Sunteți pe pagina 1din 10

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/230644354

Effect of canopy shape on physical load when


holding an umbrella
Article in Applied ergonomics August 2012
DOI: 10.1016/j.apergo.2012.06.003 Source: PubMed

CITATION

READS

472

3 authors:
Lottie Kuijt-Evers

Reinier Knemann

The Hague University of Applied Sciences

TNO

19 PUBLICATIONS 480 CITATIONS

9 PUBLICATIONS 56 CITATIONS

SEE PROFILE

SEE PROFILE

Susan Hallbeck
Mayo Clinic - Rochester
144 PUBLICATIONS 799 CITATIONS
SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,


letting you access and read them immediately.

Available from: Lottie Kuijt-Evers


Retrieved on: 22 August 2016

Applied Ergonomics xxx (2012) 1e9

Contents lists available at SciVerse ScienceDirect

Applied Ergonomics
journal homepage: www.elsevier.com/locate/apergo

Effect of canopy shape on physical load when holding an umbrella


Lottie F.M. Kuijt-Evers a, *, Reinier Knemann a, M. Susan Hallbeck b
a
b

TNO, PO Box 718, NL-2130 AS Hoofddorp, The Netherlands


Mechanical and Materials Engineering, University of Nebraska, W342 NH, Lincoln, NE 68588-0526, USA

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 8 April 2010
Accepted 13 June 2012

The aim of the current study was to investigate the effect of the canopy shape of an umbrella on physical
load when holding the umbrella in different circumstances. For this purpose, thirteen subjects participated in this study and muscle activity of seven muscles of the upper limb (including the forearm) was
measured for 5 wind speeds (4, 5, 6, 7, 8 Bft) and two wind directions (front and side). From the results, it
was seen that for the p50 value of the muscle activity, the umbrella with the asymmetric canopy required
62% and for the p90 value of the muscle activity 74% of the muscle activity, on average, over all wind
speeds e compared to the traditional umbrella. Based on these results, we can conclude that the physical
load of holding the traditional umbrella is signicantly higher than holding the umbrella with the
asymmetric canopy shape in windy conditions.
2012 Elsevier Ltd and The Ergonomics Society. All rights reserved.

Keywords:
Umbrella
Grip force
Upper limb
Muscle activity
%MVC
Physical load

1. Introduction
Umbrellas have been used for thousands of years. In the early
days, they were used as a parasol in warmer climates as protection
against the sun (Sengster, 1855). The umbrella was introduced in
the south of Europe during the Renaissance. In this time period, it
was a status symbol, as pale skin was associated with the upper
class who did not need to work outdoors. In due course, people
started to use the umbrella against the rain in France and somewhat later in England. Around 1775, the umbrella came into general
use in England and as a consequence, there were numerous
improvements. Due to the substitution of silk and gingham for the
heavy oiled silk and the ribs and frames being made much lighter,
the transition to the present portable form was made (Sengster,
1855). Today, one of the problems of using an umbrella in bad
weather is the effect of the wind on the canopy. The problem of the
reversed/inverted canopy and broken umbrella ribs was already
recognized by Koller (1957) who invented and patented one of the
rst wind-resistant reversible/invertable umbrellas. After that
patent, 245 more patents were granted on windproof umbrellas
(www.espacenet.com). Most umbrella canopy designs are fairly
standard, with symmetrical panels connected to an opening
mechanism. Most of the windproof umbrellas differ from the
standard design by employing a two layer canopy. The wind under

* Corresponding author. Tel.: 31 88 86 65 331; fax: 31 88 86 68 774.


E-mail address: lottie.kuijt@tno.nl (L.F.M. Kuijt-Evers).

the canopy can escape between the two layers without inverting
the umbrella.
Recently, a new design for the umbrella canopy shape and
staves/ribs has been patented, which purports to be non-invertable
in a storm (resists inverting at 10 Beaufort (Bft) or 24.5e28.4 m/s)
and easier to hold in windy conditions. Although, the user experiences indicate that the umbrella is easier to hold, it has not yet been
objectively proven. Therefore, the aim of this study is to examine
whether the physical load of the upper limb differs for various wind
directions and speeds between the new canopy shape and a traditional umbrella equipped with the same handle was undertaken.
The physical load when holding an umbrella is the grip force
required to hold the umbrella upright against the elements (wind/
rain) and muscle activity of the upper arm and shoulder to keep the
umbrella in position above ones head in windy circumstances.
Although umbrella holding has neither the duration nor magnitude
of exposure as many repetitive occupational tasks, designing
umbrellas to reduce muscle force demands has the potential to
increase comfort and usability. The decreased demands may also
increase user performance by allowing the user to hold the canopy
over their head during high winds. The relationship between
comfort/discomfort and gripping force is shown by a recent study
of Kong et al. (2012) who found high coefcients of determination
(R2) of gripping force on comfort/discomfort. These coefcients of
determination varied from .975 to .996, depending on the rating
scale. Another advantage is that people with reduced physical
capacity (e.g., weak or elderly) are able to hold the umbrella more
easily when less muscle activity is needed.

