Blood Cell Adhesion on a Silicone Heart Valve Leaflet Processed using Magnetic Abrasive Finishing

MISSING IMAGE

Material Information

Title:
Blood Cell Adhesion on a Silicone Heart Valve Leaflet Processed using Magnetic Abrasive Finishing
Physical Description:
1 online resource (83 p.)
Language:
english
Creator:
Boggs,Taylor A
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Mechanical Engineering, Mechanical and Aerospace Engineering
Committee Chair:
Greenslet, Hitomi
Committee Members:
Sarntinoranont, Malisa

Subjects

Subjects / Keywords:
abrasive -- adhesion -- blood -- cell -- finishing -- heart -- magnetic -- platelets -- silicone -- valve
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre:
Mechanical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Heart valve prosthetics replace damaged, malfunctioning valves in the hope of extending and improving a patient's quality of life. Current mechanical valves are durable but suffer from thrombogenicity and flow separation, and cause blood damage leading to coagulation. While bioprosthetic valves have better haemodynamic function than mechanical valves, the valves suffer from tears due to inflammation and collagen degradation. The absence of living tissue leaves them unable to repair themselves, and their antigenicity must be masked. Polymeric valves have the potential to produce improved haemodynamic function without the complications associated with bioprosthetic valves, which include calcification, hydrolysis, and durability. The goal of this project is to develop a trileaflet polymeric heart valve to overcome the shortcomings of current mechanical and bioprosthetic valves. The valve should replicate a mold surface texture and reduce blood cell adhesion. Blood cell adhesion and subsequent thrombus formation is a major concern facing the development of replacement heart valves; complications due to thrombosis occur between 1.5% and 3% per year for current mechanical and bioprosthetic valves. Studies investigating the effects of surface texture on cell adhesion are well established and cells, including, but not limited to platelets and red blood cells, fibroblasts, and bladder smooth muscle cells, have responded to changes in surface texture and roughness. For the initial trials, silicone was chosen as the valve material; it is inert and biostable, easy to manufacture, and has shown the ability to replicate surface features at the micro- and nanometer scales. To study the effects of surface texture on cellular adhesion, silicone leaflets were developed from finished brass molds. The mold surface is fabricated using a magnetic abrasive finishing (MAF) process. In the MAF process, the finished surface is controlled by the relative motion of the magnetic abrasive particles against the workpiece. The process is capable of creating various textures with varying surface roughness, and it has the ability to finish free-form surface geometries. The leaflets were secured in a polycarbonate flow chamber and flushed with whole human blood, followed by phosphate buffered saline (PBS), under low shear stress to study the conditions most susceptible to adhesion. Using light microscopy, blood cell adhesion was quantified by counting the number of adhered cells and normalizing the number to the leaflet surface area. This research uses the MAF process to finish a heart valve mold in an effort to reduce the adhesion of blood cells on the resulting leaflet surface. The MAF process is able to produce various textures on the mold surface, and the surface texture is replicated onto molded silicone leaflets. The corresponding MAF-produced smooth silicone surface reduces blood cell adhesion and aggregation.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Taylor A Boggs.
Thesis:
Thesis (M.S.)--University of Florida, 2011.
Local:
Adviser: Greenslet, Hitomi.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043457:00001


This item is only available as the following downloads:


Full Text

PAGE 1

BLOODCELLADHESIONONASILICONEHEARTVALVELEAFLETPROCESSEDUSINGMAGNETICABRASIVEFINISHINGByTAYLORA.BOGGSATHESISPRESENTEDTOTHEGRADUATESCHOOLOFTHEUNIVERSITYOFFLORIDAINPARTIALFULFILLMENTOFTHEREQUIREMENTSFORTHEDEGREEOFMASTEROFSCIENCEUNIVERSITYOFFLORIDA2011

PAGE 2

2011TaylorA.Boggs 2

PAGE 3

Tomyfamily 3

PAGE 4

ACKNOWLEDGMENTS Iwouldliketothankmyfamilyforalloftheirunendingloveandencouragement,andunwaveringsupport.Iwouldliketothankmyadvisor,Dr.HitomiGreenslet,forwelcomingmeintoherlab.Ithasbeenanhonortoworkwithherandreceiveherguidance.Additionally,Iwouldliketothankourcollaborators,Dr.RogerTran-Son-Tay,Dr.FarisAl-Mousily,andDr.CurtDeGroffforalloftheiradvice,dedication,andeffortinrealizingthisproject.IalsowanttothankDr.MalisaSarntinoranontforbeingonmycommitteeandJohnGreensletforkindlytakingthetimetoeditthiswork.IwanttothankthemembersoftheMachineToolResearchCenter(MTRC),especiallythemembersofDr.Greenslet'slab,fortheirwelcomingfriendship,advice,andsupport.Ithasbeenpleasuregettingtoknowandworkwitheveryone,yourfriendshipischerishedandappreciated. 4

PAGE 5

TABLEOFCONTENTS page ACKNOWLEDGMENTS .................................. 4 LISTOFTABLES ...................................... 7 LISTOFFIGURES ..................................... 8 ABSTRACT ......................................... 10 CHAPTER 1INTRODUCTION ................................... 12 1.1ValvularFunctionandValveDisease ..................... 12 1.2CurrentValveProstheses ........................... 12 1.3MotivationandObjectives ........................... 15 2SILICONEVALVEMANUFACTURING ....................... 18 2.1DesignConsiderations ............................. 18 2.2ValveSpecications .............................. 18 2.2.1LeaetThickness ............................ 18 2.2.2FactorsthatInuenceCellularAdhesion ............... 18 2.3ValveManufacturingProcess ......................... 22 2.4ValveDurabilityTesting ............................ 22 3MOLDFABRICATIONUSINGMAGNETICABRASIVEFINISHING ....... 26 3.1OverviewofMagneticAbrasiveFinishing .................. 26 3.2ProcessingPrinciple .............................. 26 3.3FinishingMachineDevelopment ....................... 27 3.3.1DesignandBuild ............................ 27 3.3.2MagneticFluxDensity ......................... 30 3.4MoldSurfaceAnalysis ............................. 30 3.5MoldSurfaceFinishingCharacteristics .................... 33 3.5.1UnnishedMoldSurface ........................ 33 3.5.2MoldSurfaceFinishedwithLooseDiamondAbrasive ........ 34 3.5.3MoldSurfaceFinishedwithCompositeMagneticAbrasive ..... 36 4SILICONELEAFLETFABRICATION ........................ 43 4.1ControllingLeaetThickness ......................... 43 4.2SiliconeLeaetMoldReplication ....................... 46 5BLOODCELLADHESIONTESTING ....................... 50 5.1FlowChamberDevelopment ......................... 50 5

PAGE 6

5.2ExperimentalTestingSetup .......................... 52 5.3AnalysisofBloodCellAdhesion ....................... 57 6CONCLUSION .................................... 68 6.1ConcludingStatements ............................ 68 6.2FutureWork ................................... 68 APPENDIX ABLOODCELLADHESIONDATA .......................... 71 BCALCULATIONOFBLOODSHEARSTRESS .................. 78 REFERENCES ....................................... 80 BIOGRAPHICALSKETCH ................................ 83 6

PAGE 7

LISTOFTABLES Table page 2-1Leaetshearstress ............................. 22 3-1Maximummagneticuxdensity ....................... 30 3-2Measurementconditions ........................... 33 3-3Unnishedmoldsurfaceroughness .................... 34 3-4Diamondabrasivenishingconditions ................... 35 3-5Roughnesscomparisonofmoldsnishedwithdiamondabrasive .... 35 3-6Whitealumina(WA)magneticabrasivenishingconditions ....... 39 3-7DiamondmagneticabrasivenishingconditionsphaseI ......... 39 3-8DiamondmagneticabrasivenishingconditionsphaseII ......... 39 3-9Roughnesscomparisonofcompositemagneticabrasivemolds ..... 39 4-1Leaetcuringconditions ........................... 46 4-2Opticalprolermeasurementconditions .................. 47 4-3Dataprocessingconditions ......................... 48 4-4Siliconesurfaceroughnesscomparison .................. 48 5-1Bloodowrateandshearstress ...................... 52 5-2Experimentalconditions ........................... 56 5-3Numberoftrialspersurface ......................... 57 5-4Numberofareasanalyzedpersurface ................... 58 5-5Numberofadheredbloodcells ....................... 59 A-1DatafromLeaet1 .............................. 72 A-2DatafromLeaet2 .............................. 73 A-3DatafromLeaet3 .............................. 74 A-4DatafromLeaet4 .............................. 75 A-5DatafromLeaet5 .............................. 76 A-6DatafromLeaet6 .............................. 77 A-7DatafromLeaet7 .............................. 77 7

PAGE 8

LISTOFFIGURES Figure page 1-1Diagramoftheheartandvalvefunction ...................... 13 1-2Currentreplacementvalves ............................. 14 2-1Morphologyofadheredplatelets .......................... 20 2-2Scanningelectronmicroscopy(SEM)imagesofadheredplatelets ....... 21 2-3Schematicofproposedvalve ............................ 23 2-4Siliconevalvemanufacturingprocess ....................... 23 2-5Blockdiagramofdynamictestingsystem ..................... 24 2-6Photographofdynamictestingsystem ....................... 25 2-7Pressurewave .................................... 25 3-1Schematicofmagneticabrasivenishing(MAF)processingprinciple ...... 27 3-2Designofnishingmachine ............................. 28 3-3Photographofnishingmachine .......................... 29 3-4Photographofnishingmachinecontrolbox .................... 29 3-5Diagramofmagneticeldanalysis ......................... 31 3-6Photographofsurfaceroughnessproler ..................... 32 3-7Diagramofsurfaceprolemeasurement ...................... 32 3-8Diagramofsurfaceprole .............................. 33 3-9Unnishedmoldsurfaceroughnessprole ..................... 34 3-10Diagramofnishingdirection ............................ 35 3-11Roughnesscomparisonofdiamondabrasivemolds ............... 36 3-12Roughnessprolesofdiamondabrasivemolds .................. 37 3-13Roughnessprolesofdiamondabrasivemoldsatreducedscales ....... 38 3-14Roughnesscomparisonofcompositemagneticabrasivemolds ......... 40 3-15Roughnessprolesofcompositemagneticabrasivemolds ........... 41 3-16Roughnessprolesofcompositemagneticabrasivemoldsatreducedscales 42 8

PAGE 9

4-1Moldcuringorientations ............................... 43 4-2Diagramofcuringprocess .............................. 44 4-3Diagramofthicknessvariationmeasurement ................... 45 4-4Theeffectofmoldrotationonleaetthicknessvariation ............. 45 4-5Leaetprocessing .................................. 46 4-6Diagramofopticalprolermeasurement ...................... 47 4-7Comparisonofmoldsurfaceandreplicatedsiliconeleaet ............ 49 4-8Limitationofsiliconemoldreplication ........................ 49 5-1Flowchamberconcept ................................ 50 5-2Experimentaldesign ................................. 51 5-3Designofowchamber ............................... 52 5-4Photographofowchamber ............................. 53 5-5Flowchamberopeningarea ............................. 53 5-6Flowchamberpreparation .............................. 54 5-7Experimentalsetup .................................. 55 5-8Bloodbagspike ................................... 56 5-9Diagramofcellcountingmethod .......................... 57 5-10Adheredplatelets ................................... 58 5-11Adheredredbloodcells ............................... 60 5-12Comparisonofadheredcellstodiamondabrasivemolds ............ 62 5-13Comparisonofadheredcellstocompositemagneticabrasivemolds ...... 63 5-14Comparisonofdiamondabrasivesizeonsurfacenish ............. 64 5-15Comparisonofcompositemagneticabrasivesizeonsurfacenish ....... 65 5-16Timelapseimagesofadheredbloodcells ..................... 67 9

