BENDING ANALYSIS OF A WIRE STRAND ANALYTIC MODEL

Filipe Paixão Geiger

Rogério Marczak
[email protected]
[email protected]
Mechanical Engineering Department, Federal University of Rio Grande do Sul.
Sarmento Leite Street, 435, 90050-170, RS, Brazil.

Abstract. An analytical approach, using the spatial beam theory, to determine the mechanical
response of cables is proposed in this work. Experimental results are limited and expensive to
be obtained and there is a lack of good benchmarks solutions to check the FEM models before
starting complex analysis. Therefore, in order to obtain the results the parametric spatial
description were derived for one strand, considering one wire surrounding a core, since the
response of all wires are similar. Solving the differential equilibrium equations of the spatial
beam theory and applying bending loads it was possible to obtain the mechanical response of
the wire. Six loading cases were analyzed for the wire helix analytically and numerically.
Good results were obtained when the both methodologies were compared for the cable
bending.

Keywords: Wire, Spatial Beam Theory, Differential Geometry, Parameterization, Bending.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
Bending analysis of a wire strand analytic model

1 INTRODUCTION

Cables are structures that can be used in many applications such as cable-stayed
bridges, offshore platforms and prestressed concrete structures. Its geometry is defined as a
straight core surrounded by layers of helical strands. The simplest, and most used,
configuration is the so called 1 + 6, where six helical wires are laid around the core forming
just one layer. The main reason to be used so diversely is the high strength-to-weight ratio and
the capacity to support large axial loads in comparison to bending and torsion loads.
Several authors have tried to develop theories and models to study cables. Many
numerical analyses, specially using Finite Element Method (FEM), are available and can be
used to study these structures. Although is a very powerful tool this approach demands time
and must be carefully employed. Stanova et al. (2011), Nawrocki and Labrosse (2000), and
Jiang (1999) are examples of works where this analysis was carried out using finite elements.
Analytical studies have been produced based on assumptions as curved beams theory,
purely tension wires and friction effects (Ghoreishi et al. (2007)). Some of the most know
theory are presented by Østergaard et al. (2011), Argatov (2011), Labrosse and Nawrocki
(2000) and Costelo (1990).
In order to validate the FEM or even the methodology used to represent the cable
behavior analytically, some classical experimental work are cited. The most known
procedures were performed by Machida. and Durelli (1973) and Utting and Jones (1987)

2 STRAND MODELING

In order to simulate the cable behavior the structure was modeled according to the space
curved beam theory. To do so, only one wire was considered, since the results can be used to
simulate the whole cable, this condition is valid for a simplified case where a non contact
behavior is assumed for the structure. The cable geometry is set as a helix and characterized
by its center line.
Parameterizing the curve, as Labrouse and Nawrocki (2000), to obtain the mathematical
representation of the wire as

( )

( ) (1)

where is the rod radius, is the helix angle and is the position on the helix length. With
this parameterization it is possible to determine any point in the curve that represents the
strand laid helically around a straight core.
It is also important to define the Frenet-Serret triad that defines the kinematic
properties of a particle moving along the curve. Three components can be obtained from this
theory considering only the dependence on the arc length: the tangent, normal and binormal

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
 Geiger, F. P., Marczak, R.

vectors. The first one indicates the tangent direction of the point in the curve, the second
points at the direction of the curvature radius and the latter is a cross product of the other two.
Consequently these are orthogonal to each other (Kraus, 1967).
Using the equation found in Feyrer (2007) and the description of a generic space curve
(r), where the position vector depends only on the length along the wire (S), to determine the
tangent vector and the parameterization in Eq. (1) it is possible to find

( )

( ) (2)

[ ]

and the derivative of the triad unit vectors leads to (Ramsey, 1988)

[ ] [ ][ ] (3)

where n and b denotes the normal and binormal vectors, respectively, and κ and τ represents
the curvature and torsion of the space curve. Hence, the normal vector can be determined by
solving the first equation of the system in Eq. (3) and as the triad is an orthogonal set of
vector the binormal direction can be obtained by the cross product of t and n. Therefore, this
triad represents a local reference system and can be used to study any point of the wire
individually.
Using the parameterization it is possible to obtain κ, which express the tangent vector
change rate of a point moving along the curve, and τ, that indicate the twist of the same point.
Using the Eq. 1 and the relations in Eq. 2 it is possible to define both parameters as

(4)

So, the triad can be expressed now as a function depending only on the length of the
wire, as the normal and binormal vectors written as

( )
⁄
( ) (5)

[ ]

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
Bending analysis of a wire strand analytic model

( )

( )

[ ]

Using the spatial beam theory, developed by Love (1944), to represent the wire by its
center line, where the equilibrium equations are written as

