Originalscientificpaper

Croat. j. for. eng. 37(2016)2 219

Evaluation of a New Energy Recycling

Hydraulic Lift Cylinder for Forwarders

Jussi Manner, Ola Lindroos, Hans Arvidsson, Tomas Nordfjell

Abstract

In mechanized forestry, much of the work is conducted by use of cranes, and recovering po-tential energy is a possible method to reduce energy consumption when using cranes for lift work. The objective of this study was to evaluate the capacity of a new »Energy-efficient hy-draulic lift cylinder« (EHLC), which has a secondary cylinder built into its piston rod, to store potential energy from lowering the boom in the form of pressurized hydraulic oil in an accu-mulator and using the stored energy in the next boom lift. The EHLC was mounted on a forwarder, and manipulated to enable its use also as a standard cylinder. We then compared the EHLC and a standard cylinder in terms of function and energy consumption during re-petitive boom lifts and lowerings. With the tested settings the EHLC saved up to approxi-mately 9.4% of the energy consumed during the first part of boom lifts and up to 3.2% of the total lift energy. With possible further adjustments, such as optimization of the accumulator size, enlargement of the assisting cylinder diameter, and enhancement of the accumulator pressurization, but most importantly reduction in internal leakage, the current EHLC could have commercial potential.

Keywords: weight-balancing, fluid dynamics, fluid mechanics, timber loader, mobile hydrau-lic lift devices, Boyle’s law, counterweight

1. IntroductionIncreases in energy costs and environmental con-

cernshaveintensifiedeffortstoimproveenergyeffi-ciencyrecently,bothgenerallyandspecificallyinen-gineeringresearch(e.g.EuropeanUnion2014,UnitedNations2014).Notably,severalrecentstudieshavead-dressedpossiblemethodstoimprovetheproductivityofcranesusedinharvesters,plantingmachinesandforwarders(Lindroosetal.2008,Jundénetal.2013,LaineandRantala2013,Erssonetal.2014,OrtizMo-ralesetal.2014).Cranesareusedprimarily for liftwork, inmanymobileandstationaryapplications,which involves relocating objects in such a manner thattheirpotentialenergychanges.Consequently,re-coveringpotentialenergycanprobablybeusedtore-ducetheenergyrequiredforthework(e.g.LiangandVirvalo2001a,SunandVirvalo2003,Rydberg2005,SunandVirvalo2005,VirvaloandSun2005,Linetal.2010,LinandWang2012,Minavetal.2012,Noréusetal.2013,Wangetal.2013).Forwardercranesarede-signedtoprovidelargeliftsandheights,partlyatthe

costofslowhorizontalmovements(Malmberg1981,GerasimovandSiounev1998,2000,VirvaloandSun2005).Knuckleboomcranesarenormallyusedonfor-warders, and consist of a system of hydraulic cylinders andmechanicallevers,i.e.,aswivellingcranepillar,pivotingmidandouterbooms,andanextensionboom(GerasimovandSiounev2000).Whileboomisoftenused as a synonym for crane, that usage is avoided in thispapertoavoidconfusionbetweenthesystemanditscomponents(cf.Lindroosetal.2008).Theactionofaverticalliftismainlyexecutedbyan

angularchangeinthejointbetweenthecranepillarandmid-boom,whichcausestheouterandextensionboomstorise.However,inthispapertheouterandextensionboomsaretreatedasrigidpartsofthemid-boom.Thus,boomisusedhereafterasacollectivetermforthemid-,outerandextensionbooms,withtheunderstandingthatliftsareexecutedviaactionatthejointbetweenthecranepillarandmid-boom.

During forwarder work logs are collected in the forestandcarriedtoroadsidelandings.Typically,for-

220 Croat. j. for. eng. 37(2016)2

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

warderproductivitiesrangefrom11to25m3perpro-ductive machine hour, and are negatively correlated withtransportationdistance,whilepositivelycorre-latedwithsizesofbothlogsandloads(ErikssonandLindroos2014).Logsareliftedwhileloadingandun-loadingtheforwarder’sbunk.Thetotaltimerequiredto forward a load (for which loading and unloading collectivelyaccountforabout60%)istypically45min-utes(Manneretal.2016).Typically,loadingandun-loadingwillrequireapproximately30and20liftswithafullgrapple,respectively(Manneretal.2013,Man-neretal.2016).Theboomonastandardforwardercraneislifted

withasingleactingcylinder.Pressurizedoilisdirect-ed into the cylinder, creating a force that causes the pistonrodtolifttheboom.Tolowertheboom,thepressurizedoilfromthecylinderisreleasedintothenon-pressurizedreservoir,andnoenergyisrecovered.However,thereareseveralpossiblemethodstosaveenergy,forexample,throughweight-balancing,acom-monprincipleforrecoveringpotentialenergyduringload lowering for cranes and elevators. In a weight-balancingsystem,someofthepotentialenergyisre-covered and stored during load lowering, and then usedtoassistthenextloadlift.Thepotentialenergyrecoveryprocesscreatesabrakingforcethatreducestheloadloweringspeed.Thus,theweight-balancingis a trade-offbetween the additional lift force andbraking force. In a fully balanced system, ignoring en-ergylossesthroughfriction,itistheoreticallypossibletorecoveralmostallofthepotentialenergy,allowingaloadtobeliftedandloweredwithminorenergyin-put.Other examplesof theweight-balancing tech-niqueformobiledevicesaretheuseofcounterweightsorcoilsprings(e.g.GawlikandMichałowski2008,Deepak2012,Linetal.2013).Thefewavailableenergyrecoveringliftapplica-

tionsforforestmachinecranesaretypicallybasedontheuseofahydro-pneumaticaccumulatortank(ac-cumulator),acommonweight-balancingtechnique(e.g. Liang and Virvalo 2001a, Liang and Virvalo2001b,SunandVirvalo2003,SunandVirvalo2005,VirvaloandSun2005).Anaccumulatorconsistsofavesselandbladderwhichseparatesaninertgas(e.g.nitrogen)fromhydraulicoil.Theflowofpressurizedoil into the accumulator charges the accumulator as thesealedinertgascompressesaccordingtoBoyle’slawand,similarly,theflowofoiloutfromtheaccu-mulator discharges the accumulator and releases the storedenergy(duringboomliftandboomlowering,respectively,incranework).Fastcharginganddis-charging are some of the advantages of accumulators forenergystorage(HuiandJunqing2010,Minavetal.

2012,VandeVen2013).Moreover,accumulatorsre-ducepressure spikes in thehydraulic system (e.g.Malmberg1981,Ingvast1989,Kimetal.2013,VandeVen2013).Experimentsshowthat21–59%ofpoten-tial, kinetic or rotational energy can be recovered by usingaccumulators(Zhang2011,HoandAhn2012).However, an energy recovery system based on an ac-cumulator also has limitations. Notably, in most sys-temstheaccumulatorpressuremustexceedthepres-sure in the hydraulic circuit to enable reuse of the storedenergy(cf.Einola2013),buttheaccumulatorpressuremightdecreasebelowthispressuredue,forexample, to internal leakage somewhere,which islikelytooccurinallhydraulicsystems(Manring2005).Thus,leakagecompensationtore-pressurizetheac-cumulatorisrequired.Anotherlimitationisthelowenergystoragecapacityinrelationtotheirsize(e.g.VandeVen2013).