PAGE 10

AbstractofThesisPresentedtotheGraduateSchooloftheUniversityofFloridainPartialFulllmentoftheRequirementsfortheDegreeofMasterofScienceBLOODCELLADHESIONONASILICONEHEARTVALVELEAFLETPROCESSEDUSINGMAGNETICABRASIVEFINISHINGByTaylorA.BoggsAugust2011Chair:HitomiYamaguchiGreensletMajor:MechanicalEngineeringHeartvalveprostheticsreplacedamaged,malfunctioningvalvesinthehopeofextendingandimprovingapatient'squalityoflife.Currentmechanicalvalvesaredurablebutsufferfromthrombogenicityandowseparation,andcauseblooddamageleadingtocoagulation.Whilebioprostheticvalveshavebetterhaemodynamicfunctionthanmechanicalvalves,thevalvessufferfromtearsduetoinammationandcollagendegradation.Theabsenceoflivingtissueleavesthemunabletorepairthemselves,andtheirantigenicitymustbemasked.Polymericvalveshavethepotentialtoproduceimprovedhaemodynamicfunctionwithoutthecomplicationsassociatedwithbioprostheticvalves,whichincludecalcication,hydrolysis,anddurability.Thegoalofthisprojectistodevelopatrileaetpolymericheartvalvetoovercometheshortcomingsofcurrentmechanicalandbioprostheticvalves.Thevalveshouldreplicateamoldsurfacetextureandreducebloodcelladhesion.Bloodcelladhesionandsubsequentthrombusformationisamajorconcernfacingthedevelopmentofreplacementheartvalves;complicationsduetothrombosisoccurbetween1.5%and3%peryearforcurrentmechanicalandbioprostheticvalves.Studiesinvestigatingtheeffectsofsurfacetextureoncelladhesionarewellestablishedandcells,including,butnotlimitedtoplateletsandredbloodcells,broblasts,andbladdersmoothmusclecells,haverespondedtochangesinsurfacetextureandroughness. 10

PAGE 11

Fortheinitialtrials,siliconewaschosenasthevalvematerial;itisinertandbiostable,easytomanufacture,andhasshowntheabilitytoreplicatesurfacefeaturesatthemicro-andnanometerscales.Tostudytheeffectsofsurfacetextureoncellularadhesion,siliconeleaetsweredevelopedfromnishedbrassmolds.Themoldsurfaceisfabricatedusingamagneticabrasivenishing(MAF)process.IntheMAFprocess,thenishedsurfaceiscontrolledbytherelativemotionofthemagneticabrasiveparticlesagainsttheworkpiece.Theprocessiscapableofcreatingvarioustextureswithvaryingsurfaceroughness,andithastheabilitytonishfree-formsurfacegeometries.Theleaetsweresecuredinapolycarbonateowchamberandushedwithwholehumanblood,followedbyphosphatebufferedsaline(PBS),underlowshearstresstostudytheconditionsmostsusceptibletoadhesion.Usinglightmicroscopy,bloodcelladhesionwasquantiedbycountingthenumberofadheredcellsandnormalizingthenumbertotheleaetsurfacearea.ThisresearchusestheMAFprocesstonishaheartvalvemoldinanefforttoreducetheadhesionofbloodcellsontheresultingleaetsurface.TheMAFprocessisabletoproducevarioustexturesonthemoldsurface,andthesurfacetextureisreplicatedontomoldedsiliconeleaets.ThecorrespondingMAF-producedsmoothsiliconesurfacereducesbloodcelladhesionandaggregation. 11

PAGE 12

CHAPTER1INTRODUCTION 1.1ValvularFunctionandValveDiseaseHeartvalveprostheticsreplacedamaged,malfunctioningvalvesinthehopeofextendingandimprovingthepatient'squalityoflife.Itisestimatedthat300,000valvereplacementsaremadeworldwideeachyearwith100,000beingperformedannuallyintheUnitedStates[ 1 2 ].Thisnumberisexpectedtoincreasetoapproximately850,000annualreplacementsworldwideby2050[ 1 ].Rheumaticfeverandageingaretheleadingcausesofvalvedamagerequiringreplacement.Stenosis,theincompleteopeningofthevalve,andregurgitation,leakage,areeffectsofvalvefailure[ 1 3 ].ThehearthasfourvalvesdirectingbloodowasshowninFigure 1-1 :twoatrioventricularvalves(AVvalves)directowfromtheatriumtotheventricleandtwosemilunarvalvesseparatetheventriclesfromthearteries.Thevalvesopenandcloseaccordingtopressuregradients.OftheAVvalves,themitralvalveislocatedontheleftsideoftheheartandthetricuspidvalveontheright.TheAVvalvesopenwhenthepressureintheatriumbecomeshigherthanthepressureintheventricleandclosewhenthepressuregradientisreversed.Themitralvalveisabileaetvalveconsistingoftwocuspsofconnectivetissuewhilethetricuspidvalveisatrileaetvalvefeaturingthreecusps.Theaorticvalve,asemilunarvalve,separatestheleftventriclefromtheaorta,andthepulmonaryvalveseparatestherightventriclefromthepulmonaryartery[ 4 ]. 1.2CurrentValveProsthesesTwotypesofprostheticvalvesarecurrentlyinuse:mechanicalvalves,whichincludetheball-in-cage,tilting-disk,andbileaetvalves,andbioprostheticvalvesmadefromeitherwholeporcinevalvesorbovinepericardium.ThesevalvesareshowninFigure 1-2 .Athirdtypeofprostheticvalvethatislesscommonlyusedisthecryopreservedhomograftvalve. 12

PAGE 13

Figure1-1. Diagramoftheheartandvalvefunction Theball-in-cagevalve,developedbyStarr-Edwardsin1960,iscomposedofasilasticballhousedinsideofathree-strutalloycage.Whiledurable,thesevalvescreatecircumferentialbloodowandtheballcreatesawakeofstagnantowthatcancausethromboembolism.Duetotheirlowcost,thesevalvesarestillwidelyusedindevelopingcountries[ 1 ].Thetilting-diskvalvewasdevelopedbyBjork-Shileyin1969andiscomposedofapyrolyticcarbonleaetheldinplacebyalloystrutsandaTeonring.Itcorrectedthelateralbloodowbutcreatesaregionofstagnantbloodowimmediatelydownstreamoftheminororice.Additionalproblemswiththevalveincludestrutfractureandembolismformationonthedisk,anditproducesthelargestturbulentstressesofallthemechanicalvalves,approximately150Palocatedbehindthetiltingdisk[ 1 5 ].Currently,themostimplantedprostheticvalveisthebileaetvalve.Firstintroducedin1977bySt.JudeMedical,theleaetsaremadeofpyrolyticcarbonandcoatedwithgraphite.Thevalveproducesasymmetric,non-turbulent,centralowbuthighstresses 13

PAGE 14

appearduringforwardowwhileregurgitation,leakage,andstagnantowcanoccuraroundthehinges;thehingeareaalsotendstoinducethrombusformation[ 5 ].Ingeneral,mechanicalvalvesaredurablebutsufferfromthrombogenicityduetohighshearstresses,whichcanactivateplatelets,causeowseparation,stagnation,anddamagebloodcellsleadingtocoagulation.Consequently,lifelonganticoagulationtherapyisnecessary.Mechanicalvalvesalsosufferfrompannusovergrowth,thegrowthofexcesstissueoverthesewingringthatcancauseanarrowingofthevalveoriceorleaetimmobilization.Pannusovergrowthisaprimarycauseofobstructivevalvefailurethatcanbeproducedbyinammationandachronicreactionbythepatienttothesewingringthattreatstheprostheticasaforeignbodyorinjury[ 5 ]. Figure1-2. Currentreplacementvalvesinclude:(i)bileaetvalve,(ii)ball-in-cagevalve,(iii)bioprostheticvalves,and(iv)tiltingdiskvalve[ 6 ] 14

PAGE 15

Bioprostheticvalveshavebetterhaemodynamicfunctionthanmechanicalvalvesanddonotneedlong-termanticoagulationtherapy.Porcinevalves,rstintroducedbyHancockin1970,arewholevalvesthataresewnintothevalvestructure.Alternatively,bovinevalvesaremadefrombovinepericardiumthatiscuttoformthevalveleaets;thevalvesarexedinglutaraldehydeandoftenafxedtoastentbeforebeingsewnintothestructure.Bovinevalveshaveatheoreticaladvantageoverporcinevalvesinthattheleaetsaremadelargerandcanaccommodatetheshrinkagethatoccursoverthevalve'slifeonceimplanted.Additionally,theleaetopeningisalsomoresymmetric,improvinghaemodynamics,andthecollagencontentishigher,improvingdurability.Currently,itisunclearifthetheoreticaladvantagestranslatetothepatient[ 1 ].Tearsduetoinammation,collagendegradation,andtheabsenceoflivingtissuearetheprimarycauseoftissuevalvefailurewithcalcicationbeingthecauseoflessthanhalf[ 5 ].Theabsenceoflivingtissueleavesbioprostheticvalvesunabletorepairthemselvesandthereisaneedtomaskthevalve'santigenicity.Similarproblemsexistwithcryopreservedhomograftvalves[ 7 ]. 1.3MotivationandObjectivesPolymericvalveshavethepotentialtoexhibitimprovedhaemodynamicperformanceovermechanicalvalveswithoutthecomplicationsassociatedwithbioprostheticvalvesandcanhaveeitherbileaetortrileaetdesigns.Leaetvariationsincludehemi-cylindricalcusps;half-openleaets;variable-curvatureleaets;andelliptical,hyperbolic,andconicalshapes.Thedevelopmentofpolymervalvesalsooffersthepossibilityofpercutaneousimplantation,allowingthepatienttoavoidopenheartsurgery.BoudjemlineandBonhoefferperformedthistechniqueusingabioprostheticvalve[ 8 ].Currentissuesassociatedwithpolymericvalvesincludecalcication,hydrolysis,anddurability.Todate,themostwidelystudiedmaterialsforpolymericheartvalveshavebeenpolyurethanesandsiliconeelastomers.Polyurethaneshavegoodviscoelasticityandhavedemonstratedgoodhaemodynamicsandbiocompatibility, 15

PAGE 16

resistancetocyclefatigue,andhightensilestrength.Somepolyurethanes,suchaspolyetherurethane(PEU),showgoodresistancetoacidhydrolysisbuthavealowresistancetooxidation,whileanother,polycarbonateurethane(PCU),hasgoodresistancetohydrolysisandoxidationbutispronetocalcication.Problemswithbiodegradationandmineralization,though,havebeenamajordrawbacktousingpolyurethanesasaheartvalvematerial.Meanwhile,siliconehasshowngoodexibilityandbiocompatibilitybuthassufferedfromlowdurability.Designandmaterialimprovements,includingimprovedfabricationtechnologies,haveincreasedthedurabilityandfatigueresistanceofpolymericvalves,butthereisstillhighvariabilityinlife-cycle[ 5 ].Complicationsduetothrombosisoccurbetween1.5%and3%peryearforcurrentmechanicalandbioprostheticvalves[ 9 ].Polyurethanevalveshaveshownalowerthrombogenicityratethanbileaetmechanicalvalves,andaPCUvalvehasacomparablethrombosisratetobioprostheticvalvesinanimalstudies[ 2 10 ].Polymericvalveshaveshownsimilarowcharacteristicstobioprosthesesandhavegoodhaemodynamicproperties.Bloodcelladhesionandsubsequentthrombusformationisamajorconcernfacingthedevelopmentofreplacementheartvalves[ 11 ].Theprojectobjectiveistodevelopapolymericheartvalvecapableofreplicatingamoldsurfacetexturethatreducesthrombosformation.Theinitialstudyfocusesonmedical-gradesilicone(additiontype,platinumcatalyst)andinvestigatestheeffectofsurfacetextureandroughnessonbloodcelladhesiontosiliconeheartvalveleaetsprocessedusingmagneticabrasivenishing(MAF).MAFwaschosenforitsabilitytoimpartvarioussurfacetexturesatvaryingsurfaceroughnessvaluesinasinglesetupbychangingtheprocessparameters.Siliconeisinertandbiostable,easytomanufacture,andhasshowntheabilitytoreplicatesurfacefeaturesatthemicro-andnanometerscales[ 12 ].ThedesireistoidentifytheroughnessandtextureproducedbyMAFthatmostreducesbloodcell 16

