Comparative Study of Flank Cams Manufactured by WEDM and Milling Processes

metals

Article Comparative Study of Flank Cams Manufactured by WEDM and Milling Processes

, Irene Buj-Corral * † , Enrique Zayas-Figueras † and Àngels Montaña-Faiget Department of Mechanical Engineering, School of Engineering of Barcelona (ETSEIB), Universitat Politècnica de Catalunya (UPC), Av. Diagonal, 647, 08028 Barcelona, Spain; [email protected] (E.Z.-F.); [email protected] (À.M.-F.) * Correspondence: [email protected]; Tel.: +34-934-054-015 Equal contribution to this paper. † Received: 29 July 2020; Accepted: 22 August 2020; Published: 27 August 2020

Abstract: Cam-follower mechanisms are usually employed in different machines, like combustion engines, sewing machines, machine tools, etc. In the present paper, the option to manufacture cams utilizing wire electrical discharge machining (WEDM) has been considered. For this, surface roughness and shape error of cam profiles manufactured by the processes of milling and wire electrical discharge machining (WEDM) are presented. The methodology used covers different stages: design, prototyping, manufacturing, and measurement of the cams. As a reference, a cam-follower mechanism from a motorcycle internal combustion engine has been used. A reverse engineering process has been performed to determine the geometrical parameters of the mechanism, which are used for the synthesis of the profile of the cam and its subsequent design. The manufacturing process of the cams has been assisted by CAD-CAM (Computer Assisted Drawing-Computer Assisted Manufacturing) software. Using fused filament fabrication (FFF), a physical prototype of the cam has been manufactured, in order to validate the goodness of the design. Finally, the roughness and shape parameters have been measured on the contour surface of the cams. The arithmetical mean roughness Ra value of the milled cam was 0.269 µm, below the requirement of 0.4 µm, and shape error was 18 µm, below 50 µm. Shape error of the WEDM cam of 48 µm meets the requirements for cams. However, the Ra value of 1.212 µm, exceeded the limit. For this reason, a finish operation is recommended in this case. Some advantages of WEDM cams over milled cams are that different conductive materials can be employed, more complex shapes can be obtained, and that, in rough operations, the amount of material to be removed in subsequent operations is considerably reduced.

Keywords: cam proﬁle design; milling; WEDM; surface roughness; shape error

1. Introduction The cam-follower mechanisms have wide applications in different kinds of machines such as alternative combustion engines, sewing machines, machine tools, etc., due to their advantages. For instance, they are simple and compact mechanisms, they are cheap, they have few mobile parts, and the follower displacement law is easy to design. The cam mechanisms have three parts: a cam, a follower, and a fixed frame [1]. The most common cam-follower mechanism that is employed in combustion engines consists of a disk cam and a follower with translating or oscillating movement (Figure1a). The cams for combustion engines are usually symmetric, while the ones that are used in high speed applications or aeronautics are asymmetric [2]. The cams for combustion engines can be classified into three categories: convex flank cams, tangential flank cams, and concave flank cams. In the present work, a tangential flank cam has been used (Figure1b).

Metals 2020, 10, 1159; doi:10.3390/met10091159 www.mdpi.com/journal/metals Metals 2020,, 10,, 1159x FOR PEER REVIEW 2 of 18

(a) (b)

Figure 1. (a) Parts of a disk camcam mechanismmechanism with an oscillatingoscillating roller follower;follower; ((b)) typicaltypical tangentialtangential ﬂankflank cam from a combustion engine.

In the past,past, thethe availableavailable technologiestechnologies made it diffifficultcult to obtainobtain cams with preciseprecise geometriesgeometries and surface finish.finish. At present, thanks to the numericalnumerical control (NC) machines, this phase has been successfully overcome. Today Today the the te tediousdious calculations calculations of of the the cam design process are assisted by software thatthat allowsallows toto obtain obtain the the profiles profiles and and the the graphics graphics in in a fastera faster and and more more precise precise way way [2]. [2]. In the In 70sthe of70s the oflast thecentury, last century, Grant Grant has analyzed has analyzed the diff erentthe different methods methods to manufacture to manufacture cams [3]. cams He classified [3]. He themclassified into threethem categories:into three categories: methods for methods original camsfor original (using conventionalcams (using conventional machines and machines manual finish and operations),manual finish methods operations), for copied methods cams for (using copied milling cams machines(using milling with indexingmachines heads with indexing and/or copying heads mechanisms,and/or copying as mechanisms, well as grinding as well machines), as grinding and machines), methods and that methods used numerical that used controlnumerical machines. control Thesemachines. machines These have machines evolved have considerably evolved considerably over time until over today, time until and itstoday, usehas and become its use widespreadhas become inwidespread the field of in cam the manufacturing. field of cam On manufacturing. the other hand, On for largethe other batches, hand, other for manufacturing large batches, processes other aremanufacturing employed, like processes forging are [4], employed, sintering [5 like], etc. forging Wire Electric [4], sintering Discharge [5], Machiningetc. Wire Electric (WEDM) Discharge consists ofMachining using a thin (WEDM) wire of consists a conductive of using material a thin wire like brass,of a conductive which is wound material between like brass, two which spools, isin wound order tobetween cut a conductivetwo spools, materialin order usingto cut electricala conductive discharges. material Ausing dielectric electrical fluid discharges. such as water A dielectric allows isolatingfluid such the as wirewater from allows the isolating workpiece the towire achieve from athe high workpiece current to density, achieve it a cools high the current heated density, surface it ofcools the wirethe heated and it removessurface theof the material wire particles.and it re Duemoves to thethe high material temperatures particles. reached, Due to the the surface high layertemperatures melts and reached, then it the quickly surface cools layer because melts and of the then fluid, it quickly leading cools to thebecause so-called of the recast fluid, layer leading [6]. Heatto the canso-called also induce recast layer tensile [6]. residual Heat can stresses, also indu cracks,ce tensile or microporosityresidual stresses, on cracks, the parts’ or microporosity surface [7,8]. Someon the advantages parts’ surface of WEDM [7,8]. Some processes advantages are that of they WEDM allow processes machining are di ffthaterent they conductive allow machining materials regardlessdifferent conductive of their hardness materials [9 ],regardless for example, of their titanium hardness alloys [9], used for example, in biomedical titanium applications alloys used [10 in], andbiomedical that it is applications appropriate [10], for delicate and that or fragileit is appropriate materials, becausefor delicate it eliminates or fragile the materials, mechanical because stresses it thateliminates are present the mechanical in traditional stresses machining that are processes present [11 in]. traditional Besides, complex machining shapes processes and thin-walled [11]. Besides, parts cancomplex be obtained, shapes and such thin-walled as noncircular parts gears, can be including obtained, elliptical such as and noncircular oval gears gears, [12] orincluding inclined elliptical surfaces thatand areoval used gears in tooling [12] or that inclined requires surfaces draft angle that [ 13ar].e Moreover,used in tooling WEDM that avoids requires the need draft to manufacture angle [13]. theMoreover, pre-shaped WEDM electrodes avoids the that need are necessaryto manufacture for sink the electricalpre-shaped discharge electrodes machining that are necessary (SEDM) [ 14for]. Insink recent electrical years ,discharge WEDM has machining started to(SEDM) replace [14]. conventional In recent machiningyears, WEDM in some has started rough operations,to replace forconventional example, formachining the manufacture in some rough of blades, operations, with the for advantageexample, for that the the manu amountfacture of of material blades, towith be removedthe advantage in subsequent that the amount operations of material is considerably to be removed reduced in [15 subsequent]. The main operations parameters is of considerably the WEDM processreduced are:[15]. dischargeThe main current,parameters servo of the voltage, WEDM pulse proc oness and are: odischargeff time, and current, wire feed,servo amongvoltage, others. pulse Severalon and authorsoff time, haveand studiedwire feed, the among effect ofothers. WEDM Several parameters authors on have surface studied roughness. the effect For of example, WEDM Hegabparameters et al. on [16 ]surface optimized roughness. material For removal example, rate, Hegab wear electrode et al. [16] ratio, optimized and average material surface removal roughness rate, Rawearas electrode a function ratio, of machining and average on-time, surface discharge roughness current, Ra as a voltage, function total of machining depth ofcut on-time, and percentage discharge incurrent, weight voltage, of carbon total nanotube depth of composites cut and percentage (CNT) added in weight to an of aluminum carbon nanotube base. Sonawane composites et al.(CNT) [17] consideredadded to an WEDM aluminum parameters base. Sonawane such as pulse-onet al. [17] time,considered servo voltage,WEDM parameters pulse-off time, such peak as pulse-on current, time, servo voltage, pulse-off time, peak current, feed rate, and cable tension and responses like surface roughness, overcut and material removal rate, in Nimonic-75 alloy. Magabe et al. [18]

