(c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

AOO-45051
2000-01-5585

Fastners Modeling for MSC.Nastran Finite

Element Analysis
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Alexander Rutman and Adrian Viisoreanu

The Boeing Company

John A. Parady, Jr.

MSC. Software Corportation

2000 World Aviation Conference

October 10-12, 2000
San Diego, CA

1 The Engineering Society

' For Advancing Mobility
f Land Sea Air and Space^
AM A A
INTERNATIONAL American Institute of Aeronautics
SAE International and Astronautics
400 Commonwealth Drive 370 L'Enfant Promenade, S.W.
Warrendale, PA 15096-0001 U.S.A. Washington, D.C. 20024
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics or SAE International.
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

Published by the American Institute of Aeronautics and Astronautics (AIAA) at 1801 Alexander Bell Drive,
Suite 500, Reston, VA 22091 U.S.A., and the Society of Automotive Engineers (SAE) at 400
Commonwealth Drive, Warrendale, PA 15096 U.S.A.

Produced in the U.S.A. Non-U.S. purchasers are responsible for payment of any taxes required by their
governments.

Reproduction of copies beyond that permitted by Sections 107 and 108 of the U.S. Copyright Law without
the permission of the copyright owner is unlawful. The appearance of the ISSN code at the bottom of this
page indicates SAE's and AlAA's consent that copies of the paper may be made for personal or internal
use of specific clients, on condition that the copier pay the per-copy fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923. This consent does not extend to other kinds of
copying such as copying for general distribution, advertising or promotional purposes, creating new
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

collective works, or for resale. Permission requests for these kinds of copying should be addressed to
AIAA Aeroplus Access, 4th Floor, 85 John Street, New York, NY 10038 or to the SAE Publications Group,
400 Commonwealth Drive, Warrendale, PA 15096. Users should reference the title of this conference
when reporting copying to the Copyright Clearance Center.

ISSN #0148-7191
Copyright © 2000 by The Boeing Company. Published by American Institute of Aeronautics and
Astronautics, Inc. and SAE International with permission.
Copyright ©2000 by MSC.Software Corporation. Published by American Institute of Aeronautics and Astronautics
Inc. and SAE International with permission.
All AIAA papers are abstracted and indexed in International Aerospace Abstracts and Aerospace
Database.
All SAE papers, standards and selected books are abstracted and indexed in the Global Mobility Data-
base.
Copies of this paper may be purchased from:
AlAA's document delivery service
Aeroplus Dispatch
1722 Gilbreth Road
Burlingame, California 94010-1305
Phone: (800) 662-2376 or (415) 259-6011
Fax:(415)259-6047

or from:

SAExpress Global Document Service

c/o SAE Customer Sales and Satisfaction
400 Commonwealth Drive
Warrendale, PA 15096
Phone: (724) 776-4970
Fax: (724)776-0790

SAE routinely stocks printed papers for a period of three years following date of publication. Quantity
reprint rates are available.

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise,
without the prior written permission of the publishers.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE
or AIAA. The author is solely responsible for the content of the paper. A process is available by which
discussions will be printed with the paper if it is published in SAE Transactions.
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

2000-01-5585

Fasteners Modeling for MSC.Nastran Finite Element Analysis

Alexander Rutman and Adrian Viisoreanu
The Boeing Company

John A. Parady, Jr.

MSC.Software Corporation

Copyright 2000 by The Boeing Company. Published by SAE International and the American Institute of Aeronautics and
Astronautics, Inc. with permission.
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Copyright 2000 by MSC. Software Corporation. Published by SAE International and the American Institute of Aeronautics and
Astronautics, Inc. with permission.

