ANSI AWS A9.5-2013 Guide for Verification and Validation in Computation Weld Mechanics (1st Edition).pdf

1、Guide forVerification andValidation inComputation Weld MechanicsAWS A9.5:2013An American National StandardAWS A9.5:2013An American National StandardApproved by the American National Standards Institute October 30, 2012Guide for Verification and Validation in Computation Weld Mechanics1st EditionPrep

2、ared by the American Welding Society (AWS) A9 Committee on Computerization of Welding InformationUnder the Direction of the AWS Technical Activities CommitteeApproved by the AWS Board of DirectorsAbstractThis standard provides guidelines for assessing the capability and accuracy of computational wel

3、d mechanics (CWM)models. This standard also provides general guidance for implementing verification and validation (VInternet: .AWS A9.5:2013iiAWS A9.5:2013iiiStatement on the Use of American Welding Society StandardsAll standards (codes, specifications, recommended practices, methods, classificatio

4、ns, and guides) of the AmericanWelding Society (AWS) are voluntary consensus standards that have been developed in accordance with the rules of theAmerican National Standards Institute (ANSI). When AWS American National Standards are either incorporated in, ormade part of, documents that are include

5、d in federal or state laws and regulations, or the regulations of other governmen-tal bodies, their provisions carry the full legal authority of the statute. In such cases, any changes in those AWS standardsmust be approved by the governmental body having statutory jurisdiction before they can becom

6、e a part of those laws andregulations. In all cases, these standards carry the full legal authority of the contract or other document that invokes theAWS standards. Where this contractual relationship exists, changes in or deviations from requirements of an AWS standard must be by agreement between

7、the contracting parties.AWS American National Standards are developed through a consensus standards development process that bringstogether volunteers representing varied viewpoints and interests to achieve consensus. While AWS administers the process and establishes rules to promote fairness in the

8、 development of consensus, it does not independently test,evaluate, or verify the accuracy of any information or the soundness of any judgments contained in its standards.AWS disclaims liability for any injury to persons or to property, or other damages of any nature whatsoever, whether spe-cial, in

9、direct, consequential, or compensatory, directly or indirectly resulting from the publication, use of, or reliance onthis standard. AWS also makes no guarantee or warranty as to the accuracy or completeness of any information publishedherein.In issuing and making this standard available, AWS is neit

10、her undertaking to render professional or other services for oron behalf of any person or entity, nor is AWS undertaking to perform any duty owed by any person or entity to someoneelse. Anyone using these documents should rely on his or her own independent judgment or, as appropriate, seek theadvice

11、 of a competent professional in determining the exercise of reasonable care in any given circumstances. It isassumed that the use of this standard, and its provisions is entrusted to appropriately qualified and competent personnel.This standard may be superseded by new editions. This standard may al

12、so be corrected through publication of amend-ments or errata, or supplemented by publication of addenda. Information on the latest editions of AWS standards includ-ing amendments, errata, and addenda is posted on the AWS web page (www.aws.org). Users should ensure that they havethe latest edition, a

13、mendments, errata, and addenda.Publication of this standard does not authorize infringement of any patent or trade name. Users of this standard accept anyand all liabilities for infringement of any patent or trade name items. AWS disclaims liability for the infringement of anypatent or product trade

14、 name resulting from the use of this standard.AWS does not monitor, police, or enforce compliance with this standard, nor does it have the power to do so.Official interpretations of any of the technical requirements of this standard may only be obtained by sending a request, inwriting, to the approp

15、riate technical committee. Such requests should be addressed to the American Welding Society,Attention: Managing Director, Technical Services Division, 8669 Doral Blvd., Suite 130, Doral, FL 33166 (see Annex C).With regard to technical inquiries made concerning AWS standards, oral opinions on AWS st

16、andards may be rendered.These opinions are offered solely as a convenience to users of this standard, and they do not constitute professional advice.Such opinions represent only the personal opinions of the particular individuals giving them. These individuals do notspeak on behalf of AWS, nor do th

17、ese oral opinions constitute official or unofficial opinions or interpretations of AWS. Inaddition, oral opinions are informal and should not be used as a substitute for an official interpretation.This standard is subject to revision at any time by the AWS A9 Committee on the Computerization of Weld