Metals 2020, 10, 1159 3 of 18 feed rate, and cable tension and responses like surface roughness, overcut and material removal rate, in Nimonic-75 alloy. Magabe et al. [18] investigated the effect of spark gap voltage, pulse on-time, pulse off-time, and wire feed on the material removal rate and surface roughness of a Ni-Ti shape memory alloy. Chaudhary et al. [19] considered pulse-on time, pulse-off time, and current as process parameters, whereas material removal rate (MRR), surface roughness, and micro-hardness as responses, in superelastic Nitinol shape memory alloy. Measurement of the shape error of cams is important because better accuracy of the cam profile obtained will favor the compliance of the follower with the motion-law and, consequently, a better operation of the mechanism. Most authors specify a range of tolerances for the manufacture of a cam by means of geometric tolerances, like the shape error, that is, they define a band along the entire contour that validates the functionality of the piece if the measured points on the contour are within the marked band. For example, Norton stated that shape error of 25 µm or lower (total error of ± 50 µm) is appropriate for cams [2]. Kang and Han [20] measured form deviation errors of a disk cam on a profile-measuring machine. They found a total form deviation error of 0.1874 mm. Hsieh and Lin [21] proposed and validated a method to carry out the automatic measurement of a cylindrical cam on a coordinate measuring machine (CMM); the method uses a homogenous transformation matrix to derive the CMM ability function matrix and the measuring probe location matrix. The cam was machined on a four-axis vertical machine tool. The maximum deviation between the experimental and theoretical profiles was 21 µm. Other authors, such as Chang and Wu, specify measurements based on dimensional tolerances [22]. The maximum radial dimension of a cam was set to 19.05 mm in this case. Taking into account IT9 for WEDM and IT7 for CNC machining processes, tolerance amplitude values of 52 µm and 21 µm, respectively, were defined. In another example, Le and Nguyen Tank [23] obtained dimensional tolerances for a milled planar cam (with a maximum dimension of 150 mm) of at least 0.018 mm, corresponding to IT5. Regarding the shape error of WEDM curved surfaces. Werner [24] manufactured internal contours and measured shape errors of the profile with a coordinate measuring machine (CMM). He found deviation values between +0.008 mm and 0.010 mm with respect to the − theoretical curves. In a comparative study regarding miniature brass gears manufactured using either WEDM and hobbing [25], it was observed that WEDM gears had higher quality than hobbed gears. Form error found was 5.4 µm for the WEDM gears [26]. As for surface roughness of cams, Islam et al. [27] made a numerical approach based on mixed lubrication concept considering roller sliding to predict the friction, taking into account the friction from cam/roller contact, cam bearings, valve/guide, needle roller bearing, and the effect of asperity interaction and roller sliding. A reduction in friction was obtained with a rise in the camshaft speed. Furthermore, they observed that an increased asperity interaction at high operating temperature yielded to relatively higher friction magnitude, especially in the nose of the cam profile, due to a lower curvature radius. Torabi et al. [28] presented a thermo-elasto-hydrodynamics analysis to study the behavior of the cam-and- follower contact, particularly during the running-in period. Their results have shown that the rate of flattening of surface roughness is a crucial factor influencing thermal effects during the running-in period. Therefore, according to the literature cited here, it is important to control the asperity (surface roughness) in the cam profile. Surface finish also influences the lubrication and aesthetic appearance of the cam. In contacts such as those between cam and follower, high surface roughness can lead to high local pressure and a reduction in the real contact area between cam and follower [29]. For this reason, low roughness is required in this case. Agulló and Cardona [30] stated that, when the unit cost is important, an appropriate surface finish for steel cams would be Ra < 0.4 µm. Traditionally, a grinding operation was necessary in order to achieve such roughness values [2]. In a comparative study for cams, Ra = 1.09 µm were obtained in the outer surface of a milled cam, and Ra = 1.42 µm in a WEDM cam, both made of AA6063 aluminum [31]. In another comparative study between hobbing and WEDM for the manufacture of gears, better surface finish was found for WEDM [26]. The average surface roughness was 1 µm, and maximum surface roughness Metals 2020, 10, 1159 4 of 18

wasMetals 6.4 2020µ,m 10, in x FOR this PEER case. REVIEW In another work about gear prototypes, Ra = 0.42 µm was reported for4 of the 18 WEDM parts [32]. Machining processesprocesses have have traditionally traditionally been been used used to to obtain obtain cams. cams. However, However, other other technologies technologies like electriclike electric discharge discharge machining machining could replacecould conventionalreplace conventional cutting processes cutting withprocesses important with advantages, important suchadvantages, as the possibility such as the to usepossibility hard materials, to use hard the possibilitymaterials, tothe design possibility more to complicated design more shapes, complicated and the reductionshapes, and of mechanicalthe reduction stresses. of me Untilchanical now, stresses. few works Until are now, known few about works the are manufacture known about of cams the usingmanufacture WEDM of [31 cams]. The using main WEDM objective [31]. of the The present main paperobjective is to of consider the present the possibility paper is to manufactureconsider the tangentialpossibility flankto manufacture cams by means tangential of WEDM, flank and cams to compare by means them of WEDM, with milled and cams, to compare in terms them of surface with finishmilled andcams, shape in terms error of of surface their external finish and contours. shape erro Ther geometry of their external of a cam contours. profile hasThe been geometry designed, of a takingcam profile as reference has been a realdesigned, cam from taking a camshaft as reference of a a motorcycle real cam from engine. a camshaft The structure of a motorcycle of the paper engine. is as follows:The structure Section of 2the describes paper is theas follows: methods Section that are 2 describes used to manufacture the methods the that cams, are used by means to manufacture of either millingthe cams, or by WEDM, means as of well either as themilling way theor WEDM, roughness as andwell shapeas theerror way measurementsthe roughness wereand shape performed. error Sectionmeasurements3 shows were the resultsperformed. of visual Section inspection, 3 shows the surface results roughness, of visual andinspection, shape errorsurface of roughness, the cams; Sectionsand shape4 and error5 contain of the cams; the discussion Sections 4 and and conclusions, 5 contain the respectively. discussion and conclusions, respectively.

2. Materials and Methods Figure2 2 presents presents thethe generalgeneral methodologymethodology of of the the present present work. work.

Figure 2. The general methodology of the paper.

The exposedexposed methodology methodology consists consists of six of steps, six step whichs, which will be explainedwill be explained in the following in the subsections. following First,subsections. a reference First, cam-follower a reference mechanism cam-follower was selected. mechanism Then, was using selected. reverse engineering,Then, using a newreverse cam proﬁleengineering, was designed. a new cam Next, profile a three-dimensionalwas designed. Next (3D), a three-dimensional printed prototype (3D) was printed obtained prototype to validate was theobtained design. to validate Then, the the manufacturing design. Then, processthe manufa of thecturing cams, process obtained of the by eithercams, millingobtained or by WEDM, either wasmilling deﬁned. or WEDM, The camswas defined. were manufactured, The cams were and manufactured, both roughness and and both shape roughness error ofand their shape external error contourof their external were measured. contour were Finally, measured. conclusions Finall werey, conclusions drawn from were the results.drawn from the results.

2.1. Selection of a Reference Cam-Follower Mechanism As a start point, a reference cam-follower mechanism was selected, a real cam from the camshaft of aa single-cylinder single-cylinder internal internal combustion combustion engine engine of a motorcycle of a motorcycle (Figure3 ),(Figure available 3), at available the Department at the Department of Mechanical Engineering (DEM) of the School of Engineering of Barcelona (ETSEIB). Because of the absence of geometric data of the reference cam and follower mechanism, as well as about their displacement law, reverse engineering was used to determine the required start

Metals 2020, 10, 1159 5 of 18 of Mechanical Engineering (DEM) of the School of Engineering of Barcelona (ETSEIB). Because of theMetals absence 2020, 10, ofx FOR geometric PEER REVIEW data of the reference cam and follower mechanism, as well as about5 their of 18 displacementMetals 2020, 10, x law,FOR PEER reverse REVIEW engineering was used to determine the required start information for5 of the 18 designinformation process for of the the cam.design Di ﬀprocesserent parameters of the cam. of theDifferent real cam-follower parameters mechanism of the real were cam-follower measured information for the design process of the cam. Different parameters of the real cam-follower (seemechanism Figure4 wereand Table measured1), with (see a MitutoyoFigure 4 and digital Tabl micrometer.e 1), with a Mitutoyo digital micrometer. mechanism were measured (see Figure 4 and Table 1), with a Mitutoyo digital micrometer.

Reference cam Reference cam

Figure 3. Camshaft of a real cam-follower mechanism used as a reference to the study. Figure 3. CamshaftCamshaft of of a a real real cam-follower mechanis mechanismm used as a reference to the study.