ABSTRACT INTRODUCTION

The distribution of loads between the components of a The common practice in aircraft structural analysis is the
structural assembly depends not only on their creation of large finite element models with a coarse
dimensions and material properties but also on the mesh with further extraction of separate parts along with
stiffness of fasteners connecting the components. So, the applied loads for hand analysis or for preparation of more
accuracy of the finite element analysis is influenced much detailed models. As a rule, these parts are connected in
by the fastener representation in the model. large models rigidly, i.e. they share the common grid
points. However, the distribution of loads between
This paper describes an approach designed specifically structural parts depends not only on the parts dimensions
for joints with connected plates modeled by shell and mechanical properties of selected materials, but also
elements located at plates mid planes. The procedure is on the stiffness of connecting elements, such as bolts
based on definition of independent components of a and rivets.
fastener joint flexibility, analysis of each component, and
their assembly to represent a complete plate-fastener With increase of computers speed along with the volume
system of the joint. of available memory, the trend for creation of more
detailed models has arisen. These models more
The proposed modeling technique differs from the realistically represent not only structural parts but also
traditional approach where all the connected plates are their interaction including fastener joints.
modeled coplanar. The traditional approach is based on
calculating a single spring rate for a particular The widely used method of fastener joints modeling is
combination of fastener and plate properties. The the joining of co-linear or co-planar finite elements of
application of this approach is limited by single shear joint connected structural parts with elastic elements
of two plates or symmetric double shear joint of three representing fasteners. The stiffness of these elastic
plates. It cannot be used for other joint configurations elements, or springs, is calculated using formulae
and for joints with larger number of connected plates. developed by empirical or semi-empirical methods. As a
The proposed procedure is free of those limitations. rule, these formulae consider the combination of
mechanical and geometric properties of a fastener and
Considering each fastener requires the creation of joined plates. Their application is usually limited to single
additional nodes and elements, it is obvious the manual shear and double shear symmetric joints.
use of this procedure is practically impossible for large
models of aircraft structures that could have thousands of With developing models more closely representing
fasteners. A new MSC.Patran utility that automates the structures but still consisting of plate elements, the joined
fasteners modeling was written and is described in the
paper. It takes advantage of the CBUSH element
formulation in MSC.Nastran and provides a user friendly
and efficient tool that creates fasteners connecting a
selected group of nodes.
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)1 Sponsoring Organization.

elements are no longer located in the same plane. In this

Fastener
case, the single elastic element cannot fully reflect the
work of a fastener joint.
=l
A
st
The procedure for modeling of fastener joints for detailed 1 plate
finite element models with non-coplanar joined parts was 2nd plate
described in the paper presented at the First MSC
Conference for Aerospace Users [1]. However, the ) 3rd plate di
practical use of this method showed some of its /
disadvantages, which will be discussed later. ———— ——————
j 4th plate
1
The approach to 3-dimensional modeling of fastener I— I
joints is based on definition of each deformation
component contributing to a joint flexibility and modeling
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

them by corresponding finite elements. Combination of

these elements represents the complete work of a 1L
fastener joint. Some relative displacements in the model
of a fastener joint were limited to ensure the compatibility
Figure 1. Fastener joint.
of deformations.

This paper presents an updated method for the finite

element modeling of fastener joint for MSC.Nastran and
an example of a model with fasteners. It also describes a where ECp. - compression modulus of plate / material;
new MSC.Patran utility for fastener joints modeling. The
method does not consider the effect of fastener tp. - thickness of plate /.
pretension and fit. Following the aerospace industry
common analysis practice, the friction between joint parts The fastener translational bearing flexibility at plate /
was not taken into account.

STIFFNESS OF FASTENER JOINT C

btf, = P l.
>=cf p;
In a fastener joint (Figure 1) the following stiffness
components are considered: where Ecf - compression modulus of fastener
translational plate bearing stiffness; material.
translational fastener bearing stiffness;
rotational plate bearing stiffness; Combined fastener and plate translational bearing
rotational fastener bearing stiffness; flexibility at plate /
fastener shear stiffness;
=
fastener bending stiffness. Cbtj Cbtpj + Cbtfj

Combined translational bearing stiffness at plate /

Under load, the plates slide relative to each other. This
causes the translational bearing deformations of joined
1
plates and a fastener. The translational bearing flexibility
of plate / is:

1 The relative rotation of the plate and fastener creates a

moment in the plate-fastener interaction (Figure 2). The
bearing deformations caused by this relative rotation are
assumed distributed linearly along the plate thickness
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

x
A

72
Pi
2

The rotational bearing flexibility of plate /

dx
12
rUhn -V -
"- M~ F ,3

The fastener rotational bearing flexibility at plate /

Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

C =
^' -^-r

Combined fastener and plate rotational bearing flexibility

at plate /
Figure 2. Rotational bearing stiffness definition.
C
br, = Cbrpj + Cbrfi

Combined rotational bearing stiffness at plate /

8 = X(p
1
where x - coordinate along the plate thickness; S
br, =
<P - angle of relative rotation of the plate and
fastener.
The bearing stiffness is modeled by elastic elements.
Stiffness of a dx thick slice of plate / is: The shear and bending stiffness of a fastener are
represented by a beam element.
=EcPidx
MODELING OF A FASTENER JOINT

Load on dx thick slice of plate / caused by the plate Modeling of a fastener joint is illustrated here using
bearing deformation MSC.Nastran.

j =x<pEcpdx REPRESENTATION OF A FASTENER JOINT

Moment of cF force about the plate /center line Idealization of a plate-fastener system includes the
following:
dM - xdF = Ecp<px2dx • Elastic bearing stiffness of a plate and fastener at
contact surface;
Moment in the plate-fastener contact caused by the plate • Bending and shear stiffness of a fastener shank;
deformation • Compatibility of displacements of a fastener and
connected plates at the joint.

