• Non ci sono risultati.

Parametric design, fabrication and validation of one-way polymeric valves for artificial sphincters

N/A
N/A
Protected

Academic year: 2021

Condividi "Parametric design, fabrication and validation of one-way polymeric valves for artificial sphincters"

Copied!
11
0
0

Testo completo

(1)

ContentslistsavailableatScienceDirect

Sensors

and

Actuators

A:

Physical

jo u r n al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / s n a

Parametric

design,

fabrication

and

validation

of

one-way

polymeric

valves

for

artificial

sphincters

Tommaso

Mazzocchi

a,∗

,

Leonardo

Ricotti

a

,

Novello

Pinzi

b

,

Arianna

Menciassi

a aTheBioRoboticsInstitute,ScuolaSuperioreSant’Anna,VialeR.Piaggio,34,56025,Pontedera,PI,Italy

bDipartimentodiUrologia,UniversitàdegliStudidiSiena,ViaBanchidiSotto,55,53100Siena,SI,Italy

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received5January2015

Receivedinrevisedform22June2015 Accepted7July2015

Availableonline15July2015

a

b

s

t

r

a

c

t

Thedesignofartificialsphinctersrequiresanaccuratedimensioningofdedicatedvalves,normallymade

ofpolymericmaterials.Thiseffortisalsointerestingfordevelopingfluidandpressureregulatingsolutions

relatedtootherbiomedicalandnon-biomedicalfields.Inthisarticlewefocusedontheparametricdesign

ofpolymericvalves,bytakinginspirationfromcommerciallyexploitedsolutionsusedinthefoodindustry

andperformingappropriatescalinginordertomakethemsuitableforartificialorgansandcomponents.

Inaddition,differentmaterialswithdiversemechanicalpropertieswereconsidered,focusingona

low-costfabricationapproach.Finiteelementmodelanalyseswereconductedtosimulatethebehaviorof

differentvalveprofilesandtopredictthevalveopeningpressure.Simulationresultswerevalidatedby

comparingthemwithexperimentalresults,obtainedbyfabricatingandtestingdifferentvalvetypes.This

polymericvalveparametricanalysismaybeexploitedforthedesignofartificialsphincters,havingthe

potentialtotackleurinaryincontinence,adiseasethataffectsabout350millionpeopleworldwide.

©2015ElsevierB.V.Allrightsreserved.

1. Introduction

Approximately350millionindividualsworldwideexperienced

urinaryincontinence (UI) in2008 andthis number is expected

toincrease upto423millionby 2018[1].UI isa disorder that

produces an involuntary urine loss caused by an uncontrolled

overactivebladderorapelviperinealmuscledeficitthatcanarise

as a surgery complication in male subjects (e.g. after prostate

surgery), while in female subjects it may occur after

child-birthorafterurogenitalorganssurgery.Inhospitals,UIdiseases

arenormallymanagedbyusingabsorbent materialsorurethral

cathetersprovided witha bag (tiedtothepatient’s leg).

Extra-luminal sphincters (ES) are devices able to control fluid flow

throughan elasticduct, by modulating thecompressionof the

elastic duct wall itself. They need to be implanted via.

surgi-cal procedures,so theyare rather invasive. The mostcommon

ESthat are commercially available totackle UI are the ProAct

system(Uromedica,MN,USA)andtheAMS800(American

Med-icalSystems, Minnetonka, MN, USA). These solutions require a

surgicalprocedurefortheirinstallationand theyareonly

avail-ableformales[2,3].ProActandAMS800areexpensiveandcan

lead to complications such as erosion, infection and

mechani-∗ Correspondingauthor.Fax:+3950883101. E-mailaddress:t.mazzocchi@sssup.it(T.Mazzocchi).

cal failure. Therefore, periodic revisions(rather complicated to

be performed) are mandatory. In addition, if replacement is

needed, a second invasive surgicalprocedure needstobe

per-formed.

Ontheotherhand,intraluminalsphincters(IS)donotrequire

asurgeryproceduretobeimplanted:theyareusuallyinstalledby

usingtheductorificeandbroughttothesiteofinterestthrough

apushingcatheter.Atpresent,few commercialISareavailable,

suchastheFemSoft (RochesterMedical Corporation,MN,USA)

ortheInFlowdevice(Vesiflo,Inc,WA,USA)[4].ISarenot

com-pletelyinternaldevices:asmallportionofthesystemisinevitably

visible, thus raising psychological discomfort in certain

situa-tions.

Recently,advancedandrathercomplexmechatronicsolutions

havebeenproposedtotackleUI.Lamraouietal.reportedthedesign,

developmentand in vivotesting ongoats of a 60cm3 artificial

sphincter(AS)prototypecomposedof standardmechanical and

electricalcomponents[5].Morerecently,aretro-compatibledevice

wasproposedbyHachedandcolleagues.Thesystemwasbased

onahydraulicmoduleandacontrolunitandshowedgood

effi-cacyinbothbenchtestsandexvivotests[6].Thesameauthors

alsorecentlydevelopedawirelesslycontrolledandadaptiveAS,

capableofpostoperativeocclusivecuffpressuremodification[7].

Alltheabove-mentionedmechatronicsolutionsareratherinvasive

andtheirlong-termsafety andefficacystill havetobe

demon-strated.

http://dx.doi.org/10.1016/j.sna.2015.07.005

(2)

Fig.1.Commercialvalveusedinmanywaterbottles.Thewholestructureandasection,arereported(A),togetherwithitsparameterization(B).Thevalveaxisandtheload distributionareshowninCandD,respectively.

Theauthorsalsorecentlyhighlightedapossiblenewdesignfor

IS,basedonmagneticallyresponsivemechanismsabletoengagea

unidirectionalvalveandthusmodulateitsstiffness[8].Atpresent,

aUIsolutionsuitableforbothsexesfeaturinghighstabilityandlow

invasiveness(withoutalteringthebodyscheme)isnotavailable.

