ISBS 2000 !
Hong Kong
their compositions general, fUll '
:; values of angular velocity an~
1 vanatlon among the subject
s,
mIs, more specific correlation
:; phase portraits in the corpora~
1t the real possibilities of phase
n of the running technique in
~ Training. In: GAMBETTA, V.
udy of intra limb coordination in
malrsis. Annual Conference of
hysIcal Activity. Asilomar, CA
me~t of walking as a limit cycle
I1S In development: application.
Analysis of Overground and
mce in Sports and Exercise , 8,
de Carrera Lanzada en las
va, 31, pp 129-140,
110.65 km/hr (68.73 mph), established in 1992 by a single rider on a recumbent bicycle
tic C::>mparison of Overground
named the Cheetah (Kor, 1992), it becomes questionable whether a more aerodynamically
xerclse. 27, 98-115,
effective human powered vehicle can be designed. If future speed records are to be
ng Systems and Infant Motor
ltics and Electromiography of
International Journal of Sports
es to Deep Water Running in
Imics of Disordered Human
,vior, 21 (4), 373-391.
is valuable cooperation in the
'ancia! assistance.
93
EFFECT OF CHANGES IN CRANK ARM LENGTH AND LOAD ON POWER
PRODUCTION IN RECUMBENT CYCLING
Danny Too, Chris Williams, Edgar J. Wakayama 1 and Gerald E. Landwer
State University New York at Brockport, New York, USA
lVirginia Commonwealth University, Richmond, Virginia, USA
2University of Nevada, Las Vegas, Nevada, USA
The purpose of this study was to determine the trend in peak power (PP), mean power
(MP), and minimum power (MINP) with changes in load when cycling in a recumbent
position. Fifteen female participants were randomly assigned to one of three crank arm
length (CAL) conditions (110, 180, or 250 mm) and tested on a Monark Cycle ergometer
with 5 loads varying from 75-165 gmlkg of body mass. The Wingate Anaerobic Cycling
test was performed in a recumbent position (75 0 seat tube angle, backrest perpendicular
to the ground), Curve estimation with regression analysiS on incrementing loads revealed:
(1) a quadratic trend in PP; (2) a quadratic trend in MP and MINP for the 110 and 180 mm
CAL; and (3) a linear trend in MP and MINP for the 250 mm CAL. These trends suggest
there is an optimal load for different CALs to maximize power production.
KEY WORDS: recumbent, cycling, crank-arm length, load, power
INTRODUCTION: It is well documented that recumbent human powered vehicles with an
aerodynamic fairing, having a smaller drag coefficient and cross-sectional area, are faster
than the standard racing bicycle (Kyle, 1982). However, with the current speed record of
attained, it is necessary to not only focus on the aerodynamics, but also to examine the
variables that affect power production in recumbent cycling and the intel'actions that would
maximize power. Investigations of recumbent cycling and power production have examined
changes in seat-tube angle (Too, 1991), trunk/backrest angle (Too, 1994), seat to pedal
distance (Too, 1993), and CAL (Too, 1996).
Too (1991), examining a systematic change in seat tube angle (0°, 25°, 50°,75°, and 100°)
with a load of an 85 gm/kg of body mass (BM), reported a 75° seat tube angle to yield the
largest PP and MP, with a parabolic curve (quadratic trend) to best describe the change in
PP and MP with changing seat tube angles. Using a 75° seat-tube angle and an 85 grnlkg
load, Too (1994) investigated the effect of 3 trunk angles (60°, 90°, and 120°) on power
production. A 90° trunk angle was reported to yield the largest PP and MP, and that a
parabolic trend best described PP and MP with changes in trunk angle. Using a 90 0 trunk
angle, a 75° seat tube angle, and an 85 gm/kg load, Too (1993) examined the effect of seat­
to-pedal distance (90%, 95%, 100%, 105%, and 110% of total leg length) on power
production. A quadratic and linear function was reported to best describe the trend in PP
and MP with changing seat-pedal-distance. In addition, Too (1996) examined changes in
CAL (110, 145, 180,230, and 265 mm) with a 75° seat tube angle, a 90 0 trunk angle, a 100%
seat-to-pedal distance, and a load of 85 gm/kg BM. A 110 and 180 mm CAL was reported to
yield the largest PP and MP, respectively; and that a linear and quadratic trend was reported
to best describe the change in PP and MP, respectively, with increasing CAL (Too, 1996).
Based on muscle force-length and force-velocity-power relationships, changes in CAL and
load will affect joint angles, muscle length, force, torque, and power production in cycling.
Since the literature involving traditional cycling positions and power output have reported an
interaction between CAL, load, and pedaling cadence (Hull & Gonzalez, 1988; Yoshihiku &
Herzog, 1990), it can be assumed that similar interactions will occur in a recumbent cycling
Position. Therefore the purpose of this investigation was to determine what trends exist (if
any) with systematic changes in load on power production with changes in CALs. Any trends
94

