scieee Science in your language
[en] (orig)

Developing Soft Bio-Inspired Cooperation Methods and Mechanisms for Flapping Wings Aerial Robots (FWAR)

Author: Pérez-Sánchez, Vicente
Year: 2023
Source: https://idus.us.es/bitstreams/804c55dc-d53c-455c-8ca4-2362daa81371/download
P oyec o Fin de Ca e a
Ingenie ía de Telecomunicación
Fo ma o de Publicación de la Escuela Técnica
Supe io de Ingenie ía
Au o : F. Ja ie Payán Some
Tu o : Juan José Mu illo Fuen es
Dep. Teo ía de la Señal y Comunicaciones
Escuela Técnica Supe io de Ingenie ía
Uni e sidad de Se illa
Se illa, 2013
Tesis Doc o al
Ingenie ía Au omá ica, Elec ónica y de
Telecomunicación
De eloping So Bio-Inspi ed
Coope a ion Me hods and Mechanisms
o Flapping Wings Ae ial Robo s (FWAR)
Au o : Vicen e Pé ez Sánchez
Di ec o es: Begoña C. A ue Ullés
Anibal Olle o Ba u one
Ingenie ía de Sis emas y Au omá ica
Escuela Técnica Supe io de Ingenie ía
Uni e sidad de Se illa
Se illa, 2023
Tesis Doc o al
Ingenie ía Au omá ica, Elec ónica y de Telecomunicación
De eloping So Bio-Inspi ed Coope a ion Me hods and
Mechanisms o Flapping Wings Ae ial Robo s (FWAR)
Au o :
Vicen e Pé ez Sánchez
Di ec o es:
Begoña C. A ue Ullés
P o eso Ti ula
Anibal Olle o Ba u one
Ca ed á ico
Ingenie ía de Sis emas y Au omá ica
Escuela Técnica Supe io de Ingenie ía
Uni e sidad de Se illa
2023
Tesis Doc o al:
De eloping So Bio-Inspi ed Coope a ion Me hods and Mechanisms o
Flapping Wings Ae ial Robo s (FWAR)
Au o : Vicen e Pé ez Sánchez
Di ec o es: Begoña C. A ue Ullés
Anibal Olle o Ba u one
El ibunal nomb ado pa a juzga la Tesis a iba indicada, compues o po los siguien es
doc o es:
P esiden e:
Vocales:
Sec e a io:
acue dan o o ga le la cali icación de:
El Sec e a io del T ibunal
Fecha:

A mi amilia
A la amilia que e da la ida
A mis compañe os
A mis p o eso es
A mis amigos
A Ana
Acknowledgemen s
De eloping
a Ph.D. hesis is challenging and implies yea s o wo k and s uggle. Fi s ,
hank he G upo de Robó ica Visión y Con ol (GRVC) o unding my esea ch.
I can no imagine his p ocess wi hou my ad iso s, Anibal Olle o and Begoña A ue,
p o iding hei eedback and helping me o imp o e my esea ch esul s. I wan o hank
my lab colleagues. They made i possible o a o d his pe iod wi h good momen s and
con ibu ed o my esea ch wi h hei opinions. I wan o gi e a special men ion o my old
collabo a o s: Ale,Malgani, Rica do, Manu, Emanuela, Ba is a, Ma co, and Pablo. They
pu up wi h my bad momen s and encou aged me o con inue, con ibu ing conside ably
o my wo k. The pe iod o he hesis is leng hy, and also new colleagues suppo he las
pa o his disse a ion: hanks o Alejand o, Albe o, Ca los, Jo ge, Diana, Hong Fan,
Pol, Kike and Juanma. I wan o hank also o my colleagues a Labo a o y o In elligen
Sys ems (LIS) a he École Poly echnique Fédé ale de Lausanne (EPFL). The ime I
spen was amazing, lea ning om di e en cul u es and esea che s. I wan o hank he
esea che s who p o ided a solid baseline o ob ain he esul p esen ed in his esea ch. I
would also like o hank all he eache s who con ibu ed o inc easing my knowledge.
The doc o al hesis is jus a pe iod o li e. Howe e , you educa ion gi es you he
esilience o ace i . I wan o hank my amily o p o iding me wi h he educa ion ha
makes i possible and suppo ing me o e he yea s. Thanks o my a he o eaching
me he meaning o wo k. Thanks o my mo he o s essing he impo ance o he s udy.