0003-6870/$ e see front matter 2012 Elsevier Ltd and The Ergonomics Society. All rights reserved.
http://dx.doi.org/10.1016/j.apergo.2012.06.003

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

Grip force has been measured on hand dynamometers and


with force sensing resistors or imbedded load cells during many
tasks such as cutting or using a screwdriver or nutrunner (Kong
et al., 2008; Lin et al., 2006; McGorry and Lin, 2007; Morse
et al., 2006). The disadvantage of these techniques is that either
the measurement equipment cannot be applied to an umbrella or
the handle has to be instrumented. Another way to indicate the
physical load indirectly (without affecting the umbrella), is to
measure the muscle activity of the forearm muscles (Hoozemans
and van Dien, 2005; Duque et al., 1995). Hoozemans and van
Dien (2005) measured EMG of 6 forearm muscles (m. extensor
carpi radialis longus, m. extensor carpi radialis brevis, m. extensor
digitorum, m. extensor carpi ulnaris, m. exor digitorum supercialis, and m. exor carpi radialis). In a static situation (like
holding a handle), the normalized EMG of these forearm muscles
gives a reliable prediction of the grip force (Hoozemans and van
Dien, 2005). Moreover, when at least 3 of these muscles are
assessed, a sufcient level of validity of the grip force prediction is
arrived; thus, it seems to be irrelevant which 3 of those 6 forearm
muscles is assessed (Hoozemans and van Dien, 2005). In a later
study Hoozemans et al. (2006) showed that even the location of
the electrodes on the forearm, does not affect the prediction of the
grip force by EMG, if the wrist is in a neutral and static position
and EMG is recorded using 4 pairs of electrodes. Hence, EMG
measurements are reliable method to assess grip force when
holding an umbrella, as this is a static task with the wrist in
neutral position.
The aim of the current study was to investigate whether the
muscle activity of the upper limb and the forearm muscles differs
between holding the umbrella with the new canopy shape and the
umbrella with the traditional shape for ve wind speeds and two
wind directions, in order to indicate the difference in physical load
between the umbrella with the new canopy shape and the traditional shaped umbrella. Before the study is described a biomechanics analysis is performed to establish from which muscles EMG
should be measured.

2. Biomechanics analysis
A situational sketch of a person holding an umbrella is shown in
Fig. 1a. The forces on the umbrella are illustrated in the Free Body
Diagram in Fig. 1b. The force caused by the wind velocity is the
aerodynamic force. This force can be resolved into two components: The drag force that is the component parallel to the direction
of the relative motion (i.e. the wind velocity) and the lift force that
is the component perpendicular on the relative motion (i.e. the
wind velocity). Furthermore, the gravity force applies on the
umbrella. In order to keep the umbrella in the same position, the
hand of the person holding the umbrella, should compensate for
these forces in the opposite direction.
The drag force is proportional to the relative motion velocity (i.e.
in our case the wind velocity as the umbrella does not move), the
frontal area, the air density and the drag coefcient. The drag
coefcient depends on the shape, the material and the surface
characteristics of the canopy. The lift force is proportional to the
relative motion velocity, the air density, the lift coefcient of the
angle of attack, and the planform area (i.e. the area of shape of the
umbrella viewed from above; Fig. 2).
The lift coefcient depends on the angle of attack and is
therefore caused by the direction in which the umbrella is held. The
main difference between the tested umbrellas is the shape of the
canopy, which affects both the planform area and the drag coefcient. These differences are responsible for the differences in
physical load when holding the umbrellas.
The magnitude of the force applied by the hand on the handle,
depends on the magnitude of the aerodynamics forces and the
gravity force. The direction of the force applied by the hand in the
vertical direction depends on the direction of the net force of the lift
force and the gravity force. When the lift force exceeds the gravity
force, the resultant force will direct upwards, when the lift force is
smaller than the gravity force, the resultant force will pull downwards. The horizontal force of the hand on the handle will always
be directed forward (opposite to the draft force). The magnitude of

Fig. 1. (a) Situational sketch and (b) free body diagram of the umbrella in the wind.

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

Fig. 2. Traditional (right) and asymmetrical (left) umbrella with their planform area.

the moment in the contact point of the hand and handle (Fig. 3a) is
directed in opposite direction of the net moment caused by the
aerodynamics forces and the gravity force, as the umbrella is held in
the same (static) position. It depends in the way the umbrella is

held whether this moment is directed to the left or the right. When
the umbrella is held slightly backwards, the direction of the rotational moment caused by the resultant aerodynamics force and the
gravity force is to the right. Hence the moment in the contact point

Fig. 3. (a) Large moment in the contact point of the hand (Mh), as both the gravity force and the resultant force cause a rotational moment in opposite direction. (b) The moment in
the contact point (Mh) is smaller due to the fact that the gravity force causes a rotational moment opposite to the resultant of the aerodynamics force.