Passiveunidirectionalpolymericvalvesareakeycomponentof

devicesdesignedtotackleUIdiseases.Theyareusefultocontrol

theliquidpressurebetweentwocompartments,thusensuringa

unidirectionalflowwhenacertainpressurethresholdisovercome.

Biocompatiblepolymersareobviouslyneededandindependence

fromenergysourceswouldreducetheintrinsichazardsrelatedto

theuseofsuchsystems.

The achievement of an artificial IS suitable for both sexes

requires an accurate dimensioning of ad hoc one-way passive

valves.Tothispurpose,theanalysisofavailablesolutionsforother

biomedicalor non-biomedicalfields canprovidesome

interest-inghints.Manyliteraturestudieshavefocusedonthecorrelation

betweengeometryandvalveperformance,highlightingthe

impor-tanceofnumericalsimulationssuchasnon-linearfiniteelement

models(FEM)inordertooptimizethedesignand thesizing of

valveprofiles.Inthebiomedicalfield,thisapproachhasbeen

exten-sivelyusedtodesign heart valves.Differentvalve profileshave

beenreviewed by Mohammadiand Mequanint [9],in terms of

bothmodellinganddesignstrategies.Recently,Burriesciand

col-leaguesfocused ona noveldesign strategy for polymericheart

valves(supportedbynumericalsimulations),inordertoincrease

valveperformancesandreducestresslevels[10].Another

inter-estingstudywascarriedoutbyLabrosseetal.,whichdevelopeda

geometricmodelofanaortictrileafletvalve,forthepurposeof

eval-uatinghowmuchthedimensionsoftheaorticvalvecomponents

couldbevaried,whilestillmaintainingpropertargetperformances

[11]. More recent studiesfocused onapplyingFEM analysesto

properlydesignleafletmaterialsandtopredictthehydrodynamic

behaviorofnanocomposite-basedpolymericaorticvalves[12,13].

However,theaforementionedstudiescannotbeeasilygeneralized

andappliedtothedesignofaurinarysystemvalve.Thisismainly

duetotherathercomplicateddesignandmanufacturingprocesses,

neededtobuildheartvalveswhichrequireseveralstepsinorder

tocombineametallicstentwiththepolymericleaflets.

A differentapproach mustbe adopted when facingscalable

manufacturinginthepresenceofhighpressurerangesandsmall

flowrates(␮l/min).Althoughtheseconditionsarenottypicalof

theurinarysystem,thisapproachhasbeenwidelyadoptedforthe

designofmicro-valvestobeusedinmicrofluidicsystems.The

geo-metricalparametersofthevalve,includingshape,fluidchannel

widthandmembranethicknessareusuallyexamined.Snakenborg

et al. designed a polydimethylsiloxane (PDMS) membrane

pro-videdwithanincisioninordertofabricateacheckvalveableto

controltheflowrate[14].Mohanetal.reportedinteresting

consid-erationsconcerningthedesignandtheapplicationofelastomeric

microvalves,focusingonactuationpressureminimization,reliable

operationandconvenient integrationinto complexmicrofluidic

devices[15].Jahanshahietal.focusedonthedesignofamicro-valve

characterizedbyahighpressurerangeandarelativelyhigh

seal-ingcapability(noflowupto0.6MPareversepressure)[16].Hilbert

andcolleaguesintroducedaninnovativemicro-valve

manufactur-ingtechniquebyusingPDMSloadedwithneodymiumparticles,to

adjusttheswitchingpointofeachsinglevalve[17].

Hickersonand colleaguesdeveloped a valve design strategy

(3)

tech-niquesthatallowthedevelopmentoflow-costdisposablesystems.

Thereportedresultsmaybeexploitedformanyapplicationsthat

requirespecificvalveperformance.However,theyarenotusableas

componentsforartificialurinarysphincters,duetothehigh

pres-surerange(12kPa–120kPa)andtothesmallflowrates[18].

Uohashiand colleaguestookintoaccount differentone-way

valve profiles, achievable by exploiting standard molding

pro-cesses.Theyensuredhighflowrates(ml/min).Interestingresults,

concerning the influence of material stiffness on valve

perfor-mance,werereported.However,anaccuraterelationshipbetween

valve geometrical parameters and performance was not

high-lighted[19].

Inthisarticle,wediscusstheparametricdesign,fabricationand

validationofaone-waypolymericvalveascomponentofAS,

tak-ingasamodelacommonvalveprofileavailableinwaterbottles.

Low-costfabricationtechniquesandabiocompatibleand

stiffness-tunablematerialwereconsideredforvalvedesignandfabrication.

Simulationresultsevidencedthemostsuitableprofilesin

corre-spondencetocertaindesiredopeningpressures.Forapplicationin

theurinarysystem,hermeticclosureshouldbeassured:sudden

coughsandinvoluntaryabdominalmusclecontractionscan

gener-atepressuresupto12kPa.However,thisphysiologicalparameter

dependsonthepatient’sanatomicalfeatures,thereforeatunable

openingpressure(OP)isdesirable[20–22].Ashighlightedbelow,

thiscanbeachievedbymakingsmallchangestothevalveprofile

geometryorbyslightlychangingmaterialstiffness.

2. Experimental

2.1. Valveprofileanditsparameterization

Toselectthevalveprofile,wetookinspirationfromthefood

industry,especiallyfromthevalvesintegratedinmanywater

bot-tles.Normally,theyareusedtoguaranteehermeticsealingupto

acertainpressure,afeaturethatmakesiteasiertodrinkliquids

duringsportactivities,forexample.Weconsideredacommercial

valvethathadavaultshapeandasurfacefeaturingasix-pointed

stardiecut.Thenotchesallowthevalvetoopenwhenthepressure

appliedonthetopsurfaceisaboveacertainthreshold.Weanalyzed

thevalveprofileandextracted9parametersabletodescribeit,as

showninFig.1.

