lab. manual me670/441 - department of mechanical

Copyright 2006 by Old Dominion University

Lab. Manual

For

ME670/441

Computer-Aided Design of

Mechanical Systems

Dr. Gene Hou

Department of Mechanical Engineering Old Dominion University

Copyright 2006 by Old Dominion University 1.1

I. INTRODUCTION TO UNIX SYSTEM

I.1 INTRODUCTION In this course, you will be introduced to a new operating system the UNIX operating system. UNIX is the most popularly used operating system and is being widely used in the industry. I.1.1. Logging in All the terminals display a login prompt when no one is logged on to the machine. To login, at the prompt: login: Enter your login-ID here. password: Enter your password here (you cannot see it on the screen). Once you log on, the default UNIX prompt is as shown below. acadia:/home/rxtg001] where acadia The machine or the terminal name. home Indicates the present working directory (home directory when you initially log in) rxtg001 Your user id.(username) After you login, on the IBM terminals you can work on the X-windows environment by just typing X(Upper-Case) at the UNIX prompt. X-windowing environment (discussed in the next section) gives you multiple windows to work on. I.1.2. Change Password Just type passwd at the Unix prompt and you will be prompted as Old Password: Enter your old Password here. New Password: Enter your desired Password here (you cannot see it). Retype new Password: Re-enter your new password here (for confirmation). The computer will give you a message indicating that your password has been changed. I.1.3. Logging out: After exiting from the X-window environment, type logout at the Unix prompt. I.2 OVERVIEW OF THE CAE-LAB RESOURCES. I.2.1. Computer Hardware: 20 Sun Blade workstations are housed in Kaufman Room 124 Network: lions2.odu.edu. (Engineering Network) Address: sol-login.lions2.odu.edu I.2.2. Printer: a)one HP laser jets.


Name: Yoda (This printer is the default printer and any print command issued would be routed to this printer unless alternate printer name is specified). I.2.3.Softwares: NASTRAN (version 2004.0.0). PATRAN (2004). Pro/Engineering WILDFIRE 2.0 Pine (version 3.91) Mail software. I.2.4. Stationary: The stationary (paper) for the printer can be obtained from the Dept. of Mechanical Engineering. The paper size is A4 (8.5in x 11in). The relevant manuals and references related to the above packages are available on the Lab shelves and can be used for reference. I.3 X-WINDOWS ENVIRONMENT Once in the X-windows environment, you will have multiple windows to work with. Each window acts like a separate UNIX terminal. The windows can be resized, moved, cancelled, or deleted. The mouse left button plays a very important role in the above operations. In general, the left mouse button is functionally more active and is used to select or pick an entity on the window itself. Each of the mouse buttons is functional and holding them down against the background of the screen can invoke a set of menus. A short description of them is given below. I.4 BASIC UNIX COMMANDS Most of the UNIX commands are similar to the DOS (PC's) commands (cmd). The commonly used commands are listed below with a description of the command. Note that the UNIX commands are case-sensitive and also a space is necessary between the command and the argument that follows the command. Some of the commands are also shown with relevant examples. Note that all the commands listed below are executed at the UNIX prompt. I.4.1. User Commands finger <id> Display finger info about <id> who Who is logged on


rlogin machine-name Remote logging onto another machine. mail Read mail mail <id> Send mail Examples a) finger [email protected] Finger the user rxtg001. b) mail [email protected] Send mail to user with id rxtg001. c) ssh [email protected] remote logging in to the Unix System d) PUTTY software: - you can download this software from university software resources and use it instead of X-Windows software. Putty link: http://www.chiark.greenend.org.uk/~sgtatham/putty/download.html. Enter Host name as: acadia.engr.odu.edu or beercreek.engr.odu.edu Check the options shown in the snapshot and click on “Open”.

It will take you to the following window:


Login as your Unix user ID and enter the password (you won’t be able to see the password). Hit enter. That will take you to your home directory. You can now start using Unix account remotely. You can open any number of PUTTY windows. They work independent of each other I.4.2. Process Commands ps -au List all user processes kill -9 <pid> Kill process with process-id pid (use ps to get pid) C-c Interrupt a cmd C-s Stop screen display (locks keyboard) C-q Resume screen display where C- indicates that you hold down the CONTROL key while pressing the appropriate letter. I.4.3. Help man cmd Show on-line manual page for cmd whereis Locate directory that file is in which cmd Display path name of cmd Examples


man f77 Shows the manual entry for the f77 compiler, and gives the necessary information about the command ,its usage and available options.

I.4.4. File Utilities cp filel file2 Copy file l to file 2 (overwrites file 2) rm file Remove a file mv file1 file2 Move file1 to file2 (rename) Note that the above three commands can also be interactively executed by using the -i option. cp -i filel file2 Copies file1 to file2 interactively. diff file1 file2 Display line-by-line differences grep string file Search file for string pattern chmod permissions file Change file permissions: the options provided are: u(self), g(group), or o(others). +r(read) is to add read, -w(write) is to remove write, =x(execute) is to assign the execute explicitly. Examples cp temp1 temp2 Copies the file contents of temp1 to file temp2. If the file temp2 exists then it copies or overwrites over the existing file rm -i temp1 Interactively removes file, i.e. you are prompted to make sure if you really want to remove the file. If you enter y (yes) then the file is deleted else if you enter n (no) then the file is not deleted. I.4.5. Directories ls Lists the contents of your account in short format. ls -l List contents of directory list in long format (show file Permissions, date, groups, size, etc.) pwd Displays present working directory


cd dirname Change to another directory dirname cd .. Change to previous (parent) directory . cp dir1 dir2 Copy directory dirl to dir2 mkdir dirname Make a directory dirname. rmdir dirname Remove the directory dirname.( The directory must be empty for this command) rm -r dirname Remove entire directory tree. (The Directory need not

be empty and the above command removes the directory including its files.)

mv dir1 dir2 Move dirl to dir2 Examples pwd /home/rxt100g/ mv Dtemp1 Dtemp2 Moves or renames the contents of the directory

Dtemp1 to directory Dtemp2 rmdir Dtemp1 removes an empty directory Dtemp1 NOTE: The Directory can be emptied for the “rmdir” command by first emptying all the files in the directory by using rm filename. I.4.6 Printing a file to the Printer: lpr filename Send file to the default printer lpr -Pprinter Send file to printer specified lpq Display files in the default printer queue lpq -Pprinter Display files in printer queue for specified printer lprm <pid> Remove print job with process pid enscript –Pprinter -rGh -f Courier9 filename

This is a print command to convert the normal text format into a postscript format. -rGh -f are the print output options. The option -r rotates the file to print in the landscape mode, and "h" provides the header for the printout with the filename and date and -f is the font option with Courier9 being the font with 9 its corresponding size.


Examples lpr temp1 Prints the file temp1 to the default printer. lpr -Pr2d2 temp1 Prints the file temp1 to the printer named r2d2. lpq -Pr2d2 Displays the files in the r2d2 printer queue. lprm 455 Removes the job with the process id (pid = 455) from the printer queue. Note that the results of the NASTRAN results file i.e. (filename.f06) are printed using the enscript print command. Examples enscript –Pcae-yoda-rGh -f Courier9 temp.f06 Prints the NASTRAN results file in full in the

landscape mode in the Courier font of size 9. I.4.7 Editor/Applications emacs filename Start emacs editor vi filename Start vi editor f77 Compile FORTRAN program cc filename Compile C program X Start X windows p3 Start Patran(p3) cae-nastran2004 filename.bdf Run NASTRAN. acad Start AUTO CAD -R12. pine Invoke the pine mailtool I.5 EMACS- EDITOR I.5.1. Conventions C—Character indicates that you hold down the Control key while pressing the appropriate letter. M—Character indicates that you press and release the ESC key then press the appropriate letter. <CR>- Character indicates that you hit the return key after the corresponding command. I.5.2. Overview


When the Emacs Editor is invoked an editor screen with the file-contents appears. If the file being opened is new then a blank screen is the consequence that can be used to enter information. The four Menu-Bars are Buffer, File, Edit, Help. A file is also referred to as a buffer. Multiple files can be opened on the same screen using the open file option in the FILE menu. One can switch the buffers by selecting the corresponding buffer from the Buffer menu. One or more screens or frames can be opened for the same or a different buffer using the New Frame option in the File menu. Picking the Save File option in the File menu can also save files and Buffers. Similarly unwanted buffers and frames can be eliminated by selecting Kill Buffer (Frame) from the File menu. Further the file can be edited, modified and saved under a different name using the Save as.. option from the File menu, thus maintaining the contents of the original file. The file or the buffer can be edited selectively from the Edit menu bar and by following the following Editor commands. I.5.3. Access Emacs Editor is invoked by typing emacs filename at the UNIX Command prompt and is user friendly with a set of Menu-Bars to provide interactive access to the Editor functions. 1.5.4. Files C-x C-f read a file into emacs C-x C-s save a file back to disk C-x C-v replace this file with the contents of a file you want C-x C-w write buffer to specified file I.5.5. Cursor Movement In the Emacs Editor the cursor movement is controlled by the arrow keys on the right-hand side of the keyboard. The cursor movement is governed by the corresponding arrow-keys. Alternately the following commands also achieve the same effect. C-f move forward one character C-b move backward one character C-n move to next line C-p move to previous line C-a move to beginning of line C-e move to end of line M-f move forward one word


M-b move backward one word C-v move forward one screen M-v move backward one screen M-< go to beginning of file M-> go to end of file I.5.6. Killing and Deleting <DEL> delete backward one character C-d delete forward one character M-<DEL> delete backward one word M-d delete forward one word C-k kill to end of line C-y yank back last thing killed I.5.7. Exiting emacs C-x C-c exits the emacs editor I.5.8. Additional Commands a) Incremental Search M- % interactively replaces a test string M-x go to-line <CR> Goto-line: (line number) moves the cursor to the specified line number C-M-s regular expression search M escape incremental search C-s search forward (press M to exit) C-r search backward (press M to exit) The search for a character can also be done interactively using the Edit Menu on the Emacs Editor-Screen. b) Error Recovery C-g abort command M-x recover a file "filename" lost by a system crash C-x C-u undo an unwanted change


c) Marking Marking of a particular region in a file can be accomplished with the following commands. Marking is useful to edit a region of file as a whole. Alternately marking of a particular region can also be accomplished by holding the left Mouse button and dragging it over the region to be edited. C-@ or C-<SPC> set mark here C-x C-x exchange point and mark M-h mark paragraph d) Getting Help Help for a particular command or functions can be attained from the Help menu bar on the Emacs-Editor screen. The following are some useful commands related to the help window. C-x 1 get rid of help window M-C-v scroll help window NOTE: The description above is not inclusive and provides a brief description of the commonly used Editor commands. The reader can refer to the manual pages on Emacs for more information or type man emacs on the Unix prompt to refer to the manual page entry for Emacs. I.6 APPLICATIONS: I.6.1. Compiling a FORTRAN program: All the FORTRAN source files should have a .f extension. A Typical example of a FORTRAN file would be temp.f To compile a FORTRAN program, at the Unix prompt: f77 filename.f By default, an executable file named a.out will be created. To execute this file, just type a.out at the Unix prompt. If you with to redirect the executable file into a different file, then at the prompt at the time of compiling f77 -o executable_filename filename.f This will create an executable file (executable_filename) and to executethis file, just key in the filename. For example: f77 -o temp temp.f This will create a executable file temp, and the results of the compiled FORTRAN can be obtained by typing


temp(executable filename) at the prompt. I.7 LABORATORY ASSIGNMENT I.7.1. CLASSWORK In the classwork assignment today, you will now learn how to copy files, compile and run them. At the UNIX prompt type: cp ~sbawab/beam.f beam.f (Remember there is a space between the commands and the subsequent arguments) This will copy the file beam.f from Dr.Bawab's account to your account.


Problem Statement: The file you have copied is program written in program that computes the maximum deflection of a beam which is simply supported at three locations and which has a continuous loading. The approach used was superposition. The beam model is shown in Fig (1.1).

20"15"y

x

.5".25"

q = 40 lb/in E = 3 x 10 psi7

a b

(') (")+

RF

R R R1 2 11 22

x

y

q

Fig. 1.1 Beam Under Distributed Load

The intermediate support was treated as an intermediate force. With the beam broken up into two problems, as shown below, it was possible to use formulae from the deflection tables in a design or solid mechanics text. The intermediate force F was found by setting the deflection equal to zero at the location of the intermediate support, X = 15" and solving for F. The equations used were as follows:

y = y'+y"

y'ab =Fbx6EIl

x2 + b2 − l2( )

y'bc =Fa(l − x)

6EIlx 2 + a2 − 2lx( )

y"ac = − qx24EI

2lx 2 − x 3 − l3( )

y'AB +y"AC x= a= 0

In order to compile and run this example, at the UNIX prompt type: f77 beam.f (This will compile the FORTRAN program and create an executable file, beam)


In order to see the results and execute this program at the UNIX prompt type: ./a.out beamout This will execute the program. Then you will be prompted with a few questions ENTER LENGTHS L1, L2 AND L3 (inches) 35, 15, 20 ENTER BASE AND HEIGHT OF BEAM CROSS SECTION (inches) 0.5, 0.25 ENTER ELASTIC MODULUS (psi) AND LOADING (lb/inch) 3e7, 40.0 HOW MANY TEST POINTS? USE LESS THAN 201 200 DO U WANT THESE RESULTS TO BE WRITTEN INTO A FILE (1=YES, 2=NO) 1 (for yes) This will create a file named beamout, which contains the results. You can look at this file by using the Emacs Editor. At the UNIX prompt type: emacs beamout PROGRAM EXPLANATION Predefined functions DEF1, DEF2 and DEF3 were used in this program and correspond to the deflections yab, ybc and yac respectively. The only difference is that DEF1 and DEF2 do not include F because it has not yet been computed. Note also the negative sign in yac. The direction of the y-axis is

reversed in some textbooks. Because both y and F were reversed with respect to the reference used for this problem yab and ybc were unaffected.