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

is turned to the left (away from the body). However, when the
umbrella is held forwards, the direction of the rotational moment
by the gravity force is opposite to the direction of the rotational
moment caused by the aerodynamics forces (Fig. 3b). In this way,
changing the orientation of the umbrella can affect the physical
load, as the moment in the contact point of hand and handle will
become smaller as the moment caused by the gravity force
compensates for the moment caused by the aerodynamics forces.
When the moment caused by the gravity force exceeds the moment
caused by the resultant aerodynamics force, the moment in the
contact point of hand and handle will be to the right (towards the
body). However, in windy circumstances, it is more unlikely that
the moment caused by the gravity force will exceed the moment
caused by the resultant of the aerodynamics force, keeping in mind
that the canopy should be above ones head.
In order to analyze the forces on the body, we have to draw
a new Free Body Diagrams of the upper extremity for both the wind
from the front and aside (Fig. 4).

retroection (extension from the triceps brachii caput longum). The


umbrella does not cause a moment in the top view, as the drag force
is directed through the joints (and the moment arm is zero).
2.2. Side wind
When the wind comes from the side (Fig. 4, side view), the
umbrella causes a pronation in the elbow, which is compensated
for by a (isometric) supination of the arm muscles (like the m.
biceps brachii caput longum). In the shoulder, the umbrella causes
an abductional moment (side view) as well as an internal rotation
(top view). This is compensated for by an isometric adduction and
external rotation (Ms) (e.g., m. triceps brachii caput longum). The
dorsal exion in the wrist (Mw) that compensates the moment the
umbrella causes, will be very small, as the drag force on the
umbrella has a very small moment arm regarding the contact point
h (Fig. 4).
3. Methods

2.1. Frontal wind or headwind


The moment that the umbrella causes on the body (Mu) is
directed opposite to the moment of the hand on the umbrella (Mh)
of the FBD of the umbrella and equal of magnitude. When the wind
is from the front (Fig. 4; side view), the moment of the umbrella on
the hand, causes a radial deviation in the wrist. In order to hold the
umbrella in the same position, the forearm muscles have to
compensate for this by an (isometric) ulnar deviation force exertion
(Mw) (e.g. by the m. extensor carpi ulnaris and m. exor carpi
ulnaris). The elbow extending moment is needed (Me) to
compensate for the bending moment caused by the moment of the
umbrella. The triceps brachii will be activated to achieve this
moment. In the shoulder the umbrella causes an anteexion
(exion), which is compensated by muscle action causing (static)

In this experiment, the asymmetrical shaped umbrella canopy


was compared to a traditional umbrella based on physical load for
ve different wind speeds (4, 5, 6, 7 and 8 Bft) and for two wind
directions (frontal and sideward wind). As indication of the physical
load, muscular activity was measured for 7 muscles of the upper
extremity.
3.1. Participants
Thirteen subjects (4 male, 9 female, age 37 years (SD 13), body
mass 63 kg (SD 8), body height 167 cm (SD 9)) participated in the
experiments. All participants gave informed consent prior to the
experiments and reported no upper extremity complaints or other
muscle complaints.

Fig. 4. Free body diagrams of the upper extremity from top and side view for both frontal and side wind.

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

3.2. Experimental settings and equipment


3.2.1. Umbrellas
Two umbrellas with a similar canopy area (approximately
330 cm2) equipped with the same handle for a consistent grip
(0.45 cm thick neoprene over hard plastic with overall handle
dimensions of 3 cm in diameter and 6 cm in length), as shown in
Fig. 2, were employed for testing. Participants were instructed to
hold the umbrellas in the same upright manner and not let them
rest on their shoulder, etc.
3.2.2. Wind tunnel
A wind tunnel was built using two 6 m long double door cargo
containers, totalling 12 m in length. Wind was simulated by using
a wind generator to blow through the containers. The wind
generator operated at one wind speed. Therefore, wind speed was
varied by the subjects distance from the wind generator (between
5 and 10 m away). Five different wind speeds were dened using
the Beaufort scale, used to dene storms, and translated into
average wind speed (m s1), as shown in Table 1. The wind speed
was measured with a wind metre and the locations of the ve
different wind speeds were marked with a line on the oor. The
wind had a non-uniform ow, with an average wind force of the
mentioned numbers, which is similar to outdoor situations. We did
not use a wind tunnel that is normally used for aerodynamic
measurements, since in wind tunnels, the wind is perfectly laminar,
which is not representative for the real life situation when walking
outside in a storm with an umbrella.
3.2.3. EMG
Muscle activity was measured by means of surface electromyography (EMG, Porti 16/ASD system, TMS, Enschede, NL). Following
suitable skin preparation, bipolar AgeAgCl (M-00-S, MedicoTest
Inc., lstykke, DK) surface electrodes were placed with an interelectrode distance of 25 mm over 7 muscles at the subjects
dominant side: wrist extensors: m. extensor carpi radialis (ECR)
and m. extensor carpi ulnaris (ECU); wrist exors: m. exor digitorum supercialis (FDS) and m. exor carpi ulnaris (FCU); upper
arm muscles: m. triceps brachii caput longum (TCL) and m. biceps
brachii caput longum (BCL); and a shoulder muscle group, m.
deltoideus caput clavicularis (DELT). The electrodes were positioned on the skin in accordance with positions recommended by
Basmajian (1989) in Franssen (1995). The forearm muscles were
measured as they indicate the grip force that is needed to hold the
umbrella. The upper arm and shoulder muscle activity is needed to
hold the umbrella in position and remain the prescribed posture in
stormy circumstances. A reference electrode was placed on C7
spinous process. EMG signals were sampled at 1000 Hz during 10 s
for each wind speed. All EMG signals were band pass ltered
(10e400 Hz) and stored on a computer. EMG signals were full-wave
rectied and low-pass ltered at 5 Hz (fourth order Butterworth)
(Hoozemans and van Dien, 2005) using MATLAB (MathWorks, Inc.,
V 2006). For each of the muscles, the ltered EMG signals were
normalized to the maximum value found in the processed EMG of