The presented method creates the plate-fastener system

illustrated in Figure 3.
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

NOTES:

• CBAR element
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

• Orientation vector is
parallel to Y axis of
fastener coordinate system
•REAR element
• Dependent DOF's are ijk
•Arrow points to
dependent node

•CBUSH element
•Stiffness DOF's 2356
•X axis is aligned with
fastener axis

Nodes Npj and Nfj are

coincident, but shown offset for
clarity.
The analysis coordinate system of
all nodes is the fastener coordinate
system.

X axis of fastener
w coordinate system

Figure 3. Fastener joint modeling.

 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

FASTENER MODELING fastener axis and defining the fastener shear plane and
rotational stiffness about the same axes.
A fastener is modeled by CBAR or CBEAM elements [2]
with corresponding PBAR or PBEAM cards for properties For the modeling of the bearing stiffness, two sets of
definition. For the CBAR or CBEAM elements coincident grid points mentioned above are used. Each
connectivity, a separate set of grid points coincidental pair of coincident grid points, i.e. the plate node and
with corresponding plate grid points (Figure 3) is created. corresponding fastener node, is connected by CBUSH
This set also includes grid points located on intersection element [2] or combination of CELAS2 elements with
of the fastener axis and outer surfaces of the first and equal translational stiffness along the axes normal to the
last connected plates. fastener axis and equal rotational stiffness about the
same axes. The connectivity card CBUSH must be
All CBAR or CBEAM elements representing the same accompanied by PBUSH card defining the stiffness. The
fastener reference the same PBAR or PBEAM card [2] CELAS2 card accomplishes both functions, but 4
with following properties: CELAS2 elements are required to replace one CBUSH
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

element. However it is difficult to interpret CELAS2

• MID to reference the fastener material properties. element forces.
• Fastener cross-sectional area
For correct definition of a fastener shear plane and its
f axial direction, a coordinate system with one of its axis
parallel to the fastener axis must be defined in the bulk
where df - fastener diameter. data. This coordinate system must be used as analysis
coordinate system for both sets of grid points.
Moments of inertia of the fastener cross section
An example of the bearing stiffness modeling using the
CBUSH and PBUSH cards is given in Table 3. It is
/ r = / 2 = - 64 assumed in the example that the fastener axis is parallel
to x-axis of corresponding coordinate system. An
Torsional constant
alternative method for the bearing stiffness modeling
using CELAS2 elements is shown in Table 4.

COMPATIBILITY OF DISPLACEMENTS IN THE JOINT

32
The fastener joint model was designed under the
• Area factors for shear of circular section following assumptions:
/C, = K2 = 0.9
• The plates are incompressible in transverse
An example of CBAR element and its properties direction;
definition for .375" dia. fastener is shown in Table 1. An • The plates mid planes stay parallel to each other
alternative form of CBAR properties definition is under the load;
presented in Table 2. • Planes under the fastener heads stay parallel to the
plate mid planes under the load.
Definition of a fastener using CBEAM and PBEAM cards
is similar to that shown in Table 2 for CBAR and PBAR These goals are reached by using REAR elements.
with small differences described in Reference [2].
An example of a group of RBAR elements satisfying the
MODELING OF INTERACTION BETWEEN FASTENER above compatibility conditions is given in Table 5. It is
AND JOINED PLATES also assumed in this example (Figure 3) that the fastener
axis is parallel to the x-axis of the corresponding
The interaction between a fastener and plate results in coordinate system.
bearing deformation of all parts of the joint on their
surfaces of contact. The bearing stiffness of a fastener The first RBAR card forces the plane under the fastener
and connected plates is defined in Section "Stiffness of head to stay parallel to the first plate mid plane under the
fastener joint". The bearing stiffness is presented as load. It also prevents the fastener movement as a rigid
translational stiffness in direction of axes normal to the body. The middle RBAR cards support the first two
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

assumptions. They keep the constant distance between rotation of plates keeping them parallel to each other.
the plate mid planes, i.e. assume that plates are The last card forces the plane under the other head of
incompressible. They also guarantee zero relative the fastener to stay parallel to the last plate mid plane.

CBAR EID PID GA GB XI X2 X3

CBAR 21 206 1011 2011 1.0 0.0 0.0

PBAR PID MID A 11 12 J NSM

PBAR 206 2 .11 9.7E-4
Cl C2 Dl D2 El E2 Fl F2
0.0 0.0
Kl K2 112
0.9 0.9
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Table 1. Example CBAR and PBAR cards.

