1、sed use of composites in primary structures,accident investigators will likely encounter failed composite structures with increasing frequency in the coming decades Why would these composite structures fail?First,we are building composite structures on a scale never before achieved.ite structures th
2、rough relatively new,automated techniques rather than relying on traditional methods of constructing composites by handAnd third,our inspection and maintenance requirements will no longer be driven by fatigue and corrosion performance,as they are for metallic structures,because composites are not as
3、 susceptible to these failure mechanisms.Instead,accidental subsurface damage and subsequent failure progression will be more importantThe ductility of metal structures provides macrostructurallregarding an accident(Wanttaja,1994)aft composites are not ductile;they are brittle,which means they under
4、go relatively minor permanent deformation prior to final failure What evidence would be produced by a failed composite structure?microstructuralevidence becomes paramount Microstructuralevidence refers to relatively local deformation and changes in the structure,such as fracture surfaces,that typica
5、lly require close visual or microscopic analysis.Indications of fatigue cracking in the lower right wing spar cap of the Chalks Ocean Airways Grumman Mallard G73 that crashed during takeoff December 19,2005.This is an example of using accumulated knowledge and experience with metallic structures to
6、identify possible factors in an accident(NTSB,2005).Tension failure in composites.Macroscopically,even simple tension canproduce fractures with a wide variety of features.Microscopic analysis isparamount(Ginty and Chamis,1987).拉伸拉伸压缩压缩弯曲弯曲冲击分层冲击分层FatigueA300-600 composite vertical stabilizer that fa
7、iled during American Airlines flight 587 in November 2001(NTSB,2004).er of the Airbus A300-600 is one of the largest composite principal structural elements in commercial aviation（before A 380）The vertical stabilizer of the A300-600 is attached to the fuselage by three pairs of composite lugs-forwar
8、d,middle,and aft Analysis of flight recorder data by the NTSB indicates that the aircraft was subjected to a violently changing oscillatory sideslip motion,causing loads in excess of the ultimate design loads of the stabilizerthe right rear lug of the stabilizer suffered a tensile overload failure t
