TECHNIQUES

by H. Usabiaga, M. Ezkurra, M.A. Madoz, and J.M. Pagalday

EXPERIMENTAL TEST FOR MEASURING THE NORMAL

AND TANGENTIAL LINE CONTACT PRESSURE
BETWEEN WIRE ROPE AND SHEAVES

igure 1 depicts schematically how a traction elevator

operates. The system is basically composed of an elevator car (1), a counterweight (2), a traction sheave or
drive (3), and some ropes (4) that are disposed in parallel and that join the car with the counterweight passing
through the traction drive.
In this kind of elevator, it is crucial to ensure an appropriate
adherence between the traction sheave and the hoisting ropes
because if sufficient adherence is not guaranteed, the rope
could start bodily slipping. This may cause the falling of the
elevator car or of the counterweight to the pit floor of the
hoistway, involving significant property damages as well as
loss of human life.
Traditionally, conventional elevators are dimensioned in
order to ensure that between rope and sheave enough adherence region always remains during duty service. This adherence region represents a safeguard against bodily slipping
between rope and sheave as was introduced by Schulz.1 Even
so, the present tendencies in the lifting sector lean toward the
reduction of the traction sheave diameter since it, in turn,
reduces the required nominal torque of the drive (note that
the nominal torque can be written in a simplified manner as
2max T1 2 T2

; see Fig. 1).
D
With a lower requirement of torque, a smaller hoisting machine is possible and thereby the drive fits in the hoistway,
and there is no need for additional room in the building in
order to house the drive. These types of elevator solutions
are called machine roomless elevators and they, at present,
represent a competitive advantage in the hoisting sector.
However, the reduction of the ratio between sheave and rope
diameter, D/d, beyond a determined value may entail, on one
hand, a drastic reduction of rope life and on the other, a reduction of the adherence capacity of the hoisting drive (see
Heller2 and Nabijou3) and therefore of safeguard level in the
elevator. Under these circumstances, the sector is demanding
a more precise and accurate comprehension about the mechanical interaction between rope and traction drive in order to
combine a more reduced traction sheave diameter design with
enough safeguard level.
The authors have detected in this sense that little experimental work has addressed rope and traction drive interaction. The
studies made by Wiek4 and later by Haberle5 are probably the
most remarkable, and even in those articles, the study is
focused on a particular operating situation. Therefore, in order
H. Usabiaga is a PhD student in engineering and an assistant researcher and
J.M. Pagalday is an engineer and head of the Mechanical Department at Ikerlan.
S. Coop., Arrasate-Mondragon Gipuzkoa, Spain. M. Ezkurra is an engineer and assistant professor at the Mondragon Unibertsitatea, Arrasate-Mondragon Gipuzkoa,
Spain. M.A. Madoz is a PhD and head of the Mechanical Department at Orona S.
Coop., Hernani Gipuzkoa, Spain.

34

EXPERIMENTAL TECHNIQUES September/October 2008

Downloaded from http://www.elearnica.ir

to acquire a more profound comprehension, and in order to

verify the level of the agreement between testing and the current theoretical models, a new testing procedure is discussed in
this article. In the first part of it, a brief theoretical background
and previous experimental work is introduced. Next, the testing machine and procedures are presented and, finally, some
preliminary results are discussed.

THEORY
RopeSheave Interaction Longitudinal Models
The longitudinal models are bidimensional theoretical models
that in a simplified manner characterize the interaction between the rope and the sheave along the plane that contains
the principal axis of the rope when the latter is bent over the
traction sheave. The first longitudinal model was developed
by Eytelwein6 but should probably be attributed to Euler (see
Chaplin7) with whom Eytelwein worked for some time in
Switzerland. The most popular formula derived from this
model is the one called Capstan formula, which is still nowadays the most widely used expression in order to dimension
traction drives. The Capstan formula:

qc 5

T2
5 emu ;
T1

defines a limiting ratio, qc, between the rope tension, T1 and T2,
at both sides, sides 1 and 2, of D diameter traction sheave. m is
called apparent friction coefficient of the sheave and u is the
total winding angle. Theoretically, if T2/T1 surpass the ratio
defined by Eytelwein, the rope slips bodily through the sheave.
Following the basis of the model, it also derived the normal
line pressure distribution, p(f) (note that this model being
bidimensional, the normal and frictional contact pressure are
not written in terms of [N/m2] but in terms of [N/m]), as well
as the rope tension, T(f), for any angle f, which denotes a
specific point of the winding arc.
Tf 5 T1 emf :