The nine parameters that allow reproduction of the

com-mercial valve were: P1c=3.7mm, P2c=50◦, P3c=0.20mm,

P4c=83◦, P5c=1.28mm, P6c=0.37mm, P7c=97◦, P8c=1.00mm

andP9c=1.70mm(whereC=commercial).

We aimedat fabricatingone-way valveswitha specificOP,

bytaking into accountthe fabricationconstraints. Theanalysis

ofthenineparametersledtoaredundantproblem.However,all

parameterswerevaried,inordertohighlighthowtheirdifferent

valuesinfluencedoverallvalveperformance.Foreachparameter,

fiveequidistantvalueswereidentified(Table1).Moreover,they

werecombinedwithsixdifferenttypesofmaterialstiffness.

Thefirstanalysisaimedat assessinghowvalve performance

changedwhenapplyinganisotropicscaling(increasingor

decreas-ingvalvesizebyascalefactorthatwasthesameinalldirections).

Tothisaim,P1wasconsideredbecauseitisrelatedtotheradiusof

thevaultvalve.Avalveradiussuitableforurologicalapplications

mustbesmallerthan6mm[23].Thescalingfactors(SF)analyzed

inthisworkwere0.35,0.67,1.00,1.32and1.64.Theycorresponded

toP1valuesof1.3,2.5,3.7,4.9and6.1mm,respectively.Differently

scaledvalvesmayfindapplicationsindifferentfields.

Anotherimportantaspectoftheanalysisconcernedhowvalve

performancechangedinfunctionoftheothereightparameters,

onceacertainvalvesizewasselected.Thisallowedustoidentify

themostrelevantparametersforvalveperformance.

2.2. FiniteElementModel(FEM)simulations

Matlab®R2013a(MathWorks)andAbaqus6.13(Dassault

Sys-temes)wereusedtomanagedataandtoperformFEMsimulations,

respectively.

Matlab®environmentwasusedtomanagethedifferent

param-eters. Once the desired valve profile was generated, Matlab®

generatedanappropriatefile,readablebyAbaqus,whichloaded

thedesiredvalvegeometry,fordetectingtheOPoftherespective

profile.Thisallowedustohandlealargenumberofsimulationsin

asimplemanner.ThevalvemodelwasimplementedusingPython

language.

WhenAbaqushadprocessedthesimulation,itgeneratedafile

inwhichalltheresultswerereported.Analgorithmwas

devel-opedtocompareeachsimulationandtoidentifythevalveOP.The

algorithmtrackedalltheextremepoints(A8andA7inFig.1)for

eachvalvestrip:whenthedistancebetweensuchextremepoints

wasdoublethanthevalvethickness(P3),thevalvewasconsidered

open.Oncethesimulationwascompleted,theextremepointsof

thevalvestripreachedduringthesimulationwerestoredinafile,

fromwhichMatlab®wasabletoindentifythevalveOP.

AftercreatingthevalveintheAbaqusenvironment,

experimen-taldata(stress-straincurves)ofthematerialusedwereimportedby

exploitingtheHyperplasticmodulesupportedbytheprogram.In

theAbaqusStepmodule,anexplicitdynamicbehaviorwasdefined,

byimposingamassscalingequalto100.Appropriatetieconstraints

betweenthedescribedelementswereimplemented,thus

repro-ducingthecommercialvalveofFig.1A.

Thevalvemodelwasprovidedwithafixedconstraintas

bound-arycondition,appliedonthesurfaceobtainedfromthesegment

(A5A6)revolutionaroundthevalve(A7A8)axis.Thetopsurfaceof

thevalvewasloadedwithapressure(normaltothevalvesurface)

whichlinearlyincreased,untilreachingthecrackingpressure.

Thevalve design wascomposedofthree elements. Thefirst

onerepresentedthevalvebasement,obtainedfromthe

revolu-tion(360◦)aroundthevalveaxisoftheareadelimitedbysegments

A6A5,A5A4,A4A2,A2A1,A1A3andA3A6(Fig.1B).Thesecondvalve

elementwasobtainedfromthecompleterevolution(360◦)around

thevalveaxisoftheareadelimitedbysectionsA1A2,A2A9,A9A10

andA10A11.Thethirdelementreferredtothevalvestripsobtained

whenthediecutwasappliedonthetopsurfaceofthepolymeric

valve.Asinglestripwasobtainedfromarevolutionof60◦around

thevalveaxisoftheareadelimitedbysectionsA10A8,A8A7,A7A11

andA11A10.

Thefirstelement(valvebasement)wasmeshedwith2160cells,

thesecond(valvehead)wasmeshedwith1224cells.Thethird

ele-ment(valvestrip)wasmeshedwith81cells.AstructuredHexmesh

wasusedforthethreebodies.Todemonstratethatthemeshwas

suitableforthetargetapplication,wecarriedoutmeshrefinement

tests(Supplementarymaterials,TableS1). Thementionedmesh

sizevaluewasidentifiedasagoodcompromisebetweenadherence

ofthepredictedOPvaluestotherealonesandthetimeneededto

completeeachsimulation.Oncethemeshgridofthevalvemodel

wasdefined,a setofnodeswasgeneratedinordertostorethe

segment(A8A7)displacementvaluesforeachvalvestrip,during

theentiresimulation.Therefore,12meshnodeswereselectedto

detectvalveopening.VonMisesstresseswerealsorecordedfor

eachsimulation.

2.3. Polymersynthesisandmechanicaltesting

The material selected for valve fabrication was PDMS

(Syl-gard184,DowCorning).Thismaterialisparticularlysuitablefor

biomedicalapplications,duetoitshighbiocompatibility,its

tun-ablemechanical properties, and itsgood thermal and chemical

(4)

oxy-Table1

ParametervaluesconsideredduringFEMsimulations.

genplasma,formodifyingsurfacepropertiesandmechanicaland chemicalperformances[24],[25].

