MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS

FOR INSTRUMENTED IMPLANTS

Towards the Development of a Smart Hip Prosthesis

M. Soares dos Santos

, Jorge A. F. Ferreira

, A. Ramos

, Ricardo Pascoal

, Raul Morais dos Santos

Nuno M. Silva

, Jos´e A. O. Sim˜oes

, M. J. C. S. Reis

2,3

, Camila N. Boeri

, Ant´onio Festas

and Paulo M. Santos

Department of Mechanical Engineering, Centre for Mechanical Technology & Automation, University of Aveiro,

Campus Universit´ario de Santiago, 3810-193, Aveiro, Portugal

Engineering Department, University of Tr´as-os-Montes e Alto Douro, Quinta de Prados, 5001-801 Vila Real, Portugal

IEETA - Instituto de Engenharia Electr´onica e Telem´atica de Aveiro/UTAD, 5001-801 Vila Real, Portugal

Keywords:

Smart hip prosthesis, Electric bio-generator, Energy harvesting, Self-powered systems, Biomedical implants.

Abstract:

Very few developments have been done to provide electric power supply of instrumented hip prosthesis.

Actually, vibration-powered generators are the most appropriate mechanisms for this kind of application’s

environment. This paper describes the ﬁrst attempt to develop the concept of energy harvesting from multiple

energy sources applied in the same hip implant. Exploiting the potential of the three angular movements over

the femoral component, namely in the abduction-adduction, ﬂexion-extension and inward-outward rotation

axes, three inboard vibration-based mechanisms were developed in order to ensure electric power supply from

multiple energy sources. A total of 53.7 µJ/s was harvested by a translation movement-based electromagnetic

energy generator when a sinusoidal function with an amplitude of 40 mm and a frequency of 4 Hz was applied.

A rotation movement-based electromagnetic energy generator has harvested 0.77 µJ/s when a sinusoidal

function with an amplitude of 60 and a frequency of 2.5 Hz was used. The piezoelectric energy harvester has

achieved 0.6 µJ/s with the application of a sinusoidal function with an amplitude of 200 N and a frequency of

4 Hz. Besides, its ability of being fully autonomous, operating without expiry and maintenance, while oﬀering

safety during its entire lifetime are relevant features. This paper should provide the basis for the development

of smart hip prosthesis with the ability to ﬁx the aseptic implant loosening problem.

1 INTRODUCTION

Loosening of the prosthetic stem and cup is a

serious complication of the Total Hip Replacement

Arthroplasty (THR), being referred that more than

80% of the non-success surgical procedures are due

to implant loosening (Alpuim et al., 2008). Generally,

the revision rate is about 10% in the case of prosthesis

implanted 10 years before. However, a growth of

about 100% is estimated in the revision procedures in

the EUA by 2030 (Kurtz et al., 2007). The progressive

bone loss surrounding the implant is considered the

main cause of the THR failure. The revision surgeries

are more complex, more expensive, more painful and

present a non-success rate higher than primary THRs.

Taking into account the increase of the average life

expectancy and the number of THRs applied in young

patients, the development of durable hip prosthesis

is imperative. The current instrumented prosthesis

proposals have only been designed to collect forces

and kinematics data acting in vivo, in order to promote

the continuous optimization of such implants (Damm

et al., 2010; Heinlein et al., 2009; Westerhoﬀ et al.,

2009; Rohlmann et al., 2008; Heinlein et al., 2007;

Graichen et al., 1999). The expertise of these implants

does not allow continuous real-time problem solving,

but only provides data to support new research. To

avoid the need of revision surgeries, a new concept

of prosthesis is emerging to diagnose and contribute

to ﬁx the loosening problem: methodology based on

the use of mechanical micro-stimulation to promote

the remodelling of the bone surrounding the implant,

considering that bone resorption and deposition are

strongly related to mechanical stimuli (Frias et al.,

Soares dos Santos M., A. F. Ferreira J., Ramos A., Pascoal R., Morais dos Santos R., M. Silva N., A. O. Simões J., J. C. S. Reis M., N. Boeri C., Festas

A. and M. Santos P..

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards the Development of a Smart Hip

Prosthesis.

DOI: 10.5220/0003792100710081

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2012), pages 71-81

ISBN: 978-989-8425-91-1

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

2010). Therefore, the smart hip prosthesis should

allow the osteointegration monitoring in the critical

regions of the bone-implant interface (self-diagnosis

function) and mechanically micro-stimulate the local

region under loosening, in order to remodel the bone

surrounding the hip prosthesis (adaptive function).

The failure detection in hip implants and the regions

of the implant where it occurs with time are

currently being studied (Marschner et al., 2009;

Rowlands et al., 2008; Alpuim et al., 2008;

Puers et al., 2000). To yield an architecture to

accommodate energy harvesting systems, activation

circuits to wake up deep sleep electronics, who

relies on batteries to operate, have been developed

to instrument hip prosthesis (Morais et al., 2009),

but they have been powered inductively by external

coils (Morais et al., 2009; Marcelli et al., 2007).