2T1 emf
:
D

pf 5

On one hand, note that Eqs. 2 and 3 show that the two variables obey an exponential distribution. On the other, note that
Eq. 1 suggests there is not any dependence of the D/d in the
drives limiting T2/T1 ratio. However, as was mentioned
before, the influence of the ratio D/d in the critical imbalance
ratio, qc, has been reported (e.g., Nabijou3).
doi: 10.1111/j.1747-1567.2007.00294.x
2007, Society for Experimental Mechanics

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

Fig. 1: Main components of an elevator hoisting system

Heller2 developed an equivalent expression to the Eytelweins

formula which shows a diameter ratio dependence. The most
significant difference between the Eytelwein and the Heller
models is that the former only considers the axial forces in the
rope, whereas the latter, besides considering the axial forces,
also considers radial shear forces of the rope.
The expression for T(f) derived by Heller is represented in
Eq. 4.
Tf 5 T2 e 2 Bf




B
sinh Cf 1 cosh Cf ;
C

where
D

11
2m

r
D
B2 1 :
d

B5d
and
C5

Note that, as was mentioned before, Hellers formula is dependent on the ratio D/d since B and C depend on this ratio.
Besides this, note that Hellers approach for tension gives as
well a nearly exponential distribution.
Figure 2 depicts the critical imbalance ratio, qc, estimated by
the Heller and Eytelwein models against D/d for three different apparent friction values. As illustrated in figure, the two

models lead to similar results if D/d is high. However, when

this ratio decreases, according to Heller, the shear internal
forces cannot be neglected and, as a result, Eytelweins model
diverges significantly from Hellers.

Rope and Sheave Slip Stages

The previous two expressions consider only the situation
where the sheave is about to slip bodily through the sheave.
Therefore, this situation implies theoretically the maximum
ratio at which the traction drives could operate. However, the
traction drive system must always operate under lower T2/T1
ratios. The literature describes three different interaction
stages as a function of the applied imbalance ratio, that is,
q 5 T2/T1. These three stages are schematically represented
in Fig. 3.
Following the same assumption than Eytelwein made in his
model, it can be derived that when an equal load, T1 5 T2 5 T0,
is applied at both ends of the rope, a constant contact pressure
of value
p0 5

2T0
D

as well as constant tension act respectively along the contact

arc and along the rope (see Fig. 3a). However, this uniform
situation is not initially attained in practice during the loading. Some sheave rotation is required in order to distribute the
uniform tension though the whole winding arc as was shown
in Usabiaga et al.8 Nabijou3 named the final uniform situation
nonslip stage (NSS) because, under this stage, no slip occurs
between the rope and sheave.
September/October 2008 EXPERIMENTAL TECHNIQUES

35

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

Imbalance ratio (q)

d=7.2mm
Heller =0.2
Eytelwein =0.2

Heller =0.4
Eytelwein =0.4
Heller =0.6

Eytelwein =0.6

1
10

20

30

40

50

60

70

80

Diameters ratio (D/d)

Fig. 2: Critical imbalance ratio, qc, versus diameters ratio, D/d, according to Eytelwein6 and Heller2
If we start reducing the load at one end of the rope, an imbalance between both ends of the rope is prescribed. When this
happens, part of the contact arc still remains under the NSS
situation. However, on the remaining arc length, slipping as
well as friction force start arising between the rope and
sheave, leading, in this manner and as it is depicted in Fig. 3,
to exponential rope tension and normal contact pressure profile. This stage is called partial slip stage (PSS).
When we keep increasing the applied imbalance ratio, the arc
with exponential profile contact pressure progressively
enlarges and, accordingly, the arc length that exhibits a constant profile arc length reduces. The increase of imbalance
ratio finally leads to a fully sliding contact arc and, accordingly, to a fully exponential normal contact line pressure and
rope tension distribution. This limiting ratio could be estimated, as mentioned before, by means of Capstan formula
and the associated situation is called full slip state (FSS)