(a) (b) (a) (b) Figure 4. (a(a) )Measured Measured parameter parameter values values in inthe the reference reference cam-follower cam-follower mechanism mechanism (length (length units units are Figure 4. (a) Measured parameter values in the reference cam-follower mechanism (length units are aremm); mm); (b) (segmentsb) segments of the of theprofile proﬁle and and maximum maximum follower follower angular angular displacement. displacement. mm); (b) segments of the profile and maximum follower angular displacement.

Table 1.1. Numerical values of the start design parametersparameters (according toto FigureFigure4 4a,b).a,b). Table 1. Numerical values of the start design parameters (according to Figure 4a,b). ParameterParameter Value Value Parameter Value Maximum angle of displacement of the rotating θ Maximum angle of displacement of the rotating follower θmáx12.13(°) 12.13 Maximum angle of displacementfollower θ of( )the rotating follower máx(°) 12.13 Base radius of mtheáx cam◦ Rb (mm) 14.04 BaseBase radius radius of theof the cam camR (mm) Rb (mm) 14.04 14.04 Number of sections of the profile (accordingb to the movement law) 3 Number of sections of the proﬁle (according to the Number of sections of the profile (according to the movement 3law) 3 Length movementof the follower law) arm lb (mm) 24.25 Length of the follower arm lb (mm) 24.25 LengthRadius of the of follower the roller arm Rlbr(mm) (mm) 24.25 10.00 Radius of the roller Rr (mm) 10.00 DistanceRadius between of the rollerrotationRr (mm) centers d (mm) 10.00 32.82 DistanceDistance between between rotation rotation centers centersd (mm) d (mm) 32.82 32.82 From these parameters, the design process of the mechanism has been carried out, in order to From these parameters, the design process of the mechanism has been carried out, in order to obtain the corresponding cam profile to be manufactured. obtain the corresponding cam profile to be manufactured.

Metals 2020, 10, 1159 6 of 18

From these parameters, the design process of the mechanism has been carried out, in order to obtain the corresponding cam profile to be manufactured. Metals 2020, 10, x FOR PEER REVIEW 6 of 18 2.2. Design of a New Cam Profile 2.2. Design of a New Cam Profile The design process of a cam-follower mechanism consists of three steps [33]: The design process of a cam-follower mechanism consists of three steps [33]: (a) to design the follower displacement law, (a) to design the follower displacement law, (b) to obtain the cam profile according to the designed displacement law, and (b) to obtain the cam profile according to the designed displacement law, and (c) to check that the curvature of the cam profile guarantees a correct cam-follower contact. (c) to check that the curvature of the cam profile guarantees a correct cam-follower contact. InIn this this work, work, the the Dynacam Dynacam 9.5.0.153 9.5.0.153 software software (Norton (Norton Associates Associates Engineering, Engineering, Norfolk, Norfolk, MA, MA, US) US) waswas used used to to design design the the cam cam [ 34[34].]. After After initial initial analysis, analysis, it it was was decided decided to to choose choose the the Polydyne Polydyne function function asas the the movement movement law law of of the the follower, follower, which which is is frequently frequently used used in in the the design design of of camshafts camshafts of of internal internal combustioncombustion engines engines and and in in cams cams of of textile textile machines machines [ 35[35].]. FigureFigure5 5shows shows the the kinematic kinematic graphics graphics obtainedobtained with with the the Dynacam Dynacam software. software. The The analysis analysis of of the the graphics graphics allows allows checking checking the the continuity continuity of of thethe movement movement law law until until the the second second derivative derivative [ 34[34].]. Then, Then, the the cam cam profile profile has has been been generated, generated, and and it it hashas been been checked checked if if the the cam cam curvature curvature was was appropriate appropriate to to ensure ensure good good performance performance (Figure (Figure6). 6).

FigureFigure 5. 5.Kinematic Kinematic graphics graphics that that have have been been obtained obtained with with Dynacam. Dynacam.

Metals 2020, 10, 1159 7 of 18 Metals 2020, 10, x FOR PEER REVIEW 7 of 18

(a) (b)

Figure 6. ((aa)) Cam-follower Cam-follower mechanism mechanism that that has has been been obtained obtained with with Dynacam; Dynacam; ( (b)) limit limit configurations conﬁgurations with the indications of the base andand nose circles inin thethe camcam proﬁle.profile.

The two performance quality indices of the mechanism,mechanism, pressure angle, and cam proﬁleprofile curvature radius, have been checked with Dynacam (Table(Table2 2).).

Table 2. Quality indices of the cam-follower mechanism.

ParameterParameter ValueValue Maximum pressure angle βmax (°) 27.7 (limit value = 35) Maximum pressure angle βmax (◦) 27.7 (limit value = 35) cmin MinimumMinimum radius radius of of curvature curvature of the camcam proﬁle profileR cminR (mm) (mm) 17.89 17.89 (greater (greater thanthan the the radius radius of of the the follower) follower)

The profile of the designed cam consists of two arcs of circles and two small segments that join The profile of the designed cam consists of two arcs of circles and two small segments that join them. These segments, called flanks, correspond to two circle arcs with a large radius of curvature them. These segments, called flanks, correspond to two circle arcs with a large radius of curvature that, that, at first glance, appear to be straight sections. These two flanks have been adjusted to two small at first glance, appear to be straight sections. These two flanks have been adjusted to two small tangent tangent straight segments with respect to the base circle (larger radius) and the nose circle (smaller straight segments with respect to the base circle (larger radius) and the nose circle (smaller radius) to radius) to simplify the manufacturing processes. Therefore, a tangential flank cam has been designed simplify the manufacturing processes. Therefore, a tangential flank cam has been designed (Figure1b). (Figure 1b). 2.3. Manufacturing of a Prototype of the Cam and Validation of the Design 2.3. Manufacturing of a Prototype of the Cam and Validation of the Design For the 3D design, a thickness value of 13 mm and a hole diameter of 12 mm have been selected, correspondingFor the 3D todesign, the measurements a thickness value of theof 13 model mm an cam.d a hole The diameter dimensions of 12 of mm the have keyway been have selected, been correspondingdefined according to the to DINmeasurements 6885 [36], whichof the ismodel equivalent cam. The to the dimensions withdrawn of standardthe keyway ISO have R773 been [37] (Figuredefined7 a).according to DIN 6885 [36], which is equivalent to the withdrawn standard ISO R773 [37] (FigureBased 7a). on the profile obtained, a virtual prototype of the cam has been created using SolidWorks 2016Based (Dassault on the Syst profileèmes obtained, SE, Vélizy-Villacoublay, a virtual prototyp France).e of the Subsequently, cam has been a created prototype using of theSolidWorks cam has 2016been (Dassault printed in Systèmes polylactic SE, acid Vélizy-Villacoublay, (PLA) with the fused France). filament Subsequently, fabrication a (FFF)prototype or fused of the deposition cam has beenmodeling printed (FDM) in polylactic technology acid (Figure (PLA)7b). with the fused filament fabrication (FFF) or fused deposition modelingFrom (FDM) a visual technology inspection, (Figure it was found7b). that the physical prototype was similar to the model cam, thus validating the goodness of the design.

Metals 2020, 10, 1159 8 of 18 Metals 2020, 10, x FOR PEER REVIEW 8 of 18

(a)

(b)

FigureFigure 7. 7. ((aa)) Cam Cam designed designed dimensions dimensions (l (lengthength units units are are mm); mm); and and ( (bb)) 3D 3D printed printed prototype. prototype. 2.4. Design of the Manufacturing Processes to Obtain the Cams From a visual inspection, it was found that the physical prototype was similar to the model cam, thus validatingThe cams the have goodness been of manufactured the design. with NC milling and with WEDM, respectively. For both processes, the machines, tools, and ﬁxtures to be used have been deﬁned, as well as the 2.4.cutting Design conditions. of the Manufacturing Table3 presents Processes the main to Obtain characteristics the Cams of the two processes.