PEARL PID MID GROUP TYPE

PBARL 206 2 ROD
DIM1 NSM
.375

Table 2. Example PBARL card.

CBUSH EID PID GA GB GO/X1 X2 X3 CID

CBUSH 210 12 1005 2005 0

PBUSH PID "K" Kl K2 K3 K4 K5 K6

PBUSH 12 K 1.6E7 1.6E7 5.2E3 5.2E3

Table 3. Example CBUSH and PBUSH cards.

CELAS2 EID K Gl Cl G2 C2
CELAS2 210 1.6E7 1005 2 2005 2
CELAS2 211 1.6E7 1005 3 2005 3
CELAS2 212 5.2E3 1005 5 2005 5
CELAS2 213 5.2E3 1005 6 2005 6

Table 4. Example CELAS2 cards.

REAR EID GA GB CNA CNB CMA CMB

REAR 310 905 1005 123456 1456
REAR 311 1005 2005 123456 156
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
REAR 314 5005 6005 123456 156
REAR 315 6005 915 123456 56

Table 5. Example RBAR cards for compatibility of displacements in the joint.

 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

MODELING EXAMPLE Figure 5 presents the analysis results. The

displacements at a fastener location consist of the
A symmetric double shear joint was modeled as an fastener movement as a rigid body, the combined plates
example (Figure 4). The modeled structure consists of and fastener bearing deformations, and the fastener
three aluminum plates and two titanium fasteners. The bending and shear deformation. The results of analysis
thickness is 0.15" for outer plates and 0.2" for inner plate. are in good agreement with the expected behavior of the
The fastener diameter is 0.25". The inner plate is loaded joint under load.
by a distributed load of 5000 pound/in. The model is
constrained at outer plates. The bulk data file is given in
Appendix.

Constraints
Outer plates
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Inner plate

5.00+03

5.00+03 Distributed
^ Load

Fasteners

Figure 4. Example of finite element model with fasteners.

 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)1 Sponsoring Organization.

135 Deformed
Undeformed
model model
1235

135
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Fastener
bending and shear
deformation

Rigid elements
(RBAR's)

Fastener
before
deformation
Fastener
after
deformation

Inner plate
bearing
deformation

Outer plate
bearing
deformation

Figure 5. Displacements of example finite element model.

 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

COMPARISON OF MODELING TECHNIQUES fastener-plate contact are distributed through the model
structural parts and causes additional stresses not
To compare the modeling technique described in this existing in real structure.
paper with one developed in Reference [1] the finite
element model with fine mesh was created (Figure 6). Figure 7 illustrates the behavior of a fastener joint
This model is the same example model presented in modeled using the both discussed techniques. Plates in
previous section with the only difference in mesh density. the joint modeled using the proposed technique have
The fine mesh was employed to show deformation of only in-plane deformations. If the Reference [1] technique
fasteners and particularly the joined plates. is employed, plates have clear out-of-plane deformations.

Outer plate Deformation of joint modeled using proposed techniques

Fastener
Reactions Inner plate before Fastener after
deformation deformation
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

P/2

Applied P/2
Fasteners loads Plates deform only in their planes

Deformation of joint modeled using Reference Ml techniques

Figure 6. Fine mesh finite element model for comparison Fastener
of modeling techniques. before Fastener after Deformed
deformation deformation model
Undeformed
model
P/2
Under the load, the joined plates slide along each other
due to combined plates and fastener translational
bearing deformation and the fastener bending and shear
deformation. The fastener deformation causes change of P/2
angle between fastener and plate or in other words their
relative rotation. This relative rotation results in non- Figure 7. Comparison of results obtained by two
uniform distribution of bearing stress through the plate modeling techniques.
thickness. The resultant load transferred through the
contact area between the fastener and plate consists of a
force in the plate mid plane and out-of-plane moment. In
the structure, the moment is reacted by loads on the MSC.PATRAN UTILITY
plate contact surfaces and does not cause the plates
local bending. This section presents the algorithm of the newly
developed MSC.Patran utility, the data input forms
The proposed modeling technique takes this (Graphical User Interface) and an example of fastener
phenomenon into account ensuring the plates mid planes joint modeling using this utility.
stay parallel to each other under load. This is reached by
use of rigid elements RBAR's connecting the plate nodes UTILITY DESCRIPTION
at the fastener location and forcing them to keep the
same angle of rotation during the deformation. Extraction of plate nodes for connection by fastener

