Helical Spring 7

HelicalSpring

By,

UnitTitle

SECTION1

Evaluationof the Design of Compression Spring

Ahelical spring is a major component of a compression spring (Mohamed&amp Adham, 2013). Even though the compressive force is appliedaxially, the helical spring still offers resistance. The spring iscurved at a steady breadth. The springs are used not only to resistforce but also store energy when they are loaded. The type of loadingdepends on upon the application (Aqida, Ghazali &amp Hashim, 2004).The following are the parts of the compression spring.

Storingmechanical energy is the basic function of the spring. At the start,the spring is elastically distorted and then energy is released asspring subsequently recoils (Joirdovi’c, Nedeljkovi’c, Mutrovic’,&amp Zivanic’, 2014). To test for mechanical properties of helicalspring, the following formula is used to show the relationshipbetween functionality and the material properties (Mohamed &ampAdham, 2013). The properties include shear modulus as illustratedbelow.

Dimensional terminology for helical compression spring

Inthe diagram above, Fis the compressive force

Isthe displacement

d:the diameter of the wire

D:the diameter of the coil

G: Theshear modulus of the material of which the spring is constructed.

:Canbe calculated as

Whereis the spring deflection and is the number of effective length of the spring.

Theabove equation can be modified into the following formula used tocalculate shear modulus directly from spring constant. K =, G =.

Aspring with the same length L diameter D and d but various should return the same G as they have the same material (Thomas&amp Scott, 2016).

Springindex: Thisis the ratio of mean diameter (D) to the diameter of the wire (d).

Freelength:Free length of a helical spring is the average length of a spring inan unloaded position (Joirdovi’c,Nedeljkovi’c, Mutrovic’, &amp Zivanic’, 2014).

Typesof spring ends:Spring ends are available in a variety of designs. They include plainend and ground end. Reducing the buckling squarenes during operationrequires bearing surface to be at least 270 degrees (Aqida,Ghazali &amp Hashim, 2004).

Solidheight:It’s the length of a spring when all coils are closed. Thesubstantial height of ground loops is the diameter of the coilmulHIPlied by the wire diameter while, for ungrounded springs, it’sthe number of coils plus one (Mohamed&amp Adham, 2013).

Types of ends for helical compression spring

Springrate:For a helical spring, the term spring rate is used to denote thechange in load per unit deflection. It is expressed as shown below:

K= = The equation is only valid when the angle of pitch is less. . than

Torsionalstress:This is the stress a helical spring is exposed to during torsion.

S=

Whencompression springs are arranged in parallel, the composite rate isthe sum of the rate for individual springs. But when compressionsspring are in series, the rate is calculated from the followingformula (Joirdovi’c et. al., 2014).

K=

Bucklingof Compression Spring

Compressionsprings can be buckle when length of the spring is four times greaterthan its diameter (Aqida, Ghazali &amp Hashim, 2004).

LaserMelting Process

Theselective laser melting is a fabrication technique used to producegeometrically complex components. Metal powders of very fine sizegrains are melted using a high power laser beam. Then a 3D part isbuilt from the melted metal based on a CAD model (Thomas&amp Scott, 2016). Components produced by the process of lasermelting are much different in strength, durability and performancethan components generated in any other conventional manner. Theexhibit a higher or at least comparable mechanical properties thancast parts (Mohamed &amp Adham, 2013).

Asa means of facilitating the improvement of the performance of thecomponents of a helical spring, the EOSINT 280 technology is adapted.This system has approximately 400 watts solid state laser and argonprotected atmosphere. The blasting after manufacturing ensures thatthe surface is complete (Joirdovi’c et. al., 2014). All parts aremade of Ti-6Al-4V alloy powder, commercially named as “EOS TitaniumTi64”. The powder is optimized for processing.

Thepowder is contained in a dispenser bed (Ganesh Babu, &amp Srithar,2010). A prescribed dose of the powder is elevated as illustratedbelow. A thin layer of powder is then spread over the surface of thebuild plate by a recoater mechanism which consists of a hardscrapper(Childs, 2004).

TheProcess of Laser Melting

Atypical laser melting process is shown above.