(a) NSS

Besides these stages, another remarkable feature is the

sheave turning sense influence in the placement of the adherence contact arc. This feature is evidenced for example in
Schulz.1 After the sheave has turned in one sense for some
time in a PSS situation, the placement of the adherence contact arc seems to depend exclusively on the sheave turning
sense. According to Schulz, if the sheave has been turning left
at constant speed, the constant pressure segment is at the
right side of the sheave (see Fig. 3) and, on the contrary, if
it has been turning right, the constant pressure places on the
opposite side, namely on the left. Theoretically, it can be demonstrated that the opposite statement can never happen using
the reduction ad absurdum as Johnson9 made for similar
issues.

(b) PSS

Fig. 3: Slip stages

36

(see Fig. 3c). Finally, a greater decrease in load at one end

produces a bodily slipping of rope around the sheave, promoting a very dangerous situation for the traction drive systems.

EXPERIMENTAL TECHNIQUES September/October 2008

(c) FSS

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT
Finally, the authors (see Usabiaga et al.8) have recently developed a numerical model that is mainly based on Eytelweins
assumption but allows as well to analyze the mechanical conditions along the traction drive under any slipping stage.
Moreover, the model considers the development in time of
the mechanical conditions when the sheave turning sense
changes or when the load at one or both ends varies. Note that
these two situations are very common in elevator suspension
systems.
In particular, the results presented in that paper demonstrate
that when the sheave is evenly loaded, a uniform contact
pressure and rope tension are not initially achieved. Instead,
two symmetrical exponential distributions are attained for
line tension and normal contact line pressure. However, the
model also demonstrates that when the sheave starts turning, this situation gives way progressively to a completely
uniform line tension and normal contact line pressure.
According to the model, this last uniform situation remains
until load changes at one or both ends. Considering this, when
the objective of the test is to record the contact normal line
pressure under a completely uniform line tension and uniform
normal line contact situation, after installing and loading the
rope and before recording pressure, the sheave must turn at
least an angle equivalent to the total contact arc angle. Note
that this fact has important consequences for the design of the
testing procedure and machine because the testing machine
must allow turning the sheave and running enough rope on
the sheave in order to arrive to the uniform situation.

EXPERIMENTAL PROCEDURE
The key of the procedure introduced here lies in measuring
not the normal and frictional contact pressures, p(f) and f(f),
which may be rather complicated, but the resultant contact
forces along a large contact arc angle. However, the pressures
can be derived from the measured quantities by means of
some mathematical manipulation we will introduce later.
The testing machine developed throughout this article is divided into
x the testing frame and
x the sensing sheave.

Testing Frame Description

The bed frame of the testing machine is depicted in Fig. 4. It
basically consists of what in the elevator sector is described as
2:1 hoisting suspension. Two variable mass counterweights
(1,2) are used, in order to prescribe an average rope tension
as well as an imbalance ratio between both ends of the rope.
Note that the selected suspension scheme provides a gain of
two in the ratio between the stroke length of the rope and the
stroke length of the counterweights, so that a ls length of
counterweight stroke involves 2ls length of rope running the
sheave. This type of hoisting suspension is found particularly
useful when the stroke of the counterweight is limited. Note
that when the counterweights start and end a stroke, the
sheave must change the turning sense and therefore, for some
time in each stroke, the sheave must respectively accelerate
and decelerate. If the interest of the test lies in recording the
interaction conditions under constant sheave turning speed,
and if the counterweight stroke is short, the length of rope

Fig. 4: Bed frame scheme

stroke that runs the sheave under constant speed could be too
short if a 1:1 suspension system is considered. A 2:1 suspension system could represent an alternative solution in those
cases.
The two counterweights are guided by two vertical rails (7),
as is illustrated in Fig. 4. Special care has been taken in
order to minimize the friction between the rails and the shoe
guides (8) installed at the counterweights. Note that this
friction generates additional imbalance force in the hoisting
system; therefore, it must be minimized as far as possible. To
this end, very low friction glide shoes which are commonly
September/October 2008 EXPERIMENTAL TECHNIQUES

37

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

used at the elevator sector have been selected for guiding the
counterweight along the rails. The stroke of the counterweights is 12.5 m, which implies that the stroke of the rope
(4) along the driving sheave (1) is 25 m. In order to turn the
sheave, a permanent magnet synchronous motor was
selected. This machine together with an appropriate controller can govern the turning speed of the traction sheave as
required.

