Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 81 (2015) 618 627

69th Conference of the Italian Thermal Machines Engineering Association, ATI2014

A dynamic model for the performance and environmental analysis

of an innovative e-bike
Carmelina Abagnale*, Massimo Cardone, Paolo Iodice, Salvatore Strano, Mario Terzo,
Giovanni Vorraro
Dipartimento di Ingegneria Industriale Universit degli Studi di Napoli Federico II

Abstract

The paper describes a performance and environmental analysis of an innovative e-bike by means of a dynamic model equipped
with a suitable control for the management of the assisted drive of the e-bike.
The main advantage of this approach consists of the opportunity to simulate different tracks, e-bikes characterized by different
parameters (total mass, wheels moment of inertia) and different control strategies.
The authors have also conducted an environmental analysis of the studied vehicle, particularly comparing the e-bike with a
thermal moped, in terms of environmental impact.

2015
2015TheTheAuthors.
Authors.Published
Publishedby by
Elsevier Ltd.Ltd.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Selection and peer-review under responsibility of ATI NAZIONALE.
Peer-review under responsibility of the Scientific Committee of ATI 2014
keywords: pedelec; e-bike; environmental analysis; dynamic model

1. Introduction

Strong problems related both to the air quality and to the use of petroleum [1] have been caused, in the recent
years, by the increasingly vehicular traffic. Particularly, the most of consumption and pollution are attributable to the
great mass of the vehicle and not to the mass of the handled passengers. Under this point of view, a vehicle as the
electrical assisted bicycle (e-bike) [2]-[4] can be considered a promising alternative vehicle for both personal
mobility and goods delivery, especially for small and medium distances.

* Corresponding author. Tel.: +390817683297; fax: +390812394165.

E-mail address: c.abagnale@unina.it

1876-6102 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of ATI 2014
doi:10.1016/j.egypro.2015.12.046
 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627 619

The e-bikes are normally powered by rechargeable battery [5][8], and their driving performance is influenced by
battery capacity, motor power, road types, operation weight, control, and particularly the management of assisted
power.
A classification of these BEVs (battery electric vehicles) is necessary. A first kind is represented by a pure
electric bike [9][11], which integrates electric motor into bicycle frame or wheels, and it is driven by motor force
only using a handlebar throttle. A second kind is a power-assisted bicycle, frequently called e-bike or pedelec [12],
which is a humanelectric hybrid bicycle [13] that supports the rider with electric power only when the rider is
pedalling.
A number of aspects favour the use of electric bicycles in different situations. These include lower energy cost
per distance travelled for a single rider; savings in other costs such as insurance, licenses, registration, parking;
improvement of the traffic flow; environmental friendliness; and the health benefit for the rider.
This paper investigates the performance and the environmental impact of an innovative electrically assisted
bicycle by means of a dynamic model equipped with a suitable control for the management of the assisted power.
The proposed approach allows to simulate different tracks, e-bikes characterized by different parameters (total mass,
wheels moment of inertia) and different control strategies. An environmental analysis (well-to-wheel) has been
performed taking into account a comparison with the tailpipe exhaust emissions of the regulated pollutants of a
thermal moped by using the simulated speed-time profiles. This study, then, gives informations related to some
advantages on air quality that could be obtained, in urban contexts, by using electrical assisted bicycles instead of
thermal two-wheeler.

2. The e-bike prototype

The dynamic model adopted for the numerical analysis refers, in terms of the adopted parameters, to an
innovative e-bike [14, 15] that will be briefly recalled in this section. The innovative prototype is based on the
following solutions:
- the electrical motor location;
- the battery pack located into the frame;
- the mechanical transmission;
- the low cost measurement system of the driving torque.
Differently from a common approach, in which the electric motor is located on one of the three hubs of the
bicycle, a basic idea of the e-bike prototype consists of a central motor located in a bottle, as shown in Fig. 1, while
a gearbox is located in a central position, between the pedals, as pointed out in Fig. 2.
620 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627

Fig. 1. The innovative e-bike.

Fig. 2. The electrical motor and the gear box.

Global technical specifications of the innovative electrically assisted bicycle are reported in Table 1. In this
bicycle the level of electrical assistance is determined by the user input and the motor assistance is only available
when the user is pedalling, in accordance with the Italian law requirements. This e-bike is equipped with a central
motor with two different gearboxes: the first one is a planetary gearbox and the second one is a bevel gear.
 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627 621

Table 1. Technical specifications of the e-bike.