Finelygrounded metal powder is contained in a dispenser bed. The focusedenergy of the beam is used to fuse the powder together on alayer-wise basis. This process is called the select layer melting.The powder, in this case, is metal, and the process is direct metallaser melting. The thickness of powder layer is typical between 10μmand 10μm (Joirdovi’c et. al., 2014). To prevent oxidation of themetal during the fusion process, the procedure is conducted in anatmosphere with relatively lower oxygen content as in the case ofnitrogen or argon atmosphere (Ganesh Babu, &amp Srithar, 2010). Toobtain a finished part of a layer-by-layer fashion, the process isrepeated to a buildup. Several parameters are put in place to ensurethe final product is of the desired quality (Ganesh Babu, &ampSrithar, 2010). They include building environment parameters, powderproperties and recoating parameter and laser and scanning parameters.The laser serves as an energy source in the melting process(Joirdovi’c et. al., 2014).

Duringmelting, the diameter of the molten zone is always larger than thelaser diameter. To compensate for the dimensional error, it isnecessary for the laser beam to be shifted by half the width. This isdone from the contour to the inside to ensure that the shape of thelatter part corresponds to the original CAD data (Childs, 2004). Thecorrection is referred to as beam offset. The value of offset beamdetermines the quality of the melt. The particle of the irradiatedregion may be over-melted or not melted. The laser beam istransmitted from line to line severally to ensure that the meltingprocess can be complete (Ganesh Babu, &amp Srithar, 2010). Thehatching distance which is about a quarter of the laser beam is thedistance between the lines. The final parameter which can lead to themelting process interruption is the layer thickness of the powder.Extreme high values between the single layers can be obtained whenthe melting depth is not strong enough. The mechanical tensiongenerated through this layer can lead to detachment (Ganesh Babu, &ampSrithar, 2010). Tearing off of the structure can also happen duringthe recoating process. The tearing-off maybe due to the selection ofa value which is too small (Mohamed &amp Adham, 2013).

Themost complex metal geometries can be built layer by layer usingdirect metal laser melting. The lengths of such layers can be asthin as two microns. The process is done directly from 3D CAD data(Thomas&amp Scott, 2016).In order to be able to produce a fine, homogeneous, fully densestructure, the metal powder is melted entirely. Due to the uniquegeometry of freedom in design, direct metal laser melting can be usedto build incredibly complex geometry of metal components. Thecomplexity of design proves difficult with the conventional machiningmethods

(Ruipeng,Lei, Jie, Rui &amp Bernie, 2015). The products of direct lasermelting have excellent mechanical properties. They are equivalent towrought materials with high detail resolution and exceptional surfacequality.

Toenhance the functionality of components, AlSi10Mg, a typical castingalloy with excellent casting properties is utilized (Thomas&amp Scott, 2016).It is preferable in casting thin walls, parts, and complex geometry.It offers superior strength, hardness, and dynamic propertiesqualities that are primary requirements of compression springsubjected to high loads (Robert,2011).The need for parts which can withstand high-temperature conditionsand are little weight in the motor vehicle makes Parts in EOSaluminum ideal.

Difficultyin the Process of Direct Laser Melting

Difficultyin processing is the primary factor affecting the use of aluminum indirect laser melting process. Aluminum is hard to cast or weld.Alloys of the metal are relatively harder to weld. Even with laserthe casting in most cases, result in high degree porosity. This isnot an exception to direct laser melting. The mechanical propertiesof the final products are greatly affected by porosity (Thomas&amp Scott, 2016).

Porosityis mainly caused by two main factors:

1.Surface Oxidation

2.Moisture pick-up

SurfaceOxidation

Aluminumrapidly forms an oxide layer when exposed to air. The result of thisbehavior proves that aluminum powder favors porosity (Mohamed&amp Adham, 2013).The defects noticed in laser melted Aluminum parts are the same asthose found in conventional casting because oxidation is a surfacephenomenon. This effect is particularly encouraged in Aluminum LaserMelting because the powder feedstock has a higher surface-to-volumeratio (Robert,2011).

Theeffect of oxidation in laser melting include the following:

  1. Diffusion is hindered by the oxidation on the surface of the particles. This refrains sintering between the un-melted particles. As a result of this region of weakness develop in the component.

  2. Oxide films can form at relatively low oxygen concentration due to their adherent.

  3. When the molten metal is stirred into, oxide formed entrap gasses. These gasses end up in pores within the material.

  4. Oxide films form between the laser hatches at every layer of the aluminum parts. Formation of oxide films on both solid and liquid metal surface occurs in the process. Pores are formed where the two oxide films meet (Ruipeng et al., 2015).

