Piston-Cylinder Aseembly

8/20/2019 Piston-Cylinder Aseembly

1/13

374 Ventil 18 /2012/ 5

■ 1 Introduction

Among all types of positive displa-cement machines piston machinescan be designed for very high ope-rating pressures (above 40 MPa).Piston machines have the potenti-al to achieve very high efcienciesand allow for variable displacementunits. These three attributes form astrong basis for component selecti-

on for highly efcient and compactuid power drive systems. SinceRamelli’s rst axial piston pump de-signed at the beginning of the 17 th century many generations of pistonmachines have been invented, de-signed and produced. The designshave been continuously simpliedover the last decades. Most of thecurrent manufactured piston pumpsand motors have much lower num-

ber of parts than machines desi-gned 50 years ago. A successfuldesigned piston machine requireselegant solutions for many parts,but especially for the highly loadedsealing and bearing interfaces. Mostpiston machines have at least threedifferent types of sealing and bea-ring interfaces as shown in Figure1 exemplarily for swash plate typeaxial piston and radial piston ma-

chine with external piston support.The piston cylinder interface is themost complicated interface in termsof balancing high pressure forces.The piston cannot be hydrostatical-ly balanced like the slipper/swashplate or slipper/outer ring and thecylinder block/valve plate interface,which can be designed as combi-ned hydrostatic/hydrodynamic be-aring. The successful design of thepiston cylinder interface determinesthe achievable maximum opera-ting pressure, maximum speed andmaximum swash plate angle. A verychallenging design goal is to avoidwear by creating a sufcient load

carrying ability of the uid lm andto minimize energy dissipation ina wide range of operating conditi-ons in order to achieve a very highefciency. Many researchers havestudied the behaviour of slippers inaxial and radial piston machines aswell as the cylinder block valve pla-te interface of axial piston machi-nes in order to nd optimal designsolutions. However the amount of

research conducted on the pistoncylinder interface is much larger.The piston cylinder interface repre-sents a very complex hydrodynamicbearing, which has to perform si-multaneously as a sealing elementto seal the displacement chamberfrom case pressure. This doublefunction is especially difcult to sol-ve in those piston machines, wherethe moment generation leads to ahigh side force acting on the piston.Consequently the majority of pastresearch related to piston cylinderstudies focus on designs involvinghigh side load of the piston. The si-tuation is different in bent axis ma-

The Piston Cylinder Assembly in Pi-

ston Machines – a long Journey ofDiscoveryMonika IVANTYSYNOVA

Abstract: This paper summarizes the main contributions of researchers and engineers to the discovery ofphysical phenomena dening the uid lm properties and the operational conditions of the piston cylinderinterface in hydrostatic piston machines. The main focus of this paper is the piston cylinder assembly of desi-gns, where the design principal is based on torque generation requiring a large side load of the piston. Since1965 more than 20 dissertations have been completed on theoretical and/or experimental studies of thepiston cylinder interface. Listing all of the papers published worldwide on analysis of piston kinematics anddynamics, modelling and simulation of piston/cylinder interface and experimental studies would exceed theallowable length of this paper. Therefore only major milestones in discovery will be discussed.

Keywords: hydrostatic piston machines, uid lm properties, theoretical and experimental studies, lubricati-on gap, swash plate, piston kinematics, forces, test rig, numerical model, measurement, piston design

Prof. dr. Monika Ivantysynova,Maha Fluid Power ResearchCenter, Purdue University, USA

HIDROSTATIČNI POGONI


2/13

375 Ventil 18 /2012/ 5

chines and radial piston machinesutilizing special mechanism to mi-nimize the piston side load. Pistonmachines with low side force canutilize piston rings like combusti-on engines. Piston rings allow for amuch better sealing of the displace-ment chamber and also help to re-duce piston friction. Much researchhas been conducted on the pistoncylinder interface for combustion

engines. Many of those research re-sults have been translated and utili-zed in hydrostatic piston machineslike bent axis machines, which canuse piston rings. This might expla-in that there has been less researchconducted on piston/cylinder as-sembly for bent axis machines. Inthis paper only work related to pi-ston machines with high side forceswill be reported.

■ 1 The piston motionAs it will be explained in more detailin this chapter the piston conductsa very complex periodical motion,which can be divided in two parts:the piston macro and the pistonmicro motion. The piston macromotion can be derived analyzingthe piston kinematics.

1.1 Kinematics of thepiston

The kinematics of the piston can bedetermined from the basic mecha-nism describing the rotating group

of each design. A comprehensiveoverview can be found in Ivantysynand Ivantysynova (1993). Figure 2 shows the basic kinematic relation-ship for swash plate type machines.

The piston stroke s K

is dependent on

pitch radius R , swash plate angle and

angular position of the block .

(1)

The piston stroke H K

, which repre-sents the displacement of the pistonfrom outer dead center (ODC) to in-ner dead center (IDC) yields:

(2)

Based on the maximum piston stro-ke, the piston velocity v

K and the

piston acceleration a K

can be de-rived as follows:

Therefore, for a given pump or mo-tor design, the piston sliding velo-city and acceleration are a functionof the angular position and angularvelocity of the shaft.