(6)

where and are the load and distributed load, respectively, in each direction. For the
equilibrium of moments, the system is

(7)

where and are the moment and distributed moment, respectively, in each direction.
Therefore, the response of the cable due to the loads applied can be calculated.
Solving both systems (Eq. 6 and 7) it is possible to obtain the cable mechanical behavior.
Considering that the wire is clamped at one end, the boundary conditions, in order to
determine the 6 integration constants, can be used in the solutions. Since the 5 parameters are
know (t, n, b, and ) one can describe the wire geometrically at any point.
To do so, the loads applied in the structure are decomposed into the local system,
using the triad as the transformation matrix written as

[ ] (8)

so, each component of the force is evaluated in any point of the helix. This procedure is
reversible, that means, the Frenet-Serret formulas makes it possible to transform any
information, regarding the local system, back to the global system, and vice-versa. All the

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
 Geiger, F. P., Marczak, R.

study developed here is focused on the global analysis, hence the diagrams are presented
regarding the fixed system in the center of the cable.
The triad is used to transform not only the force vector but also the moments. So the
position vector (r) determines the distance for the moments and the cross product aligns the
boundary condition to the triad.

3 RESULTS

12
Mx Num
My Num
10
Mz Num
Mx Ana
8
My Ana
Mz Ana
Moments [Nm]

-2
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 1 – (a) Case 1 studied, Vx parallel to the x1 direction and (b) moments diagram.

Using the formulation proposed, one turn and one wire of the cable was simulated.
This methodology provided results that can be used to characterize the role structure. Setting
the helix angle as 75º and the radius 50mm the maximum value of S is calculated, due to the
parameterization (Eq. 1), as . With the geometry defined, the loads were changed
and the moments diagram, due to bending, analyzed. Six cases of bending were studied. The
first case is a force applied in one end of the wire, the other is clamped in all the cases, in the
x1 direction as shown in Fig. 1(a).
The moments diagram (Fig. 1(b)) presents the numerical solution ( , and
) and the model results ( , and ). Initially, at the clamped end
( ), the moment in the x2 direction is maximum and is evaluated as , the
linear behavior presented by this component was already expected. The moment has the
maximum equals, in modulus ( ), to the minimum value at and
, respectively. As the load applied is parallel to the x1 direction, there is no moment
generated in that component in any point, therefore, .

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
Bending analysis of a wire strand analytic model

-2

Moments [Nm]
-4

-6
Mx Num
My Num
-8
Nz Num
Mx Ana
-10 My Ana
Mz Ana
-12
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 2 - Case 2, applied parallelly to the direction and (b) the diagram of moments.

In order to analyze the quality of results, a FEM model was developed, using a
commercial software, and the responses compared. For the first case, all the curves are
coincident to the numerical points generated by the finite element analysis. This result can be
used to validate the model.
For the second case, Fig. X(a), the load applied is now parallel to the x2 direction and
set again as . Once more the bending maximum value is , it is also
linear, but is negative, due to the notation used. For and it is possible to note that the
curves have a similar behavior as and in the first case, because of the same reasons.
When two loads are applied to the wire, and , as demonstrated in Fig. 3(a), the
bending behavior has the same features as cases one and two. This is not only expected but
also indicates that the methodology used is respecting the superposition effect, since the
system solved is linear and must present this characteristic. Adding the responses generated
by each load alone provide the same result when both loads are analyzed together.
It is interesting to observe that the moment in the x3 direction is generated by the
resultant of the loads, indicated as V. Therefore the maximum moment occurs not at one of
the axis, but in the middle. The higher value is set at 135° while the angle where it is zero is
270°, those points are at the vector V direction (no distance) and parallel to it (top distance),
respectively.
Case 4, in Fig. 4, demonstrates that due to the component is no longer linear.
Once again the superposition is present. Considering the first case, since does not produce
moment in the direction, adding this load only generates effects in and . In other
words, does not change when is taken into account.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
 Geiger, F. P., Marczak, R.

15

10

Moments [Nm]
0

Mx Num
-5 My Num
Mz Num
Mx Ana
-10
My Ana
Mz Ana
-15
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 3 - Case 3 with and applied parallelly to the plane and (b) the diagram of moments.

12
Mx Num
My Num
10
Mz Num
Mx Ana
8
My Ana
Mz Ana
Moments [Nm]

-2
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 4 - Case 4, and applied parallelly to the plane and (b) the moments diagram.

For the fifth case where and are applied, see Fig. 5, the component is not
linear for the same reasons of the forth case. Once more, the analysis of two loads were
identical to the addition of separated cases, which demonstrates the linearity of the system.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
Bending analysis of a wire strand analytic model

-2

Moments [Nm]
-4

-6
Mx Num
My Num
-8
Mz Num
Mx Ana
-10 My Ana
Mz Ana
-12
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 5 - Case 5, and applied parallelly to the plane and (b) the moments diagram.