in power production will provide information regarding loads and CALs to maximize power sed on regression analys.is of
production during recumbent cycling. Baveral trends were determined
~ three CAL, although the tren
METHODS: Fifteen healthy volunteer female participants (mean age = 22.3 ± 4.25 yr., a CAL is best described by a
weight 60.6 ± 9.45 kg, height 165 ± 9.8 cm) were randomly assigned to one of three CAL ~:L is best characterized by a
conditions (110 mm, 180 mm, or 250 mm) and tested with 5 different loads. The 5 loads CAL is best represented by ~ q
used in the 110 mm CAL condition were 75, 90, 105, 120, and 135 gm/kg 8M. The loads CAL is best described by a linE
used in the 180 mm CAL condition were 90,105,120,135, and 150 gm/kg 8M; whereas the . the various measures of power
loads used in the 250 mm CAL condition were 105, 120, 135, 150, and 165 gm/kg 8M. All
five participants in each CAL condition were tested with the designated loads according to a peak power: . _
110 mm CAL (quadratic trend, p ­
randomly determined sequence and with a minimum of 24 hours rest between test sessions.
, 180 mm CAL (quadratic trend, p
The recumbent cycling position, used for all test sessions, was defined by a 75° angle 250 mm CAL (quadratic trend, p

formed between the bicycle seat tube and a vertical line passing through the crank spindle
Mean power: . _

(Too, 1991). To obtain this seating position, a variable seating .apparatus, allowing for 110 mm CAL (quadrat~c trend, p =

manipulations in seat tube angle, backrest angle, and seat-to-pedal distance, was used and 180 mm CAL (quadratic trend, p ­
interfaced to a Monark cycle ergometer (Model 814E). The seat backrest was kept 250 mm CAL (linear trend, P =0.(

perpendicular to the ground. and the seat-to-pedal distance was adjusted to 100% of total leg Minimum power:

length for each subject as measured from the right femur to the ground (Too, 1991). Two 110 mm CAL (quadratic trend, p ,

adjustable crank arms, allowing for manipulations from 0 to 300 mm, were used for the 110, 180 mm CAL (quadratic trend, p

180, and 250 mm CAL conditions (Too, 2000). 250 mm CAL (linear trend, p O.

Each subject was strapped to the seat-backrest at the trunk and hips, and pedal toe-clips
were used. The test protocol involved a computerized 30-second Wingate Anaerobic Cycling ....
Test. To initiate the test, the subject pedaled the cycle ergometer with no load. Once the "..
..c "t
ergometer's inertial resistance had been overcome, the appropriate load was instantaneously <0­ f­ """-
k
applied using calibration weights, and the subject pedaled as fast as possible for 30 seconds. -B-- Craokann 110 mrn

A Sports Medicine Industry (SMI) opto-sensor (Model 2000) with a sampling rate of 50 Hz, 0
Crankann 160 mrn

interfaced with a Zenith 386 micro-computer and used in conjunction with 16 reflective 4-- Oankann 250 mm