The detection/micro-stimulation system’s operation

demands for an autonomous bio-generation system

that electrically must supply not only the telemetry

but also the mechanical bio-stimulation systems.

Only a vibration bio-generator, designed using the

electromagnetic principle, was proposed to supply

smart hip prosthesis (Morais et al., 2011; Morais

et al., 2010), but the multi-source methodology

of the vibration-based energy harvesting system in

the same hip implant was not approached. When

the availability of the electric power supply is

jeopardized, the reliability of the power supply

mechanism decreases. This paper reports an electric

vibration-based bio-generator from multiple energy

sources for smart hip prosthesis. It is composed of

a piezoelectric and two electromagnetic generation

mechanisms in a same hip joint prosthesis. This

methodology should serve as a reference for future

work on smart hip prosthesis with the purpose of

avoiding revision procedures.

This paper is organized as follows: after this

introductory section, the three generators that make

up the prototype are described, from the theoretical

background to the characterization of the practical

design. The simulated and experimental generation

results are presented and discussed in section 3.

Finally, in section 4 conclusions and ﬁnal remarks of

this work are given.

2 MULTI-SOURCE ELECTRIC

POWER BIO-GENERATORS

2.1 Hip Prosthesis Prototype

To validate the multi-source self-powered electrical

supplying methodology, the commercially available

Metabloc

T M

straight stem system (Zimmer

Corporate, Warsaw, Indiana, EUA) was used as

a model to design a passive hip prosthesis prototype.

The size 10 of this stem system was modiﬁed to

obtain a hollow hip prosthesis prototype. The fatigue

resistance of instrumented hollow bone implants has

already been guaranteed by fatigue tests (Westerhoﬀ

et al., 2009; Heinlein et al., 2007). Figure 1 focuses

on the full active prototype of the hip prosthesis

with vibration-based electric power bio-generators.

The electromagnetic-based transduction mechanisms

were positioned respectively in the body and

in the upper half of the head, whereas the

piezoelectric-based transducer was mounted in

the lower half of the femoral head.

Figure 1: Electric power bio-generators from multiple

energy sources for a hip prosthesis prototype.

2.2 Energy Harvesting Systems

There are many possibilities to harvest electric energy

from the surrounding environment. Biofuel Cells,

magnetic induction, thermoelectric and vibration are

some of the main sources used to harvest energy.

High-quality articles and books have been published

about this subject and highlight how important such

mechanisms can be in the development of smart

bone implants (Lu et al., 2011; Ka´zmierski and

Beeby, 2011; Carmo et al., 2010; Zhu et al., 2010;

Beeby and White, 2010; Westerhoﬀ et al., 2009;

Priya and Inman, 2009; Kerzenmacher et al., 2008;

Wei and Liu, 2008). Vibration-based generation is

currently the most appropriate solution to convert

mechanical vibrations into electrical energy in order

to electrically supply the active elements of the smart

hip prosthesis (Morais et al., 2011; von B¨uren et al.,

2006).

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

2.3 Translation Movement-based

Electromagnetic Energy Harvester

Using mechanical accelerations, obtained from

human gait activities, as the energy source for energy

scavengers, an electromagnetic power transducer

(TEEH) was designed in the body of the hip

prosthesis to take advantage of the movements in

the abduction-adduction and ﬂexion-extension axes

through the hip joint (Winter, 2009; Whittle, 2007).

2.3.1 Theory Background

The principle of the electromagnetic generation of

electric energy is based on Faraday’s law, who

formulated that the relative movement of a coil

within a constant magnetic ﬁeld, generated by a

permanent magnet, induces an electromotive force

in that coil through the change in the magnetic ﬂux

(Ida, 2004). The theory of the linear behaviour

of the generality of the vibration-based generators,

is already well developed (Ka´zmierski and Beeby,

2011; Priya and Inman, 2009; Cook-Chennault et al.,

2008; Gilbert and Balouchi, 2008; Beeby et al.,

2006). As with inertial resonant generators, they

use mechanical movement as the physical condition

to harvest electric energy. Linearly, they can be

modelled as a single degree-of-freedom mechanical

damping system represented by a second-order

mass-spring-damper system, because the damping

mechanism is proportional to the kinetic energy. The

general resonant generator theory states that:

(i) The mechanical input is the external mechanical

vibration (represented below by y t , as in ﬁgure

2). Generally, the human gait analysis has shown

frequency movements between 0.5 Hz and 2 Hz.