2. Materials and methods

2.1 The energy-efficient hydraulic lift cylinder (EHLC)Technicalprinciplesandclaims foraflawlessly

functioningEHLC(Fig.1)mountedonaforwarder

Fig. 1 Thordab AB’s patented »Energy-efficient hydraulic lift cylinder« (EHLC) with a pressure accumulator tank (accumulator), a movably arranged secondary piston (a) that divides the cylinder system into primary (b) and secondary cylinders (c). The secondary cylinder is built inside the primary piston rod (d) and connected to the accumulator. Load cell1 (p1) measured the pressure in the standard cylinder and primary cylinder of the EHLC. Load cell2 (p2) measured the secondary cylinder pressure and accumulator pressure in the EHLC

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 221

crane are described below based on the manufactur-er’sinformation(cf.WIPOPatentWO/2011/075034).Inthestartingposition,thesecondarypiston(a)is

insidetheprimarypistonrod(d),basicallyformingastandardcylinder(Fig.2:panel1).Immediatelyafterthemachinehasstartedandduringthefirstboomliftthepressureinthehydrauliccircuitoftheaccumulatorwill be the same as in the circuit of the standard cylin-der. However, during boom lowering the accumulator isfurtherpressurized,thecheck-valvepreventsanout-flowandthestandardcylinderturnsintoprimary(b)and secondary (c)cylinders(Fig.1andFig.2:panel2)

duetothepressuredifferences.Asaresult,thesecond-arypistonwillmoveinthedirectionofthelowerpres-sure at any given time, since both of the secondary pistonendshavethesamearea(Figs.1–2).Hence,thefollowingliftworkwillbeconductedwithassistancefromtheaccumulator,providedthatthepressureinthe accumulator and secondary cylinder (p2)ishigherthanintheprimarycylinder(p1).Ifnot,thecylinderwill function as a conventional cylinder.

Given that p2 is higher than p1,theproductofp2 and thesecondarypistonarea(A2)createasecondarycyl-inder force (F2)(seedetailsinFigs.1–2).Asthesecond-arypistonthruststheprimarycylinderheadcontinu-ously with a F2 that depends on p2, it creates an assistingforceduringboomliftsandabrakingforceduringboomlowering.Duringboomlifting,theac-cumulator discharges and decreases p2, while a boom lowering charges the accumulator and p2 increases (Fig.3).Thus,F2decreasestheEHLC's needforexter-nalenergyinput.Somelossesinp2 are likely to occur due, for in-

stance,tooilleakagefromthesecondarytothepri-mary cylinder. If p1exceedsp2thecheck-valveopensand the accumulator will be charged from the hydrau-liccircuitsystem(Fig.1).Therefore,ahigherpressuremust be maintained in the accumulator and secondary cylinderthanintheprimarycylindertogetanassist-ing force during a boom lift. Occasional pressurespikesinthehydrauliccircuitpassthecheck-valveandloadtheaccumulator,to25–30MPaaccordingtothepatent.Thus,occasionalpressurespikesthatalwaysappearinordinaryhydraulicsystems(Manring2005),andhenceoccasionalshort-termcheck-valveopeningsareessentialcontributionstotheEHLC'sfunctionalitybecause they maintain a higher p2 than p1 during boom lifts,andalsoprovideleakagecompensation(Fig.3).In addition, the secondary cylinder work (W2)increas-es with increasing p2, thereby decreasing the need for anexternalenergyinput.Thus,check-valveopeningscausedbyoccasionalpressurespikesshouldnotbeconfoundedwith»malfunctions«of theEHLC, i.e.regularlyopencheck-valve.Thestandardandsecondarypistondiametersofthe

studiedEHLCcylinderwere124.5mmand32.0mm,respectively(RogerGustavsson,ThordabAB).Thisprovidesastandardpistonarea(Asp)of12,174mm2, A2 of804mm2,andaprimarypistonarea(A1)of11,370mm2 (Fig.2).Theaccumulatorvolumewas1.0litre(RogerGustavsson,ThordabAB).Basedonthesespecifications,thetheoreticalen-

ergysavingis6.6–6.9%,giventhatA2/Asp≈0.066,andp2isintherangeof1.00to1.05timesp1 during the entireboomlift.

Fig. 2 Panel 1: a standard cylinder. Panel 2: EHLC. In both panels: a is a movably arranged secondary piston, b is the primary cylinder, c is the secondary cylinder and d is the primary piston rod. The standard piston area (Asp) is the sum of primary piston area (A1) and secondary piston area (A2)

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

222 Croat. j. for. eng. 37(2016)2

2.2 Test and measure procedureAnEHLCwasmountedonastandardGremo1050

Fforwarder(enginepower120kW,hydraulicworkingpressure 23.5MPa) equippedwith aCranabFC80crane(boommass890kg),astandardrotator(56kg),andaCranabCR280grapple(200kg).Theexperimenttookplacebetweenthe11thand13thofJuly2011inUmeå,NorthernSweden.Themidandextendingboomsweremechani-

callyblockedinpositionsthatgaveaconstantcranereachof5.7m.Duringtheexperiment,theboomwasliftedandloweredbyactuatingtheliftcylinderdi-rectional control valve using a joystick. Three set-tingsforthevalveresponsetothejoystickactuationswereused,resultingindifferentdirectionalcontrolvalveopeningspeedsforthesamejoystickmove-ment and hence different acceleration and boomspeeds,designatedslow,mediumandfast(Table1).Aboomliftwith»slow«valvesettingwasfollowedbyaboomloweringwith»slow«valvesettingandsoon.Whenactuated, the joystickwaspushedtoextremitywithafastmovement.Atenddestinationsoftheboom(liftedandlowered),the joystickwasreleasedbacktotheneutralpositionforat least5secondstoensurepressurestabilizationinthehy-draulic circuit.

Theboomtippositionvariedfromapproximately1.5–4.5mabovegroundduringaliftcycle.Theopera-tor(a31year-oldmalewithnopreviousexperienceofworkwithheavymachinery) tried to keep the liftheightsconstantthroughouttheexperiment.Initialandfinalcylinderlengths(Fig.1)weredocumented(Table1),andthepossibleeffectsoftheirvariationontheresultswerereducedstatistically(seechapter,»2.3Statisticalanalysis«).Byclosingandopeningcertainvalvesinthehy-

draulic system, the accumulator could be overridden. Whenitwasoverridden,theEHLCfunctionedasastandardhydraulicliftcylinder(standardcylinder)providingareferencecylinder forcomparisonsofenergyuse(Fig.2:panel1).

Pressure observations (p1 and p2)wererecordedbyBoforsTDS-1loadcells(error±500kPa)(Fig.1).Bothof the loadcellsusedwerecalibratedwithaBarnetInstrumentsdeadweighttesteronceat thebeginningoftheexperiment.Thecylinderlengthsen-sorused(error±2.3mm)wasbasedona10turn1kΩpotentiometer.Datafromloadcellsandthecylinderlengthsensorwererecorded105timespersecondus-ingaDEWE2520dataloggerwithanintegratedcom-puterrunningtheDEWEsoft6.5program.