9、hat caused the progressive failure of the remainder of the attachment points.tensile failures in composites generally produce rough fracture surfaces.The rough appearance of this fracture surface helped the NTSB determine that the lug failed under tensile loads.Similar rough fracture surfaces were f
10、ound on the other two lugs on the right side of the stabilizer.As a result,the NTSB concluded that the lugs on the right side of the stabilizer failed due to overstress under tensile loading.Along with several other fractures,the fractures of the right aft lug were rough,consistent with tensile load
11、ing(NTSB,2002).alysis by the NTSB,after the lugs on the right side failed,the damaged stabilizer deflected from right to left,loading the lugs on the left side of the stabilizer in bending.In bending,tension developed on the inboard side of the lugs and compression developed on the outboard side of
12、the lugs.The NTSB identified evidence consistent with tension failure on the inboard side and compression failure on the outboard side of the lugs on the left side of the stabilizer.Fractures in multiple locations exhibited chop marks(marked with a“c”)on theends of fractured fibers,consistent with c
13、ompressive loading and buckling offibers(NTSB,2002).Interlaminar fractures in multiple locations exhibited hackles,consistent withfailure in shear(NTSB,2002).on the left aft lugcted to compressive loads,they can buckle and the fracture surface on the end of a failed fiber may indicate chop marks.The
14、 left aft,left center,and left forward lugs of the failed stabilizer each contained fractured fibers with chop marksAlso found on the left aft lug were hackles associated with shear failure in the matrix-rich region between pliesHackles were found on the left forward lug as well.It must be noted tha
15、t the NTSB did not find any indication of fatigue damage in the vertical stabilizerproach,however,without understanding the changes of internal structure of the material over time and awareness of the consequences of those mechanisms upon structural integrity would prove disastrous.PROGRESSIVE FAILU
16、RE ATESuctures is a progressive series of events.Failure often starts as a tiny crack between the fibers and matrix These cracks decrease the stiffness of the matrix causing the fibers or surrounding plies to carry a higher stress than they normally would.Capturing stress redistribution is the key t
17、o realistic simulation of composite structures failurePROGRESSIVE FAILURE ATESmonly underutilized in industry because of the difficulty in achieving a solution the analyst can have confidence in.This difficulty stems from convergence problems in general purpose finite element codes and inaccurate me