The international ISO 14242 standard speciﬁes, for a

frequency of 1 Hz, a pattern of the relative angular

movement between articulating components and a

pattern of the applied force for orbital bearing type

wear testing machines of hip prosthesis. However, the

real loading and displacement parameters are much

more complex than the standardized ones, involving

highly nonlinear functions with several parameters,

such as the frequency of movements, weight, age,

bone structure, health conditions, type of activity,

muscle acceleration imposed by the patient’s activity,

among others;

(ii) The mechanical structure can be modelled as an

inertial frame (ﬁxed referential) where a suspended

seismic mass is attached, representing a magnet,

coupled to a spring, which in turn is coupled to

a damping element, representing the sum of the

comprising parasitic losses and the electrical energy

extracted by the transducer (Beeby et al., 2006).

External movements are transmitted by the inertial

frame, producing a relative displacement z

t between

the mass and the frame (as represented in ﬁgure

2), which is the amount of mechanical vibration to

be converted into electric energy. The diﬀerential

equation of this second order system is described as:

z t cz t kz t my t (1)

in which m is the seismic mass, k the stiﬀness of the

spring and c is the damping coeﬃcient. When this

damping system is excited by an external sinusoidal

vibration y

t Ysin ωt , the solution of (1) is given

by:

z t

2ζω

Ysin ωt arctg

2ζω

(2)

in which ω

k m is the natural frequency and

c 2mω

is the total damping ratio of the

resonant generator system. The damping coeﬃcient

resulting from electromagnetic transduction c

can be approximately achieved (considering only

movements from high to zero magnetic ﬁelds (Beeby

et al., 2006)) by:

NlB

load

coil

jωL

coil

(3)

l is the perimeter of one turn of the coil, N is the

number of turns of the coil, B is the ﬂux density to

which the coil is subjected, and R

load

, R

coil

and L

coil

are respectively the load resistance, coil resistance

and coil inductance;

(iii) The average power output, applying y

Ysin ωt , is deﬁned as:

P t

mζ Y

Ω

1 Ω

2ζ Ω

(4)

in which Ω

ω ω

is a dimensionless parameter

that denotes the diﬀerence between the real frequency

and the natural frequency of the generator system.

Considering V

2PR, the average voltage output

can be estimated by:

BlYω

2ζω

(5)

(iv) In order to optimize the generation of electric

power, the frequency of the hip kinematics must

match the resonant frequency of the generator.

Taking into account that the external vibration,

which is the input of the bio-generator of smart hip

prosthesis, cannot be controlled by the generator,

the matching must be performed by tuning the

resonant frequency of the generator. Although the

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards

the Development of a Smart Hip Prosthesis

vibration-based transducers are currently considered

the best methodology to develop bio-generators from

human gait at hip localization, a continuous matching

of the frequency range of the hip kinematics of

patients with the resonant frequency of the generator

is a diﬃcult task to do, because the duration and

frequency of every-day human activities and the

frequency of the activity itself are unpredictable

(Morlock et al., 2001). With constant resonant

frequency, the best result is obtained when the human

being controls the duration and frequency of the

activity. Even so, a perfect match is not achievable.

This linear model does not take into account

the real mechanical input data (translational and

rotational movements of the hip joint), and several

other forces acting in the system, such as friction and

gravity forces. However, it is a solid basis to start the

study of the dynamic behaviour of this kind of system.

2.3.2 Permanent Magnet Vibration Power

Generator Prototype

The generator prototype is composed of an extension

coil spring (K 2.45 N/m, 5 mm of diameter, 10

mm of steady-state length and 0.2 mm

of wire

section) and two neodymium disc magnets N35 (6

mm of diameter, 6 mm of height, 1.21 g of weight

and 1.22 T of magnetic ﬁeld), which are suspended

inside a Teﬂon tube (c

0.04) where enamelled

copper wire (0. 1 mm of diameter, 27 mm of length

and 1.72

Ωm of electrical resistivity) was

wound (N 2000 turns, 124.4 Ω of total wire

resistance), which in turn was attached to the hip

prosthesis ﬁxture. The coil and the prosthesis make

up the inertial frame. Two magnets were used to

make larger the separation between the poles of the

magnet, in order to ensure that the magnetic ﬁeld

lines were nearly perpendicularto the wiring direction

of the coil, which leads to better results, according

to Morais at al. (2011). Figure 2 shows a 2-D

representation of this bio-generator. The coil was

located as close as possible to the magnetic ﬁeld lines

in order to maximize the change in magnetic ﬂux

and reduce friction between the tube and the magnet.

Taking into account the volume restrictions inside the

prosthesis, the coil was lengthened to maximize the

number of coil turns. The spring was chosen to be

nonferrous to avoid the induction of a current into

the spring from the magnetic ﬁeld generated by the

magnet, preventing magnetic attraction between them

that would stop the relative movement between the

magnet and the coil during the human gait. With this

design, the natural frequency of the electromagnetic

transducer is 4.98 Hz and the total damping ratio is

0.2556.

