Influence of Manufacturing Process of Warp-knitted Vascular Prosthesis on the Wall Homogeneity

Journal of Donghua University (Eng. Ed. ) Vol1.20, No.2 (2003) 9Influence of Manufacturing Process of Warp-knitted VascularProsthesis on the Wall HomogeneityWANG Lu(王璐), DING Xin(丁辛)TS|ACollege of Texriles, Donghua Uniersity ,Shanghai, 200051This paper reports the evolution of textile structureprostheses, out of bundreds of retrieval ones, wih longitudinaland mechanical properties of vascular prosthesis inruptures in remeshing line and guide line were observed. Thesethe level of the whole prosthesis and the constituentfailures are probably related t0o such factors, as the improperfilaments with respect to the manufacturing process.design of textile stucture , change of fibrous architecture of theThe tubular wall of the prosthesis is dividedprosthesis during manufacturing， traumatic lesion of thecircumferentially into three zones: basic line ( BL),prosthesis during implantation, and the physico chemicalremeshing line (RL) and guide line ( GL). Someeffects in vivo when the prosthesis was exposed to the chronicheterogeneity has been observed on the tubular wallinflammalory reaction of foreign body and to the systolo-in :erms of stitch structure of the prosthesis anddiastolic arterial stresses'llinear density of the constituent filaments. TheIn order to understand the mechanismn of the deteriorationsbreaking positionthprosthesisund sensidive factors relaing to the rupture, a protocol ofcircumferential tensile localizes preferentially instudying the virgin prosthesis, the same type as the explantedremeshing line that is the weakest zone by warpprostheses with longitudinal rupture, was set up. The tubularknitting with double needle bed. Furthermore, thewall homogeneity of the prosthesis before its implantation wasstatistical differences of the mechanical properties ofevaluated by physical and chemical analyses. In this paper, thethe filanents of zone RL, GL and BL have beenevolution of the mechanical properties of one type of warpconfirmed too. It is predictable that the deteriorationknitted polyester vascular prosthesis in terns of manufacturingof prosthesis, under physiological loads ( periodicalprocesses is approached.pulse blood pressure etc. ), could happen firstly inthe weaker zone in vivo.ExperimentalKeywords: Vascular prosthesis ,stnucture,Circumferential tensile properties, Manufacturing process1 Manufacturing processManufacturing processes of the prosthesis include fiveprinciple stages listed in Table 1. The twbular wall of theIntroductionprosthesis can be divided circumferentially into three zones;basic line (BL), remeshing line (RL) and guide line (GL),Alhough textile vascular prostheses have been widelyas shown in Fig.1. Guide line refers to the zone in which dyedused in vascular surgery, some complications as the type ofblack yamns are kited to avoid twisting during surgicaldilatations, ruptures and false aneurysms were reported in theimplantation. Remeshing line refers to the sides' connection ofliteratures'-” in the past decades. In a joint research under thethe wall, in which the guide bars lap the threads on the frontframe of the European Collaborative Retrieval Program,and back needle bar in sequence during warp kniting process.deteriorations appeared on 20 explanted warp-knitted polyesterThe basic line is the rest zone of the prosthesis.Table 1 Principle manufacturing process ( diameter of virgin prosthesis: 10 mm)ProcessFunctionPurposeRemarksWarp knitingForming tubular shapeReverse locknit (Fig. 2)Washing and dry Removing oil and impurity70°C hot water with detengentsCompactionReducing porosityHigh pesure, high temperature with chemical agentsCrimpingStrengthening longitudinal compliance and transversal nigidityCrimp formation mechanicallyThermosetingStabilizing the crimping shapeThermnosetting with certain pressureFunded by the Shanghai Pust Dxtoral Foundation, Overseas Retumed Scholars' Foundation of Fducation Ministury and the Shanghai Key DisciplineProjectReceired Frb, 14, 2003中国煤化工HCNMHG0Journal of Donghua University (Eng. Ed. ) Vol. 20, No.2 (2003)struclure and mechanical properties in RL and GL with respectto those in zone BL to evaluate the wall homogencity of theReneshingprosthesis.Guide line(GL)，Basic line(BL)line(RL,)1 Stitch