PDMSstiffnesscanbemodulatedbyvaryingtheratiobetween

the monomer and its curing agent. In this work, 6 different

monomer/curingagentratioswereanalyzed:5:1,10:1,20:1,30:1,

40:1and50:1(w/w).

SamplemechanicalpropertieswereevaluatedwithanINSTRON

(5)

performingtractiontests.Sampleswerecutinto20×5×2mm3

slicesandallocatedbetweentwoaluminiumclamps.Allsamples

werepulled atthe constantspeed of 5mm/min, untilreaching

samplefailure.Datawererecordedatafrequencyof100Hz,the

stresswascalculatedastheloaddividedbythecross-sectionarea

oftensilespecimens,whilethestrainwascalculatedastheratio

betweentheextensionandtheinitiallengthoftensilespecimens.

Thetensilemodulusforeachtestedsamplewascalculated

start-ingfromthestress/straincurve,accordingtoastandardprocedure

[26].Mechanicaltestswererepeatedonthreedifferentsamples,

foreachsampletype.

2.4. Valvefabrication

Inorder tovalidate thesimulation results, polymericvalves

werefabricatedbyexploitingdedicatedmolds.Themoldswere

composedoftwoparts,providedwiththroughholes,neededfor

aligningthepartsandclosingthemold(usingM3screwsandnuts).

Toachievethediecutonthevalvesurface,aproperframe(designed

toconstrainthepolymericvalve)andacustombladewereused.

Bothcuttingframeandmoldswerefabricatedbyusinga3D

printer(HD3000,ProJet).Thecustombladewasobtainedby

mod-ifyingthecuttingedgeofanoff-the-shelfsurgicalblade(CHIMO,

size22).

Thevalvefabricationprocesswasbasedonthefollowingsteps:

(1)PDMSpreparation:monomerandcuringagentweremixedwith

thedesiredratio,thenthemixturewasdegassedfor20minina

vacuumchamber.(2)Moldsurfacetreatment:a thinteflonfilm

(LOCTITE8192) wasdeposited byspraying it onthe mold

sur-face,tofacilitatesubsequentPDMSremoval fromthemold.(3)

PDMScasting:oncethemixturewaswelldegassed,itwascasted

intotheteflon-coatedmold.(4)Polymerization:afterPDMS

cast-ing,themoldwasclosedandthermallytreatedat90◦Cfor3h,to

allowfullvalvepolymerization.(5)Valveextractionandcutting:

themoldwasopenedand thepolymericvalveremoved.Itwas

thensubjectedtodiecuttingbythededicatedframeandcustom

blade.

2.5. Valvetesting

Totestthepolymericvalvefunctionality,adedicated

experi-mentalset-upwasdeveloped(Fig.2).

Theset-up consistedof a motorized pistonprovided with a

finecontrolofboth horizontalstrokeand speed.Thehorizontal

strokekeptconstanttheheightofthewatercolumndownstream

of thevalve during the tests,which wasneeded to detectthe

valveOP.Speedcontrolwascompiledintoadevelopmentboard

(STM32F407,ST)exploitingSimulink(MathWorks)environment

withasamplingfrequencyof1kHz.

TomonitortheOP,thedevelopmentboardwasconnectedtoa

pressuresensor(HCX001D6V,purchasedfromSensorTechnics)for

wetenvironments,capableofmeasuringpressuresupto100kPa.

Themechanicalparts ofthetesting set-upincludeda Screw

Ball(BSSC1510-300-SC10,diameter15mm,stride:10mm,stroke:

300mm)andappropriatetoolsforsupportingit(BSW12,BUN12,

BNFB1505C-30 produced by MISUMI company). These

com-ponents were assembled on a derlin-polycarbonate frame. A

DC-micromotor(Series2342012CR,Faulhaber) providedwitha

PlanetaryGearheadtoincreasethemotor torquewitha

reduc-tionratioof43:1(Series30/1S3,Faulhaber),ahighperformance,

low-cost,three-channelopticalincrementalencoder(HEDS-5540,

AvagoTechnologies)andafullH-Bridge(L298,ST)weremounted

onthesystem.

Thetestingset-upallowedasyringe(5ml)anditsplunger,

con-nectedwiththepistonpusher,tobemounted.Athree-wayT-joint

Table2

showsthedifferentPDMSmonomer/curingagentratiosandthecorresponding Young’smoduli,obtainedbywayoflinearfittingofthestress–straincurvesinthe initial(linear)region.

Monomer/curingagentratio Young’smodulus[MPa]

5:1 4.95±0.33 10:1 3.55±0.35 20:1 1.26±0.16 30:1 0.41±0.06 40:1 0.14±0.02 50:1 0.08±0.01

connectedthevalveseat,thepressuresensorandthesyringe out-put.Thevalveopeningdirectionwashorizontal,toavoidgravity effects.

Six polymeric valve types (four valves for each type) were fabricatedby using three different molds (thus,three different geometries)andtwodifferentmaterialtypes(10:1and20:1PDMS). Thefirstvalvegeometrytobefabricatedfeatured aP1valueof 3.7mmandaP3valueof0.2mm.Thesecondvalvegeometry fea-turedaP1valueof2.5mmandaP3valueof0.25mm.Thethird valvegeometryfeaturedaP1valueof2.5mmandaP3valueof 0.3mm. The otherparameters, common for thedifferent valve geometries,were: P2=50◦, P4=83◦, P5=0.86mm, P6=0.25mm, P7=97◦,P8=0.67mmandP9=0.13mm.

Oncethe twenty-fourpolymericvalveshad been fabricated, theyweretestedinordertomeasuretheirOP(detectedusingthe dedicatedsensortomeasurepressuredrop).Thesyringewasfilled withairordistilledwater(d-H2O)andthepistonspeedwassetat 5mm/s.Eachvalvewastestedthreetimes,andsixcyclesof open-ing/closurewereperformed,withineachtest.Thus,eighteenOP valueswererecordedforeachtestedvalve.