Figure 2: Scheme of the TEEH transducer.

2.4 Rotation Movement-based

Electromagnetic Energy Harvester

Another electromagnetic transducer was designed

using the modular ball head of the hip prosthesis

and an acetabular component, in order to exploit the

potential of the rotation movement through the hip

joint (REEH transducer).

2.4.1 Theory Background

According the Faraday’s law, the total induced

electromotive force (V

emf

) in a circuit due to motion

is given by the integral of the electric ﬁeld intensity

along the desired contour for the electromotive force,

as given by equation 6 (Ida, 2004).

emf



v B d l (6)

in which

v B is the induced electric ﬁeld intensity,

C is the desired contour (in this case a circle of

4 mm of radius). The electromagnetic energy

harvester based on the rotation movement of the

hip prosthesis uses this same principle, which is

also used to design alternating current generators.

The total harvested energy is the total sum of the

energy that can be harvested from the rotation around

the ﬂexion-extension axis and the energy acquired

from the rotation around the inward-outward or the

abduction-adduction axes, according to the following

expression of the total induced electromotive force:

ˆxˆz

emf

πR

dα

ˆz

sin

ˆz

πR

dα

ˆx

sin

ˆx

(7)

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

Equation 7 demonstrates that the voltage generation is

maximized with the increase of the rotation change,

which means that the higher the frequency the higher

the output voltage. Because the amplitude of rotation

around the ﬂexion-extension is greater than the other

rotations, the induced electromotive force harvested

due to this rotation (V

ˆx

emf

) is predominant.

2.4.2 Permanent Magnet Vibration Power

Generator Prototype

The ball head was hollowed to allow the installation

of a circular winding of enamelled copper wire (AWG

42, 0.063 mm of diameter), which was coiled (N

4710 turns, 682 Ω of total wire resistance, 117.1 m of

total length of the coil, 7.92 mm of average diameter)

around a Teﬂon tube (5.8 mm of diameter, 12 mm

of length) whose core was designed to be a steel

cylinder (4 mm of diameter, 14 mm of length, 100 of

relative permeability). Twenty-four neodymium disc

magnets N52 (6 mm of diameter, 2 mm of height,

0.43 g of weight and 1.48 T of magnetic ﬁeld) were

put inside the structure of an acetabular component

of high density polyethylene to set the magnetic ﬁeld

lines over the volume of the upper half of the ball

head of the hip prosthesis, according to ﬁgure 3.

Six groups of two magnets, positioned equidistantly,

were settled symmetrically in the acetabulum with

six other groups of two magnets, also positioned

equidistantly. Therefore, the rotation around the

ﬂexion-extension axis and around the inward-outward

axis or the abduction-adduction axis (depending on

the location of the magnets) are used as the vibration

source to harvest electric energy. Figure 4 presents

the electromagnetic-based generator to be applied to

the human hip implant.

Figure 3: Acetabular component.

2.5 Piezoelectric Energy Harvester

A piezoelectric power generator (PEH) was designed

to exploit the potential of the axial load over the

hip joint in the harvester process. Figure 5 provides

a representation of the piezoelectric-based generator

applied to the human hip implant.

Figure 4: Scheme of the REEH transducer.

Figure 5: PEH system.

2.5.1 Piezoelectric Power Generator Prototype

A piezoelectric ceramic diaphragm (ref. 7BB-12-9,

muRata Corporate, Kyoto, Japan) with 9 mm of

diameter and 0.22 mm of thickness (12 mm of plate

size, 0.1 mm of plate thickness and 9.0

1.0 kHz

of resonant frequency) was put in the lower half of

ball head of the hip prosthesis, in order to transduce

the large axial load changes during the gait cycle

over it into electric energy (Priya and Inman, 2009).

The maximization of the electric power harvested

requires the matching of the frequency range of the

axial load over the hip joint of patients with the

resonant frequency of the transducer. However, the

unpredictability of the frequency of the axial load

makes this task hard to implement.

2.5.2 Theory Background

General 1-D piezoelectric vibration energy harvesters

can also be modelled as a single degree-of-freedom

mechanical damping system, represented by a

second-order mass-spring-damper system (Renno

et al., 2009; duToit et al., 2005). These

standard models are inaccurate for our purpose

since the total mechanical damping ratio is not

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards

the Development of a Smart Hip Prosthesis

equal to the mechanical damping ratio of the

second-order mass-spring-damper system, because

the piezoelectric element is attached to the hip

prosthesis structure, which settles a new mechanical

damping ratio and a proof mass very diﬃcult to

ﬁnd due to the geometry of the prosthesis. An

artiﬁcial neural network model (Kalogirou, 2000)

was used to overcome this ill-deﬁned problem,

oﬀering an alternative way to predict the power and

energy conversion of this transducer mechanism. A

multilayer ’feed-forward’ neural network (ANN) was

trained to perform the matching between the input

data (a series of pairs of the frequency and amplitude

of sinusoidal axial forces over the head of the hip

prosthesis) and the target data (average power and

peak-to-peak voltage), acquired from experimental

tests. The ANN consists of one input layer, with two

neurons, two hidden layers, with seven neurons each,

and one output layer, with two neurons, as shown in

ﬁgure 6 and equation 8. The Levenberg-Marquardt’s

algorithm was used as the training algorithm and the

mean square error of 1.0

as the convergence

criteria for the network training. Sigmoid functions

(Tansig) for the hidden layers and linear function

(Purelin) for the output layer were used as the transfer

functions.