Fig. 3 Schematic visualization of flawless work by EHLC. A lift starts by pressurizing the primary cylinder (a), and the boom starts to lift (b) as the primary cylinder pressure ( p1) reaches a required threshold. The secondary cylinder pressure ( p2) decreases (c) until the end of boom lift (d) as the increasing secondary cylinder volume discharges the accumulator (Fig. 1). During the boom lowering (e) p2 increases (f) until the end of boom lowering (g) as the decreasing secondary cylinder volume recharges the accumulator (h). p2 creates an assisting lift force, which is inversely proportional to the lift cylinder length. There are risks of p2 and p1 intersecting without a pressure surplus from the occa-sional pressure spikes. However, no occasional pressure spike is shown in the figure, only the pressure surplus

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 223

Ageneralmodelfordeterminingliftworkduringa given time interval (Wt®t+1)(Eq.1)wasusedasa start-ingpointforcalculatingtheworkconductedbythetwo cylinders. Wt®t+1 =0.5´ (pt + pt+1)´ Apiston ´ (Lt+1–Lt) (1)Where:

p thecylinderpressureattimet or t+1,Apistonpistonarea,L is the cylinder length at time t or t+1.Theworkconductedduringagivenboomliftwas

calculatedbydividingtheliftinto105 time intervals persecondandsummingtheworkforalltimeinter-vals.Thestandardcylinder'sliftworkduringtimeinterval

from t to t+1 (Ws, t®t+1)wascalculatedbasedonEq.1,withp1usedforpressure(i.e.thepressureobservationsretrieved from load cell1,Fig.1),andAspaspistonarea(Fig.2:panel1).CalculationofthefunctioningofEHLC’sliftworkinatechnicallyperfectstate(WEHLC,t®t+1)wasbased on p1 and A1aspistonarea(Fig.2:panel2).How-ever,theEHLCmightnotworkperfectly;apossiblecondition being when p1mightexceedp2duringpartofthelift.

Inthisstudy,thepartofaliftwherep1<p2 is referred toastheEHLC's»successfulliftphase«(Fig.4:t1®t2),becauseduringthisphasetheEHLCistheoreticallycapableofcontributingtotheliftwithrecoveryenergy.Similarly,the»unsuccessfullifephase«referstothepartoftheliftafterwhichp1>p2forthefirsttimeandtheEHLCwillnotbeabletocontributetoreductionsin energy use.LiftworkcalculationsfortheEHLC’ssuccessfullift

phasewereidenticaltothecalculationsforflawlessfunctioningoftheEHLC.However,workcalculationsforEHLC’sunsuccessfulliftphasevarieddependingon whether p1 or p2washighest.Fortimeintervalswithp1>p2theEHLCwasassumedtofunctionasastandardcylinderandworkwasdeterminedcorrespondingly(Ws, t®t+1).Fortimeintervalswithp1<p2 during the un-successfulliftphase,theEHLC'sworkwasdeterminedas the sums of WEHLC,t®t+1andthesecondarycylinderliftwork (W2, t®t+1),giventhatregainingtheEHLC'sfunc-tionality was a result of accumulator being loaded from thehydrauliccircuitduringthelift(i.e.consumingen-ergy),andnotbyusingtherecoveredpotentialenergy.W2, t®t+1 wascalculatedaccordingtoEq.1withp2 used aspressureandA2asthepistonarea(Fig.2:panel2).

Table 1 Initial and final lift cylinder lengths, stroke lengths, lift times and average piston velocities during boom lifts for each combination of lift cylinder model (cylinder model), payload and directional control valve setting (valve setting). Mean values, with standard deviations in parenthesis

Payload kg

Valve setting

Cylinder model

Observations n

Initial cylinder length mm

Final cylinder length mm

Stroke length mm

Lift time s

Piston velocity mm/s

0

SlowStandard 25 1093 (1) 1213 (3) 121 (3) 4.6 (0.1) 26.1 (0.3)

EHLC 27 1091 (1) 1211 (3) 120 (3) 4.6 (0.1) 26.1 (0.3)

MediumStandard 34 1093 (3) 1223 (5) 130 (4) 1.7 (0.1) 77.4 (2.5)

EHLC 32 1091 (3) 1219 (8) 128 (8) 1.7 (0.1) 76.7 (2.7)

FastStandard 57 1092 (6) 1223 (5) 131 (5) 1.5 (0.1) 85.5 (2.4)

EHLC 34 1096 (8) 1224 (9) 128 (12) 1.5 (0.1) 84.1 (3.2)

264

SlowStandard 21 1113 (1) 1228 (2) 116 (2) 7.7 (0.2) 15.0 (0.2)

EHLC 22 1113 (1) 1229 (2) 117 (2) 7.9 (0.2) 14.8 (0.2)

MediumStandard 20 1111 (3) 1234 (10) 123 (12) 1.8 (0.1) 68.8 (3.0)

EHLC 27 1114 (5) 1233 (6) 119 (9) 1.8 (0.1) 64.7 (1.5)

FastStandard 26 1037 (12) 1173 (8) 136 (15) 1.9 (0.2) 71.3 (2.6)

EHLC 27 1093 (9) 1236 (10) 144 (16) 2.0 (0.2) 72.6 (1.9)

513

SlowStandard 15 1095 (1) 1213 (1) 118 (2) 20.6 (0.6) 5.7 (0.1)

EHLC 18 1094 (3) 1213 (1) 119 (3) 19.3 (1.0) 6.2 (0.2)

MediumStandard 20 1060 (9) 1220 (8) 160 (13) 3.9 (0.4) 41.6 (2.6)

EHLC 20 1066 (6) 1229 (9) 162 (9) 3.5 (0.2) 46.2 (1.8)

FastStandard 26 1057 (18) 1216 (11) 159 (15) 3.7 (0.8) 44.4 (6.8)

EHLC 22 1053 (12) 1233 (8) 179 (9) 3.7 (0.3) 49.0 (2.9)

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

224 Croat. j. for. eng. 37(2016)2

TheenergysavingsfortheEHLC'ssuccessfulliftphaseweredeterminedasthesecondarycylinderproportionof theEHLC's totalworkduring thesuccessful liftphase,whichinturnwasdeterminedasthesumoftheprimaryandsecondarycylinderwork.