the MVC tests. The average of the ECR, ECU, FDS and FCU activity
was also examined, as the EMG of forearm muscles is an indication
of the grip force performed (Hoozemans and van Dien, 2005;
Hoozemans et al., 2006). The amplitude probability distribution
function (APDF) was calculated to determine the 50th and 90th
percentile (p50 and p90) of the EMG signal and employed as
dependent variables for each condition and muscle (group).
3.3. Tasks and procedures
The participants were rst informed about the experiment and
their vital data were recorded. Then they practised the measurement protocol with and without wind (to get accustomed to the
feeling). All participants gave their written informed consent.
EMG sensors were then placed on the skin after the skin was
prepared. For normalization of the EMG signals to the maximum
EMG, the maximum voluntary contraction (MVC) was assessed for
each muscle. During MVC assessment, the participants stood in the
same standardized position that is the body posture like they were
holding the umbrella; with the upper arm straight along the body
and the elbow exed in 90 . The participants performed maximum
voluntary contractions in accordance of the functional characteristics of the muscles for which EMG was recorded. This means, for
instance, that the participants isometrically performed elbow
exion and upper arm anteexion in order to capture the MVC of
the m. biceps brachii caput longum. The subjects exerted the force
against external resistance from the experimenter. Furthermore,
MVC values were also recorded when they applied their maximum
grip force on the umbrella handle. The participants had to build up
to their maximum exertion in 3 s and hold it for 3 s. All MVC
measurements were performed twice. The maximum contraction
measurement of each muscle from the MVC measurements was
used to normalize the EMG signal of the test trials. Therefore, the
EMG output of the umbrella hold test measurements can be
expressed as percentage of the MVC (%MVC, i.e. the normalized
EMG signal).
After the MVC EMG values were recorded for each muscle group,
the test measurements started. The participants held the umbrella
in a standardized position, i.e. standing, upright holding the
umbrella in their dominant hand with the elbow exed (about 90
included angle) as if it were raining. They started at their rst
position (wind force 4 Bft) looking towards the wind generator
(front) and 10 s of EMG data were recorded. After that, they held the
umbrella in the non-dominant hand to relax the dominant arm for
approximately 3 s. They then again held the umbrella in the
dominant hand and turned their whole body 90 towards the
sideward wind position (dominant hand towards the wind generator). When holding the asymmetric umbrella canopy, the canopy
was kept in the same direction, long axis in line with the wind
direction, as shown in Fig. 5. Again, they held the umbrella in the
dominant hand for 10 s while EMG for the 7 muscle sites was
recorded. After that, they moved towards the wind generator (5 Bft)
facing forward, while holding the umbrella in the non-dominant
hand for approximately 3 s, then shifted the umbrella to the

Table 1
Wind speed description.
Beaufort value

Description

Effect

Wind speed equivalent

Distance from the wind generator

4
5
6
7
8
10

Moderate breeze
Fresh breeze
Strong breeze
Near Gale
Gale
Storm

Paper blows from street


Small leafy trees begin to sway
Large branches in motion
Some difculty walking in the wind
Difculty walking
Trees uprooted

5.5e7.9 m/s
8.0e10.7 m/s
10.8e13.8 m/s
13.9e17.1 m/s
17.2e20.7 m/s
24.5e28.4 m/s

10 m
8.5 m
7m
6m
5m
NA

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

Fig. 5. Experimental setting from top view.

dominant hand and held it against the wind for 10 s while EMG was
recorded for all 7 muscle groups. This procedure and the postures
were repeated for both umbrellas over all ve wind speeds and
both wind directions while EMG was recorded for all 7 muscle
groups. The order of the presentation of the umbrellas was
systematically varied among the participants to avoid order effects;
half the participants starting with the traditional and half starting
with the asymmetric canopy.
3.4. Statistical analysis
A full-factorial ANOVA on the means for the 50p and 90p %MVC
values, with blocking on subjects nested within gender, was performed using SAS (V9.1). The independent variables were umbrella
(2 levels), wind speed (5 levels), direction of wind (2 levels), gender
(2 levels) and muscle tested (7 levels). Post-hoc (Tukey) and
simple-effects F-tests were performed on all signicant (p < 0.01)
main effects and interactions of the independent variables with
more than two levels.