(8)

in which y

is the output 2

1 matrix, i

is the 2 1

input matrix, IW

is a input weight 7 2 matrix, LW

and LW

are respectively layer weight 7

7 and 2 7

matrices, and b

, b

and b

are respectivelybias 7 1,

1 and 2 1 matrices. f

and f

are respectively

linear and sigmoid functions.

Figure 6: Architecture of the ANN used in modelling for

the average power and peak-to-peak voltage of the PEH

transducer.

3 RESULTS AND DISCUSSION

All experimental procedures were performed with

a mechanical testing machine used to study the

tribological behaviour of materials for hip joint

prosthesis (Santos et al., 2011). Each generator was

independently tested. The generated instantaneous

voltage signal was acquired from the combination of

the amplitude of several rotational and translational

movements and the associated frequency. For

each harvesting element and for each combination

amplitude/frequency, the experimental generated

average and peak power, energy and peak-to-peak

voltage were analysed. These experimental results

were compared with the models reported in sections

2.3.1, 2.4.1 and 2.5.2.

3.1 TEEH Results

A load resistance of 979 Ω was used to enable the

energy transfer of this bio-generator when sinusoidal

input vibrations with amplitudes in the range 10 mm

to 40 mm and frequencies in the range 0.5 Hz to

4 Hz were applied to the generator. Figures 7 and

8 show respectively the results of the experimental

and simulated average power, whereas ﬁgure 9

highlights respectively the results of the experimental

and simulated peak-to-peak voltage. The maximum

energy harvested was 53.7 µJ/s when the sinusoidal

function has an amplitude of 40 mm and a frequency

of 4 Hz. With the same amplitude but with a

frequency of 2.5 Hz, 12.7 µJ/s can still be harvested.

It is clear that energy production is increased by

increasing amplitude and frequency. This harvester

is able to provide 567.4 µW of instantaneous peak

power when the input is excited with an amplitude

of 40 mm and a frequency of 3 Hz.

Figure 7: Experimental average power harvested from the

TEEH transducer.

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

Figure 8: Simulated average power harvested from the

TEEH transducer.

Figure 9: Simulated (dashed line) and experimental (dotted

line) voltage harvested from the TEEH transducer.

3.2 REEH Results

A load resistance of 8.98 kΩ was used to enable

energy transfer of this bio-generator when sinusoidal

rotations in the ﬂexion-extensionaxis with amplitudes

in the range 50

to 70 and frequencies in the range

0.5 Hz to 2.5 Hz, were applied to the generator.

Figures 10 and 11 show the experimental results.

Figures 12 and 13 highlight the simulated results

using 80 mT as the magnetic ﬁeld in the winding

(measured at the ends of the winding). The maximum

energy harvested was 0.77 µJ/s when a sinusoidal

function with an amplitude of 60

and a frequency

of 2.5 Hz was applied. With an amplitude of 70

and a frequency of 1.5 Hz, 0.39 µJ/s can still be

harvested. The increased energy production with

increasing amplitude/frequency is also veriﬁed. This

harvester provides 3.1 µW of instantaneous peak

power when the input is excited with an amplitude

of 60

and a frequency of 2.5 Hz.

Figure 10: Experimental average power harvested from

the REEH transducer (the plus sign refers to peak-to-peak

amplitudes in the range

10 mm to 60 mm, 10 mm

to 50 mm and

10 mm to 40 mm; the square refers to

peak-to-peak amplitudes in the range

20 mm to 50 mm,

20 mm to 40 mm and 20 mm to 30 mm).

Figure 11: Experimental voltage harvested from the REEH

transducer (the plus sign refers to peak-to-peak amplitudes

in the range

10 mm to 60 mm, 10 mm to 50 mm

and

10 mm to 40 mm; the square refers to peak-to-peak

amplitudes in the range

20 mm to 50 mm, 20 mm to 40

mm and

20 mm to 30 mm).

3.3 PEH Results

External sinusoidal forces with amplitudes in the

range 100 N to 250 N and frequencies in the range 0.5

Hz to 4 Hz were applied to the generator. A load of

1 MΩ was used to enable the energy transfer. Figures

14 and 15 show the experimental results, whereas

ﬁgures 16 and 17 highlight the validation results of

the ’feed-forward’ neural network using only data not

used in the training process. The maximum energy

harvested was 0.6 µ J/s when the sinusoidal function

has an amplitude of 200 N and a frequency of 4 Hz.