Due to the rotation of the shaft andthe cylinder block or in case of in-verse kinematics due to the rotationof the swash plate (wobble plate de-sign), the piston has to move along

the piston axis to accomplish the pi-ston stroke. This axial piston motionrepresents a forced motion. Besidesthat the piston has the freedom toturn about its own axis. The pistonturning/spinning motion is not for-ced and therefore depends on thefriction between cylinder and pistonand between piston and slipper. Se-veral researchers have studied thepiston relative motion (Renius 1974,Regenbogen 1978, Hooke and Ka-

koullis 1981, Harris et al 1993, Fangand Shirakashi 1995, Wieczorekand Ivantysynova 2002 and Lasaar2003). Renius (1974) accomplisheda very comprehensive experimen-tal study and conrmed the pistonspin motion in swash plate type axi-al piston machines through measu-rements of piston friction forces inaxial and circumferential direction.Hooke and Kakoullis (1981) studiedexperimentally the dependence ofpiston spin motion on pump speed

Figure 1. Lubricating gaps in piston machines

Figure 2. Swash plate type axial piston pump kinematics

(3)

(4)



3/13

376 Ventil 18 /2012/ 5

and operating pressure and con-cluded that increasing operatingpressure supports the piston spinmotion due to high piston forces

pressing the piston into the slipperball joint and preventing any relativemotion between piston and slipper.Ivantysynova (1983) considered thepiston relative motion in her nume-rical model. She compared calcula-ted pressure elds with measuredpressure proles, which explainedthat the piston spin motion con-tributes to hydrodynamic pressurebuilt up. Fang and Shirakashi (1995)measured the piston spin motionon a modied wobble plate pumpand explained that the spin motioncontributes to the load carrying abi-lity of the piston and therefore helpsto prevent mixed friction. Also Lasa-ar (2003) conrmed with his pistonfriction force measurements using aspecially designed swash plate typemachine that the piston spin moti-on is present and its value dependson operating conditions.

In addition to the described axial

motion and rotation about its axisthe piston is conducting a complexmicro motion within the cylinderbore. The micro motion can be de-rived from the force balance of thepiston.

1.2 Forces acting on thepiston

Forces acting on the piston havebeen studied by many authors

(Brangs 1965, Yamaguchi 1976, vander Kolk 1972, Dowd and Barwell1974, Renius (1974), Hooke andKakoullis 1981, Ivantysynova 1983,

Harris, Edge and Tilley 1993, Ivan-tysyn and Ivantysynova 1993, Fangand Shirakashi 1995, Kleist 1995,Jang 1997, Manring 1999, Sanchen

1999, Olems 2001, Wieczorek 2002,Scharf and Murrenhoff 2005). Fi-

gure 3 shows a free body diagramof the piston. The gure shows allexternal forces acting on the piston.When the pump or motor runs un-der pressure the largest force is usu-

ally the pressure force F DK

.

The pressure force is determined bythe instantaneous cylinder pressure

and therefore changes periodicallyover one shaft revolution. Similar

the piston inertia force F aK

changesperiodically with the piston accele-ration. Dependent on the uid lmconditions and the resulting owbetween piston and cylinder the pi-

ston friction force F TK

will also acton the piston. Finally the reaction

force of the swash plate F SK

and

the piston centrifugal force will

act on the piston. As shown in Fi-

gure 3 the component F SKy

of swa-sh plate reaction force representsa radial force acting on the piston.In addition the y-components ofthe piston centrifugal force and the

slipper friction force F TG

will con-tribute to the resulting side load of

the piston. The resultant force F RK , which acts on the piston perpen-dicular to the piston axis at point Aof the ball joint, can be determined

by vectorial addition of the forces

F SKy

, and F TG

:

(5)

Under the assumption of full uidlm lubrication and absence of me-

tal to metal contact, the force F RK

must be compensated by the uidlm between piston and cylinder,i.e. resulting pressure force andmoment generated from the uid

pressure eld. In other words, thehydrodynamic pressures built up inthe uid lm must develop uid for-ces and moments sufcient to ba-lance the external forces. To illustra-te the amount of force to be carriedby the uid lm lets take a swashplate pump with a 20 mm pistondiameter and a swash plate angle

of 21 degree. The side force F RK

acting on a single piston is approxi-mately 5000 N when the pump runsat 45 MPa pressure.

Because of the nature of the basicoperation of piston machines, mostof the external forces change perio-dically with the rotating angle of theshaft. The periodically oscillatingexternal forces lead to a complexmicro motion of the piston, wherethe piston changes its inclinationangle in the cylinder bore till theresulting hydrodynamic pressure -

eld generates the required resultinguid force and moments to balancethe external forces. Fang and Shira-kashi (1995) rst time proposed a

Figure 3. Piston free body diagram

FTG

A

FRK

y

x

ump o e

ODP

ϕ

FSKz

FSKF

SKy

p

FDK

FTKs

FωK

F'ωK

FaK

FRKy

z

vK

0

y

A

F'ωK



4/13

377 Ventil 18 /2012/ 5

method to calculate the changinginclined positions of the piston du-ring one shaft revolution that would

balance the external piston forces.Kleist (1995) developed a methodto calculate the piston micro motionin order to balance external forces.Similar models were used by San-chen (1999), Olems (2001) and Wi-eczorek (2002). Figure 4 shows theresulting inclined piston position ofthe piston in the cylinder bore at agiven angular position of the shaft.

■ 2 Experimental work

Many researchers world wide builtspecial test rigs using different sim-plied versions of the piston cylin-der assembly or modied pumps(van der Kolk 1972, Yamaguchi1976, Dowd and Barwell 1975, Re-nius 1974, Regenbogen 1978, Hoo-ke and Kakoullis 1981, Ivantysyno-va 1983, Koehler 1984, Ezato andIkeya 1986, Fang and Shirakashi1995, Donders 1998, Manring 1999,Sanchen 1999, Tanaka 1999, Olems

2001, Ivantysynova and Lasaar 2000,Oberem 2002, Ivantysynova, Huang,and Behr 2005) to study the pistoncylinder interface. Most of the testrigs were based on a single pistondesign and inverse kinematics. Oneof the rst piston cylinder test rigswas built by van der Kolk (1972).The design did not allow for any axi-al piston motion. A side force wasapplied from outside. The arrange-ment simplied the conditions to a

tilted journal bearing with increasedoutside pressure. The test rig builtby Renius (1974) was also basedon a single piston assembly using

inverse kinematics. The design al-lowed for a more realistic study ofthe condition of the piston cylinder

interface by accounting for complexpiston motion (axial and spin) un-der varying pressures in the displa-cement chamber. Renius measuredpiston friction forces in axial and cir-cumferential direction. His extensiveexperimental work formed a majormilestone in the discovery of physi-

cal behavior of the piston cylinderinterface. The test rig used by Dowdand Barwell (1975) was proposedto study the impact of materialproperties and surface roughnesson friction and wear of the piston

cylinder interface. As shown in Fi- gure 5 the test rig was also basedon inverse kinematics. In additionto the missing centrifugal force alsothe impact of transient external sideload and piston spin motion on thehydrodynamic pressure built up areneglected in the experimental studyconducted by Dowd and Barwell.