Anadditionaltestwascarriedout,inordertoevaluatethe tight-nessofthefabricatedvalves.Duringthistest,thepistonspeedwas setat0.25mm/sanditwasstoppedbeforereachingthevalveOP. Differentconstantpressurevalueswerekeptfor10s.Thisallowed thevalvepressurelosstobemeasuredwhenrelativelyhigh pres-sureswereappliedandkeptconstant.

3. Resultsanddiscussion

3.1. PDMSmechanicalproperties

Thestress–straincurvesexperimentallyobtainedforthePDMS sampleswithdifferentmechanicalpropertiesasshowninFig.3.

ThesecurveswereimportedintheAbaqusenvironment,inorder

toproperlysimulatematerialmechanicalbehavior.

Table2Young’smoduliofthePDMSsamplestested,fordifferent

monomer/curingagentratios.Threesamplesweretested,foreach

sampletype.

3.2. Simulationresults

Fig.4showssomeframescapturedfromvalveFEMsimulations,

inwhichprogressivevalvevaultcompressionandopeningphases

arereported.Adynamicsimulationofvalveopeninganda

compar-isonbetweentherealandthesimulatedvalveareshowninVideo

S1(seeSupplementarymaterials).

Theanalysisofthe9parametersgenerated9datasets.Ineach

dataset,the5parametervaluesofTable1werecombinedwith

6materialstiffnessvalues,thusgenerating30simulationresults.

Asimpleapproachthatimmediatelyhighlightsmacroscopic

dif-ferencesbetweentheobtaineddatasets isdesirable. We fitthe

simulationresultsandcomparedthecoefficientsobtainedfromthe

(6)

Fig.2.Valvetestingset-upanditscomponents.

Fig.3.Representativestress-straincurvesofPDMSsamplesshowingdifferentstiffness.

Fig.4.FEMsimulationframes(toppictures:topview,bottompictures:bottomview),showingvalvebehaviorwhenanincreasingpressurewasappliedonthetopvalve surface.DifferentcolorsstandfordifferentvonMisesstressvalues.

thevalveperformance.Themathematicalexpressionusedfordata

fittingwasanexponentialfunction:

P(x,y)=a×xb×yc (1)

wherexwasthematerialstiffnessandywastheparametertobe

evaluated(fromP1toP9).Thecoefficientsa,bandcwereobtained

byusingcftool(Matlabtoolbox),andfittingtheopeningpressure

P(x,y),obtainedfromFEMsimulations,asafunctionofthe

mate-rialstiffnessandtheparameteranalyzed.Theresultsareshownin

Table3.

Byfocusingontheobtainedcoefficientvalues,especiallywith

regardtotheexponentialcoefficientc,animmediatecomparison

betweenparameters(P1andP9)ispossible.Thevalvethickness

(P3)isthemostsensitiveparameter,beingcharacterizedbya c

valueof1.6920.AnotherparticularlysensitiveparameterisP4(c

(7)

Fig.5.Valveperformances(intermsofopeningpressure)incorrespondencetodifferentmaterialstiffnessanddifferentP1values.Alltheotherparameterswerethoseofthe commercialvalve.Angularparameters(P2,P4andP7)werekeptconstant(Pi=Pic),whilelength-relatedparameters(P3,P5,P6,P8andP9)wereproperlyscaledaccording tothehomothetictransformation(Pi=Pic×SF).

Fig.6.Valveperformances(intermsofopeningpressure)incorrespondencetodifferentmaterialstiffnessanddifferentP3values.Alltheotherparameterswerethoseofthe commercialvalve,properlyscaledaccordingtothehomothetictransformationinordertohaveaP1equalto2.5mm(P2=50◦,P4=83◦,P5=0.86mm,P6=0.25mm,P7=97◦, P8=0.67mmandP9=0.13mm).

previouslymentioned,thedesignofapolymericvalveshowinga

specificOPisaredundanttask.Therefore,itcanbeusefultoselect

adesignprotocolaimingatreducingtheoverallsimulationeffort.

Byvaryingvalvethicknessandmaterialstiffness,alargerangeof

OPvaluesisachievable.Anefficientdesignprotocolmayinclude:

(1)thedefinitionofthevalvesizeandtheidentificationofa

suit-ablescalingfactor;(2)thetuningofvalvethicknessand/ormaterial

stiffness.Alltheotherparameters(P2,P5,P6,P7,P8andP9)playa

minorrole,ashighlightedbythesmallcoefficientvaluesreported

inTable3.

Fig.5showsthesimulationresultsobtainedbyapplyinga

homo-thetictransformation(i.e.byvaryingP1).Fig.6showsthevalve

performanceswhenvalvethickness(P3)varied.

Asmentioned,theotherparametersarenotcrucialfor

deter-miningvalvebehaviour.Inanycase,thecorrespondentsimulation

resultsarereportedinFig.7.

Table3

Coefficientsa,bandcderivingfromthefittingofFEMdata,byusingEq.(1).

Parameteranalyzed Coefficientvaluesderivedfromdatafitting R2value

a b c P1 0.0052 0.6956 0.0517 0.9858 P2 0.0055 0.7273 0.0114 0.9898 P3 0.1363 0.8298 1.6920 0.9961 P4 0.0009 0.6991 0.4127 0.9878 P5 0.00786 0.7716 −0.1892 0.9926 P6 0.0050 0.7319 −0.0183 0.9887 P7 0.0043 0.7481 0.0471 0.9942 P8 0.0057 0.7274 0.0057 0.9890 P9 0.0054 0.7039 −0.0052 0.9871

(8)

Fig.7. Valveperformances(intermsofopeningpressure)incorrespondencetodifferentmaterialstiffnessanddifferentparametervalues.Whenaparameterwasvaried, alltheotherangularparameterswerekeptconstantandequaltothoseofthecommercialvalve(Pi=Pic),whilelength-relatedparameterswerekeptequaltothoseofthe

commercialvalve,properlyscaledaccordingtothehomothetictransformationthatallowedP1=2.5mm(Pi=Pic×0.67).