2.3 Statistical analysisThestatisticalanalysisincludedonlyboomlifts,as

boomloweringswereexcluded.Analysisofcovari-ance(ANCOVA)wasusedtoevaluateeffectsofthreefixedfactors(liftcylindermodel,payload,andvalvesetting)ontwodependentvariables:totalworkperliftand the initial p1, retrieved from a load cell1(Fig.1).Thefactorcylindermodelhadtwolevels‒EHLCandstandardcylinder,whilethefactorpayloadhadthreelevels‒objectswiththemass0,264and513kg‒heldintheboomtip.Finally,thefactorvalvesettinghadthreelevels:slow,mediumandfast(Table1).Intotal,thisresultedin18treatments,whichwereeachrepli-catedseveraltimes(15≤n≤57,Table1).Thethree-fac-torialmodelcontainedallpossibleinteractioneffectsbetween factors.Covariateswereusediftheysignificantlycontrib-

uted to the model, and were considered logical and notriskedtobeconfoundedwithtreatmenteffects.Inthisexperiment,thecontinuousvariables‒initialandfinalcylinderlength,strokelengthandlifttime‒wereusedascovariatesforeachboomlift.Toavoidarank

deficiency,theinitialandfinalcylinderlengthswereprioritizedoverthestrokelength.Thestrokelengtheffectwastestedonlyif theinitialorfinalcylinderlengthhadnoeffect.Inadditiontothedependentvariablesmentioned

above,theEHLCwasalsoevaluatedseparatelytoad-dressitsfunctionality(i.e.withoutcomparisonbetweencylindermodels).Forsuchanalysesall thecylindermodel-related terms were removed from the three-way ANCOVA,resultinginatwo-wayANCOVAwiththefixedtwo-wayinteractioneffect:payload´ valve set-ting.Thedependentvariablesanalyzedwererelatedtopressureintheprimaryandsecondarycylindersaswellasworkandtimeduringthesuccessfulliftphase.Theorderbetweentreatmentswasrandomized,

butallreplicateswithinatreatmentwereconductedsequentially.However,thefirst10boomliftswithineachtreatmentwereexcludedfromtheanalysistoen-surethatthesystemhadstabilizedintermsofoilpres-sureandtemperatureduringthedatacollection.Agenerallinearmodel(GLM)wasusedtoanalyze

theanalysisofvariance(ANOVA)andANCOVAmod-els.DuringtheGLMprocedure,pair-wisedifferenceswereanalyzedwithTukey’ssimultaneoustestofmeans.ThenormalityofresidualswasevaluatedbytheAnder-son-Darlingtest.Differencesininitialcylinderpres-sureswithintheEHLCweretestedfordeviationfrom

Fig. 4 Example of observed pressures and cylinder lengths as a function of time for one whole EHLC boom lift cycle (t1®t7) with the valve setting »medium« and payload of 0 kg. The boom lift, as well as the successful lift phase, starts at time t1. The successful lift phase ends at t2 when p2 exceeds p1 for the first time. The boom lowering, as well as charging of the accumulator, starts at t4. At t5®t6, p2 drops rapidly because the pressure relief valve opens and oil flows to the oil reservoir (Fig. 1). At t7, a new lift starts

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 225

zerobyuseofaone-samplet-test. The critical level of significancewassetto5%.Minitab16(MinitabLtd.)was used for all analyses.

3. ResultsTheEHLC'sfunctionalityisdependentonthepres-

sureinthesecondarycylinderexceedingthepressureintheprimarycylinder(p1<p2,Figs.1–2).However,thiswasfoundtoneveroccurduringafulllift,butonlyduringvariousintervalsofthefirstpartoftheboomlift,whichvariedfromrepetitiontorepetition(Fig.4:t1®t2).Moreover,p2 and p1werepracticallyidenticalduring thewhole liftwhenusing thevalvesetting»slow«(nodatashown).

Forvalvesettings»medium«and»fast«,theEHLC'sp2levelsurpassedp1inthebeginningofalift(Fig.4).Duringthis»successfulliftphase«(Fig.4:t1®t2),theaccumulator contributed recovered energy from the precedinglift.Duringthenextboomlowering,theac-cumulator was loaded as p1<p2(Fig.4:t4®t5).Whenp2 exceeded approximately 30MPa, the accumulatorstoppedcharging,indicatingthatthepressurereliefvalvewasreleasedatthatpressure(Fig.4:t5®t6).

3.1 Comparison of cylinder models for whole liftsThethreemainfactors–thecylindermodel,pay-

load,andvalvesetting–significantlyaffectedbothofthetwodependentvariablesinitialp1 and the energy

Table 2 Levels of significance ( p-values) and explained variance (R2 adjusted values) obtained from the analysis of variance (ANOVA) of: effects on the dependent variables listed in the first column of the factors cylinder model (a), payload (b) and valve setting (c); their fixed interaction effects (a ´ b, a ´ c, b ´ c and a ´ b ´ c); and effects of the covariates initial lift cylinder length (d), final lift cylinder length (e), lift time (f) and stroke length (g)

p-valueAdj. R2

%n

Dependent variablesFactor Covariate

a b c a×b a×c b×c a×b×c d e f g

Energy consumption per whole lift <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 – 99.7 473

Initial p11) <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 Excl. – – – 83.8 473

Initial p21) – <0.001 <0.001 – – <0.001 – <0.001 – – – 99.6 229

Difference between initial p2 and initial p1 – <0.001 <0.001 – – <0.001 – Excl. – – – 94.4 229

Number of p2 and p1 intersections per lift – <0.011 <0.001 – – <0.001 – <0.001 Excl. Excl. Excl. 73.3 229

W21) during the successful lift phase – <0.001 <0.001 – – <0.001 – Excl. – – – 76.6 162

W11) during the successful lift phase – <0.001 <0.001 – – <0.001 – Excl. – – – 78.2 162

Relative energy saving for successful lift phase2) – <0.001 0.068 – – <0.001 – Excl. – – – 43.7 162

Successful phase stroke length – <0.001 <0.001 – – <0.001 – Excl. – – – 76.0 162

Successful phase proportion of total stroke length

– <0.001 <0.001 – – <0.001 – Excl. 0.005 – Excl. 76.5 162

Successful phase lift time – <0.001 <0.001 – – <0.001 – 0.011 – – – 88.6 162

Successful phase proportion of the total lift time

– <0.001 <0.001 – – <0.001 – Excl. 0.002 – Excl. 78.3 162

n£229 when only the EHLC treatments were included, and n=162 when only the EHLC treatments with medium and fast directional control valve settings were included– independent variable was not testedExcl. covariate term was tested, but excluded from the model because it had no effect (p>0.05) or it decreased adj. R2-value1) p1 primary or standard cylinder pressure (depending on cylinder type)p2 secondary cylinder pressure (only EHLC)W2 secondary cylinder workW1 primary cylinder work2) W2 during the successful lift phase/(W2 during the successful lift phase+W1 during the successful lift phase)

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

226 Croat. j. for. eng. 37(2016)2

consumption per full lift (three-way ANCOVA,p<0.001,Table2:rows1–2).Asexpected,mostofthediscrepanciesbetweenthetwodependentvariableswereexplainedbythepayload,whilethevalvesettingandcylindermodelonlyhadminoreffects(datanotshown).The initialp1wasnotaffectedby theonlytestedcovariate‒theinitialcylinderlength(three-wayANCOVA,p=0.774,Table2:row2,completedatanotshown).Ontheotherhand,theenergyconsumptionperliftwasaffectedbyalltherecordedcovariates:theinitialandfinalcylinderlength,andlifttime(three-wayANCOVA,p<0.001,Table2).Overall, the totalenergyconsumptionmodelwasimprovedmostbyinclusion of the initial cylinder length, followed by the finalcylinderlength.Thelifttimehadleasteffect(datanotshown).