With an amplitude of 100 N and a frequency of 2.5

Hz, 0.2 µ J/s can still be harvested. Regarding the

instantaneous peak power, this generator can harvest

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards

the Development of a Smart Hip Prosthesis

Figure 12: Simulated average power harvested from the

REEH transducer.

Figure 13: Simulated voltage harvested from the REEH

transducer.

9.1 µW for frequencies of 3.5 Hz and 4 Hz.

Figure 14: Experimental average power harvested from the

PEH transducer.

3.4 Discussion

The concept of energy harvesting from multiple

energy sources was proved in this study as a reliable

Figure 15: Experimental voltage harvested from the PEH

transducer.

Figure 16: Validation of the average power harvested from

the PEH transducer (dash-dot line refers to the network

output).

Figure 17: Validation of the voltage harvested from the PEH

transducer (dash-dot line refers to the network output).

methodology to suﬃce the electrical power needs

of smart hip prosthesis. According to Morais et

al. (2011), a total energy consumption of about

360 µJ with an average power of 1.21 µW is

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

required to power a telemetric system of a hip

prosthesis for a working period of 300 seconds.

Three transducers provide electric energy to supply a

telemetry system, but they also ensure the availability

of the electric supply, underlining the development

of optimized electric power harvesting elements from

multiple energy sources, towards the design of a

new concept of smart hip prosthesis based on its

lifetime extension. However, experimental results

show that each transducer must be optimized in

order to maximize electric generation during typical

walking speeds, namely in the range between 0.5

Hz and 2 Hz, and to allow the osteointegration

monitoring as well as mechanical micro-stimulation.

Due to the tracking performance of the control

operations of the testing machine, a small loosening

between the acetabular component and the femoral

head occurs when the current force is near to zero

in the tracking of sinusoidal trajectories. This

explains the higher values of the voltage harvested

from the PEH transducer for 100 N of amplitude

over higher amplitudes for the same frequency. The

developed models for the electromagnetic transducers

do not perfectly represent all possible dynamics of

the electric power generation process. More accurate

models are being designed to carry out optimization

programs. The piezoelectric transducer must be

redesigned to remain attached to the hip prosthesis

but its dynamic behaviour should not be signiﬁcantly

aﬀected by the mechanical properties of the implant.

Also note that the energy proﬁles must be simulated

and measured under in-vitro and in-vivo realistic

conditions.

Although the power harvested from the TEEH

transducer is dominant over the other transducers, all

the three bio-generators are being optimized in order

to maximize its ability to harvest energy from the

human motion. The development of eﬃcient power

management modules is outside the scope of this

study.

3.4.1 Optimization of the TEEH Generator

The major problem of this transducer is the matching

between the frequency range of the hip kinematics

of patients and the resonant frequency of this

bio-generator. A possible solution to this problem

is the development of a broadband energy harvesting

that must carry out the power maximization over

all the frequency range of the hip joint kinematics.

Another problem of this transducer is the use of a

coil spring, because the spring constant decreases

over time during operation, which can jeopardize the

autonomy of the smart hip prosthesis. Magnetically

levitated generators are a potential solution for this

speciﬁc situation, which must however be designed

for a broadband application.

3.4.2 Optimization of the REEH Generator

A constant magnetic ﬁeld and an average radius

for the coils were assumed in the model presented

in section 2.4.1. However, the magnetic ﬁeld,

produced by the several magnets put in the acetabular

component, is neither constant nor uniform, which

makes expression 7 for the voltage generation a

highly nonlinear function. On the other hand, the

magnets are at a distance from the winding such that

the magnetic ﬁeld over the winding is very low.

Several issues must be considered in order to

maximize the generation of electric power by this

transducer: (1) minimize the distance between the

magnets and the winding; (2) maximize the radius of

the winding and the number of wire turns; (3) setup

concave magnets in order to ensure a uniform

magnetic ﬁeld; (4) setup a core of a material with

very high relative permeability; (5) development of

a transducer design based on a broadband approach.

3.4.3 Optimization of the PEH Generator

The piezoelectric transducer was designed with

only a single piezoelectric element, which resonant

frequency is much higher than the frequency of the

hip joint kinematics. The presented piezoelectric

harvesting is not based on a cantilevered broadband

vibro-impacting power transducer methodology,

which performs high vibration frequencies after

mechanical impacts. Besides, the installation of a

stack of piezoelectric elements with much lower

resonant frequency and performing as a broadband

energy harvesting must be considered in order to

multiply the generation of electric energy.

