), regardlessas theIt ofcan screw the be screw seenpreload that preload. increases: the most Nonetheless, aloaded preload pin limit theis that of magnitude Fv on = the 20 kNdetection of is theassumed loadrod side bornebased (closer on by the theto provisionsthe pins slider), decreases of as theregardless screwStandard preload of [13] the for screw increases: M8, preload.8.8 class apreload Nonetheless,screws. B limitased the ofon magnitude Fvdifferent = 20 kN preload ofis the assumed load levels, borne it based is byalso the onpossible pins the decreases provisions to extract of Standardasa the plot [ 13screw of] forthe preload M8,von 8.8Mises increases: class stresses screws. a onpreload the Based most limit on loaded of different Fv =parallel 20 preloadkN pin,is assumed as levels, shown based it in is Figure also on the possible 12. provisions to extract of a plot ofStandard the von [13] Mises for M8, stresses 8.8 class on thescrews. most Based loaded on different parallel preload pin, as shownlevels, it in is Figurealso possible 12. to extract a plot of the von Mises stresses on the most loaded parallel pin, as shown in Figure 12.

(a) (b) (c)

Figure 12. von Mises stress plot on the detection rod side parallel pin at a screw preload of (a) 10 kN; (a) (b) (c) (b) 15 kN; (c) 20 kN. Figure 12. von Mises stress plot on the detection rod side parallel pin at a screw preload of (a) 10 kN; Figure 12. von Mises stress plot on the detection rod side parallel pin at a screw preload of (a) 10 kN; (bThe) 15 kN equivalent; (c) 20 kN stress. calculated by FEA on the most loaded parallel pin is compatible with the (bfailure) 15 kN; event, (c) 20 which kN. took place during the experiment at a total load of F = 160 kN. The strength of theThe pins equivalent could be modified stress calculated by changing by FEA the oncoupling the most system, loaded increasing parallel pinthe isinterference compatible or with adopting the Thefailurea different equivalent event, coupling which stress took technique. calculated place during Based by the FEA on experiment the on literature, the most at a a loadedtotal valid load alternative parallel of F = 16 pin could0 kN is compatible. beThe making strength use with of of the failuretheanaerobic event, pins could which or be epoxy modified took adhesives, place by duringchanging which the the wo experiment couplinguld make system, it at possible a totalincreasing loadto significantly ofthe F interference = 160 increase kN. or The adopting the strength actual of the pinsa matingdifferent could area becoupling modiﬁedwith atechnique. positive by changing outcome Based on the in the terms coupling literature, of the system, overalla valid increasing strength.alternative This the could interferencepoint be hasmaking been or use tackled adopting of a differentanaerobicexperimentally coupling or epoxy in technique. papers adhesives, [14– Based which16], which onwould thealso make literature,provide it possible tips a regarding valid to significantly alternative the proportioning increase could be theof making t actualhe joint use matingupon itsarea design. with a positive outcome in terms of the overall strength. This point has been tackled of anaerobic or epoxy adhesives, which would make it possible to signiﬁcantly increase the actual experimentally in papers [14–16], which also provide tips regarding the proportioning of the joint mating area with a positive outcome in terms of the overall strength. This point has been tackled upon4. Conclusion its design. s experimentally in papers [14–16], which also provide tips regarding the proportioning of the joint From a designer’s standpoint, the present work achieved a twofold result: (i) an experimental upon4 its. Conclusion design. s setup has been designed, manufactured and calibrated, which is novel and original and will be useful 4. Conclusionsfor Fromsubsequent a designer’s experimentations standpoint, onthe other present products work ofachieved the same a family;twofold (ii) result: numerical (i) an toolsexperimental have been setupdeveloped has been and designed, validated, manufactured with respect and to calibrated, experimental which data. is novel These and models original allow and thewill designerbe useful to Fromforevaluate subsequent a designer’s the effect exper ofimentations standpoint, structural onchanges the other present productsearly,work hence of the achievedreducing same family; the a twofold time (ii) to numerical mark result:et of (i)tools new an havemachines. experimental been setupdeveloped has been designed,and validated, manufactured with respect and to experimentalcalibrated, which data. isThese novel models and original allow the and designer will be to useful for subsequentevaluate the experimentations effect of structural onchanges other early products, hence of reducing the same the family; time to (ii)mark numericalet of new toolsmachines. have been developed and validated, with respect to experimental data. These models allow the designer to evaluate the effect of structural changes early, hence reducing the time to market of new machines. Machines 2018, 6, 6 9 of 9