Instrumented Sheave
The main goal of the experiment consists in measuring the
resultant normal and tangential force acting onto an arc portion of the sheave. Properly postprocessed, this value can be
converted to its respective local value, namely p(f) and f(f).
Based upon this principle, a classical elevator cast iron driving sheave (1) with a nominal diameter of 0.250 m and 7.5-mm
diameter U-type groove was selected for the experiment (see
Fig. 5). A 158 and 40-mm deep sheave portion was machined
with an electrical discharge machining device taking special
precaution in order to preserve the cut part.
After this, a triaxial piezoelectric load cell (reference: PCB
260A01, PCB Piezotronics, Inc., Depew, NY) was selected
for measuring the applied load on the removed part. These
transducers are ceramic made and therefore very stiff. The
decision to select a very stiff transducer was made after running several finite-element methodbased calculations of the
assembly and afterward concluding that any displacement of
the transducer or the detached part could generate undesired
influence on the measured magnitudes. The selection of
a ceramic transducer should avoid this.
Figure 6 depicts the final assembly of the sensing sheave. The
transducer (2) was placed in the cavity left for this purpose in
the sheave. Above it, the grooved side of the removed part,
which from here on we will call detachable part (3), was
mounted again. Note that this part had to be carefully
machined in order to take out from it the same volume is
now occupied by the transducer.

The transducer together with the detachable part were fixed

to the sheave with a CopperBeryllium screw rod (4) which is
also very stiff and therefore should prevent any disturbance of
the normal and tangential contact pressure distribution. The
fixing nut (5) prestresses the screw rod, avoiding in this manner any nonlinear behavior between forces and relative displacements of the sensing sheave parts.
It should be remarked that a very strong effort was made
to place the demountable part of the sheave on the same
level as the rest of the sheave surface. High-definition geometric measuring machines were used and corrective measures were taken until the groove was placed according to
jD1 2 D3 j < 0:005 mm and jD2 2 D4 j < 0:005 mm3 (see Fig. 6).
The load cell signal was finally redirected to an electronic
signal conditioning and amplifier device and stored later on
disk.

The Leakage Effects of the Transducer

Piezoelectric transducers usually show a kind of leakage effect under static and nearly static situations. In this procedure, this phenomenon was successfully compensated based
upon the principle that the leakage variation in time, namely
the variation in time of the difference between the compensated, Qc, and the transducer signal, Qt, was proportional to
the transducers signal, Qt. That is to say:
dQc 2 Qt
5 cQt :
dt

The constant c was calculated by the least square method

adjusting the signal generated by the transducer under a
time-independent load with respect to the solution of the differential equation shown in Eq. 8.

Calibration
The transducer x- and y-axis signal was calibrated (see Fig. 6)
in order to estimate respectively the resultant normal and
frictional contact forces applied in the detachable part. For
that purpose, successive precisely calibrated weights were
suspended in the mid-length groove of the detachable part
respectively in x- and y-axis directions. The linearity in both
axes was very remarkable, Rx 5 0.988 and Ry 5 0.995. The
calibration also showed that the cross talk between normal
and tangential components was not significant.

Conversion to Local Normal and Tangential Pressure

As it is depicted for the normal case in Fig. 7, the transducer
measures in every instant i not the pressure, p(f) or f(f), but
the forces P,
fi 1

Pfi 5

Dut
2

pfdf;

Du
fi 2 2 t

Fig. 5: Snapshot of the instrumented sheave installed in the

testing frame

38

EXPERIMENTAL TECHNIQUES September/October 2008

and F, which are the resultant forces generated by the former normal and tangential pressures applied on the transducer arc portion of angle ut. However, local values, that is,
p(f) and f(f), are more meaningful variables for describing