4 CONCLUSIONS AND FINAL

REMARKS

The development of smart prosthesis is a rising trend

of the concept of instrumented prosthesis. In the

case of the hip prosthesis, expertise methodologies

are claimed to be developed in order to preventively

ﬁx the loosening problem during the lifetime of

the implant in order to avoid revision procedures.

The ﬁrst demand to achieve this goal is the design

of energy harvesting elements to electrically supply

the telemetric systems and the design of active

mechanisms that can preventively ﬁx the loosening

problem. This paper reports the ﬁrst study about

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards

the Development of a Smart Hip Prosthesis

the development of an energy harvesting system

from multiple sources for smart hip prosthesis.

Considering the energy obtained from the movement

as the most abundant in the human body, three

energy harvesting power bio-generators, namely a

piezoelectric-based and two electromagnetic-based

harvesting elements, were designed to harvest energy

from several movements over the femoral component.

They were able to produce energy to supply the

power needs of a telemetric system. This approach

ensures the availability of the electric power supply

and operates autonomously, safely and without

maintenance during the lifetime of the hip prosthesis.

An ongoing optimization of the harvesting elements

is being conducted in order to improve the electric

power bio-generation up to levels required by the

active actuators that would prevent the aseptic

loosening.

ACKNOWLEDGEMENTS

The present work was conducted with the support of

the FCT - Foundation for Science and Technology,

under the PTDC/EME-PME/105465/2008 project.

REFERENCES

Alpuim, P., Filonovich, S., Costa, C., Rocha, P., Vasilevskiy,

M., Lanceros-Mendez, S., Frias, C., Marques, A.,

Soares, R., and Costa, C. (2008). Fabrication of

a strain sensor for bone implant failure detection

based on piezoresistive doped nanocrystalline silicon.

Journal of Non-Crystalline Solids, 354:2585–2589.

Beeby, S., Tudor, M., and White, N. (2006). Energy

harvesting vibration sources for microsystems

applications. Measurement Science and Technology,

17:R175–R195.

Beeby, S. and White, N. (2010). Energy Harvesting for

Autonomous Systems. Artech House, Norwood, MA,

EUA.

Carmo, J., Ribeiro, J., Silva, M., Goncalves, L., and

Correia, J. (2010). Thermoelectric generator and

solid-state battery for stand-alone microsystems.

Journal of Micromechanics and Microengineering,

20:1–8.

Cook-Chennault, K., Thambi, N., and Sastry, A. (2008).

Powering mems portable devices a review of

non-regenerative and regenerative power supply

systems with special emphasis on piezoelectric energy

harvesting systems. Smart Materials and Structures,

17:1–33.

Damm, P., Graichen, F., Rohlmann, A., Bender, A., and

Bergmann, G. (2010). Total hip joint prosthesis for

in vivo measurement of forces and moments. Medical

Engineering & Physics, 32:95–100.

duToit, N., Wardle, B., and Kim, S. (2005). Design

considerations for mems-scale piezoelectric

mechanical vibration energy harvesters. Integrated

Ferroelectrics, 71:121–160.

Frias, C., Reis, J., e Silva, F. C., Potes, J., Sim˜oes, J., and

Marques, A. (2010). Polymeric piezoelectric actuator

substrate for osteoblast mechanical stimulation.

Journal of Biomechanics, 43:1061–1066.

Gilbert, J. and Balouchi, F. (2008). Comparison of energy

harvesting systems for wireless sensor networks.

International Journal of Automation and Computing,

5(4):334–347.

Graichen, F., Bergmann, G., and Rohlmann, A. (1999).

Hip endoprosthesis for in vivo measurement of joint

force and temperature. Journal of Biomechanics,

32:1113–1117.

Heinlein, B., Graichen, F., Bender, A., Rohlmann, A.,

and Bergmann, G. (2007). Design, calibration and

pre-clinical testing of an instrumented tibial tray.

Journal of Biomechanics, 40:S4–S10.

Heinlein, B., Kutzner, I., Graichen, F., Bender, A.,

Rohlmann, A., Halder, A., Beier, A., and Bergmann,

G. (2009). ESB clinical biomechanics award 2008:

Complete data of total knee replacement loading for

level walking and stair climbing measured in vivo with

a follow-up of 610 months. Clinical Biomechanics,

24:315–326.

Ida, N. (2004). Engineering Electromagnetics. Springer,

New York, 2nd edition.

Kalogirou, S. (2000). Applications of artiﬁcial

neural-networks for energy systems. Applied

Energy, 67:17–35.

Ka´zmierski, T. and Beeby, S. (2011). Energy Harvesting

Systems - Principles, Modeling and Applications.

Springer, New York, EUA.

Kerzenmacher, S., Ducr´ee, J., Zengerle, R., and von

Stetten, F. (2008). Energy harvesting by implantable

abiotically catalyzed glucose fuel cells. Journal of

Power Sources, 182:1–17.