Dowd and Barwell (1975) concludethat “the clearance between pistonand cylinder is extremely small andthe slope of the asperities is too sli-ght to provide sufcient hydrodyna-mic lift.” They also wrote in theirconclusion that elastohydrodynamiceffects might balance the side force.Yamaguchi (1976) built a single pi-ston test rig to study the piston mo-

Figure 4. Inclined position of the piston in the cylinder

Figure 5. Test rig used by Dowd and Barwell (1975)



5/13

378 Ventil 18 /2012/ 5

tion and uid lm conditions of pi-

ston machines. His test rig simulatesthe conditions in bent axis machinesusing a design with connecting rod.He measures gap heights betweenthe piston and cylinder under verylow pressures (maximum operatingpressure is 5 MPa). Several moreauthors built test rigs to measure

the piston friction force (Ezato and

Ikeya 1986, Donders 1998, Ma-nring 1999, Ivantysynova and Lasa-ar 2000). The test rig proposed andbuilt by Ivantysynova and Lasaar(2000) rst time allowed to measu-re piston friction forces on a swa-sh plate machine under machineconditions comparable to standarddesigns. The specially designedpump allows measurements of axialand circumferential friction force inpumping and motoring mode in awide range of operating conditions.

Figure 6 shows a cross section of thetest pump using a 3 component pi-ezoelectric force sensor mounted toa hydrostatically beard bushing.

The instantaneous cylinder pressureis measured with another piezoelec-tric pressure sensor. A telemetry isused to transfer the measured datawireless from the rotating cylinderblock to the data acquisition sy-stem. Results of a study of impact

of surface roughness as well asmeasurements using a half-barrellike piston shape on piston fricti-on force are summarized in Lasaar

(2002). Donders (1998) used a spe-

cial single piston test rig built at theInstitute for Fluid Power Drives andControl in Aachen to investigate thepiston friction force developmentunder different operating condi-tions using HFA uids and mineraloil. The test rig allows direct mea-surement of the piston friction for-ce using a force sensor mounted tothe piston. Also this test rig uses ahydraulic cylinder to apply a pistonside force. Oberem (2002) used thesame test rig later for a comprehen-

sive study of different piston desi-gns and the inuence of clearanceon the axial piston friction force.Ivantysynova (1983, 1985) used amodied swash plate type pumpwith inverse kinematics to measurethe temperature eld between pi-ston and cylinder. 30 thermocoupleswere implemented around a singlepiston. In addition the pressure inthe gap between piston a cylinderwas measured using 3 piezoelectric

pressure sensors. Figure 7 shows theimplemented thermocouples andthe pressure sensors.

Figure 6. Cross section of the test pump

Figure 7. Test pump with imple-mented thermocouples and pressu-re sensors



6/13

379 Ventil 18 /2012/ 5

Figure 8. Test rig for direct measurement of the pressure eld in the uid lm between piston and cylinder

The goal of the measurements con-ducted in pumping and motoringnode was to verify the developednon-isothermal gap ow model.Olems (2001) implemented 65 ther-mocouples around the cylinder su-rface and in the cylinder block tomeasure the surface temperaturebetween piston and cylinder and theheat transfer through the cylinder

block into the housing. The test rigshown in Figure 8 allows rst timedirect measurement of the pressureeld in the uid lm between pistonand cylinder.

The idea was experimentally to con-rm that elasto-hydrodynamic pres-sure built up is present in this inter-face. The test rig uses a single pistonand inverse kinematics. A speciallocking devise allows turning andxing the block in 180 predenedpositions. Therefore dynamic pres-sure can be measured at a grid oftotally 1620 measurement locationsaround the piston using 9 pressure

sensors only (Ivantysynova, Huangand Behr 2005). The test rig also al-lows measuring the temperature -eld between piston and cylinder ona ne grid of 1620 grid points using9 thermocouples and the describedlocking device.

In summary, the above briey de-scribed experimental work allowed

for major steps in the discovery ofwhat effects are involved and howthose effects contribute to the re-sulting piston motion and uid lmbehavior. Many of the test rigs werebuilt to support and conrm mode-ling of the piston cylinder interface.The next chapter will summarizemajor milestones in the area of mo-deling and simulation.