Fig.8.(1)Closedmold,castedwithPDMSandthermallytreated;(2)moldopening;(3)valveextraction;(4)frameusedtopunchthetopsurfaceofthevalve;(5)valve positioningintotheframeusedfordiecutting;(6)bladeinsertionintotheframe;(7–9)valveprototypeandinsertionintoavalveseat,toconstrainthepolymericvalve withinthetestingset-up.

(9)

Fig.9.OPvaluesexperimentallyobtainedfromthetestingoffour10:1valvesandfour20:1valvesforeachgeometryprofile(dataareexpressedasmeanvaluesandstandard deviations)andcomparisonwithFEMpredictions.Themeanabsoluteerror(MEA)valuesarereportedforeachsampletype.

3.3. Fabricationandvalidation

Fig.8showsapolymericvalveobtainedbyusingthededicated

moldandcuttingframe.

Thevalveprofilesselectedtovalidatethesimulationshadthe

followingparametersincommon:P2=50◦,P4=83◦,P5=0.86mm,

P6=0.25mm, P7=97◦,P8=0.67mm and P9=0.13mm. The first

testedvalvegeometryfeaturedP1=3.7mmandP3=0.2mm.The

secondtestedvalvegeometryfeaturedP1=2.5mm(suitablefor

urinarysystems)andP3=0.25mm.Thethirdtestedvalvegeometry

featuredP1=2.5mmandP3=0.3mm.

Foreachvalvegeometry,fourPDMSvalvesmadeof10:1PDMS

andfourvalvesmadeof20:1PDMSwerefabricatedandtestedby

usingtheset-upofFig.2.

ThetargetOPscanbeextrapolatedfromEq.(1).FEM

simula-tionspredictedthatthefirstvalvegeometryshouldhaveanOPof

13.6kPaifmadeof10:1PDMSandanOPof6.6kPaifmadeof20:1

PDMS.ThesecondvalvegeometryshouldhaveanOPof37.4kPaif

Fig.10.(A)Pressureprofilesovertime(sixopeningcycles)fora10:1polymericvalve(P1=2.5mm,P3=0.30mm)testedwithad-H2Oflowproducingacontinuously

increasingpressure.(B)Pressureprofileovertimeforthesamevalve,obtainedbyraisingthepressureuntilreachingthevalveOP.Differentconstantpressurevalueswere keptfor10s.Thesetestsallowedustoevaluatevalveleakage.

(10)

madeof10:1PDMSandanOPof15.7kPaifmadeof20:1PDMS.

ThethirdvalvegeometryshouldhaveanOPof50.9kPaifmadeof

10:1PDMSandanOPof21.4kPaifmadeof20:1PDMS.

Inagreementwiththeconditionssetinthesimulations,the

fab-ricatedvalvesweresolicitedbyahomogeneouspressurenormalto

thevalvesurface.Thiswasachievedbyloadingthemwithaliquid

flow.Thepistonadvancedwithaconstantspeedof5mm/sandthe

pressuredetectedbythepressuresensorincreaseduntilthe

open-ingpressurewasreached.Afterwards,thepressureimmediately

fell.Theactualperformancesofthefabricatedvalvesarereported

inFig.9.

Theresultsobtaineddemonstratethattheperformancesofthe

fabricatedunidirectionalvalves,intermsofOP,areingood

agree-mentwiththepredictionsderivedfromFEMsimulations,witha

maximumerrorof23%.ThisallowsustouseEq.(1)asanefficient

toolforthedesignofunidirectionalvalvesfordifferentapplications,

byplayingonthegeometricandmechanicalfeatures.Concerning

thedesignofanAScapableoftacklingUIdiseases,thesecondvalve

geometrytested(P1=2.5mmandP3=0.25mm)anda20:1PDMS

monomer/curingagentratioseemparticularlysuitable.

Fig.10Ashowspressureloading/unloadingcyclesperformedon

avalveprototype.Fig.10Bdemonstratesthatthevalveshowsa

goodtightness:thevalvewasloadedwithrelativelyhigh

pres-surevalues,keptconstantonthevalvefor10s,withsmallpressure

losses.ForpressurevaluesuptoabouthalfofthevalveOP,almost

no losses were detected(max pressure loss: 0.7% in 10s). For

higherpressurevalues,upto90%ofvalveOP,thepressureloss

increased,butitdidnotovercomeinanycase2.4%in10s.This

fur-therdemonstratesthesuitabilityofthistechnologicalsolutionfor

thedevelopmentofanAS,aswellasforotherapplications.

Thisresearchworkcanbeconsideredworthyofattentionfor

differentreasons.Themain purposeofthis article istolaythe

groundworkforthedesignandmanufacturingofanASthatcan

tackleUI[27](seeSupplementarymaterials,Fig.S1).Theresults

reportedinthispaperpermitthedesignofapolymericvalve

show-ingspecificperformancesintermsofOP.Thevalveprofilehasbeen

described by univocallydefining the correspondent parameters

(thesedetails,inmanyarticles,arepartial,omitted,ordescribed

inanonparametricmanner)[9,10,16,18,19].Anotherimportant

noveltycomparedtothestate-of-the-artistheaccurateparameter

analysis,conductedwiththeaimofhighlightingwhichparameters

weremoresignificantforvalveperformance.Finally,the

valida-tionresultsdemonstrate that relativelysimple and widespread

manufacturingtechniques[28]canbeusedtofabricate—ina

reli-ablewayandwithrelativelysmalldeviationsfromthesimulation

results—one-wayvalvesthatareabletoensureatightseal-up

pres-sureintheorderof10kPamagnitude.