Total Weight 18 kg
Motor Type and Power Brushless dc machine, 250 W
Motor Assembly Hub
Motor Placement Central hub
Levels of electrical assistance three Modes
Battery Type Li-Po, 36V, 10Ah
Charging Time 4h
Cycles of charge/discharge Up to 1000
Tire Type 26 x 1,8"

The innovative prototype has been tested on different on-road test tracks in urban area of Naples to the aim to
acquire experimental data essential to tuning and validate the dynamic model described in the next section.

3. Dynamic model description

With reference to Fig. 3, the bicycle longitudinal dynamics can be expressed as

dv
M  Fw  Frr  F  Ft (1)
dt

where M is the total mass of the pedelec (including the rider), v is the pedelec longitudinal velocity and t is the time.
The term Fw in (1) is the aerodynamic drag force:

1
Fw As Cw airv 2 (2)
2

where As is the reference area of the bicycle-rider system, Cw is the aerodynamic drag coefficient, air is the air
density.
The term Frr (with reference to Fig. 3, Frr is the sum of its two components Frr,r and Frr,f, acting on the rear and front
wheels ) is the rolling resistance force:

Frr Crr Mg cos() (3)

where g is the gravitational acceleration, is the slope of the road, Crr is the rolling resistance coefficient.
The gradient resistance force F is formulated as

F Mg sin() . (4)

Finally, Ft is the total thrust force provided by the motor and the rider.
622 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627

Fig. 3. Longitudinal components of the forces acting on the pedelec.

The bicycle longitudinal dynamics can be expressed in function of the total driving torque Td,w applied to the rear
wheel

dv
Td ,w r M  Fw  Frr  F . (5)
dt

where r is the nominal radius of both the bicycle wheels.

Introducing the gear ratio g and the efficiency g of the bicycle gearbox

p
g
w
, (6)
Td , w w Td , w
g
Td , p p Td , p g

it is possible to obtain the expression of the driving torque Td,w relative to the pedal shaft:

r dv
Td , p M  Fw  Frr  F Th  Tm, p (7)
g g dt

where w is the angular velocity of the bicycle wheels, Th and Tm,p are the human and the motor torque applied to the
pedal shaft, respectively.
Considering the aforementioned gear train between the motor and the pedal shaft, characterized by a gear ratio m
and an efficiency m defined as

m
m
p
(8)
Tm , p p Tm , p
m
Tm m Tm m

the torque equilibrium at the motor shaft can be derived:

 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627 623

dv g m g m g g
M  Fw  Frr  F Tm  Th (9)
dt r r

where m is the motor angular velocity and Tm is the torque provided by the electric motor, applied to its shaft.
The assistance method is based on the pedelec velocity control (PVEC). The proposed control method is based on
the electric motor torque control law designed to compensate external disturbances and to minimize the bike
velocity tracking error. The control torque consists of a feedforward torque integrated with a feedback one. The
feedforward contribution is a nonlinear torque based on the pedelec model, while the feedback torque has the
function to compensate the tracking error due to the model uncertainties and the unknown disturbances.
The control action consists in the motor torque Tm and the control feedback is the pedal angular velocity p,
obtainable with a proximity sensor. The target to the PVEC is the desired velocity vd imposed by the rider that is
relative to a desired assistance level. For example, a target pedelec velocity equal to 25 km/h corresponds to the
maximum motor assistance level (100 %), instead, a desired velocity of 15 km/h involves a 60 % assistance level.
The rider pedalling torque, estimated via the chain force measurement, is used only for the activation of the
assistance, in particular, whenever the rider torque exceeds a limit value, the additional motor torque will be
provided by the controlled electric motor.
The performance of the controller can be achieved by suitably choosing the parameters of the PID control. The
tuning of the PVEC parameters has been performed in order to achieve a balance between tracking performance,
stability and robustness.
The rider has been modelled with a constant torque of 15 Nm that will be reduced to a value of 2 Nm when the
pedelec velocity exceeds the desired velocity.
The pedelec specifications and environmental parameters adopted for the simulation are listed in Table 2.

Table 2. Model parameters.

Model parameter Value
M (kg) 90
r (m) 0.3
m (-) 0.95
g (-) 0.98
m (-) 25
g (-) 0.47
Cw (-) 0.9
Crr (-) 0.003
As (m2) 0.4
air (kg/m3) 1.225
g (m/s2) 9.81

The slope profile for testing the PVEC for urban riding within 7 km is shown in Fig. 4.
624 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627

Fig. 4. Slope profile.

Two different simulation tests have been performed, the first one with a target velocity of 16 km/h (TestA) and the
second one with a target velocity of 22 km/h (TestB).
The target and the effective velocity of TestA and TestB are presented in Fig. 5 and Fig. 6, respectively.

Fig. 5. Target and effective pedelec velocity, TestA.

 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627 625

Fig. 6. Target and effective pedelec velocity, TestB.