The cams have Tablebeen 3.manufacturedMain parameters with of theNC manufacturing milling and processwith WEDM, of the cam. respectively. For both processes, the machines, tools, and fixtures to be used have been defined, as well as the cutting conditions. ProcessTable 3 presents the main characteristics NC Milling of the two processes. WEDM Machine Haas VM-2 Machining Center ONA UE205 TableEnd 3. Main milling parameters cutter of diameter of the 20 mmmanufacturing and 3 teeth process of the cam. Tools (rough contour milling) + end milling cutter of Brass wire of diameter 0.25 mm diameter 10 mm and 4 teeth (finish contour milling) Process NC Milling WEDM Fixture system Conventional vice + adaptable fixture Clamp + magnets Machine Haas VM-2 Machining Center ONA UE205 Cutting speed 120 m/min, feed 0.15 mm (rough), feed End milling0.05 mmcutter (finish), of axialdiameter depth of 20 cut mm 1 mm and (rough), 3 teethVoltage 130 V, feed speed 1.75 mm/min, Cutting conditions Brass wire of diameter Tools (rough contouraxial depth milling) of cut 14 mm + (finish),end milling radial depth cutter of cut of cutting speed 2.25 mm/min 0.15 mm (finish) 0.25 mm diameter 10 mm and 4 teeth (finish contour milling) Fixture A commonConventional high temperature vice structural+ adaptable steel fixture was used to manufacture Clamp the cams, + magnets C45E4 from ISO system 4957 [38]. Both the millingCutting and speed the WEDM 120 m/min, process feed are 0.15 described mm (rough), in more feed detail inVoltage the following 130 V, subsections.feed Cutting 0.05 mm (finish), axial depth of cut 1 mm (rough), axial speed 1.75 mm/min, 2.4.1.conditions Machining depth Process of cut 14 mm (finish), radial depth of cut 0.15 cutting speed 2.25 mm (finish) mm/min The CAM (Computer Aided Manufacturing) software Cimatron 11.0 (3D Systems, Rock Hill, SC, USA) has been used to generate the g-code for the machining canter. The raw material was a block of A common high temperature structural steel was used to manufacture the cams, C45E4 from ISO 4957 [38].

Metals 2020, 10, x FOR PEER REVIEW 9 of 18

Both the milling and the WEDM process are described in more detail in the following subsections.

2.4.1. Machining Process MetalsThe2020 ,CAM10, 1159 (Computer Aided Manufacturing) software Cimatron 11.0 (3D Systems, Rock Hill,9 ofSC, 18 USA) has been used to generate the g-code for the machining canter. The raw material was a block of dimensions 50 mm × 50 mm × 22 mm. The software simulates the programmed stages to evaluate × × possible tool-piece intersections or errors in the trajectory followed by the tool. Two operations were carried out, rough and ﬁnishfinish contourcontour milling.milling.

2.4.2. WEDM Because no simulation software was available for the WEDM process, the tool path of the wire and the contour contour of of the the cam cam have have been been manually manually programmed programmed using using the the assist assisteded programming programming of the of themachine. machine. For this, For the this, tangency the tangency points points of the ofprofile, the profile, that is, thatthe points is, the that points join that the different join the disectionsfferent sectionsof the cam of profile, the cam have profile, been havedetermined, been determined, and either circular and either or linear circular interpolation or linear interpolationhas been applied. has beenBefore applied. programming, Before programming, the dimensio thens dimensionsof the material, of the a material, 50 mm a× 5050 mm mm ×50 16 mm mm 16block, mm block,were × × wereintroduced introduced in the in program. the program. An excessively An excessively high cu hightting cutting speed speed value value was avoided was avoided to prevent to prevent the wire the wirefrom frombreaking. breaking. The wire The wirediameter diameter used usedwas 0.25 was mm. 0.25 mm. Only rough machining was used, because it was notnot possible to machine the full shape without reclamping it, which would have inducedinduced alignmentalignment errorserrors [[32].32].

2.5. Roughness and Shape Measurement

2.5.1. Roughness Measurement The surface finish ﬁnish of the cams has been measured and analyzed with a Taylor Hobson contact roughness meter—Formmeter—Form TalysurfTalysurf Series Series 2 2 model—using model—using a a contact contact inductive inductive sensor sensor with with a resolutiona resolution of 18of 18 nm nm in ain height a height range range of 1of mm. 1 mm. The cut-ocut-offﬀ value employed was 0.8 mm, with a measurement length of 4 mm, according to the ISO 42884288 standardstandard [ 39[39].]. One One measurement measurement was was carried carried out onout each on each of the of four the areas four of areas the cam of the contour cam (seecontour Section (see 2.5.2 Section), along 2.5.2), the longitudinalalong the longitudinal direction of direction the cam. Arithmeticalof the cam. Arithmetical mean height parametermean height Ra wasparameter considered. Ra was considered.

2.5.2. Shape Measurement 2.5.2. Shape Measurement The cam profile was measured in a Mitutoyo Euro CA5-44 coordinate measuring machine (CMM), The cam profile was measured in a Mitutoyo Euro CA5-44 coordinate measuring machine with Mcosmos v3.0 software (Mitutoyo Corporation, Kawasaki-shi, Kanagawa, Japan). The shape (CMM), with Mcosmos v3.0 software (Mitutoyo Corporation, Kawasaki-shi, Kanagawa, Japan). The error is defined as the distance between the highest and the lowest distortion along the complete shape error is defined as the distance between the highest and the lowest distortion along the measured profile. complete measured profile. The coordinate origin (xc1, yc1) was defined at the center of the cam hole, from which the positions The coordinate origin (xc1, yc1) was defined at the center of the cam hole, from which the positions (xi, yi, zi,) of the profile points were determined (Figure8). Later, 354 points of the contour of the piece (xi, yi, zi,) of the profile points were determined (Figure 8). Later, 354 points of the contour of the piece have been measured at a vertical distance of zi = 5 0.001 mm from the upper face of the cam. have been measured at a vertical distance of zi = 5 ±± 0.001 mm from the upper face of the cam.

Figure 8.8. ExampleExample ofof the the determination determination of of the the tangency tangency points points inthe in milledthe milled cam cam profile profile (units (units are mm). are mm). In the measurement process, the cam profile was divided into four segments (Figure9). Segment 1 corresponds to the base circle (in blue); segment 3 corresponds to the nose circle (in green); and, segments 2 (in orange) and 4 (in yellow), correspond to the straight flanks. Metals 2020, 10, x FOR PEER REVIEW 10 of 18

In the measurement process, the cam profile was divided into four segments (Figure 9). Segment Metals1 corresponds2020, 10, 1159 to the base circle (in blue); segment 3 corresponds to the nose circle (in green);10 and, of 18 segments 2 (in orange) and 4 (in yellow), correspond to the straight flanks.

Figure 9. Segments into which the measured cam profiles have been divided in the measurement process. Figure 9. Segments into which the measured cam profiles have been divided in the measurement Thisprocess. division of the measured profile in sections might introduce some geometrical errors. In order to minimize them, each tangency area has been measured twice, from the two corresponding This division of the measured profile in sections might introduce some geometrical errors. In adjacent areas, and the measurement having greater error has been discarded. order to minimize them, each tangency area has been measured twice, from the two corresponding Next, the procedure used in order to determinate the difference between the measured and the adjacent areas, and the measurement having greater error has been discarded. theoretical profiles is explained. From the CAD file, the four geometric entities that constitute the Next, the procedure used in order to determinate the difference between the measured and the theoretical cam profile have been defined: the base circle, the nose circle, and the two straight flanks. theoretical profiles is explained. From the CAD file, the four geometric entities that constitute the The circles are defined using three parameters: the (xc, yc) coordinates of the cam hole center and its theoretical cam profile have been defined: the base circle, the nose circle, and the two straight flanks. radius Rc. The straight sections have been defined using the A, B, and C parameters, according to the The circles are defined using three parameters: the (xc, yc) coordinates of the cam hole center and its equation of the line Ax + By + C = 0. radius Rc. The straight sections have been defined using the A, B, and C parameters, according to the On the other hand, the coordinates (x , y ) of the measured points of the cam profiles have been equation of the line Ax + By + C = 0. i i obtained with the CMM. Subsequently, a transformation (rotation and translation) of the measured On the other hand, the coordinates (xi, yi) of the measured points of the cam profiles have been points required, in order to make a superposition of the measured and the theoretical profile, so as to obtained with the CMM. Subsequently, a transformation (rotation and translation) of the measured minimize the differences between them. This minimization problem is solved with a code written in points required, in order to make a superposition of the measured and the theoretical profile, so as Matlab R2015b (Mathworks, Natick, MA, US). This code calculates the values of the parameters (x’, y’) to minimize the differences between them. This minimization problem is solved with a code written of the translation and the angle θ of the rotation that minimizes the distance between the measured in Matlab R2015b (Mathworks, Natick, MA, US). This code calculates the values of the parameters points and the theoretical profile, which is, the sum of the point-straight line and point-circle distances (x’, y’) of the translation and the angle θ of the rotation that minimizes the distance between the (Equation (1)). measured points and the theoretical ! profile," which is, #the sum! of the !point-straight line and point- x cos θ sin θ x x circle distances (Equation (1)). 0 = − i + 0 (1) y0 sin θ cos θ · yi y0 𝑥′ cos −sin 𝑥 𝑥 The equation of the perpendicular = distance between∙ + a point—straight line, dpoint-line(1)is 𝑦′ sin cos 𝑦 𝑦 (Equation (2)):

The equation of the perpendicular distance Abetweenx + By a+ point—straightC line, dpoint-line is (Equation d = (2) (2)): point line p − A2 + B2 |A𝑥 + B𝑦 +C| The equation of the perpendicular distance between a point—circle, dpoint-circle is (Equation (3)): 𝑑 = (2) √A +B q 2 2 The equation of thedpoint perpendicularcircle = (x distance0 xc) + between(y0 yc )a point—circle,Rc dpoint-circle is (Equation (3)):(3) − − − − 𝑑 =𝑥 −𝑥 + 𝑦 −𝑦 −𝑅 (3) where xc and yc are the coordinates of the center of the circle. whereTable xc and4 shows yc are the the values coordinates of x’ translation, of the centery’ translation, of the circle. and rotation θ that must be applied to the measuredTable points4 shows in the order values to minimize of x’ translation, the square y’ valuetranslation, of the and distance rotation between θ that both must of be the applied proﬁles, to whichthe measured is, the measured points in error. order to minimize the square value of the distance between both of the profiles, which is, the measured error.