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

Fig. 6: Final sensing sheave assembly

the experimental conditions along the contact arc. Therefore,
in the following, a procedure is introduced in order to compute pressure values from the transducer-measured resultant forces.
In this sense, note that the operation that performs the transducer when it collects the resultant forces along the whole
contact arc can be reinterpreted as a mathematical operation called convolution between the local pressure function,
namely, p(f) and f(f), and a particular transfer function.
Mathematically the convolution is written as follows:

hgt 5 ht 2 tgtdt;

10

where hg represents, for this particular case, the resultant

force measured by the transducer, h; the local pressure; and
g, the test function.
The test function can be computed bearing in mind
x the measuring rate of the signal acquisition device,
x the arc length that occupies the detachable part, and
x the prescribed turning speed of the sheave.
The inverse of this mathematical operation is called deconvolution and can be used together with the previously introduced transfer function in order to invert the process, so
that the pressures can be computed from the resultant forces
recorded by the transducer.

Tested Ropes and Testing Conditions

Regarding the tested ropes, two elevator rope constructions
were considered:
x a 7.2 2 8 3 19S-FC (the designation of wire rope follows
ISO17893:2003) rope
x a polyurethane (PU)-jacketed 6.5-mm diameter 4.8 2 6
3 19S wire strand core (WSC) rope.

The comparison between jacketed and conventional rope

should contribute to determining if the literature hypotheses
are also acceptable in order to model jacketed ropes.
Regarding the imbalance of the prescribed load, during the
test, the imbalance ratio was varied. For the conventional
rope, one counterweight was load with 2000 N, whereas the
second counterweight was load with 2000 N, 1800 N, 1660 N,
and 1560 N, applying in this manner respectively an imbalance ratio of q 5 1, q 5 1.1, q 5 1.2, and q 5 1.3.
Due to the higher adherence friction coefficients between the
jacket and the sheave, higher imbalance ratios were used with
this second rope construction. In that case, one counterweight
was loaded with 2000 N, whereas the second was loaded with
2000 N, 1660 N, 1430 N, and 1250 N, applying this time an
imbalance ratio of q 5 1, q 5 1.2, q 5 1.4, and q 5 1.6.
Finally, regarding the sheave-turning speed, all the tests
were carried out under constant 0.8rad/s sheave turning
speed. The relatively low speed was selected keeping in mind
that quasistatic hypothesis are considered by Eytelwein. The
selected speed together with keeping sheave speed constant
should prevent any significant influence of inertial forces on
the traction drive interactions and should help make a much
clearer comparison between theory and tests.

RESULTS
Test Carried Out with Conventional Rope
Figure 8 shows some normal contact pressure distribution
results for q 5 1, q 5 1.1, q 5 1.2, and q 5 1.3 imbalance
ratios. Not including the first peaks that arise during the
running off and winding region (this phenomena will be discussed later in Normal Pressure Peaks Near Winding and
Running Off Points section), the test we carried out shows
that for q 5 1 (see Fig. 8), a uniform constant pressure distribution is attained along the sheave contact arc. However,
September/October 2008 EXPERIMENTAL TECHNIQUES

39

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

Fig. 7: Conversion to local quantities

when the applied imbalance ratio increases progressively, a
nearly exponential-like distribution starts arising on the right
side of the sheave according to what Schulz pointed out in his
article about sheave-turning sense influence on the adherence
region placement. Finally, q 5 1.1 and q 5 1.2 show that the
constant pressure gives way progressively to an exponential
pressure distribution until the highest imbalance ratio, q 5
1.3, shows a nearly FSS situation.

Tests Carried Out with PU-Jacketed Rope

Some disagreements between the theoretical model and the
experimental test arose with the jacketed rope test. In this
case, even in a balanced situation, the experimental measures
show not a constant distribution but an increase of pressure
toward the middle arc of the sheave (see Fig. 9a).
For the conventional rope case, the friction force between rope
and sheave should be the responsible of any rope tension
change. For the jacketed rope case, the friction distributions
shows that friction force is distributed, in the case of q 5 1,
evenly symmetrical and oriented from the outer limits of the

winding arc toward the center. Therefore, the recorded friction force supports the idea that tension is not constant and
that it increases toward the center of the contact arc in the
carried out tests.
In the case of applying an imbalance between wire rope and
sheave, a similar behavior is attained. However, this time the
distribution is not symmetric. Note that this lack of symmetry
is required, since the resultant friction between sheave and
rope must balance the prescribed imbalance.

