MSC 1994 World Users' Conference Proceedings

The conference proceedings for the 1994 World Users' Conference are now available on-line in Adobe Acrobat PDF format. (The Adobe Acrobat Reader software is available for free download from Adobe's web site at www.adobe.com) When printed, these PDF files will produce a better quality image than the one shown on your computer screen. All papers have been categorized by topic.

Categories

Civil Structures

ANALYSIS OF HYPERBOLIC SHELL NATURAL DRAUGHT COOLING TOWERS USING MSC/NASTRAN (Acrobat 623KB) #3994
V. S. N. Raju -- A.P.S.E Board/Hyderabad, Andhra Pradesh
 
ABSTRACT:Natural Draught Cooling Towers are very common in modern day thermal and nuclear power stations. These towers with very small shell thickness are exceptional structures by their sheer size and sensitivity to horizontal loads. This paper discusses the analysis of R.C.C. hyperbolic shell of a Natural Draught Cooling Tower using Quad8/Quad4 elements of MSC/NASTRAN. Analysis has been carried out for different cases by varying the mesh size and aspect ratio of Quad8/Quad4 elements and the results have been compared. Not much variation in results has been found for these cases except at the base of the shell. At the base of the shell the Quad8 results are much at variance from quad4 and SAPIV results.
 
THE INFLUENCE OF PHASE-DIFFERENCE EFFECTS ON EARTHQUAKE RESPONSE OF CABLE-STAYED BRIDGES (Acrobat 294KB) #3794
Hongye Zhou -- China Academy of Railway Sciences
Youping Chen -- China Academy of Railway Sciences
 
ABSTRACT: In this paper, a model of cable-stayed bridge for earthquake response analysis is established. The earthquake response analysis is performed by using a large-stiffness technique with MSC/NASTRAN. The influence of phase-difference effects on the earthquake response is discussed.
 
REFINED ANALYSIS FOR SOIL-PIPE SYSTEMS (Acrobat 765KB) #3894
Yohchia Chen, Ph.D., P.E.-- The Pennsylvania State University
 
ABSTRACT: This paper discusses a finite element-based refined approach for buried pipes under severe earthquake effects. Modeling and computation techniques using MSC/NASTRAN are described in detail. A number of realistic pipes are investigated. The ref med analysis results are compared with those obtained from a simplified method which is the currently accepted method. Special features and advantages of the program are demonstrated. Recommendations for analyzing and designing a buried straight pipe are also made.

DMAP Applications

DMAP ALTERS FOR CALCULATING AND SUMMARIZING RESULTANT FORCES FROM THE APPLICATION OF SPC AND AUTOSPC CONSTRAINTS (Acrobat 983KB) #994
Grant Parker -- Northrop Corporation
John Brown -- Northrop Corporation
Ted Rose -- The MSC.Software Corporation
Jim Swan -- The MSC.Software Corporation
 
ABSTRACT: MSC/NASTRAN DMAP (Direct Matrix Abstraction Programs) alter programs have been written to identify and summarize all resultant forces from single point constraints (SPC) and AUTOSPC constraints. This capability was developed primarily as an enhancement to previous DMAPs for model checking and identifying grounding problems and constraint checks through the use of strain energy calculations on the stiffness matrix.

Discussion and examples of the use of the DMAPs for model verification are presented.

STIFFNESS-GENERATED RIGID-BODY MODE SHAPES FOR LANCZOS EIGENSOLUIION WITH SUPORT DOF VIA A MSC/NASTRAN DMAP ALTER (Acrobat 694KB) #1094
Ayman A. Abdallah -- Analex Corporation
Alan R Barnett -- Analex Corporation
Timothy W. Widrick -- Analex Corporation
Richard T Manella -- National Aeronautics and Space Administration
Robert P. Miller -- National Aeronautics and Space Administration
 
ABSTRACT: When using all MSC/NASTRAN eigensolution methods except Lanczos, the analyst can replace the coupled system rigid-body modes calculated within DMAP module READ with mass orthogonalized and normalized rigid-body modes generated from the system stiffness. This option is invoked by defining MSC/NASTRAN r-set degrees-of- freedom via the SUPORT Hulk Data card. The newly calculated modes are required if the rigid-body modes calculated by the eigensolver are not "clean" due to numerical roundoffs in the solution. When performing transient structural dynamic load analyses, the numerical roundoffs can result in inaccurate rigid-body accelerations which affect steady-state responses. Unfortunately, when using the Lanczos method and defining r-set degrees-of-freedom, the rigid-body modes calculated within DMAP module REIGL are retained. To overcome this limitation and to allow MSC/NASTRAN to handle SUPORT degrees-of-freedom identically for all eigensolvers, a DMAP Alter has been written which replaces Lanczos-calculated rigid-body modes with stiffness-generated rigid-body modes. The newly generated rigid-body modes are normalized with respect to the system mass and orthogonalized using the Gram-Schmidt technique. This algorithm has been implemented as an enhancement to an existing coupled loads methodology.
 