The modeling technique presented in Reference [1] The MSC.Patran utility for modeling of fastener joints is
assumes that plates follow locally the fastener applied to a group of nodes selected by user in the model
deformation. It means the fastener guides the connected area where the group of fastener joints must be created.
plates and it results in bending of plates and interference
between them. The plates bending moments in the
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

The program extracts sub-groups of nodes from the Fastener Coordinate System
entire group. Each sub-group is associated with one
fastener. The criterion for node subgroups creation: the Two options of the fastener coordinate system definition
distance between any two nodes of the subgroup must are available to the user: manual and automatic.
be smaller than or equal to the fastener length supplied
by user. The fastener length chosen by user should be If the user selects the manual option the program
bigger than the longest fastener but smaller than the requires ID of one of previously defined coordinate
distance from any of subgroup nodes to grid points not systems and ID of coordinate axis parallel to the fastener
belonging to the subgroup to avoid creation of axis. This coordinate axis will be addressed as reference
undesirable elements. This condition can influence not axis of the fastener coordinate system.
only the user's definition of fastener length but also the
selection of initial group of nodes. If the automatic option is chosen, the program either
selects one of previously defined coordinate systems or
The procedure assumes that fasteners in considered creates a new one. With the automatic option, the X-axis
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

group have the same diameter and material. Fasteners of the fastener coordinate system is always directed
with different diameters or material cannot be combined along the fastener axis. The direction of X-axis of the
in one group. In this case separate groups of nodes must fastener coordinate system is defined as weighted
be selected. average of normals of all elements adjacent to the
fastener nodes:
Plate properties
n n/
Thickness and material properties of plate elements II
using a subgroup node for connectivity are extracted i=1 k=1
from the MSC.Patran database and do not require the
user's input.

If connected structural parts include tapered plates, the

thickness of plate elements adjacent to the fastener can
be different. Moduli of elasticity for those elements can where X - X-axis vector;
also differ if the influence of temperature distribution R, - distance from node NP! to centroid of
along the structure is considered in the analysis. To take element / adjacent to node Npi;
these phenomena into account the thickness and
modulus of elasticity for bearing stiffness analysis are Nk - normal of element /;
calculated as weighted average of plates adjacent to
- length of vector Nk ;
node /\L:
n - number of plate nodes NP!.

14
k=1
The program performs an alignment check before
computing the direction of the X-axis. If the angle
between normal of element k adjacent to node / and
—t normal of element 1 adjacent to node 1 is greater than
90°, then direction of normal of element k is reversed for
computational purposes.
where tpl - average plate thickness at node NPI;
Epi - average plate modulus at node NP!; To reduce the number of coordinate systems in the
t'k - thickness of element k adjacent to node model, the program checks the MSC.Patran database for
NP,; existing coordinate systems that could be used to define
R'k - distance between centroid of element k and the orientation of the current fastener with the following
test:
node A/p, ;
E'k - modulus of element k adjacent to node
NP!;
HI - number of plates adjacent to node NPI.

10
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

Model symmetry

<sina The symmetry coefficient 77 is used to scale the

X properties of fasteners located on symmetry planes.

If the structure has one symmetry plane and only half of

where X; - X axis vector of existing coordinate the structure is modeled, 77 = 0.5 for the fasteners located
system C/D;; on the symmetry plane.
X j X X - vector product of X} and X ;
For structures with two symmetry planes, when only
a - tolerance angle, default value 1°. quarter of the structure is modeled, a fastener located on
the intersection of the two symmetry planes has 77 = 0.25.
If an existing coordinate system satisfying the above test For other fasteners located on symmetry planes, 77 = 0.5
is found, the program uses it to define the axis of the is used.
current fastener. Otherwise a new coordinate system is
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

created and committed to the database. Coordinates of For fasteners not located on symmetry planes, rj = 1.0.
the new coordinate system origin are
Modeling of fasteners
II

I.*' Each fastener is represented by a group of CBAR

Y - elements. The program creates CBAR elements between
n ~ n n nodes NH1, A/F, ..., NFn, NH2(see Figure 3). The orientation
vector of the CBAR elements is given by one of axes
where X:, Y,, Zi - coordinates of plate nodes A/P|. normal to the reference axis of the fastener coordinate
system. Calculation of CBAR elements section properties
Two other axes of the fastener coordinate system must is described in Section "Modeling of a fastener joint".
be in plane normal to the fastener axis. Area, moments of inertia, and torsional constant of
fasteners located on symmetry planes are multiplied by
Alignment of plate nodes coefficient 77.