MetalsMetals 20202020,, 1010,, 1159x FOR PEER REVIEW 1111 ofof 1818

Table 4. Values of x’ translation, y’ translation, and rotation that must be applied in order to minimize Tablethe square 4. Values of the of distancex’ translation, betweeny’ translation, theoretical and rotationmanufactured that must profiles. be applied in order to minimize the square of the distance between theoretical and manufactured proﬁles. Manufacturing Process x’ Translation (mm) y’ Translation (mm) Rotation θ (°) ManufacturingMilled Processcam x’ Translation−0.508 (mm) y’ Translation−0.153 (mm) Rotation114.342θ ( ◦) MilledWEDM cam cam 0.508 0.018 0.153 0.040 114.342118.227 − − WEDM cam 0.018 0.040 118.227 3. Results 3. ResultsIn the present section, first, the results of a visual inspection are presented, followed by the roughnessIn the presentmeasurements section, first, (micro the resultsgeometric of a visualerrors) inspection and the shape are presented, measurements followed (macro by the roughnessgeometric measurementserrors). (micro geometric errors) and the shape measurements (macro geometric errors).

3.1.3.1. Visual Inspection OnceOnce thethe two two cams cams have have been been obtained, obtained, visual visual inspection inspection has been has performed been performed in order in to evaluateorder to theevaluate appearance the appearance of the diff oferent the machiningdifferent machining processes. processes. InIn the milled cam, cam, two two main main flaws flaws have have been been detected. detected. The The first first one one corresponds corresponds to the to themark mark left leftby the by theadaptable adaptable clamp clamp that that was was used used to fix to the fix thepart part in the in themachining machining center. center. The Themark mark does does not notcomprise comprise the thewhole whole thickness thickness of ofthe the piece, piece, but but only only a alocal local area area (Figure (Figure 10).10). In In the the future, future, it isis recommendedrecommended toto placeplace aa smallsmall aluminumaluminum plateplate betweenbetween thethe clampclamp andand thethe workpiece,workpiece, inin orderorder toto minimizeminimize thethe eeffectffect ofof thethe clamp.clamp. On the other hand,hand, the contoured surface presentedpresented slightslight scratchesscratches thatthat cancan bebe attributedattributed to to an an incorrect incorrect setting setting of of the the tool tool or or tool tool wear wear [40 [40],], although although it wasit was apparently apparently in goodin good condition. condition.

Marks of the Mark of the clamp tool (scratches) (flat area)

FigureFigure 10.10. Picture of the cam surface withwith aa ﬂatflat markmark andand slightslight scratches.scratches.

InIn thethe WEDMWEDM cam,cam, aa geometricgeometric imperfectionimperfection hashas beenbeen observedobserved duedue toto thethe entryentry andand exitexit ofof thethe wirewire (Figure(Figure 11 11).). This This kind kind of of shape shape error error is inherent is inhere tont this to typethis oftype process of process if the pieceif the does piece not does contain not anycontain outer any corner outer at 90corner◦. For at this 90°. reason, For this it has reason, been decidedit has been to begin decided and endto begin the passes and end tangentially the passes to thetangentially surface of to the the piece, surface at one of the of thepiece, points at one of change of the points of shape, of change in order of to shape, reduce in as order much to as reduce possible as itsmuch influence as possible on the its quality influence of the on cam. the quality Figure 11of showsthe cam. a detailFigure of 11 the shows workpiece, a detail where of the the workpiece, expected shapewhere isthe shown expected in red. shape The is real shown surface in red. is slightly The real diff surfaceerent because is slightly the forcedifferent applied because by the the fixing force magnetapplied wasby the not fixing sufficient magnet to keep was the not piece sufficient completely to keep straight, the piece and completely its weight straight, has made and it tilt its slightly.weight Duringhas made the it cuttingtilt slightly. process, During a clamping the cutting nerve process, is maintained a clamping in thenerve area is (Figuremaintained 12). in the area (Figure 12).

Metals 2020, 10, 1159 12 of 18 Metals 2020, 10, x FOR PEER REVIEW 12 of 18 Metals 2020, 10, x FOR PEER REVIEW 12 of 18

2mm 2mm

Figure 11. Detail of the contour of the cam profile in the area where the wire starts and finishes cutting. Figure 11. DetailDetail of of the the contour contour of of the the cam cam profile proﬁle in in the the ar areaea where where the the wire wire starts starts and and finishes ﬁnishes cutting. cutting.

Figure 12. Detail of the required clamping nerve of the WEDM cam. Figure 12. DetailDetail of of the the required required clampi clampingng nerve of the WEDM cam. The WEDM surfaces offer a different appearance than the milled surface, with craters instead of The WEDM surfaces offer offer a different different appearance thanthan the milled surface, with craters instead of cutting tool marks [6], which are produced by the electrical discharges. This gives the parts a matte cutting tool marks [6], [6], which are produced by the electrical discharges. This gives the parts a matte color that is quite different from the bright appearance of the milled parts. color that is quite didifferentfferent from thethe brightbright appearanceappearance ofof thethe milledmilled parts.parts. 3.2. Roughness 3.2. Roughness Roughness

3.2.1. Roughness of the Milled Cam 3.2.1. Roughness Roughness of of the Milled Cam Table 5 shows the roughness results of the milled cam. The highest Ra value has been found in Table5 5 shows shows thethe roughnessroughness resultsresults ofof thethe milledmilled cam.cam. TheThe highesthighest Ra value has been found in segment 2, where the tool starts and ends the machining operation. For this reason, it does not really segment 2, where the tool starts and ends the machin machininging operation. For For this this reason, reason, it it does does not really really correspond to surface finish, but to a surface mark. correspond to surface ﬁnish,finish, but to a surface mark.

Table 5. Roughness Ra in the different segments of the milled cam. TableTable 5. Roughness Ra inin the the different different segments of the milled cam. Segment 1 2 3 4 Average SegmentSegment 1 1 22 3 34 Average 4 Average Ra (μm) 0.144 0.409 0.229 0.295 0.269 Ra (µm)Ra (μ 0.144m) 0.144 0.4090.409 0.229 0.229 0.295 0.295 0.269 0.269 The average value of parameter Ra of the four segments is 0.269 μm, which is lower than the The average value of parameter Ra of the four segments is 0.269 μm, which is lower than the required roughness value below 0.4 μm for cams [30]. This is due to the fact that a finish operation requiredThe roughness average value value of below parameter 0.4 μRam forof thecams four [30]. segments This is due is 0.269 to theµm, fact which that a is finish lower operation than the has been performed, with a small feed per tooth value [41]. hasrequired been roughnessperformed, value with belowa small 0.4 feedµm per for tooth cams [value30]. This [41]. is due to the fact that a finish operation has As an example, Figure 13 depicts the measured profile in Segment 1. beenAs performed, an example, with Figure a small 13 feeddepicts per the tooth measured value [41 profile]. in Segment 1. As an example, Figure 13 depicts the measured profile in Segment 1.