18、thods for predicting multiple failure modes.In turn,this had led industry to use very conservative first-ply-failure solutions,negating many of the optimization advantages that can be gained from composite materials.PROGRESSIVE FAILURE ATESdesire to answer such questions as:Where does damage occur a
19、nd how does it affect ultimate integrity of the structure?How and when does damage initiate?How much tolerance exists between initial,localized failure through ultimate failure?How do different load conditions affect the structural response?How do environmental factors affect the structural response
20、?How can the design be optimized to improve performance and efficiency?Inaccurate Analytomogeneous,constituent nature of composite materials,they fail in manners quite different than linear elastic,homogeneous materials(metals)No Convergent Solution The Elusive Final Answer Without accurate knowledg
21、e of composite behavior，Uncertainty Leads to Excessive ConservatismThe keys to this solution are the combined applications of Multicontinuand the Intelligent Discrete Softening(ISD)Methodrs have recognized the need for multiscalestress or strain information in order accurately capture the failure re
22、sponse of the constituents in a composite However,a fundamental challenge in doing so is to efficiently cross multiple scales to capture micro-structural information where failure initiatesThe keys to this solution are the combined applications of Multicontinuand the Intelligent Discrete Softening(I
23、SD)Methodssful failure theories developed for composite laminates recognize that different failure criteria apply for the fiber and matrix within a composite material.Hashinproposed failure criteria for both the fiber and matrix materials within a composite based on the composite stress fields.In th
24、e spirit of Hashin,the multicontinuumtheory employs independent failure modes of the fiber and matrix constituents.However,rather than utilize composite stresses to predict constituent failure,MCT utilizes constituent stresses to predict constituent failure of the fiber and matrix.Continuum level co
25、nstituent information may be generated in a numerically efficient manner utilizing a multiscaledecomposition originally developed by HillProgressive Failure Analyructureposite structure is loaded primarily in bending.Failure of the structure is shown to be a series of events beginning with matrix cr