PAGE 17

adhesionandsubsequentthrombusformation.Studiesinvestigatingtheeffectsofsurfacetextureoncelladhesionarewellestablishedandcells,including,butnotlimitedtoplateletsandredbloodcells,broblasts,andbladdersmoothmusclecells,haverespondedtochangesinsurfacetextureandroughness[ 13 18 ].RoughnessandtopographyhavealsobeenshowntoaffectthesettlementofUlvazoosporestosilicone[ 19 20 ].Duetothesizeoftheplatelets(1-3m),itisdesiredthattheleaetshavearoughnessoflessthan1mRztopreventtheplateletsfrombecominglodgedbetweenthepeaks.Forthisstudy,themoldroughnesswasvariedfrom2-3mto0.1mRz. 17

PAGE 18

CHAPTER2SILICONEVALVEMANUFACTURING 2.1DesignConsiderationsAccordingtoGhanbarietal.thedesignrequirementsfordevelopingaheartvalveprostheticarelistedbelow[ 10 ]: 1. Mustthostanatomy 2. Mustprovideminimumresistancetoforwardow 3. Mustpreventregurgitation 4. Shouldminimizedamagetobloodcellsandreducethrombogenicity 5. ShouldminimizestressandimprovedurabilityThedesignmustallowunobstructedforwardowattheinitializationofaminimumtransvalvularpressurefollowedbyvalvesealingattheappropriatebackpressure.Additionally,thevalveshouldminimizeblooddamage,thrombogenicity,andstressonthevalvecomponents.Ithasbeenshownthatthedesignandgeometryofthevalveleaetsinuencesthefunctionofthevalvesuchasopening,closing,effectiveoricearea(EOA),leakage,andstress;themanufacturingprocessaffectsvalvedurabilityandhaemodynamicfunction[ 10 ]. 2.2ValveSpecications 2.2.1LeaetThicknessLeaetthicknessisamajorcontributortovalvedurability.Cyclefatiguetestingfeaturingpolyurethanevalvesshowedthatleaetslessthan50mthicklastedlessthan100millioncycles,100mthickleaetslasted600millioncycles,leaetsbetween150-200mthicklastedaround800millioncycles,and300mthickleaetslasted1billioncycles[ 2 ].Subsequently,ourtargetleaetthicknesswillbebetween200-300m. 2.2.2FactorsthatInuenceCellularAdhesionPartofthedesignistoimpartinthevalveasurfacetexturethatreducesbloodcelladhesionandthromgenicity.Surfacecharacteristicshavebeenshowntoaffectthe 18

PAGE 19

biocompatibilityofmaterialsandcancontrolinteractionsbetweenthematerialanditsenvironment.Propertiesthatinuencecellularadhesionarelistedbelow. 1. Hydrophobicity/hydrophilicity 2. Surfaceenergy 3. Morphology 4. TopographySurfacetopographycanbeintroducedtoamaterial,intentionallyoraccidentally,affectingcellularadhesion,selectivecellproliferation,differentiation,andapoptosis[ 10 ].Commonfabricationtechniquesusedtointroducepatternsortopographyincludephotolithography,casting,ablation,andembossing.Foragivenpattern,featuredepth,width,orientation,andfrequencyarethoughttoaffectcellresponse[ 21 ].Modicationstothesurfacecanalterbiocompatibilitywithoutaffectingbulkmaterialproperties.Studieshaveshownthatbroblaststendtoalignalongridgesorparalleltocollagenbers[ 22 ].Plasmaimmersionionimplantationhasbeenusedtoimprovecellattachmentonsyntheticmaterialsurfacesandcholesterol,andpeptidemodicationhasbeenusedtoenhanceendothelialcellafnity.Additionally,endothelialcellattachmentcanmaskthevalvefrombeingconsideredaforeignbody,increasingbloodcompatibility.Creatingnanoscaletopographycouldincreasebiocompatibilityandhaemodynamicperformanceofpolymericheartvalves[ 10 ].Whenplateletsencounteranimplantedprostheticorforeignobject,theyrespondbyinducinghemostasis,whichistheprocessusedtostopbleedingataninjurysite,andcaninducethrombosisinaheartvalveprosthetic[ 23 ].Kuwaharaetal.usedaowchambertostudymorphologychangesinadheringplatelets.AnimageofthesemorphologychangesisshowninFigure 2-1 .Asplateletscomeintocontactwithasurfacetheybeginrollingbeforeenteringarmbutreversibleadhesion.Thenextstageintheadhesionprocessisspreading,whichisirreversible. 19

PAGE 20

Figure2-1. Morphologyofadheredplatelets[ 23 ] Parketal.studiedtheeffectofdifferentsurfacenishesonplateletadhesiontocommerciallypuretitanium[ 24 ].Plateletadhesionwasmeasuredusingalactatedehydrogenase(LDH)assayondualacidetched(DAE),abraded(320grit),machined,andpolishedtitanium.Scanningelectronmicroscopy(SEM)wasusedtoqualitativelyviewtheadheredplateletsafterxingwithglutaraldehyde.Allthesurfaceswereextensivelysonicatedinacetone,deionizedwater,andethanolpriortocontactwithplatelet-richplasma.Figure 2-2 showsSEMimagesofadheredplateletsonthefourtitaniumsurfaces;intheimagesthescalebaris6m.TheabradedandDAEsurfacesproducedbetween85000and95000adheredplateletspersquaremillimeter,whilethemachinedandpolishedsurfacesproducedbetween45000and42000adheredplateletsoverthesamearea.Thedifferencesbetweenthesurfacesweresignicant(p<0.05)suggestingthattopographyandsurfaceroughnessofthetitaniuminuencesplateletadhesion.Milneretal.texturedpolyether(urethaneurea)(PUU)withsub-micronpillars(700and400nm)viaatwo-phasemoldingprocesstodecreaseplateletadhesionbychangingsurfacetopographywithoutaffectingthesurfacechemistryofthesubstrate[ 14 ].Sampleswereplacedinarotatingdisksystemthatradiallyvariedtheshearstress.Theresultsfromthestudysuggestthatplateletadhesion,andtheeffectofsurfacetextureonplateletadhesion,ishighestatlowshearstresses. 20

PAGE 21

Figure2-2. SEMimagesofadheredplateletsto(i)DAE,(ii)abraded,(iii)machined,and(iv)polishedtitaniumafter30minofexposuretoplateletrichplasma[ 24 ] Valvularshearstressvarieswidelybasedontheowconditions,measurementmethod,andpositionofmeasurement,withstudiesindicatingthestressesrangefrom3-180Pa[ 25 28 ].Thelargestshearstressesareoftenrecordeddownstreamofthevalveduetothehighvelocityjetsproducedfromtheclosingvalveleaets;veryfewstudieshaveaccuratelymeasuredorestimatedtheshearstressespresentontheleaets.Westonetal.attemptedtoestimatetheshearstressesonanaorticvalveleaetsurfaceduringpeaksystoleundersteadyowconditions[ 25 ].Themaximumshearstressrecordedinthestudywas79dyne/cm2(7.9Pa)foraowrateof22L/minatthetipoftheleaet;theminimumshearstressrecordedwasjustunder1Pa,2mmupstreamoftheleaettip.Themostcommonshearstressfoundonthevalveleaetwasapproximately3Pa,locatedatvariouspositionsovertheleaetsurfaceundermultipleowconditions.Table 2-1 liststhelocationsandmagnitudesofshearstressespresentontheleaetsurface.The0positionrepresentstheleaetcommissure,60identiesthecenteroftheleaet,and30lieshalfwayinbetween.Additionally,the0mmpositionidentiestheleaettip,whileallpositivevaluesindicateapositionupstreamofthetip.Fromthese 21

PAGE 22

studies,thephysiologicallyrelevantshearstressrangeforbloodcelladhesionontheleaetswouldbebetween1-3Pa. Table2-1. Leaetshearstress[ 25 ] PositionfromleaettipOrientationFlowrateShearstress(mm)(deg)(L/min)(Pa)030227.97.63022646022,1534-7022,1530-70,607.51 2.3ValveManufacturingProcessDipcasting,lmfabrication,andinjectionandcavitymoldinghavebeenusedtomanufacturevalves,withthepolymerchoiceoftendictatingthemanufacturingprocess.Intrials,dipcastingwasshowntocreatevalveswithbetterdurabilityandhaemodynamicperformancethanvalvesmadeusinglmfabrication[ 10 ].Anewfabricationmethodisproposedtocreateasiliconeheartvalvethatwillfeatureatrileaetdesign,withtheleaetsspacedevenlyaroundthecircumferenceofthevalveat120.AschematicoftheproposedvalveisshowninFigure 2-3 .Tocreatethisvalve,auniquemanufacturingprocesswasdevelopedandisshowninFigure 2-4 .First,siliconeiscuredinacylindricalmoldthatisbetween30and35mminlengthcreatingasiliconetubethatis200-250mthick.Next,thenewlyformedsiliconetubeispulledawayfromthemoldandandadheredtogetheratthreelocations120apart.Theleaetsarethenpushedbackintothemoldandadheredtotheinnerwallofthesiliconetubetoformthevalve.Thecurrentvalveisdesignedtobe22mmindiameterand10-15mmdeep;thefabricationprocesscanbealteredtocreatesmallerorlargervalves,asnecessary,byalteringtheinnerdiameterandlengthofthemold. 2.4ValveDurabilityTestingTostudyvalvedurability,adynamictestingsystemwasdesignedtosimulatethepulsatileowproducedbytheheartandtoprovidethevalvewithanappropriateamount 22

PAGE 23

Figure2-3. Schematicofproposedvalve Figure2-4. Siliconevalvemanufacturingprocess 23

PAGE 24

ofbackpressure.AblockdiagramofthesystemisshowninFigure 2-5 .Inthecurrentiterationofthedesign,apneumaticpumpisconnectedinlinewiththevalvealongwithtwopressuretransducerspositionedbeforeandafterthevalve.Aowmeterisplacedbetweenthevalveandthereservoirtorecordtheowexitingthevalve.Areservoirisheldabovethevalveanduidislledtoaline540mmabovethevalvecenterline.Thiscorrespondstoaconstantbackpressureof40mmHg.ApictureofthecompletedsetupisshowninFigure 2-6 Figure2-5. Blockdiagramofdynamictestingsystem Duringtesting,prototypesaresecuredintothetestingsystemwiththecuspsorientedsothatthesuppliedbackpressurecausestheleaetstocloseandseal.Witheachforwardstrokeofthepump,uidpulsesthroughthevalve.Fortesting,duetolimitationsofthecurrentpump,20mLofuidwaspumpedat50pulse/min.ThepressurewaveswererecordedandareshowninFigure 2-7 .Apeakpressureofapproximately120mmHgisachievedandthevalvesareabletoholdapressureof40mmHgwithoutleakage. 24

PAGE 25

Figure2-6. Photographofdynamictestingsystem Figure2-7. Pressurewave 25

PAGE 26

CHAPTER3MOLDFABRICATIONUSINGMAGNETICABRASIVEFINISHING 3.1OverviewofMagneticAbrasiveFinishingThemagneticabrasivenishing(MAF)processwaschosenforitsabilitytocreatevarioustextureswithvaryingsurfaceroughnessusingasinglenishingsetup.Additionally,ithastheabilitytonishfree-formsurfacegeometries[ 29 32 ].MAFusesamagneticeldtoactuatemagneticabrasiveagainstaworkpiecesurface.Magneticabrasivesarecomposedofferrousabrasiveparticlesorferrousparticlescombinedwithlooseabrasive.Themagneticux,producedbyeitherelectromagnetsorpermanentmagnetsafxedtothenishingsetup,inuencesthemotionofthemagneticabrasiveagainsttheworkpiecesurfacewithoutevercontactingthesurface.Anishedsurfaceisproducedwhenrelativemotionisachievedbetweenthesurfaceandthemagneticabrasive,whichispressedagainstthesurfacebythemagneticforce.MAFproducesahigh-precisionsurfacenishthatcanbeadaptedin-processtoaltertheroughnessortextureforagivenareabychangingtheprocessparameters.Theprocessisconsideredpressure-copying,allowingmanycomplex,mostlyinaccessiblesurfacesbyconventionalnishingoperationstobepolished[ 33 ].Yamaguchietal.usedrotatingandtranslatingpermanentmagnetstonishtheinternalsurfacesofbenttubes[ 30 ]. 3.2ProcessingPrincipleAschematicofthenishingprocessisshowninFigure 3-1 .Magneticabrasiveisheldagainstthemoldsurfaceaccordingtothemagneticeldlinesproducedbythepermanentmagnetsconguredonthemagneticyoke.Asthemoldrotates,themagneticabrasiveisheldatthenishingsite;whenrelativemotionisachieved,materialisremoved.Ideally,themotionofthenishingunitalsocausesfreshabrasivetobeintroducedtothesurface,providingmoreefcientmaterialremoval.Toachieverelativemotion,themagneticforceholdingthemagneticabrasiveagainstthesurfacemustbe 26