3. Environmental analysis

In relation to the examined e-bike, in this paper, an environmental assessment has been performed considering
the relationship with the emissive behaviour of a moped under real driving situations. Mopeds and motorcycles are
broadly used and represent a large share of motorized vehicles, above all in Asia and Europe. Determination of
emissions from two-wheelers is then very important in order to estimate the relevant contribution to total emissions
attributable to road transport, taking also into account that their impact to air pollution is generally growing,
particularly in urban contexts [16 - 18].
Starting from these considerations, in the last years experimental investigations on 2-wheel vehicles emissive
behaviour have been executed by the Department of Industrial Engineering of the University of Naples Federico II
in the Istituto Motori emission laboratory (National Research Council).
For the particular purpose of the present study, the experimental results of roller test bench measurements
performed on a moped [19] have been used in order to collect the relevant emissive data during real driving
conditions. In this investigation, in order to characterize the emissive behaviour of the moped during different
driving cycles, CO, HC and NOX emissions (regulated pollutants by law) have been evaluated in the exhaust of this
moped, that belongs to the Euro 2 legislative category, and is equipped with a 4-stroke engine, with a displacement
of 50 cm3. The emission performance of the tested moped has been determined on the statutory driving cycles for
Europe (ECE-47) and on different real-world driving cycles both during cold-start and hot phases.
Each driving cycle test of this experimental activity has been at first divided into basic kinematic sequences
(speed-time profiles amid two successive stops) under warmed-up engine conditions, and analysed by using the
continuous emission results. Subsequently, each basic sequence has been categorized with several kinematic
parameters in order to better characterize its driving behaviour. The considered kinematic parameters are above all
the average speed v_mean, and the mean product m_va_pos of instantaneous speed v and acceleration a
(calculated for a>0). This product, in fact, characterizes the traction power per unit mass needed to overcome the
vehicle inertia, then its a factor decisively correlated to the emissions of the moped [20].
Eleven elementary sequences (Table 3), that cover different situations present on urban road, have been
identified; mean value of emission factors obtained and measured during several repetitions of these driving
sequences are displayed in the same table.
626 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627

Table 3. Kinematic parameters of the moped driving patterns and moped Emission factor mean of measurements.

m_va_pos
v_mean

(m2/s3)
(km/h)
Driving CO NOX HC
pattern (g/km) (g/km) (g/km)

A 25.13 2.38 3.18 0.13 0.71

B 21.50 3.27 2.10 0.32 1.03
C 21.43 2.65 3.66 0.19 0.61
D 18.37 5.11 4.91 0.15 0.72
E 35.44 3.29 2.96 0.18 0.47
F 10.41 6.23 4.03 0.37 1.16
G 13.44 3.37 4.67 0.24 1.10
H 34.03 3.59 2.67 0.24 0.39
I 38.97 2.08 2.75 0.20 0.29
L 17.59 4.05 5.83 0.13 0.81
M 20.60 4.53 8.47 0.11 0.88

This extensive database of the emission factors of the regulated pollutants (for different eleven driving patterns)
has been used to assess the environmental impact of a thermal moped against the e-bike, under the same driving
conditions of the simulated speed-time profiles. By means of this procedure, it has been achievable to obtain the
total emissions of each regulated pollutant that could be emitted, for the simulated speed profiles, utilizing a thermal
moped instead of a e-bike. The results of this elaboration are shown in Table 4.

Table 4. Savings of emissions for the modelled speed-time profiles utilizing an E-BIKE in place of a thermal moped.

Pollutant CO [g] NOX [g] HC [g]

Speed-profile TestA 32.7 1.7 7.7

Speed-profile TestB 25.6 1.3 4.3

Conclusion

This paper has described the environmental analysis of an electrically assisted bicycle under real driving conditions
of simulated speed-time profiles.
In this study, experimental results of roller test bench measurements performed on a thermal moped were
employed in order to collect the pertinent emissive data during real driving conditions. The environmental
assessment was performed taking into account a comparison with the emissive behavior of this moped by using
kinematic parameters that describe the simulated driving dynamics; a clear benefit of e-bike compared to thermal
mopeds was shown and quantified in terms of emissions saved of CO, HC and NOX.

Acknowledgements

This research has been financially supported by MIUR (Ministero Istruzione Universit e Ricerca) under the
grant named PON04a3_00408 "Bicicli e Tricicli elettrici a pedalata assistita di nuova generazione.