I.7.2. HOMEWORK Modify the program from the above example by replacing the left boundary condition with a fixed end and name the file as beam_mod.f. Turn in a printout of your program and your output file.


II NASTRAN STATIC ANALYSIS

II.1 OBJECTIVE The objective of this lab session is to familiarize the user with NASTRAN Finite Element Analysis (FEA) program. This session will deal primarily with static analysis. The concepts and commands section in this session will contain most of the basic information necessary for the use of NASTRAN. Students may refer to a free website for NASTRAN tutorial materials: http://www.mscsoftware.com/support/university/olduniversity_tutorials.cfm II.2 INTRODUCTION NASTRAN or NAsa STRuctural ANalysis was developed initially in 1966 under the sponsorship of NASA based on known requirements of the aerospace industry for structural analysis. The version used here, MSC/NASTRAN, has been enhanced and maintained by the MacNeal-Schwendler Corporation which has been involved with NASTRAN since its inception. MSC/NASTRAN is a general purpose FEA program capable of solving a wide variety of engineering problems. The program is mainly written in FORTRAN and currently contains over 700,000 program statements. Some of the solving capabilities of MSC/NASTRAN include: Linear static analysis Static analysis with geometric and material nonlinearities Transient analysis with geometric and material nonlinearities Vibrational and buckling analysis Direct and modal complex eigenvalue analysis Direct and modal frequency analysis and random response Direct and modal transient analysis - including response spectral analysis Linear static and vibration analysis with cyclic symmetry Linear and nonlinear steady-state heat transfer Transient heat transfer Aeroelasticity Multilevel superelements Design sensitivity Optimization/resizing


II.3 RUNNING NASTRAN NASTRAN always uses an extension of a filename to differentiate between files containing data and commands. For example,

filename.bdf The NASTRAN input file. filename.f06 The NASTRAN output results file in text-format. filename.res The NASTRAN results file in binary format. filename.op2 the NASTRAN results file in binary format(read by p3 during Translation).

To execute NASTRAN, you can basically assign any filename. However, as a default you have to put a period sign (.) between the filename and its extension; e.g. temp.bdf Actually the extension (bdf in this case) can also be anything (like dat). For example the input file for xnastran is of the form *.dat where * stands for a filename. Since later on we will be dealing with lots of different files with each of them having a different meaning, it is advisable to use the extension, '.bdf', for the NASTRAN input file. This will be called the bdf file. Once the problem to be analysed is defined in the input temp.bdf file the job is submitted to NASTRAN for analysis. At the UNIX prompt, type either

NASTRAN temp.bdf or

xnastran This will cause a dialogue box to open displaying various options, and then the file temp.bdf can be selected and submitted for analysis. Wait until NASTRAN finishes processing your job. Once done, the screen will indicate MSC/NASTRAN job "temp" complete. At this moment NASTRAN should have produced three new files; temp.f06, temp.f04, temp.log. Among those, the most important one is temp.f06 (or the f06 file) which lists the results of the Edit this file using vi or emacs editor and directly go to the very end of the file where the following should be written:

*** END OF JOB*** Now take a look to the lines just above this statement, if there is no mention of the word ERROR then your file temp.bdf has been successfully analyzed by NASTRAN and temp.f06 file contains all the outputs desired by the user such as the displacements and stresses. This file can be later used in viewing the results (called post-processing) using PATRAN.


If there is any error, then something is wrong with your input file (or the bdf file). Edit your bdf file to correct the mistakes and re-run it. The .f06 file often points out the source of error and must be corrected.


II.4 CONCEPTS AND COMMANDS The approach in FEA is to let the distributed properties of a structure be represented by a finite number of simple substructures or elements. These elements are interconnected by grid points which can have loads and constraints applied to them. The elements contain physical properties such as stiffness, inertia, conductivity, etc. NASTRAN Input Data File The NASTRAN input file (the .bdf file) is made up of five sections, two of which are optional. The last three sections are bounded by the delimiters CEND, BEGIN BULK and END DATA. The sequence and the input file structure is as follows:

NASTRAN Definitions (Optional) ...

File Management Statements (Optional ...

Executive Control Statements ...

CEND Case Control Commands ...

BEGIN BULK Bulk Data Entries ...

ENDDATA CEND in the above file structure designates the end of the Executive Control Section while BEGINBULK designates the beginning of the data entries. The final line in your Input file must read ENDDATA. The NASTRAN definition is only used for unusual actions. The File Management section provides for database initialization and management, along with job identification and restart conditions. One will use NASTRAN file management to create an output file to be used by PATRAN for graphic display of results, or post-processing. Executive Control This section is used to select the type of solution to be run. It also controls other general job functions such as maximum job time and analysis alterations to a DMAP. A DMAP (Direct Matrix Abstraction Programming) is basically used to input into matrices, manipulate and output data contained in the NASTRAN data file for the proper type of solution. Some executive control statements are given below.


ALTER: Used to designate the addition or deletion or both of statements in an existing SUBDMAP.

CEND: Used to designate the end of the Executive Control section. This statement is required unless an ENDJOB FMS statement appears in the Executive Control Section.

COMPILE: Used to specify which NDDL, or SUBDMAP, to compile. ECHO: Used to control the executive section echo. ENDALTER: Used to designate the end of an ALTER. ID: Used to designate a job-ID for a specific run. SOL: Specifies the DMAP main program to be executed in this run. TIME: Sets the maximum execution time of a job. Case Control The Case Control Section has several basic functions; specifically it: • selects the loads and constraints. • requests printing, plotting, and/or punching of Input and Output data( Plotter commands are

described in section 4 of the MSC/NASTRAN User's Manual ) • defines the subcase structure for the analysis. The whole purpose of the case control section is to allow several subcases of loadings and boundary conditions for the same model. Subcases are begun by the command SUBCASE. All items placed ahead of the first subcase will be used for all following subcases unless overridden in the individual subcase. For static analysis it is possible to combine subcases. This is useful for studying different loading combinations. Some of the Case Control commands we will use are listed below: SUBCASE n Example: SUBCASE 2 Description: Delimits and identifies a subcase. LOAD = n Example: LOAD = 300 Description: Selects the external static load set to be applied to the structural model where n corresponds to at least one load entry in the Bulk Data section. SPC = n Example: SPC = 400 Description: Selects the single-point constraint set to be applied to the structural model. Parameter n refers to a constraint entry in the Bulk Data section. A single-point constraint applies a fixed value to a


translational or rotational component at a geometric grid point or to a scalar point. These are usually used to specify boundary conditions by eliminating unwanted degrees of freedom. Also, certain elements may not offer resistance to motion in certain directions, causing the stiffness matrix to be singular. SPC's can be used to eliminate these degrees of freedom. Another way to do this is by entering PARAM, AUTOSPC, YES in the Bulk Data section. This will eliminate all singularities found by the Grid Point Singularity Processor (GPSP). However, this option must be used with caution as it may cover up true design flaws. MPC = n Example: MPC = 500 Description: Selects the multipoint constraint set to be applied to the structural model. A MPC is described by a single equation, which gives a linear relationship between two or more degrees of freedom. Output Control TITLE = text Example: TITLE = Analysis for a Cantilever Beam Description: The text will appear on the first line of every page. If placed inside a subcase, then the title will appear for that subcase only. Example: Analysis for a cantilever beam SET n= a, b, c, THRU, s, EXCEPT, h, i, j, THRU, q Example: SET 1 = 5, 6, 7, 8, 9, 10, THRU, 55 Description: Defines lists of point numbers, element numbers, frequencies or time steps. These sets can be used with the Physical Set Output Requests. Example: SET 1 = 5, 6, 7, 8, 9, 10, THRU, 55 DISPLACEMENT = n or ALL Example: DISPLACEMENT = 20 Description: Requests output of displacements for all physical points or certain points defined by a SET. FORCE = n or ALL Example: FORCE = 10 Description: Requests output of forces for structural elements. STRESS = n or ALL Example: STRESS = 30


Description: Requests output of stresses for structural elements. Default output is Von Mises stresses. STRESS (MAXS) = n or ALL gives maximum shear stress. Notice that n in the above requests refers to the n in the SET definition. This could be used if, for example, only the output of the displacement of a few nodal points was desired instead of the entire model. Other useful output requests are: GPSTRESS: Stresses at grid points STRAIN: Strain for plate elements GPFORCE: Force balance for a set of grid points SPCFORCES: Single-point forces of constraint for a set of points PRESSURE: Hydroelastic pressure for a set of points You will be able to see in the examples that Case Control commands usually reference one or more data entries in the Bulk Data section. Bulk Data The Bulk Data section is begun by the delimeter BEGIN BULK. Bulk Data entries contain all the information necessary to define the model. This includes geometry, element connectivity, material properties, constraints, loadings, etc. The user must specify the geometric location of each grid point. This is normally done with respect to a global rectangular coordinate system. It is also possible to define local coordinate systems, which are directly or indirectly related to the global system. The six possible local coordinate systems are CORD1R, CORD2R, CORD1C, CORD2C, CORD1S and CORD2S. These refer to rectangular, cylindrical and spherical systems. The following sections explain in details the Bulk Data entries.


Grid Points Grid points are used to specify model geometry, degrees of freedom at a point, locations of constrained displacements, locations of applied loads and locations of desired output quantities. Each GRID entry refers to two coordinate systems. The user can also specify constraints for that particular GRID. The first coordinate system is used for locating the grid point. The second coordinate system is used to define displacements, degrees of freedom, constraints and output. The degrees of freedom are 1, 2, 3, 4, 5, and 6. Degrees 1, 2, and 3 correspond to translation in the x, y, and z directions respectively. Degrees 4, 5, and 6 correspond to rotation about the x, y, and z directions respectively.

Bulk Data Entry: GRID- Grid Point Description: Defines the location of a geometric grid point of the structural model, the directions of its displacement, and its permanent single-point constraints Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

GRID ID CP X Y Z CD PS SEID

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

GRID 2 3 1.0 -2.0 3.0 316 or GRID,2,,1.0,-2.0,3.0 (This is the commonly used format when you enter the values in the .bdf file.) Field Contents ID Grid point identification number (1,000,000 > Integer > 0) CP Identification number of coordinate system in which the location of the grid point is defined

(Integer 0) X,Y,Z Location of the grid point in coordinate system CP (Real) CD Identification number of coordinate system in which the displacements, degrees of freedom,

constraints, and solution vectors are defined at grid point (Integer 0 or blank*) PS Permanent single-point constraints associated with grid point (any of the digits 1-6 with no

imbedded blanks) (Integer 0 or blank*) SEID Super element identification number (Integer 0 or blank)


Single Point Constraints The function of the SPC was discussed in the Case Control section. The SPC entries in the Bulk Data section contain the data necessary to define how certain grids will be constrained. Note that the same grid can be referenced by two or more SPC entries with completely different constraining conditions. However only one SPC entry would be referenced at a time. Each subcase may use different constraining conditions, as would be the case if the purpose for having subcases were to try different boundary conditions on a model. Notice that SPC and SPC1 listed below are bulk data entries. These are separate from the SPC requests mentioned for the Case Control section. The SPC in the Case Control section has a number n, which corresponds to SID in the entries below. This allows different constraints for different subcases. Subcases will be used in the example problems. Bulk Data Entry: SPC - Single-Point Constraint Description: Defines sets of single-point constraints and enforced displacements Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC SID G C D G C D

Examples:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC 2 32 436 -2.6 5 or SPC,2,32,436,-2.6,5 (Which imposes the condition that the z-displacement of node 32 is kept as –2.65) Field Contents SID Identification number of single-point constraint set Integer > 0) G Grid or scalar point identification number (Integer > 0) C Component number of global coordinate (6 Integer 0; up to six unique digits may be placed in the field with no embedded blanks). D Value of enforced displacement for all coordinates designated by G and C (Real)


Bulk Data Entry: SPC1 - Single-Point Constraint, Alternate Form Description: Defines sets of single-point constraints Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC1 SID C G1 G2 G3 G4 G5 G6

G7 G8 G9 etc..

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC1 3 2 1 3 10 9 6 5

2 8

Alternative Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC1 SID C GID1 THRU GID2

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

SPC1 313 12456 6 THRU 32 or SPC1,313,123456,6,THRU,32 (Which states that the six degrees of freedom of nodes from 6 to 32 are completely fixed) Field Contents SID Identification number of single-point constraint set (Integer > 0) C Component number of global coordinate (any unique combination of the digits 1-6 (with no

embedded blanks) when point identification numbers are grid points; must be null if point identification numbers are scalar points)

Gi,GIDi Grid or scalar point identification numbers (Integer > 0)

Note that SPC1 is able to handle more grids per entry than SPC and is thus more generally used. SPC is needed only when enforced displacements are used. If you want to use a large number of enforced displacements, see the SPCD entry in the MSC/NASTRAN User Manual.