TRANSIENT ANALYSIS MODE PARTICIPATION FOR MODAL SURVEY TARGET MODE SELECTION USING MSC/NASTRAN DMAP (Acrobat 1.14MB) #894
Alan R. Bamett -- Analex Corporation
Omar M. Ibrahim -- Analex Corporation
Timothy L. Sullivan -- Analex Corporation
Thomas W. Goodnight -- National Aeronautics and Space Administration
 
ABSTRACT: Many methods have been developed to aid analysts in identifying component modes which contribute significantly to component responses. These modes, typically targeted for dynamic model correlation via a modal survey, are known as target modes. Most methods used to identify target modes are based on component global dynamic behavior. It is sometimes unclear if these methods identify all modes contributing to responses important to the analyst. These responses are usually those in areas of hardware design concerns. One method used to check the completeness of target mode sets and identify modes contributing significantly to important component responses is mode participation. With this method, the participation of component modes in dynamic responses is quantified. Those modes which have high participation are likely modal survey target modes. Mode participation is most beneficial when it is used with responses from analyses simulating actual flight events. For spacecraft, these responses are generated via a structural dynamic coupled loads analysis. Using MSC/NASTRAN DMAP, a method has been developed for calculating mode participation based on transient coupled loads analysis results. The algorithm has been implemented to be compatible with an existing coupled loads methodology and has been used successfully to develop a set of modal survey target modes.

Dynamic Analysis

ACCURATE ENFORCED MOTION ANALYSIS USING MSC/NASTRAN SUPERELEMENTS
(Acrobat 666KB) #2594
Christopher C. Flanigan -- SDRC Engineering Services Division, Inc.
 
ABSTRACT: The standard approach for performing an enforced motion analysis in MSC/NASTRAN uses very large masses and forces to obtain the desired motion at selected locations. This approach can lead to inaccurate results if the large masses are too large or too small.

An alternate approach for enforced motion analysis is presented in this paper. The alternate method uses the Craig-Bampton superelement capability in MSC/NASTRAN to form the required matrices for a direct solution of the equations of enforced motion. The need for large masses is eliminated, resulting in improved accuracy. In addition, the enforced motion analysis is performed directly, eliminating the need for Lagrange multipliers.

A rigid format alter for performing the new enforced motion analysis method is included in the paper. An example problem is presented to demonstrate the new method and to illustrate some of the pitfalls of enforced motion analysis.

ACOUSTIC SENSITIVITY ANALYSIS USING BOUNDARY ELEMENTS AND STRUCTURAL DYNAMIC RESPONSE (Acrobat 737KB) #794
Nickolas Vlahopoulos -- Automated Analysis Corporation
S.T. Raveendra -- Automated Analysis Corporation
Chris Mollo -- Automated Analysis Corporation
 
ABSTRACT: Structural/acoustic sensitivity analysis can provide information on how structural design modifications can affect the noise emitted from vibrating structures. MSC/NASTRAN S0L108 and SOLl1 can be used to compute structural dynamic frequency response sensitivities. This information can be coupled with Boundary Element (BE) results to derive the acoustic sensitivity with respect to structural design variables.

In this work two BE formulations, the Direct and Indirect, have been coupled to MSC/NASTRAN design sensitivity results to compute the acoustic sensitivity with respect to structural design variables. The theoretical background of the coupling algorithm is presented along with some examples of noise prediction in interior and exterior noise radiation.

AEROELASTIC MODELLING OF AN AIRPLANE WITH STAND-BY ACTUATOR USING MSC/NASTRAN FOR FLUTTER ANALYSIS (Acrobat 481KB) #3394
Yan Mursal -- Pt.IPTN/Indonesia Aircraft Industry Ltd.
Moh. Risdaya Fadil -- Pt.IPTN/Indonesia Aircraft Industry Ltd.
 
ABSTRACT: The aircraft control surfaces that were considered in this paper are driven by a pair of hydraulically powered servo actuators. One actuator is normally in an active mode and the other is normally in a stand-by mode. This paper describes an aeroelastic modeling technique where a control System has two hydraulic failures, the active mode actuator is failed (e.g. a structural disconnect) and no hydraulic power comes to the stand-by mode actuator. In this situation the stand-by mode acts as a hydraulic damper.

In the failure condition, the stand-by actuator must provide sufficient damping in order the airplane still maintain flutter free condition. To perform the aeroelastic analysis of the System, the generalized mass, stiffness and damping of the plant (airplane) equation must be modified. These tasks were done using the combination of EPOINT, TF MSC/NASTRAN bulk data entry and DMAP. The EPOINT entry was used to add one generalized coordinate. In this case it is due to the moment of the actuator introduced to the airplane. The Nastran TF bulk data was used to introduce the diagonal terms of the MHH, BHH and KHH matrices. A small DMAP routine was created to add off-diagonal terms of these matrices and to perform the analysis automatically. The calculation was done using SOL 145. Some results are presented as an example and also compared with another method.