References

[1] Onoda T and Gueret T, Fuel efficient road vehicle non-engine components: Potential saving and policy recommendations, Int. Energy
Agency Inf. Paper, pp. 124, Oct. 2007
 Carmelina Abagnale et al. / Energy Procedia 81 (2015) 618 627 627

[2] Seki H, Ishihara K, and Tadakuma S, Novel regenerative braking control of electric power-assisted wheelchair for safety downhill road
driving IEEE Trans. Ind. Electron., vol. 56, no. 5, pp. 13931400, May 2009.
[3] Tsai CC, Huang HC, and Lin SC, Adaptive neural network control of a self-balancing two-wheeled scooter, IEEE Trans. Ind. Electron.,
vol. 57, no. 4, pp. 14201428, Apr. 2010.
[4] Emadi A, Lee YJ, and Rajashekara K, Power electronics and motor drives in electric, hybrid electric, and plug-in hybrid electric vehicles,
IEEE Trans. Ind. Electron., vol. 55, no. 6, pp. 22372245, Jun. 2008.
[5] Muetze A and Tan YC, Electric bicycles: A performance evaluation IEEE Ind. Appl. Mag., vol. 13, no. 4, pp. 1221, Jul./Aug. 2007.
[6] Tanaka Y and Murakami T, A study on straight-line tracking and posture control in electric bicycle, IEEE Trans. Ind. Electron., vol. 56, no.
1, pp. 159168, Jan. 2009.
[7] Defoort M and Murakami T, Sliding-mode control scheme for an intelligent bicycle IEEE Trans. Ind. Electron., vol. 56, no. 9, pp. 3357
3368, Sep. 2009.
[8] Lomonova EA, Vandenput AJA, Rubacek J, dHerripon B, and Roovers G, Development of an improved electrically assisted bicycle in
Conf. Rec. IEEE IAS Annu. Meeting, 2002, vol. 1, pp. 384389.
[9] Fairley P, Chinas cyclists take charge: Electric bicycles are selling by the millions despite efforts to ban them, IEEE Spectr., vol. 42, no. 6,
pp. 5459, Jun. 2005.
[10] Somchaiwong N and Ponglangka W, Regenerative power control for electric bicycle, in Proc. IEEE Int. Joint Conf. SICE-ICASE, 2006,
pp. 43624365.
[11] Yang MJ, Jhou HL, Ma BY, and Shyu KK, A cost-effective method of electric brake with energy regeneration for electric vehicles IEEE
Trans. Ind. Electron., vol. 56, no. 6, pp. 22032212, Jun. 2009.
[12] Schneider D, Easy rider: Convert your bicycle to a humanelectric hybrid, IEEE Spectr., vol. 46, no. 9, pp. 2627, Sep. 2009.
[13] Du W, Zhang D, and Zhao X, Research on battery to ride comfort of electric bicycle based on multi-body dynamics theory, in Proc. IEEE
Int. Conf. Autom. Logistics, Aug. 2009, pp. 17221726.
[14] Abagnale C, Cardone M, Iodice P, Strano S, Terzo M, Vorraro G. Theoretical and Experimental Evaluation of a Chain Strength
Measurement System for Pedelecs. Engineering Letters. 2014; 22: 3, pp. 102 108.
[15] Abagnale C, Cardone M, Iodice P, Strano S, Terzo M, Vorraro G. Analysis of a new measurement system of the chain strength for
electrically assisted bicycles. In proceedings of the ASME 2014 12th Biennial Conference On Engineering Systems Design and Analysis
ESDA2014, Copenhagen, Denmark, 2014.
[16] Iodice P, Abagnale C, Cardone M, Strano S, Terzo M, Vorraro G. Performance evaluation and environmental analysis of an electrically
assisted bicycle under real driving conditions. In proceedings of the ASME 2014 12th Biennial Conference On Engineering Systems Design and
Analysis ESDA2014, Copenhagen, Denmark, 2014.
[17] Iodice P, Senatore A. Experimental-analytical investigation to estimate an emission inventory from road transport sector. IAENG
Transactions on Engineering Sciences - Special Issue of the International MultiConference of Engineers and Computer Scientists, IMECS 2013
and World Congress on Engineering, WCE 2013, 2014, 141-149
[18] Iodice P, Senatore A. Evaluation of Dispersion Models for Predicting Carbon Monoxide Concentrations from Motor Vehicles in a
Metropolitan Area. International Review on Modelling and Simulations; 2013; Vol. 6, Issue 6: p. 1928-1932
[19] Iodice P, Senatore A. Influence of ethanol-gasoline blended fuels on cold start emissions of a four-stroke motorcycle. Methodology and
results. SAE Technical Paper 2013-24-0117. 11th International Conference on Engines and Vehicles, ICE 2013; Capri, Naples; Italy
[20] Iodice P, Senatore A. Cold start emissions of a motorcycle using ethanol-gasoline blended fuels. Energy Procedia; 2014; Vol. 45, p. 809818