structure and linear densityThe denseness of warp-kitted fabrics can been defined byhe stich length, which is related to the disance of wales, theFig.1 Three zones of vascular prosthesisdistance of courses, the length of underlap, and etc.. Fig. 3The prosthesis was knitted with flat yarn of 44 circularshows the variations of stich lengths in the hree zonesfilaments (52 dtex) forming a 1 to l lapping movementfollowing the manufacturing process. After treatment incompaction process ( stage C )，the most significant( abbreviation 1 - 1) and textured yarn of 88 circular filaments(105 dtex) forming a2 to 1 lapping movement ( abbreviationmodification of stitch length can be observed. The stitch1-2) in the zones of BL and RL (indicated by BLI -1,lengths decrease by 22% and 16% respectively for BL 1 - 1RL1-1, BLl -2 and RLI -2 respectively). The black yarnsand BL 1-2.of 44 circular filaments (107 dtex) made a 2 to 1 lapping4.00movement in the GL zone ( abbreviation GLI - 2). The3.75影=厦- +RL1-1 - B-RL1-2detailed architecture of the knited yam is ilustrated in Fig. 2.- GL1-23.50.3.25 .1-2中3.00.言275。一Fig.2 Schemnatic diagram of reverse locknit2.50.intemal view of prosthesis)2 Testing Methods2.25.(1) Stitch structure and linear density of yarns: Stitch2.00lengdh and linear density of yam were measured by a method' 'modified from the French standard NF G 07 - 316. Then theManufacturing processlinear density of yam is divided by the number of constituentFig.3 Stich length variation following thefilaments to obtain the linear density of monofilament. AI1 ofhe yams and monoflaments extracted from the vascularmanufacturing processprosthesis should be conditioned in a standard environment forRefening the remeshing line, the stitch length of RLI - 2at least 24 hours before measurements. The yams andis shorter than that of BL1 - 2 after knitting process indicating amonofilaments were sampled randomly from three vasculartighter loop structure of remeshing line than that of basic line.prostheses labeled in the sarme lot and the testing results wereFollowing the stress relaxation in washing process, the stichthen averaged.length of RL1-2 is still shorter than BL1 -2. However, be(2 ) Circumferential tensile property; The lest wascompaction process appears to give a less strong retraction tocompleted in an MTS dynamometer referming to the IsOthe yarns of RLI - 2 leading to a reverse result of comparison719881 with a uniform extension rate of l00 mn/ min. Theafter stage C. For the stitch length of RLI-I and BLI-1, itdiameter of a semi -cylinder holder on the jam adopters to thedoes not show any significant difference after compactiondiarneter of the testing sarmples in order not to apply extra tensionprocess.on the surople. Testing results from 3 samples were averaged.Regarding the guide line, the stitch length of GL1 - 2(3) Tensile property of monoflament: The test wasafier compacton process is much shorter than that of BLI - 2conducted in a dynamometer ( Lhomargy) with a unifomand RL1 -2. It means that the black yams suffer a strongextension rate of 34 mm/ min. The initial distance of the tworetraction after compaction process.jams was 25 mm. 30 filarnents extracted from three vascularYamn linear density is another important factor afctingprostheses were sampled and the average results calculated.he porosity of the prosthesis wall similar to stitch length. Theporosity predominates the loss of blood in surgery and theResults and Discussiongrowth of tssue on the wall of prosthesis in vivo. Thcormpaction process makes a great contribution to linear densityIn view of the facts of longitudinal ruptures in zones RLof yarns. This phenomenon is more distinct for the yam inand GL of the explanted prostheses after long term implantation中国煤化工.in vivo, this study will focus on the comparison of the stichMHCNMHGJournal of Donghua University (Eng. Ed.) Vol. 20, No.2 (2003) 1(a) Circurmfcrential tensile test130.This circumferential tensile test evaluates the breaking120.strength and breaking deformation according to ISO 719868] .Because of the circumferential beterogeneity of stitch structure110as mentioned above, a reference line is marked in the middle100.position between the two guide lines, indicated in Fig. 5(a).The angle θ is defined, as shown in Fig. 5(b), to indicate theinitial