(p < 0.0001), direction (p < 0.0001), gender by direction


(p 0.0051), umbrella by muscle (p < 0.0001), and umbrella by
muscle by wind speed (p < 0.0001).
4.1.1. Umbrella
The mean (standard deviation) muscle activity of the asymmetric umbrella shape was 7.0 (5.8)%MVC and for the traditional
umbrella it was 11.3 (8.8)%MVC. Post-hoc (Tukey) tests show that,
overall, the asymmetrical umbrella canopy requires only the 62% of
the muscle activity required for the traditional umbrella over all
wind speeds, a signicant difference (p < 0.01).
4.1.2. Wind speed
The higher the wind speed, the heavier the physical load. The
post-hoc (Tukey) test shows that the wind speeds of 4 and 5 Bft did
not differ but were lower than, 6 and 7 Bft which did not differ from
each other and were lower as a group from and 7and 8 which did
not differ in the magnitude of physical load (%MVC over all muscle
groups) while holding the umbrellas. These data, separated by
umbrella canopy type, are shown in Fig. 6.

4. Results
4.1. 50p of normalized EMG signal
The signicant effects for the dependent variable 50p %MVC
were umbrella (p 0.0035), wind speed (p < 0.0001), muscle

4.1.3. Muscle
The seven muscle groups were loaded differently during this
task. The Tukey groupings for the %MVC for the muscles showed
that the ECU, TCL, FCU, FDS and ECR did not differ, but this group
differed from the Tukey grouping of FCU, FDS, DELT and BCL with

Fig. 6. Asymmetric vs traditional umbrella canopy shape for average p50 and p90 %MVC for all muscles and the grip muscles (ECR, ECU, FDS, FCU) over 10 s by wind speed.

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

this latter group having signicantly lower loading than the


previous muscle grouping. This effect differed by umbrella; therefore, more will be presented in the muscle by umbrella analysis
Section 4.1.6 and Table 3.
4.1.4. Direction
The direction of the wind created a signicantly lower physical
load when coming from the front (mean 8.0, SD 6.2) than from
the side (mean 10.2, SD 8.9). Thus, it required only 78% of the
muscle activity to hold the umbrella facing the wind than when the
wind came from the right side while holding the umbrella in the
right hand.
4.1.5. Gender by direction interaction
The average physical load for males and females by direction is
shown in Table 2. There were 3 women who could not complete the
umbrella-holding task at wind speed 7 and 6 women who couldnt
complete the trial at wind speed 8, all with sideward wind. Only
one participant couldnt hold the umbrella when the wind came
from the front at wind speed 7, but 3 participants were unable with
wind speed 8. This created an interaction of gender by direction due
to unequal male female ratios and the drop out of the weaker
women (higher %MVC in all trials) at higher wind speeds, especially
in the sideward wind condition. The front/side% ratio illustrates the
interaction of gender by wind direction, as this ratio differs
between men (57%) and women (88%).
When the data were analyzed by gender (simple-effects F-test
by gender), there was no effect of direction of wind for males
(p 0.248), but a signicant effect of direction for females
(p 0.0077) showing a drop-out of these women who could not
complete the high wind trials. Therefore, the data were reanalyzed
for wind speeds of 4, 5 and 6 Bft only and for those data there was
not a signicant effect of the direction of the wind for males and
females (p 0.3268).
4.1.6. Umbrella by muscle
The physical load by muscle differed by umbrella canopy shape.
Table 3 shows that the triceps (TCL) was the most loaded muscle for
the traditional umbrella, signicantly higher in %MVC than all other
muscles for that umbrella; however, this was one of the lowest
percentage muscles while using the asymmetric canopy (24% of the
traditional). The deltoid (DELT) was also much lower for the
asymmetric compared to the traditional canopy (20%), and the only
muscle that was signicant when the umbrella was examined by
muscle (p 0.0004). The bicep also was somewhat less active when
employing the asymmetric canopy compared to the traditional
(65%). As you can see, the orders were slightly different between
the two umbrellas for the muscles activation level, but the overall
trend was that the conventional umbrella canopy required higher
activation at the 50p level.
4.1.7. Umbrella by muscle by wind speed interactions
The muscle activity of the 7 muscle grouping locations varied by
umbrella and wind speed. When analyzed by muscle, all the
umbrella  wind speed interactions were signicant (p < 0.001).