markers on the ergometer flywheel, was used to monitor and record flywheel revolutions 0
75 90 105 120 135 150 16
during the test. PP was calculated from the highest average flywheel speed during any Load (gm/kg)
consecutive 5 seconds; MP was determined from the mean flywheel speed for the entire 30­
second test, and MINP was calculated from the lowest mean flywheel speed during any Figure 1· Predicted pow
consecutive 5 seconds. Curve estimation with regression analysis on incrementing loads 250 mm CAL.
was used to determine the trend in PP, MP, and MINP with the 110, 180, and 250 mm CAL.
From Table 1 and the trer
RESULTS: With changes in load, the mean ± SO values of PP, MP, and MINP for the 3 generally greater than MP, '
CALs are presented in Table 1. greater with the 110 mm CP
is greater with the 250 mm (
Table 1 Peak Power, Mean Power, and Minimum Power with Changes in Load as load increases, power pi
and Crank arm Length trend of PP, MP, and MINP
gmlkg of body mass), It WI
CAL Power 75 90 105 120 135 150 165 power production.

577 ± 216
DISCUSSION: In this inv!
678 ± 155 647 ± 82 606 ±80
consistent with the interact
110 MP 345 ± 101 370 ± 116 353 ± 101 279 ± 110 170 108 reported in the upright cyc
MINP 223 ± 96 222 ± 128 211 ± 120 175 ± 107 48 ± 76 With increasing load for
pedaling rate can be reac
PP 488 57 542 ± 51 511 75 462 ± 101 435 ± 91
value and the maximal I
180 MP 317 ± 57 312 ± 59 291 ± 58 225 ± 98 156 ± 81 decrement in PP as obsei
MINP 217 ± 42 199 ± 50 197 ± 43 125 ± 115 34 ±54 pedaling rate was not dir'
recorded flywheel revolut
PP 488 ± 79 514 ± 71 512 117 490 ± 79 479 ± 113 with increasing CAL (fro!
250 MP 366 ± 63 359 ± 72 362 ± 82 330 ± 69 317 97 assumed that the load Ii'.
MINP 277 ± 41 262 ± 60 265 ± 58 243 ± 66 220 + 62 greater loads (such as th
 ISBS 2000 I Hong Kong 95

and CALs to maximize POWer
mean age 22.3 ± 4.25 yr., .
I assigned to one of three CAL'
j different loads. The 5 loads
nd 135 gm/kg BM. The loads
ld 150 gm/kg BM; whereas the
, 150, and 165 gm/kg BM. All
;:)signated loads according to a
rs rest between test sessions.
Based on regression analysis of power production with incrementing loads and different CAL,
everal trends were determined (see Figure 1). PP is best described by a quadratic trend for
5 three CAL, although the trends were not significant (Q > 0.05). MP with the 110 and 180
11
~m CAL is best described by a quadratic equation (Q < 0.01), whereas MP with the 250 mm
CAL is best characterized by a linear equation (Q < 0.05). MINP with the 110 and 180 mm
CAL is best represented by a quadratic function (Q < 0.05), whereas MINP with the 250 mm
CAL is best described by a linear function (Q < 0.05). The specific regression equations for
the various measures of power with incrementing loads, using different CALs are as follows:
Peak power: 2
110 mm CAL (quadratic trend, p == 0.29): PP =-0.091 x + 18.8x - 311 (standard error =35)

was defined by a 75° angle 180 mm CAL (quadratic trend, p =0.17): PP =-0.058x2 + 12.6x - 165 (standard error == 24)

sing through the crank spindle 250 mm CAL (quadratic trend, p = 0.17): PP = -0.050X2 + 7.9x 1 (standard error 9)

;;ating apparatus, allowing for Mean Power:

2
110 mm CAL (quadratic trend, p 0.001): MP -0.103x + 18.7x 480 (standard error == 3)

-pedal distance, was used and

180 mm CAL (quadratic trend, p = 0.004): MP -0.055x2 + 10.5x 184 (standard error 6)

The seat backrest was kept 250 mm CAL (linear trend, p = 0.03): MP == -0.851 x + 462 (standard error = 10)

IS adjusted to 100% of total leg

Minimum Power:

the ground (Too, 1991). Two 110 mm CAL (quadratic trend, p = 0.03): MINP -0.088x2 + 15.9x - 479 (standard error 18)

)0 mm, were used for the 110, 180 mm CAL (quadratic trend, p 0.02):'MINP == -0.069x2 + 13.6x - 456 (standard error = 14)