PAGE 27

largerthanthefrictionforcebetweenthesurfaceandabrasive.Themagneticforceactingonthemagneticabrasivecanbecalculatedusingthefollowingequation:F=VHrH (3)whereFisthemagneticforce,Visthevolumeofthemagneticabrasive,isthemagneticsusceptibilityoftheabrasive,Histhemagneticeldstrength,andrHisthegradientofthemagneticeld.Thematerialremovalcorrespondstotheamountofabrasivepressedagainstthemoldsurfaceandtheprocessparameters. Figure3-1. Schematicofmagneticabrasivenishing(MAF)processingprinciple 3.3FinishingMachineDevelopment 3.3.1DesignandBuildTonishtheinternalmoldsurface,amachineneededtobedeveloped.Basedonpreviousmachinesdesignedforsimilarnishingprocesses,thedesignneededtobeabletoholdaminimumoffourpermanentmagnetsorientedat90angles,rotatetheheartvalvemold,andoscillatethenishingunitrelativetotherotatingmold. 27

PAGE 28

Acomputer-aideddesign(CAD)modelofthenishingmachinewascreatedandisshowninFigure 3-2 .Thedesignfeaturesonemotortoproducemoldrotationandasecondmotortooscillatethenishingunitwhichhouseseightpermanentmagnets,groupedinnorth-south(N-S)pairsontheyoke,secured90apartviaalinkandshaftcoupling.Thenishingunitismountedatopalinearslidethat,combinedwiththelinkandshaftcoupling,convertsthemotorrotationintolinearvibrationalongthemoldaxisdirection.Operatedinunison,themotorsarecapableofproducingthemotionrequiredforthenishingprocessoutlinedinFigure 3-1 Figure3-2. Designofnishingmachine AphotographofthecompletednishingmachineisshowninFigure 3-3 .ThemachineisoperatedbythecontrolsshowninFigure 3-4 .Initially,thedesignwasfortwoouterchuckstobeconnectedviaabeltandpulleysystemtothecentershaft.Duetomachinevibrationcausedbytensioninthepulleysandinstabilityinthebearings,thedesignwasreducedtoasinglecenterchuckandshaftcollarswereaddedtostabilizethecenterbearing.Thismodicationreducedtheunwantedvibrationsandwasthenalmodicationtothemachine. 28

PAGE 29

Figure3-3. Photographofnishingmachine Figure3-4. Photographofnishingmachinecontrolbox 29

PAGE 30

3.3.2MagneticFluxDensityTocharacterizethemagneticeld,aHallprobewasusedtomeasurethemagneticuxdensity.AdiagramoftheHallprobeandthemeasurementsetupisshowninFigure 3-5 .Twomeasurementsweremadeoftheuxbetweenthemoldandthemagnetface.Intherstmeasurement,magneticparticleswereintroducedintothemold;inthesecondmeasurement,noparticleswerepresent.TheHallprobewasinsertedinthe1mmgapbetweenthemoldandthemagnetwiththesensorfacingthemagnet.Themeasureduxwasreducedbymorethanhalf.TheresultsofthemeasurementsareshowninTable 3-1 .Themeasurementcontainingthemagneticabrasiveproducedthelargestamountofux,whentheabrasivewasremovedtheuxreducedbymorethanhalf. Table3-1. Maximummagneticuxdensity MagneticabrasivepresentMagneticabrasiveabsent 0.7T0.235T 3.4MoldSurfaceAnalysisTwosurfaceanalysismethodsemployedinthisstudyincludeusingastylus-typesurfaceprolerandanopticalsurfaceproler.Tocharacterizethemoldsurfacesastylus-typeprolerwasusedandispicturedinFigure 3-6 .Totakeameasurement,thestylusismovedoverthesurfaceforthespeciedevaluationlength(maximumdistanceof1inch)whilethediamondstylus(tipradius:2m)encountersthesurfacefeatures.Theverticaldisplacementofthetipcreatesananalogsignal(forcedetector:0.75mN)whichisconvertedtoadigitalsignalandanalyzedtocreatea2-dimensionalsurfaceprole.TheprocessingunitcanbecontrolledfromacomputerviaaRS-232cable,importingtheproleforanalysis.AdiagramofasurfaceprolemeasurementisshowninFigure 3-7 .Fromthisprole,thepeak-to-valleydistanceRz,averageroughnessRa,androot-mean-square(rms)roughnessRqcanbecalculated.AnexampleofasurfaceproleisshowninFigure 3-8 .Thepeak-to-valleydistanceisthelargestdistance 30

PAGE 31

Figure3-5. Diagramofmagneticeldanalysisfeaturing(i)aHallProbemeasuringthemagneticuxdensityproduced(ii)withferrousparticlesand(iii)without betweenthehighestandlowestpointsoftheproleforagivenevaluationlength.Theaverageroughnessiscalculatedastheaveragedistanceoftheprolefromthecenterlineandthermsistakenastheroot-meansquareoftheproledistancefromthecenterline.AllmeasurementsusedaGaussianband-passlterfollowingtheISO4287:1997standard.Thehigh-frequencylterisbasedonthediamondtipradius,whilethelow-frequencylterisdeterminedbythecut-offlengthdenedinthestandard.Allmeasurementswereperformedrstwiththelongercut-offlengthbeforebeing 31

PAGE 32

Figure3-6. Photographofsurfaceroughnessproler Figure3-7. Diagramofsurfaceprolemeasurement 32

PAGE 33

Figure3-8. Diagramofsurfaceprole stepped-downiftheroughnesswasfoundtobelessthantheminimumvaluefortheperformedevaluation.Table 3-2 liststheconditionsusedforeachmeasurementinthisreport.Themeasurementspeedwasalsoreducedwitheachsteptoincreasethemeasurementresolutionatlowerroughnessvaluesandshorterevaluationlengths.Foreachmold,fourmeasurementsweretakenrandomlyaroundthecircumferenceofthemoldandaveraged. Table3-2. InternationalOrganizationforStandardization(ISO)standardsandmeasurementconditions Cut-offlengthEvaluationlengthRoughnessrangeMeasurementspeed(mm)(mm)(mRz)(mm/s) 0.840.5-100.50.251.250.1-0.50.1 3.5MoldSurfaceFinishingCharacteristics 3.5.1UnnishedMoldSurfaceFortheheartvalvemold,abrass260seamlesstubewasused.Theinitial,unnishedmoldsurfacehadaroughnessbetween2-3mRz.RepresentativeroughnessvaluesfortheunnishedmoldareshowninTable 3-3 .ArepresentativemoldsurfaceroughnessproleisshowninFigure 3-9 .Theunnishedmoldsurfaceis 33

PAGE 34

labeledMold1andcorrespondstoLeaet1studiedinthebloodcelladhesiontests.Thisnamingsystemwillcontinueforeachmoldgeneratedbynishing. Table3-3. Unnishedmoldsurfaceroughness RoughnessparametersMold1 Rz(m)2.59Ra(m)0.34Rq(m)0.38 Figure3-9. Unnishedmoldsurfaceroughnessprole(Mold1) 3.5.2MoldSurfaceFinishedwithLooseDiamondAbrasiveThreemoldsweregeneratedusingloosediamondabrasivetostudybloodcelladhesionatvarioussurfaceroughnesses.ThenishingconditionsforthesemoldsarelistedinTable 3-4 .AllofthemoldsnishedusingtheMAFprocesswererotatedat2000min-1whilethenishingunitvibratedatanamplitudeof5mmatafrequencyof1.33Hz.Duetothehighrotationalspeed,thenishingdirectionthedirectionofcuttingmarksproducedduringnishingisatanangleof0.17relativetothetubesradialdirection;adiagramofthenishingdirectionisshowninFigure 3-10 .Foreachmold,2.7gof330mmeandiameterelectrolyticironwascombinedwith0.3gofdiamondabrasive.Thediamondabrasiveusedwas0-1mpowder,forMold2,and0-0.5mpaste,forMolds3and4.Molds2and3werenishedfor5minwhileMold4wasnishedfor20min.TheroughnessofeachmoldafternishingispresentedinTable 3-5 .Tobettervisualizethenishingresultsthepeak-to-vallyroughnessofeachmoldispresentedinFigure 3-11 34

PAGE 35

Table3-4. Diamondabrasivenishingconditions Mold2Mold3Mold4 Mold260Brasstube(?25.4?2290mm)Moldrotation2000min-1PoleNd-Fe-Brare-earthpermanentmagnet(25.412.712.7mm)PolemotionAmplitude:5mm;Frequency:1.33Hz(80min-1)Clearance1mmIronparticlesElectrolyticironparticles:2.7g(330mmeandiameter)AbrasiveDiamondabrasive:0-1m0-0.5m0-0.5m0.3gpowderpastepasteLubricantSoluble-typebarrel3mL3mL4mLnishingcompoundFinishingtime5min5min20min Figure3-10. Diagramofnishingdirection Table3-5. Roughnesscomparisonofmoldsnishedwithdiamondabrasive RoughnessMold2Mold3Mold4 Rz(m)0.280.790.11Ra(m)0.040.060.01Rq(m)0.050.110.02 35

PAGE 36

Figure3-11. Roughnesscomparisonofdiamondabrasivemolds AcomparisonoftheroughnessprolesispresentedinFigure 3-12 .Mold3,nishedfor5minwith0-0.5mdiamondpaste,hadtheroughestsurfacewithlargefeaturespresentfromtheoriginalsurfaceremaining.ForMolds2and4,theprolesareviewedatareducedscaleinFigure 3-13 .Afternishingfor5minwith0-1mdiamondpowder,Mold2hadthegreatestreductioninroughnessfortheshortnishingtime.Largegapsexistbetweenthepeakscorrespondingtothelargerabrasivesize.Mold4wasnishedforthelongesttimewiththesmallerabrasiveproducingthesmoothestsurface,featuringeven,shortcuttingmarks. 3.5.3MoldSurfaceFinishedwithCompositeMagneticAbrasiveThreemoldsweregeneratedusingcompositemagneticabrasivesforthebloodcelladhesiontrials.Thenishingconditionsforthemoldproducedusingwhitealumina(WA)magneticabrasive(Mold5)arelistedinTable 3-6 .Mold5wasnishedfor5minwith2.4gof330mmeandiameterelectrolyticironcombinedwith0.6gofthemagneticabrasive(80mFe;<10mWA).ThenishingconditionsforthemoldsproducedfromthediamondmagneticabrasivearelistedinTables 3-7 and 3-8 forMolds6and 36

PAGE 37

Figure3-12. Roughnessprolesofdiamondabrasivemolds 37

PAGE 38

Figure3-13. Roughnessprolesofdiamondabrasivemoldsatreducedscales 7,respectively.Mold6wasnishedfor15minwith2.4gof330mmeandiameterelectrolyticironcombinedwith0.6gofthediamondmagneticabrasive(0-0.75mFe;0-0.25mdiamond).Mold7underwentthesamenishingprocessasMold6withanadditionalnishingphase(Table 3-8 )of5minwith3gofthediamondmagneticabrasiveandnoironparticles.Allofthemoldswererotatedat2000min-1whilethenishingunitvibratedatanamplitudeof5mmatafrequencyof1.33Hz.TheroughnessofeachmoldafternishingispresentedinTable 3-9 .TheresultsarepresentedgraphicallyinFigure 3-14 .AcomparisonoftheroughnessprolesispresentedinFigure 3-15 atthesamescaleastheunnishedmoldsurface.Mold5,nishedfor5minwithWAmagneticabrasive,hadtheroughestsurfacewithdeepgougescreatedfromthelargeabrasive;Molds6and7producedevensurfacenishes.Theprolesareshown 38