Multi-point Constraint Bulk Data Entry: MPC - Multipoint Constraint Description: Defines a multipoint constraint equation of the form

A ju j

j∑ = 0

where uj represents the degree of freedom at grid or the scalar point Gj. The MPC card defines a linear

constraint involving more than one degree of freedom. Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MPC SID G1 u1 A1 G2 u2 A2 blank

G3 u3 A3 -etc.-

Example 1:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MPC 3 28 3 6.0 2 4 -4.0

1 5 -2.5

In the above example, A1=6.0, A2=-4.0 and A3= -2.5 while u1 is the z-displacement at node 28, u2 is the x-rotation at node 2 and u3 is the y-rotation at node 1. The MPC card implies that

6.0u1-4.0u2-2.5u3 = 0

or MPC, 3, 28, 3, 6.0, 2, 4.0, -4.0, ,,ABC +BC, 1, 4, -2.5 Field Contents SID Set identification number (Integer > 0) Gj Identification Number for the Grid-point or the scalar Point. (Integer > 0) Cj Component numbers (Any one of the integers 1 thru 6 that represents the translational and

rotational components for grid-points; blank or zero for the scalar points.) Aj Coefficient (Real; Default = 0.0 except A1 must be Non-zero)


Remarks 1. Multipoint Constraint sets must be selected with Case Control command MPC= SID. 2. The first degree of freedom (G1, C1) in the sequence is defined to be the dependant degree of

freedom. A dependent degree of freedom assigned by one MPC entry cannot be assigned as a dependent one by another MPC entry or by rigid element.

3. The m-set degrees of freedom specified on this entry may not be specified on the other entries that define mutually exclusive sets.

Example 2: (extra, not available in the manual that Rajesh has) In this example, grid 3 is constraint to slide along the line AB.

1

3

x

y

θ

L

u

u

x

y

A

B

Fig. 2.1 MPC for Crank-Slider According to the MPC equation, the result is

u x

u y

=L cos θ( )L sin θ( ) or − ux sin θ( )+ uy cos θ( )= 0

where

ux = u1, uy = u2, sin(θ) = A1, and cos(θ) = A2.


Loads There are many different types of loading entries that can be used for the static loading case. The most commonly used entry is FORCE. This entry simply applies a force at a specified grid. Bulk Data Entry: FORCE - Static Load Description: Defines a static load a grid point by specifying a vector Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

FORCE SID G CID F N1 N2 N3 AXI

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

FORCE 2 5 2.9 0.0 1.0 0.0

or FORCE,2,5,,2.9,0.0,1.0,0.0 The above is an example of the force of 2.9 units being applied in the positive Y-direction at grid point 5. Note that the CID entry is blank when the only single reference coordinate system is used. Field Contents SID Load set identification number (Integer > 0) G Grid point identification number (Integer > 0) CID Coordinate system identification number (Integer 0, or blank; Default = 0) F Scale factor (Real) N1,N2,N3 Components of Vector measured in coordinate system defined by CID (Real; must have at least one Non zero component) AXI Indicates an axisymmetric loading (BCD: "AXI" or blank)


Bulk-Data Entry: MOMENT - Static torque ot twisting moment. Description: Load entries of the type torsion and twisting moments can be solved using the MOMENT card at the corresponding Grid point. Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MOMENT SID G CID M N1 N2 N3

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MOMENT 2 5 2.9 0.0 1.0 0.0

or MOMENT,2,5,,2.9,0.0,1.0,0.0 The above is an example of the moment of 2.9 units being applied along the Y-axis.. Field Contents SID Load set identification number (Integer > 0) G Grid point identification number (Integer > 0) CID Coordinate system identification number (Integer 0, or blank; Default = 0) M Scale factor (Real) Remarks: 1. The static force applied to the grid-point is given by

r f = F

r N

and

r m = M

r N

where r N is the vector defined in the fields 6, 7 and 8. The magnitude of

r f is equal to F times the

magnitude of r N for the FORCE load card and similarly the magnitude of

r m is equal to M times the

magnitude of r N for the MOMENT load-card.


2. In static solution sequences the load set ID must be selected by the case control command LOAD= SID.

3. A CID of zero or blank (default) references the basic coordinate system. The following is a summary of loading types and the available Bulk Data entries. Type of LOAD Bulk Data Entries

Forces applied directly to Grid points FORCE, FORCE1, FORCE2

Moments applied directly to Grid points MOMENT, MOMENT1, MOMENT2

Loads on scalar points SLOAD

Loads on line elements PLOAD1

Pressures and Tractions on Surfaces PLOAD, PLOAD2, PLOAD3, PLOAD4, PLOADX

Gravity GRAV (Plus Specification of Mass)

Centrifugal Force RFORCE (Plus Specification of Mass)

Thermal Expansion TEMP, TEMPD, TEMPP1, TEMPP3, TEMPRB (Plus Specification of Thermal Expansion Coefficients)

Enforced Extensional Deformation of Line- DEFORM elements(Except BEND)

Enforced Displacement at Grid Points SPCD, SPC, GRID

Linear Combination of Load Sets. LOAD The user must understand that a comprehensive explanation of the above loads is beyond the scope of the manual. However the user can obtain sufficient information regarding their usage in Volume I of the MSC/NASTRAN User's Manual.


Material Properties MSC/NASTRAN allows for the definition of structural and thermal properties of isotropic, anisotropic and orthotropic materials. Definitions are also available for plastic and fluid materials. The most commonly used entry for STATIC analysis is the MAT1 entry. Bulk Data Entry: MAT1 - Material Property Definition, Form 1 Description: Defines the material properties for linear, temperature-independent, isotropic materials Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MAT1 MID E G NU RHO A TREF GE

ST SC SS MCSID

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

MAT1 17 3.e+7 0.33 4.28 6.5e-6 5.37e+2 .23

20e+4 15e+4 12e+4 1003

or MAT1,17,3.e+7,,.33,4.28,6.5e-6,5.37e+2,2.23,ABC +BC,20e+4,15e+4,12e+4,1003 Field Contents MID Material identification number (integer >0) E Young’s Modulus (Real or Blank) G Shear Modulus (Real or Blank) NU Poisson’s Ratio (-1.0 < Real ≤ 0.5 or blank) RHO Mass density (Real) A Thermal Expansion Coefficient.(Real) TREF Reference temperature for the calculation of a) thermal loads, or b) a temperature- dependant thermal expansion coefficient. GE Structural element damping coefficient.


ST,SC,SS Stress Limits for tension, compression, and shear (Real). Used only to compute the margins of safety in certain elements; they have no effect on the computational procedures. NOTE: Margin of Safety = Factor of Safety - 1.

MCSID Material coordinate system identification number (used only in CURV module processing; Integer ≥ 0 or blank) Remarks: 1. The material identification number must be unique for all MAT1, MAT2, MAT3, and MAT9

entries. 2. The following rules apply when E, G, and NU are blank. • E and G may not both be blank. • If NU and E, or NU and G, are both blank, then both are set to 0.0. • If only one E, NU, or G is blank, then it will be computed from the equation:

G =

E2(1 + υ)

Filed 2 or MID is referenced by an element property card, which defines the geometric or sectional properties of a structural element.


Structural Elements There are many different types of elements available in MSC/NASTRAN. We will introduce the connection and property entries for the 1-D bar element and the 2-D quadrilateral element. These elements require further explanation, which will be given in the lecture. Bulk Data Entry: CBAR - Simple Beam Element Connection Description: Defines a simple beam element (BAR) of the structural model Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

CBAR EID PID GA GB X1,G0 X2 X3

PA PB W1A W2A W3A W1B W2B W3B

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

CBAR 2 39 7.0 3.0 13

513

or CBAR,2,39,7.0,3.0,13,,,,ABC +BC,513 (The above statement defines beam element 2 with properties defined by PBAR card number 39. The beam connects nodes 7 and 3. The orientation vector

r V , is a position vector from node 7 to node 13.

Furthermore there is no axial force, z-direction shear force and y-direction bending moment transmitted though node 3.) Field Contents EID Unique element identification number (Integer > 0). PID Identification number of a PBAR entry (Integer > 0) GA, GB Grid point identification numbers of connection points. (Integer > 0; A

≠ GB). X1, X2, X3 Components of the vector

r V , from GA, in the displacement

coordinate system at GA (Real).


G0 Alternate method to supply the orientation of the vector r V using grid

point G0. Direction of r V is from GA to G0 (Integer > 0 ).

PA, PB Pin Flags for the bar ends A and B respectively. Used to remove the connections between the grid -point and the selected degrees of freedom of the bar. The degrees of freedom are defined in the element coordinate system. (See Figure 2.2) The bar must have stiffness associated with PA and PB degrees of freedom to be released pin flags. For example if PA=4 is specified the PBAR entry must have a value for J, the torsional stiffness. (up to 5 of the unique integers 1 through 6 anywhere in the field with no

embedded blanks; Integer > 0).

W1A, W2A, W3A Components of the offset vectors r

W a and r

W b , respectively (see W1B, W2B, W3B Figure 2.2) in the displacement coordinate systems at points GA and GB, respectively. (Real or blank) Remarks: 1. Element identification numbers must be unique with respect to all other element identification

numbers. 2. The figure 2.2 defines the bar element geometry 3. The continuation may be omitted if there are no pin flags or offsets. 4. For additional information please refer to the MSC/NASTRAN User's Manual Section 1.2.1.


Y

Z

Ze

Xe

Ye

Y

Z

Grid Point A

Grid Point Bθx

θy

θ z

Plane 2

Plane 1

End A

End B

Orientation Vector V

Offset W b

Offset Wa

12

3 4

X

X

Fig. 2.2(a) Coordinate System of a Beam (CBAR) Element

xe

ye

ez

P

P

1

2

GGA

B

x

yz

Global System

Fig. 2.2(b) Local and Global Coordinate Systems


Bulk Data Entry: PBAR - Defines the property of a simple beam element (CBAR entry). Description: Defines the properties of a simple beam (CBAR), which is used to create bar elements Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

PBAR PID MID A I1 I2 J NSM

C1 C2 D1 D2 E1 E2 F1 F2

K1 K2 I12

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

PBAR 39 6 2.9 5.97

2.0 4.0

or PBAR,39,6,2.9,,5.97,,,,,2.0,4.0 (The above example requires stresses be reported at point 2.0,4.0 according to the Ye-Ze axes on the cross section at both the ends of element 39) Field Contents PID Property Identification Number (Integer > 0) MID Material Identification number. (Integer > 0) A Area of bar cross section. (Real) I1, I2, I12 Area moments of Inertia. (Real I1 ≥ 0.0 , I2 ≥ 0.0 , I1*I2 > I122 ), (I1=IZeZe,

I2=IYeYe)

J Torsional Constant. (Real) NSM Non-structural mass per unit length. (Real) K1, K2 Area factor for shear in planes 1 and 2. See Remark 3. (Real) Ci, Di, Ei, Fi Stress Recovery Coefficients where stress values at theses locations are

displayed. (Real; Default = 0.0). Remarks: 1. For structural problems, PBAR entries may only reference MAT1 material entries. 2. For heat transfer problems, PBAR entries may reference MAT4 or MAT5 material entries.


3. The transverse shear stiffness per unit length in plane 1 and plane 2 (Shown in Figure 2.2) are K1xAxG and K2xAxG, respectively where G is the shear modulus. The default values for K1 and K2 are infinite; in other words the transverse shear flexibilities are set equal to zero. K1 and K2 are ignored if I12 ≠ 0.

4. The stress recovery coefficients C1, C2, D1, D2 etc... are the Y and Z coordinates in the bar element coordinate system of a point at which the stresses are computed. The stresses are computed at both ends of the bar element.

The shear displacement yv of a beam is

yv =

VLKAG

where

1KAG

- the shear flexibility of the beam

V - is the transverse shear load in the element z direction L - beam length K - shear factor whose value depends on the geometry of the cross section, A - beam cross section area G - beam shear modulus

NASTRAN uses two values; K1 and K2 to define shear factors of a beam with a nonsymmetric cross-section. Value K1 contributes to the shear resisting transverse force in the element ye-axis direction and value K2 contributes to the shear resisting transverse force in the element ze-axis direction. The values of K1 and K2 are infinite by default; i.e. beam is infinitely rigid in shear.

The value of K

K =

1F

where

F = 1 +

3(D 22 − D1

2

2D23

t 2

t1

− 1

4D2

2

10r2

and D1 - distance from the neutral axis to the nearest surface of the flange D2 - distance from the neutral axis to extreme fiber t1 - thickness of web (or webs in box beams) t2 - width of flange

r - radius of gyration of section with respect to the neutral axis.


The torsional constant J and the values of F for regular shapes are listed in Fig. 2.3.