■ 3 Modeling of pistoncylinder interface

Gerber (1968) presented the rstmodel for the calculation of gapow and friction between piston

and cylinder. Gerber does not con-sider any hydrodynamic pressurebuilt up and proposes to use dif-ferent coefcients to calculate thefriction between piston and cylinder.Van der Kolk (1972) rst time calcu-lated the pressure eld between pi-ston and cylinder by solving the Re-ynolds equation. He neglected theaxial piston motion and calculated

the pressure eld based on the spinmotion of the piston and an inclinedpiston position, i.e. conditions simi-lar to a tilted journal bearing. Dueto limitations in the available com-puting power a very rough com-puting grid of just 48 elements wasused to describe the uid lm ow.Yamaguchi (1976) studied the moti-on of the piston in piston machinesbased on the hydrodynamic lubri-cation theory solving the Reynoldsequation. He considered alreadythe impact of change of gap heightwith respect to time due to chan-ging pressure in the displacementchamber. Based on his calculations



7/13

380 Ventil 18 /2012/ 5

he proposes a tapered piston shapeto stabilize the piston motion, i.e.to achieve force equilibrium betwe-en uid forces and external forces.Ivantysynova (1983, 1985) presen-ted a non-isothermal model, wherein addition to the Reynolds equa-

tion the energy equation is solvedin order to consider the impact ofchange of viscosity due to tempera-ture and pressure on gap ow con-ditions. The presented model solvesthe pressure and temperature eldbetween piston and cylinder for anassumed inclined position. The re-sulting uid forces and momentsare compared with external forces,but no additional iteration loop isadded to the simulation routine todetermine the micro motion of the

piston. Fang and Shirakashi (1995)rst time proposed a calculationmethod to determine the pistonposition considering force balancebetween external and uid forces byvarying the piston position. In thecase that no force balance can be fo-und they assume contact forces andmixed friction. Fang and Shirakashineglect the inuence of squeezelm effect on hydrodynamic pres-sure built up. Kleist (1995) presen-ted a new approach to calculate thepiston position within the cylinderbore based on the balance betweenexternal and uid forces and mo-ments. He considers also the micro

motion of the piston and its effect onhydrodynamic pressure built up. Kle-ist uses a modied Reynolds equa-tion, where the inuence of surface

roughness is consideredusing a statistical model,which was originally pro-posed by Patir and Cheng(1979). The model was la-ter integrated by Sanchen

(1999) in the simulationprogram PUMA. Olems(2002) used a non-isother-mal model together withNewton Raphson’s appro-ach to solve for the forcebalance between externaland uid forces and to de-termine the piston micromotion and resulting addi-tional load carrying abilityof the uid lm. Wieczorek(2002) integrated Olem’s

piston cylinder model intothe simulation programCASPAR. CASPAR calcula-tes all three connected lu-bricating gaps of a swash

plate type axial piston ma-chine based on non-isothermal gapow models and micro-motion ofmoveable parts.

Figure 10. Flow chart of implemented numerical algorithms

Figure 9. Piston cylinder coupled physical domains

Figure 10 shows the ow chart describing the implemented numerical al-gorithms.



8/13

381 Ventil 18 /2012/ 5

Figure 11. Comparison of measured and calculated piston friction forces

Figure 12. Comparison of measured and simulated temperature eld between piston and cylinder at

175 bar p∆ = , 1500n = , 51.8 CcaseT = ° , 45.8 C HP T = ° , and 43.1 C LP T = °

Figure 12 shows a comparison of measured and simulated temperature eld between piston and cylinder.

The investigation of elastohydrodyna-mic effects within the piston cylinderassembly was inspired from variousresearch in the eld of tribology, whe-re various procedures for solving thecomplex elastohydrodynamic lubri-cation problem were reported alrea-dy in the sixties. For example Dowsonand Higginson (1959) described aniterative procedure that solved the

elastohydrodynamic problem for linecontacts. Knoll (1974) studied theload carrying ability of journal bearin-gs under elastohydrodynamic lubri-cation. Dowson et al (1982) studiedthe elastohydrodynamic lubricationof piston rings in the combustionchamber of an internal combustion

engine. Also Reiners (1987) studiedelastohydrodynamic effects on pistoncylinder assembly of combustion en-gines under the consideration of pri-mary and secondary piston motion.

Ivantysynova and Huang (2002) pro-posed rst time to include an elasto--hydrodynamic model for the simu-lation of the gap ow between piston

and cylinder of a swash plate type axi-al piston machine. It is assumed thathigh pressure peaks developed in theuid lm lead to local elastic surfacedeformation of the piston and cylin-der. This will lead to a local change oflm thickness and therefore impactthe resulting pressure eld and all

other gap ow parameter includingfriction. The change of uid lm thic-kness due to elastic deformation ofpiston and cylinder was calculatedusing an inuence matrix approach.The inuence matrices were calcu-

lated using ANSYS. A similar pistoncylinder simulation model conside-ring elastohydrodynamic effects waspresented by Fatemi et al (2008). Thismodel coupled commercial FEM andmulti-body simulation software withan in-house uid lm model. A fullycoupled uid-structure thermal andmulti-body dynamics simulation mo-del for the piston cylinder interfacewas rst time published by Pelosi andIvantysynova (2009). The model doesnot only consider elastic deforma-tion of the piston and cylinder dueto pressure load, but also considersthermal expansion due to energydissipation in the uid lm. The pro-posed algorithm includes a completeheat transfer model of the cylinderblock and piston assembly. Figure 9 illustrates the numerical coupling ofdifferent physical domains conside-red in this new piston cylinder model.

The accuracy of gap ow predic-

tion of this new fully coupled u-id structure interaction model hasbeen veried through comparisonof measured and calculated pistonfriction forces, as shown in Figure11. The measurements were madeusing the tribo test rig developedby Lasaar (2003).



9/13

382 Ventil 18 /2012/ 5

The measurements were made usingthe test pump shown in Figure 8.More results have been published inPelosi and Ivantysynova (2011). Thenew model was further used to in-vestigate the deformation of pistonand cylinder due to pressure andtemperature for a standard swashplate unit with brass bushing pres-sed into the cylinder block madefrom steel. It was discovered that

the resulting elastic deformation ofthe brass bushing due to thermal adpressure load leads to a waviness ofthe surface. Simulation results showsurface deformations due to ther-mal loading and pressure of the pi-ston and bushing for a 75 cc swashplate type pump running at 35 MPadifferential pressure, 1500 rpm andmaximum swash plate angle. Portand case temperatures were takenfrom steady state measurements of

the same pump, i.e. inlet temperatu-re 49.7°C, temperature at high pres-sure port 54.6°C and case tempera-ture 65°C. As shown in Figure 13 the

elastic deforma-tion of the pistonand bushing leadto change of re-sulting gap heightin order of severalmicrons.