18、ingInformation. It must be reviewed every five years, and if not revised, it must be either reaffirmed or withdrawn. Comments(recommendations, additions, or deletions) and any pertinent data that may be of use in improving this standard arerequired and should be addressed to AWS Headquarters. Such c

19、omments will receive careful consideration by the AWSA9 Committee on the Computerization of Welding Information and the author of the comments will be informed of theCommittees response to the comments. Guests are invited to attend all meetings of the AWS A9 Committee on theComputerization of Weldin

20、g Information to express their comments verbally. Procedures for appeal of an adverse decisionconcerning all such comments are provided in the Rules of Operation of the Technical Activities Committee. A copy ofthese Rules can be obtained from the American Welding Society, 8669 Doral Blvd., Suite 130

21、, Doral, FL 33166.AWS A9.5:2013ivThis page is intentionally blank.AWS A9.5:2013vPersonnelAWS A9 Committee on the Computerization of Welding InformationS. S. Babu, Chair The Ohio State UniversityS. N. Borrero, Secretary American Welding SocietyF. Brust EMC2D. J. Dewees The Equity Engineering Group, I

22、ncorporatedZ. Feng Oak Ridge National LaboratoryJ. A. Fleming Bridgestone AmericasJ. Goldak Goldak Technologies Inc.S. P. Khurana Axon Innovations LLCP. Michaleris Pennsylvania State UniversityC. Schwenk BMW GroupG. Sonnenberg Huntington Ingalls Industries, IncorporatedW. Zhang Oak Ridge National La

23、boratoryAdvisors to the AWS A9 Committee on the Computerization of Welding InformationA. J. Buijk Simufact-Americas, LLCR. Ganta Westinghouse Electric CompanyJ. E. Jones EnergYnTech/N.A. Tech, Inc.D. Killian Areva NP, IncorporatedJ. S. Noruk Servo Robot CorporationH. Porzner ESI GmbHE. F. Rybicki Un

24、iversity of TulsaB. T. Alexandrov The Ohio State UniversityF. Arnold SIMULIA Erie Region (Abaqus)P. Dong University of New OrleansJ. C. Kennedy Engineering Mechanics Corp of ColumbusP. F. Mendez University of AlbertaD. H. Roarty Westinghouse Electric CorporationAWS A9.5:2013viThis page is intentiona

25、lly blank.AWS A9.5:2013viiForewordThis foreword is not part of AWS A9.5:2013, Guide for Verification and Validation in Computation Weld Mechanics, but is included for informational purposes only.A task group was formed in 2007 under the AWS technical committee structure to investigate the need for c

26、omputationalweld mechanics standards. The task group was reorganized as the AWS A9 Technical Committee on the Computerizationof Welding Information and began work in 2008. This is the first standard publication by this committee with morerelated topics on computational weld mechanics (CWM) planned.P

27、rogram managers need assurance that computational models of weld mechanics are sufficiently accurate to support pro-grammatic decisions. As there are multiple acceptable approaches to analyzing the welding process using computationalmodels, a step-by-step Verification and Validation (V yet, the use

28、of computational weld mechanics (CWM) has not. It has been suggested that the same level ofconfidence in CWM analyses does not exist due to relative newness of the tools and the lack of experience in their use. Incomparison CWM is quite complex involving a coupled phenomena of thermal and nonlinear,

29、 transient structural analyses.Information regarding material responses due to thermal inputs, microstructure evolution, and to stresses and strains areneeded to perform this type of analysis. It is for these reasons that CWM has emerged about two decades later than CSM.The process to develop confid

30、ence in computational modeling can be expedited by a process called verification and validation (V” extensive user programming is typically required to approach the level of detail captured inmany welding-specific software packages.AWS A9.5:20136Source: Reprinted, with permission, from Scandinavian

31、Journal of Metallurgy, Wiley 2Figure 3Schematic Illustration of Integrated Computational Weld Mechanics ApproachThe development of integrated thermo-mechanical-metallurgical models is indeed challenging due to the complex inter-action between physical processes during welding. Some of these physical

32、 processes include heat and mass transfer,phase transformations, electro-magnetic phenomenon, plastic strain, and reactions with environment during welding.Researchers have developed a framework for linking thermo-mechanical histories to microstructure development andmechanical heterogeneity in weld