ANALYSIS AND MODAL SURVEY TEST INTELSAT VIIA DEPLOYED SOLAR ARRAY (Acrobat 469KB) #2794
K. C. Jiang -- Space Systems/Loral
R. Gahart -- Space Systems/Loral
 
ABSTRACT: A nonlinear normal modes analysis and a modal survey test have been performed to determine the frequencies and mode shapes of a large flexible deployed solar array for the Intelsat VIIA series communication satellites under production at Space Systems/Loral. Test parameters such as gravity, air mass and test fixture stiffness are included in the analysis. MSC/NASTRAN Solution 105 was used to generate the gravitational stiffness due to 1-g gravity and Solution 103 was used to calculate the frequencies. A DMAP Alter was updated for MSC/NASTRAN Version 67 to combine the Solution 105 gravitational stiffness matrix and Solution 103 structural stiffness matrix. The modal survey test was performed to measure frequencies and mode shapes for correlation with the analytical results. The deployed solar array was hung vertically. The solar array was pulled and released suddenly to excite transient vibrations. The decaying vibration response was measured and analyzed to deduce the frequencies. The test results show good correlation with the predictions.
 
DYNAMIC DESIGN ANALYSIS METHOD (DDAM) USING MSC/NASTRAN (Acrobat 314KB) #3194
Pam Barber -- Newport News Shipbuilding
Kevin Arden -- Newport News Shipbuilding
 
ABSTRACT: Components on naval ships are currently analyzed for shock loads due to hostile attacks using the Dynamic Design Analysis Method (DDAM). DDAM estimates the dynamic response of a component to a beam excitation resulting from the motion of the ship's hull.

This paper gives a brief description of the history and use of DDAM, presents an overview of the mathematics, and demonstrates the use of DDAM in designing a typical submarine component.

The demonstration problem uses MSC/NASTRAN for the dynamic analysis and modal summations, and shows how DDAM can be performed using MSC/NASTRAN Solution 103.

The opinions herein are that of the authors' and do not necessarily represent the opinion of Newport News Shipbuilding.
 
AN EFFICIENT PROCEDURE FOR DATA RECOVERY OF A CRAIG-BAMPTON COMPONENT
(Acrobat 506KB) #2694
Edward T. Tong -- Rockwell International
Craig C.J. Chang -- Rockwell International
 
INTRODUCTION: Dynamic analyses of large and complex space structures are generally performed using the modal approach. Component modal syntheses are used frequently so that various companies, each designing their own hardware, can supply the component models to the integrator for System dynamic analysis. Of the various techniques of component modal reduction, the Craig-Bampton procedure is most commonly used. The deliverable data usually consists of the component Craig-Bampton generalized mass and stiffness matrices, and the Craig-Bampton transformation matrix. Linear Transformation Matrices (LTMs) associated with the model generated by the component companies are used to facilitate data recovery. Sizes of these LTMs range from a few key bad indicators which are of interest to the integrator to an extensive data recovery set that drives the component's design. Component modal synthesis and data recovery therefore form a tandum in dynamic analysis of large space structures. The two widely used procedures to compute the LTMs are the mode displacement approach and the mode acceleration approach. Although the mode acceleration approach is generally perceived to be an improvement over the mode displacement approach, there are considerations when choosing a data recovery method.

This paper will formulate a simplified procedure to construct the Craig-Bampton LTMs using either the mode acceleration approach or the mode displacement approach. The procedure will be formulated for the general case of statically indeterminant structures. Further simplification can be made for statically determinant Craig-Bampton components. A general purpose DMAP routine of MSC/NASTRAN version 67 encompassing the tandum has been implemented. Sample problems to demonstrate the procedure with the DMAP are included. Although the DMAP is written for a Craig Bampton component, it can be extended to a general modally reduced model.

A disadvantage of the mode acceleration recovery is the costly computation to construct the LTMs when an applied load at the interior set is present. One of the recent applications is the plume implingement on the solar arrays during Shuttle docking with the Space Station. This paper will discuss a cost saving technique and suggest an approach which reduces cost by utilizing a 'mixture' of the two approaches.
 
SHAKE TEST SIMULATION USING MSC/NASTRAN (Acrobat 744KB) #3294
Randal H. Visintainer -- Ford Motor Company
Farhang Aslani -- Automated Analysis Corporation
 
ABSTRACT: Advances in Computer-Aided Testing technology have led to the Simulation of Service Operation and test track loads in the laboratory. The controlled environment of the lab creates an ideal Situation for attempts to simulate the lab test. Furthermore, the concept of up-front engineering using analytical prototype has pushed the demands on the CAE analysts to develop methodology to simulate operating conditions, durability events, and laboratory tests. To address this need, the frequency response solution (SOL 111) of MSC/NASTRAN is used to simulate the vehicle shake test of a heavy truck. The displacement-controlled lab test was performed between the frequency range of 20 to 50 Hertz (Hz). The MSC/NASTRAN simulation of the test and the correlation between the test and analysis results revealed several important points that should be considered in a computer simulation using the analytical prototype.
 