MoisturePick-Up

Poreformation in molten aluminum is directly caused by moisture pick-up.Molten liquid dissolves hydrogen, but as it solidifies, hydrogentends to degas as the solidification front passes. The presence ofany hydrogen-bearing species in the powder leads to porosity as theywill be released in the molten pool.

SECTION2

Powdermetallurgy (PM), also known as the Ti–6Al–4V alloy is the productof hot isostatic pressing (HIPing). The HIPing process uses the gasatomized powder. When investigating the thermal response of thealloy, various conditions of hot isostatic pressing as well as heattreatment are extensively utilized (Mohamed &amp Adham, 2013). Thereare significant changes in the microstructure of the powder underdifferent heat treatments. These changes occur with no variation intensile features. Porosity is induced by the effect of temperatureafter annealing at 930oc. However, the hot isostatic pressing processhas no influence on the toughness of the structure and tensileimpact. In addition to this, it has no effect on the components ofpowder compact (Mohamed &amp Adham, 2013).

Titaniumand its alloys exhibit unique combination of chemical and mechanicalproperties. The alloys have outstanding corrosion resistance (Robert,2011). To fabricate titanium alloys, extensive costs are incurred.This is considered to be the case in instances where the processinvolves parts with bulky proportions or compound geometry.

Materialsand methods

Thefirst material taken into consideration is powder. The sphericalTi–6Al–4V powder is formed by a process known as the electrodeinduction melting gas atomization. The completion of the procedure ismassively influenced by the utilization of a crucible-free system(EIGA 50-500). This is necessitated by the chemical features of EIGATi–6Al–4V that are regarded to have an extensively diminishedinterstitial.

HIPingand Heat Treatment

Accordingto this process, particles of the powder are immersed into low carbonsteel containers. Thereafter, the components are hot degassed.additional HIPing experiments are conducted to establish the natureof associations by the HIPing parameters. To be specific, this phasefocuses on the final stage of powder densification (Collins, Busby, &ampStaab, 2010). The densities of the as-HIPed samples are measuredusing the Archimedes approach. When conducting an investigation onthe impacts of heat treatments on various variables such asmicrostructure and mechanical properties of PM Ti–6Al–4V alloy,numerous treatments are taken into consideration. To complete theprocess, diverse cooling rates are utilized after annealing at 930°C/1 h (Collins, Busby, &amp Staab, 2010).Some of the techniquesused in cooling include furnace cooling and water quenching.

Microstructure

Opticalmicroscopy and electron microscopy are some of the techniques used toobserve the microstructure of as-HIPed and heat-treated PM Ti–6Al–4Valloys (Collins, Busby, &amp Staab, 2010). The samples to be testedare taken through a series of procedures that involve grinding,polishing and chemical etching. The latter process is accomplished byusing Kroll`s reagent (2%HF+4%HNO3).The SEM framework is used toexamine any fracture surfaces of tensile and fatigue specimens in thestudy. The process also integrates the use VersaXRM-500 X-raymicro-computerized tomography to evaluate the porosity of as-HIPedand heat-treated powder compact (Pinjarla &amp Lakshmana, 2012).

Testingof Chemical Properties

Thisprocess utilizes an INSTRON 5582 testing machine to conduct tests onroom temperature tensile. The machine is set to have an initialstrain rate of 2×10−4 s−1 and 2×10−3 s−1 after yielding.This is in agreement with ASTM E8-08. The study utilized betweenthree and six tensile specimens to facilitation the repetition ofresults for each underlying condition. On the other hand, SAN-ZBC2452-C testing machine is used to conduct the U-notched Charpyimpact tests whereas an Amsler axial resonance pulser with afrequency of 110–116 Hz and a sinusoidal waveform is used toconduct the high cycle fatigue (HCF) tests (R=0.1). Finally, anMTS810 testing machine with a frequency of 10 Hz and a sinusoidalwaveform with an R-ratio of 0.1 is used to pre-pack the fracturetoughness testing specimens (Ruipenget al., 2015).

