ABAQUS中Truss单元预应力的两种施加方法

ABAQUS中预应力Truss单元的两种实现方法例题：100m长钢缆水平放置从x=0到x=100。两端固定。无初始拉力，计算下垂量。截面；A=0.01539m2，Density:r=7800kg/m3,g=9.8m/s2,E=2.1e+11N/m2Analyticalsolutionofmaximumdisplacement(u2)atx=50m:U2_max=-((3*r*g*L^4)/(64*E))^(1/3)=-1.194944005m方法一.沿trusselement加沿长度方向初始拉应力(seejob-1.inp)此文件中使用了initialcondition,type=stress方法加初始拉应力。因工程上无此初应力,更好的方法是使用降温法。算完后再升温。用降温法。算完后再升温。NOTE:降温法施加预应力（激活钢绞线）。温度=-力/（膨胀系数*弹模*钢绞线面积）1、第一步，在truss单元中施加一个初始应力，让计算处于初始平衡状态；初始应力设置过小，可能不收敛，应多次试算，找到一个合理的应力值。一般情况下，这个初始值对最终值的影响不大，可以忽略。2、第二步，施加truss单元的自重荷载，打开非线性开关（nlgeom=YES）考虑几何非线性问题；3、本例中初始值采用0.1Mpa。自重作用下缆索的拉应力约为80Mpa。最大位移为-1.195m，与理论计算值吻合得很好。*HeadingCableapplinggravityloadwithinitailstressThemaximumAnalyticaldisplacementwithoutinitailstress(atx=50m)U2=-1.194944005meter****Method1.Using*initialcondition,type=stressmethod***Preprint,echo=NO,model=NO,history=NO,contact=NO*Node1,0.,0.101,100.,0.*NGEN,NSET=NALL1,101,1*Element,type=T2D21,1,2*ELGEN,elset=ELALL1,100,1*ELSET,ELSET=EL_OUT1,51,100*SolidSection,elset=ELALL,material=steel0.01539,***Nset,nset=Left1,*Nset,nset=right101,*Nset,nset=mid51,****MATERIALS***Material,name=steel*Density7800.,*Elastic2.1e+11,0.3*initialcondition,type=stress**Note:thesolutionwillnotconvergeastheinitialstress100,000N/m^2ELALL,100000*BoundaryLeft,1,2Right,1,2*Step,name=Step-0,inc=1000Initialstressequilibrium*Static1,1.,1e-05,1.*Output,field,variable=PRESELECT*Output,history,variable=PRESELECT*Nodeprint,nset=mid,freq=1000U,*ELPRINT,ELSET=EL_OUT,freq=1000S*ENDSTEP**----------------------------------------------------------------****STEP:Step-1***Step,name=Step-1,nlgeom=YES,inc=1000Applygravityload*Static0.01,1.,1e-05,1.**Name:GRAVITY-1Type:Gravity*DloadELALL,GRAV,9.8,0.,-1.****OUTPUTREQUESTS***Restart,write,numberinterval=1,timemarks=NO*Output,field,variable=PRESELECT*Output,history,variable=PRESELECT*Nodeprint,nset=mid,freq=1000U,*ELPRINT,ELSET=EL_OUT,freq=1000S*EndStep方法二.使用STABILIZEparameteronthe*STATIC.(seejob-2.inp)“stabilization”在结构上附加artificialviscousdamping（粘滞阻尼），使得计算结果togobeyondtheinstabilitypoint。但计算结果必须验证，并必须保证ALLSD比内能ALLIE小很多。NOTE:1.第一步用*Static,stabilize=2E-10。笫二步不用stabilize(相当于*Static,stabilize=0)。2.使用nlgeom=YESinthesteptoapplythegravityload.3.最终拉应力=8E7N/m2与方法一相等。4.Themaximumdisplacement(atnode51)equalstheanalyticalresult.5.ChecktheALLSDandALLIE.TheALLIEisgreaterthanALLSD.(Seefigure1)6.ThedeformationshapeofthecablecanbeexaminedbyCAE.Itmayneedtosetthedeformationscalefactortoalargenumber(10–1000).7.使用此法必须极端谨慎。稍微不慎，结果会完全不对。Forexample,用*Static,stabilize=2E-4(defaultvalueofthestabilizeparameter)重算此题。其结果如下；.ChecktheALLSDandALLIE.TheALLIEislessthanALLSD.(Seefigure2)Thedeformationshapewithdeformationscalefactor1000isshowninfigure3.Inthefigure,onlythefirstandlastelementsaredeformed.Themaximumdeformationvalueisnotcorrect.使用stabilizeparameter学问很多，一般是越小越好。因为stabilizeparameter是artificialvalue,无法确定理论上的最佳值。我是用试错法。从开始defaultvalue(2.0e-4)往下减(2.E-6,2.0E-8,..)，直到不收敛(2.0E-12).经过验证结果(seethenote4,5,and6)，我决定在计算中使用2.0E-10。*HeadingCableapplygravityloadusingstabilizeparameterThemaximumAnalyticaldisplacementwithoutinitailstress(atx=50m)U2=1.194944005meter****Method2.Using*Static,stabilizemethod***Preprint,echo=NO,model=NO,history=NO,contact=NO*Node1,0.,0.101,100.,0.*NGEN,NSET=NALL1,101,1*Element,type=T2D21,1,2*ELGEN,elset=ELALL1,100,1*ELSET,ELSET=EL_OUT1,51,100*SolidSection,elset=ELALL,material=steel0.01539,***Nset,nset=Left1,*Nset,nset=right101,*Nset,nset=mid51,****MATERIALS***Material,name=Steel*Density7800.,*Elastic2.1e+11,0.3*BoundaryLeft,1,2Right,1,2****STEP:Step-1***Step,name=Step-1,nlgeom=YES,inc=1000Applygravityload**thedefaultvalueofstabilizevalueis2.0E-4**Static,stabilize=2E-4*Static,stabilize=2E-100.01,1.,1e-05,1.**Name:GRAVITY-1Type:Gravity*DloadELALL,GRAV,9.8,0.,-1.****OUTPUTREQUESTS***Restart,write,numberinterval=1,timemarks=NO*Output,field,variable=PRESELECT*Output,history,variable=PRESELECTALLSD,ALLIE*Nodeprint,nset=mid,freq=1000U,*ELPRINT,ELSET=EL_OUT,freq=1000S*EndStep***Step,name=Step-2,nlgeom=YES,inc=1000Recovery*Static0.01,1.,1e-05,1.*EndStep