Kurtz, S., Ong, K., Lau, E., Mowat, F., and Halpern,

M. (2007). Projections of primary and revision hip

and knee arthroplasty in the united states from 2005

to 2030. The Journal of Bone and Joint Surgery

(American), 89(4):780–785.

Lu, M., Zhang, G., Fu, K., Yu, G., Su, D., and Hu, J. (2011).

Gallium nitride schottky betavoltaic nuclear batteries.

Energy Conversion and Management, 52:1955–1958.

Marcelli, E., Scalambraa, F., Cercenelli, L., and Plicchi,

G. (2007). A new hermetic antenna for wireless

transmission systems of implantable medical devices.

Medical Engineering & Physics, 29:140–147.

Marschner, U., Grtz, H., Jettkant, B., Ruwischa, D.,

Woldt, G., Fischer, W., and Clasbrummel, B. (2009).

Integration of a wireless lock-in measurement of hip

prosthesis vibrations for loosening detection. Sensors

and Actuators A: Physical, 156:145–154.

Morais, R., Frias, C., Silva, N., Azevedo, J., Serˆodio, C.,

Silva, P., Ferreira, J., Sim˜oes, J., and Reis, M. (2009).

An activation circuit for battery-powered biomedical

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

implantable systems. Sensors and Actuators A:

Physical, 156:229–236.

Morais, R., Silva, N., Santos, P., Frias, C., Ferreira, J.,

Ramos, A., Sim˜oes, J., Baptista, J., and Reis, M.

(2010). Permanent magnet vibration power generator

as an embedded mechanism for smart hip prosthesis.

Procedia Engineering, 5:766–769.

Morais, R., Silva, N., Santos, P., Frias, C., Ferreira,

J., Ramos, A., Sim˜oes, J., Baptista, J., and Reis,

M. (2011). Double permanent magnet vibration

power generator for smart hip prosthesis. Sensors

andActuatorsA: Physical, In Press, Corrected Proof.

Morlock, M., Schneider, E., Bluhm, A., Vollmer, M.,

Bergmann, G., M¨uller, V., and Honl, M. (2001).

Duration and frequency of every day activities in total

hip patients. Journal of Biomechanics, 34:873–881.

Priya, S. and Inman, D. (2009). Energy Harvesting

Technologies. Springer, New York, USA.

Puers, R., Catrysse, M., Vandevoorde, G., Collier, R.,

Louridas, E., Burny, F., Donkerwolcke, M., and

Moulart, F. (2000). A telemetry system for the

detection of hip prosthesis loosening by vibration

analysis. Sensors and Actuators A: Physical,

85:42–47.

Renno, J., Daqaq, M., and Inman, D. (2009). On the optimal

energy harvesting from a vibration source. Journal of

Sound and Vibration, 320:386–405.

Rohlmann, A., Graichen, F., Bender, A., Kayser, R., and

Bergmann, G. (2008). Loads on a telemeterized

vertebral body replacement measured in three patients

within the ﬁrst postoperative month. Clinical

Biomechanics, 23:147–158.

Rowlands, A., Duck, F., and Cunningham, J. (2008). Bone

vibration measurement using ultrasound: Application

to detection of hip prosthesis loosening. Medical

Engineering & Physics, 30:278–284.

Santos, M., Rolo, D., Ferreira, J., Ramos, A., and Boeri,

C. (2011). Fuzzy control of a servopneumatic wear

simulator of acetabular components of hip prosthesis

- implementing nonlinear controllers for the angular

movement of three coupled. In Proceedings of the 8th

International Conference on Informatics in Control,

Automation and Robotics, pages 155–160.

von B¨uren, T., Mitcheson, P., Green, T., Yeatman, E.,

Holmes, A., and Tr¨oster, G. (2006). Optimization

of inertial micropower generators for human walking

motion. IEEE Sensors Journal, 6(1):28–38.

Wei, X. and Liu, J. (2008). Power sources and electrical

recharging strategies for implantable medical devices.

Frontiers of Energy and Power Engineering in China,

2(1):1–13.

Westerhoﬀ, P., Graichen, F., Bender, A., Rohlmann, A.,

and Bergmann, G. (2009). An instrumented implant

for in vivo measurement of contact forces and contact

moments in the shoulder joint. Medical Engineering

& Physics, 31:207–213.

Whittle, M. (2007). Gait analysis - An introduction.

Butterworth Heinemann Elsevier, 4th edition.

Winter, D. (2009). Biomechanics and motor control of

human movement. John Wiley & Sons, 4th edition.

Zhu, D., Tudor, M., and Beeby, S. (2010). Strategies for

increasing the operating frequency range of vibration

energy harvesters: a review. Measurement Science

and Technology, 21:1–29.

MULTI-SOURCE ENERGY HARVESTING POWER GENERATORS FOR INSTRUMENTED IMPLANTS - Towards

the Development of a Smart Hip Prosthesis