Effectsof HIPning Parameters

Thefigure below shows the relative concentration of particle compactnessa series of HIPing cycles. In this case, the value increases from theinitial density of 68% to 99% (Pinjarla &amp Lakshmana, 2012). Fromthe graph, it can be concluded that both HIPing temperature andpressure have a massive influence on the powder consolidation.However, there is an inverse relationship between relative densityand HIPing temperature once the latter reaches approximately 900 °C.As the temperature of HIPing increases above 900 °C, the relativedensity will substantially decrease. (Collins, Busby, &amp Staab,2010). In this regard, in instances where the relative density ismore than 99.37% or 99.54%, the powder compacts with a relativedensity of more than 99.5% is regarded as the full mass (Ruipenget al., 2015).The HIPing pressure have to be raised to more than 110 Mpa whenconsidering the attaining a powder compact with full density(Collins, Busby, &amp Staab, 2010). The figure below shows therelationship between relative density, pressure and temperature. AtHIPed of between 920–940 °C or over 120 Mpa, the powder compactattains the highest level of relative density.

MicrostructureEvolution

Thefigure below shows the evolution of the microstructure during heattreatment. The figure indicates the microstructure of a HIPedTi–6Al–4V alloy when the temperature is at 940 °C/120 Mpa for 3hours. equiaxed and lath-like α phase and acicular β phase are themajor components of a fully dense microstructure. The α phase isdark while the β phase is light (Ruipenget al., 2015).The volume percent of β phase is about 5% (Goetsch, 2005). There arethree major steps of powder densification during HIPing process. Theyinclude plastic yielding, power-law creep, and diffusion bonding. Asa result of severe distortion and recyclization of powder during theprocess, the equiaxed microstructure is produced at the powderboundaries of massive particles (Aqida, Ghazali &amp Hashim, 2004).

Theplot below shows the relationship between the ultimate tensilestrength and elongation of Ti–6Al–4V. The marked area in thefigure is known as “safe area”. It is at this point that thespecification ASTM B348 for all wrought material is met by thetensile data (Ruipenget al., 2015).Based on the outcomes of the procedure, it can be stated that thetensile properties of PM Ti–6Al–4V alloy are relatively better incomparison to those of as-cast alloy. However, between 920 Mpa and1080 Mpa, the tensile strength of PM alloy is likely to change.During this period, the elongation falls between 14% and 20%. Therange is significantly higher for the tensile data of wroughtmaterials (Goetsch, 2005). It is for this reason that PM Ti–6Al–4Valloy has a lower heat response when compared to that of wroughtalloy during heat treatments. As such, it is most appropriate for theproduction of PM Ti–6Al–4V components.

Impactfracture and HCF Properties

Suchtests are conducted for the primary reason of determining thefeatures of powder compacts. To achieve the desirable outcomes, theprocess is conducted under room temperature (Collins, Busby, &ampStaab, 2010). The fatigue strength and aging of Ti-6Al- 4v is biggerthan that of as-HIPed at 510mpa after annealed at 930oCfor one hour (Pinjarla &amp Lakshmana, 2012). The fatigue ratiorefers to the relationship between the fatigue strength and ultimatetensile strength of both as-HIPed and heat-treated samples are at0.5–0.6 level (Collins, Busby, &amp Staab, 2010). The values ofthe impact and toughness are relatively lesser than those of as-HIPedsample. However, there are similarities between the fatigue strengthof both as-HIPed and heat-treated samples and as-forged materialswith bimodal microstructure. On the other hand, even in theexistence of the heat-treated condition, the fracture toughnessas-forged materials is lower than that of powder compact (Pinjarla &ampLakshmana, 2012).

Thefigure below is a plot of S–N curves of as-HIPed and heat-treatedpowder compact. They indicate that the fatigue strength of powdercompact increases at Nf&gt107 cycles after heat treatment (BhandarI,2010). The slope of the curve is also steeper in the as-HIPedsamples in comparison to the heat-treated samples (Goetsch, 2005).Moreover, the HIPed sample demonstrates better fatigue propertieswithin the shorter fatigue lives regime from the S–N curve.

Discussion

Themechanical properties of powder metal titanium alloy aresignificantly affected by powder compact porosity. X- ray’stomography is used to indicate the occurrence of porosity and HIPfollowing heat treatment (Childs, 2004). Research indicates that theHIP will depend on a series of factors that include the grain size,time as well as the amount of entrapped inner gas and annealingtemperature. HIP in powder compact can also be caused by the leaks ofpowder container (BhandarI, 2010). The result is composed of twocomponents namely the effect of heat treatment on mechanicalproperties, and the optimism of HIPing parameters (Collins, Busby, &ampStaab, 2010). Various guidelines are integrated into the process tocontrol porosity (Collins, Busby, &amp Staab, 2010).