Thanks o my sis e o being an example o me. I wan o make a special men ion o my
g andpa en s, who passed on o me he meaning o being a good pe son. I would also like
o hank he amily ha gi es you li e. Thanks o my iends. I wan o apologize o he
ime my esea ch s ole om me wi h hem.
Finally, I wan o hank Ana o suppo ing me du ing his long pe iod. I has no been
easy a imes, bu we ha e made i . We ne e know wha he u u e has in s o e o us, bu
his doc o al hesis is o you.
Vicen e Pé ez Sánchez
Se illa, 2023
III
XAc onyms
LiPo Li hium Polyme . 87
LIS Labo a o y o In elligen Sys ems. 12, III
LMPA Low Mel ing Poin Alloy. 7
MAV Mic o Ae ial Vehicles. 131, 132
MFC Mac o Fibe Composi e™. 8, 33–35, 92–104, 106, 108, 142, 159, 167, VII
MoCap Mo ion Cap u e. 139, 145
MOSFET Me al–Oxide–Semiconduc o Field-E ec T ansis o . 86–88, 105, 134
NASA Na ional Ae onau ics and Space Adminis a ion. 8, 33
PD P opo ional De i a i e. 134
PDMS Polydime hylsiloxane. 32
PLA Polylac ic Acid. 59, 67–71, 159
RC Radio Con ol. 104, 105, 143, 145
RPAS Radio Pilo ed Ai c a Sys ems. 1
SMA
Shape Memo y Alloys. 7, 8, 29, 32, 33, 40–43, 46, 47, 50, 57, 73–75, 78–84,
86–88, 119, 122, 127, 129, 135, 158, 163, VII
SMP Shape Memo y Polyme s. 7
TPU The moplas ic Polyu e hane. 59, 62, 67–71, 73, 76, 127, 159
UAV Unmanned Ae ial Vehicles. 1–5, 8, 11, 27, VII
WWI Wo ld Wa I. 1
WWII Wo ld Wa II. 1

Sho Con en s
Resumen V
Abs ac VII
Ac onyms IX
Sho Con en s XI
1 In oduc ion 1
1.1 Mo i a ion and Objec i es 1
1.2 Thesis F amewo k 12
1.3 Thesis Ou line 13
1.4 Con ibu ions 14
2 Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR) 19
2.1 Bioinspi a ion 19
2.2 O ni hop he s Flapping Wings Ae ial Robo s(FWAR) 26
2.3 So Robo ics Technologies 31
3 Bio-Inspi ed Manipula o s o O ni hop e s 37
3.1 Shape Memo y Alloy(SMA) A i icial Muscles 41
3.2 Analysing he Fo ces In ol ed in he Pe ching 51
3.3 Claws Mo phology 55
3.4 Leg Mo phology 72
3.5 Ac ua ion Me hod and In eg a ion 82
4 Bio-Inspi ed Mo phing Su aces 91
4.1 Mo phing Tail 92
4.2 Mo phing Wing 108
5 Expe imen al Valida ion 121
5.1 Bio-Inspi ed Claws Expe imen s 121
5.2 So Ae odynamic Su aces Expe imen s 139
6 Conclusions and Fu u e Wo ks 157
6.1 Conclusions 157
XI
XII Sho Con en s
6.2 Fu u e Wo ks 162
Lis o Figu es 163
Lis o Tables 167
Bibliog aphy 169
Con en s
Resumen V
Abs ac VII
Ac onyms IX
Sho Con en s XI
1 In oduc ion 1
1.1 Mo i a ion and Objec i es 1
1.2 Thesis F amewo k 12
1.3 Thesis Ou line 13
1.4 Con ibu ions 14
2 Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR) 19
2.1 Bioinspi a ion 19
2.1.1 Bi ds Manipula ion Capabili ies 21
2.1.2 Bi ds Pe ching Capabili ies 23
2.1.3 Bi ds Human In e ac ion Capabili ies 25
2.2 O ni hop he s Flapping Wings Ae ial Robo s(FWAR) 26
2.3 So Robo ics Technologies 31
3 Bio-Inspi ed Manipula o s o O ni hop e s 37
3.1 Shape Memo y Alloy(SMA) A i icial Muscles 41
3.1.1 SMAs P ope ies 41
3.1.2 Modeling SMA Muscles 44
Muscles dynamic model 44
Muscles elec ic and ene gy model 46
3.1.3 De eloping SMA Muscles 50
3.2 Analysing he Fo ces In ol ed in he Pe ching 51