More important isthe generation of waviness in cir-cumferential direction of the bu-shing surface and the generationof a wedge in axial direction of thegap between piston and cylinder.Both the circumferential wavinessand the wedge effect will contributeto additional pressure built up andtherefore increased load-carryingability of the uid lm. This disco-

very is important for the understan-ding of physical phenomena con-tributing to uid lm conditions. Itwill also allow further optimizationof the piston cylinder assembly tofurther reduce energy dissipation

and increase power density of themachines.

■ 4 Design studies of thepiston- cylinder assembly

Dowd and Barwell wrote in theirpaper published in ASLE Transac-tions in 1974:” Conventional high--pressure hydraulic pumps require adelicate balance between clearance,surface nish, and pump materials if

a long, efcient life is to be maintai-ned.” Thanks to the research effortsof so many researchers over the last40 years we now know a little moreabout the important contributionof different physical phenomena toload carrying ability of the uid lm,but still need to continue to studythe sensitivity of the different de-sign options to the resulting uidlm behavior. In this chapter the re-search activities related to study theimpact of piston and cylinder shapeas well as material options will bediscussed. Figure 14 shows severalstandard piston designs, which havebeen and are still used by many ma-nufactures worldwide.

Whereas piston # 1 is too heavy andwill minimize pump speed, piston 2introduces too high compressionloss due to high dead volume. De-pendent on the material used insi-

Figure 14. Current standard piston designs

Figure 15. Piston shape proposed by Lasaar and Ivantysynova

Figure 13. Change of resulting gap height



10/13

383 Ventil 18 /2012/ 5

de the hollow piston piston #4, thepiston will have similar problemslike piston # 1 or #2. The ring gro-oves on the outer surface of piston# 5 are clearly destroying the hy-drodynamic pressure built up in the

groove area and are therefore nothelping to increase load carryingability of the piston, i.e. such de-sign should not be used at all. Thebest design considering a cylindri-cal piston shape represents piston# 3. However when studying thegoverning equation describing theuid lm behavior, it is obvious thatshaping of the piston and or cylin-der is a very helpful design elementto increase load carrying ability ofthe piston cylinder interface, redu-ce energy dissipation and to avoidwear and mixed lubrication. Yama-guchi (1976) rst time proposed atapered piston shape to utilize animproved hydrodynamic pressurebuilt up with axial piston motion.Ivantysynova (1983) proposed abarrel like piston with a large cur-vature radius leading to a diameterreduction of 4 micrometer at bothpiston ends. The idea of optimi-zing the piston shape to minimize

energy dissipation within the pistoncylinder interface was continued byLasaar (2003), who proposed a halfbarrel like piston as shown in Figure15.

Kleist (1997) proposed and testeddifferent shaped pistons for radialpiston machines. The shaped pistonshowed lower friction and compa-rable leakage than the cylindricalpiston. Murrenhoff et al (2010) pre-

sented several research works onshaping of the piston and bushingtogether with piston and bushingcoatings as well as rst investiga-tions into surface texturing. Thepaper reports reduction of pistonfriction force for coated and sha-ped pistons, however the presentedmeasurement results show impro-vements only for the low pressu-re/suction stroke. These results arevery different from the results Lasa-ar obtained for his shaped piston,

where reduction of axial frictionforce was obtained especially du-ring the high-pressure stroke. Thenewest version of the fully coupled

uid structure interaction model ofthe piston cylinder interface deve-loped by Pelosi and Ivantysynova(2011) was used to conduct a simu-lation study for waved pistons. A USpatent application has been led forthe waved piston shown in Figure16 . Recent simulation results havedemonstrated a huge potential toachieve major reduction of energydissipation for the piston cylinderinterface. Table 1 shows a compari-son of average reduction of powerloss achieved for all eight simula-ted operating conditions for threedifferent waved piston designs. Thesimulations were made for pumpoperation at high and low speed(1000 and 3000 rpm), high and lowpressure (400bar and 100 bar) and

100% and 20% swash plate angle.The waved design with 2.5 sin wavesand an amplitude of +/- 6 microme-ter showed reduction of power lossfor all operating conditions with thehighest of 80% at simulated opera-ting point 4 (3000 rpm, 400 bar and20% swash plate angle). These veryrecently obtained research results

have not been published elsewhere.Besides shaping and surface struc-turing several researchers haveinvestigated the use of new ma-terials and coatings for the piston

and cylinder. Feldmann and Bartelt( 2003) reported results of testingof ceramic pistons and bushingsin a swash plate type axial piston

machine. After 600h testing atmaximum pressure of 320 bar andspeed of 2000 rpm piston and bu-shing showed no wear , but the ove-rall efciency of the unit was lowerthan the standard one. Another unit

equipped with ceramic pistons fai-led after short operation. Scharfand Murrenhoff (2004) concludedafter they conducted a compre-hensive study on different physicalvapor deposition (pvd) coatings ofdifferent pump parts that the pistonreveals the weak point of ZirconiumCarbid coating because the pistoncylinder interface represents themost complex interface within thedisplacement unit.

The main idea of most of the pre-vious research on new materialcombinations and coatings for thepiston cylinder assembly was toprevent wear and reduce friction.The newest nding of the behaviorof the brass bushing pressed in tothe steel cylinder block reported inchapter 3, reveal that the piston andcylinder material needs to supportthe load carrying ability of the uidlm. This is the only successful way

of increasing load carrying abilityof the uid lm, which will result inreduction of energy dissipation andfriction and with that also avoidwear. Dowd and Barwell (1974) con-cluded in their paper that the sideload of the piston must be balan-ced by hydrodynamic lift generatedby microaspirities and elasto-hy-

drodynamic contact between thetwo surfaces. Our newest ndingsconrm that elasto-hydrodynamiceffects play a major role and helpcurrent designs to run successful.