Acknowledgments: The authors gratefully acknowledge Eng. Marcello Andrenacci, Eng. Leonardo Bozzoli, Eng. Francesco Muscatello and Eng. Francesca Sopranzetti at Alstom Ferroviaria SpA for having made this research possible. The authors would also like to acknowledge Eng. Francesco Vai, laboratory director at the Department of Industrial Engineering, University of Bologna, for his fundamental contribution to the experimental activities. Author Contributions: D.C. and M.D.A. conceived and designed the experiments; S.F. and M.D.A. performed the experiments; M.D.A., S.F., and F.R. performed the numerical analyses; G.O. analyzed the data; S.F., and M.D.A. provided reagents, materials and analysis tools; G.O. and M.D.A. wrote the paper. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Niemann, G.; Winter, H.; Hohn, B.R. Maschinenelemente: Band 1: Konstruktion und Berechnung von Verbindungen, Lagern, Wellen; Springer-Verlag: Berlin, Germany, 2005. 2. De Agostinis, M.; Fini, S.; Olmi, G. The influence of lubrication on the frictional characteristics of threaded joints for planetary gearboxes. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 2016, 230, 2553–2563. [CrossRef] 3. Hot Rolled Products of Structural Steels, Part 2: Technical Delivery Conditions for Non-Alloy Structural Steels; UNI EN 10025-2:2005; British Standards Institution: London, UK, 2005. 4. Eurocode 3, Design of Steel Structures, Part 1-8: Design of Joints; UNI EN 1993-1-8:2005; Europian Comittee for Standardisation: Brussels, Belgium, 2005. 5. Croccolo, D.; De Agostinis, M.; Vincenzi, N. Structural analysis of an articulated urban bus chassis via FEM: a methodology applied to a case study. Stroj. Vestnik J. Mech. Eng. 2011, 57, 799–809. [CrossRef] 6. Zhao, X.; Li, F.; Fan, Y.; Liu, Y. Fatigue behavior of a box-type welded structure of hydraulic support used in coal mine. Materials 2015, 8, 6609–6622. [CrossRef][PubMed] 7. Non Destructive Testing—Inspection by Strain Gauges: Terms and Definitions; UNI EN 10478-1:1996; Italian Standards: Milano, Italy, 1996. 8. Non Destructive Testing—Inspection by Strain Gauges: Selection of Strain Gauges and Accessory Components; UNI EN 10478-2:1998; Italian Standards: Milano, Italy, 1996. 9. Non Destructive Testing—Inspection by Strain Gauges: Strain Gauge Installation and Checking; UNI EN 10478-3:1998; Italian Standards: Milano, Italy, 1998. 10. Parallel Pins, of Hardened Steel and Martensitic Stainless Steel; ISO 8734; International Organization for Standardization: Geneva, Switzerland, 1997. (Dowel pins) 11. Croccolo, D.; De Agostinis, M.; Fini, S.; Olmi, G. Tribological properties of bolts depending on different screw coatings and lubrications: An experimental study. Tribol. Int. 2017, 107, 199–205. [CrossRef] 12. Eccles, W.; Sherrington, I.; Arnell, R.D. Frictional changes during repeated tightening of zinc plated threaded fasteners. Tribol. Int. 2010, 43, 700–707. [CrossRef] 13. Mechanical Properties of Fasteners Made of Carbon Steel and Alloy Steel—Part 1: Bolts, Screws and Studs with Specified Property Classes—Coarse Thread and Fine Pitch Thread; ISO 898-1; International Organization for Standardization: Geneva, Switzerland, 2009. 14. Croccolo, D.; De Agostinis, M.; Mauri, P.; Olmi, G. Influence of the engagement ratio on the joint strength of press fitted and adhesively bonded specimens. Int. J. Adhes. Adhes. 2014, 53, 80–88. [CrossRef] 15. Croccolo, D.; De Agostinis, M.; Fini, S.; Olmi, G. Influence of the engagement ratio on the shear strength of an epoxy adhesive by push-out tests on pin-and-collar joints: Part I: Campaign at room temperature. Int. J. Adhes. Adhes. 2016, 67, 69–75. 16. Croccolo, D.; De Agostinis, M.; Fini, S.; Olmi, G. Influence of the engagement ratio on the shear strength of an epoxy adhesive by push-out tests on pin-and-collar joints: Part II: Campaign at different temperature levels. Int. J. Adhes. Adhes. 2016, 67, 76–85. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).