Cross Section Torsional Constant Value of F

r i

ro

2a

2a

2b

t

twh

b

J =

12

π r2

J =

12

π ro2 − ri

2( )

J = 2.25 a4

J = a b3 16

3− 3.36

ba

1−b4

12a 4

J =

13

2 t3b + tw3 h( )

10/9

2

6/5

6/5

~A/A Major Axisweb

~1.2A/A Minor Axisf

Fig. 2.3 Torsional and Area Factors for Shear of Regular Shapes

where

A - Beam Cross Sectional Area Af - Area of Flange Aweb - Area of Web


Bulk Data Entry: CQUAD4 - Quadrilateral Element Connection Description: Defines a quadrilateral plate element (QUAD4) of the structural model. This is an isoparametric membrane-bending element. Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

CQUAD4 EID PID G1 G2 G3 G4 THETA/MCID ZOFFS

T1 T2 T3 T4

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

CQUAD4 111 203 31 74 75 32 2.6 0.3

1.77 2.04 2.09 1.80

or CQUAD4,111,203,31,74,75,32,2.6,0.3,ABC +BC,,,1.77,2.04,2.09,1.80 Field Contents EID Unique element identification number (Integer > 0). PID Identification number of a PBAR entry (Integer > 0 or blank) G1, G2, G3, G4 Grid point identification numbers of the connection points. (integers >0, all unique) THETA Material property orientation angle in degrees. (Real; Default = 0) MCID Material coordinate system identification number. The X-axis of the material coordinate system is determined by projecting the X- axis of the MCID coordinate system (defined by the CORDij entry or zero for the basic coordinate system) onto the surface of the element. (Integer ≥ 0; if blank then THETA = 0.0 is assumed). ZOFFS Offset from the surface of the grid points to the element reference plane. T1, T2, T3, T4 Membrane thickness of element at grid points G1 through G4. (Real ≥ 0.0 or blank, not all zero.)


x

y element

materialelement

x

G

G

G

G2

3

4

1

θ

γβ

α

α

zelement

α = −−−−−β + γ

2

Fig. 2.4: Coordinate System for CQUAD Element

Remarks: 1. Element identification numbers must be unique with respect to all other element identification

numbers. 2. Grid points G1 through G4 must be ordered consecutively around the perimeter of the element. 3. All the interior angles must be less than 180 degrees. 4. The continuation or the second line in the CQUAD4 entry above is optional. If it is not supplied,

the T1 through T4 will be set equal to the value of T on the PSHELL entry. 5. The stresses are output in the element coordinate system. 6. The elements maybe offset from the connection points by means of ZOFFS. Other data, such as

material matrices and stress fiber locations, are given relative to the reference plane. A positive value of ZOFFS implies that the element reference plane is offset a distance of ZOFFS along the positive Z-axis of the element coordinate system. The use of ZOFFS will result in the incorrect resullts in buckling analysis and differential stiffness. If the ZOFFS field is used, then MID1 and MID2 fields must be specified on the PSHELL entry specified by the PID.


Bulk Data Entry: PSHELL - Shell Element Property Description: Defines the membrane, bending, transverse shear, and coupling properties of thin shell elements. Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

PSHELL PID MID1 T MID2 12I/T3 MID3 TS/T NSM

Z1 Z2 MID4

Examples:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

PSHELL 203 204 1.90 205 1.2 206 0.8 6.32

+.95 -.95

or PSHELL,203,204,1.90,205,1.2,206,0.8,6.32,ABC +BC, +0.95,-0.95 Field Contents PID Property identification number (Integer >0) MID1 Material identification number for the memrane (Integer ≥ 0 or blank) T Default value for the membrane thickness. (Real) (For a membrane that

is uniform in its thickness ) MID2 Material identification number for bending. (Integer ≥ -1 or blank) 12I/T3 Bending stiffness parameter. (Real >0.0; Default = 1.0) MID3 Material identification number for transverse shear. (Integer ≥ 0 or blank unless MID2 > 0) TS/T Transverse shear thickness divided by the membrane thickness. (Real > 0.0; Default = .833333) NSM Non structural mass per unit area. (Real). Z1, Z2 Fiber distances for the stress calculations. The positive direction id determined by the right hand rule and the order in which the grid points are listed on the connection entry. See Remark 9 for defaults. (Real or blank)


MID4 Material identification number for membrane-bending coupling. See Remark 6 and 13. (Integer > 0 or blank, must be blank unless MID1>0 and MID2 >0, may not equal MID1 or MID2). Remarks: 1. All PSHELL property must have unique identification numbers. 2. The structural mass is calculated from the density using the membrane thickness and membrane

material properties. 3. The results of leaving the MID field blank (or MID2 = -1) are : MID1 No membrane or coupling stiffness. MID2 No bending, coupling, or transverse shear stiffness. MID3 No transverse shear flexibility. MID4 No membrane coupling unless ZOFFS is specified on the connection entry.

See remark 6 . NOTE: MID1 and MID2 must be specified if the ZOFFS field is also specified on the connection

entry. 4. The continuation entry is not required. 5. The structural damping (for dynamic solution sequences) uses the values defined for the MID1

material. 6. The MID4 field should be left blank if the material properties are symmetric with respect to the

middle surface of the shell. If the element centerline is offset from the plane of the grid points but the material properties are symmetric, the preferred method for modeling the offset is by use of the ZOFFS field on the connection entry. Although MID4 may be used for this purpose, it may produce ill conditioned stiffness matrices(negative terms on the factor diagonal) if done incorrectly.

See the MSC/NASTRAN User's Manual Section 1.3.5 for further information. 7. For structural problems the PSHELL entries may reference MAT1, MAT2, or MAT8 material

property entries. 8. For heat transfer problems, the PSHELL entries may reference the MAT4 or MAT5 material

property entries. 9. The default for Z1 is -T/2 and for Z2 is +T/2. T is the local plate thickness defined by T on this

entry or by membrane thicknesses at the connected grid points, if they are input on the connection entries.

10. For plane strain analysis, Set MID2 = -1 and set MID1 to reference a MAT1 entry.


II.5 EXAMPLES: In the following examples, we will create the entire input file including the Executive Control section, the Case Control section and the Bulk Data section. For more complicated geometries, modeling software, such as PATRAN, will be used to create the model. Then, using a translator, the .bdf file will be automatically generated and ready for analysis. Example 1 In this example, we will model a cantilever beam by placing several BAR elements end-on-end. We will apply different loadings to display the use of subcases. The beam to be modeled is shown below.

x

y

z

F

1 2 3 4 518" 18" 18" 18"

1 2 3 42"

4"y

z

Fig. 2.5 First Beam Example

Load on the bar: 10 lbs., 20 lbs. and 30 lbs. are applied separately at node 5. Model: To model the beam we will use four 18-inch BAR elements connected end-to-end. Open a new file, and name it beam1.bdf using the EDIT utility. Enter the following data in this file. $ EXECUTIVE CONTROL SECTION ID MEM441, BEAM1BDF SOL 24 TIME 10 CEND $CASE CONTROL SECTION TITLE=CANTILEVER BEAM USING BAR ELEMENTS DISPL=ALL STRESS=ALL SPC=100 SUBCASE 1 LOAD=100 SUBTITLE=10 LB. LOAD CASE SUBCASE 2


SUBTITLE=20 LB. LOAD CASE LOAD=200 SUBCASE 3 SUBTITLE=30 LB. LOAD CASE LOAD=300 SUBCASE 4 SUBTITLE = GRAVITATIONAL LOADING CASE LOAD=400 SUBCASE 5 SUBTITLE = PRESCRIBED DISPLACEMENT PROBLEM SPC = 200 BEGIN BULK $ BULK DATA SECTION SPC1,100,123456,1 FORCE,100,5,,10.,0.,0.,-1. FORCE,200,5,,20.,0.,0.,-1. FORCE,300,5,,30.,0.,0.,-1. GRAV, 14, , 386.14, 0., 0., -1.0 LOAD, 400, 1.0, 1.0, 14, 1.0, 300 SPC1,200,246,2,3,4,5 SPC1,200,123456,1 SPC, 200, 5, 3, -0.04 GRID,1,,0.,0.,0. GRID,2,,18.,0.,0. GRID,3,,36.,0.,0. GRID,4,,54.,0.,0. GRID,5,,72.,0.,0. CBAR,1,100,1,2,1.,0.,1. CBAR,2,100,2,3,1.,0.,1. CBAR,3,100,3,4,1.,0.,1. CBAR,4,100,4,5,1.,0.,1. PBAR,100,10,8.,10.667,2.667,13.334,,,123 +23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456 +56,.8333,.8333 MAT1,10,3.+7,,0.33 ENDDATA


Explanation Executive Control Section: The DMAP, or solution sequence, used is SOL24 for STATICS. TIME 10 implies that the job is set to terminate if the solution is not reached in 10 CPU minutes. This job should take less than one minute. Case Control Section: Outputs requested for all subcases are the displacements at all the grids and the stresses in all the elements. This explains the statements DISPL = ALL and STRESS = ALL. The same single-point constraints will be used for all subcases. Here the two SPC1 entries in the bulk data section whose ID numbers are 100 are referenced. Each subcase refers to a different FORCE entry. For example, in SUBCASE 3 of the Case Control Section, LOAD=300 references the FORCE,300 entry in the Bulk Data Section. Unlike the first three subcases, the fourth subcase refers to the LOAD entry in the Bulk Data Section. Thus, the LOAD = n, entry in the Case Control Section can refer to any static loading condition defined in the Bulk Data Section. In the last subcase, an enforced displacement solution is sought. In this subcase, the SPC card 200 is selected from the Bulk Data Section. The previous SPC number (= 100) will be ignored. Bulk Data Section: GRID: The second field of each GRID entry gives the grid ID. The third field, which is empty, gives a second coordinate system ID. The last three fields give the x, y and z global coordinates of the grid. FORCE: The second field of each FORCE entry gives a unique ID number, one for each subcase. The third field gives the grid to which the force is applied. The fourth field is a coordinate system ID. Here, the default system which is the global coordinate system is used. The fifth field is the force magnitude. The last three fields give the components of the force vector. GRAV: Used to define the gravitational forces applied to the structure. The second field of the GRAV entry gives a unique ID to the entry. This GRAV entry, with an ID = 14, defines a gravitational acceleration constant of 386.14 units being applied in the negative z direction.


LOAD: Used to combine two or more load case entries. It must be used to apply the gravitational force. The LOAD card here is combining the gravitational load and the FORCE entry 300. The multiplication factors for both the gravitational force and the force entry are 1.0. This means that the forces defined by the FORCE entry 300 and the GRAV card are multiplied by 1 and then added together. The load equation is of the form

p = S Σ (SiPLi)

where S = 1.0, S1 = 1.0, PL1 = 14, S2 = 1.0, L2 = 300.

SPC1: The second field of the SPC1 entry is the ID referred to in the case control section. The third field gives the degrees of freedom that are to be constrained. The fields, which follow, give the grid numbers to which the SPC applies. Notice that grid one is constrained in all six degrees of freedom making the end fixed. These are directions in which no displacement or rotation takes place. All of the other grids have not been constrained. SPC: Used to specify displacements on degrees of freedom. This card here is different from SPC1, which specifies only zero displacements. In this example, a displacement of 0.04 units is specified on node 5 in the z-direction. CBAR: The second field of the CBAR entry is the ID for the BAR. The third field refers to the ID number of a PBAR entry. The fourth and fifth fields give the grid-points that are the two ends of the BAR. The last three entries are the components of a vector, which defines a local coordinate system for each bar. This coordinate system is needed to define moments of inertia, shear factors and points on the cross section for stresses to be output. The figure below shows how the local coordinate system is set up. Grid points A and B are shown at offset positions and with local coordinate systems defined. The local coordinate system for the BAR element is shown by xe, ye and ze. The CBAR element geometry is

defined as follows:

(1) Element X-axis: Always defined positive from GA to GB as given on the CBAR entry. (2) Element X-Y plane: Determined by V which lies in the xe-ye plane. Plane1 is always the xe-ye

plane. (3) Element Z-axis: Cross-product xe x V. Plane 2 is always the xe-ze plane.


The vector V is defined with respect to the global coordinate system. In our example, the last three fields of the CBAR entries define the element coordinate system as shown below. Notice that field three of the CBAR entries refers to the PBAR entry. PBAR: The PBAR entry defines all the geometric properties for the BAR element. Field three references the material property entry MAT1. Field four is the cross-sectional area. Fields five and six give the area moments of inertia for bending in plane one and plane two respectively. In the example, these planes are defined by the element coordinate system. Notice that the PBAR entry is continued for two more lines. The first continued line gives the y and z coordinates of points on the BAR cross-section where stress outputs are desired. These points are given as C, D, E and F. Please refer to the PBAR entry description. The next continued line gives the area factors for shear in planes one and two. The continuation entries are in field one of the continued lines. These fields contain the symbol + followed by the same seven characters found in field 10 of the continuing line. (Note: Each field can contain only eight characters. The first character in field ten of the line to be continued is ignored). MAT1: In the example, a Young's modulus of 3 x 107 psi and a Poisson's ratio of .33 are used. Shear modulus is left out and will automatically be calculated when the program is run as explained in the MAT1 entry earlier. After entering all of the data, check to make sure that all real number data fields have decimal points. Degrees of freedom and ID numbers are integers and do not need decimals. Execution: To execute your file in NASTRAN, exit your file and enter the following at the unix prompt:

cae-nastran2004 beam1.bdf scr = yes, news = no The computer will indicate that a batch job has started, and a temporary name will be given to it. Since the job is run in batch mode you can logoff while the job runs. This is convenient for large files. You will be prompted on the completion of the job on the screen and then you can proceed to read and interpret the results file. You will notice that many new files will be automatically created. After the job is executed you may delete all of these files except your original file (the .bdf file) and the output file, which has the extension .f06. To examine the NASTRAN output, type


emacs beam1.f06 OUTPUT: The first piece of output information is epsilon and external work. The quantity external work is the work done by the external load. External work terms greater than 500 could indicate incompatible stiffnesses in adjacent regions of a model. Epsilon is the work done by the unbalanced forces. When this term is greater than 10-6 it could mean that elements are poorly modeled. Bar output: For the displacement results vector columns, T1, T2 and T3 give translation displacements in the x, y and z directions respectively. Quantities R1, R2 and R3 give rotation displacements about the x, y and z directions respectively. Notice that there is one set of output for each subcase. The first four columns of the output for stresses in bar elements give normal stresses at the four points on the bar cross-section defined in the PBAR entry. There are two normal stress values given for each BAR. These stresses are at grid A and grid B of each BAR element. For example, in this problem, SA1 and SB1 are the stresses at point C on the cross-sections at grids A and B of each element. The sixth and seventh columns give maximum and minimum stresses. These are combined bending and axial stresses. The eighth column gives the margin of safety for tension and compression. This defaults to one unless stress limits are specified on the material property entry. If margin of safety is less than zero, then failure has occurred. Margin of safety is defined as the difference between the ratio of the calculated stress and the yielding stress and 1. Example 2 In this example we will model the same cantilever beam from example one, but we will use the CQUAD4 element, which is a 2-D plate element. The elements will have different cross-sectional dimensions, but the moments of inertia will be equivalent. For this problem, open a file called beam2.bdf and enter the following data. The model is shown below.

x

y

zF

0.6"

6"y

z

18" 18" 18" 18"1 4 6 8 10

2 3 5 7 9

1 2 3 4


Fig. 2.6 Second Beam Example

$ EXECUTIVE CONTROL SECTION ID MEM441, BEAM2BDF SOL 24 TIME 10 CEND $CASE CONTROL SECTION TITLE=CANTILEVER BEAM USING QUAD4 ELEMENTS SET 1=9,10 DISPL=1 STRESS=ALL SPC=100 SUBCASE 1 SUBTITLE=10 LB. LOAD CASE LOAD=100 SUBCASE 2 SUBTITLE=20 LB. LOAD CASE LOAD=200 SUBCASE 3 SUBTITLE=30 LB. LOAD CASE LOAD=300 SUBCASE 4 SUBTITLE = PRESSURE LOAD SPC = 200 LOAD = 500 SUBCOM 5 SUBTITLE = COMBINATION LOADS SUBSEQ = 1.0, 1.0, 0.0, 0.0 $ BULK DATA SECTION BEGIN BULK GRID,1,,0.,0.,-3. GRID,2,,0.,0.,+3. GRID,3,,18.,0.,3. GRID,4,,18.,0.,-3. GRID,5,,36.,0.,3.