Optimizingof HIPing Parameters

TheHIPing temperature are excepted to be in the range of 900–940 °C,with the pressure over 100 Mpa. However, changes to either parametercan lead to a fall in the relative density. For example, powdercompact with a relative density &gt99.54% can be processed with aHIPing pressure of over 100 Mpa as well as temperatures between 880°Cand 900°C (Pinjarla &amp Lakshmana, 2012). A negative relationshipexist between the flow stress of powder Ti–6Al–4V andtemperature. This is caused by the shielding effects of the capsule.As such, when the flow of stress decreases, there will be acorresponding increase in HIPing temperature (Collins, Busby, &ampStaab, 2010). As a result, increased pressure on powders leads to ahigher relative density (BhandarI, 2010). The trend is howeverdifferent once the HIPing temperature increased from 900 °C. In thiscase, the relative density will decline (Jindal, 2010). The declinecan be attributed to the existence of powder particles with gasbubbles that arise as a result of gas atomization.

HeatingEffect on Mechanical Properties on AlSi10Mg

Theelimination of residual porosity leads to an increase in UTS. Thetensile features are in close proximity to the fully solid particlesat a relative density of over 99%. The porosity of as-HIPed andheat-treated, that is, annealed at a temperature of 930 °C per hour,AC+aging samples is 0.0022% and 0.0199% respectively(Jindal, 2010).

Porosityof powder compact increases more than ten-fold as a result of heattreatment. However, the porosity is still considered to be relativelylow (Kampen et al., 2012). When conducting the tensile test, thestress concentration around pores can always be ignored. Attemperatures of approximately 930 °C and above, the microstructureof powder compact moves towards the bi-model microstructure(BhandarI, 2010). This is similar to the cases of the UTS andelongation of powder compact as a result of heat treatments (Kampen,Thijis, Humbeeck &amp Kruth, 2012). The microstructure of powdercompact undergoes a massive transformation after various heattreatments.

Thesignificant change results to no apparent variations in tensilefeatures. This is particularly suitable for the production of powdercomponents with unit components (Jindal, 2010). The as-HIPedlath-like structure could globularize without deformation orpost-deformation heat treatments in comparison to a typical cast andwrought Ti–6Al–4V alloy (Kampen, Thijis, Humbeeck &amp Kruth,2012). HIPed is therefore, microstructure and considered asnon-equilibrium. This states the reason as to why the microstructurehas no apparent impact on tensile components even in the existence ofdifferent heat-treated states (BhandarI, 2010).

Aset of parameters that culminate in a maximal density is utilized inthe promotion of optimum mechanical properties of AlSi10Mg parts.Such parts are produced by SLM. A laser power of 200W is used in theoptimization of such parameters (Krishna, 2010). It also involves ascan speed of 1400mm/s. In addition to this, there is a spacing of105μm amid the scan tracks. The result is a relative density of atleast 98.5% that can be elevated to 99.8% by re-melting every layerwith similar parameters (Jindal, 2010).In the latter procedure,alternate directions of over 90° are adopted (Pinjarla &ampLakshmana, 2012).

MechanicalProperties

Tensiletests are conducted on separate directions to obtain mechanicalproperties (Krishna, 2010). In most instances the outcome is theaverage value for 2 or 3 specimens with 95% confidence intervals(Jindal, 2010). In this regard, high pressure die casting is utilizedin achieving the optimum properties amongst the casting processes.One stress-strain curve is considered for effective visualization(Kampen et al., 2012).