TheEHLC'sinitialp1increasedsignificantly(p<0.001)withincreasingpayloadacrossallthethreevalveset-tings(Table3).However,thestandardcylinder'sinitialp1 increased significantly (p<0.001)with increasingpayloadonlyacrossthevalvesetting»fast«.Inaddi-tion,therewasalackofstatisticallysignificantdiffer-encesbetweenthetreatmentswhenthepayloadef-fectswerecomparedoverthevalvesettings»slow«and»fast«.Thisresultedinsignificant(p<0.001)two-andthree-wayinteractioneffects(Tables2–3).Asexpected,theenergyconsumptionperliftfor

bothcylindermodelsincreasedsignificantly(p<0.001)withincreasingpayloadacrossallthreevalvesettings(Table3).Thevalvesettingssignificantly(p<0.05)af-fectedEHLC's energyconsumptionwithpayloadsof0and513kg,buthadnosignificanteffectwiththe264kg

Table 3 Pressure and energy consumption for a whole lift with EHLC and standard cylinder for each combination of payload, valve setting and cylinder model. Mean values, with standard deviations in parenthesis

Payload kg

Valve setting

Cylinder model

Initial p11)

kPaInitial p2

kPa

Difference between initial p2 and initial p1

2), kPa

Number of p2 and p1 intersections during

the whole lift, n

Energy consumption per whole lift, J

Observations n

0

SlowStandard 14,182DE (304) – – – 18,992EF (346) 25

EHLC 13,068E (98) 12,591G (129) –479F (81) 47.5A (18.9) 19,330E (410) 27

MediumStandard 13,264E (2617) – – – 18,257F (551) 34

EHLC 10,464F (1670) 24,374ABC (465) 13,909A (1912) 5.1C (2.3) 17,669G (1340) 32

FastStandard 13,105E (2698) – – – 18,400F (712) 57

EHLC 9791F (777) 21,637D (400) 11,767B (762) 11.7BC (4.4) 18,488EF (1573) 34

264

SlowStandard 18,328BC (524) – – – 28,514C (388) 21

EHLC 17,588C (208) 16,863F (195) –1053F (93) 11.8BC (3.1) 28,186C (465) 22

MediumStandard 21,647A (2163) – – – 27,390C (2423) 20

EHLC 14,985D (1154) 24,516AB (180) 9178C (1208) 23.4B (3.4) 27,185C (1663) 27

FastStandard 15,167D (936) – – – 23,064D (3010) 26

EHLC 15,262D (576) 24,152C (211) 8870C (506) 14.5BC (6.0) 27,172C (3294) 27

513

SlowStandard 19,859AB (145) – – – 36,164AB (475) 15

EHLC 20,043AB (115) 19,868E (121) –227F (33) 18.4B (7.9) 35,642AB (841) 18

MediumStandard 21,142A (1725) – – – 32,773A (3214) 20

EHLC 19,928AB (2417) 24,655A (282) 5119D (2081) 24.4B (11.0) 32,097B (2107) 20

FastStandard 20,518A (2047) – – – 32,763A (4298) 26

EHLC 21,662A (1896) 23,974BC (244) 2909E (1373) 58.3A (21.9) 32,720A (2269) 22

Within columns, different superscript letters indicate significant differences (p<0.05). Statistical models are described in Table 21) p1 pressure recorded in load cell1 (Fig. 1), i.e. in the primary EHLC cylinder and standard cylinderp2 pressure recorded by load cell2, i.e. in the accumulator circuit and the secondary cylinder2) Each of the nine EHLC mean values in the column were statistically significantly different from zero (one-sample t-test, p<0.001)

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 227

payload.Incontrast,thestandardcylinder'sliftenergyconsumptionwassignificantlyaffectedbythevalvesettingswiththe264kgpayload,butnotwiththe0and 513 kg payloads. This resulted in significant(p<0.001) two- and three-way interaction effects(Tables2–3).

3.2 Evaluation of the EHLC's successful and unsuccessful lift phases

In addition to the initial p1andenergyconsump-tionperfulllift,theEHLC'ssuccessfulliftphasewasfurtheranalyzedwithsupplementaldependentvari-ables(Table2:numberofobservationsintherangeof162–229).Thetwomainfactors,payloadandvalveset-ting,significantly(p<0.001)affectedallthesupplemen-taldependentvariables,whendataforallthreevalvesettingswereincludedintheanalyses(Table2).Theinitial p2, and number of intersections of p2 and p1perlift,werealsosignificantlyaffectedbythecovariateinitial cylinder length (p<0.001)(Table2).Thesignifi-cant (p<0.001)interactioneffectsshowedthatthefac-torseffectsvariedbetweenthecomparedtreatments,andoccasionally,theinteractioneffectwastheresultofalackofdifferencesbetweenthecomparedtreat-ments(Tables2–3).Withthevalvesetting»slow«,theinitialp1 was sig-

nificantlyhigherthanp2(Table3),thusthepossibilitiesof energy recovery were eliminated even before the boomwaslifted.Consequently,onlydataobtainedwiththevalvesettings»medium«and»fast«werefur-theranalyzed.Withthesesettings,theinitialp2 was substantially higher than the initial p1, which enabled energyrecovery(Table3).Thepayloadandvalveset-

tingsignificantly(p<0.05)affectedallexceptonede-pendentvariableinEHLC'ssuccessfulliftphase(Ta-ble4).Theexception,whichfelljustoutsidethesetlevelforsignificance,wasthattherelativeenergysavingsforEHLC'ssuccessfulliftphasewasnotaffectedbythevalvesetting(p=0.068)(Table2).Inaddition,thevalvesettingshadasignificant(p<0.05)effectonthedepend-entvariableswiththe0kgpayload,butnotwiththe264or513kgpayloads(Table4).Thisresultedinsig-nificanttwo-wayinteractioneffects(p<0.001,Table2).TheEHLC'ssuccessfulliftphasecorrespondedto

9.3–10.2%ofthetotallifttimewhenliftingapayloadof0kgwiththevalvesetting»fast«,orwhenliftingapayloadof264kgwitheither»fast«or»medium«valvesettings.However,giventhatthesuccessfulliftphasecoveredonlyanaccelerationphase,thesuc-cessfulliftphaseproportionofthetotalstrokelengthwasonly0.3–3.1%.Thus, W2wasonly17–34J(Table4).TheEHLC'ssuccessfulliftphaseconstituted31.9–33.7%ofthetotallifttimewhenliftingapayloadof513kgwitheithervalvesettings,correspondingto13.7–17.7%of the total stroke length and W2of340–407J(Table4).Whenusingthevalvesetting»medium«andliftingapayloadof0kg,thesuccessfulliftphasewas51.9%ofthetotallifttime.Thiscorrespondedto52.3%ofthe total stroke length and resulted in a W2of1033J(Table4).