SIMULATION OF PROVING GROUND EVENTS FOR HEAVY TRUCK CABS USING ADAMS, MSC/NASTRAN, and P/FATIGUE (Acrobat 862KB) #594
Farhang Aslani -- Automated Analysis Corporation
Manicka Yatheendar -- Automated Analysis Corporation
Randal H. Visintainer -- Ford Motor Company
David S. Rohweder -- Ford Motor Company
Juan Lopez de Alda -- Mechanical Dynamics Inc.
 
ABSTRACT: The concept of up-front engineering using analytical prototype has pushed the demands on the CAE analysts to develop methodologies which simulate operating conditions, proving ground events, and laboratory tests well ahead of the availability of structural prototypes of the vehicle. One purpose of such analysis is to provide fatigue life prediction at an early stage of the design to permit the incorporation of design modifications prior to initial tooling release. The stress history resulting from the operation of the vehicle on the proving ground events can be simulated by enforcement of the profile of proving ground events using transient analysis in MSC/NASTRAN However, some of the proving ground events for heavy truck cabs include tire patch displacement of 12 inches that can not be simulated in MSC/NASTRAN directly. A methodology has been developed for use of the ADAMS vehicle model which provides accelerations at the cab mounts under those events. The acceleration response from the ADAMS model was used to perform transient analysis in MSC/NASTRAN to predict the time history of stresses in the cab. The time history of the stresses are input to the P/FATIGUE program to predict fatigue life of the structure under specific proving ground events.
 
A STUDY ON THE ANALYSIS AND IMPROVEMENT OF THE ACOUSTIC CHARACTERISTICS OF THE MUFFLER WITH COMPLEX GEOMETRY (Acrobat 466KB) #694
Jin-Yong Mo -- SAMSUNG ELECTRONICS Co
Man-Sun Huh -- SAMSUNG ELECTRONICS Co
Jung-Rae Kim -- SAMSUNG ELECTRONICS Co
Heui-Jong Kang -- SAMSUNG ELECTRONICS Co
Sang-Kyoung Oh -- SAMSUNG ELECTRONICS Co
 
ABSTRACT: An acoustic transmission loss analysis method for mufflers with complex geometry is developed using MSC/NASTRAN on the basis of acoustic-structural analogy and two-microphone method. In this study, mufflers with simple and complex shapes are analyzed using this method and compared with theoretical and experimental results for verification.

Applying this method to design of suction muffler of reciprocating compressor, we could obtain 4dB(A) total noise reduction, especially more than 10dB(A) noise reduction at 500Hz.

Interfaces to MSC/NASTRAN

AN ADVANCED POST PROCESSING METHODOLOGY FOR VIEWING MSC/NASTRAN GENERATED ANALYSES RESULTS (Acrobat 529KB) #2194
Michael McMeekin -- Swales and Associates
Paul Kirchman -- Swales and Associates
 
ABSTRACT: This paper presents an advanced post processing methodology for viewing analysis results generated from MSC/NASTRAN normal modes analyses, and transient response analyses. In both cases the analysis results are added as a function of time to the basic finite element model geometries through specified coordinate transformations, creating a new deformed geometry at each time step. The final deformed geometries are rendered for each time step and animated through a public domain software utility on a personal computer for a final visualization of the analysis results. The deformed geometries can include adjunct super-elements or Craig-Bampton substructures. The same methodology has been successfully applied to time history data from a commercial kinematic software package using the basic finite element model geometry from MSC/NASTRAN.
 
CREATING A FINITE ELEMENT MODEL WITH THE MSC.Aries APPLICATION MODULES (Acrobat 993KB) #2394
Marianne L. Kronenberg -- The MSC.Software Corporation
 
ABSTRACT: Software generated solid models provide complete and unambiguous visual and mathematical definition of design objects. The MSC.Aries Solids software application module affords an easy to use, highly intuitive way of creating solid models to capture the design intent of engineers and analysts. The objectives of this presentation are to:
  • introduce the techniques applied to build and modify solid models.
  • describe the transition from the geometric model to the Finite Element model.
  • display the results of a linear static finite element analysis.
USING VERSION 67.5 OUTPUT2 NEUTRAL FORMAT OPTION FOR FILE TRANSFER ACROSS DISSIMILAR MACHINES (Acrobat 297KB) #2494
John Brown -- Northrop Corporation
Grant Parker -- Northrop Corporation
Ted Rose -- The MSC.Software Corporation
ABSTRACT: A new option in Version 67.5 provides the ability to write OUTPUT2 files in a machine-independent format. Northrop's Aircraft Division F/A-18 E/F Program has migrated a large portion of its engineering computer use from a large mainframe to a multi-node client/server networked workstation environment. Finite element modeling and analysis software, both proprietary and commercial, have been migrated to the workstation - including MSC/NASTRAN.

Special features in Northrop's mainframe based proprietary pre and post processing code (NCASA) has required its continued use. This mainframe program requires OUTPUT2 data from MSC/NASTRAN. Prior to the introduction of workstations, both NCASA and MSC/NASTRAN were resident on the mainframe, and transfer of files between the programs was not a problem.