When different structural parts are modeled separately If a PBAR property card with same data already exists,
and sometimes by different modelers, it is possible that the program associates the current CBAR to the existing
some plate nodes, which should be connected by a property. Otherwise a new PBAR record is created.
single fastener, are not collinear. If the maximum
deviation of those nodes normally to the fastener Fastener-plate interface
direction is smaller than the tolerance established by
user, the program defines the fastener axis as passing CBUSH elements created between plate nodes NPi and
through point (X0, Y0, Z0) parallel to the reference axis of their corresponding fastener nodes A/F/ represent
the fastener coordinate system. If the deviation of plate fastener-plate interface. An example of expressions for
nodes exceeds the established tolerance, program stops the CBUSH stiffness coefficients when the fastener axis
and the corresponding message is displayed. is parallel to X-axis of the fastener coordinate system:

Creation of fastener nodes

S1 =0 S2 = S3 =
The new fastener nodes NFI are created as duplicates of
corresponding plate nodes NPi. In addition to these
nodes, the program creates the fastener head node NH1 -op. -cf
at distance tp1 / 2 from node NF1 and fastener node NH2 at
distance tpn from node NFn along the fastener axis, as .3
shown in Figure 3. The analysis coordinate system of all S4 = 0 Pi
fastener nodes is the fastener coordinate system.
1- i]
4\ OP
E
,
+£
cf

11
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

Where S1-6 - CBUSH stiffness coefficients method has been chosen, the user is required not only
(Reference [2]) identify the existing coordinate system as a fastener
coordinate system but also to tell the program which axis
The program then checks the MSC.Patran database for of the system is parallel to the fastener axis. When the
existence of a property card PBUSH with the same data. automatic option was selected, the user is not required to
If such PBUSH record is found, then the current CBUSH supply any additional information. In this case, the
element is associated with the existing property. program either selects an existing coordinate system
Otherwise a new PBUSH record is created. according to established criterion or creates a new one.
When the fastener is on intersection of two symmetry
Compatibility of displacements planes identification of coordinate system is not required
and the fastener axis selection panel is dimmed (Figure
Compatibility of displacements in the fastener joint is 10).
enforced by REAR elements. The RBAR elements are
created as shown in Figure 3. If the fastener is on a The user has to identify the region where fasteners will
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

symmetry plane, the degrees of freedom already be created by giving the program the list of plate nodes.
constrained by symmetry are eliminated from the It is not necessarily the program will use all this nodes for
dependant set of the RBAR elements. connection by fasteners. The fasteners will be created
only between nodes located not further from each other
MSC.PATRAN INPUT PANEL than the established by user maximum fastener length.

Figure 8 shows the input form for the MSC.Patran utility.

The starting ID of the new nodes and elements can be
selected, as it is usually done in majority of input forms.
On the symmetry coefficient panel the user has three
choices:

• 1.0 - for fasteners not located on symmetry

planes (Figure 8);

• 0.5 - for fasteners belonging to one plane of

symmetry (Figure 9);

• 0.25 - for fasteners located on the intersection of

symmetry planes (Figure 10).

If the fasteners are on a symmetry plane (Figure 9), the

user must identify the coordinate system with one of
coordinate planes coplanar with the symmetry plane.
This coordinate plane is indicated by perpendicular to its
coordinate axis.

When the fastener axis is on the intersection of two

symmetry planes (Figure 10), the user must to identify
the coordinate system with one axis collinear wit
intersection line. Two other axes are in symmetry planes.

The fastener diameter and material are self-explanatory.

The fastener material listbox (Figure 8) contains all the
materials currently defined in the MSC.Patran database.
The fastener material must be created before this utility is
executed.

The user has two options for the fastener coordinate

system definition: manual and automatic. If the manual

12
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

Help

Starting Node to .- ID of first new

J1108 fastener node
pos
Starting £)e»n to
ID of first new fastener Starting Bern ID
jiSO -*\~~~~~~~ element ,__
Node Alignment Tolerance (Deg.) \
Node alignment Node Alignment Toterance (Deg.)]
PS.O ' '• -.
tolerance SiS———————
Symmetry Coefficient
Symmetry Coefficient
Symmetry panel
I Jl-0
I » 0.5
Fastener diameter I J0.25 Coordinate axis
normal to symmetry
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Select Sjfm, Wane Mormal plane

: .Coord 0,2?
Fastener not to exceed
length
Fastener Information
Diam«ter length
Fastener material 'CMS! i^^.-..—
MatertaJ
Fastener axis
manually selected

Fastener Axis Method

,i Automatic

Select fastener Axis

Coord 3.25

j Select Moct«s Selected

plate nodes
11 Mode 2:10:41 „ Auto Execute
! Select Nodes
Node2:10:4l

Apply Cancel
Figure 8. MSC.Patran utility input panel when fasteners
are not on symmetry plane.