PAGE 39

Table3-6. Compositemagneticabrasivenishingconditions:WAmagneticabrasive(Mold5) Mold260Brasstube(?25.4?2290mm)Moldrotation2000min-1PoleNd-Fe-Brare-earthpermanentmagnet(25.412.712.7mm)PolemotionAmplitude:5mm;Frequency:1.33Hz(80min-1)Clearance1mmIronparticlesElectrolyticironparticles:2.4g(330mmeandiameter)AbrasiveWAmagneticabrasive:0.6g(80mFe;<10mWA)LubricantSoluble-typebarrelnishingcompound:3mLFinishingtime5min Table3-7. Compositemagneticabrasivenishingconditions:diamondmagneticabrasive(Mold6) Mold260Brasstube(?25.4?2290mm)Moldrotation2000min-1PoleNd-Fe-Brare-earthpermanentmagnet(25.412.712.7mm)PolemotionAmplitude:5mm;Frequency:1.33Hz(80min-1)Clearance1mmIronparticlesElectrolyticironparticles:2.4g(330mmeandiameter)Abrasivediamondmagneticabrasive:0.6g(0-0.75mFe,0-0.25mdiamond)LubricantSoluble-typebarrelnishingcompound:3mLFinishingtime15min Table3-8. Compositemagneticabrasivenishingconditions:diamondmagneticabrasive,secondnishingphase(Mold7) Mold260Brasstube(?25.4?2290mm)Moldrotation2000min-1PoleNd-Fe-Brare-earthpermanentmagnet(25.412.712.7mm)PolemotionAmplitude:5mm;Frequency:1.33Hz(80min-1)Clearance1mmAbrasivediamondmagneticabrasive:3g(0-0.75mFe,0-0.25mdiamond)LubricantSoluble-typebarrelnishingcompound:3mLFinishingtime15min Table3-9. Roughnesscomparisonofcompositemagneticabrasivemolds RoughnessparametersMold5Mold6Mold7 Rz(m)0.640.180.18Ra(m)0.060.030.02Rq(m)0.090.030.03 39

PAGE 40

Figure3-14. Roughnesscomparisonofcompositemagneticabrasivemolds atareducedscaleinFigure 3-16 .Mold6and7hadsimilarroughnessvalues,butMold7appearstohaveshorterpeak-to-peakdistancesthanthesurfaceproducedbyMold6correspondinglyMold7hadaslightlysmalleraverageroughness,0.02mRacomparedto0.03mRaforMold6.Themoldsurfaceroughnessiscontrolledbythesizeoftheabrasiveandthenishingtime.Mold5wasproducedusinglargerabrasivethanMolds6and7andwasnishedforashorteramountoftime.Theroughnessvaluesofthethreemoldsproducedusingthecompositemagneticabrasivesaresimilartothoseproducedusingtheloosediamondabrasive,allowingforcomparisonstobemadeontheeffectofabrasivechoiceandtextureonbloodcelladhesion. 40

PAGE 41

Figure3-15. Roughnessprolesofcompositemagneticabrasivemolds 41

PAGE 42

Figure3-16. Roughnessprolesofcompositemagneticabrasivemoldsatreducedscales 42

PAGE 43

CHAPTER4SILICONELEAFLETFABRICATION 4.1ControllingLeaetThicknessInitially,thesiliconewascuredinthemoldwhilethemoldwasinaverticalorientation,showninFigure 4-1 .Duetothelowviscosityofthemedical-gradesilicone,theleaetswereunevenandlostasignicantamountofsiliconeduringthecuringprocess.Toremedythis,itwasdecidedtorotatethemoldinahorizontalorientationtocounteracttheeffectsofgravityandreduceleaetvariation.ThecuringsetupisshowninFigure 4-2 .Todevelopleaetsfromthemolds,themoldswerecutto35mmandsiliconewasappliedtothemold,whichwasthensecuredinthechuckandrotatedfor24hr. Figure4-1. Moldcuringorientations Tostudytheeffectofthemoldrotationonleaetthickness,ameasurementmethodwasdevised,Figure 4-3 ,thatallowedthethicknesstobemeasuredacrosstheentireleaet.Aftercuring,thesiliconetubewasremovedfromthemoldandcutopendownitslengthcreatinga7035mmleaet.Theleaetwasthencutinto6strips,5mmwide.Alongeachstrip,13measurementsweremadeusingdigitalverniercalipers.Atotalof78measurementsweremadeacrossthesurfacetomeasuretheleaetthicknessandvariation.Therotationalspeedwasvariedfrom75to1100min-1andtheamountof 43

PAGE 44

Figure4-2. Diagramofcuringprocess appliedsiliconewaseither0.6or0.9g.Thesevalues,basedonthespecicgravityofthecuredsilicone(1.09),wereexpectedtoproduceleaetsbetween0.22and0.34mthick.TheresultsfromtherotationtrialsareshowninFigure 4-4 .Leaetvariationwasquantiedbyusingthestandarddeviationofthemeasurements.LookingatFigure 4-4 ,thereisahighamountofthicknessvariationatlowerspeeds(<200min-1),whichdecreasesasthespeedisincreased.Aminimumisreachedbetween700and1100min-1ofapproximately20%ofthetotalleaetthickness.Atthesespeeds,leaetsproducedusing0.6gofsiliconehadthicknessvaluesatorbelow200m,whileleaetsproducedusing0.9gofsiliconehadthicknessvaluesrangingfrom200m(1100min-1)to250m(700min-1).Itwasobservedthatathigherrotationalspeeds,siliconewasejectedfromthemoldreducingleaetthickness.Fromtheseresults,curingconditionswerechosentobe0.9gat700min-1.TheleaetcuringconditionsthatproducedthemostdesireableresultsareprovidedinTable 4-1 ,theseconditionswereusedonallsubsequentleaets.Todevelopleaetsforanalysisandbloodcelladhesiontesting,theleaetswereprocessedasshowninFigure 4-5 .Aftercuring,theleaetswereremovedfromthemoldasasiliconetubeandcutdowntheirlengthandopened.Theedgesoftheleaetwere 44

PAGE 45

Figure4-3. Diagramofthicknessvariationmeasurement Figure4-4. Theeffectofmoldrotationonleaetthicknessvariation 45

PAGE 46

Table4-1. Leaetcuringconditions Molddimensions?25.4?2235mmSiliconeMedicalgradeWeightamount0.6-0.9gCuringtime72hr(rotating:24hr)Moldrotation700min-1 trimmed(7025mm)andmountedtoastandardmicroscopeslide(7625.41mm)usingmedical-gradesiliconeadhesive. Figure4-5. Leaetprocessing 4.2SiliconeLeaetMoldReplicationToexaminethereplicabilityofsilicone,theleaetsurfacewasreplicatedfromnishedbrassmoldsandviewedunderanopticalsurfaceproler,whichusesscanning 46

PAGE 47

whitelightinterferometry.TheopticalprolercreatesamapofthesurfaceundermagnicationandusesthedatatocalculateRz,Ra,andRq.Forsiliconeleaets,the20magnicationwasused,measuringanarea176133m.AdiagramoftheopticalprolermeasurementisshowninFigure 4-6 .ThemeasurementconditionsarelistedinTable 4-2 ,andtheprocessingconditionsarelistedinTable 4-3 .Toapplyaltertothedata,a100100mmaskwasappliedtothetotalmeasuredarea.AGaussianbandpasslter(low:20m,high:0.828m)wasthenapplied. Figure4-6. Diagramofopticalprolermeasurement Table4-2. Opticalprolermeasurementconditions Numberofaverages3FDAresolutionHighScanlength20m Leaets6and7,weremeasuredusingtheopticalproler.TheresultsareshowninTable 4-4 .Toillustratetheeffectsandlimitationsofthesiliconeleaetreplication,acomparisonwasmadebetweenMolds6and7(producedfromthediamondmagneticabrasive)andtheirrespectiveleaetsinFigure 4-7 .WhileMold6andMold7have 47

PAGE 48

Table4-3. Dataprocessingconditions Removed4thorderwavinessMasked100100mFilterGaussianbandpassLowfrequencylter20mHighfrequencylter0.828m similarroughnessvalues(0.18mRz)theireffectonthedevelopedleaetsisdrasticallydifferent.Leaet6,producedfromMold6,haslong,evenlyspacedridgeswhosedirectioncorrespondstothemoldnishingdirection(Figure 3-10 ).Leaet7hasnovisibleridges.Bothleaetshavesimilarpeak-to-valleydistances,1.18mcomparedto1.14mforLeaets6and7,respectively,buttheaverageroughnessofLeaet7islessthanhalfofLeaet6.Basedontheleaet'sobliqueplotsthisistobeexpected;becausethepeak-to-valleydistanceisthegreatestdistanceovertheentiremeasuredarea,thespikelocatedonLeaet7skewstheroughnessvalue. Table4-4. Siliconesurfaceroughnesscomparison RoughnessLeaet6Leaet7 Rz(m)1.1821.138Ra(m)0.4350.192Rq(m)0.7930.138 Thedisparitybetweenthetwosiliconeleaetsliesinthedifferenceinpeak-to-peakdistancebetweenfeaturesonthetwomolds.Duetothesurfacetensionofthesilicone,ifthepeak-to-peakdistancebecomestoosmallthesiliconewillnotcompletelyreplicatethesurfacefeature.ThiseffectisshowninFigure 4-8 andcanbecontrolledbytheabrasivesizeandthenishingprocessparametersinadditiontothechoiceofpolymer.Inthisstudy,thefocushasbeenonthepeak-to-valleydistanceinanefforttoproducesurfacetexturesthatpreventplateletsfrombecominglodgedinsurfacefeatures.Sincethepeak-to-valleydistanceisthedistancebetweenthehighestpeakandthelowestvalley,thisparameterisgreatlyinuencedbyrandomspikesandsurfacefeaturesandmaynotaccuratelyreecttheactualsurfacetexture.Amoreaccuratepredictormaybetheaverageroughness,whichaccountsfortheentiremeasuredsurfacearea. 48

PAGE 49

Figure4-7. Comparisonofmoldsurfaceandreplicatedsiliconeleaet Figure4-8. Limitationofsiliconemoldreplication 49

PAGE 50

CHAPTER5BLOODCELLADHESIONTESTING 5.1FlowChamberDevelopmentInSection 2.2.2 itwassuggestedthatthephysiologicallyrelevantshearstressrangeforbloodcelladhesionontheleaetswouldbebetween1-3Pa.Theshearstress,ormoredirectly,theshearratecanbecontrolledbytheowrateoftheuidandthegeometryoftheowchamber.Flowchambersarewidelyusedtostudycelladhesion[ 13 15 18 34 ];theconceptispresentedinFigure 5-1 .Asuspensionentersthroughthetopplateattheentranceportandoodsthesamplesurfacebeforeexitingthechamberthroughtheexitport.TheheightHandwidthWintheinnerchamberinuencestheshearrateinducedonthesuspensionforagivenowrate. Figure5-1. Flowchamberconcept TheexperimentaldesignforbloodcelladhesiontestingisdetailedinFigure 5-2 .Bloodwouldbeushedovertheleaetsurfaceinasinglepassandcollectedafterexiting.Theleaetwouldneedtobesecuredinaclearowchamberthatcouldbesealedanddisassembledeasilyandprovidethecellswiththeappropriateshearstress.ACADmodelofaowchamber(Figure 5-3 )wascreatedthatwouldallowtheleaet,mountedonamicroscopeslide,tobeviewedunderdynamicowconditions. 50