In the workstation environment, the binary incompatibility with the mainframe precludes the use of MSC/NASTRAN on the workstation if NCASA post processing capabilities are required. The new OUTPUT2 option in V67.5, which produces a machine-independent compressed ASCII format, allows all MSC/NASTRAN processing to be accomplished on the workstation which results in large cost savings and productivity gains for Northrop.

The implementation of the OUTPUT2 Neutral Format Option, the RCOUT2 MSC supplied FORTRAN conversion program on the mainframe, and use of the MSC/NASTRAN Data Definition Language (NDDL) are presented.

VARIATION ANALYSIS BY MONTE CARLO RANDOMIZATION OF LOAD VARIATION SENSITIVITIES (Acrobat 1.38MB) #2294
Mark J. Stone -- Eastman Kodak Company
 
ABSTRACT: Most analysis considers only the nominal loads acting on a structure, but there may be significant impact due to the variation or error in the loads as well. When there are multiple load sources, the effect from the combination of these load variations is difficult, if not impossible, to predict. This paper describes the use of a Monte Carlo randomization method applied to the displacement results generated from MSC/NASTRAN analyses using sensitivity loads. A Monte Carlo process is used to efficiently obtain a statistical distribution of possible results from the random combination of load variations. Using the method presented minimizes the number of analyses which must be run in order to obtain a population of results from which accurate conclusions can be drawn. The model used represents the High Resolution Mirror Assembly (HRMA) for NASA's Advanced X-ray Astrophysics Facility-Imaging (AXAF-I). The variation analysis discussed considers the impact from support induced load variations during alignment and assembly of the AXAF-I mirrors to the mounting structure.

Nonlinear Analysis

BUCKLING AND GEOMETRIC NONLINEAR ANALYSIS OF A TIE ROD IN MSC/NASTRAN VERSION 68
(Acrobat 630KB) #1594
George Campbell -- Ford Motor Company
Wen Ting -- Ford Motor Company
Peyman Aghssa -- The MSC.Software Corporation
Claus C. Hoff -- The MSC.Software Corporation
 
ABSTRACT: The improved geometric nonlinear capability in MSC/NASTRAN Version 68 is tested on a large scale finite element model of a tie rod. The static buckling load of a tie rod is analyzed. The results of the finite element model are compared with experimental results. The analysis is performed in three steps. First, linear buckling is analyzed with SOL 105. Second, a nonlinear static analysis with arc-length method is performed in SOL 106 to determine the instability behavior of the structure. In the last step, a nonlinear buckling analysis is done with restart into SOL 106 to determine the nonlinear buckling load. The tie rod has a strongly nonlinear behavior which is due to material yield and geometric nonlinear effects. It is shown that MSC/NASTRAN's computed buckling load agrees well with the experimental buckling load.
 
A NEW METHOD DEVELOPMENT TO PREDICT BRAKE SQUEAL OCCURRENCE (Acrobat 453KB) #1494
Lajos I. Nagy -- Ford Motor Company
James Cheng -- Ford Motor Company
Yu-Kan Hu -- BQUAD Engineering, Inc.
 
ABSTRACT: A new method to predict brake squeal occurrence was developed by MSC under contract to Ford Motor Company. The results indicate that the stability characteristics of this disc brake assembly are governed mainly by the frictional properties between the pads and rotor. The stability is achieved when the friction coefficient of the pads is decreasing as the contact force increases. Based on the results, a stable brake system can be obtained without changing the brake structure by incorporating the appropriate frictional coefficient in the brake system. The method developed here can be also used as a tool to test the quality of any brake design in the early design stage.
 
NONLINEAR ANALYSIS OF FREE FLIGHT ROCKETS USING MSC/NASTRAN (Acrobat 511KB) #3694
David S. Livshits -- TAAS - Israel Industries Ltd.
David Saltoun -- MSI Engineering Software Ltd
 
ABSTRACT: A finite element model for nonlinear dynamic analysis of a free flight rocket is developed in this paper. The rocket response time history is calculated as a result of aerodynamic loads, dynamic imbalance and thrust misalignment. The aerodynamic loads are calculated using combinations of NOLINs. The model can perform aeroelastic stability analysis and loads distribution calculations as a function of time. A nonlinear aerodynamic behavior can be included in the model for large angles of attack.
 
PRODUCTION ORIENTED NONLINEAR ANALYSIS OF SOLIDS AND STRUCTURES (Acrobat 917KB) #394
H. D Hibbitt -- Hibbitt, Karlsson & Sorensen, Inc.
 
ABSTRACT: Requirements for modeling nonlinear effects in routine analysis applications have grown to the point where "general purpose" finite element-based programs are expected to offer significant nonlinear modeling capabilities. One result of this market demand has been this year's announcement of the establishment of a long term relationship between The MSC.Software Corporation and Hibbitt, Karlsson Sorensen, Inc. ("HKS"), whereby MSC will package a substantial set of the capabilities offered by HKS's ABAQUS/Standard program with MSC.Aries, to supplement the nonlinear capabilities of MSC/NASTRAN and MSC/DYTRAN for applications in solid and structural analysis. MSC will provide full support of these capabilities, as it does for its other analysis products.