GRID,6,,36.,0.,-3. GRID,7,,54.,0.,3. GRID,8,,54.,0.,-3. GRID,9,,72.,0.,3. GRID,10,,72.,0.,-3. SPC1,100,123456,1,2 SPC1,100,56,3,THRU,10 SPC1,200,123456,1,2 SPC1,200,5,3,THRU,10 PLOAD2, 500, 10., 4 FORCE,100,9,,10.,0.,0.,-1. FORCE,200,9,,20.,0.,0.,-1. FORCE,300,9,,30.,0.,0.,-1. CQUAD4,1,100,1,4,3,2 CQUAD4,2,100,4,6,5,3 CQUAD4,3,100,6,8,7,5 CQUAD4,4,100,8,10,9,7 PSHELL,100,10,.6,10 MAT1,10,3.+7,,0.33 ENDDATA Explanation Executive Control Section: There are no changes here. It is similar to the one explained in Example 1 Case Control Section: Here we define SET 1 as grids 9 and 10. The next line indicates that we want displacement output for SET 1. This will cause the displacements to be output only at grids 9 and 10. In the subcase 4, a new set of LOAD, the pressure load is applied and a new set of SPC is also selected. The final subcase SUBCOM is a method to combine the loads. SUBCOM is a subcase, but gets its loads from the previously defined subcases. The coefficients are selected from the SUBSEQ statement. In the example problem, the loads defined in subcase 1 and subcase2 are combined. You might want to compare the results of the SUBCASE 3 and SUBCOM 5. Bulk Data Section: SPC1:


Notice here that for the second SPC1 entry we used the THRU option. We specified that for grids 3 through 10 we wanted to constrain degree of freedom 5. Notice that we had to constrain the Y direction. This was necessary because the QUAD4 element offers no resistance to rotation about the direction perpendicular to its surface. CQUAD4: The CQUAD4 entry refers to the element property entry PSHELL. The last four fields define which grids will be G1, G2, G3 and G4. For purposes of interpreting the output, the element coordinate system is shown below. The element x-axis is defined at the plate centroid and parallel to the line connecting G1 and G2. The element y-axis is perpendicular to the x-axis and is in the plane of the plate. PSHELL: As with the PBAR entry, this entry refers to a property entry. The only numerical data in this entry is the plate thickness of .6. The MAT 1 entry has not changed from example 1. PLOAD2: This load statement applies uniformly distributed pressure load acting normal to the plane of the element. The PLOAD2 defined in the example problem defines a load of 10 units acting normal to the plane of the CQUAD4 element number 4. For further information about the PLOAD2 command, refer to the MSC/NSATRAN user's manual to get information about the fields and its usage. Execution Run this job as you did for the first example. Once delete files not needed and look at your results in the f06 file. QUAD4 OUTPUT: The displacement output for the QUAD element is similar to that of the CBAR element (explained in the previous example) In stress output, the first column gives the distances of the surfaces from the center of the plate thickness. The next three columns give stresses as defined in the element coordinate system. The principal stresses are also printed out. The last column gives Von Mises stress. Example 3 Two new concepts are applied in the following example. The first is the use of offset for the BAR element. The second is the use of subcases for different boundary conditions instead of different


loadings. The figure shows a plate with a stiffener subjected to loading. Constraints for the first case is shown below with one end fixed and one end free.

X

Y

Z

1 2 3

654

1 2

30 lbs

20 lbs

Z

Y

e

e

.5".5"

4"

2.25"

3"

.5"

Fig. 2.7 A Stiffened Plate

The second case adds a roller support to the loaded end and a simple support to the opposite end. The nastran input file below pertains to the above problem and you can copy and execute the program. $EXECUTIVE CONTROL SECTION ID NAME, FILENAME TIME 10 SOL 24 CEND $CASE CONTROL SECTION DISP=ALL


STRESS=ALL LOAD=100 TITLE = PLATE WITH A STIFFENER SUBCASE 1 SUBTITLE = FIXED - FREE CASE SPC = 100 SUBCASE 2 SUBTITLE = SIMPLE - ROLLER CASE SPC = 200 BEGIN BULK $BULK DATA SECTION SPC1, 100, 246, 1, 2, 5 SPC1, 100, 123456, 3, 4, 6 SPC1, 200, 2346, 1, 2, 5 SPC1, 200, 12346, 3, 4, 6 FORCE, 100, 2,,20.,0.,0.,-1. FORCE, 100,2,,30.,-1.,0.,0. GRID,4,,0.,0.,0. GRID,1,,36.,0.,0. GRID,2,,36.,12.,0. GRID,3,,0.,12.,0. GRID,5,,36.,24.,0. GRID,6,,0.,24.,0. CBAR,1,100,2,3,1.,0.,1.,,+B1 +B1,,,0.,0.,-2.25,0.,0.,-2.25 CQUAD4,1,100,1,2,3,4 CQUAD4,2,100,2,5,6,3 PBAR,100,10,5.5,10.375,2.2812,12.6562,,,+P1 +P1,-2.,0. PSHELL,100,10,.5,10 MAT1,10,3.+7,,.33 ENDDATA Explanation: Case Control Section:


Notice carefully that the two force entries with ID 100 are referenced by LOAD = 100 in the Case Control section. This is done before the subcase level so these two forces are applied for all cases. In addition to constraining the degrees of freedom dictated by the boundary conditions, the degree 6 is also constrained. Direction 6 is constrained for the CQUAD4 elements because, from the theory of FEA, these elements fail to handle in-plane rotation. Directions 2 and 4 may be constrained to prevent translation in y-direction and rotation about x-direction. Bulk Data Section: On the second line of the CBAR entry, offset vectors are defined at each end of the beam. Without offset, the local coordinate system of a CBAR is defined by GA and GB on the CBAR entry. If this were the case in our example, then the neutral axis of the beam would be defined on the central plane of the plate. This would happen because the CBAR is defined using grids 2 and 3. These grids are also used in defining the CQUAD4 elements. So, the offset vectors act as rigid links between the grids and the neutral axis of the beam (ie. where you want the beam ends in the true model.). Notice that both offset vectors have the components 0,0, -2.25. Notice that the CBAR and PBAR entries have continuation fields in the tenth field of their first line and in the first field of the continued line. The fields are eight columns long. In field one of the continued card, the symbol + must be added in column one. The first letter in field ten is ignored. So, the CBAR entry could also have been given as: CBAR,1,100,2,3,1.,0.,1.,,AB1 +B1,,,0.,0.,-2.25,0.,0.,-2.25 The A in AB1 is ignored. Also remember that two continuation fields anywhere in the Bulk Data section cannot be the same. On the PSHELL entry a material ID of 10, referring to MAT1,10, was added because this plate needs stiffness for bending. (See MID2 field on PSHELL definition).


II.6 HOMEWORK Homework 2 Consider the two beams shown in Figs. 2.8 and 2.9. Prepare a common Bulk Data Deck for both the beams and distinguish them in your case control deck. Generate two sets of data. In the first set, let the length of each element be 7.5 inches and divide the beams uniformly. In the second case, let the length of each element be 2.5 inches. Find the following: 1. NASTRAN input and output files for the 7.5 inch elements 2. NASTRAN input and output files for the 2.5 inch elements 3. Disp. .vs. length graph for beam in Fig. 2.8. 4. Disp. .vs. length graph for beam in Fig. 2.9. 5. Stress .vs. length graph for beam in Fig. 2.8. 6. Stress .vs. length graph for beam in Fig. 2.9 7. Find the factor of safety of the beams with 2.5 inch elements through NASTRAN.

20"15"y

x

.5".25"

q = 40 lb/in E = 3 x 10 psi7

AB

C

Fig. 2.8 Continuous Simply Supported Beam

20"15"y

x

.5".25"

q = 40 lb/in E = 3 x 10 psi7

A BC

Fig. 2.9 Continuous Clamped - Simply Supported Beam


Homework 3 Model the plate shown below. Take advantage of the symmetry so that only a quarter of the plate is constructed. Make two plates, one using two elements and one using eight elements. In constructing the eight elements use your own judgment for their size and placement. Find stresses in all elements and displacements only along the right edge. Run each plate for two subcases of loadings, one for P=10 lb/in and one for P=30 lb/in. Find the factor of safety.

10"20"

20"

P P

E = 3x10 psi ν = .33

Thickness = 0.4"

7

Fig. 2.10 A Plate with a Circular Hole


Homework 4 Consider the two stiffened plates shown in the Fig. 2.11. One end is fixed while the other is free. The thickness of the plates as well as the beams is .5". The dimension of the plates are 38"x26". Each of the plates are divided into 4 equal CQUAD4 elements.

1. Find the factor of safety. 2. Assuming the factor of safety not to exceed 2, suggest briefly how to optimize the solution and why.

0.4"

0.6"

0.4"

0.6"

X

Y

Z

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Fig. 2.11 Angled Plates with Stiffners


III. MSC/PATRAN III.1 OBJECTIVE The objective of this manual is to introduce the users to the finite element modeling capability of MSC/PATRAN and to give hands on, end_to_end guided tour of the software. The manual shows users how to start MSC/PATRAN, create the geometry model, create the finite element mesh, define the element and material properties, create load case and the loads and boundary conditions, prepare data for the analysis, and finally postprocess the results. This manual does not replace the MSC/PATRAN User's manual (which can currently be browsed through in the online help menu), but provides sufficient information to start with. III.2 INTRODUCTION The finite element method is a proven technique for using computers to model a wide variety of engineering problems. Its application in the real world was hindered, however, by the amount of time spent both in producing the raw data to feed a Finite Element Analysis (FEA), and in interpreting the usually large volumes of results from the analysis. MSC/PATRAN is a software developed to provide a systematic approach towards making FEA modeling fast and accurate. It uses a simple step-by-step approach that helps to create, analyze and interpret a mathematically realistic model of the structure. This approach is built around geometric modeling, interactive computer graphics, and current finite element theory. III.2.1 Engineering Functionality The capabilities of MSC/PATRAN include: (i) A full set of tools for the creation of parameterized model geometry. In addition,

MSC/PATRAN has Single Geometric Model (SGM) capability. SGM accesses geometry data, topology, and evaluators from the CAD (Computer Aided Design/Drafting) system without transformation and establishes and maintains associativity with the corresponding MSC/PATRAN finite element entities throughout the entire design and analysis process.

(ii) Finite modeling tools for analysis model creation and verification, including mapped meshing,

automatic surface meshing, and automatic tetrahedral solid modeling. (iii) A complete set of functional (loads, boundary conditions, and material/element properties)

assignment capabilities, including the capability to assign these directly to the geometry or the finite element model. In application, this means that the finite element mesh can be deleted and the geometry remeshed without having to reapply the functional assignments. All functional assignments can be collected into load cases and named, modified or deleted at the user's discretion.

(iv) PATRAN Command Language (PCL) for the customization of MSC/PATRAN, the

performance of variance and design sensitivity studies, and automation of routine, repetitive


tasks. III.2.2 Graphical User Interface

The MSC/PATRAN user interface is form driven and truly intuitive, minimizing the time required to learn the system. At startup, after creating a new database or opening an existing one, the MSC/PATRAN user interface consists of a control panel, graphics window (view port), and a command/history window. The control panel controls the parameters for system tasks such as managing files, setting preferences, manipulating groups and graphics, as well as switching between interactive applications. The information displayed in the graphics window includes geometry, finite elements, applied loads, boundary conditions, analysis results and user defined annotation. The command/history window contains a record of all commands executed during a MSC/PATRAN session, as well as enabling commands to be entered from the keyboard. III.2.3 Files MSC/PATRAN modeling begins with opening a file (new or existing) using the FILE pick in the upper left corner of the main menu. File management options available under the FILE pick include: create a new database, open or close an existing database, access CAD geometry, import IGES models or PATRAN Neutral Files, record and playback session/journal files. The MSC/PATRAN relational database defaults to a ”.db” extension and contains a complete record of user created geometry, finite elements, load cases, load and boundary conditions, materials, element properties, preferences, and results. MSC/PATRAN database creation is memory expensive. Each database requires approximately 6MB of disk space. Therefore old database should be compressed using the Compact Database option under the FILE pull down menu. The record of commands shown in the command/history window during a MSC/PATRAN modeling exercise are stored in editable files that can later be run through MSC/PATRAN to perform parametric or sensitivity analyses. Two files are created by MSC/PATRAN to store these commands: ·A session file (.ses extension) which contains the record of all PCL commands for each

interactive session. ·A journal file (.jou extension) which contains a record of all PCL commands for a particular

PATRAN database.