250 mm CAL (linear trend, p = 0.02): MINP = -0.881-x + 372 (standard error == 9)

and hips, and pedal toe-clips

_800 ~800
md Wingate Anaerobic Cycling [800 ~ -£- 110mmCAL
neter with no load. Once the .... 60(1 '-- 600 W600
.riate load was instantaneously Q)
(» ;;:
3: 0
:1st as possible for 30 seconds. ~.oo 0400
0...
a. 400
a.. E
Nith a sampling rate of 50 Hz, :J
~
ro 200,
-.f;- Crankarrn 150 mm
~ 200 E 200
conjunction with 16 reflective (l) -+- Crafikarrn 250 mOl (» ·c
ld record flywheel revolutions a.. ~ 0
~ 0
0
~e flywheel speed during any 75 90 105 120 135 150 165 75 90 105 120 135 150 165 75 90 105 120 135 150 165

wheel speed for the entire 30­ Load (gm/kg) Load (gm/kg) Load (gm/kg)
m flywheel speed during any
nalysis on incrementing loads Figure 1 - Predicted power production with incrementing load for the 110, 180, and
110, 180, and 250 mm CAL. 250 mm CAL. ,.

PP, MP, and MINP for the 3 From Table 1 and the trends in Figure 1, several obselVations can be made: (1) PP is
generally greater than MP, and MP greater than MINP regardless of CAL and load; (2) PP is
greater with the 110 mm CAL than with the 180 or 250 mm CAL regardless of load; (3)MINP
)wer with Changes in is greater with the 250 mm CAL than with the 110 or 180 mm CAL regardless of load; and (4)
as load increases, power production appears to be favored with longer CALs. Based on the
trend of PP, MP, and MINP for the different CALs with comparable loads (105, 120, and 135
gm/kg of body mass), It would appear that there is an interaction between CAL, load, and
135 150 165 power production.

77 ± 216 DISCUSSION: In this investigation, the trends obselVed in power production appear to be
70 ± 108 consistent with the interactions of load, pedaling rate, CAL, and power output that have been
~8 ± 76 reported in the upright cycling literature (Hull & Gonzalez, 1988; Yoshihiku & Herzog, 1990).
With increasing load for any given CAL, greater PP is produced if the same maximal
32 ± 101 435 91 pedaling rate can be reached and maintained. If the load exceeds some optimal or critical
25 ± 98 156 == 81 value and the maximal pedaling rate cannot be attained or sustained, there will be a
decrement in PP as obselVed in the trends for the 110, 180, and 250 mm CALs. Although
25 ± 115 34 ± 54
pedaling rate was not directly measured in this investigation, it can be calculated from the
recorded flywheel revolutions. When Too (1996) reported a decreasing linear trend in PP
12 117 490 ± 79 479 ± 113
·62 ± 82
with increasing CAL (from 110 to 265 mm) using a fixed load (85 gm/kg BM), it can be
330 ± 69 317 ± 97
assumed that the load was large enough to maximize PP only for the 110 mm CAL. If
'65 ± 58 243 + 66 220 ± 62 greater loads (such as those in this investigation) were used, a quadratic trend in PP would
 96 ISBS 2000 I Hong Kong ISBS 2000 I Hong Kong .._
-~- .. - - -..- ­