Nonlinear effects introduce a broad range of issues which might deter the analyst who is unfamiliar with this type of problem from trusting such modeling as a basis for achieving design goals and schedules. Once nonlinearity is introduced into a model, uniqueness and stability of the solution may be (and often are) lost, and issues of convergence, choice of nonlinear solution algorithm, etc. must be considered. Nevertheless, the analyst may have no choice but to face up to these problems: he cannot analyze a design for certain events, or design the manufacturing process to create a product, without considering nonlinearity. The viewpoint taken in this paper is that, with mature software such as the ABAQUS-based products that MSC now offers, some nonlinear effects of practical importance can be modeled on a routine, production, basis. The spectrum of difficulty ranges from such cases all the way to problems that are still research topics. One purpose of this paper is to suggest what level of difficulty might be anticipated in modeling various nonlinear effects that are commonly encountered, thus providing some guidance to the analyst in determining the extent to which expertise is needed in order to utilize nonlinear analysis software.
 
SPACE STATION FREEDOM SOLAR ARRAY WING NONLINEAR TRANSIENT ANALYSIS OF PLUME IMPINGEMENT LOAD (Acrobat 767KB) #3594
C. C. Tang -- Lockheed Missile & Space Company
 
ABSTRACT: This paper describes the nonlinear transient analysis of the Space Station Freedom Solar Array (SSFSA) wing for on-orbit plume impingement bad due to Space Shuttle berthing. Design features and the finite element model of the SSFSA wing are briefly described. Nonlinear transient analysis is performed using MSC/NASTRAN SOL 99 (V67) with blanket tensioning accomplished by restarting with static solution results. The blanket tensioning is from tension mechanisms. Transfer function (TF), scalar point (SPOINT), nonlinear load (NOLIN1), and damper (CDAMP2) are used to describe the nonlinear characteristics of the tension mechanisms. Stiffness updates for capturing the nonlinear geometrical stiffness changes due to tension variation in the solar array blankets is utilized in the iterative nonlinear solution. Results, when compared to that from linear transient analysis, showed that the beam-column effect for the slender mast of the solar array wing is insignificant, and the assumptions made in the linear transient analysis are acceptable.
 
THREE-DIMENSIONAL SLIDELINE CONTACT IN VERSION 68 (Acrobat 553KB) #1694
Rakesh Allahabadi -- The MSC.Software Corporation
Dianxiang Xiang -- The MSC.Software Corporation
Detlef Liebe -- MSGmbH
Glenn C. Grassi -- The MSC.Software Corporation
Peyman Aghssa -- The MSC.Software Corporation
Behrooz K. Shahidi -- Ford Motor Company
 
ABSTRACT: Slideline contact was first introduced in Version 67.5 of MSC/NASTRAN through a standalone DMAP alter - 'contact.v675' available in the /misc/sssalter directory. This DMAP alter is applicable only for solving static problems. Starting with Version 68, the slideline contact capability is available for solving both nonlinear static (SOL 106) and nonlinear transient (SOL 129) problems as a standard feature. The use of the DMAP alter is no longer required.

This paper shares the experience of using the MSC/NASTRAN slideline contact capability. It (a) demonstrates the capability through real life applications and (b) provides guidelines for effective usage.

VULNERABILITY ANALYSIS OF AIRCRAFT FUSELAGE SUBJECTED TO INTERNAL DETONATIONS
(Acrobat 990 KB) #3494
Geetha Bharatram -- Wright Patterson AFB
Capt. Scott A. Schimmels -- Wright Patterson AFB
Dr. Vipperla B. Venkayya -- Wright Patterson AFB
 
ABSTRACT: The Air Force, in support of the FAA's (Federal Aviation Administration) Aircraft Hardening Program (AHP), is conducting an extensive test program involving simple cylinders to full scale aircraft such as the B-52s as well as representative commercial airplanes. The purpose of this program is to ascertain the extent of the damage caused by internal explosions and to develop strategies to protect the safety of the passengers. The effects of the internal explosion are very complex, and the tests alone can not provide an adequate understanding to develop protection strategies. The purpose of this paper is to present the analysis results of the B-52 aircraft subjected to internal explosions. In addition analysis results are compared to those obtained from the test program conducted at Davis Monthan Air Force Base.
 
Two types of analysis will be addressed in this paper:
 
1. Fluid structure interaction (blast pressures and airframe interaction) using MSC/DYTRAN.
2. Joint and buckling analysis of a B-52 panel using MSC/DYTRAN.
 
The ultimate goal of the proposed analysis is to develop a vulnerability map of the entire fuselage. This map can be used to make cost effective decisions on hardening of the aircraft against bomb blasts.