This discovery however opens thedoor to interesting new research ininvestigating new materials, coatin-gs and shapes for this heavy loaded

Figure 16. Waved Piston

Table 1. Average reduction in power loss due to piston shaping

Design Clearance[µm]

% Reduction

in Power LossLeakage

% Reduction

in Power LossFriction

% Reduction

in Total PowerLoss

6µm-2 peaks 34µm 8.65% 8.3% 8.35%

6µm-2 peaks 18µm 80% 61% 68%

6µm-2.5 peaks 18µm 78.8% 56.8% 64.7%



11/13

384 Ventil 18 /2012/ 5

interface.References

[1] Brangs, E. 1965. Ueber die

Auslegung von Axialkolben-

pumpen mit ebenem Steuer-

spiegel. Dissertation RWTHAachen, Germany.

[2] Dowd, J. R. and Barwell, F. T.

1975. Tribological interacti-

on between piston and cylin-

der of a model high pressure

pump. Transact. ASLE, 18, pp.

21–30.

[3] Donders, St. 1998. Kolbenma-

schinen für HFA Flüssigke-

iten - Verlustanteile einer

Schrägscheibeneinheit. Dis-sertation RWTH Aachen, Ger-

many.

[4] Donders, St., Kane, B. and Se-

ifert, V. 2004. Einsatz von

Kunststoffen und Keramik in

Hydraulikkomponenten. Proc.

of the 4th International Fluid

Power Conference, Dresden,

pp. 453–460.

[5] Ivantysynova, M. (at that time

Berge, M.) 1983. An investi-

gation of viscous ow in lu-bricating gaps. (In Slovak).

Dissertation SVST Bratislava,

Czechoslovakia.

[6] Ivantysyn, J. and Ivantysynova,

M. 1993. Hydrostatische Pum-

pen und Motoren. Wurzburg.

Vogel Buchverlag.

[7] Ivantysynova, M. 1985. Tem-

peraturfeld im Schmierspalt

zwischen Kolben und Zylinder

einer Axialkolbenmaschine.Maschinenbautechnik 34, pp.

532–535.

[8] Ivantysynova, M. and Lasaar, R.

2000. Ein Versuchsträger zur

Messung der Reibkräfte zwi-

schen Kolben und Zylinder in

Axialkolbenmaschinen. Kon-

struktion 52 (2000) 6, pp. 57

- 65.

[9] Ivantysynova, M. 1985. Tem-

peraturfeld im Schmierspalt

zwischen Kolben und Zylindereiner Axialkolbenmaschine.

Maschinenbautechnik 34, pp.

532–535.

[10] Ivantysynova, M. and Huang,

Ch. 2002. Investigation of the

gap ow in displacement ma-

chines considering the elasto-

hydrodynamic effect. 5th JFPS

International Symposium on

Fluid Power. Nara, Japan. pp.219-229.

[11] Ivantysynova, M. and Lasaar,

R. 2004. An investigation into

Micro- and macro geome-

tric design of piston/cylinder

assembly of swash plate ma-

chines. International Journal

of Fluid Power, Vol. 5 (2004),

No.1, pp. 23-36.

[12] Ivantysynova, M., Huang, C.

and Behr, R. 2005. Measure-ments of elastohydro-dynamic

pressure eld in the gap bet-

ween piston and cylinder. Bath

Workshop on Power Transmis-

sion and Motion Control PTMC

2005, Bath, UK, pp. 451 - 465.

[13] Ivantysynova, M. 2011. Positive

displacement machine piston

with wavy surface form. US

patent application. Serial No.

13/260,649.

[14] Ezato, M. and Ikeya, M. 1986.Sliding friction characteristics

between piston and cylinder.

Starting and low speed condi-

tions in the swash plate type

axial piston motor. 7th Interna-

tional Fluid Power Symposium,

Bath, UK.

[15] Fang, Y and Shirakashi, M.

1995. Mixed lubrication cha-

racteristics between piston

and cylinder in hydraulic pi-ston pump-motor. Journal

of Tribology, Trans. ASME ,

Vol.117, pp. 80-85.

[16] Fatemi, A., Wohlers, A. and

Murrenhoff, H. 2008. Simula-

tion of Elastohydrodynamic

Contact between Piston and

Cylinder in Axial Piston Pumps.

Proc. of the 6th Internatio-

nal Fluid Power Conference,

Dresden, pp. 539–552.

[17] Feldmann,D.G. and Bartelt, Chr.2003. Development of ceramic

pistons and bushings for axial

piston units. 8th Scandinavian

Internatinal Conference on

Fluid Power, SICFP’03, Tampe-

re, Finland, pp. 1289-1300.

[18] Harris, M.R., Edge, K.A. and

Tilley, D.G. 1993. The spin

motion of pistons in a swash

plate type axial piston pump.3rd Scandinavian International

Conference on Fluid Power ,

Finland.

[19] Hooke, C.J. and Kakoullis, Y.P.

1981. The effects of centrifu-

gal load and ball friction on

the lubrication of slippers in

axial piston pumps. 6th Fluid

Power Symposium, Cambridge

ISBN 0906085, pp. 179-191.

[20] Kleist, A. 1997. Design of Hy-drostatic Bearing and Sealing

Gaps in Hydraulic Machines

– a new Simulation Tool. Fifth

Scandinavian International

Conference on Fluid Power.

SICFP”97, Linkoeping, Sweden,

pp. 157-169.

[21] Kleist, A. 2002. Berechnung

von Dicht-und Lagerfugen in

hydrostatischen Maschinen.