PAGE 51

Figure5-2. Experimentaldesign Theowchamberwasfabricatedwithacentralpockettohousetheleaetandcouldbesealedusingvacuumgreaseand6M6bolts.Thecompletedowchamber,picturedinFigure 5-4 was1507025.4mmandfeaturedanopeningonthebottomplatetogivethemicroscopeobjectiveclearanceforlocatingandfocusingonthesurface.Theowchamberwasdesignedtohaveanopeningarea,showninFigure 5-5 ,of0.225.4mm.Themicroscopeslideis1mmthickandtheleaetthicknessrangedfrom200-250m.Basedontheowchamberopeningarea,aowrateof50mL/hrusinga20mLsyringewaschosen.Theowrateisconvertedtoashearrate(219s-1)andthentoashearstress(0.83Pa)byusingthedynamicviscosityofblood(0.038mPas);aowrateof100mL/hrcorrespondstoashearrateof438s-1orashearstressof166Pa.A 51

PAGE 52

Figure5-3. Designofowchamber calculationisincludedinAppendix B .FlowratesandtheircorrespondingshearstressesinthegivenrangeareshowninTable 5-1 Table5-1. Bloodowrateandshearstress Flowrateofblood(mL/hr)Shearrate(s-1)Shearstress(Pa) 502190.831004381.661506562.492008753.32 5.2ExperimentalTestingSetupPriortotesting,alltheleaetswerecuredforaminimumof72hrinadvanceandmountedonthemicroscopeslide24hrpriortotestingtoallowthesiliconeadhesivetocure.Theowchamberwasthenpreparedforthersttrial.Vacuumgreasewasrstappliedtotheowchamberseatlocatedonthebottomplate,showninFigure 5-6 ;the 52

PAGE 53

Figure5-4. Photographofowchamber Figure5-5. Flowchamberopeningarea 53

PAGE 54

ledgemustbecompletelylledbeforetheleaetcanbepressedintoplace.Thebottomplatewasthenippedoverandvacuumgreaseliberallyappliedtothecontactingedgeofthemicroscopeslideandtheowchambertoeffectivelysealthebaseoftheowchamber.Athincoatingofthevacuumgreasewasnextappliedtotheinterfacebetweenthetopandbottomplates.Thetopplatewasthencenteredandpressedintoplacewiththe6boltsusedaslocatorswhilecompletingtheseal.Thecompletedowchamberwasmountedtothemicroscopestageusingtheowchamberclamp,showninFigure 5-7 ,andinletandexittubesmadeofNalgene50siliconewereattached.TheinlettubecanbeconnectedtoasyringeviaaLuer-Lokwhiletheexittubeallowstheuidtobecollectedina50mLtesttube. Figure5-6. Flowchamberpreparation Twohoursbeforetesting,thebloodwascollectedandbroughttothelabonice.Onceinthelab,a20mLsyringelledwith15mLofphosphate-bufferedsaline(PBS)wasconnectedtotheinlettubeandsecuredinthesyringepumptoushthesurface 54

PAGE 55

Figure5-7. Experimentalsetup andchecktheowchamberseal.Fortheinitialushing,theowratewassetto100mL/hrortwicetheexperimentalowrate.Whilethesurfacewasbeingushed,thebloodwasremovedfromthecoolerandpiercedwithadecanter,showninFigure 5-8 ;20-30mLofbloodweredrainedintoabeakerbeforedrawing15mLofbloodintoa20mLsyringe.Initially,thebloodwaskeptoniceduringalloftheexperiments,buttoextendtheblood'slife,aprotocolwasestablishedwhereafterllingtherstsyringe,theremainingbloodinthebloodbagwastobestoredintherefrigeratorbetween2and4C.Betweentrials,thebloodwouldbebroughtoutandpreparedforthenextleaetintimeforthebloodtobebroughtbacktonearlyroomtemperature(22C).Oncetheinitialushingwascomplete,thePBSsyringewasexchangedforonewithbloodtobegintheexperiment.TheexperimentalconditionsarelistedinTable 5-2 .Bloodwaspushedovertheleaetsurfaceinasinglepassat50mL/hrfollowedbyasinglepassofPBSatthesameowrate.Duringthisnalush,theleaetsurfacewasviewedusingaZeissAxiovert100microscopeanda33objective.Imageswere 55

PAGE 56

Figure5-8. Bloodbagspike recordedbyaCCDcameraandsimultaneouslyviewedonatelevisionmonitor.Whilerecording,differentareasofthesurfacewereviewedbymovingtheowchamber. Table5-2. Experimentalconditions Volumeofblood15mLNumberofpasses1Bloodowrate50mL/hrVolumeofsaline15mLSalineowrate50mL/hr Aftercompletingeachtrial,theleaetwasremovedfromtheowchamberandtheowchamberandallofthecomponentsweresoakedinbleachfor30min.Oncecleaned,thecomponentswererinsedwithPBSandanydisposablecomponentsandtubingwerediscarded.Fortheinitialtrials,thesametubingwasusedforallexperimentscompletedinasingleday,butinthenalprotocol,newtubingwastobeusedforeachtrial.Thesecondleaetwasmountedintheowchamberandtheprocedurewasrepeated.Effortsweremadetorunupto4trialseachday(eachtrialtakesapproximately2hrfromsetuptocleanup),butitwasobservedthatthebloodwouldbegintodrasticallydeteriorateafterthesecondtrialmakingitdifculttoviewtheleaetsurface.Subsequently,thetrialswerelimitedtotwoperday.AtablelistingthenumberofexperimentsperformedpersurfaceisshowninTable 5-3 56

PAGE 57

Table5-3. Numberoftrialspersurface SurfaceNumberoftrials Leaet12Leaet21Leaet31Leaet42Leaet52Leaet63Leaet73 5.3AnalysisofBloodCellAdhesionOncevideooftheleaetsurfacewascapturedandstoredonacomputer,agridwasoverlaidsectioningtheeldofviewinto16quadrants.AdiagramofthevideogridisshowninFigure 5-9 .Cellswerecountedineachquadrantandaddedtogethertogetthetotalnumberofcellsperarea.MultipleareaswereviewedoneachsurfaceandarelistedinTable 5-4 Figure5-9. Diagramofcellcountingmethod 57

PAGE 58

Table5-4. Numberofareasanalyzedpersurface SurfaceAreasanalyzed Leaet17Leaet23Leaet36Leaet47Leaet53Leaet65Leaet75 ThenumberofadheredplateletsandredbloodcellsforeachsurfaceispresentedinTable 5-5 ;thecompletedatacanbefoundinAppendix A .TheresultsareshowngraphicallyforplateletsandredbloodcellsinFigures 5-10 and 5-11 ,respectively. Figure5-10. Adheredplatelets Aftercountingthecellsforeachareaandnormalizingthenumbertothetotalarea,thestandarddeviationwascalculatedandaddedtothebiasusingtheroot-mean-squareerror(ErrorRMS)showninthefollowingequation:ErrorRMS=q 2Bias+2StDev (5) 58

PAGE 59

Table5-5. Numberofadheredbloodcells Leaet1Leaet2Leaet3Leaet4Leaet5Leaet6Leaet7 Platelets Average(Cells)650550462334576803664ErrorRMS13988496870439124Area(mm)0.220.17AdheranceIndex,AI1738014687123578919154102147117765ErrorRMS(AI)37172353131018181872117383316 Redbloodcells Average(Cells)1610.330.14710.2ErrorRMS7.410.50.48.71.70.4Area(mm)0.220.17AdheranceIndex,AI4282794187275ErrorRMS(AI)1982713112334511 59

PAGE 60

Figure5-11. Adheredredbloodcells whereBiasisthebiasimpartedbytheresearcherandStDevisthestandarddeviationbetweentheareas,whensquaredthistermbecomesthevariance.Thisisthetotalerrorforeachsurfacestudied.TheresultsfromthebloodcelladhesiontrialsarepresentedinFigures 5-12 and 5-13 correspondingtomoldsgeneratedfromloosediamondabrasiveandcompositemagneticabrasive,respectively.Themolds'surfaceRzroughnessvaluesareincludedforcomparison.Itcanbeseenfromtheguresthattheunnishedsurfacehasahighnumberofadheredplateletsandredbloodcells,whilemoldsnishedwithloosediamondabrasivehavefewadheredredbloodcells;thenumberofadheredredbloodcellsincrementallyreducesformoldsnishedwithcompositemagneticfollowingdecreasesinsurfaceroughness.Thenumberofadheredplatletsdoesnotfollowasimilartrend.Ofthemoldsproducedfromloosediamondabrasive,Mold3,whichhaslessadheredplateletsandredbloodcellsthanMold2,hasahigherroughness,0.79mRzcomparedto0.28mRzforMold2.Thisisexplainedbythesizeofabrasiveusedtonishthetwomoldsand 60

PAGE 61

thesizeofthesubsequentcuttingmarks.Figures 5-14 and 5-15 showacomparisonbetweenallofthenishedmoldsbasedonthesizeoftheabrasiveusedtonisheach.Themoldsarecomparedbasedona300msectionofeachmold'ssurfaceprole;thereducedscalewasusedtoexaminethesurfacetextureofthemold.FromFigure 5-14 ,Mold2,whichwasnishedwith0-1mdiamondpowder,haslargersurfacefeaturesthanMold3whichwasnishedwith0-0.5mdiamondpaste.Bothmoldswerenishedfor5min,withthelargerabrasiveremovingmorematerialandcreatingamoreevensurfacenishcomparedtothesmallerabrasivewhichwasunabletoremovesomeofthelargesurfacefeaturespresentontheoriginalsurface.Theseremnantfeaturesinuencetheoverallsurfaceroughness,butonlyoversmallareas,likethoseobservedduringthebloodcelladhesiontrials,thesmallercuttingmarksmadebythesmallerabrasivereducedbloodcelladhesion.ThistrendcontinuesastheMold4wasnishedwithsamesizeabrasive,0-0.5mdiamondpaste,butforalongerperiodoftimetocreateanevensurfacenishofthesmallercuttingmarks.Thissurfacehadthefewestnumberofadheredbloodcells.Allofthesurfacesproducedusingdiamondabrasivehadfewadheredredbloodcells,whilethesurfaceproducedbyWAmagneticabrasive,thelargestofthecompositeabrasivesstudied,hadalargenumberofadheredredbloodcellsandlessvisiblyadheredplateletsthanthesmootherdiamondmagneticabrasivemolds;thesurfacehadonlyslightlylessadheredplateletsthantheunnishedmold.Mold5,showninFigure 5-15 andnishedwiththeWAmagneticabrasive,haslargesurfacefeaturesrelativetothemoldsproducedfromthediamondcompositeabrasive,suggestingthatthelargersurfacefeaturesinuencedtheadhesionofredbloodcells.Fromthevideosusedtoquantifytheresults,itwasapparentthattheadheredredbloodcellsandmanyoftheplateletsaggregatedinlargegroupsonroughersurfaces.Oftencellsknockedawaybyotherowingcellsemptiedaspaceonlytouncoveranothercellunderneath.Duetothisobservationandthatmanyofthecellsbecomeadheredaroundorwithinlarge 61