Optimization

NONLINEAR ADAPTIVE ANALYSIS VIA QUASI-NEWTON APPROACH WITH MSC/NASTRAN
(Acrobat 802KB) #1394
Ortwin .cfmer -- California State University, Long Beach
 
ABSTRACT: The Quasi-Newton method has proven to be the most efficient optimization method. The purpose of this paper is to apply this numerical procedure for optimization problems as well as large deflection analysis and animation. A FORTRAN program developed to calculate constrained optimization problems is used as the basic code within an iterative nonlinear adaptive analysis. The new numerical procedure calculates the displacements of an elastic structure due to given loading conditions. Then the displacements are added to the joint coordinates. In the deformed position the degrees of freedom of the structure are supported and the negative displacements are applied as loadings, to move the structure back to the old undeformed position. The difference of the reaction forces in both positions specifies the geometric nonlinear adaptive loading conditions. These additional forces are applied in an iteration procedure, until equilibrium is achieved. The software ME-BANK (Mechanical Engineering Program-Bank), written in C-language, was developed to execute MSC/NASTRAN and a constrained optimization FORTRAN-code via the SYSTEM-function within an iteration procedure.
 
SHAPE PARAMETERIZATION AND OPTIMIZATION USING THE BOUNDARY SHAPES CONCEPT
(Acrobat 1.10MB) #1194
Hemant D Patel -- The MSC.Software Corporation
 
ABSTRACT: The concept of boundary shapes for parameterization of design boundaries in shape optimization is introduced. In this concept the boundary definition of the finite element geometry is designed. In addition the requirement for use of shape basis vectors with low mesh distortion properties is ideally satisfied by interpolating domain point sensitivity by applying the boundary shapes or forms as enforced displacements and updating the shape basis vectors at each design cycle based on current geometry. This results in smooth mesh changes and minimizes the need for intermediate remeshing for small to moderate design changes. To support ease of use and provide flexibility in the prescription of boundary shapes the concept of auxiliary boundary models as been incorporated in MSC/NASTRAN as an integral part of the analysis model. The boundary shapes are generated with auxiliary boundary model analysis by exploiting available options in static analysis of applied loadings and multiple boundary conditions. Basic design examples demonstrating the power of the boundary shape approach are presented.
 
STRUCTURAL OPTIMIZATION WITH SOLUTION 2001 IN THE DESIGN PROCESS (Acrobat 2.57MB) #494
Ingo Raasch -- BMW AG, Munich, Germany
 
ABSTRACT: The paper will explain the initiation and capabilities of S0L2001, a MSC/NASTRAN DMAP for structural optimization. The reminder will describe example problems, where 50L2001 has successfully used. Two examples show the achievements in concept design, followed by an example of a car body in white. Finally some shape optimization problems of engine components will be shown. A short outlook to the optimization capabilities of Version 68 will be given. In the conclusions it will be stressed that optimization is a very valuable (even in the meaning of $$) tool in the design process.
 
TOPOLOGY OPTIMIZATION USING MSC/NASTRAN (Acrobat 990KB) #1294
B. P. Wang -- The University of Texas at Arlington
C M. Lu -- The University of Texas at Arlington
R.J. Yang -- Ford Motor Company
 
ABSTRACT: Recently, Bendsoe and Kikuchi developed a homogenization method which can be applied to find the optimal topology of a continuum in a fixed domain. The homogenization approach is based on an artificial but physical micro-structure whose properties are homogenized. Alternatively, it has been demonstrated that the solution of the optimum material distribution problem can be considerably simplified by employing a density-dependent isotropic material without a specific physical micro-structure. In this paper, topology optimization for minimum compliance under static loading and for maximum eigenvalue using this approach has been implemented using MSC/NASTRAN. Optimal topology for a plate under in-plane and bending loads is presented. Optimal material distribution for a plate to maximize the first frequency is also presented.

Stress Analysis

THE EFFICIENT CALCULATION OF STRESS CONCENTRATION FACTORS (Acrobat 891KB) #3094
D. L. Dewhirst -- Ford Motor Co.
Wai Ho -- The MSC.Software Corporation
 
ABSTRACT: Certain stress concentration problems recur often in the design of torque-carrying shafts. We desired to construct a set of tables similar to those in Peterson's Stress Concentration Handbook to permit rapid manual preliminary sizing of these shafts. Because a large number of parameters were involved, the efficiency of the modeling process was an issue. This study compares a variety of modeling techniques with emphasis on the p-adaptive approach available in MSC/NASTRAN, version 68.
 
ELASTIC-PLASTIC ANALYSIS AROUND A CIRCULAR HOLE STRESS CONCENTRATION (Acrobat 752KB) #2894
Robert P. Thacker, Jr. -- Boeing Computer Support Services
 
ABSTRACT: This paper takes a look at the ability of the 8 node solid (CHEXA) and the 4 node shell (CQUAD4) to calculate the plastic surface around a circular hole, when the plate is loaded bi-axially. The MSC calculated stresses will be compared to two analytical solutions; one by Galin, the other by Sokolov. The solution by Galin, which is exact, calculates the boundary of the plastic surface for a plane strain condition. The solution by Sokolov, which is approximate, calculates the boundary of the plastic surface for a plane stress condition. First, the CHEXA will be used in a plane strain condition with two load cases, and 3 different meshes in each load case. Second, the CHEXA and CQUAD4 will be used in a plane stress condition with the same two load cases, with 3 meshes for load case 1 and 1 mesh for load case 2. Load case 1 will produce an elliptical plastic surface, load case 2 will produce a circular plastic surface. The version of MSC/NASTRAN' is 67.7. The material property is elastic-perfectly plastic. Reading from the OUTPUT2 file, the nodes and their von Mises stress, along angles 0, 45 and 90 degrees, from the hole to the edge of the plate, will be determined. The von Mises stress for these nodes will be plotted, such that the nodes which are plastic and the nodes which are elastic are displayed. These nodal values will be compared with the analytical solutions of Galin and Sokolov. All of the models were run on a Cray Y-MP 8I, located at Marshall Space Flight Center.
 