Figure 9. MSC.Patran utility panel when fasteners are

on symmetry plane.

13
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)1 Sponsoring Organization.

EXAMPLE OF MSC.PATRAN UTILITY APPLICATION

Show Description...
Figure 11 shows a fragment of example model consisting
Starting Node ID
of three plates made from different materials. The shown
|:10S plate nodes were selected for connection by a fastener.
Starting Bern ID

Node Alignment Tolerance (DegJ

Symmetry coefficient
J-1,0 ' \
J0.5 I
Coordinate axis parallel
;I 0.2S
with the intersection of
Select Fastener Axis the two symmetry
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

-v™ « /"™y<««««^
planes
jCoordO.1

Fastener Information
Diameter Length Aluminum Place

Figure 11. Example model before fastener creation.

Fastener Axis M<ftho<J

The fastener axis is X-axis of local coordinate system 1.

Figure 12 shows the fastener created between the three
S«?tecf Fastener Axis selected nodes. Degrees of freedom of CBUSH and
REAR elements refer to the same local coordinate
system 1.
..* Auto Execute
j select Nodes
II Node 2:10:41

Apply Cancel

Figure 10. MSC.Patran utility panel when fasteners are

on intersection of two symmetry planes.

Figure 12. Example model with created fastener.

14
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

REFERENCES

1. Multi-Spring Representation of Fasteners for

MSC/NASTRAN Modeling. A. Rutman, J. B. Kogan.
Proceedings of The First MSC Conference for Aerospace
Users, Los Angeles, CA, 1997
2. MSC.Nastran Version 70.5 Quick Reference Guide, The
MacNeal-Schwendler Corporation, Los Angeles, CA 1998

CONTACT

Alexander Rutman, Ph.D. Adrian Viisoreanu, Ph. D. John A. Parady, Jr., P.E
Principal Engineer Principal Engineer Sr. Application Engineer
The Boeing Company The Boeing Company MSC.Software Corporation
P.O. Box 7730, MS K89-04 P.O. Box 3707, MS 9U-RF 1000 Main Street #190
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

Wichita, KS 67277 Seattle, WA 98124-2207 Grapevine, TX 76051

alexander.rutman @ boeing.com adrian.viisoreanu@boeing.com Phone:817-481-4812
john.parady@mscsoftware.com

15
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

APPENDIX. BULK DATA FILE FOR EXAMPLE MODEL

$ NASTRAN input file created by the MSC MSC/NASTRAN input file

$ translator ( MSC/PATRAN Version 8.5 ) on May 26, 2000 at
$ 12:37:02.
$
ASSIGN OUTPDT2 = 'dsh_new_course.op2', UNIT = 12
$
$ Linear Static Analysis, Database
$
SOL 101
TIME 600
CEND
$
SEALL = ALL
SUPER = ALL
TITLE = MSC/NASTRAN job created on 17-Feb-OO at 17:37:42
ECHO = NONE
MAXLINES = 999999999
$
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

SUBCASE 1
$ Subcase name : Tension
SUBTITLE=Tension
SPC = 2
LOAD = 2
DISPLACEMENT (SORT1, REAL) =ALL
SPCFORCES (SORT1, REAL) =ALL
OLOAD (SORT1, REAL) =ALL
STRESS (SORT1, REAL, VONMISES, BILIN) =ALL
FORCE (SORT1, REAL, BILIN) =ALL
$
$
BEGIN BULK
$
$
PARAM POST -1
PARAM PATVER 3.
PARAM AUTOSPC YES
PARAM INREL 0
PARAM ALTRED NO
PARAM COUPMASS -1
PARAM K6ROT 10.
PARAM WTMASS 1.
PARAM,NOCOMPS,-!
PARAM PRTMAXIM YES
$
$ Elements and Element Properties for region : pshell.1
$
PSHELL 1 1 .2 1 1
$
CQUAD4 1 1 12 11 15 14
CQUAD4 2 1 14 15 19 18
CQUAD4 3 1 18 19 23 22
CQUAD4 4 1 11 25 27 15
CQUAD4 5 1 15 27 31 19
CQUAD4 6 1 19 31 35 23
$
$ Elements and Element Properties for region : pshell.2
$
PSHELL 2 1 .15 1 1
$
CQUAD4 72 36 37 39 38
CQUAD4 8 2 38 39 43 42
CQUAD4 9 2 42 43 47 46
CQUAD4 10 2 37 49 51 39
CQUAD4 11 2 39 51 55 43
CQUAD4 12 2 43 55 59 47
CQUAD4 13 2 60 61 63 62
CQUAD4 14 2 62 63 67 66
CQUAD4 15 2 66 67 71 70
CQUAD4 16 2 61 73 75 63
CQUAD4 17 2 63 75 79 67
CQUAD4 18 2 67 79 83 71
$
$ Elements and Element Properties for region : pbar.4