Metals 2020, 10, 1159 13 of 18 Metals 2020, 10, x FOR PEER REVIEW 13 of 18 Metals 2020, 10, x FOR PEER REVIEW 13 of 18 m) μ m) μ Height ( Height (

Length (mm) Length (mm) Figure 13. Graphical representation of the measured profile of the milled cam in Segment 1. Figure 13. Graphical representation of the measured profile of the milled cam in Segment 1. Figure 13. Graphical representation of the measured profile of the milled cam in Segment 1. The profile is quite regular, with consecutive tool marks of approximately 0.05 mm width, correspondingThe profile to the is feedquite value regular, employed, with consecutiv and peak toe tool valley marks distance of approximately of around 0.5 µ m0.05 as mm a general width, correspondingThe profile to isthe quite feed valueregular, employed, with consecutiv and peake totool valley marks distance of approximately of around 0.5 0.05μm asmm a general width, trend.corresponding Some irregularities to the feed are value observed, employed, mainly an deeperd peak to valleys, valley which distance can of be around attributed 0.5 μ tom scratchesas a general duetrend. to tool Some wear. irregularities The maximum are ob peakserved, to valley mainly distance deeper is valleys, around 2.4whichµm. can be attributed to scratches duetrend. to toolSome wear. irregularities The maximum are ob peakserved, to valleymainly distance deeper valleys,is around which 2.4 μ m.can be attributed to scratches due to tool wear. The maximum peak to valley distance is around 2.4 μm. 3.2.2. Roughness of the WEDM Cam 3.2.2. Roughness of the WEDM Cam 3.2.2.Table Roughness6 shows the of roughnessthe WEDM results Cam of the WEDM cam. Table 6 shows the roughness results of the WEDM cam. Table 6 shows the roughness results of the WEDM cam. Table 6. Roughness Ra in the different segments of the WEDM cam. Table 6. Roughness Ra in the different segments of the WEDM cam. SegmentTable 6. Roughness 1 Ra in 2 the different 3 segments of 4 the WEDM Average cam. Segment 1 2 3 4 Average RaSegment(µm) 1.0751 1.1322 1.5333 1.1084 1.212Average Ra (μm) 1.075 1.132 1.533 1.108 1.212 Ra (μm) 1.075 1.132 1.533 1.108 1.212 SegmentSegment 3 corresponds3 corresponds to to the the highest highestRa Ravalue. value. The The average averageRa Ravalue value of of the the four four segments segments is is Segment 3 corresponds to the highest Ra value. The average Ra value of the four segments is 1.2121.212µm. μm. This This value value is higheris higher than than that that of of the the milled milled cam cam as as expected, expected, taking taking into into account account that that only only 1.212 μm. This value is higher than that of the milled cam as expected, taking into account that only roughrough WEDM WEDM had had been been performed, performed, and and it isit higheris higher than than the the requirement requirement of 0.4of 0.4µm μ [m29 [29].]. rough WEDM had been performed, and it is higher than the requirement of 0.4 μm [29]. AsAs an an example, example, Figure Figure 14 14shows shows the th measurede measured profile profile in Segmentin Segment 1. 1. As an example, Figure 14 shows the measured profile in Segment 1. m) μ m) μ Height ( Height (

Length (mm) Length (mm)

Figure 14. Graphical representation of the measured proﬁle of the WEDM cam in Segment 1. Figure 14. Graphical representation of the measured profile of the WEDM cam in Segment 1. Figure 14. Graphical representation of the measured profile of the WEDM cam in Segment 1. Figure 14 corresponds to a more irregular profile, with higher valleys than peaks, which correspondFigure to14 the corresponds craters that to are a producedmore irregular by the prelectricalofile, with discharges higher ofvalleys the WEDM than peaks,process. which As a correspond to the craters that are produced by the electrical discharges of the WEDM process. As a

Metals 2020, 10, 1159 14 of 18

Metals 2020Figure, 10, 14 x FOR corresponds PEER REVIEW to a more irregular proﬁle, with higher valleys than peaks, which correspond14 of 18 to the craters that are produced by the electrical discharges of the WEDM process. As a general trend, generalalong the trend, proﬁle along peak the to profile valley peak distance to valley has a distance value of has around a value 6 µ m.of around The maximum 6 μm. The peak maximum to valley peakdistance to valley is close distance to 9 µm. is close to 9 μm.

3.3.3.3. Shape Shape Error Error

3.3.1.3.3.1. Shape Shape Error Error of of the the Milled Milled Cam Cam FigureFigure 1515 presentspresents the shapeshape errorerror ofof the the milled milled cam. cam. It It corresponds corresponds to to the the numerical numerical values values of theof thediﬀ differenceserences between between the the measured measured and and the the theoretical theoretical points points of the of the cam cam proﬁle, profile, for eachfor each angle. angle.

FigureFigure 15. 15. ShapeShape error error of of the the milled milled cam. cam.

AlongAlong the the profile, profile, errors errors are are mainly mainly positive, positive, with with some some small small areas areas where where they they are are negative. negative. ThisThis means means that that the the milled milled piece piece is is slightly slightly larger larger than than the the theoretical model. model. The The highest highest positive positive differencedifference within within segment 1 isis obtainedobtainedat at 90 90°,◦, which which corresponds corresponds to to the the end end of of the the cam cam in in the the area area of theof thebase base circle. circle. At theAt the end end of segment of segment 2, there 2, there is a sharpis a sharp decrease decrease in the in valuethe value of the of shape the shape error error up to up 0 µ tom, 0followed μm, followed by an increaseby an increase up to 15 upµm to at 15 the μm beginning at the beginning of segment of 3.segment Those extreme 3. Those values extreme correspond values correspondto the entry to and the exitentry of and the millingexit of the tool milling during tool the during machining the machining process of process the contour. of the The contour. total errorThe range is 18 µm. It is lower than the shape error that was reported by Norton, around 25 µm for total error range is 18 μm. It is lower than the shape error that was reported by Norton, around± ±25 μpartsm for that parts were that manufactured were manufactured with standard with standa millingrd tools milling [4], andtools similar [4], and to thesimilar value to reported the value by Rothbart for high-speed machines, in the 8 µm range [40]. For camshafts of combustion engines with reported by Rothbart for high-speed machines,± in the ±8 μm range [40]. For camshafts of combustion ground lobes, lower tolerances are obtained of 2.5 µm [14]. engines with ground lobes, lower tolerances are± obtained of ±2.5 μm [14]. 3.3.2. Shape Error of the WEDM Cam 3.3.2. Shape Error of the WEDM Cam Figure 16 presents the shape error of the WEDM cam. Figure 16 presents the shape error of the WEDM cam. In this case, the points belonging to the transition region between segments 3 and 4 have In this case, the points belonging to the transition region between segments 3 and 4 have been been removed, because the area that corresponds to entry and exit of the wire had been removed, because the area that corresponds to entry and exit of the wire had been manually polished, manually polished, and it is not relevant to the WEDM process. However, polishing these points is and it is not relevant to the WEDM process. However, polishing these points is important to important to guarantee a smooth transition between two adjacent segments of the cam profile and thus guarantee a smooth transition between two adjacent segments of the cam profile and thus avoid avoid discontinuities in the follower acceleration law and keep its jerk at normal values [2]. As the discontinuities in the follower acceleration law and keep its jerk at normal values [2]. As the authors authors Angeles and López-Cajún expose in their book [1], a cam-follower mechanism can produce Angeles and López-Cajún expose in their book [1], a cam-follower mechanism can produce the the desired motion program with an accuracy that is only limited by that of the machine used to cut desired motion program with an accuracy that is only limited by that of the machine used to cut the the cam [1]. cam [1]. The obtained piece is oversized because only a rough WEDM operation has been performed. By default, the WEDM machine leaves a material surplus, which would later require to be removed in a finishing stage. This, finishing stage would improve the accuracy and surface finish of the part and would allow for better compliance with the displacement law of the follower and its derivatives

Metals 2020, 10, 1159 15 of 18

(shown in Figure5). Slightly negative values have been observed in the transition area between Sections1 and4. The shape error increases at 90 ◦and 270◦, corresponding to both ends of the cam with a curved shape, at the base circle, and the nose circle, respectively. This suggests that, as it is usual in WEDM machines, especially at high cutting speed [42], the machine has a higher deviation between the actual and programmed trajectory in corners than in straight areas. This is due to the fact that, in corners, where there is a change in the direction of movement of the guides, ﬂexion and vibration produce deformation of the wire, with wire-lag, and the so-called back-wheel eﬀect. The total shape error is 48 µm. It lies within the usual range between 25 µm and 127 µm for

WEDMMetals 2020 parts, 10, xin FOR rough PEERphases REVIEW [ 43], although it is higher than the shape error of the milled cam.15 of 18

Figure 16. Shape error of the WEDM cam.