Normal Pressure Peaks Near Winding and

Running Off Points
In all the tests that we carried out, near the winding on and
running off points, the pressure peaks always rise diverging
from what was expected from the theoretical point of view.
Wiek4 and later Haberle5 describe the same peaks in the tests
that they carried out.
As it is depicted in Fig. 10a, due to its bending stiffness, the
ropes need certain transition length before taking the same

25

p0

Normal line pressure (p)[N/mm]

20

p0

15

10

D=200mm
U groove (dg=7.5mm)
q= 1.0
q = 1.1
q=1.2
q= 1.3

60

80

100

120

140

160

180

200

220

240

260

Winding angle ( ) [degrees]

Fig. 8: Normal line contact pressure, p(f), versus sheave winding angle, f, for q 5 1, q 5 1.1, q 5 1.2, and q 5 1.3 with
conventional rope

40

EXPERIMENTAL TECHNIQUES September/October 2008

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

Normal line pressure (p) [N/mm]

20

p0

20

p0

15

15
4.8619 SWSC (jacketed dj=6.5mm)
D=200mm
U groove(dg=7.5mm)

10

10

Normal line pressure

Tangential line pressure

5
50

50

100

150

200

Tangential line pressure (f) [N/mm]

25

25

5
250

Winding angle ( ) [degrees]

(a)
30

30
25

20

20
4.8619 SWSC (jacketed dj=6.5mm)

15

15

D=200mm
U groove(dg=7.5mm)

10

10

Normal line pressure

Tangential line pressure

5
50

Tangential line pressure (f) [N/mm]

Normal line pressure (p) [N/mm]

25

5
0

50

100

150

200

250

Winding angle ( ) [degrees]

(b)
Fig. 9: Normal and tangential normal line contact pressure, p(f) and f(f), versus winding angle, f, for (a) q 5 1 and (b) q 5 1.6
with jacketed rope

radius of curvature as the sheave. This transition curvature

was theoretically estimated by Feyrer.10 As it is depicted in
Fig. 10a, the transition length makes the rope meet the
sheave curvature beyond the point it would, if such transition
length would not exist when winding on and previously when
running off, which creates, in turn, additional moments and
increasing peaks at the external regions of the sheave contact
arc (see Fig. 10b).
Haberle5 has made thorough considerations with respect to
this phenomena. By means of a regression analysis, the

experimental test that he carried out provides expressions

for estimating the peak level as well as the delay and lead
angle, nA and nB, of, respectively, the winding on and running off contact points. Feyrer11 gives some theoretical
expressions as well in order to estimate n and the reaction
force Q.
The test that we carried out shows also a slight winding
angle difference with respect to the theoretical 1808, so the
reason argued by the former authors is consistent in our
tests.
September/October 2008 EXPERIMENTAL TECHNIQUES

41

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

ing theories. The stretch of the rope in the mid-length arc

should not obligatorily mean that the polymer surface in contact with sheave should be as well stretched in the same way.
Therefore, it does not necessary imply slipping between jacket
and sheave.

(a)

(b)

Fig. 10: Source of the apparition of the peaks at the winding

on and running off points of the winding contact arc

DISCUSSION
According to the test performed during this study, the hypothesis assumed by the theory characterize correctly the phenomena, at least under constant sheave-turning speed and
for conventional (not jacketed) ropes.

Thanks to the polyurethane flexibility, the polymer jacket

could be subjected to shear deformation in such a way that
the inner surface of the sheave is deformed according to
the rope deformation (which in turn might be produced
by the effective sheave diameter variation), and on the contrary, the outer part remains in adherence with regards
to the sheave. Note that if the sheave is in adherence with
the jacket, the recorded values are still compatible with
Amontons law and that in this case, the tangential force
we are measuring is strictly the applied shear force in
the jacket.
However, note that the thickness of the jacket is very thin,
0.85 mm. Therefore, there might be reasonable doubts
for this particular case whether the effective diameter
change promoted by the jacket uneven compression along
the winding arc could produce such significant effects in
the interaction.