3.3 Energy savingsWhenevaluatingafullboomlift,theEHLCfunc-

tionedbestforpayloadsof513kgand0kgwiththevalvesetting»medium«.Undertheseconditionslift-ingwiththeEHLCconsumedsignificantlylessenergy

Table 4 EHLC's energy savings (mean values with standard deviations in parenthesis) for the successful lift phase

Payload kg

Valve setting

Cylinder work 1)

Energy saving 2)

%

Successful lift phase stroke length

mm

Successful lift phase proportion of the total

stroke length, %

Successful lift phase lift time

s

Successful lift phase proportion of the total

lift time, %

Observations nW2, J W1, J

0Medium 1033A (423) 9422A (4191) 9.4A (1.0) 69.6A (31.4) 52.3A (22.0) 0.84A (0.32) 51.9A (17.5) 32

Fast 34C (15) 396C (192) 8.4B (1.3) 2.0C (0.9) 0.3C (0.11) 0.19B (0.03) 9.3C (2.2) 34

264Medium 26C (14) 318C (140) 8.2B (0.9) 1.4C (0.6) 2.6C (1.1) 0.25B (0.06) 10.0C (3.4) 27

Fast 17C (9) 191C (87) 8.5B (0.9) 0.8C (0.3) 3.1C (0.5) 0.17B (0.03) 10.2C (1.5) 27

513Medium 407B (30) 5326B (365) 7.1C (0.1) 22.1B (1.6) 13.7B (1.3) 0.98A (0.07) 33.7B (1.3) 20

Fast 340B (29) 4496B (386) 7.0C (0.0) 29.1B (6.6) 17.7B (4.2) 0.98A (0.12) 31.9B (1.7) 22

Within columns, different superscript letters indicate significant differences (p<0.05). Statistical models are described in Table 21) W2 secondary cylinder work, and W1 = primary cylinder work2) W2 during the successful lift phase/(W2 during the successful lift phase+W1 during the successful lift phase)×100

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

228 Croat. j. for. eng. 37(2016)2

than lifting with the standard cylinder (2.1–3.2%,p<0.001)(Table3).However,whenlifting264kgwiththevalvesetting»fast« theEHLCdidnot functionflawlesslyatall,andconsumedsignificantlymoreen-ergyforafullliftthanthestandardcylinder(17.8%,p<0.001)(Table3).Duringthesuccessfulliftphase,theEHLCsaved

approximately 7.0–9.4% of the energy consumed(p<0.001).Therelativeenergysavingswerealsosig-nificantlyhigherwithalowerpayload(Table4).

4. Discussion

4.1 Comparison with previous studiesDuringthesuccessfulliftphaseEHLCsavedupto

9.4%oftheenergyconsumed.However,onaveragethesuccessfulliftphaseaccountedforonlyasmallproportionofthewholeboomlift,andsometimesdur-ingthefollowingunsuccessfulliftphasetheEHLCevenincreasedenergyrequirements.Inaddition,thesavingsduringthesuccessfulliftingphasewereinthelowerendofrangesofsavings(ca.5–65%)reportedinpreviousevaluationsofothersolutionsforreducingenergy requirementsofmobile liftingdevices (e.g.LiangandVirvalo2001a,SunandVirvalo2003,Ryd-berg2005,SunandVirvalo2005,VirvaloandSun2005,Linetal.2010,LinandWang2012,Minavetal.2012,Noréusetal.2013,Wangetal.2013).Furthermore,theEHLCdidnotgiveanyenergyrecoveryatallwiththevalvesetting»slow«,presumablybecauseofinternaloilleakagebetweentheprimaryandsecondarycylin-ders.Theboomliftingandloweringtimeswereatleasttwiceaslongwiththe»slow«valvesettingthanwiththesettings»medium«or»fast«,andinevitablyinter-nal leakageandthus thepressuredecreasewillbelarger if the duration is longer (assuming all other variables remainconstant).Anexampleof leakagefromthesecondarycylinderisshowninFig.4,wherep2 is decreasing during the time interval t6®t7 and the onlypossiblereasonforthedecrease(afterthepres-surereliefvalvehasclosed)isleakage.Thefindingthatdegrees of leakage are correlated with its duration is consistentwithpreviousreports(Wangetal.2013).AnapparentproblemwiththetestedEHLCisthat

the accumulator gas volume seems to be too small in relationto thesecondarycylinderoilflowvolumeduringtheboomliftingsandlowerings.ThisissuecanbeseeninFig.4,wheretheEHLC'spressurecurvedecreases sharplyas a functionof cylinder length(t1®t2),thenincreasesrapidlyuntilthepressurereliefvalveopens(t4®t5).AccordingtoBoyle’slaw,increas-ing the accumulator gas volume could solve this

problemasitwouldstabilizep2orreduceitspeak-to-peakpressureamplitude.Themainproblemhereisnottoolowmaximump2, which is already regulated bythepressurereliefvalve,asdescribed,butthatp2 decreasestoorapidly.

4.2 Strengths and weaknesses of the studyThestudywasconductedinanexperimentalset-

ting,withstandardizedworkproceduresandhighdatarecordingfrequencies,facilitatingisolationoftheeffectsofcylindertypeundervariousworkcondi-tions,andpotentially,theidentificationofusefulgen-eralproceduresforassessingforwardercranesenergyrequirements,forcesandfunctionalparametersdur-ingliftwork.ConsideringtheEHLCasanearlypro-totype,wehave focussedmainly onworkphaseswhere the EHLCmade a successful contributionwhenanalyzingthedata.

The forwarder used was a standard forwarder equippedwithaloadsensingsysteminthehydraulicsystem,sothehydraulicoilflowandpressuredeliv-eredfromthehydraulicpumpdependedonthepow-errequiredatthemoment(seee.g.Schereretal.2013).The load sensing system was not overridden during theexperiment.Thus,effectsfromtheloadsensingsystem could have been confounded with the tested factoreffects,whichmayhaveaffectedtheANOVAresults.However,toourknowledgeanyeffectsoftheload sensing system should be negligible for the tech-nicalevaluationoftheEHLC.Whendeterminingtheenergyconsumptionofthe

EHLCforawholeboomlift,itwasassumedthatthesecondarypistonthrustseithertheprimaryorthesec-ondarycylinderhead.However,thisassumptionwasnotentirelyvalidbecausethesecondarypistoncouldalsomoveinthedirectionoflowerpressure.Never-theless,thiserrorsourcesonlyapplieswhenthesuc-cessfulliftphaseiscompleteandshouldnotthereforeimpacttheanalysisoftheEHLC'ssuccessfulliftphase.