26、acking through fiber failure and ultimate failure.Failure of the structure initiates early on in the loading at 260%Flight Limit Load(FLL),but does not collapse until 850%FLL!Progressive Failure AnalyructureProgressive Failure AnalyructureProgressive Failure AnalyructureWiseTek概概以碳以碳/碳、碳碳、碳/酚醛等为代表的酚
27、醛等为代表的多向编织防热复合材料是洲际战略导弹弹多向编织防热复合材料是洲际战略导弹弹头、固体火箭发动机喷管等关键热结构部头、固体火箭发动机喷管等关键热结构部件无可替代的防热件无可替代的防热/结构材料，要承受超高结构材料，要承受超高温、高压和多相粒子流的高速冲刷等极为温、高压和多相粒子流的高速冲刷等极为恶劣的服役环境，对其使用规范、可靠性恶劣的服役环境，对其使用规范、可靠性能否做出准确估算将取决于对该过程力学能否做出准确估算将取决于对该过程力学现象的理解深度现象的理解深度碳碳/碳碳3D4D5D碳碳/碳碳碳碳/碳碳碳碳/碳碳单单、微微结结构构复复杂杂的的单单元元素素多多相相体体材材料料体体系系，在
28、在微微观观、细细观观和和宏宏观观尺尺度度上上具具有有各各自自明明显显的的材材料料结结构构特特征征，在在不不同同尺尺度度上上控控制制材材料料性性能能和和破破坏坏的的因因素素也也不不同同“现现在在人人们们已已经经认认识识到到，对对于于多多晶晶体体而而言言，至至少少存存在在着着宏宏观观、细细观观和和微微观观三三个个主主要要层层次次。它它们们之之间间并并不不存存在在着着从从微微观观可可以以推推导导出出宏宏观观性性质质的的关关系系，或或者者说说，宏宏观观并并不不是是微微观观的的简简单单演演绎绎”。（郑郑哲哲敏敏，周周垣垣，张张涵涵信信，黄黄克克智智和和白白以以龙龙等等五五位位院院士士）碳碳/碳碳Macr
29、o meshMeso meshMicro meshMacroMicroMesoM3methodGsxyW1,GTraction多尺度、多场耦合本本构构本本构构本本构构),.,()()(621 kjifSgCfSgCCkijkkijkij 0001)(,)(,)(kkkfiffiffSgiiiiiiiiiiiiiiiiiiiiiiiibaababbaba/)()(/)/(421Z 向：向：a=2.75 X 10-3 cm3/Kg，b=8.3 X 10-5 cm3/KgXY向：向：a=3.42 X 10-5 cm3/Kg，b=7.63 X 10-5 cm3/Kg本本构构4545简单拉伸试验，材料主
30、方向应力简单拉伸试验，材料主方向应力比比1 1：1 1，不存在剪应力，不存在剪应力力力学学性性能能径径就就是是通通过过材材料料的的性性能能试试验验，真真实实、可可靠靠的的材材料料性性能能数数据据来来源源于于科科学学的的测测试试评评价价方方法法 长长期期以以来来多多向向编编织织碳碳/碳碳复复合合材材料料性性能能测测试试方方法法不不能能科科学学反反映映材材料料应应有有的的本本征征特特性性，例例如如材材料料非非均均质质性性、各各向向异异性性，纤纤维维连连续续性性、完完整整性性以以及及增增强强纤纤维维的的空空间间取取向向和和分分布布等等特特点点不不能能在在材材料料试试样样取取样样和和测测试试中中给给予
31、予完完全全的的反反映映，缺缺乏乏组组分分材材料料的的原原位位测测试试技技术术，复复合合后后的的纤纤维维强强度度保保持持率率以以及及基基体体相相结结构构和和性性质质等等重重要要参参数数难难以以掌掌握握力力学学性性能能力力学学性性能能破坏压缩应力与破坏压缩应力与试件长厚比的关系试件长厚比的关系当当L/hL/h由由7 7变化到变化到5050时，压时，压缩破坏强度可从缩破坏强度可从1500MPa1500MPa下下降到降到125MPa125MPa，相差，相差1212倍，倍，此外压缩破坏的形式还有此外压缩破坏的形式还有脱层和脱层屈曲，这时候脱层和脱层屈曲，这时候的破坏应力的破坏应力XcXc也是很难测也是很
32、难测准的，测试条件稍有变化，准的，测试条件稍有变化，测得的测得的XcXc可能产生几十至可能产生几十至几倍的变化几倍的变化力力学学性性能能力力学学性性能能力力学学性性能能力力学学性性能能力力学学性性能能力力学学性性能能力力学学性性能能测试界面强度测试界面强度的纤维束顶出的纤维束顶出试验装置试验装置力力学学性性能能力力学学性性能能论论分分析析，设设计计所所关关注注的的复复杂杂应应力力状状态态（包包括括拉拉拉拉、拉拉压压、压压压压、拉拉剪剪等等）实实验验状状态态，确确定定载载荷荷比比例例 解解决决第第二二向向联联合合加加载载、试试件件标标距距区区域域的的应应力力状状态态是是否否满满足足设设计计需需要
33、要，载载荷荷比比例例是是否否同同步步可可控控等等问问题题 获获得得复复杂杂应应力力状状态态下下材材料料破破坏坏的的载载荷荷数数据据，采采用用二二次次函函数数分分区区拟拟合合实实验验数数据据，建建立立材材料料强强度度准准则则，并并给给出出不不同同强强度度的的包包络络线线力力学学性性能能形形，厚厚度度为为3 3-5 5m mm m，四四个个夹夹紧紧端端，贴贴有有垫垫片片，成成燕燕尾尾状状，使使便便于于施施加加拉拉力力。对对于于受受压压试试件件，采采用用矩矩形形形形状状试试样样，厚厚度度稍稍厚厚 对对于于拉拉拉拉试试件件，在在十十字字形形的的凹凹角角处处会会发发生生应应力力集集中中，为为了了测测得得