CONCLUSIONS
In particular, for the q 5 1 case, the results attained in those tests
are very similar from the ones obtained by Haberle.5 In both
cases, the agreement between the experimental value and the
theoretical value estimated from Chaplin7 is very remarkable.
However, the tests performed for the jacketed rope case show
remarkable differences between test and theory. This may
indicate that, for the jacketed rope, Eytelweins or Hellers
hypothesis is not appropriate for measuring the phenomenon.
In this sense, note that, although frictional distribution
matches the normal pressure distribution, suggesting that
rope is not constant, as it is depicted in Figs. 9a and b, the
frictional pressure distribution is not proportional to the normal pressure distribution as would be expected if Amontons
law was fulfilled and sliding between the polymer jacket and
the sheave would occur.
These disagreements may be attributed to the local compression of the polyurethane jacket. Due to the winding and running off peaks, the jacket may be unevenly compressed
between the outer regions and the mid-length region of the
winding arc. Note that this difference of compression exists
also for the conventional rope case; however, the compression
deformation level is not as significant for the jacketed rope
case because the jacket stiffness is remarkably lower than
the stiffness of rope cross-section.
The higher compression of the polyurethane jacket along the
outer region of the winding arc might locally reduce the effective sheave diameter and this in turn might stretch the rope
in the mid length of the winding arc.
The relation between the recorded frictional and the normal
pressure distribution could also be consistent with the exist-

42

EXPERIMENTAL TECHNIQUES September/October 2008

The experimental procedure described in this article is demonstrated to be particularly robust and useful in order to
determine experimentally the interaction conditions
between the rope and the sheave. This is particularly true
for the ease of analysis for certain rope and sheave configurations to determine the effect of the
x prescribed mean load,
x prescribed imbalance load, and
x sheave turning speed
in the mechanical conditions of the hoisting drive.
The carried out preliminary tests show that conventional rope
could exhibit a very close behavior from what was theoretically expected. However, further studies should be carried out
in order to certify the explanation that is given for understanding the disagreements between the theoretical hypothesis and the result we attained.

References
1. Schulz, S., Braking Equipment for Friction Hoists, The
South African Mechanical Engineer 28:426433 (1978).
2. Heller, S.R., The Contact Pressure Between Rope and
Sheave, Naval Engineering Journal 4957 (1970).
3. Nabijou, S., Frictional Behaviour and Fatigue Performance of Wire Ropes Bent Over Small Diameter Sheaves, PhD
Thesis, Imperial College of Science, Technology and Medicine,
London, UK (1990).
4. Wiek, L., The Distribution of the Contact Forces on Steel
Wire Ropes, OIPEEC Bulletin 44:1025 (1982).
5. Haberle, B., Pressung zwischen Seil und Seilrille, PhD
Thesis, Universitat Stuttgart, Stuttgart, Germany (1995).
6. Eytelwein, J.A., Ausgaben grobtentheils aus der angenwandten mathematik, Freidrich Manrer, Berlin, Germany (1793).

SHEAVEWIRE ROPE INTERACTION CONTACT

PRESSURE MEASUREMENT

7. Chaplin, C., Rope Tension in Double Drum Traction

Winches, OIPEEC Conference 2003: Experiences with Ropes,
pp. 201212, Lenzburg, Switzerland (2003).
8. Usabiaga, H., Ezkurra, M., Madoz, M., and Pagalday, J.,
Mechanical Interaction between Wire Rope and Sheaves, Oipeec
Conference 2006: Trends for Rope, pp. 157166, Athens, Greece (2006).

9. Johnson, K.L., Contact Mechanics, Cambridge University

Press, Cambridge, UK (1985).
10. Feyrer, K., Effect of Bending Length on Endurance of Wire
Ropes, Wire World International 23:115119 (1981).
11. Feyrer, K., Wire Ropes: Tension, Endurance, Reliability, 1st
Edition, Springer-Verlag, Berlin, Germany (2006). n

September/October 2008 EXPERIMENTAL TECHNIQUES

43