4.3 Improvements and future studiesTheresultsofthisstudyindicateseveralpossible

improvementsfortheEHLC.First,theinternalleak-ageproblemhastoberesolved.Second,inadditiontoenlargingtheaccumulator,itspressurizationshouldbe optimized so that the EHLC works flawlesslythroughoutthewholeliftingphase.Thissuggestionispromptedbytheobservationthatthegeneratedpres-surespikesdonotappeartobesufficienttomaintainpressurizationslongenoughtoreduceenergyneedswithinthecurrentaccumulatorsystem.Consequently,itisessentialtoredesignthesystemtoprovideano-tablyhigherinitialaccumulatorpressure.Duringthe

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 229

»unsuccessfulliftphase«(afterthefirstp2 and p1 inter-section),thehydraulicpumppressurizedtheprimarycylinder as well as the accumulator and secondary cylinder(Fig.4:t2®t3).However,flawlessfunctioningoftheEHLCdependsontheactualboomlengthandload, since if p2 is too high and the boom too short energy will actually be needed to lower the boom. This reflectsageneral trade-offforallweight-balancingsystems,andrequiresoptimizationaccordingtothegivencranedimensionsandtheloadsliftedandlow-ered. Third, A2shouldbeenlarged.Overall,theex-perimentshouldbereplicatedwithalargeraccumula-tor,arefinedaccumulatorpressurizationsystem,andpossiblyalargerA2.

4.4 Potential energy savings for all forwarder workThecurrentEHLCcansaveupto3.2%ofenergy

forafullboomliftunderoptimalconditions.How-ever, if its technical weaknesses can be resolved, the savingsshouldbeatleast6.6%,asexplainedintheIntroduction.Moreover,ifsomeadditionaltechnicalimprovementsaremade,additionalenergysavingscouldbeachieved.Forinstance,iftheaccumulatorpressurizationsystemisredesignedtoensurethatp2 isconsistentlyatleast20%higherthanp1, the energy savingsforawholeboomliftshouldtheoreticallybeat least 7.9%. Inaddition, increasingA2 would in-creasetheenergyrecyclingpotentialoftheEHLC.Forexample, increasingsecondarypistondiameterby20%wouldincreasethepossibleenergysavingsfrom6.6%to9.5%.Therefore,theEHLC'senergyrecyclingpotentialcouldbeenhancedbyincreasingbothsec-ondarypistondiameterandp2by20%.Withthesetwoimprovements,theEHLCcouldtheoreticallyrecycleatleast11.4%ofpotentialenergy,ignoringpossibleleakage.However,energy-efficientliftingdeviceshavenot

yetacquiredanyofthemarketsharesforforwardercranes.Sofar,theEHLC'smarketconsistsprimarilyofhydraulicliftdevices,wherethefullenginepowerisusedforliftwork,whichisnotthecaseforforwarders.Currentlyminorenergysavingscanbegainedfromenergy-efficientliftcylindersforastandardforwarder,asdrivingthemachinerequiressubstantiallymoreforceandpowerthantheboomlifting(cf.Löfgren1999,Edlundetal.2013,Table3).Thus,theresearchpriority shouldbeplacedondecreasingpower re-quirementsduringthedrivingphase,i.e.improvingpowertrainefficiency,asinafewrecentstudies(e.g.Edlundetal.2013,SwedishEnergyAgency2014).Overall,combiningenergy-efficientliftingdevices

withahybridpowertraincouldbeinterestingforfu-

turestudies.Withanenergy-efficientliftingdevicereducing the energy needed for the crane work, more powerfromthecombustionenginecouldbedirectedto loading thebattery throughout thecranework.This electric energy, stored during the crane work, would then be available for use during driving, when the power input requirement is highest. Thus, anenergy-efficientliftdevicecouldimprovethebatteryloadingefficiencyofahybridsystemduringcranework.Suchtechnologycouldenabletheuseoflesspowerful, i.e. less fuelconsuming,combustionen-gines in forwarders.

AcknowledgementThisstudywasfundedbyStoraEnsoSkogABand

theForestIndustrialResearchSchoolonTechnology(FIRST).WethankSees-editingLtdforrevisingtheEnglish.

5. ReferencesDeepak,S.R.,2012:Staticbalancingofrigid-bodylinkagesandcompliantmechanisms,Doctoralthesis.IndianInstituteofScience.Thefacultyofengineering.Departmentofme-chanicalengineering.Bangalore,India.168p.

Edlund,J.,Bergsten,U.,Arvidsson,H.,2013:Aforestma-chinebogiewithabearingcapacitydependentcontactarea:accelerationandangularorientationwhenpassingobstaclesanddrawbarpullforceandfreerollingresistanceonfirmground.SilvaFennica47(3):articleid1017,7p.

Einola,K.,2013:PrestudyonPowerManagementofaCut-To-LengthForestHarvesterwithaHydraulicHybridSys-tem.In:The13thScandinavianInternationalConferenceonFluidPower,Linköping,Sweden,71–83.

Eriksson,M.,Lindroos,O.,2014:ProductivityofharvestersandforwardersinCTLoperationsinNorthernSwedenbasedonlargefollow-updatasets.InternationalJournalofForestEngineering25(3):179–200.DOI:10.1080/14942119.2014.974309.

Ersson,B.T.,Jundén,L.,Lindh,M.,Bergsten,U.,2014:Simu-latedproductivityofconceptual,multi-headedtreeplantingdevices.InternationalJournalofForestEngineering25(3):201–213.DOI:10.1080/14942119.2014.972677.

EuropeanUnion,2014:EnergyEfficiencyDirective.http://ec.europa.eu/energy/efficiency/eed/eed_en.htm.Accessed9March2015.

Gawlik,A.,Michałowski,S.,2008:Conceptofactivecoun-terweightsystemformediumsizeexcavators.In:5thFPNI-PhDSymposium,Cracow,Poland,462–465.

Gerasimov,Y.Y.,Siounev,V.S.,1998:Forestmachinerycranecompoundschemesynthesis:»Skeleton«part.JournalofForestEngineering9(1):25–30.DOI:10.1080/08435243.1998.10702708.

Gerasimov,Y.Y.,Siounev,V.S.,2000:Forestmachinerycranecompoundschemesynthesis:Optimizationofhydrauliccyl-

J. Manner et al. Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231)

230 Croat. j. for. eng. 37(2016)2

inderoperatingmechanisms.JournalofForestEngineering11(1):73–79.DOI:10.1080/08435243.2000.10702746.

Ho,T.H.,Ahn,K.K.,2012:Designandcontrolofaclosed-loophydraulicenergy-regenerativesystem.AutomationinConstruction22:444–458.

Hui,S.,Junqing,J.,2010:Researchonthesystemconfigura-tionandenergycontrolstrategyforparallelhydraulichy-bridloader.AutomationinConstruction19:213–220.

Ingvast,H.,1989:Hydraulikensgrunder(Basichydraulics)Utgåva2. Institutet för tillämpadhydraulik (ITH).Örn-sköldsvik,Sweden,243p.

Jundén,L.,Bergström,D.,Servin,M.,Bergsten,U.,2013:Simulationofboom-corridorthinningusingadouble-cranesystemanddifferent levels of automation. InternationalJournal of Forest Engineering 24(1): 16–23. DOI:10.1080/14942119.2013.798131.

Kim,J.,Yoon,G.H.,Noh,J,Lee,J.,Kim,K.,Park,H.,Hwang,J.,Lee,Y.,2013:Developmentofoptimaldiaphragm-basedpulsationdamperstructureforhigh-pressureGDIpumpsystemsthroughdesignofexperiments.Mechatronics23(3):369–380.

