You are on page 1of 5

Appendix D - Tapered Beam Stiffness Matrix ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

Appendix D Derivation of Tapered Beam Stiffness Matrices
In ADAMS/WT, we create a tapered beam effect using the ADAMS FIELD element, which allows us to directly input the terms in the stiffness matrix. To derive the FIELD matrix terms which correspond to a linearly tapered beam, we go to the finite element method. Consider the transverse deflection of a single tapered beam element, shown below with the local coordinate system and the nodal displacements. Axial and torsional response will be considered later. q1 q2 q3 q4

w

x x=L

As is standard practice in finite element analysis (for example, see Bathe1i), we assume that the deflected shape can be approximated by a cubic polynomial for which the nodal displacements are the boundary conditions. In order to determine the element deflected shape from the nodal displacements, we assume that

w( x ) = ∑ Hi ( x )qi
1

4

A little bit of algebra then gives us the definitions for the shape functions, Hi :
x 2. 3 L x L. 3 L 2. x
3 3 3 3

H1( x, L )

x 3. 2 L x
2 2

2

1

H2( x, L )

2.

x L

L

H3( x, L )

L

x 3. 2 L x
2 2

2

H4( x, L )

x L. 3 L

3

L

1

Bathe, K.-J., Finite Element Procedures in Engineering Analysis, Prentice-Hall, NY, 1982

–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– ADAMS/WT 2.0 User's Guide D-1

we follow the same methodology. Again. we expand the deflection in terms of the shape functions: u( x ) = ∑ Hi ( x )qi 1 4 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– ADAMS/WT 2. but instead of a cubic assumed displacement over the length of the element. and that we know the values at the end points. this is standard practice. both for finite element analysis and for the ADAMS BEAM. we find the definition of the terms of the stiffness matrix is: kij = ∫ L 0 2 d 2 Hi d H j EI dx dx 2 dx 2 We then assume that the variation in EI is linear over the length of the element. Looking at the axial extension (u) problem (torsion is exactly analogous).Appendix D . we find q1 q2 u L x Similarly to the bending case. L) = EI o +  L x   L We can then complete the integrals symbolically to get only those terms which are needed for the FIELD statement.Tapered Beam Stiffness Matrix –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Returning to the finite element method. These are k 33 = k 34 k 44 6 (EI L + EI o ) L3 −2 = 2 (2 EI L + EI o ) L 1 = (3EI L + EIo ) L For the axial and torsional degrees of freedom. we assume a linear function.  EI − EI o  EI ( x.0 User's Guide D-2 .

ADAMS computes the forces on the I marker due to displacements of the I marker in the J marker’s reference frame. where these values are the running properties (per unit length) of in-plane and out-plane bending stiffness. because ADAMS’ formulation uses the displacement difference between the markers. That is. only those terms listed should be incorporated into the matrix. and that it is the combined properties which are assumed to vary linearly of the element length. GJ at the element boundaries. extensional stiffness and torsional –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– ADAMS/WT 2. just substitute GJ for EA. EA. it is important to remember that for each FIELD element. we find the definition of these terms in the stiffness matrix is: k ij = ∫ EA L 0 dH i dH j dx dx dx We then assume that the variation in EA is linear over the length of the element.  EA − EAo  EA( x. EIz. L ) 1 x L x L H2 ( x . where the assumption of globally small displacements allows everything to be assembled in one base frame.Tapered Beam Stiffness Matrix –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– where the shape functions in this case are H1 ( x . Moreover. Note that the input to this element are the combined stiffness properties at each end of the beam. L) = EAo +  L x   L We can then complete the integral symbolically to get the term which is needed for the FIELD statement. k 44 = 1 (GJ L + GJ o ) 2L In order to create the correct FIELD matrix. not cross section) beam is assumed to be straight in the unloaded configuration. and that we know the values at the end points. not the separate material and cross section properties. This is k11 = 1 (EAL + EAo ) 2L For the torsion degree of freedom. It is also important to note that this linearly tapered (in stiffness. L ) Again returning to the finite element method. you must specify EIy.Appendix D . This is different from finite element methods.0 User's Guide D-3 .

As with all ADAMS force elements. y. the inertial properties associated with the beam go separately into ADAMS PART s (see Appendix E . For convenience. the rows correspond to x. followed by rotations about the x. To maintain consistency with the approach used for the ADAMS BEAM element.Tapered Part Derivation). z x y In the FIELD matrix. with the z-axis toward the leading edge and the y-axis in the positive flap direction (toward the suction side of the airfoil. as shown below. k11 = 1 ( EAL + EAo ) 2L k21 = k31 = k41 = k51 = k61 = 0 k12 = 0 6 k22 = 3 ( EI ZL + EI Zo ) L k32 = k42 = k52 = 0 −2 k62 = 2 ( 2 EI ZL + EI Zo ) L k13 = k23 = 0 6 k33 = 3 ( EIYL + EIYo ) L k43 = 0 2 k53 = 2 ( 2 EIYL + EIYo ) L k63 = 0 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– ADAMS/WT 2.Appendix D . y and z displacements.Tapered Beam Stiffness Matrix –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– stiffness. the coordinate axes are set so that the blade runs along the positive x-axis from root to tip. and z axes.0 User's Guide D-4 . all terms are listed here in the same order which they would go into the FIELD statement (column-major order).

0 User's Guide D-5 .Tapered Beam Stiffness Matrix –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– k14 = k24 = k34 = 0 k44 = 1 (GJ L + GJo ) 2L k54 = k64 = 0 k15 = k25 = 0 2 k35 = 2 ( 2 EI YL + EI Yo ) L k45 = 0 1 k55 = ( 3EIYL + EIYo ) L k65 = 0 k16 = 0 −2 k26 = 2 ( 2 EI ZL + EI Zo ) L k36 = k46 = k56 = 0 1 k66 = (3EI ZL + EI Zo ) L –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– ADAMS/WT 2.Appendix D .