Table 2
Interaction of gender and wind direction for the dependent variable p50.
Wind direction

Front
Side
Front/side%

Male

Female

Mean (%MVC)

Std dev

Mean (%MVC)

Std dev

5.6
9.9
57%

4.2
10.1
e

9.1
10.3
88%

6.7
8.2
e

However, due to the great difference in the mean loading of the TCL
(very high) and the DELT (very low) with respect to the grip
muscles, these data were divided into the resisting and handle
holding muscle groups. The muscles of primary interest for
retaining the umbrella in strong wind are the grip muscles (ECR,
ECU, FCR and FCU). Therefore, the activity of these 4 muscles was
averaged and plotted by wind speed and canopy shape, as shown in
Fig. 6. All were signicant (p < 0.0001).
4.2. 90p of normalized EMG signal
The signicant effects for the dependent variable 90p of the
normalized EMG signal were umbrella (p < 0.0001), wind speed
(p < 0.0001), muscle (p < 0.0001), umbrella by muscle (p < 0.0001),
and umbrella by muscle by wind speed (p < 0.0001).
4.2.1. Umbrella
The mean (standard deviation) physical load of the asymmetric
umbrella canopy shape was 25 (18)%MVC and for the traditional
umbrella it was 33 (26)%MVC. Post-hoc (Tukey) tests show that,
overall, the asymmetrical umbrella canopy requires only the 74% of
the force required for the traditional umbrella over all wind speeds.
4.2.2. Wind speed
The higher the wind speed, the heavier the physical load. The
post-hoc (Tukey) test shows that the wind speeds of 4 and 5 Bft did
not differ but were lower than, 6 (middle grouping) and the
grouping of 7 and 8 did not differ in the magnitude of physical load
(%MVC over all muscle groups) while holding the umbrellas. These
data, separated by umbrella canopy type, are shown in Fig. 6.
4.2.3. Muscle
The seven muscle groups reacted differently to the wind load on
the umbrellas. The Tukey groupings for the %MVC for the muscles
showed that the lowest muscle activity was in the DELT, alone in
a grouping with the next higher grouping being FDS, FCU, ECR, BCL
and TCL, with the highest physical load on the grouping of ECU, FDS,
FCU and ECR (the grip muscle groups). Since the muscle effect
differed by umbrella canopy, these results will be presented in the
umbrella by muscle Section 4.2.4.
4.2.4. Umbrella by muscle interaction
The physical load by muscle differed by umbrella canopy shape.
Table 3 shows that the grip muscles (ECU, ECR, FDS and FCU) have
the highest muscular activity compared to the shoulder and upper
arm muscles. The differences between umbrellas, while signicant,
is lower in percentage than the 50p %MVC data and only the bicep
and TCL differed signicantly between the umbrella types
(p 0.0136 and p < 0.0001, respectively). The DELT was the least
loaded muscle group for both umbrellas (Asymmetric/Traditional
(A/T) percentage 82%). The tricep (TCL) was also much lower for
the asymmetric compared to the traditional canopy (52% A/T). The
bicep also was somewhat less active when employing the asymmetric canopy compared to the traditional (58% A/T). The Tukey
groupings also differed for the muscles by umbrella canopy shape,
as shown in Table 3. As you can see, the orders were slightly
different between the two umbrellas for the muscles activation
level, but the overall trend was that the conventional umbrella
canopy required higher activation at the 90p level than the asymmetrical canopy.
4.2.5. Umbrella by muscle by wind speed interactions
The muscle activity of the 7 muscle grouping locations varied by
umbrella and wind speed. When analyzed by muscle, all the
umbrella  wind speed interactions were signicant (p < 0.001).

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

Table 3
Post-hoc analysis of 50p and 90p %MVC, muscle by umbrella.
Muscle

P50

P90

ECU
FCU
FDS
ECR
DELT
BCL
TCL
ECU
ECR
FDS
FCU
BCL
TCL
DELT

Asymmetric

Traditional

Tukey grouping

Mean %MVC

Tukey grouping

A
A
A
A

10.3
8.7
9.8
8.7
2.1
4.8
4.4

C
C
C
C
C

B
B
B

B
B
B
D
D

A
A
A
A
C
C

36.1
32.2
31.6
30.0
20.2
15.4
8.1

However, due to the great difference in the mean loading of the TCL
(very high) and the DELT (very low) with respect to the grip
muscles, these data were divided into the resisting and handle
holding muscle groups. The muscles of primary interest for
retaining the umbrella in strong wind are the grip muscles (ECR,
ECU, FCR and FCU). Therefore, the activity of these 4 muscles was
averaged and plotted by wind speed and canopy shape, as shown in
Fig. 6. These data, while signicant at the 0.01 level, appear fairly
parallel for the grip muscle groups (p < 0.0001).
Overall, seven out of thirteen participants were not able to hold
the traditional umbrella during wind speed 7 and/or 8. They simply
could not hold it above their head. These seven participants were all
female with a mean body height of 162 cm and a mean body weight
60 kg.
5. Discussion
In this study, the effect of canopy shape on the muscle activity of
the upper extremity for different wind speeds was examined. The
tests were performed in a self-built wind tunnel with a wind
generator. This resulted in a non-uniform wind ow, which is more
like the real situation compared to a wind tunnel used for aerodynamics testing with a perfect laminar wind ow. However, the
primary problem for most people holding an umbrella, is a sudden
gust of wind. Wind gusts cause highly dynamic and unanticipated
changes in forces acting on the umbrella and require a quick
response of the person who holds the umbrella. In the current
experiment, it was difcult to simulate that situation by using the
wind generator. Therefore, we choose for a more or less constant
wind speed, with little uctuations. From the results, it was seen
that even with a constant wind ow, some of the people could not
hold the umbrella above their head. This would even be worse, if
that wind speed occurs unexpectedly like in a wind gust.
5.1. Drop of %MVC at wind speed at 7 and 8
The drop in %MVC at wind speed 8 is likely due to six (side) and
three (front) direction trials out of thirteen participants, and at
wind speed 7 due to the three (side) and one (front) trials out of
thirteen participants, who could not hold the umbrella. Hence, only
the strongest participants remained. For them, the %MVC used to
hold the umbrella is lower compared to the weakest group that
could not complete the task. It is remarkable that the p90 %MVC is
just about 50% of the MVC of the grip force muscles and under 50%
on average for all muscles together. The reason that the p90 %MVC
is still around the 50% is due to the fact, that this p90 value is an