Laine,T.,Rantala,J.,2013:Mechanizedtreeplantingwithanexcavator-mountedM-Planterplantingdevice.Internation-al Journal of Forest Engineering 24(3): 183–193. DOI:10.1080/14942119.2013.844884.

Liang,X.,Virvalo,T.,2001a:Developmentandresearchofanenergysavingdriveinahydrauliccrane.In:7thScandi-navianInternationalConferenceonFluidPower,Linköping,Sweden,151–161.

Liang,X.,Virvalo,T.,2001b:Energyreutilizationandbalanceanalysisinahydrauliccrane.In:The5thInternationalCon-ference on Fluid Power Transmission and Control,Hangzhou,China,306–310.

Lin,P.Y.,Shieh,W.B.,Chen,D.Z.,2013:Atheoreticalstudyofweight-balancedmechanismsfordesignofspringassis-tivemobilearmsupport(MAS).MechanismandMachineTheory61:156–167.

Lin,T.,Wang,Q.,Hu,B.,Gong,W.,2010:Researchontheenergyregenerationsystemsforhybridhydraulicexcava-tors.AutomationinConstruction19(8):1016–1026.

Lin,T.,Wang,Q.,2012:Hydraulicaccumulator-motor-gen-erator energy regeneration system for a hybrid hydraulic excavator.ChineseJournalofMechanicalEngineering25(6):1121–1129.DOI:10.3901/CJME.2012.06.1121.

Lindroos,O.,Bergström,D.,Johansson,P.,Nordfjell,T.,2008:Cuttingcornerswithanewcraneconcept. InternationalJournal of Forest Engineering 19(2): 21–27. DOI:10.1080/14942119.2008.10702564.

Löfgren,B.,Granlund,P.,Brunberg,T.,1999:Testav trestoraskotare‒dragkraft,bromsar,rullmotståndochbränsle-förbrukning.SkogforskResultat,nr.2,1999.StiftelsenSkog-bruketsForskningsinstitut,Uppsala,Sweden,4p.

Malmberg,C.A.,1981:Terrängmaskinen.Del2(Theoff-roadvehicle.Volume2).461p.

Manner,J.,Nordfjell,T.,Lindroos,O.,2013:Effectsofthenumber of assortments and log concentration on time con-sumptionforforwarding.SilvaFennica47(4):19p.articleid1030.DOI:http://dx.doi.org/10.14214/sf.1030

Manner,J.,Palmroth,L.,Nordfjell,T.,Lindroos,O.,2016:Loadlevel forwarding work element analysis based on automatic follow-updata.SilvaFennica50(3):19p.articleid1546.DOI:http://dx.doi.org/10.14214/sf.1546.

Manring,N.D.,2005:Hydrauliccontrolsystems.JohnWiley&SonsInc.NewYork,USA,464p.

Minav,T.A.,Virtanen,A.,Laurila,L.,Pyrhönen.J.,2012:Stor-ageofenergyrecoveredfromanindustrialforklift.Automa-tioninConstruction22:506–515.

Noréus,O.,Hägglund,M.,Emlén,M.,2013:MeasurementsandSimulationstoEvaluateStrategiesforImprovedEnergyEfficiencyofaReachStacker,ForwarderandWheelLoader.In:The13thScandinavianInternationalConferenceonFluidPower,Linköping,Sweden,319–325.

OrtizMorales,D.,Westerberg,S.,LaHera,P.X.,Mettin,U.,Freidovich,L.,Shiriaev,A.S.,2014:Increasingthelevelofautomationintheforestryloggingprocesswithcranetrajec-toryplanningandcontrol.JournalofFieldRobotics31(3):343–363.DOI:10.1002/rob.21496.

Rydberg,K.E.,2005:Hydraulicaccumulatorsaskeycompo-nentsinenergyefficientmobilesystems.In:6th Internation-alConferenceonFluidPowerTransmissionandControl,Hangzhou,China,124–129.

Scherer,M.,GeimerM.,WeissB.,2013:ContributiononCon-trolStrategiesofFlow-On-DemandHydraulicCircuits.In:13thScandinavianInternationalConferenceonFluidPower,SICFP2013,Linköping,Sweden,531–540.

Sun,W.,Virvalo,T.,2003:Accumulator-pump-motorasen-ergysavinginhydraulicboom.In:8thScandinavianInterna-tionalConferenceonFluidPower,Tampere,Finland,297–309.

Sun,W.,Virvalo,T.,2005:Simulationstudyonahydraulic-accumulator-balancing energy-saving system in hydraulic boom.In:50thNationalConferenceonFluidPower,LasVe-gas,USA,371–381.

SwedishEnergyAgency,2014:ElforestTechnologies.http://www.energimyndigheten.se/en/Innovations-R--D/Business-Development-and-Commercialisation/Invest-in-CleanTech-Energy/El-forest.Accessed9March2015.

UnitedNations,2014:UNFCCCandtheKyotoProtocol.http://www.un.org/wcm/content/site/climatechange/pages/gateway/the-negotiations/the-un-climate-change-conven-tion-and-the-kyoto-protocol.Accessed9March2015.

VandeVen,J.,2013:Constantpressurehydraulicenergystoragethroughavariableareapistonhydraulicaccumula-tor.AppliedEnergy105:262–270.

Evaluation of a New Energy Recycling Hydraulic Lift Cylinder for Forwarders (219–231) J. Manner et al.

Croat. j. for. eng. 37(2016)2 231

Virvalo,T.,Sun,W.,2005:Improvingenergyutilizationinhydraulicbooms–whatitisallabout.In:6th International Conference on Fluid Power Transmission and Control,Hangzhou,China,55–65.

Wang,T.,Wang,Q.,Lin,T.,2013:Improvementofboomcontrolperformanceforhybridhydraulicexcavatorwithpotentialenergyrecovery.AutomationinConstruction30:161–169.

WIPOPatentApplicationWO/2011/075034,2011:WorldIn-tellectualPropertyIndicators.9p.

Zhang,L.,2011:Anenergy-savingoildrillingrigforrecover-ingpotentialenergyanddecreasingmotorpower.EnergyConversionandManagement52(1):359–365.

Received:May12,2015Accepted:December24,2015

Authors’address:

Researcher,JussiManner,PhD.*e-mail:jussi.manner@skogforsk.seSkogforskUppsalaSciencePark75183UppsalaSWEDEN

Assoc.prof.OlaLindroos,PhD.e-mail:ola.lindroos@slu.seSwedishUniversityofAgriculturalSciencesDepartmentofForestBiomaterialsandTechnology90183UmeåSWEDEN

HansArvidsson,seniortestengineere-mail:hans.arvidsson@smp.sp.seSwedishMachineryTestingInstitute90403UmeåSWEDEN

Prof.TomasNordfjell,PhD.e-mail:tomas.nordfjell@slu.seSwedishUniversityofAgriculturalSciencesDepartmentofForestBiomaterialsandTechnology90183UmeåSWEDEN

*Correspondingauthor

Personalcommunication:Roger Gustavssone-mail:roger.gustavsson@thordabCEOThordabABArnäsvallSWEDEN