34、材材料料的的真真实实强强度度，必必须须控控制制失失效效的的位位置置，使使之之发发生生在在试试件件中中央央的的标标距距区区内内。为为此此要要把把中中心心区区铣铣薄薄 为为了了尽尽量量不不损损伤伤纤纤维维，试试件件制制备备的的关关键键是是要要使使试试件件表表面面恰恰巧巧在在三三向向碳碳碳碳的的纤纤维维束束平平面面内内，要要尽尽量量降降低低一一切切应应力力集集中中力力学学性性能能拉拉、拉压试件压压试件力力学学性性能能力力学学性性能能拉剪双轴载荷试验及试件拉剪双轴载荷试验及试件力力学学性性能能本本构构关关系系非非常常重重要要 许许多多商商业业材材料料性性能能试试验验机机已已经经能能够够解解决决1 16
35、 60 00 0C C以以下下的的力力学学性性能能、热热物物理理性性能能的的测测试试 超超高高温温（2 20 00 00 0）材材料料性性能能测测试试技技术术需需要要研研究究专专用用设设备备 辐辐射射、通通电电、感感应应加加热热方方式式 尺尺寸寸、加加载载方方式式受受到到限限 高高温温应应变变测测量量依依然然是是大大问问题题 美美、乌乌、法法、日日、中中HITCCM力力学学性性能能500-15001500-2800HITCCM力力学学性性能能力力学学性性能能强强度度强强度度平平面面或或空空间间中中，强强度度准准则则可可分分别别用用一一条条曲曲线线或或曲曲面面来来表表示示，这这就就是是强强度度曲
36、曲线线或或曲曲面面，它它们们应应该该是是连连续续的的而而且且是是封封闭闭的的，不不但但函函数数本本身身连连写写而而且且其其切切线线或或切切面面也也应应该该是是连连续续变变化化的的，但但有有时时候候强强度度准准则则给给出出的的曲曲线线或或曲曲面面，只只能能保保证证函函数数本本身身的的连连续续，其其切切线线和和切切面面会会出出现现突突变变 从从破破坏坏机机理理来来分分析析，应应力力场场的的微微小小变变化化只只可可能能引引起起强强度度曲曲线线或或曲曲面面的的微微小小变变化化，切切线线或或切切面面方方向向也也不不应应该该突突然然大大幅幅度度变变化化，不不过过在在不不了了解解机机理理的的情情况况下下，不
37、不能能过过分分强强调调这这一一点点，与与试试验验数数据据的的附附和和才才是是更更重重要要的的，因因为为唯唯象象论论的的强强度度准准则则，或或多多或或少少存存在在着着误误差差强强度度果果所所获获得得的的现现象象出出发发而而确确定定的的，缺缺乏乏严严密密的的理理论论背背景景，它它只只能能预预测测材材料料破破坏坏时时的的强强度度值值，具具有有较较大大的的实实用用价价值值，但但不不能能确确切切知知道道破破坏坏形形式式和和破破坏坏机机理理，不不能能推推广广 唯唯象象论论的的强强度度准准则则与与破破坏坏机机理理之之间间没没有有直直接接和和必必然然的的联联系系，材材料料的的破破坏坏有有它它的的机机理理和和规
38、规律律性性的的，因因此此反反映映材材料料能能力力的的强强度度准准则则和和破破坏坏机机理理之之间间总总应应该该有有些些联联系系，探探讨讨破破坏坏机机理理，将将有有助助于于强强度度理理论论的的研研究究强强度度ABCDABCD差别不大差别不大EFGHEFGH差别很大差别很大延性材料延性材料脆性材料脆性材料强强度度强强度度强强度度强强度度强强度度强强度度能能量量有有关关的的准准则则，考考虑虑了了应应力力的的联联合合作作用用和和相相互互作作用用对对破破坏坏所所起起的的影影响响，比比一一阶阶强强度度理理论论更更合合理理和和精精确确 最最大大剪剪切切应应变变能能准准则则适适用用于于延延性性材材料料，其其精精
39、度度高高于于最最大大剪剪应应力力准准则则，属属于于二二阶阶强强度度理理论论 最最大大应应变变能能准准则则近近似似地地适适用用于于脆脆性性材材料料，其其精精度度大大多多数数情情况况下下高高于于最最大大主主应应力力准准则则和和最最大大主主应应变变准准则则，也也属属于于二二阶阶强强度度理理论论 蔡蔡-吴吴二二阶阶强强度度准准则则 改改进进的的蔡蔡-吴吴二二阶阶强强度度准准则则 简简化化的的蔡蔡-吴吴二二阶阶强强度度准准则则强强度度强强度度1222222321222222 yzyzxzxzxyxyxyzxyxzyxSSSKKKZYX 强强度度强强度度强强度度破破坏坏机机理理破破坏坏机机理理，其其发发展
40、展和和国国际际上上出出现现许许多多低低应应力力下下的的脆脆断断事事故故密密切切相相关关 美美国国北北极极星星导导弹弹发发动动机机壳壳体体在在试试射射中中爆爆炸炸，所所采采用用的的D D6 6A AC C高高强强钢钢，屈屈服服强强度度为为1 16 60 0k kg gf f/m mm m2 2，而而实实际际使使用用才才只只有有7 70 0k kg gf f/m mm m2 2，引引起起了了材材料料工工作作者者和和科科学学家家的的高高度度重重视视 断断裂裂力力学学是是固固体体力力学学的的重重要要分分支支，专专门门研研究究带带有有裂裂纹纹材材料料或或构构件件的的强强度度以以及及裂裂纹纹扩扩展展规规律
41、律，处处理理材材料料抗抗断断裂裂问问题题，可可以以说说它它是是连连续续介介质质力力学学应应用用材材料料或或构构件件破破坏坏的的一一门门应应用用科科学学，已已成成为为当当今今设设计计师师在在进进行行材材料料安安全全设设计计时时不不可可缺缺少少的的重重要要数数据据脆脆性性FW应应力力？应应力力 21021/tma2102/tma（for long crack）衡量裂纹尖端的衡量裂纹尖端的应力增长幅度应力增长幅度应力集中因子应力集中因子21002/tmtak应应力力应应力力0 0s s)提提出出所所有有的的材材料料中中都都存存在在很很小小的的缺缺陷陷，非非常常小小的的微微观观缺缺陷陷会会降降低低脆脆
42、性性材材料料的的强强度度，这这是是为为什什么么脆脆性性弹弹性性固固体体很很难难达达到到理理论论结结合合强强度度(-E E/1 10 0)，而而延延性性材材料料对对应应力力集集中中源源不不太太敏敏感感，是是由由于于当当最最大大应应力力超超过过屈屈服服强强度度时时，它它们们在在裂裂纹纹尖尖端端发发生生屈屈服服 G Gr ri if ff fi it th h提提出出所所有有的的脆脆性性材材料料包包含含一一定定数数量量的的不不同同尺尺寸寸、形形状状和和取取向向的的微微小小裂裂纹纹和和缺缺陷陷，当当在在这这些些缺缺陷陷的的尖尖端端的的应应力力超超过过理理论论结结合合强强度度时时，破破坏坏就就会会发发生