4. Conclusion

Thispaper aims at describing thesteps needed todesign a

polymericvalvesuitableforapplicationinartificialsphincter

sys-temsemployedfortacklingurinaryincontinencediseases.Aftera

preliminaryanalysis,acommercialvalveprofileusedinthefood

industrywaschosen asreference.Byanalyzing thecommercial

valvecrosssection,nineparameterswereextracted,whichdefined

thevalvedesign. Then, aparametricanalysiswasconductedto

identifywhichfactorsweremoresuitableforachievinga

signif-icanttuningofthevalveperformance.Thisapproachallowedusto

defineadesignstrategyaimedatminimizingdesignefforts.After

settingthevalvediameter,themostinfluentialparameterswere

valvestiffnessandthickness.Byvaryingtheseparameters,opening

pressuretuninginthephysiologicalrangewaseasilyachievable.

Appropriatemoldsandframesformanufacturingandtesting

polymericvalvesweredesignedandfabricated.Twenty-four

poly-mericvalves(dividedinsixdifferentexperimentalgroups)were

testedinordertovalidatethesimulations.ExperimentalResults

revealedamaximumerrorof23%comparedtosimulation

predic-tionswithageneralgoodagreementbetweentheenvisionedand

theactualvalveperformance.Thisworkwillconstitutetheground

forfutureadvancementsonthedevelopmentofbothpassiveand

controllableartificialurinarysphincters.

Acknowledgment

The authors would like to thank the Fondazione Cassa di

RisparmiodiLucca(Lucca,Italy),forprovidingfinancialsupport

forthisstudy,intheframeworkoftheSUAVESproject(Artificial

UrinarySystembasedonbladderandsphincterendoprostheses).

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in

theonlineversion,athttp://dx.doi.org/10.1016/j.sna.2015.07.005

References

[1]D.E.Irwin,Z.S.Kopp,B.Agatep,I.Milsom,P.Abrams,Worldwideprevalence estimatesoflowerurinarytractsymptoms,overactivebladder,urinary incontinenceandbladderoutletobstruction,BJUInt.108(2011)1132–1138.

[2]C.R.Bauer,M.Gozzi,W.Hübner,V.W.Nitti,G.Novara,A.Peterson,C.G.Stief, Contemporarymanagementofpostprostatectomyincontinence,Eur.Urol.59 (2011)985–996.

[3]E.Chung,M.Ranaweera,R.Cartmill,Newerandnovelartificialurinary sphincters(AUS):thedevelopmentofalternativestothecurrentAUSdevice, BJUInt.110(2012)5–11.

[4]D.S.Elliott,T.B.Boone,Urethraldevicesformanagingstressurinary incontinence,J.Endourol.14(1)(2000)79–83.

[5]H.Lamraoui,A.Bonvilain,G.Robain,H.Combrisson,S.Basrour,A. Moreau-Gaudry,P.Cinquin,P.Mozer,Developmentofanovelartificial urinarysphincter:aversatileautomateddevice,IEEE/ASMETrans.Mech.15 (6)(2010)916–924.

[6]S.Hached,O.Loutochin,J.Corcos,A.Garon,M.Sawan,Novel,remotely controlled,artificialurinarysphincter:aretro-compatibledevice,IEEE/ASME Trans.Mech19(4)(2014)1352–1362.

[7]Z.Hached,O.Saadaoui,A.Loutochin,J.Garon,M.Corcos,Novel,wirelessly controlled,andadaptiveartificialurinarysphincter,IEEE/ASMETransMech. (2015),http://dx.doi.org/10.1109/TMECH.;1;2015.2389254

[8]N.Pinzi,T.,Mazzocchi,G.Giuliani,Astructureofartificialendo-urethral sphincter,PatentNo.WO2013144770.

[9]H.Mohammadi,K.Mequanint,Prostheticaorticheartvalves:modelingand design,Med.Eng.Phys.33(2011)131–147.

[10]G.Burriesci,F.C.Marincola,C.Zervides,Designofanovelpolymericheart valve,J.Med.Eng.Technol.33(2010)7–22.

[11]M.R.Labrosse,C.J.Beller,F.Robicsek,M.J.Thubrikar,Geometricmodelingof functionaltrileafletaorticvalves:developmentandclinicalapplications,J. Biomech.39(2006)2665–2672.

[12]A.N.Smuts,etal.,Applicationoffiniteelementanalysistothedesignoftissue leafletsforapercutaneousaorticvalve,J.Mech.Behav.Biomed.Mater.4 (2011)85–98.

[13]B.Rahmani,S.Tzamtzis,H.Ghanbari,G.Burriesci,A.M.Seifalian,

Manufacturingandhydrodynamicassessmentofanovelaorticvalvemadeof anewnanocompositepolymer,J.Biomech.45(2012)1205–1211.

[14]D.Snakenborg,H.Klank,J.P.Kutter,Polymermicrovalvewithpre-stressed membranesfortunableflow–pressurecharacteristics,Microfluid.Nanofluid. 10(2011)381–388.

[15]R.Mohan,B.R.Schudel,A.V.Desai,J.D.Yearsley,C.A.Apblett,P.J.Kenis,Design considerationsforelastomericnormallyclosedmicrofluidicvalves,Sens.Act. BChem.160(2011)1216–1223.

[16]A.Jahanshahi,P.Salvo,J.Vanfleteren,PDMSselectivebondingforthe fabricationofbiocompatibleallpolymerNCmicrovalves,J.Microelectromech. Syst.22(2013)1354–1360.

[17]W.Hilbert,B.Jakoby,Controlledliquidflowinamicrofluidicnetworkwith pressuresensitivevalvesbasedonpolydimethylsiloxane(PDMS)/neodymium (NdFeB)composites,ProcediaEng.47(2012)382–385.

[18]A.I.Hickerson,H.W.Lu,K.Roskos,T.Carey,A.Niemz,Disposableminiature checkvalvedesignsuitableforscalablemanufacturing,Sens.Act.APhys.203 (2013)76–81.

[19]Y.Uohashi,K.Suzumori,H.Taniguchi,Fabricationandevaluationofvarious typesofmicroone-wayvalvesthroughmicrorubbermoldingprocess,J. Mech.Sci.Technol.24(2010)219–222.