III.2.4 Analysis Preferences MSC/PATRAN provides for the user unique analysis preferences i.e. a user can readily switch between commercial analysis solvers and proprietary in-house codes. Thus the user is not faced with having to learn the command and syntax of each particular analysis solver. Upon setting the new preference, MSC/PATRAN automatically updates common element properties, material properties, loads, boundary conditions, and analysis code forms to those of the newly selected solver. Note that in this manual, our analysis solver will be MSC/NASTRAN. III.2.5 Results Visualization MSC/PATRAN post-processing is state-of-the-art in its ability to display, sort, combine, scale, and query in a general way a single results database. After execution, analysis results are loaded directly into the MSC/PATRAN relational database and can be sorted by time step, frequency, temperature or spatial location. MSC/PATRAN post-processing enables the engineer to filter engineering results by material, element type and property, node and element IDs, thresholds of results, etc., simultaneously. Display types include but are not limited to: deformed, fringe, vector, tensor, and engineering X_Y plots. Shear and bending moment diagrams are available for beam results and a sophisticated text report writer is available to print out all results in a user defined sorting sequence and format. The insight application within MSC/PATRAN condenses mountains of raw numerical data into graphical tools and displays for complete, accurate interpretation of finite element analysis results. III.2.6 On-Line Help/Documentation The entire manual of MSC/PATRAN can be browsed through by clicking the left mouse button on the HELP button of the Top Menu Bar of the Control Panel. All interactive features, functions, and applications in MSC/PATRAN can be obtained from the completely topical and context sensitive help on-line. Hypertext links throughout the on-line system allow for instant retrieval of complex information. The MSC/PATRAN on-line help system eliminates the need for printed documentation, because the MSC/PATRAN printed documentation is identical to on-line help files. III.3 LAB. SESSION 1 III.3.1 Things to know before you start How to start MSC/PATRAN: After you are in the X-Window environment, just type 'cae-patran2004' at the UNIX prompt. Hardcopy Printout: You can get a black and white printout through the printer 'cae-yoda'. The procedures for printing are shown in Section III.3.3. Generally the user prints into a postscript file. Then the file is sent to a printer of choice by using the 'lpr' command, i.e. lpr -Pcae-yoda filename


where 'printer' is the name of the printer and 'filename ' is the postscript file generated. How to quit MSC/PATRAN: To quit MSC/PATRAN, click the left mouse button on the 'File' button of the control panel and select 'Quit' from the pulled down menu. If your want to start work on another database without quitting P3/PATRAN, you can select 'Save ' instead of 'Quit' to save the current database. Next we will go through the following: ·The menu components of MSC/PATRAN. ·How to use the three-button mouse. ·A highlight of how to modify the view of a model. ·How to print the contents of a viewport.


III.3.2 Menu Components of MSC/PATRAN MSC/PATRAN Control Panel MSC/PATRAN’s Control Panel appears at the top of the screen (example below). The selections on this panel are divided by a horizontal line. The selection above the line is MSC/PATRAN’s Top Menu Bar, which includes MSC/PATRAN’s heartbeat and the on-line help system. The items below the line are a group of radio buttons and a group of Control Panel icons.

The Menu Bar selections is a pull-down menu, which will be explained later in using the mouse (p.3.7) Radio buttons are mutually exclusive choices in that only one radio button within a group can be pressed in at one time. They work like the radio buttons in your car. Only one button on your car radio can be pressed in at one time. MSC/PATRAN’s heart tells you if it is ready to execute an option or if it is busy executing.

Green means MSC/PATRAN is ready and waiting to execute a form. Blue means it is busy, but it can be interrupted with the Hand icon. Red means it is busy and it cannot be interrupted.


MSC/PATRAN’s on-line help is an easy to use, hypertext-based documentation system. You can enter the on-line help either by pressing Help on the Menu Bar, or by placing the cursor anywhere on a MSC/PATRAN form and pressing the F1 key to get help on that particular form. If your cursor selects anything that is in magenta in the on-line help, MSC/PATRAN will quickly “jump” to another page in the on-line help that it is referencing. The Control Panel icons are explained below:

Refresh Icon Redisplays (refreshes) all of the graphics viewports

Display Cleanup Icon Removes all fringe and marker plots, all automatic titles, highlighting and deformed shape plots. Repaints the viewport in wireframe mode

Interrupt Icon Interrupts a command in progress. This is useful when you want to abort out of an executing MSC/PATRAN form

Undo Icon Undoes the last executed action of a MSC/PATRAN form when an “Apply” was pressed MSC/PATRAN Command Window MSC/PATRAN also has a Command Window that appears at the bottom of the screen (see below). You can manually enter commands in this window, but mostly this window is used to view the commands MSC/PATRAN generates when a menu form is executed, and to view errors or information messages.

The command line. Commands Can optionally be typed in here. The history window. You can use the scroll

bars at the bottom and right sides to scroll back or to scroll to the right.


MSC/PATRAN Application Form and Subordinate Form The radio buttons on the Control Panel will bring up a vertical form on the right side of the screen. These forms are called the Application forms . There are also smaller forms that are sometimes displayed which are associated with the specific Application form, and they are called Subordinate forms.


III.3.3 Using the Mouse MSC/PATRAN uses a three-button mouse. Depending on what you are doing, each of the three buttons will accomplish a specific task in MSC/PATRAN. The following is a summary of what you will need to know to complete the guided tour. Using Pull-Down or Option Menus To use a pull-down or option menu, with the left mouse button, click on menu item. While holding down the mouse button, drag the mouse to the appropriate selection and then, release the mouse button. Selecting From a Listbox A List box has a title at the top and a list of contents. To select an item, use the left mouse button and highlight the item. To deselect the item, use the left mouse button and highlight the item. To deselect the item, use the left mouse button and click on the highlighted item. Overview of the Three Buttons The left mouse button is used to select model entities. To select more than one entity, hold down the Shift key while selecting the entities with the left mouse button. The right mouse button is used to deselect an entity that was selected in error. To deselect more than one entity, hold down the Shift key while selecting the entities with the right mouse button. The center mouse button is used to change the view by rotating, translating, or zooming. Keep the button pressed down and drag the mouse to move the model. Cursor Selecting Shortcuts To cursor select model entities with a rectangular box, hold down the left mouse button at the upper left hand point and drag the cursor to the lower right hand point. Then, release the mouse button. To cursor select entities with a polygon shape box, hold down the Control Key (Ctrl) and use the left mouse button and start picking at arbitrary points A, B, C, D, … to define a closed polygon box.


III.3.4 Modifying the View There are several ways to alter the view of your model, and below are some of the more common ways. Using the Middle Mouse Button As mentioned earlier, you can rotate, translate, or zoom in on your model by dragging the mouse, while holding down the middle mouse button. To switch the mouse function to rotate, translate, or zoom, enter View/Mouse Settings… View/Select Corners… You can zoom in on a particular part of your model, and at the same time, define a new viewing center by selecting from the Menu Bar:


MSC/PATRAN will turn the cursor into a cross-hatch. Define a new “viewport” by dragging the cursor diagonally and defining a rectangular box: Drag the cursor along a diagonal direction to select a portion of the model you want to view up close. Select View/Fit View… if you want to return the view back to the entire model. View/Transformations… You can rotate or translate the view incrementally by selection from the Menu Bar. A subordinate Transformations form will appear with a matrix of buttons. By pressing any one of the buttons, you can translate, zoom, or rotate your model incrementally. For example by pressing a button, MSC/PATRAN will rotate your model about the screen or model’s Y-axis To perform a “Fit View”, where MSC/PATRAN will resize the view of the model to fit within the viewport, you would simply press the assigned button. More view/transformation buttons are shown on the next page.


Show the entire model with hidden lines Show the entire model without hidden lines Show the shaded solid view of the model Show the end_view in the Y_X plane

Show the end_view in the Z_X plane

Show the end_view in the Z_Y plane Show the isometric view Show all the labels (set labels on) Do not show all the labels (set labels off)


III.3.5 How to print the contents of the current viewport. To print a hardcopy output of the contents of a viewport, the user should click the left mouse button on the 'File' option of the top menu bar of the control panel and select 'Print' in the pulled_down menu. The following application form is shown on the screen: The user then selects ”Postscript Default” under available printers and ”Postscript” under Driver. The 'Page Setup...' option provides another form where the either landscape or portrait paper orientation can be selected. A third form is obtained by clicking on 'Options...' menu. This form, which is shown below, asks for the Format, Background, Lines & Text, etc. The form shown here has been completed for the generation of a postscript file named 'manual.ps' for black & white output.


To obtain a color postscript file, the 'Color' option is selected for 'Format', the 'White' option for 'Background' and the 'Actual' option for 'Lines and Text'. To print the generated postscript file, simply type at the UNIX prompt: lpr -Pprinter filename where 'printer' is the name of the printer and 'filename ' is the postscript file generated. Use 'cae-yoda' for black and white printout. III.3.6 Creating the Geometry Each student should go through this section as part of Lab. Session 1. To enter MSC/PATRAN, just type ”cae-patran2005” at the UNIX prompt. Wait until the system displays the control panel window and the command history window (see p. 4.4 & 4.5) for <SC/PATRAN. The first step is to create a new database file. This is done by clicking the left mouse button on ”FILE” in the top menu bar of the control panel. This will show a pull down menu from which ”CREATE A NEW DATABASE” is selected. The result is the form shown below:


Note that one should click on “Modify Preference” to check whether MSC/NASTRAN is selected as the analysis before proceeding. In the form shown above, a new database name “manual” has been entered. The user can supply his/her own name. The OK button is selected With the left mouse to create a new database named “name.db” (in this example, manual.db). After the database has been created, the following form for model preferences is automatically displayed: This form is completed for tolerance limit, approximate maximum dimension of the model to be created, analysis code of choice and analysis type. The form shown here has been completed for structural analysis with MSC/NASTRAN as the analysis code. As said earlier (see p.2), our analysis code will be MSC/NASTRAN. After the form is completed, ”OK” button is selected. The next step is to start the creation of a geometrical model. In this lab session, the reader is taken through creation of basic geometrical entities such as point, curve (1_D), surface (2_D), and solid (3_D). The creation of all geometrical entities is initiated by clicking the left mouse button on the ”GEOMETRY” radio button of the control panel (see p. 4.4). A typical form that needs to be completed is shown below (The left mouse button is used to move through the form). Note that in MSC/PATRAN, the input format of the coordinate of a point, (x, y, z) is given by [x, y, z] , which is different from that of a vector given by < x, y, z >.


The above form has been completed for the creation of a point at [x,y,z] = [0,0,0]. It should be mentioned here that there are several options under ”Action”, ”Object”, and ”Method”. The choice of combination of options for these three is based on the result desired. Further examples will be seen in the course of this manual. After a geometrical entity has been created, the figure that lists the visible entities, shown above on the right, is seen at the bottom of the screen. This figure highlights geometrical entities that can be selected by clicking of the left mouse button on the desired entity. The second point is at [5 0 0].


The figure below shows the view port after two points have been created.

The next stage is to create a curve. The form below has been completed for the creation of a 2_point curve using the two points created earlier:


Next, we illustrate the creation of a 3_point arc. The third point is at [2.5, 2.5, 0]. For this purpose the completed form is shown below: Don’t click on “Auto Execute.” Instead, click on “Apply”. It gives user a little room to correct typo errors. The result of applying the above form is the arc (curve 2) shown in the figure below:


The next stage is to create a surface. A surface can be created from 2, 3, 4, or N curves where N is a number to be supplied by the user. An example is given here of how the surface can be created from two curves. First, the 3_point arc should be deleted by selecting ”delete” under ”action” and then clicking on the 3_point arc in the viewport. Next, the following form is completed for creating a second curve by translating the first 2_point curve: In the above form, ”<0 5 0>” is entered for the ”Translation Vector”, and ”Curve 1” is entered for the ”Curve List”. The result of applying this form is shown below:


After the two curves have been created, the form below is completed for the creation of a surface:


Applying the above form, we have the result shown below: The next stage is to create a solid. Again, there are several ways of creating a solid. In this example, we are using the 2_surface option. Hence we need to create another surface. This is done by translating the previous surface 5 units of distance in the z_direction (i.e. Translation Vector: <0 0 5>). The form completed for surface translation is similar to that of curve translation with ”Surface” as the ”Object”. After surface translation, we have the result below:


The rectangular solid is created by completing the form shown below: Applying the above form, we have the result shown below. To write a note or legend on the figure, click on “Display,” then “Title” and follow the instructions to fill out the proper information. Click on “Apply”. In the following picture, the title is given as Solid_Model.


III.4 LAB SESSION 2: Plate problem In this lab session, try to create the plate shown below. The steps used in generating the plate are illustrated. You are not limited to using these steps. You can obtain the same results using any other combination of steps if you are confident of doing so.