% Asym/Trad
Mean %MVC
B
B
B
B
B
A

B
B
B
B
B

A
A
A
A
A
C

12.8
11.2
9.8
9.4
10.5
7.3
18.4

81%
78%
100%
92%
20%
65%
24%

49.2
34.5
37.9
37.2
35.1
29.7
9.8

73%
93%
83%
81%
58%
52%
82%

average over all muscles, during a period of time. If only one muscle
exceeds the 100 %MVC for a split second, the umbrella cannot be
held, whilst the other muscles have lower %MVC. Therefore, the 50p
and 90p values of the %MVC were reduced at wind speeds 7 and 8,
due to drop outs of the weaker participants (and a reduced data
set).
5.2. 50p of normalized EMG
The asymmetric umbrella canopy signicantly reduced the
muscular load for all wind speeds and both wind directions for both
males and females. Specically, the triceps group for the traditional
umbrella canopy required a very high muscular load, especially as
compared to the asymmetric canopy (4.2 times the %MVC). The
traditional canopy shape required 1.6 times more muscular activity
over the seven muscle groups tested at the 50p level, signicantly
higher than the asymmetric umbrella shape. The main muscular
load for both canopy shapes was in the hand grip muscle groups.
For these muscles the 50p load was between 8.7% and 12.8% of their
maximum. The average muscle load for the grip muscles was
between 78% and 100% for the asymmetric umbrella when
compared to the traditional umbrella canopy shape.
Surprisingly, the average muscular load of the grip muscle
groups did not decrease as much as the other muscle groups. A
reason for this may be that the subjects are accustomed to holding
a traditional umbrella which requires a specic grip force. Automatically, the subjects applied the same grip force to the asymmetric umbrella although it might not be necessary to hold it so
tight, as one of the subjects mentioned after the test.
5.3. 90p of normalized EMG
At the 90p over all muscle groups, the traditional canopy shape
required 1.3 times more muscular activity than the asymmetrical
shape. The 90p muscular activity of the grip muscle groups
demonstrates that for all wind speeds the traditional canopy
required signicantly higher activity than the asymmetric canopy.
The grip muscles had the highest %MVC, signicantly higher than
the shoulder for the traditional canopy and the upper arm and
shoulder for the asymmetric canopy shaped umbrella. This illustrates that it is plausible that the grip force muscles were the
weakest link. Although, none of the participants let go off the
umbrella, some were not able to hold it above their head. However,
by moving the umbrella in such way that forces caused by the wind
on the hand decreases, they were able to hold it in their hand.
Another explanation of the difculties that people had with holding