test position to observe the breaking position of the--RL1.1 奋RL1-2samples.The circumferential tensile tests, shown in Table 2,一GL1-2demonstrate that the position of rupture is in correlation withg70the initial test position of the prosthesis. In addition, thebreaking occurs systematically in the zone RL after stage C, D60and E. These results show that rL appears to be an area of50$==￥weakness for this type of prostheses. The loop structureevolution in zone RL differs from that in zone BL duringcircumferential tensile test is also confirmed by the imageManufacturing prucessprocess analysis'?l. However, the evolution of the loopstructure in zone GL does not sbow any significant difference(a) Yarmby a comparison to zone BL.Rcmeshing line RLzGuide line GLRemeshing lime RL2:Reference line2.(a) Reference line+81-1 - BL1.22.0七-RL1-1廿RL1-2Strelching1.8←GL1-2directio, Reference lincTop holder[↓1.HorizontalBoom hlded几StretchingdirctionManufacuring process(b) Monofilament(b) lnitial pusition of sampleFig.4 Linear density of yarn and consitucnt flamnentFig. 5 Initial position of sample during theater manufacturing processcircurmferenial tensile testTo the leve1 of filaments, Fig. 4(b) indicates that theTable 2 Breaking position samples in circumfereniallinear density of the black filament (GL1 -2) is much highertensile testthan the others. In addition, the linear density of the flat (1 -Breaking position1) and lexlured filaments (1 -2) shows a signifcant difrence6/(°)after compaction process ( stage C). This indicates that thereStageA StageB StageC SlageD Stage Eare dfferet responses to the cormpaction treatments between two0BBLRL2RLtypes of filaments. Among all manufacturing processes, it is theRIcompaction process that produces the biggest increase of linear135RL RL,RL: RL,density of both flat and textured filaments.1803LRL1RL:2 Mechanical properties25RL]The mechanical properties are evaluated by a315RL] RL,RL, RLcircumferential teosile test for the whole prosthesis and uniaxialBreaking percentage67100100100tensile test for the constituent filaments extracted from the threein RL/[% )zones of prostheses.中国煤化工TYHCNMHG2Journal of Donghua University (Eng. Ed.) Vol.20, No.2 (2003)(b) Uniaxial censile test5% significant level between the mean values of anyThe tensile property of monoflament is characterized bymechanical properties of the tensile tests, a mark“ #”in Tablethe following parameters: the initial modulus E, the strain,3 is used. In this case. no further comparison of the meanstress and the energy in the yield point (e,. o, and w,) and invalues is made.the rupture point (E,, σ, and W,).(2) LSD test: If the dfference does exist, a pairviseThe differences in the flamenLs from the three zones aremultiple comparison is carried out by“ the least significantevaluated by a statistic analysis' 10’software SPSS 9.0.difference (LSD)”. From the results of the LSD test, theHypotheses are made thilt all of the constituent flamentsdifference between the mean values of the mechanicalinitially come from the sane population. The procedures of theproperties of filaments from two zones is obtained. Once hestatistics are as followsinl.difference between the mean values is significant, at the 5%(1) ANOVA test: If the difference does not exist at thesignificant level, the value is underlined in Table 3.Table 3 Comparison of the mean value of mechanical propertiesWw，StagecomparisoncN. dlex(%)cN. dtex-’10-*cN. dtex-'I .cN. dtex-+ 10-7cN. dex -1(GL- BL)1 -23. 400-0. 401-0.049-0.2694. 5820.25315. 8983 (RL-BL)1-10.4000. 1250.0220.171-0.967-0.076-7. 254(RL- BL)1-20.400 .0. 0990.0540. 107-1.8330. 023.(GL- BL)1-20.2280. 13810. 1530.28620. 364C (RL-BL) 1-10. 0300. 0460.029-0.674-0.022-3.648(RL- BL)1 -2-0. 4400. 019-0.039-2.118-0.174-9.198-1.3800.5210.0050.22210. 1560. 21319.984E (RL-BL) 1-1-0.0100. 0950.0120.0660. 027-0.1160.8770. 2200.1400.0250.1180.8380. 1863. 