Step 1: Enter MSC/PATRAN by typing 'cae-patran2004' at the UNIX prompt. Step 2: Create a new database named 'classwk1' (or any other name you prefer) by clicking on 'File/New Database' of the control panel. The following form is displayed. It is completed by supplying the new data base name.


Step 3: Set the model parameters by completing the model preferences form shown below. This form is automatically displayed after a new database has been created or an existing one opened. Step 4: Start geometry creation by clicking on the 'Geometry' button of the radio panel. The following form is displayed. The form is completed for point 1 (at [x y z] = [0 0 0]) as shown.


Step 5: Similar to point 1, create points 2 and 3 at [x y z] of [5 0 0] and [0 3 0] respectively. Step 6: Create point 4 by translation method. This is done by completing the form shown below. The form has been completed for creation of point 4 by translating point 3 through a vector of <5 0 0>. The Point ID List is shown to be 5 because the 'Apply' button has been clicked on and point 4 has been created. Step 7: Similar to point 4, create point 5 by translating point 2 through <0 1 0>. Step 8: Similar to point 4, create point 6 by translating point 4 through <0.5 0 0>. Step 9: Similar to point 4, create point 7 by translating point 5 through <0.5 0 0>. Step 10: Similar to point 4, create point 8 by translating point 7 through <2 0 0>. Step 11: Similar to point 4, create point 9 by translating point 6 through <2 0 0>. Step 12: Similar to point 4, create point 10 by translating point 9 through <2.5 0 0>. Step 13: Similar to point 4, create point 11 by translating point 8 through <2.5 0 0>. Step 14: Similar to point 4, create point 12 by translating point 10 through <2.5 0 0>.


Step 15: Similar to point 4, create point 13 by translating point 11 through <2.5 0 0>. Step 16: This step begins the creation of curves. Create a curve by selecting 'Curve' as 'Object' in the 'Geometry' form. The form is shown below completed and already applied (hence 2 is shown in the Curve ID List) to create curve 1. Curve 1 is created using the 2_point method and the points used are 1 and 2. Step 17: Similar to curve 1, create curves 2 through 11 by using 'Starting Point List' 3, 4, 2, 5, 6, 7, 8, 8, 11, 13 and 'Ending Point List' 4, 6, 5, 7, 9, 8, 9, 11, 13, 12 respectively. In other words, curve 11, for example, is created by joining point 13 and point 12. Step 18: Create curve 12 by using the '2D ArcAngles' option under 'Method'. A form to be completed for this is shown below. This form has been completed and applied for curve 12 (hence 13 is displayed in the 'Curve ID List'). For this curve, the following information was supplied: Radius (1.0), Start Angle (180), End Angle (225), Center Point List (Point 10).


Step 19: Similar to curve 12, create curves 13 through 15 using the same radius and center point list as curve 12 but with start angles 225, 270, 315 and end angles 270, 315, 360 degrees. Step 20: Create curve 16 by using the 'Fillet' option under 'Method'. The region of fillet is shown in the plate figure. The form displayed to be completed is shown below. This form has been completed for the fillet curve. The following information was supplied: Fillet Radius (0.5), Curve/Point 1 List (click on curve 5, then point 5), Curve/Point 2 List (click on curve 4, then point 5). After the form is completed, P3 asks if the original curves should be trimmed. A window is displayed for this (see below). The option 'Yes' or ”Yes For All' is selected. A new point 19 and a new curve 16 are created. The result is shown below:


Step 21: Trim the curve between points 19 and 5 by using “Edit” under “Action”. Select “Trim” under “method” and “point” under “option.” Select “Trim” under “method” and “point” under “option”. Select point 19 and curve 4 for “Trim Point List” and “Curve/Point List”. Click on “Apply” after the form is completed. This step involves merging of two curves 5 and 7. This is done by using the form below. The new curve formed from the merge is 17. The next two forms contain questions about the merging. The answer to the first question (Do you wish to create a duplicate curve?) is 'No'. The answer to the second question (Do you wish to delete the original curves?) is 'Yes'. Follow step 21 to trim the curve between points 5 and 7.


Step 22: Trim showed 10 Step 23: Create Surface 1 using the 2_curve option. This is done by completing the form shown below. The form has been completed and applied for surface 1 (hence it is ready for surface 2, i.e. Surface ID List = 2). The 'Starting Curve List' for surface 1 is 'Curve 16' and the 'Ending Curve List' is 'Curve 3'. Step 24: Similar to surface 1, create surfaces 2 through 7 with 'Starting Curve List' 2, 6, 12, 13, 14, 15 and 'Ending Curve List' 1, 17, 8, 9, 10, 11. For example, surface 3 is created from curves 6 and 17.


Step 25: Before we continue, we want to set the 'Display Property' such that only surface numbers will be displayed; the curve and point numbers will be hidden. This is done by clicking on the Display' button of the control panel followed by “Geometry” button. This option allows the surfaces to be displayed with the fewest number of parametric lines. Next, all label buttons in the 'Entity Types' section of the form are deactivated except the surface label button. The form is then applied.


Step 26: Finally to complete our plate geometry creation, we mirror the figure in the viewport, which is half of our geometry. This done by completing the form below. The mirror plane normal is defined as {[0 3 0][0 0 0]}. The 'Surface List' is selected to include all existing surfaces (i.e. surfaces 1 through 7). Step 27: The created geometry can then be printed by following the example shown on p. 4.11. The user is free to supply his/her own file name. Homework: Lug Problem Now as your homework, generate the geometry of the LUG shown below. It is not necessary for the patch numbers to match those given in the figure.


III.5 LAB SESSION 3: Analysis of the Plate problem In this lab session, we will analyze the plate of lab session 2. In particular we will learn how to use the following radio buttons: 1. Load/BCs: To assign boundary conditions and loads 2. Load Cases: To create and load cases 3. Finite Element: To generate finite element mesh. 4. Materials: To assign material properties. 5. Element Props : To assign element properties. 6. Analysis: Interfaces with NASTRAN via PATNAS (PATran to NAStran) to generate bulk data file and submit the file to NASTRAN for analysis automatically; and via NASPAT (NAStran to PATran) to read in and translate the analysis (OUTPUT2 format) results. 7. Results: To graphically view analysis results. To accomplish the above, we will follow the steps shown below. Load the geometry of the plate generated in lab session 2. This is done by opening the MSC/PATRAN database saved in the lab session 2 by clicking on File/Opening an existing database.... The form shown below is completed by supplying the 'Existing Database Name'; indicating the full path to the directory in which the database file is located. If P3 is entered from the directory that contains the existing database, this database is automatically displayed in 'Database List'. The database can then be selected by clicking on its name in the list.


III.5.1 Boundary Conditions The next step is to assign boundary conditions. The desired boundary condition is to fix the left edge of the plate, that is, to constrain all six degrees of freedom for the left edge. To achieve this, the 'Load/BCs' radio button is clicked on. This will bring a display of the form shown below on the left. This form is completed as indicated. The 'New Set Name' for the boundary condition is 'fixed'. The new set name is just for the user to identify the data; it is not used in NASTRAN input file. To complete the boundary condition specification, we have to 'Input Data' and 'Select Application Region'. To input data, the 'Input Data...' button is clicked on and the form above on the right is displayed. This form is completed as indicated. The 'Translations' vector is <T1 T2 T3> = <0 0 0> and the 'Rotations' vector is <R1 R2 R3> = <0 0 0>. This implies that the six degrees of freedom are constrained.


To select application region, the 'Select Application Region...' button is clicked on. This will bring a display of the form shown below. Since we are assigning the boundary conditions before the finite element mesh is generated, we select 'Geometry' under 'Geometry Filter'. Then the cursor is moved to 'Select Geometry Entities'. At this point, since we want to apply the boundary condition to the left edge of the plate, we first click on the edge icon in the 'Visible Entities' display (see page 4.14). The select menu will then state 'Select a Curve'. At this point, we cursor select the left edge of the plate by using a rectangular cursor selection. After this is done, 'Surface 2.1 9.4' (that is, edge 1 of surface 2 and edge 4 of surface 9) is shown in the 'Select Geometry Entities' window. We then click on 'Add' and 'Surface 2.1 9.4' is now shown in the 'Application Region' window. After this is completed, click on 'OK' to conclude this part.


Then press 'Apply' on the 'Load/BCs' Application Menu form to conclude the whole step. At this point, you should have the figure shown below. Notice the 'cones' at the left edge of the plate at the display lines' point location. These cones represent the three fixed translational and three fixed rotational boundary conditions. Notice that this boundary condition is applicable to any future finite element mesh generated for the plate, that is all finite element nodes on the left edge of the plate will have all six degrees of freedom constrained even though the nodes are yet to be created.

III.5.2 Applied Loads The next step is to assign loads to the plate. In this example, we will have two point (nodal) loads, one in the negative Y_direction and the other in the positive X_direction, the lines of action of both forces passing through the center of the circular hole in the plate. To achieve this we complete the same set of forms as for the boundary conditions for each of the point loads, one after the other, but with the following modifications. The 'Object' is 'Force' rather than 'Displacement'. For the first point load, the 'New Set Name' of '1000_pound_down' can be used and for the second point load, '10000_pound_right' can be used. These names are just for the user's benefit, they are not used in NASTRAN's input file. In the 'Input Data...' form, the following force vectors are provided: <F1 F2 F3> = <0 _1000 0> for the first point load and <F1 F2 F3> = <10000 0 0> for the second point load. For both loads, the moment vector is <M1 M2 M3> = <0 0 0> since no external moment is applied. In the 'Select Application Region...' form, when the cursor is in the 'Select Geometry Entities' window, the 'Point' icon is selected instead of the 'Curve' icon. For the first load, cursor-select the point at the top of the line dividing surfaces 12 and 13. For the second load, cursor-select the point at the middle of the right edge of the plate. After 'Apply' is pressed for each of the loads in the 'Application Menu' form, the figure below is the result. Notice the arrows drawn at the selected points signifying the 1000_pound force in the negative Y_direction and 10000_pound force in the positive X_direction. Each arrow is displayed as soon as the load corresponding to it is created.


Now a load case will be created that combines the two point loads and also the specified boundary conditions. This is done by clicking on the Load Cases' button of the control panel. The following form is then displayed. Initially under Èxisting Load Cases', we have the `Default' load case listed. This load case is always present in any analysis, irrespective of how many more load cases are created by the user. Then we specify the new `Load Case Name' as `lcase_1'. We make the new load case current. The user is free to give some description for the new load case under `Description'. This is optional. The next step is to go to the list of Àssigned Load/BCs Sets' and then select the combinations of loads and boundary conditions that will constitute the load case being created. For our example, we select all the loads and boundary conditions in this list to constitute our load case. This means that our load case is made up of `Displ_fixed_bc', `Force_10000_ibf' and `Force_1000_ibf'. After the form is completed, we press on Àpply' to conclude the creation of the load case. By now, we should see under Èxisting Load Cases' our newly created load case i.e. `lcase_1'.


III.5.3 Mesh Generation The next step is to create the finite element mesh for our plate. This step is divided into two; the first is to create mesh seeds for the curves bounding our geometry. Mesh seeds allow us to define exactly how many elements we want on a selected curve or edge of a surface or solid. The number of mesh seeds on a curve is equal to the number of elements on that curve. An example is given here of how to create 10 mesh seeds on curve 1 of our geometry. The radio button 'Elements' is clicked on and the form below on the left is displayed. The form is completed as indicated. Mesh seeds are created for other curves one after the other such that surfaces 2 and 9 have 5x10 elements each, surfaces 1 and 8 have 3x4 elements each, surfaces 3 to 7 and 10 to 14 have 4x4 elements each. After the mesh seeds have been created, the next thing is to create mesh for our geometry. This is achieved by selecting mesh as the 'object' in the form shown in the middle below. The 'type' of element is chosen as surface since we are dealing with a 2_D problem. The 'element topology' selected is 'Quad4', i.e., quadrilateral elements with four nodes. We select 'Isomesh' as our 'Mesher'. All these selections are then applied to all the surfaces 1 to 14 as indicated in the 'Surface List'. To select all surfaces, we can use the window cursor after we have clicked on 'Surface' icon in the 'Visible Entities' display. After the apply button is pressed, the entire geometry is subdivided into 'Quad4' elements.


Before concluding the process of mesh generation, some modifications are made to the nodes at the interface of surfaces 8 and 9 and also the interface of surfaces 1 and 2. The modification involves making, at the interface, the nodes for surface 9 coincident with the corresponding nodes for surface 8 and the nodes for surface 2 coincident with the corresponding nodes for surface 1. This is achieved by moving the nodes to their new positions. An example is shown in the form above on the right where 'Node 75' is moved to coincide with 'Node 4'.


After all the modifications are completed, the plate with the generated element mesh looks as shown below. Note the skewness of the elements at the interfaces mentioned above in surfaces 2 and 9 where some nodes have been moved to coincide with the nodes in surfaces 1 and 8. The process of mesh generation is completed by doing the 'Equivalencing' and 'Optimization'. Equivalencing forces nodes that are within some specified tolerance limit to coincide. The tolerance used here is 0.007. This means that any two or more nodes within 0.007 unit of distance from one another will be forced to coincide at the position of the node with the lowest node identification number within the group. Examples of nodes affected by ”Equivalencing” are the nodes along the symmetry line and along lines common to adjacent surfaces. The form below is completed for 'Equivalence'. Optimization is done to minimize the memory requirement of the global stiffness matrix by rearranging the elements of the matrix such that the storage bandwidth is reduced. The form below has


been completed to 'Optimize'. The table below the form is a result of the optimization when the 'Apply' button is pressed. Notice the drastic reduction in the bandwidth after optimization! Pressing 'OK' will make the table disappear.