4. DiscussionThe obtained piece is oversized because only a rough WEDM operation has been performed. By default,In the the present WEDM paper, machine the leaves option a to material manufacture surplu as, tangential which would flank later cam require by means to be of removed two different in a processes,finishing stage. milling, This, and finishing wire electric stage would discharge improve machining the accuracy (WEDM), and issurface considered. finish of Milling the part is and one ofwould the most allow usual for better processes compliance employed with in thethe serialdisplacement production law of of cams. the follower However, and WEDM its derivatives presents several(shownadvantages in Figure 5). that Slightly could allow negative the manufacture values have of been cams observed in serial production:in the transition different area conductive between materialsSection 4 and can beSection machined 1. The regardless shape error of theirincreases hardness, at 90°and more 270°, complex corresponding shapes can to be both obtained, ends andof the it avoidscam with the a use curved of pre-shaped shape, at electrodesthe base circle, that areand necessary the nose forcircle, SEDM respectively. [14]. Alternatively, This suggests WEDM that, could as it beis usual used in in WEDM the first machines, rough operation especially to reduceat high thecutting amount speed of [42], material the machine to be removed has a higher in subsequent deviation operationsbetween the [15 actual]. and programmed trajectory in corners than in straight areas. This is due to the fact that,The surfacein corners, finish where of cams there is is important, a change becausein the direction low roughness of movement reduces of friction the guides, in the flexion cam-roller and contactvibration [27 produce]. Roughness deformation has been of measuredthe wire, with with wire-lag, a contact and profile the meter,so-called as isback-wheel usual in metallic effect. parts withThe curved total surfaces shape error [44]. is Shape 48 μm. error It lies is importantwithin the sinceusual betterrange accuracybetween of25 theμm camand profile127 μm will for favorWEDM the parts compliance in rough with phases the [43], motion-law although of it the is higher follower than and, the consequently, shape error of better the milled operation cam. of the mechanism [2]. Shape error has been measured with a coordinate measuring machine, as is usual in the4. Discussion measurement process of cams and other curved surfaces [24]. AsIn the for present the milled paper, cam, the after option a finish to manufacture operation, an a tangential average Ra flankvalue cam of by 0.269 meansµm hasof two been different found. Thisprocesses, value ismilling, similar and to thatwire obtained electric discharge for miniature machining brass gears (WEDM), [26]. Moreover, is considered. the milled Milling cam is one meets of the requirementmost usual processes of Ra value employed below 0.4 inµ mthe [30 serial]. The production shape error of of cams. the milled However, cam is WEDM 18 µm, whichpresents is lower than the values of 25 µm that were reported by Norton [2], and slightly higher than those that several advantages that± could allow the manufacture of cams in serial production: different are mentioned by Rothbart [40] for high-speed cams, of 8 µm. conductive materials can be machined regardless of their± hardness, more complex shapes can be obtained,Regarding and theit avoids WEDM the cam, use subjected of pre-shaped only to a roughelectrodes operation, that anareaverage necessaryRa valuefor SEDM of 1.212 [14].µm hasAlternatively, been measured. WEDM This could suggests be used that in the first WEDM roug camh operation would require to reduce a further the amount finish of operation material to meetbe removed the requirement in subsequent of 0.4 operationsµm [30]. The [15]. shape error of the WEDM cam is 48 µm. This error value is higherThe than surface that offinish the milledof cams cam, is important, and of the because 21 µm low reported roughness by Hsieh reduces and Linfriction for spatialin the cam-roller cams [21]. However,contact [27]. it isRoughness lower than has the been value measured of 187 µ mwith that a contact was measured profile meter, by Kang as andis usual Han in [20 metallic] for marine parts enginewith curved cams, surfaces and it complies [44]. Shape with error the is requirement important since of 50 betterµm defined accuracy by of Norton the cam [2]. profile Concerning will favor the the compliance with the motion-law of the follower and, consequently, better operation of the mechanism [2]. Shape error has been measured with a coordinate measuring machine, as is usual in the measurement process of cams and other curved surfaces [24]. As for the milled cam, after a finish operation, an average Ra value of 0.269 μm has been found. This value is similar to that obtained for miniature brass gears [26]. Moreover, the milled cam meets the requirement of Ra value below 0.4 μm [30]. The shape error of the milled cam is 18 μm, which is lower than the values of ±25 μm that were reported by Norton [2], and slightly higher than those that are mentioned by Rothbart [40] for high-speed cams, of ±8 μm. Regarding the WEDM cam, subjected only to a rough operation, an average Ra value of 1.212 μm has been measured. This suggests that the WEDM cam would require a further finish operation to meet the requirement of 0.4 μm [30]. The shape error of the WEDM cam is 48 μm. This error value is higher than that of the milled cam, and of the 21 μm reported by Hsieh and Lin for spatial cams

Metals 2020, 10, 1159 16 of 18 compliance of the required motion program of the cam-follower mechanism (meeting the displacement, velocity, acceleration, and jerk), in this case, the milled cam would have better compliance, and the WEDM cam would require a ﬁnish operation in order to guarantee proper compliance. Thus, from the standpoint of quality, WEDM cams obtained with a rough operation allow obtaining the required shape error, although, a ﬁnish operation would be required in order to reduce surface roughness. From an economic point of view, a comparative study between milling, WEDM, and SEDM concluded that WEDM provides material removal rates that, although lower than those for milling, are higher than those for SEDM [15]. In the same study, it was calculated that the roughing costs of EDM are lower than the roughing costs of milling for short batches. This shows the potential of the WEDM technique for the manufacture of complex shapes, such as cams, especially in short production batches.

5. Conclusions In the present paper, the option to obtain tangential ﬂank cams by means of the WEDM process was considered. For this, two diﬀerent cams have been manufactured by means of either milling or WEDM processes. Main conclusions are as follows:

- The milled cam, obtained with subsequent rough and finish contour milling operations, meets both the requirements about surface roughness (average Ra = 0.269 µm < 0.4 µm) and shape error (shape error = 18 µm < 50 µm) for cams. - The WEDM cam, obtained with a rough operation, complies with the shape requirements (shape error = 48 µm < 50 µm), but not with the roughness requirements, with an average Ra value of 1.212. For this reason, a finish operation would be recommended, with a special clamping system to ensure the proper alignment of the part. - WEDM is a potential process for the manufacture of cams, especially in short production batches. Some advantages of this technique are the possibility to obtain complex shapes in different conductive materials, and the fact that, if WEDM is used in rough operations, the quantity of material to be removed in subsequent operations can be reduced.

The present study will help to develop the manufacture of tangential ﬂank cams and other types of parts having curved surfaces by means of WEDM.

Author Contributions: Conceptualization, E.Z.-F. and À.M.-F.; methodology, E.Z.-F. and I.B.-C.; software, À.M.-F.; validation, I.B.-C. and E.Z.-F.; formal analysis, I.B.-C.; investigation, I.B.-C. and E.Z.-F.; data curation, À.M.-F.; writing—original draft preparation, I.B.-C. and E.Z.-F.; writing—review and editing, I.B.-C. and E.Z.-F.; visualization, À.M.-F.; supervision, I.B.-C. and E.Z.-F. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: The authors thank the company Cimatech S.A.C. for their help with the Cimatron software, and CIM-UPC for lending the WEDM machine. They also thank José Antonio Travieso, Alejandro Domínguez-Fernández and Ramón Casado-Lopez for their help with the experimental work. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Angeles, J.; López-Cajún, C.S. Optimization of Cam Mechanisms; Springer: Berlin/Heidelberg, Germany, 1991. 2. Norton, R.L. Cam Design and Manufacturing Handbook; Industrial Press Inc.: New York, NY, USA, 2002. 3. Grant, B.; Soni, A.H. A survey of cam manufacture methods. J. Mech. Des. Trans. ASME 1979, 101, 455–464. [CrossRef] 4. Luis Pérez, C.J.; Luri Irigoyen, R.; Fuertes Bonel, J.P.; León Iriarte, J.; Salcedo Pérez, D.; Puertas Arbizu, I. Experimental and fem analysis of wear behaviour in AA5083 ultraﬁne-grained cams. Metals 2020, 10, 479. [CrossRef] Metals 2020, 10, 1159 17 of 18