818Note: (GL- BL)1 -2 indicutes the value of GLI -2 subracts the walue of BLI -2. If the value shown in rable is pusitive, it means the walue ofGLI -2 is greater than that ofBL! -2. (RL- BL) 1-1 and (RL- BL)I -2 have the similar implications.Firstly, the difrence of mechanical properties betweenfilarnents BL. After thecompactionproces, the differenceGL and BL is examined. In the suages B, C and E，thebetween RL and BL is decreased. However the rupture parametersbreaking parameters (E,. o,. W,) of GL flaments are alwaysof flamnents RL are still smaller than that of BL.greater than BL filaments, i. e. corresponding values in Table 3Finally, the diference of mechanical properties betweenare positive. In addition, the compaction treatment makes theRLI - 2 and BLI- 2 is examined. Regarding the rupturedifference of GL and BL greater than that of the thermosetting.parameters, the comnpaction process makes the flaments RLFor the initial modulus E, the mean value of GL is greater thanweaker than that of BL. However, after thermosetting, ththat of BL after stage B. The difference becomes smaller in the, difference between RLI -2 and BLI - 2 becomes smaller. Insuccessive process stages. After thermosetting, the mean valueaddition, it shows that the flat filaments (1 - 1) have differentof GL is stitically smaller than that of BL.responses to manufacturing processes with respect to theConcemning the elastic paraneters (e, o,, andW,).textured flaments (1 -2).those of GL flaments after stage B are smaller ( negative valuesThe results show that there is exactly heterogeneity ofin Table 3). But after compaction and thermosetting processes,mechanical properties among filaments in RL, GL and BLthe mean value of r, and W, of GL filarnents are statisticallyzones. However, the differences are quite complex in relationgreater than that of BL. This result indicates that GL flamentsto the processing stages, initial flament forms ( textured orhave a larger deformation under low stuess loads and thereforeflat) and stitch structure. From the statistical point of view ,showing a weaker elastic property. This implicates that GL zonesome weaknesses have been confirned in zones RL and GL ofcould appear in vivo the earlier dilatation of the prosthesis andthe prosthesis along the manufacturing processes.degradation of the constituent flaments.Secondly, the difference of mechanical properties betweenConclusionsRLI -1 and BLI - 1 is examined. Some statistic dfferences existfor the paruncters of e, and w, after stage B. It can be noticedThe evolution of the textile structure and mechanicalthat the diference of values, ater stage B, are positive for theproperties on the tubular wal of the prosthesis is shownelasic parameters but negative for the rupture ones, This indicatesobviously in relation to the manufacturing processes. And thethat the RL flaments suffer an additional stress with respect to中国煤化工“L contribution to the variationMYHCNMHGJourmal of Donghua Universiy (Eng-. Ed.) Vol.20, No.2(2003) 131979, 189. 741 -745of the parameters. The heterogeneity of stitch stucure in zone[2] BergerJ.H. SuvgeL.R.. Ann Sug 1981. 193. 477-R is observed. The unevenness is one of the principle causesof the preferential rupures of prostheses in rL during the[ 3 ] Guidion R.. Chakit N.. Aneurysmal dererontion of aterialcircumferential tensile test. Furthermore ，the heterogeneity ofsubstitutes. In:" Current Terapy in Vascular Surgery" ，Enstmechanical properties of the constituent filaments in zones RL,.B.. Suanley 1C. (eds.). B.C. Decker Ed.. Toronto,GL and BL has also noticed. It is predictable that thePilndelphia, 1990, 324 -328deterioration of prostheses, under the physiological loads[4] RiepeG.. LoosJ.H., Eur. J. Vasc. Endovasc. Surg，( periodical pulse blood pressure etc. )，could happen1997, 13, 540- 548peferenially in the weaker zone in vivo and sborten the[s] Chakfe N. RiepeG, DievalF, aal, Jounal of Vascuterservice life of vascular prostheses.Surgery, 2001. 33(5). 1015 - 1021[6] Pourdeyhimi B, Wagner D..J. Biomed. Maner. Res., 1986,Acknowledge20, 375 -409[7] Wang L.. Study of ageing of texile vascular prostheses,University of Haute Alsace, France, 2001 (in French)Authors are gruleful to the LPMT and GEPROVAS of[8] ISO7198， Cartiovascular implantsTubular vascularMulhouse University of France for their great help in experinentalprostheses, 1998work and especially to Prof. B Durand, D. D. Mathicu,Dr. F.[ 9 ] Guidoin R, King M, Marceau D, etal, J. Biomed. Mater,Dieval for their belpful suggestions and discussions.Recs. 1987.21,65-87[ 10] in Pibuan, Satistis Methods for Mdicine, Shanghai McdialReferencesUniversity Press, Shanghai, 1992, 50 - 83 (in Chinese)[ 1] NunnD.B.. Freeman M. H. . HudginsP.C. Ann. Sung.,中国煤化工MHCNMHG