EVALUATION OF MSC'S PARABOLIC TETRAHEDRON FINITE ELEMENT (Acrobat 571KB) #2994
Kevin E. Arden -- Newport News Shipbuilding
 
ABSTRACT: Automatic mesh generators, based on the solid model geometry, will only produce tetrahedron elements. Solids based modeling techniques allow for efficient and easy creation of solid element models. This paper evaluates the accuracy of the element by comparing closed form solutions and actual test data to solutions calculated using MSC/NASTRAN.

The test data resulted from a recent design experience where solid modeling capabilities were beneficial. This design experience is presented along with the test validation of the finite element results.

The modeler used herein is 1-DEAS VI.l provided by SDRC.

The opinions of this paper are the author's and do not necessarily represent the opinions of Newport News Shipbuilding.

Test - Analysis Correlation

MATCHING FREQUENCY RESPONSE TEST DATA WITH MSC/NASTRAN (Acrobat 490KB) #1794
Ken Blakely -- The MSC.Software Corporation
 
ABSTRACT: This paper describes the use of MSC/NASTRAN for matching frequency response test data. MSC/NASTRAN's design optimization capability (SOL 200) is used to minimize the difference between the MSC/NASTRAN results and test data. In the procedure, the model parameters are automatically updated until the analytical results match the test data. The procedure is enabled in MSC/NASTRAN via the use of a user-written equation (DEQATN Bulk data entry) that defines the difference between test and analysis. An example is shown to illustrate the method.
 
PARAMETER ESTIMATION USING FREQUENCY RESPONSE TESTS (Acrobat 795 KB) #1894
D. N. Herting -- FE Tools
 
ABSTRACT: Structural optimization techniques in MSC/NASTRAN may be adapted to improve the correlation between finite element calculations and dynamic test results. The goal of the system is to reduce the "errors" in the finite element results by predicting changes to selected structural properties. Modern methods, which minimize weighted differences between test and analytic results over many excitation frequencies, have been adapted to the MSC/NASTRAN structural design optimizer. Response amplitudes from forced sine-sweep excitations are used as the basic inputs and actual structural properties changes are the calculated results.

This approach bypasses many of the previous difficulties by using the following methods: 1) The error measures are defined directly from the solution vectors to avoid large complicated symbolic equation entries and manually transcribed data tables, 2) Frequency response solutions are used to avoid the difficult task of calculating eigenvector derivatives, and 3) Constraint equations are built into the solution to enforce test responses and produce faster convergence. A minor amount of automated preprocessing is the necessary extra effort to use the standard V68 system. Test results show the feasibility of the approach, and perhaps its practicality. Results will be shown for a classical example problem.
 
SIMULATION OF CRASH-TEST USING THE ACCELERATION AS A PARAMETER TO CHECK THEORETICAL AND EXPERIMENTAL RESULTS (Acrobat 418KB) #2094
Luiz Carlos H. Ricardo -- Autolatina Brasil S.A.
Oscar Bito Augsto -- Universidade Sao Paulo
 
ABSTRACT: A compact car was modeled using MSC/NASTRAN. The car model is composed by 6 beam elements and 2 rigid elements, also representing the engine and body masses, besides 2 other gap elements for mathematical reasons.

The hyphothesis was that a vehicle would make a crash-test at 14 m/s, with accelerometers placed in the tunnel. We have obtained, with this test, the curve acceleration versus time. These data were used for comparison with the MSC/NASTRAN model. Solution 129, transient non-linear response from version 67, was used.

Looking at the graphs and comparing the analysis, the MSC/NASTRAN versus experimental results are very close.

VALIDATION OF 3650 DWT SEMI CONTAINER SHIP FINITE ELEMENT MODEL BY FULL SCALE MEASUREMENTS (Acrobat 757KB) #1994
Ricky L. Tawekal -- PT. PAL - INDONESIA (PERSERO)
N.P. Miharjana -- PT. PAL - INDONESIA (PERSERO)

ABSTRACT: A vibration analyses of 3650 DWT Semi Container Ship using MSC/NASTRAN is presented in this paper. Measurements of the full scale ship are used to verify the numerical predictions by MSC/NASTRAN. Two finite element models are made to study the optimum size of the finite element model with adequate accuracy. The results show a good agreement between measurements and the corresponding numerical predictions where the differences of the lowest superstructure global natural frequency of the two models are less than 5%. The finite element models are then used to predict the effects of structural modification which was done to improve the vibration behaviour.