MechanicalProperties of SLM Built Parts and Cast + Aged Parts

X -+

E

GPa

UTS

MPa

%

HV

Xy direction

Z direction

57+- 6

307 +- 3

389 +- 7

5.78 +- 0.9

3.53 +- 0.92

120

Conventional cast and aged

68

312 – 334

2.6 – 3.8

73

High pressure die casting F

High pressure die casting T6

68

68

312 -381

312 – 334

3-5

3-5

95 – 105

130 -133

Thefirst observation shows that SLM AlSi10Mg parts have mechanicalproperties that are relatively significant in comparison to the castAlSi10Mg material (Jindal, 2010). The hardness of the high pressuredie cast (HPDC) AlSi10Mg in the as-cast condition is similarly highas in the case of HPDC AlSi10Mg in the old condition (Pinjarla &ampLakshmana, 2012). Moreover, the strength of the tensile in the builtAlSi10Mg SLM parts is also significantly greater when compared tothose of the HPDC taking into account both conditions. When analyzingthe Z-direction, the as-built AlSi10Mg parts is deemed to becomparable to the HPDC parts. On the other hand, the elongation forparts built in XY – direction is significantly higher (by 2%)(Jindal, 2010).

X +- s

Impact energy

J

XY direction

Z direction

3.94+- 0.5

3.69 +- 0.48

As- cast

2.5 – 3.0

Conclusion

Mechanicalproperties of Hardness, UTS, and elongation are present in SLMAlSi10Mg. Their impact on energy is relatively higher comparable tothe cast AlSi10Mg material (Jindal, 2010). This elevated effect canbe attributed to the very fine microstructure in addition toimpressive allocation of the Si phase (Aqida, Ghazali &amp Hashim,2004).

Asample of SLMs shows some anisotropy in elongation at break as aresult of the optimal density scanning strategy (Krishna, 2010).Borderline porosity is formed by the Z-oriented tensile samples dueto the scanning approach. The Z-oriented tensile sections become moresensitive to crack initiation in comparison to XY-oriented tensilespecimens due to the pores (Ganesh Babu, &amp Srithar, 2010).

Reference

Topof FormTop of Form

Aqida,N. Ghazali M, and Hashim, J.(2004). Effects of Porosity on mechanicalproperties of metal matrix composite. Jurnal teknology, 40 (A) :17-32

BhandarI,V. B. (2010). Designof machine elements.New Delhi, Tata McGraw-Hill.

Childs,P. R. N. (2004). Mechanicaldesign.Oxford, Elsevier Butterworth-Heinemann.http://public.eblib.com/choice/publicfullrecord.aspx?p=288872.

Collins,J. A., Busby, H. R., &amp Staab, G. H. (2010). Mechanicaldesign of machine elements and machines: a failure preventionperspective.Hoboken, NJ, Wiley.

GaneshBabu, K., &amp Srithar, K. (2010). Designof machine elements.New Delhi, Tata McGraw Hill EducationBottomof Form

Goetsch,D. L. (2005). Technicaldrawing.Clifton Park, NY, Thomson Delmar Learning.

Jindal,U. C. (2010). Machinedesign.Noida, India, Pearson. http://www.myilibrary.com?id=265251

Joirdovi’c,B. Nedeljkovi’c, N. Mutrovic’, J. and Zivanic’. (2014). AEffect of heat treatment on structural changes in metastableAlSi10MgAlloy. B. Jordovic’ et al. / JMM 50 (2) B 133-137

Kampen,K. Thijis, L. Humbeeck, V and Kruth, P. (2012). Mechanical propertiesof AlSi1Mg produces by selective Laser Melting. Physics Procedia 39:439 – 446

KrishnaRao, T. (2010). Designof machine elements. Vol. ll Vol. ll.New Delhi, I.K.

Mohamed,A. and Adham, E. (2013).Ti-6Al-4V Helical Spring Manufacturing viaSLM: Effect of Geometry on Shear modulus. Proceedings of theInternational Multi-conference of engineers and computer scientists’2013 volume II, IMECS 2013, March 13 – 15, 2013, Hong Kong.

Pinjarla,P. and Lakshmana, T. (2012).Design and Analysis of a shock Absorber.International journal of Research and Engineering and Technologyvolume I issue 4 Dec 2012. 578 – 591.

Robert,S.(2011). EOS GmbH Electro optical systems. Aluminum AlSi10Mg. AD MS /05.

Ruipeng,G. Lei, X. Jie W. Rui, Y and Bernie, Y. (2015). Microstructuralevolution and mechanical properties of Powder metallurgy Ti-6Al-4Valloy based on heat response. Material science and engineering A639:327 – 334

Thomas,G. and Scott, G. (2016). In-process laser sensing in selective lasermelting additive manufacturing. Spears and Gold integrating materialsand manufacturing innovation (5) 2 1-25