[20]M.M.Elmissiry,A.G.Ali,G.A.Ali,Differenturodynamicpatternsinfemale bladderoutletobstruction:canurodynamicsalonereachthediagnosis,Arab. J.Urol.11(2013)127–130.

(11)

[21]F.Schmidt,P.Shin,T.M.Jorgensen,J.C.Djurhuus,C.E.Constantinou, Urodynamicpatternsofnormalmalemicturition:influenceofwater consumptiononurineproductionanddetrusorfunction,J.Urol.168(2002) 1458–1463.

[22]W.Schäfer,P.Abrams,L.Liao,A.Mattiasson,F.Pesce,A.Spangberg,P.V. Kerrebroeck,Goodurodynamicpractices:uroflowmetry,fillingcystometry, andpressure-flowstudies,Neurourol.Urodynamics21(2002)261–274.

[23]J.Rothschild,L.ChangKit,L.Seltz,L.Wang,M.Kaufman,R.Dmochowski,D.F. Milam,Differencebetweenurethralcircumferenceandartificialurinary sphinctercuffsize,anditseffectonpostoperativeincontinence,J.Urol.191 (2014)138–142.

[24]J.C.McDonald,G.M.Whitesides,Poly(dimethylsiloxane)asamaterialfor fabricatingmicrofluidicdevices,Acc.Chem.Res.35(2002)491–499.

[25]G.G.Genchi,G.Ciofani,I.Liakos,L.Ricotti,L.Ceseracciu,A.Athanassiou,V. Mattoli,Bio/non-biointerfaces:astraightforwardmethodforobtaininglong termPDMS/musclecellbiohybridconstructs,Coll.Surf.BBiointerf.105 (2013)144–151.

[26]W.D.Callister,Rethwisch,MaterialsScienceandEngineering:AnIntroduction, sixthed.,JohnWiley&Sons,Minnesota(USA),2003,pp.113–152.

[27]N.Pinzi,T.,Mazzocchi,G.Giuliani,Astructureofartificialendo-urethral sphincter.WO2013IB52156,2013.

[28]J.M.Pearce,C.M.Blair,K.J.Laciak,R.Andrews,A.Nosrat,I.Zelenika-Zovko,3-D printingofopensourceappropriatetechnologiesforself-directedsustainable development,J.SustainableDev.3(2010)17.

Biographies

Tommaso Mazzocchi obtained a M.Sc.in Biomedical EngineeringinApril2013atUniversityofPisa(Italy)with athesisentitled:“Developmentandvalidationofsafety systemsformagneticallyactuatedendoluminalrobots”. HeiscurrentlyResearchAssistantatTheBioRobotics Insti-tuteofScuolaSuperioreSant’Anna,wherehecarriesout researcheffortsinthefieldofbiomedicaldevices,witha particularfocusonsystemsaimedtofaceurinary inconti-nenceandbladdercancer.Heisinventorof2international patentsrelatedtoartificialsphinctersandartificial blad-ders.

LeonardoRicottiisAssistantProfessorandheadofthe “Micro-Nano-BioSystemsandTargetedTherapy”Labat TheBioRoboticsInstituteofScuolaSuperioreSant’Anna. Hisresearchinterestsincludebiomechatronics,bionics, biomaterials,regenerativemedicineandbiohybrid sys-tems.Heisauthorofabout50scientificpapers(Source: Scopus),withmorethan30onISIJournalsand3book chapters.Heisalsoinventorof4patents,relatedtosmart mechatronicdevicesformedicalapplications.

NovelloPinziobtainedaM.Sc.inMedicineandSurgeryat theUniversityofPisa(Italy)in1976.In1980heobtained the specializationinUrology atthe UrologyClinicof Florence(Italy),andin1985heobtainedasecond special-izationingeneralsurgeryattheUniversityofPisa.Over thepast25years,heperformedatotalof800 interven-tions.Hismainfieldsofinterestarebladderreconstruction afterradicalcystectomyandtreatmentofurinary incon-tinencebymeansofartificialsphincters.Heisinventorof 2internationalpatentsrelatedtoartificialsphinctersand artificialbladders.

Arianna Menciassi is Full Professor of Biomedical Roboticsandleaderofthe“SurgicalRoboticsandAllied Technologies”Area,attheBioRoboticsInstituteofScuola Superiore Sant’Anna.Hermain research interests are inthe fieldsofmedical mechatronics,bio-hybrid sys-tems,biomedicalmicro-andnano-devicesandrobotic surgery.Shepossessesanextensiveexperiencein Euro-pean Projectsand internationalcollaborative projects ontopicsrelatedtoroboticandmicroroboticdiagnosis, surgeryandtherapy.Sheisauthorofabout300 interna-tionalpapers(Source:Scopus),withmorethan150onISI journals,1editedbookand6bookchaptersonmedical devicesandmicro-technologies.

Riferimenti

Documenti correlati

Magnetic resonance imaging is widely used for in vivo applications, due to its safety, spatial resolution, soft tissue contrast, clinical relevance, and ability to record anatomical

The QCD background to SR γγ S−L , SR γγ S−H , SR γγ W−L and SR γγ W−H is expected to arise from events with two real, isolated photons (diphoton QCD events) unaccompanied

Thus, we hypothesized that subjects with more severe NAFLD/NASH are those with increased adipose tissue (AT) insulin resistance (Adipo-IR, reflecting the impaired suppression

PULVIRENTI, in La giustizia penale minorile: formazione ,devianza, diritto e processo, Giuffrè, 2004, pag.329.. Sotto questo aspetto, la messa alla prova evoca un altro

A comparison of the layouts of these three documents (Table 2) shows how little they have in common in the following three categories: (1) name of the button at the foot of each

Mean values and variances of Q and D, for contingency tables whose values are uniformly distributed in [0, 1], when the number of rows and columns are

 conoscenze sulla sicurezza alimentare, sezione nella quale sono riportate domande con risposta vero o falso su intossicazioni alimentari, igiene del personale,