33、s (see Annex BPart C). These developments are again summarized in Figure 3.According to Figure 3, by integrating individual sub-models for heat, mechanical, and material models, one can predictthe overall performance of welded structures. The approach starts with a heat-transfer model that simulates

34、 temperaturedistributions in three-dimensions T = f (x, y, z, time) as a function of process parameters and time. Thermal cycle datawill be used by material models to predict the microstructure evolution and its impact on transient mechanical (- rela-tions) properties. The transient changes in tempe

35、rature and mechanical properties will be fed into a finite-element struc-tural model to predict plastic strain distribution. This information allows for the prediction of final properties, residualstress, and distortion in a complex welded geometry. This interdisciplinary approach may appear simple;

36、 however, itrequires collaboration between experts in metallurgy, finite element analyses, welding process, and computer science62, 63. To a limited extent, this vision has become a reality by pioneering work in many organizations and commercialsoftware companies (see Annex BPart D). There exist sev

37、eral conference proceedings that provide a detailed progression of integrated weld modeling and its capability for a wide range of joining processes (see Annex BPart E).In addition, there are several useful classic textbooks related to modeling of welding processes (see Annex BPart F).5.3 Key Analys

38、is Inputs. Key inputs are those that are either fundamental to the analysis, or that the analysis will beparticularly sensitive to them. A CWM may contain thermal, metallurgical and structural analyses, and inputs andassumptions are addressed with respect to each.5.3.1 Material Properties. Accuracy

39、of the prediction by CWM relies on the accuracy of thermal, physical,mechanical, and metallurgical properties used by the models. Universally accepted, codified material properties for awide range of materials do not exist as they do in other technology areas (e.g., the NIST Database of chemical pro

40、perties,or the ASME Boiler and Pressure Vessel Code Section II, “Material Properties”). This unfortunate combination ofsensitivity and uncertainty is largely counteracted by the use of a concept called the cutoff temperature; this concept isaddressed in 5.6.1. For a certain subset of analyses, great

41、ly simplified material properties have been found to be adequatewhen used in combination with test data. Examples are contained in the methods described in 5.7.Some of the properties required for a general CWM include thermal conductivity, the coefficient of thermal expansion,density, specific heat,

42、 heat of fusion or transformation, elastic modulus, Poissons ratio, yield strength, tensile strength,strain hardening exponent, total elongation or ductility, and fracture strain. It is important for the analyst to understandthe implications of material model selection. The selection of any of the a

43、bove approaches must be related to the end goalof an accurate CWM analysis. In the published literature, many approaches have been used for selection of the aboveproperties. Some of these are:(a) The most common approach is to assume that these properties do not change with ensuing temperature and s

44、train(or stress) during welding. Here the analyst has instant access to these from standard literature and minimizes materialdata development costs while reducing computational impact. However, some of the process physics may not be fullydepicted as seen in (b) and (c) and may introduce accuracy err

45、ors.(b) The second approach is to introduce additional complexity of temperature dependence of these properties.However, a much more complicated material model must be developedgreater accuracy in the results should beobtained. Generating the data requires significant investments in cost and testing

46、. However, greater accuracy may lead tohigher confidence in software results and may drive further optimization efforts. Additionally, material model selectionmay force the type of required representative model required. To accurately capture the entire phenomena that are occur-ring, 3D elements may

47、 be required, thus increasing the complexity of the modeling, the computational time, and the post-processing efforts.(c) To support the previous statement, it is well known that during welding of steels, the thermal expansion coefficientand density changes as a function of temperature and is differ

48、ent for the heating and cooling cycles. This is related to hys-teresis in ferrite to austenite (on heating) and austenite to ferrite (on cooling). These solid phase transformations thatoccur during the thermal cycle produced by welding lead to irreversible plastic deformation at the microstructural

49、scaleknown as transformation plasticity. This phenomenon is driven by the volume change during solid-state phase transfor-mations 3, 4, 5, 6, 7 and transformation plasticity (a term proportional to the deviatoric stress tensor) 3, 4, 5, 6, 7. Theabove effects are shown to be inherently 3D 8.AWS A9.5:20137The third approach relies on describing these properties as a function of the history or path dependence during welding.This approach is highly dependent on the process used. It requires model calibration to the empirically collected data. Forexample, during welding of st