Dissertation RWTH Aachen,

Germany.[22] Knoll, G. 1974. Tragfähigkeit

zylindrischer unter elastohy-

drodynamischen Bedingun-

gen. Dissertation. RWTH Aa-

chen 1974, Germany.

[23] Knoll, G.D. and Peeken, H.J.

1982. Hydrodynamic lubricati-

on of piston skirts. Transaction

of ASME Journal of Lubrication

technology. Vol. 104, pp. 504-

509.[24] Knoll, G. and Peeken, H. 1990.

Formulation of uid contact

elements for elstohydrodyna-

mic structure interaction. Proc.

Japanese International Tribo-

logy Conference, Nagoya, pp.

179-184.

[25] Lasaar, R. 2003. Eine Untersu-

chung zur mikro-und makro-

geometrischen Gestaltung der

Kolben/ Zylinderbaugruppe

von Schrägscheibenmaschi-nen. VDI Fortschrittsberichte

Reihe 1 No. 364, VDI Verlag

Düsseldorf, Germany.



12/13

385 Ventil 18 /2012/ 5

[26] Murrenhoff, H. et al. 2010. Ef-

ciency improvements of Fluid

Power Components Focusing

on Tribological Systems. 7th

International Fluid Power Con-

ference, Aachen, Germany, pp.

215-248.[27] Manring, N. 1999. Friction for-

ces within the cylinder bores

of swash plate type axial pi-

ston pumps and motors. ASME

Journal of Dynamic systems,

Measurement and Control, Vol.

121, pp. 531-537.

[28] Oberem, R. 2002. Untersuc-

hung der Tribosysteme von

Schrägscheibenmaschinen der

HFA-Hydraulik. Dissertation.RWTH Aachen, Germany.

[29] Olems, L. 2001. Ein Beitrag

zur Bestimmung des Tem-

peraturverhaltens der Kol-

ben-Zylinderbaugruppe von

Axialkolbenmaschinen der

Schrägscheibenbau-weise. VDI

Fortschrittsberichte Reihe 1

No. 348, VDI Verlag Düssel-

dorf, Germany.

[30] Olems L., 2000. Investigation of

the Temperature Behaviour ofthe Piston Cylinder Assembly

in Axial Piston Pumps. Interna-

tional Journal of Fluid Power .

[31] Patir, N. and Cheng, H.S. 1978.

An average ow model for de-

termining effects of three di-

mensional roughness on par-

tial hydrodynamiclubrication.

ASME Journal of Lubrication

Technology . Vol. 100, pp. 12-

17.[32] Pelosi, M. and Ivantysynova,

M. 2009. A Novel Thermal

Model for the Piston/Cylinder

Interface of Piston Machines.

Bath ASME Symposium on Flu-

id Power and Motion Control

(FPMC2009), [DSCC2009-2782

[22]

[33] Pelosi, M. and Ivantysynova, M.

2011. The Inuence of Pressu-

re and Thermal Deformation

on the Piston/Cylinder Inter-face Film Thickness. 52nd IFPE

Conference on Fluid Power , Las

Vegas, USA.


2009. A Novel Fluid-Structure

Interaction Model for Lubrica-

ting Gaps of Piston Machines.

Proc. of the 5th Fluid Structure

Interaction Conference, Crete,

pp. 13-24.[35] Pelosi, M. and Ivantysynova, M.

2010. A Simulation Study on

the Impact of Material Proper-

ties on Piston/Cylinder Lubri-

cating Gap Performance. Proc.

of the 6th FPNI PhD Symposi-

um, West Lafayette, IN, USA.


2011. The Inuence of Pressu-

re and Thermal Deformation

on the Piston/Cylinder Interfa-ce Film Thickness. Proceedin-

gs of the 52nd National Con-

ference on Fluid Power 2011,

NCFP I11-9.3.


2011. Surface Deformations

Enable High Pressure Operati-

on of Axial Piston Pumps. Bath

ASME Symposium on Fluid

Power and Motion Control

(FPMC2009), [DSCC2011], Wa-

shington, USA.[38] Reiners, W. 1987. Die Ausbil-

dung elastohydrodynamischer

Reaktionskräfte am Kolben-

schaft unter Beruecksichti-

gung der Prim är- und Sekun-

där-bewegung. Dissertation

RWTH Aachen, Germany.

[39] Renius, K.T. 1974. Untersuchung

zur Reibung zwischen Kolben

und Zylinder bei Schrägschei-

ben-Axialkolbenmaschinen. VDI Forschungsheft 561. VDI

Verlag Düsseldorf, Germany.

[40] Regenbogen, H. 1978. Das Re-

ibungsverhalten von Kolben

und Zylinder bei Schrägschei-

ben- Axialkolbenmaschinen .

VDI Forschungsheft 590, VDI

Verlag Düsseldorf, Germany.

[41] Rinck, St. 1992. Untersuchung

und Optimierung schnellau-

fender Axial- und Radial-kol-

benpumpn beim Betrieb mitwasserbasischen Druckues-

sigkeiten. Dissertation RWTH

Aachen, Germany.

[42] Tanaka K., Kyogoku K. and

Nakahara, T. 1999. Lubricati-

on characteristics on sliding

surfaces between piston and

cylinder in a piston pump and

motor (Effects of running-in,

prole of piston top and stiff -ness). JSME International Jour-

nal, Series C (Mechanical Sy-

stems, Machine Elements and

Manufacturing), Vol. 42 (1999)

No. 4, pp. 1031-1040.

[43] Van Bebber, D. 2003. PVD Schi-

chten in Verdrängereinheiten

zur Verschleiß- und Reibun-

gsminimierung bei Betrieb mit

synthetischen Estern. Disserta-

tion RWTH Aachen, Germany.[44] Wieczorek, U. and Ivantysyno-

va, M. 2002. Computer Aided

Optimization of Bearing and

Sealing Gaps in Hydrostatic

Machines - The Simulation

Tool CASPAR. International

Journal of Fluid Power , Vol. 3

(2002), No.1, pp. 7-20.