5. Purohit, R.; Sagar, R. Fabrication of a cam using metal matrix composites. Int. J. Adv. Manuf. Technol. 2001, 17, 644–648. [CrossRef] 6. Mouralova, K.; Zahradnicek, R.; Benes, L.; Prokes, T.; Hrdy, R.; Fries, J. Study of micro structural material changes after WEDM based on TEM lamella analysis. Metals 2020, 10, 949. [CrossRef] 7. Kumar, S.; Grover, S.; Walia, R.S. Effect of hybrid wire EDM conditions on generation of residual stresses in machining of HCHCr D2 tool steel under ultrasonic vibration. Int. J. Interact. Des. Manuf. 2018, 12, 1119–1137. [CrossRef] 8. Klink, A.; Guo, Y.B.; Klocke, F. Surface integrity evolution of powder metallurgical tool steel by main cut and finishing trim cuts in wire-EDM. Proc. Procedia Eng. 2011, 19, 178–183. [CrossRef] 9. Ahuja, N.; Sharma, N.; Hegab, H.; Khanna, R.; Khan, A.M. Bioactivity measurement of commercially pure titanium processed by micro-electric discharge drilling. Int. J. Adv. Manuf. Technol. 2020, 107, 2797–2805. [CrossRef] 10. Sharma, N.; Singh, G.; Gupta, M.; Hegab, H.; Mia, M. Investigations of surface integrity, bio-activity and performance characteristics during wire-electrical discharge machining of Ti-6Al-7Nb biomedical alloy. Mater. Res. Express 2019, 6, 9. [CrossRef] 11. Mohd Abbas, N.; Solomon, D.G.; Fuad Bahari, M. A review on current research trends in electrical discharge machining (EDM). Int. J. Mach. Tools Manuf. 2007, 47, 1214–1228. [CrossRef] 12. García-Hernández, C.; Gella-Marín, R.M.; Huertas-Talón, J.L.; Efkolidis, N.; Kyratsis, P.WEDM manufacturing method for noncircular gears, using CAD/CAM software. Stroj. Vestn. J. Mech. Eng. 2016, 62, 137–144. [CrossRef] 13. Sanchez, J.A.; Plaza, S.; De Lacalle, L.N.L.; Lamikiz, A. Computer simulation of wire-EDM taper-cutting. Int. J. Comput. Integr. Manuf. 2006, 19, 727–735. [CrossRef] 14. Ho, K.H.; Newman, S.T.; Rahimifard, S.; Allen, R.D. State of the art in wire electrical discharge machining (WEDM). Int. J. Mach. Tools Manuf. 2004, 44, 1247–1259. [CrossRef] 15. Klocke, F.; Zeis, M.; Klink, A.; Veselovac, D. Technological and economical comparison of roughing strategies via milling, sinking-EDM, wire-EDM and ECM for titanium- and nickel-based blisks. CIRP J. Manuf. Sci. Technol. 2013, 6, 198–203. [CrossRef] 16. Hegab, H.A.; Gadallah, M.H.; Esawi, A.K. Modeling and optimization of Electrical Discharge Machining (EDM) using statistical design. Manuf. Rev. 2015, 2, 21. [CrossRef] 17. Sonawane, S.A.; Kulkarni, M.L. Optimization of machining parameters of WEDM for Nimonic-75 alloy using principal component analysis integrated with Taguchi method. J. King Saud Univ. Eng. Sci. 2018, 30, 250–258. [CrossRef] 18. Magabe, R.; Sharma, N.; Gupta, K.; Paulo Davim, J. Modeling and optimization of Wire-EDM parameters for machining of Ni55.8Ti shape memory alloy using hybrid approach of Taguchi and NSGA-II. Int. J. Adv. Manuf. Technol. 2019, 102, 1703–1717. [CrossRef] 19. Chaudhari, R.; Vora, J.J.; Prabu, S.S.M.; Palani, I.A.; Patel, V.K.; Parikh, D.M.; de Lacalle, L.N.L. Multi-response optimization of WEDM process parameters for machining of superelastic nitinol shape-memory alloy using a heat-transfer search algorithm. Materials 2019, 12, 1277. [CrossRef] 20. Kang, J.G.; Han, S. Measurement and evaluation of form deviation error of disk cam with an exclusively built profile-measuring machine. Proc. Robot. Comput. Integr. Manuf. 2010, 26, 686–693. [CrossRef] 21. Hsieh, J.F.; Lin, P.D. Application of homogenous transformation matrix to measurement of cam profiles on coordinate measuring machines. Int. J. Mach. Tools Manuf. 2007, 47, 1593–1606. [CrossRef] 22. Chang, W.T.; Wu, L.I. Computerized tolerance analysis of disk cam mechanisms with a roller follower. Eng. Comput. 2009, 25, 247–260. [CrossRef] 23. Le, C.; Tan, K.N. Reconstruction of planar cam profile function and its follower displacement using B-spline curve based on inverse subdivision method and theory of contact relations—Application to cam mechanism of oscillating follower with roller. Appl. Sci. Eng. Prog. 2011, 4, 51–58. 24. Werner, A. Method for enhanced accuracy in machining curvilinear profiles on wire-cut electrical discharge machines. Precis. Eng. 2016, 44, 75–80. [CrossRef] 25. Gupta, K.; Jain, N.K. Comparative study of wire-EDM and hobbing for manufacturing high-quality miniature gears. Mater. Manuf. Process. 2014, 29, 1470–1476. [CrossRef] 26. Gupta, K.; Jain, N.K. On surface integrity of miniature spur gears manufactured by wire electrical discharge machining. Int. J. Adv. Manuf. Technol. 2014, 72, 1735–1745. [CrossRef] Metals 2020, 10, 1159 18 of 18

27. ul Islam, T.; Khurram, M.; Umar, M.; Mufti, R.A.; Akhtar, K. Experimental and theoretical evaluation of friction in roller follower valve train. Meas. J. Int. Meas. Confed. 2020, 160, 107808. [CrossRef] 28. Torabi, A.; Akbarzadeh, S.; Salimpour, M.; Khonsari, M.M. On the running-in behavior of cam-follower mechanism. Tribol. Int. 2018, 118, 301–313. [CrossRef] 29. Björklund, S. The influence of surface roughness in elliptical contacts. Tribol. Int. 2001, 34, 841–845. [CrossRef] 30. Agulló, J.; Cardona, S. Assessment of Surface Roughness (Valoración de la Rugosidad Superficial); Asociación Nacional de la Máquina—Herramienta: Spain, 1974. 31. Veera Ajay, C.; Pradeep, B.A.; Boopathi, C.; Sanjeev, R.K.; Meganathan, V. Comparison of geometrical accuracy and surface finish of cam profile generated by wire-EDM and CNC milling machine. In Proceedings of the IOP Conference Series: Materials Science and Engineering, Volume 764, International Conference on Advances in Materials Processing and Characterization, Sathyamangalam, India, 10–11 September 2019. 32. Bouquet, J.; Hensgen, L.; Klink, A.; Jacobs, T.; Klocke, F.; Lauwers, B. Fast production of gear prototypes—A comparison of technologies. Proc. Procedia CIRP 2014, 14, 77–82. [CrossRef] 33. Zayas, E.E.; Cardona, S.; Jordi, L. Analysis and synthesis of the displacement function of the follower in constant-breadth cam mechanisms. Mech. Mach. Theory 2009, 44, 1938–1949. [CrossRef] 34. Norton, R.L. Design of machinery: An Introduction to the Synthesis and Analysis of Mechanisms and Machines, 3rd ed.; Mc Graw Hill: New York, NY, USA, 2004; p. 856. 35. Zayas, E.E.; Parra, F.; Jordi, L.; Cardona, S. QtCAM: Free software for the design and the analysis of flat cam-follower mechanisms (QtCAM: Software libre para el diseño y el análisis de mecanismos planos de leva-palpador). In Proceedings of the XXII National Conference of Mechanical Engineering of Spain, Madrid, Spain, 18–21 September 2018; pp. 875–884. 36. DIN 6885-1: 1968. Key, Hub Keyway, Shaft Keyway Details and Dimensions; Deutsches Institut für Normung (DIN): Berlin, Germany, 1968. 37. ISO/R 773:1969. Rectangular or Square Parallel Keys and Their Corre((Sponding Keyways (Dimensions in Millimetres); International Organization for Standardization: Geneva, Switzerland, 1969. 38. ISO 4957:2018. Tool Steels; International Organization for Standardization: Geneva, Switzerland, 2018. 39. ISO 4288:1996. Geometrical Product Specifications (GPS)-Surface Texture: Profile Method-Rules and Procedures for the Assessment of Surface Texture; International Organization for Standardization: Geneva, Switzerland, 1996. 40. Rothbart, H.A.; Klipp, D.L. Cam Design Handbook. J. Mech. Des. 2004, 126, 375. [CrossRef] 41. Buj-Corral, I.; Vivancos-Calvet, J.; Domínguez-Fernández, A. Surface topography in ball-end milling processes as a function of feed per tooth and radial depth of cut. Int. J. Mach. Tools Manuf. 2012, 53, 151–159. [CrossRef] 42. Sanchez, J.A.; Rodil, J.L.; Herrero, A.; de Lacalle, L.N.L.; Lamikiz, A. On the influence of cutting speed limitation on the accuracy of wire-EDM corner-cutting. J. Mater. Process. Technol. 2007, 182, 574–579. [CrossRef] 43. Sommer, C. Non-Traditional Machining Handbook, 2nd ed.; Advance Publishing Inc.: Houston, TX, USA, 2009. 44. Buj-Corral, I.; Ortiz-Marzo, J.A.; Costa-Herrero, L.; Vivancos-Calvet, J.; Luis-Pérez, C. Optimal machining strategy selection in ball-end milling of hardened steels for injection molds. Materials 2019, 16, 860. [CrossRef] [PubMed]

© 2020 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/).