III.5.4 Material Selection In this section, we will create the structural material that constitutes our elements. We start by clicking on the 'Materials ' button on the control panel. The form shown below on the left side is displayed. We complete the form as shown. Our material is isotropic and the properties will be input manually. We supply the material name under 'Material Name'. After this we click on the 'Input Properties' button and we have the form shown on the right side. In this second form, we select the constitutive model, which in this case is 'Linear Elastic'. We also fill out the property sheet by supplying the shown values for 'Elastic Modulus', 'Poisson Ratio', and 'Density'. We then press 'Apply' to make the information active. We then press 'Cancel' to complete the process. We now see in the left side form under 'Existing Materials' the name 'Steel'.


III.5.5 Element Properties This section deals with specification of properties of our elements. This is started by clicking on the 'Element Props ' button on the control panel. This gives a display of the form shown on the left below. This form is completed as indicated. The elements are 2_dimensional in geometry. Hence the element type chosen is 'shell'. The name `plate_shell' is given to the property set being created. Under 'Option(s)’ the material is chosen to be `Homogeneous'. The next step is to input the element properties. This is done by clicking the left mouse button on Ìnput Properties...'. The form shown on the right below is then displayed. The form is completed as shown. The material name is 'steel' where steel is the material created in step 5. Notice that 'm:' precedes the material name. This indicates that the elements whose properties are being created are made of material `steel'. The element thickness is specified to be `0.125'. For the problem we want to solve, the information supplied so far is sufficient. Thus the element properties creation is concluded by pressing ÒK' on this form and then Àpply' in the left form.


“Select Members” box in “Application Region”. (Note that you cannot type in the member numbers directly into “Application Region” box.) One can either type in the members selected (e.g. Node 1:10, Surface 1:14, …etc.) or use the mouse to select the members by boxing out the members of concerns. Next, click on “Add” button. The ID numbers of the members selected will appear in the Application Region” box and then “Apply” in the left form.

III.5.6 NASTRAN Analysis The next step is to submit our plate to NASTRAN for analysis. It should be mentioned here that MSC/PATRAN automatically generates the NASTRAN bulk data file (.bdf) and also submit the .bdf file to NASTRAN. The way we do this is shown here. The first thing is to click on the Analysis' button of the control panel. We have the form shown on the left below. We want to analyze the `Entire Model' in `Full Run'. The `Job Name' we supplied is `classwk1'. The user is free to use any job name. Notice that the analysis code to be used is MSC/NASTRAN and the solution type is `Structural'.


To select MSC/NASTRAN as the analysis code, go to Preference menu. Choose analysis option, and then choose MSC/NASTRAN in the pop-up menu of Analysis Code .

When the `Translation Parameters...' button is clicked on, we have the form displayed on the right above. All default values in this form are accepted. Notice that we have `OP2 and Print' under `Data Output' this ensures that results will be printed both in the NASTRAN OUTPUT2 format and the .f06 ASCII format. Furthermore, the OUTPUT2 request is going to be done via the P3 built-in automatically. Clicking on the `Solution Type...' button will give the form shown below. The form is accepted as defaulted by pressing “OK”.


The next step is to create subcases for the analysis. This is done by clicking on `Subcase Create...’, which then displays the form on the left below. To make lcase_1 (i.e. the load case created earlier) a possible subcase, we click on lcase_1 under Àvailable Load Cases' and then press apply. This will place lcase_1 under Àvailable Subcases'. To make requests for the variables we want in our output for this subcase, we click on Òutput Requests...' and we have the form on the right below. In this form, we click on `Displacements', Èlement Stresses' and `Constraint Forces' one after the other under `Select Result Type'. Pressing ÒK' will place these three output requests under Òutput Requests' as indicated. The indicated output requests will be printed both in the .f06 and .op2 files. We then press 'Cancel' to conclude this selection. We then press Àpply' in the form on the left to conclude Òutput Requests' and `Cancel' to conclude `Subcase Create'.


The last step is to select the subcases (out of the subcases created above) that will be used for analysis. This is done by clicking on `Subcase Select...'. This will display the form shown below. We click on lcase_1 under `Subcases For Solution Sequence: 101' and then press `OK'. This will place lcase_1 under `Subcases Selected'. Thus lcase_1 has been selected for analysis. We then press `Cancel' to conclude this step. After all these have been concluded, we press `Apply' on the main application menu form and MSC/PATRAN starts the creation of classwk1.bdf (the bulk data file for the job `classwk1'). At this point, the user waits until this process is completed. After the creation of classwk1.bdf, MSC/PATRAN automatically submits the file for analysis to NASTRAN without the user intervention. At the conclusion of NASTRAN analysis, we will now have in the current directory, among the files generated by NASTRAN, classwk1.op2 (the binary OUTPUT2 format result file) and also classwk1.f06 (an ASCII result file). At this point we are set for post processing to see graphically our results. III.5.7 Post Processing The post processing procedure begins by translating the binary NASTRAN output2 (.op2) file using 'MSC/PATRAN'. This is done by clicking on the 'Analysis' radio button. When the required form is displayed (shown below), the option 'Read Output2' is chosen under 'Action' , 'Both' is chosen as 'Object' and 'Translate' is chosen as 'Method'.


Clicking on 'Translation Parameters...' will bring the form shown on the left below. All settings in this form are accepted as default. Clicking on 'Select Results Files' will bring the form shown on the right below. The user then supplies the .op2 file in 'Selected Results File '. This can be done by clicking on the required .op2 file under 'Available Files'. In this example, we have 'classwk1.op2' under 'Available Files'. We then press 'OK' to accept our selection and 'Cancel' to conclude this step.


When the translation process is completed, we are ready to view our results. This is achieved by clicking on the 'Results ' radio button on the control panel. We then have the form shown below on the left in which we have chosen 'Basic' as the 'Form Type '. The result for our load case, i.e., load_case_1, is automatically placed under 'Select Result Cases'. We then click on our load case name as shown on the right below and the results that can be displayed by fringe plotting are indicated under 'Select Fringe Result' while those that can be displayed line plotting are indicated under 'Select Deformation Result'.


Clicking on '2.1 _ Displacements, Translational' under 'Select Deformation Result' will lead to the display of the deformed shape shown below. Clicking on '3.1 _ Stress Tensor' under 'Select Fringe Result' will lead to a display of the fringe plot of the stress tensor superimposed on the deformed shape as shown below.


III.5.8 Homework For the lug shown on page 4.29 (whose geometry you have generated), specify the loads and boundary conditions indicated in the figure. Also generate a suitable finite element mesh and perform NASTRAN analysis. Submit a print out of the deformed shape and the fringe plot of the Von Misses Stresses.


III.6 LAB SESSION 4: 1D Beam Problem using Bar Elements The objective of this lab session is to demonstrate the use of the 'Element Props ' radio button for problems involving one-dimensional bar or beam elements. The example problem is the frame shown below which is made up of 1D straight curves. Geometry creation, boundary conditions and loads assignment, load case creation, finite element mesh generation, material properties assignment, submission of job to NASTRAN for analysis and post processing can be handled in a manner similar to the last example. Hence, enough information will only be provided to guide us through the example. Necessary detailed explanations will be given under 'Element Props ' III.6.1 Geometry Points 1 to 8 in the above figure are at (x y z) of [0 1 0], [1 1 0], [1 1 1], [0 1 1], [0 0 0], [1 0 0], [1 0 1], [0 0 1] respectively. 2_Point straight curves are generated as shown by connecting points 1 and 2, 2 and 3, 3 and 4, 1 and 5, 2 and 6, 3 and 7, 1 and 4, and finally, 4 and 8. III.6.2 Load/BCs All six degrees of freedom are constrained for points 5 through 8, that is ‘Translations’ and 'Rotations ' vectors are both equal to <0 0 0>. Force vectors of <0 _120 50> and <30 40 0> act at points 1 and 3 respectively. For both points, the moment vector is <0 0 0>. A load case with any name the user wants is created from the specified boundary conditions and the imposed point forces. III.6.3 Mesh Generation


First mesh seeds are created, one mesh seed per curve. Thus each curve is represented by one element. After this, a mesh is created to connect the nodes created by the mesh seed operation. During mesh creation, in the 'Application Menu' form, 'Curve' is chosen under 'Type ' and 'Bar2' is chosen under 'Element Topology'. Under 'Curve List', the entire geometry is chosen, i.e. Curves 1 through 8. 'Equivalencing' and 'Optimization' steps are carried out as usual. III.6.4 Material Selection An isotropic material with 'Elastic Modulus ' of 3.E7 and a 'Poission Ratio' of .33 is specified in the 'Input Properties...' form. You can specify any name for the material under 'Material Name '. III.6.5 Element Properties Proceed with element properties' specification as in the previous example. However in the 'Element Props Application Menu' form, ‘Object’ should be '1D' and 'Type ' should be 'Beam'. Under 'Options ' select 'General Section' In this example, even though the section properties of the bar elements are the same, the elements belong to three different groups, each group having its own orientation vector. Therefore, three property sets will be assigned, each set for each orientation vector. The first orientation vector is for the elements in the two curves parallel to the x_axis, the second is for the elements in the two curves parallel to the z_axis, and the third is for the elements in the four curves parallel to the y_axis. The user is free to supply 'Property Set Name ' of choice for each of these sets. For the first property set, 'Input Properties...' is clicked on and the appropriate form is displayed. A portion of this form is shown below. The information required to be supplied in the above form are listed below. Where a particular


information is applicable to our example, the required value (real scalar or vector) is given in front of the information. The values given are for the first property set, i.e. for curves parallel to the x-axis. Material Name: m:steel (Note: Material name 'steel' is used for illustrative purpose)

Bar Orientation: [1 1 0] (Note: This is the orientation vector, V. See p. 2.25 of the manual).

[Offset @ Node 1]: Not applicable, skip.

[Offset @ Node 2]: Not applicable, skip

[Pinned DOFs @ Node 1]: Not applicable, skip.

[Pinned DOFs @ Node 2]:

Area: 8.0 (Note: This is a real scalar)

[Inertia 1, 1]: 10.667 (Note: See p. 2.16 of the manual).

[Inertia 2, 2]: 2.667

[Inertia 2, 1]: Not applicable, skip.

Torsional Constant: 13.334 (Note: This is 'J'. See p. 2.17 of the manual)

[Shear Stiff, Y]: 0.8333 (Note: See p. 2.17 of the manual).

[Shear Stiff, Z]: 0.8333

[Nonstructural Mass]: Not applicable, skip.

[Y of Point C]: 2. (Note: See p. 2.17 of the manual).

[Z of Point C]: 1.

[Y of Point D]: 2.

[Z of Point D]: _1.

[Y of Point E]: _2.

[Z of Point E]: _1.

[Y of Point F]: _2.

[Z of Point F]: 1.

Station Distances: Not applicable, skip. For the curves parallel to the y-axis, the vector of 'Bar Orientation' is [1 1 0] while for the curves parallel to the z-axis, the vector is [0 1 1]. Every other information is the same as for the curves parallel to the x-axis.


III.6.6 NASTRAN Analysis Submit your job for NASTRAN analysis following the steps in III.5.6 of the plate problem. Be sure to confirm the load cases and boundary conditions before submitting job for analysis. III.6.7 Post Processing Perform the post processing of your NASTRAN .op2 result following the steps in Section III.5.7 of the plate problem. In the 'Result Menu Form' shown below, select 'Von Mises' (stress quantity to be reported) as the option under 'Result Quantity' and select '6_At Center' (the position where stress quantity will be reported) under 'Position' at point C which implies that the stress at Point C defined in Section 6.5 will be reported.


The fringe plot of the Von Mises stresses is superimposed on the plot of the deformed shape as shown below. III.6.8 Homework Construct the model Baja frame given below using MSC/PATRAN. The coordinates of the points are given. Connect these points with bar elements. The frame is subjected to the following forces: 100 lbf in the negative z-direction at point 8, 50 lbf in the positive x-direction at point 17, 120 lbf in the negative z-direction at point 17, 130 in the negative z-direction at each of points 1 and 2, and 150 lbf in the negative z-direction at each of points 24 and 26. Points 7, 10, 23 and 25 are the points of connection for the wheels. Constrain these points in the x, y, z, directions. The direction is left free because of the wheels. Use aluminum for the frame. Perform NASTRAN analysis to compute the deformation and the Von Mises stresses in the frame.


5 10 16 21 25 28

1723191373

2 612 18

22 26

2420159

41

11 17

148

YX

Z

Grid X Y Z

1 108.0 1.0 23.0 2 108.0 23.0 23.0

3 108.0 1.0 0.0

4 96.0 -2.0 26.0

5 108.0 23.0 0.0

6 96.0 26.0 26.0

7 96.0 -2.0 0.0

8 60.0 -6.0 78.0

9 72.0 -8.0 32.0

10 96.0 26.0 0.0

11 60.0 30.0 78.0

12 72.0 32.0 32.0

13 72.0 -8.0 0.0

14 36.0 -2.0 78.0

15 48.0 -4.0 28.0

16 72.0 32.0 0.0

17 36.0 26.0 78.0

18 48.0 28.0 28.0

19 48.0 -4.0 0.0

20 24.0 0.0 24.0

21 48.0 28.0 0.0

22 24.0 24.0 24.0

23 24.0 0.0 0.0

24 0.0 0.0 24.0

25 24.0 24.0 0.0

26 0.0 24.0 24.0

27 0.0 0.0 0.0

28 0.0 24.0 0.0

lab. manual me670/441 - department of mechanical

Documents