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

L.F.M. Kuijt-Evers et al. / Applied Ergonomics xxx (2012) 1e9

the umbrella above their head (whilst the 90p value of the %MVC of
the gripping muscle group is only 50%), is the unexpected increase
of forces. As explained before, the wind forces on the hand are
affected by the orientation of the umbrella canopy. A slight change
in this orientation, may instantly increase the forces on the hand
enormously. This can occur at higher as well as lower wind speeds,
however at higher wind speeds, it is more likely to occur as it is
difcult to hold the umbrella statically in the same position in that
circumstances. Even a small change in orientation, can suddenly
result in an enormously increase of force, which forces the people
to move the umbrella away. However, then the question rises why
some people are not able to hold the traditional umbrella, but still
can hold the asymmetric umbrella. It seems that this is one of the
benets of the asymmetric shape of the canopy, which is like
a stream-lined half-body. Due to this shape, the drag coefcient is
lower than the drag coefcient of the traditional umbrella. The
lower drag coefcient results in a decrease of the drag force on the
asymmetric umbrella and thus directly affects the forces and
moments applied on the person. Therefore, it is easier to hold the
asymmetric umbrella in the same position, even at high wind
speeds.
Overall, the grip muscle activity with the asymmetric shape was
only 82% of the traditional, without the weakest participants at the
wind speeds of 7 and 8. The large average activity (over 30% MVC)
of the ECU, ECR, FDS and FCU demonstrates that holding an
umbrella against wind ranging from breezes to gales can be
a relatively difcult task with either umbrella. This task was more
difcult when resisting a side wind than one from the front. Around
half the participants couldnt even hold the traditional umbrella
when the wind speed got to gale-level (17.2e20.7 m/s), winds gusts
that are likely to be encountered frequently in many areas of the
world. This can also be explained by the biomechanical model as
shown in Fig. 4a and b. When the wind comes from the front, no
moments on the upper extremity occur in the top view (as all joints
are in line with the forces of the umbrella on the hand). When a side
wind occurs, both in top view and side view, moments occur on the
upper extremity. This means that the muscles have to compensate
for moments in different directions in different joints of the upper
extremity.
5.4. Workload and hand comfort
In examining the workload in the grip and comfort, Kong et al.
(2012) found that a 65% MVC is the tipping point of hand
comfort where comfort changes into discomfort of the hand. In
their study, in general, participants rated an overall hand area as
comfortable when asked to exert less than 65% MVC, whereas they
rated it uncomfortable when asked to provide more than 65% MVC
(Kong et al., 2012). Hence, the %MVC of the grip force muscles in our
study, will not have caused real hand discomfort, although the grip
force needed to hold the traditional umbrella was higher than

holding the asymmetric umbrella, therefore it is plausible that the


hand comfort would be higher holding the traditional umbrella for
longer periods of time, especially at higher wind speeds.
5.5. Conclusions
The traditional umbrella canopy is signicantly more difcult to
hold at all wind speeds than the asymmetric canopy (34% higher at
90p). At higher wind speeds, weaker participants could not even
hold the traditional umbrella, especially when the wind came from
the right side while the umbrella was in their right hand. One can
imagine that the opposite direction (wind from the left side when
holding the umbrella in the right hand) would be even worse, as the
wind force vectors then cause an external rotation and abduction of
the upper arm compared to an internal rotation and adduction
(supported by the trunk) as it was during the experiment. In
summary, to reduce the load, especially on the grip muscles at
higher wind speeds, the asymmetric umbrella canopy shape should
be adopted.
Acknowledgements
We would like to thank Senz Umbrellas and the Dutch Ministry
of Economic Affairs for their support.
References
Basmajian, J.V., 1989. Biofeedback: Principles and Practice for Clinicians. Williams &
Wilkins, Baltimore.
Duque, J., Masset, D., Malchaire, J., 1995. Evaluation of handgrip force from EMG
measurements. Applied Ergonomics 26, 61e66.
Franssen, J.L.M., 1995. Handboek oppervlakte-elektromyograe. De Tijdstroom,
Utrecht.
Hoozemans, M.J.M., van Dien, J.H., 2005. Prediction of handgrip forces using
surface EMG of the forearm muscles. Journal of Electromyography and Kinesiology 15 (4), 358e366.
Hoozemans, M., Loos, R., Wilms, M., Dien, J.H., 2006. Exact vs. random positioning
of EMG electrodes for handgrip force prediction. In: Pikaar, R.N.,
Koningsveld, E.A.P., Settels, P.J.M. (Eds.), Meeting Diversity in Ergonomics,
Proceedings IEA 2006 Congress, 10e14 July 2006, Maastricht, the Netherlands.
Koller, M., 1957. Wind-resistant reversible umbrella. IPC A45B25/22; A45B25/00. US
2788792(A) 1957-04-16.
Kong, Y.-K., Lowe, B.D., Lee, S.-J., Krieg, E.F., 2008. Evaluation of handle shapes for
screwdriving. Applied Ergonomics 39 (2), 191e198.
Kong, Y.-K., Kim, D.-M., Lee, K.-S., Jung, M.-C., 2012. Applied Comparison of comfort,
discomfort, and continuum ratings of force levels and hand regions during
gripping exertions. Applied Ergonomics 43 (1), 283e289.
Lin, J.-H., McGorry, R.W., Dempsey, P.G., Chang, C.-C., 2006. Handle displacement
and operator responses to pneumatic nutrunner torque buildup. Applied
Ergonomics 37 (3), 367e376.
McGorry, R.W., Lin, J.-H., 2007. The effect of handle location and orientation on
power grip force production. Journal of Biomechanics 40 (Suppl. 2), S70.
Morse, J.L., Jung, M.-C., Bashford, G.R., Hallbeck, M.S., 2006. Maximal dynamic grip
force and wrist torque: the effects of gender, exertion direction, angular
velocity, and wrist angle. Applied Ergonomics 37, 737e742.
Sengster, W., 1855. Umbrellas and Their History. Efngham Wilson, Royal Exchange,
London.

Please cite this article in press as: Kuijt-Evers, L.F.M., et al., Effect of canopy shape on physical load when holding an umbrella, Applied
Ergonomics (2012), http://dx.doi.org/10.1016/j.apergo.2012.06.003

S-ar putea să vă placă și