43、生 玻玻璃璃晶晶须须因因为为缺缺陷陷少少而而接接近近其其理理论论强强度度脆脆性性材材料料的的G Gr r脆脆性性材材料料的的G Gr r破破坏坏机机理理aEGcf 破破坏坏机机理理 aKfIC断断裂裂aKICC 断断裂裂拉拉伸伸试试件件尺尺破破坏坏机机理理破破坏坏机机理理破破坏坏机机理理/C C材材料料2 2-2 2平平面面和和2 2-3 3平平面面内内的的宏宏观观试试件件的的简简单单拉拉伸伸和和偏偏轴轴拉拉伸伸试试验验，偏偏轴轴拉拉伸伸试试验验分分两两类类 0 0 1 15 5（7 75 5 9 90 0）1 15 5 7 75 5 裂裂纹纹扩扩展展规规律律 0 0 5 5以以内内，裂裂纹纹
44、自自相相似似扩扩展展 5 5 1 15 5以以内内，裂裂纹纹先先自自相相似似扩扩展展，最最后后沿沿纤纤维维束束（或或垂垂直直纤纤维维束束）剪剪断断 1 15 5 7 75 5以以内内，试试件件在在裂裂纹纹尖尖端端处处剪剪断断破破坏坏机机理理扩扩展展过过程程，所所以以一一个个试试件件的的强强度度决决定定于于裂裂纹纹尖尖端端的的那那个个纤纤维维束束的的强强度度 0 0 1 15 5以以内内，由由该该纤纤维维束束的的拉拉伸伸强强度度决决定定 1 15 5 7 75 5以以内内，由由该该纤纤维维束束的的剪剪切切强强度度决决定定 0 0 1 15 5内内，纤纤维维束束同同时时受受拉拉应应力力和和不不可可
45、忽忽视视的的剪剪应应力力，其其强强度度随随剪剪应应力力变变化化 1 15 5 7 75 5内内，纤纤维维束束受受正正应应力力不不足足以以明明显显改改变变其其剪剪切切强强度度，所所以以完完全全由由纯纯剪剪切切强强度度决决定定是是否否剪剪断断破破坏坏机机理理破破坏坏机机理理机机制制的的作作用用，提提高高了了在在孔孔洞洞和和裂裂纹纹存存在在情情况况下下材材料料断断裂裂阻阻力力和和强强度度保保持持能能力力，根根据据碳碳/碳碳材材料料的的组组分分和和内内部部结结构构，碳碳/碳碳材材料料表表现现出出三三种种损损伤伤机机制制 C Cl la as ss s I I为为破破坏坏伴伴随随着着内内部部纤纤维维断断
46、裂裂和和明明显显的的基基体体裂裂纹纹 C Cl la as ss s I II I的的复复合合行行为为是是应应力力在在多多重重基基体体裂裂纹纹周周围围再再分分布布，但但增增强强相相保保持持完完好好 C Cl la as ss s I II II I为为出出现现多多重重无无序序的的剪剪切切基基体体裂裂纹纹，应应力力在在一一个个剪剪切切损损伤伤带带内内部部再再分分布布微微结结构构表表征征微微结结构构表表征征微微结结构构表表征征微微结结构构表表征征制样方法带制样方法带来了表面的来了表面的伪真实性伪真实性微微结结构构表表征征微微结结构构表表征征微微结结构构表表征征SEMCT微微结结构构表表征征获得孔隙
47、率及统计获得孔隙率及统计分布分布C/C 3D C/C 3D 可视化可视化模型与虚拟切模型与虚拟切片片C/C/SiCSiC CT CT和和组分含量统计组分含量统计微微结结构构表表征征 基体内 的空洞 界面 脱层 黑框为 参考体 2)/ln(exp21)(2 RRRRN 基体内孔洞的概率分布曲线及分布函数 微微结结构构表表征征微微结结构构表表征征微微结结构构表表征征 有限元法理论曲线 等效夹杂法理论曲线 弹性模量 EL(拉伸)弹性模量 EL(压缩)弹性模量 ET(压缩)编织角（）杨氏弹性模量 E(GPa)ELET微微结结构构表表征征Mode LZTLMode T<Mode Z&ZLMode T
48、ZDamage modeM aximumstress-to-strengthratioAnisotropic damage model for fiber bundleIsotropicdamage modelfor matrix100010001000000001000010000100000000100010000Damage tensorZTLDDD00000022sZLZLcZtZZForFF2sTZTZFZTLZTLZTL微微结结构构表表征征Strain()Stress(MPa)TensilCompressivStrain()Stress(MPa)=28.48=21.98=38.04
49、=48.84微微结结构构表表征征 Strain()Strain(）Number of Integral Points of Damage Stress(MPa)(a)(b)Sum of Damage Pattern TZ Damage Damage of Pattern Tor ZPattern L Damage Matrix Damage DCBA微微结结构构表表征征S tra in ()Stress(MPa)E x p e rim e n ta l T h e o re tic a l=2 8.4 8 =2 1.9 8 =3 8.0 4 =4 8.8 4 拉伸应力应变曲线理论拉伸应力应变曲
50、线理论预报与实验结果对比预报与实验结果对比 Strain()Stress(MPa)TensileCompressiveTheoreticalExperimental简单拉伸及压缩理论预报与实验简单拉伸及压缩理论预报与实验结果对比（编织角结果对比（编织角=48.84=48.84）微微结结构构表表征征微微结结构构表表征征C/CC/C复合材料复合材料制备工艺制备工艺内部（纤维、基体内部（纤维、基体和界面）存在大量和界面）存在大量的孔洞和裂纹的孔洞和裂纹组分（基体、界面、组分（基体、界面、纤维）离散性纤维）离散性孔洞呈非均匀、孔洞呈非均匀、随机分布随机分布C/CC/C复合材料复合材料有效性能预报有效性