15

10

5
Moments [Nm]

Mx Num
-5 My Num
Mz Num
Mx Ana
-10
My Ana
Mz Ana
-15
0 50 100 150 200 250 300 350
Helix Angle () [°]

(a) (b)

Figure 6 - Case 6, , and applied and (b) the moments diagram.

The last case studied is an application of loads in all the directions. As expected there
are two points where the moment is maximum and they are equal in modulus because the
loads and distances are the same. Both and are set as and have the same
behavior (nonlinear) as cases 5 and 4, respectively. The other component ( ) has the same
result as in case 3, reaching its maximum value at 135° as for the same reasons.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
 Geiger, F. P., Marczak, R.

Therefore, it is possible to realize that this last case is a combination of the other cases and
that the methodology is coherent to the expected. In fact, the application of this procedure
leads to a linear system.
For all the cases the comparison between the analytic model and the numerical
analysis were presented. The methodology used to generate the results was validated by the
Fem model. No significant difference was found in any case analyzed when the solutions
were compared.

4 CONCLUSIONS

An analytical expression to study the mechanical behavior of a cable, subjected to

bending, was obtained by solving the spatial beam theory. The wire was modeled as a helix
and a parameterization was used to describe the cable geometrically. As the results of one
wire can be used to study the whole cable, since the responses are similar, only one
component was analyzed.
Using the transformation matrix to decompose the loads and moments, the boundary
conditions were used to determine the integration constants. The results obtained in the local
system were transformed back to the global system. Six cases of loading were analyzed, using
the methodology applied, and compared to a FEM model. The differences between the results
were not significant and the theory developed was validated.
Another feature of the theory was the superposition. It was expected, since the system
is linear, that this theory could be used to analyze any loading case. The third case showed
that the methodology applied could generate the same result as the first two cases combined.
The other cases also presented this characteristic.

REFERENCES
Ansys, 2006. Release 11.0 Documentation. Contact Technology Guide: ANSYS Inc.
Argatov, I., 2011. “Response of a wire rope strand to axial and torsional loads: Asymtotic
modeling of the effect of interwire contact deformations”. International Journal of Solids and
Structures, Vol. 48, p. 1413.
Costelo, A.G., 1990. Theory of Wire Rope. Springer, London, 1st edition.
Elata, D., Eshkenazy, R. & Weiss, M.P., 2003. ”The mechanical behavior of a wire rope with
an independent wire rope core”.International Journal of Solids and Structures, Vol. 41, p.
1157.
Feyrer, 2007. Wire Ropes: Tension, Endurance, Reliability. Sringer, Stuttgart, 1st edition.
Ghroreishi, S. R., Davies, P., Cartraud, P. & Messager T., 2006. ”Analytical modeling of
synthetic fiber ropes. Part I: A new continuum model for multilayered fibrous structures”.
International Journal of Solids and Structures, Vol. 44, p. 2924.
Jiang
Jolicoeur, C. & Cardou, A., 1996. “A semicontinuous mathematical model for bending of
multylayered wire strands”. Jounal of Engineering Mechanics, Vol. 122, p. 643.
Kraus, H. , 1967. Thin Elastic Shells: An Introduction to the Theoretical Foundation and the
Analysis of Their Static and Dynamic Behavior. Wiley & Sons, New York, 1st edition.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015
Bending analysis of a wire strand analytic model

Labrosse, M. & Nawrocki, A., 2000. “Frictional dissipation in axially loaded simple straight
strands”. Journal of Engineering Mechanics, Vol. 126, p. 641.
Love, A. E. H., 1944. A Treatise on the Mathematical Theory of Elasticity. Dover
Publications, New York, 4th edition.
Machida, S. & Durelli, A. J., 1973. “Response of a strand to axial and torsional
displacements”. Journal of Mechanical Engineering, Vol. 15, p. 241.
Maple, 1999. Release 5.3 documentation. Contact Technology Guide: Maplesoft..
Nawrocki, A. & Labrosse, M., 2000. “A finite element model for simple straight wire rope
strands”. Computers and Structures, Vol. 77, p. 345.
Østergaard, N. H., Lyckegaard, A. & Andreasen, J. H., 2011. . “A A method for prediction of
the equilibrium state of a long and slender wire on a frictionless toroid applied for analysis of
flexible pipe structures”. Engeneering Structures, Vol. 34, p. 391.

CILAMCE 2015
Proceedings of the XXXVI Iberian Latin-American Congress on Computational Methods in Engineering
Ney Augusto Dumont (Editor), ABMEC, Rio de Janeiro, RJ, Brazil, November 22-25, 2015