[45] Wieczorek, U. 2002. Ein Si-

mulationsmodell zur Beschre-

ibung der Spaltströmung in

Axialkolbenmaschinen derSchrägscheibenbauart. VDI

Fortschrittsberichte Reihe 1

No. 443, VDI Verlag Düssel-

dorf, Germany.

[46] Yamaguchi, A. 1976. Motion of

Pistons in Piston-Type Hydrau-

lic Machines.Bulletin of JSME,

Vol. 19, No. 130, pp. 402-419.

[47] Yamaguchi, A. 1990. Motion

of the piston in piston pumps

and motors: the case of metal-lic contact. JSME International

Journal, Series III, 33(4), 627-

633, Tokyo.

[48] Zhang, Y. 2000. Verbesse-

rung des Anlauf- und Lan-

gsamlaufverhaltens ei-

nes Axialkolbenmotors in

S chrägsche ibenbauwe i se

durch konstruktive und mate-

rial-ltechnische Maßnahmen.

Dissertation RWTH Aachen,

Germany.



13/13

386 Ventil 18 /2012/ 5

Bat v izvrtini znotraj hidrostatičnih batnihpogonov – dolga pot odkrivanja

Razširjeni povzetek

Prispevek prikazuje pomembnejše rezultate raziskovalcev in inženirjevpri odkritjih zikalnih zakonitosti, povezanih z lastnostmi lma kapljevinein delovnih pogojev bata v izvrtini znotraj hidrostatičnih batnih pogo-nov. Osredotočen je na konstrukcijsko izvedbo bata v izvrtini za primerustvarjanja potrebnega momenta za premagovanje večjih bočnih sil nabat. Od leta 1965 je bilo izvedenih preko 20 različnih doktorskih diser-tacij, ki so obravnavale teoretične in/ali eksperimentalne študije mejnihkontaktnih ploskev med batom in izvrtino znotraj hidrostatičnih batnihpogonov. Prispevek prikazuje le glavne mejnike v razvoju hidravlično--tribološkega para, bata v izvrtini.

Slika 1 prikazuje mazalne reže znotraj aksialnih in radialnih batnih črpalk.Slika 2 prikazuje osnovne kinematične razmere za primer aksialne batne

črpalke z nagibno ploščo. Slika 3 prikazuje diagram sil na bat v izvrtiniznotraj aksialne batne črpalke. Slika 4 prikazuje nagnjen položaj bata vizvrtini pri določenem položaju gredi. Na sliki 5 je prikazano eno prvihtovrstnih preizkuševališč, namenjeno raziskavam vplivov različnih mate-rialov in površinskih hrapavosti na trenje in obrabo bata v izvrtini. Preiz-kuševališče sta leta 1975 razvila Dowd in Barvell. Prerez testirane črpalkes tremi piezomerilniki sile je prikazan na sliki 6. Testirana črpalka z inte-griranimi termočleni in tlačnimi zaznavali je prikazana na sliki 7 . Na sliki 8

je preizkuševališče za neposredno merjenje tlačnih polj in lma kapljevi-ne med batom in izvrtino. Numerično združevanje različnih zikalnih do-men za nov numerični model sestava bata v izvrtini je prikazano na sliki9. Blokovni diagram izboljšanega numeričnega algoritma za preračun

razmer med batom in izvrtino kaže slika 10. Slika 11 prikazuje primerjavosile trenja bata v izvrtini med rezultati meritev in numeričnega izračuna.Na sliki 12 je prikazana primerjava med izmerjenimi in simuliranimi tem-peraturnimi polji med batom in izvrtino aksialne batne črpalke. Elastičnadeformacija bata in puše vpliva na spremembo višine reže med njima(slika 13). Na sliki 14 so prikazane obstoječe oblike standardnih batov.Slika 15 prikazuje novo, za 4 mikrometre po premeru, sodčkasto oblikobata. Slika 16 prikazuje patentiran bat z valovito površino. Preglednica 1 prikazuje nekaj vplivov na povprečno zmanjšanje izgub za osem različnihsimuliranih delovnih pogojev in za tri različno valovite površina bata. Si-mulacije so bile izvedene za črpalko pri visoki in nizki vrtilni hitrosti (3000vrt./min. in 1000 vrt./min.), pri visokem in nizkem delovnem tlaku (400bar in 100 bar) in pri 100-odstotnem ter 20-odstotnem kotu nagibneplošče. Bat s površino, ki je imela 2,5 sinusnih valov, z amplitudo +/–6mikrometrov, je pokazal največje zmanjšanje izgub (80 % zmanjšanje pri3000 vrt./min, 400 bar in pri 20-odstotnem kotu nagibne plošče).

Poleg naštetih raziskav vpliva različnih oblikovnih in površinskih izvedbdrsne površine bata so se nekateri raziskovalci osredotočili tudi na ma -teriale in prevleke batov in izvrtin. Glavni namen omenjenih raziskav, po-vezanih z materiali in prevlekami, je zmanjšanje trenja in obrabe bata vpuši aksialne batne črpalke. Rezultati omenjenih raziskav odpirajo vrataraziskavam novih materialov, prevlek in oblik drsnih površin težko obre-menjenih mejnih ploskev batnih črpalk in motorjev.

Ključne besede: hidrostatični batni pogoni, lastnosti mazalnega lma, teo-retične in eksperimentalne študije, mazalna reža, nagibna plošča, kinemati-ka bata, sile, preizkuševališče, numerični model, meritve, oblika bata


Piston-Cylinder Aseembly

Documents

Transcript of Piston-Cylinder Aseembly