PAGE 62

Figure5-12. Comparisonofadheredcellstodiamondabrasivemolds 62

PAGE 63

Figure5-13. Comparisonofadheredcellstocompositemagneticabrasivemolds 63

PAGE 64

Figure5-14. Comparisonofdiamondabrasivesizeonsurfacenish 64

PAGE 65

Figure5-15. Comparisonofcompositemagneticabrasivesizeonsurfacenish 65

PAGE 66

surfacefeatures,thepresenceoftheredbloodcellsmaybereducingthenumberofobservableadheredplatelets,bothskewingthemicroscopefocusandcoveringcellsattachedbeneaththem.Inaddition,duetothelargeeldofviewproducedbythe33objective,thelargesurfacefeaturesoftheunnishedandWAmagneticabrasivemoldsmayalsobemaskingadheredcells.Thiscouldbexedbyincreasingthemagnicationandviewingasmallerarea,allowingforamoredetailedviewandeasieridenticationofcells.Aggregationstilloccuredonsmoothersurfaces,butinsmallergroupsandlessfrequently.Thistrendvariedamongareasforeachsurfaceviewedbutwasconsistentoverall.TimelapseimagesofLeaets3and5areshowninFigure 5-16 .ImagesfromLeaet5,showninFigure 5-16 (i),showalargenumberofadheredredbloodcellsthattendtoaggregateinthetopleft-handcorneraroundalargefeatureonthesurface.Comparatively,Leaet3,hasalargeroverallpeak-to-valleyroughness0.79comparedto0.6mRzbutasimilaraverageroughness(0.06mRa)andnoadheredredbloodcells.Whenroughnessprolesofthetwomoldsarecompared,inFigures 5-14 (ii)and 5-15 (i)forMolds3and5,respectively,itisapparentthatatsmallerscalesMold5hasaroughersurfacecomparedtoMold3withlargerpeak-to-peakdistancesanddeepervalleys.SinceMold3wasnishedwithasmallerabrasive,butforthesameamountoftimeasMold5,thepeak-to-valleyisgreatlyinuencedbythelargeremnantsurfacefeaturesthatwerepresentontheunnishedmoldandnotremovedbythesmallabrasive.Theseremnantfeaturesarenotnecessarilyencounteredbythecells,whicharesmallandseemtorespondtothetextureproducedbythenishingabrasive.Additionally,itwasobservedthatthesmoothersurfaces,withloweraverageroughnessvalues,weremorelikelytolosecellsfromthesurfacetolocalshearstressesinducedbylocalizedow.Currently,duetothesmallheightoftheowchamberopening,thereisnotmuchavailablespacetoaccomodatevariationinleaet 66

PAGE 67

Figure5-16. Timelapseimagesofadheredbloodcells thickness.Itwasobservedthatthickerleaetsorleaetswithhigherthicknessvariationhadvaryingowratesandpathsaccordingtothethicknessofagivenarea.Thecorrespondingincreaseinshearstresswouldmakeitlesslikelyforthecellstoadheretoasurface.Itwasobservedthatofthesurfacessubjectedtothisincreasedshear,includingsurfacesfromMold1(unnished),Mold4(diamondabrasive),andMolds6and7(diamondmagneticabrasive),thatthesmootherthesurfacethemorelikelycellsweretobepulledoff,reversingadhesion.Thiscouldbeduetocellsbeinglodgedinthelargesurfacefeaturesoftheroughermoldswhichwouldprotectthemfromthehighershearstresses,orbecausethesurfacetopographyinducedirreversibleadhesion,discussedinSection 2.2.2 .Toreducethevariationinlocalizedowrateandshearstress,thenextowchamberwillfeatureadeeperopeningtomitigatetheinuenceofleaetthicknessvariation. 67

PAGE 68

CHAPTER6CONCLUSION 6.1ConcludingStatementsThisworkfocusedondevelopingandimpartingasurfacetextureintosiliconeheartvalveleaetsthatreducesbloodcelladhesion.Theresultsaresummarizedbelow, 1. Anishingprocesswasdevelopedandamachinewasbuiltfornishingtheinternalsurfaceofamoldusedinauniqueheartvalvemanufacturingprocess. 2. Varioussurfacetexturesatdifferentsurfaceroughnessvalueswerecreatedinthemoldsbychangingthethetypeandsizeofabrasiveparticles.Thesurfacetextureswerereplicatedontosiliconeheartvalveleaets.Itwasshownthatshorterpeak-to-peakdistancesproducedsmoothsiliconesurfacesduetotheeffectofthesiliconesurfacetensionduringreplication. 3. Anexperimentwasdesignedfeaturingthedevelopedleaetsandaowchamberallowingbloodcelladhesionontheleaetsurfacetobeobserved.Atestingprotocolwasnalizedandimplemented. 4. Theprocesswasrecorded,andplateletandredbloodcelladhesionwasquantied.Thenumberofcellswasnormalizedtotheobservedarea. 5. Celladhesionwascorrelatedtothesizeofabrasiveparticlesusedduringmoldnishing.Smallerabrasives,whichcreatesmallerpeak-to-peakdistancesinthemoldsandsmoothersiliconesurfaces,reducedbloodcellsadhesion.Moldsurfacesnishedwithloosediamondabrasive,createdshortcuttingmarksandshortmicrofeaturesreducedbloodcelladhesionmorethanmoldsurfacesnishedwithcompositemagneticabrasive. 6. Sincethecellsaresmallandareviewedoverarelativelysmallareacomparedtotheevaluationlengthsusedtomeasuresurfaceroughness,theaveragesurfaceroughnessbetterdeterminescellularadhesionthanpeak-to-valleyroughness,whichmeasuresthehighestpeaktothelowestvalleyovertheentireevaluationlength. 7. Smoothersurfacesfeaturedlesscellaggregationandweremorelikelytolosecellsduringinstancesofhighshearthansurfaceswithlargesurfacefeatures. 6.2FutureWorkThisworkisaninitialstepinthedevelopmentofapolymericheartvalve.Tobetterobservetheeffectofthevalvesurfacetextureonbloodcelladhesion,changestothetestingsetupandexperimentaldesignaresuggested.Toreducetheeffectofthickness 68

PAGE 69

variationontheexperimentalshearstress,anewowchamberwillbedevelopedwithadeeperopeningtoreduceresistancetoow.Thiswillprovidemoreconsistentexperimentalconditionswithinthechamberandreducevariation.Infuturetrials,themagnicationwillbeincreasedtobetterobserveadheredcellsandtheirmorphology.Additionally,itwasobservedthatleaetplacementonthemicroscopeslideaffectedthebehaviorofthebloodwithintheowchamber.Sincethe70mmleaetwasshorterthanthe76mmmicroscopeslide,theplacementcouldbeskewedtoonesideoftheowchamberortheother.Itwasobservedthatbloodwouldpooloverareaswheretheglassslidewasexposed.Thisaffectedthebehaviorofthebloodintwowaysdependingontheposition.First,iftheleaetwaspositionedundertheowchamberinlet,bloodwouldnotpoolinthislocationandthesurfacewouldclearrapidlyduringushing.Thiswasadvantageousbecausewiththesurfaceclearedadheredcellswereeasytoview,identify,andcount.Negatively,iftheleaetwaspositionedundertheinlet,bloodwouldpoolattheexitandclot.Thisblockedowandcausedpressurechangeswithintheowchambermakingthesurfacegoinandoutoffocus.Iftheleaetwaspositionedattheoppositeendoftheslide,theoppositewouldoccur.Thebloodwouldfreelyleavetheowchamberbutthepoolingattheinletwouldprovidetheobservedareawithaconstantstreamofcells,oftenblockingtheviewofthesurfacemakingthecellshardertoidentifyandcount.Withtheleaetcentered,bothclottingattheexitandcellstreamingwouldoccurtovariousdegrees.Toremedythis,itisproposedthatalargerdiametermoldbeusedwhendevelopingleaetsforbloodcelladhesiontesting.Thelargerdiametertubecreatedinthemoldwillbeopenedtomakealongerleaetthatspansthelengthofthemicroscopeslide.Thisshouldeliminatetheeffectsofbloodpooling.Inthisstudy,duetothegeometryoftheowchamber,thenishingdirectionwasorientedinlinewiththebloodow.Infuturetrials,celladhesionwillbestudiedwhenthebloodisowingperpendiculartothenishingdirection.Thiswillbeaccomplishedbynishingalongersectionofthemold,thusdevelopingalongerleaetthatisthen 69

PAGE 70

trimmedtotonthemicroscopeslide.Analternateowchamberwasproposedthatwouldfeaturea2525mmleaet,wherebloodcelladhesioncouldbestudiedineitherdirectionandorientingtheleaetaccordingly.Butduetothetestingsetupandsizeofthemicroscopeobjective,thepocketcreatedontheundersideoftheowchamberwouldbetoosmalltoaccomodateit.Afterrunningmanybloodcelladhesiontrials,aprotocolwasestablishedforhandlingthebloodandperformingtheexperiments.Thebloodispickeduptwohoursbeforetheinitialtestandbroughtbacktothelabonice.Onceinthelab,theinitialsamplemountedintheowchamberisimmediatelyushedwith15mLofsalineattwicetheexperimentalowrate.Thebloodbagisspikedandgentlyrockedback-and-forthfor5mintomixthebloodbefore20-50mLaredrainedfromthebag.Thegentlerockingmotionisperformedtominimizedamagetothebloodcellswhileadequatelymixingthebloodcomponents.Drainingthebloodclearsoutbloodthathascontactedthespikeandpossiblydamaged,afterthebloodisdrained,15mLaredrawnintoasyringe.Theremainingbloodisstoredintherefrigeratorat2-4Cuntilthenexttrialisbeingprepped.Beforebeginningthesecondtrial,thebloodshouldbebroughtbackoutanddrawnintoasyringeapproximately45minbeforethetrialtoallowthetemperatureofthebloodtoapproachroomtemperature(approximately22C).Itwasnoticedthattheexperimentaltemperatureofthebloodaffecteditsbehavior,ifthebloodwastoowarmortoocold,itwouldeasilyclotanddisruptthetrial.Whilethisstudyfocusedonthenishedsiliconesurface,thesurfaceexposedtoairduringcuringwillalsobestudiedandistheoreticallythesmoothestsurfaceachievablebythesilicone;itisformedbythesiliconesurfacetensionandthecentrifugalforceexertedbythemoldrotation.Theair-sidesurfacewillbecomparedwiththesurfaceproducedfromMold4toidentifyifasmootherleaetsurfaceisachievable. 70

PAGE 71

APPENDIXABLOODCELLADHESIONDATA 71

PAGE 72

TableA-1. DatafromLeaet1 Area1Area2Area3Area4Area5Area6Area7QuadrantPlRBCPlRBCPlRBCPlRBCPlRBCPlRBCPlRBC Q11182222231303440390430Q12162240330380441470483Q13441251292422471560580Q14300260230311230520390Q21292410280410400401321Q22402390372391471441421Q23641457363450421600610Q24278370430413344491490Q31393241442480650472451Q32501400301600454610454Q335502810412590502541570Q34391243400470452530420Q41251280410360231590361Q42341440410391410610590Q43322310370440450442580Q44360210281340390490420 Total5782749924554146741167417815875611Bias802802802802802802802 72

PAGE 73

TableA-2. DatafromLeaet2 Area1Area2Area3QuadrantPlRBCPlRBCPlRBC Q11170310320Q12280440390Q13NANA370390Q14NANA240230Q21320440380Q22490320460Q23490351440Q24300320260Q31440380500Q32530330440Q33460380380Q34320290290Q41270270340Q42390361450Q43340440390Q44310300171 Total511055425831Bias800800800 73

PAGE 74

TableA-3. DatafromLeaet3 Area1Area2Area3Area4Area5Area6QuadrantPlRBCPlRBCPlRBCPlRBCPlRBCPlRBC Q11150260160300260170Q12210180250140270320Q13330400310390230370Q14250160150220270310Q21340250210250350350Q2290250290330310340Q23410240480380340530Q24350390270231290480Q31210170250200300320Q32190410410250420350Q33450380280280330370Q34140240220350340310Q41151320370200210300Q42300360410340300400Q43310300360300250260Q44120130360300190210 Total400144404780446146605390Bias160160160160160160 74

PAGE 75

TableA-4. DatafromLeaet4 Area1Area2Area3Area4Area5Area6Area7QuadrantPlRBCPlRBCPlRBCPlRBCPlRBCPlRBCPlRBC Q11120150170160100160190Q1217014127025040230150Q13200240310270300300310Q14210160270170210160270Q2117017023016090260160Q22250200300190100220140Q23210280390170280310490Q2426019030014080200380Q31340220320150190270200Q32250260370170180290130Q33210270390130210100230Q341701402709090310200Q4113017018025050130100Q42160220250130140250230Q43350320180220150320170Q4420011027019014016030 Total3400324144702840235036703380Bias160160160160160160160 75