be expected with increasing CALs. However, when the same fixed load (85 gm/kg BM) was ESTIMATION OF POWE
used in upright cycle ergometry, a quadratic trend was found in PP wlth increasing CAL from AceE
110 to 265 mm (Too, 2000). This difference in PP trend (linear vs. quadratic) between an
upright and recumbent position would suggest that greater PP can be achieved in a AnUm
recumbent cycling position (when compared to a standard upright cycling position) and may Biomechan
be attributed to greater pedal forces when using a seat-backrest to push against. Tech
In this investigation, the greater PP obtained with the 110 mm CAL (when compared to the
180 and 250 mm CAL) can be attributed to the greater pedal rate that can be achieved with a The purpose of this study
shorter CAL. As the load increased from 105 to 135 gm/kg BM, the same maximum pedal knee, and from knee to
acceleration phase of spr
rate for the 110 mm CAL apparently could not be maintained resulting in a decrement in PP.
skeletal model. The energ
The trends in PP for the 3 CALs would suggest that if PP were extrapolated (using a load of was 72.6±14.5 J (34.4±5.~
165 gm/kg BM) for the 110 and 180 mm CAL, PP for the 110 mm CAL would be less than from knee to ankle was ;
that of the 180 and 250 mm CAL transfer allows high powe
This appears to be supported by the smaller MP and MINP values of tM 110 mm CAL when explanation of why the ffi'
compared to those of the 180 and 250 mm CAL with a load of 135 gm/kg BM. lhe smaller the values obtained in an i
MP and MINP values of the 110 mm CAL can be attributed to a decrement in pedaling rate
due to fatigue in the latter part of the test. As the load increased, pedal rate and power KEY WORDS: biarticular r
production decreased, and even more so with shorter CALs and with the onset of fatigue.
This interaction between load, CAL, and pedaling rate would be consistent with what would INTRODUCTION: In sport me
be expected based on muscle force-length and force-velocity-power relationships. explosive leg extensionmo
Although only females were selected (due to their relatively smaller body masses when exta'1sion moment of force, (
compared to males, and to the maximum load limit of the ergometer) to participate in this isometric and isokinetic mom(
investigation, similar trends in power production would be expected with males if similar machines. Values of maximal
loads and CALs were used. maximal running (Jacobs, Bot
jump exercises (Bobbert, Hui
CONCLUSION: The trends in PP, MP, and MINP with incrementing load for 3 CALs (110, values for maximal isometric
180, and 250 mm), would suggest there is an optimal load for. different CALs to maximize studies of (Wickiewicz. Roy,
. power production in a recumbent cycling position. For human powered vehicle competitions (Cabri, 1989) who presente(
of short duration, where maximal PP is necessary, a short CAL is recommended for use with extensor moments. The sam(
the largest load that would not result in a decrement in maximal pedal rate. For competitions 1988).
of longer duration where fatigue is a factor and greater MP and MINP becomes important, it One of the mechanisms that I
is suggested that a long CAL is used with the largest load that would not result in a muscles perform differently
decrement in maximal pedal rate. The optimal load(s) for various competitions would be testing. The major difference
dependent on the force and power production capability of each individual. length changes are depende,
the muscle. This means tt
REFERENCES:
biarticular leg extensors are c
Hull, M.L., & Gonzalez, H. (1988). Bivariate optimization of pedalling rate and crank arm
adjacent limb is in a fixed pc
length in cycling. Journal of Biomechanics, 21, 839-849.
transfer of mechanical power
Kor, J. (1992). News flash! Human Powered Vehicle News, 9 September/October, 7.
joints they cross (Prilutsky & :
Kyle, C.R. (1982). Bicycling, 23, 59-66
The purpose of this study w
Too, D. (1991). The effect of hip position/configuration on anaerobic power and capacity in
knee and from knee to anklE
cycling. International Journal of Sports Biomechanics, 7(4), 359-370.
muscles, during the acceler;
Too, D. (1993). The effect of seat-to-pedal distance on anaerobic power and capacity in
explained the difference four
recumbent cycling. Medicine and Science in Sports and Exercise, 25(5), S68. (Abstract)
and the value obtained by inv
Too, D. (1994). The effect of body orientation on power production in cycling. The_Research

Quarterly for Exercise and Sport, 65, 308-315.

METHODS: Ten elite sprinte'
Too, D. (1996). The effect of pedal crankarm length on power production in recumbent cycle
0.433±0.020 m and shank 1£
ergometry. In T. Bauer (ed.) Proceedings of the XIII International Symposium on
extension and ankle plantar
Biomechanics in Sports (pp. 350-353). Lakehead University Printing Services, Thunder Bay,
performed at the angle wher
Ontario, Canada.
when performing the maxima
Too, D. (2000). The effect of pedal crank arm length on joint angle and power production in
The same sprinters perform
upright cycle ergometry. Journal of Sports Sciences, 18(3), 153-161.
start was performed over a
Yoshihiku, Y., & Herzog, W. (1990). Optimal design parameters of the bicycle-rider system
recorded at 1KHz. The trial
for maximal muscle power output. Journal of Biomechanics, 23, 1069-1079.
selected for analysis. Simultc
plane of the ankle, knee, hip