16
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

$ 1.
PBAR 4 1 1. 1. 1.
$
CBAR 27 4 22 23 11 A
+ A 1
CBAR 28 4 23 35 11 B
+ B 1
$
$ Elements and Element Properties for region pbar.5
$
PBARL 5 2 ROD
+ C .125
$
CBAR 19 5 84 85 1. 0. 0.
CBAR 20 5 86 87 1. 0. 0.
CBAR 21 5 85 89 1. 0. 0.
CBAR 22 5 87 91 1. 0. 0.
CBAR 23 5 92 84 1. 0. 0.
CBAR 24 5 89 95 1. 0. 0.
CBAR 25 5 96 86 1. 0. 0.
CBAR 26 5 91 99 1. 0. 0.
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

$
$ Referenced Material Records
$
$ Material Record : aluminum
$ Description of Material : Date: 03-Feb-OO Time: 15:39:23
$
MAT1 1 1.05+7 .33
$
$ Material Record : titanium
$ Description of Material : Date: 08-Feb-OO Time: 17:45:11
$
MAT1 2 1.6+7 .3
$
$ Nodes of the Entire Model
$
GRID 11 1. 0. 0.
GRID 12 1. 1. 0.
GRID 14 1.5 1. 0.
GRID 15 1.5 0. 0.
GRID 18 3. 1. 0.
GRID 19 3. 0. 0.
GRID 22 4.5 1. 0.
GRID 23 4.5 0. 0.
GRID 25 1. -1. 0.
GRID 27 1.5 -1. 0.
GRID 31 3. -1. 0.
GRID 35 4.5 -1. 0.
GRID 36 0. 1. .175
GRID 37 0. 0. .175
GRID 38 1.5 1. .175
GRID 39 1.5 0. .175
GRID 42 3. 1. .175
GRID 43 3. 0. .175
GRID 46 3.5 1. .175
GRID 47 3.5 0. .175
GRID 49 0. -1. .175
GRID 51 1.5 -1. .175
GRID 55 3. -1. .175
GRID 59 3.5 -1. .175
GRID 60 0. 1. -.175
GRID 61 0. 0. -.175
GRID 62 1.5 1. -.175
GRID 63 1.5 0. -.175
GRID 66 3. 1. -.175
GRID 67 3. 0. -.175
GRID 70 3.5 1. - .175
GRID 71 3.5 0. - .175
GRID 73 0. -1. -.175
GRID 75 1.5 -1. -.175
GRID 79 3. -1. -.175
GRID 83 3.5 -1. -.175
GRID 84 1.5 0. .175
GRID 85 1.5 0. 0.
GRID 86 3. 0. .175
GRID 87 3. 0. 0.

17
 (c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

GRID 89 1 .5 0. -.175
GRID 91 3. 0. -.175
GRID 92 1 .5 0. .25
GRID 95 1..5 0. -.25
GRID 96 3 0. .25
GRID 99 3 0. -.25
$
$ Loads for Load Case : Tension
$
SPCADD 2 ^ 3
$
LOAD 2 1. 1
$
$ Displacement Constraints of Load Set : spc_2
$
SPC1 1 2 37
$
$ Displacement Constraints of Load Set : spc_l
$
SPC1 3 135 36 37 49
Downloaded by CARLETON UNIVERSITY LIBRARY on July 30, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.2000-5585

$
$ Displacement Constraints of Load Set : spc_3
$
SPC1 4 15 60 61 73
$
$ Distributed Loads of Load Set : Tension
$
PLOAD1 1 27 FYE FR 0. -5000. 1. -5000.
PLOAD1 1 28 FYE FR 0. -5000. 1. -5000.
$
$ Bearing Stiffnesses
$
PBUSH 6 K 1267925.1267925. 4226. 4226.
PBUSH 7 K 950943. 950943. 1783. 1783.
$
CBUSH 31 7 39 84
CBUSH 32 6 15 85
CBUSH 33 7 63 89
$
CBUSH 34 7 43 86
CBUSH 35 6 19 87
CBUSH 36 7 67 91
$
$ Compatibility Conditions
$
REAR 41 92 39 123456 3456
REAR 42 39 15 123456 345
REAR 43 15 63 123456 345
REAR 44 63 95 123456 45
$
REAR 45 96 43 123456 3456
REAR 46 43 19 123456 345
REAR 47 19 67 123456 345
REAR 48 67 99 123456 45
$
$
ENDDATA

18