PAGE 76

TableA-5. DatafromLeaet5 Area1Area2Area3QuadrantPlRBCPlRBCPlRBC Q11280192320Q12430201390Q13350501340Q14431270251Q21320272230Q22390294450Q23460330450Q24380460370Q31350290320Q32490421430Q33570471410Q34400410390Q41320221330Q42420232371Q43550382281Q44310300280 Total6451523175613Bias320321320 76

PAGE 77

TableA-6. DatafromLeaet6 Area1Area2Area3Area4Area5QuadrantPlRBCPlRBCPlRBCPlRBCPlRBC Q11600490480200220Q12730760750150220Q13751601790170250Q14890530640100230Q21570510610170170Q22530540530240250Q231260820740170250Q24950710700210160Q31720650440230280Q32651861740120320Q33980850790280310Q34760770690160180Q41680500380220200Q42820810700200220Q43781790770150190Q44620730540260170 Total12293109221029030303620Bias801801801161161 TableA-7. DatafromLeaet7 Area1Area2Area3Area4Area5QuadrantPlRBCPlRBCPlRBCPlRBCPlRBC Q11270350280250200Q12570370390310330Q13610450410420350Q14490400420231190Q21500380520250180Q22460510370340360Q23540500490410390Q24400460470280250Q31380350340390330Q32590490580370280Q33620660610530470Q34510430590290320Q41380370410380330Q42450460590480420Q43500340770420440Q44350400490410350 Total76206920773057615190Bias320480160800800 77

PAGE 78

APPENDIXBCALCULATIONOFBLOODSHEARSTRESSTocalculatetheowratebasedonaprescribedshearstresstheshearratewasfoundusingthefollowingrelation:=_ (B)whichbecomes_= (B)where_istheshearrate,istheshearstress,andisthedynamicviscosityofblood(0.0038Pas).Velocitycanbecalculatedfromtheshearrateusingtherelation,_=@u @y (B)whichbecomesu=_y+C (B)whereuisthevelocityproleandyisthedistancefromthecenterofthechamber.Atthewall,adistanceofh 2fromthecenterofthechamber,wherehistheheightofthechamber,thevelocityiszero.Applyingthisboundarycondition,EquationBcanbesolvedfortheconstantC.C=)]TJ /F9 11.955 Tf 10.92 0 Td[(_h 2 (B) 78

PAGE 79

Pluggingthisbackin,thevelocityisthenfoundtobe,u=_y)]TJ /F8 11.955 Tf 13.15 8.09 Td[(h 2 (B)Themaximumvelocityisfoundatthecenteroftheprole,whereyisequaltozero.umax=_h 2 (B)Assumingparabolicow,theaveragevelocityisapproximatelyhalfofthemaximumvelocity.Volumetricow,ortheexperimentalowrate,isfoundbymultiplyingtheaveragevelocitybytheowchamberopeningarea._V=umax 2A (B) 79

PAGE 80

REFERENCES [1] J.C.Sun,M.J.Davidson,A.Lamy,andJ.W.Eikelboom,Antithromboticmanagementofpatientswithprostheticheartvalves:currentevidenceandfuturetrends,TheLancet,vol.374,no.9689,pp.565,2009. [2] A.G.Kidane,G.Burriesci,P.Cornejo,A.Dooley,S.Sarkar,P.Bonhoeffer,M.Edirisinghe,andA.M.Seifalian,Currentdevelopmentsandfutureprospectsforheartvalvereplacementtherapy,JournalofBiomedicalMaterialsResearchPartB:AppliedBiomaterials,vol.88B,no.1,pp.290,2009. [3] G.Truskey,F.Yuan,andD.Katz,Transportphenomenainbiologicalsystems,ser.PearsonPrenticeHallBioengineering.PearsonPrenticeHall,2004. [4] W.GermannandC.Staneld,Principlesofhumanphysiology.BenjaminCummings,2002. [5] P.Zilla,J.Brink,P.Human,andD.Bezuidenhout,Prostheticheartvalves:Cateringforthefew,Biomaterials,vol.29,no.4,pp.385,2008. [6] P.Bloomeld,Choiceofheartvalveprosthesis,Heart,vol.87,no.6,pp.583,2002. [7] M.S.Sacks,F.J.Schoen,andJ.E.Mayer,Jr.,BioengineeringChallengesforHeartValveTissueEngineering,AnnualReviewOfBiomedicalEngineering,vol.11,pp.289,2009. [8] Y.BoudjemlineandP.Bonhoeffer,Thepercutaneousimplantableheartvalve,ProgressinPediatricCardiology,vol.14,no.1,pp.89,2001. [9] L.H.Edmunds,S.Mckinlay,J.M.Anderson,T.H.Callahan,J.H.Chesebro,E.A.Geiser,D.M.Makanani,L.V.McIntire,W.Q.Meeker,G.K.Naughton,J.A.Panza,F.J.Schoen,andP.Didisheim,Directionsforimprovementofsubstituteheartvalves:Nationalheart,lung,andbloodinstitute'sworkinggroupreportonheartvalves,JournalofBiomedicalMaterialsResearch,vol.38,no.3,pp.263,1997. [10] H.Ghanbari,H.Viatge,A.G.Kidane,G.Burriesci,M.Tavakoli,andA.M.Seifalian,Polymericheartvalves:newmaterials,emerginghopes,TrendsinBiotechnology,vol.27,no.6,pp.359,2009. [11] A.Yoganathan,Z.He,andS.Jones,Fluidmechanicsofheartvalves,AnnualReviewOfBiomedicalEngineering,vol.6,pp.331,2004. [12] S.Chung,Y.Im,H.Kim,H.Jeong,andD.A.Dornfeld,Evaluationofmicro-replicationtechnologyusingsiliconerubbermoldsanditsapplications,InternationalJournalofMachineToolsandManufacture,vol.43,no.13,pp.1337,2003. 80

PAGE 81

[13] N.Mohandas,R.M.Hochmuth,andE.E.Spaeth,Adhesionofredcellstoforeignsurfacesinthepresenceofow,JournalofBiomedicalMaterialsResearch,vol.8,no.2,pp.119,1974. [14] K.R.Milner,C.A.Siedlecki,andA.J.Snyder,Developmentofnovelsubmicrontexturedpolyether(urethaneurea)fordecreasingplateletadhesion,ASAIOJ,vol.51,no.5,pp.578,2005. [15] E.Martines,K.McGhee,C.Wilkinson,andA.Curtis,Aparallel-plateowchambertostudyinitialcelladhesiononananofeaturedsurface,NanoBioscience,IEEETransactionson,vol.3,no.2,pp.90,june2004. [16] N.J.Hallab,K.J.Bundy,K.O'Connor,R.L.Moses,andJ.J.Jacobs,Evaluationofmetallicandpolymericbiomaterialsurfaceenergyandsurfaceroughnesscharacteristicsfordirectedcelladhesion,TissueEngineering,vol.7,no.1,pp.55,2001. [17] A.Thapa,T.Webster,andK.Haberstroh,Polymerswithnano-dimensionalsurfacefeaturesenhancebladdersmoothmusclecelladhesion,JBiomedMaterRes,vol.67A,no.4,pp.1374,2003. [18] M.Stavridi,M.Katsikogianni,andY.F.Missirlis,Theinuenceofsurfacepatterningand/orsterilizationonthehaemocompatibilityofpolycaprolactones,MaterialsScienceandEngineering:C,vol.23,no.3,pp.359,2003. [19] M.L.Carman,T.G.Estes,A.W.Feinberg,J.F.Schumacher,W.Wilkerson,L.H.Wilson,M.E.Callow,J.A.Callow,andA.B.Brennan,Engineeredantifoulingmicrotopographiescorrelatingwettabilitywithcellattachment,Biofouling,vol.22,no.1-2,pp.11,2006. [20] J.F.Schumacher,M.L.Carman,T.G.Estes,A.W.Feinberg,L.H.Wilson,M.E.Callow,J.A.Callow,J.A.Finlay,andA.B.Brennan,Engineeredantifoulingmicrotopographies-effectoffeaturesize,geometry,androughnessonsettlementofzoosporesofthegreenalgaulva,Biofouling,vol.23,no.1-2,pp.55,2007. [21] A.CurtisandC.Wilkinson,Topographicalcontrolofcells,Biomaterials,vol.18,no.24,pp.1573,1997. [22] K.MilnerandC.Siedlecki,Fibroblastresponseisenhancedbypoly(l-lacticacid)nanotopographyedgedensityandproximity,IntJNanomedicine,vol.2,no.2,pp.201,2007. [23] M.Kuwahara,M.Sugimoto,S.Tsuji,H.Matsui,T.Mizuno,S.Miyata,andA.Yoshioka,Plateletshapechangesandadhesionunderhighshearow,Ar-teriosclerThrombVascBiol,vol.22,no.2,pp.329,2002. 81

PAGE 82

[24] J.Y.Park,C.H.Gemmell,andJ.E.Davies,Plateletinteractionswithtitanium:modulationofplateletactivitybysurfacetopography,Biomaterials,vol.22,no.19,pp.2671,2001. [25] M.W.Weston,D.V.LaBorde,andA.P.Yoganathan,Estimationoftheshearstressonthesurfaceofanaorticvalveleaet,AnnalsofBiomedicalEngineering,vol.27,pp.572,1999,10.1114/1.199. [26] S.Einav,D.Stolero,J.Avidor,D.Elad,andL.Talbot,Wallshearstressdistributionalongthecuspofatri-leaetprostheticvalve,JournalofBiomedicalEngineering,vol.12,no.1,pp.13,1990. [27] F.J.WalburnandP.D.Stein,Wallshearstressduringpulsatileowdistaltoanormalporcineaorticvalve,JournalofBiomechanics,vol.17,no.2,pp.97,1984. [28] D.M.Stevenson,A.P.Yoganathan,andF.P.Williams,Numericalsimulationofsteadyturbulentowthroughtrileaetaorticheartvalvesii.resultsonvemodels,JournalofBiomechanics,vol.18,no.12,pp.909,1985. [29] H.YamaguchiandT.Shinmura,Studyofaninternalmagneticabrasivenishingusingapolerotationsystem:Discussionofthecharacteristicabrasivebehavior,PrecisionEngineering,vol.24,no.3,pp.237,2000. [30] H.Yamaguchi,T.Shinmura,andA.Kobayashi,Developmentofaninternalmagneticabrasivenishingprocessfornonferromagneticcomplexshapedtubes,JSMEInternationalJournalSeriesCMechanicalSystems,MachineElementsandManufacturing,vol.44,no.1,pp.275,2001. [31] H.Yamaguchi,T.Shinmura,andR.Ikeda,Studyofinternalnishingofausteniticstainlesssteelcapillarytubesbymagneticabrasivenishing,JournalofManufac-turingScienceandEngineering,vol.129,pp.885,2007. [32] R.E.Riveros,H.Yamaguchi,T.Boggs,I.Mitsuishi,K.Mitsuda,U.Takagi,Y.Ezoe,K.Ishizu,andT.Moriyama,Magneticeldassistednishingofsiliconmemsmicro-porex-rayoptics,ASMEConferenceProceedings,vol.2010,no.49460,pp.87,2010. [33] H.YamaguchiandT.Shinmura,Studyofthesurfacemodicationresultingfromaninternalmagneticabrasivenishingprocess,Wear,vol.225-229,no.Part1,pp.246,1999. [34] N.M.K.Lamba,J.M.Courtney,J.D.S.Gaylor,andG.D.O.Lowe,Invitroinvestigationofthebloodresponsetomedicalgradepvcandtheeffectofheparinonthebloodresponse,Biomaterials,vol.21,no.1,pp.89,2000. 82

PAGE 83

BIOGRAPHICALSKETCH TaylorBoggswasbornin1986toWilliamandPattiBoggsinOrlando,Florida.SheenrolledattheUniversityofFloridain2005andgraduatedwithaBachelorofScienceinmechanicalengineeringin2009.Duringhernalyearasanundergraduate,shejoinedtheMachineToolResearchCenter(MTRC)andbeganconductingresearchwithDr.HitomiGreensletwhocontinuedtoguidehergraduateworkinJanuary2010.ShegraduatedwithaMasterofScienceinmechanicalengineeringinAugust2011. 83