3.3 Claws Mo phology 55
3.3.1 Claws Concep ualiza ion 55
3.3.2 Claws Design 58
3.3.3 Claws Design E olu ion 67
3.4 Leg Mo phology 72
3.5 Ac ua ion Me hod and In eg a ion 82
3.5.1 SMAs Ac ua ion Con igu a ion 82
XIII
XIV Con en s
3.5.2 Tendon D i en Ac ua ion 83
3.5.3 Ac ua ion Con ol 86
4 Bio-Inspi ed Mo phing Su aces 91
4.1 Mo phing Tail 92
4.1.1 MFC 93
Modeling MFC 95
4.1.2 Tail de elopmen 101
Design 102
Manu ac u ing 103
Con ol Elec onics 104
Feas ibil y Tes 105
Tail Compa isons 107
4.2 Mo phing Wing 108
4.2.1 Design 110
4.2.2 Manu ac u ing P ocess 115
5 Expe imen al Valida ion 121
5.1 Bio-Inspi ed Claws Expe imen s 121
5.1.1 G asping Expe imen a ion 122
5.1.2 Pe ching Expe imen a ion 132
5.2 So Ae odynamic Su aces Expe imen s 139
5.2.1 Bio-Inspi ed Tail Expe imen al Valida ion 139
Kinema ic analysis 139
Ou doo s expe imen s 142
5.2.2 Bio-Inspi ed Wings Expe imen al Valida ion 145
Wing es bench expe imen s 146
F ee ligh expe imen s 150
6 Conclusions and Fu u e Wo ks 157
6.1 Conclusions 157
6.2 Fu u e Wo ks 162
Lis o Figu es 163
Lis o Tables 167
Bibliog aphy 169
1 In oduc ion
Se una pe sona è pe sis en e, anche se di icile da capi e, sa à
in elligen e; e anche se è debole, di en e à più o e.
Leona do Da Vinci, (1452-1519)
1.1 Mo i a ion and Objec i es
The p esence o d ones in ou li es has inc eased exponen ially o e he las ew yea s.
A i s , d ones we e concei ed as mili a y weapons. Today, d ones a e p esen in ou
daily li es. This p esence gene a es challenging objec i es in he in e ac ion wi h his
ae ial pla o m. Howe e , hey a e c ucial in anspo , su eillance, main enance, and
indus ial inspec ion asks. Mo eo e , d ones p o ide a new ision o he wo ld, a oiding
isky si ua ions and educing anspo ime.
Mili a y applica ions we e he s a ing poin o d ones. The i s eco d o d one use
was in 1849 [
1
] when he Aus ian o ces launched 200 incendia y balloons a Venice.
Due o he lack o con ol o hese pla o ms and he wind gus s o ha day, mos did no
each hei objec i es. Ne e heless, a e his i s a emp , mos o he o ces in he wo ld
ealized he po en ial use o hese pla o ms, and sys ems e ol ed quickly.
The i s miles one o d ones’ e olu ion was he de elopmen o Radio Pilo ed Ai c a
Sys ems (RPAS). Leona do To es Que edo, in 1903[
2
,
3
], used a adio sys em o con ol
he ajec o y o an ai ship o he i s ime (see Figu e 1.1a. This concep se s he base o
d ones, and i is cu en ly main ained. Howe e , be o e his de elopmen , he absence o
con ol o e hei ajec o y made impossible hei use e ec i e.
Du ing 1914-1915, he e o made du ing he Wo ld Wa I (WWI) concludes de eloping
he base o he i s Unmanned Ae ial Vehicles (UAV). Figu e 1.1b ep esen s he i s
concep o d ones de eloped du ing he WWI.
The pla o ms we e imp o ed and gained p ominence du ing WWI and Wo ld Wa
II (WWII). As a esul , hey we e c ucial in mili a y usages like su eillance asks and
localiza ion.
All he knowledge lea ned in he mili a y use o UAVs was di ec ly e lec ed in ci il
applica ions o hese sys ems. In 2006, he Fede a ion o A ia ion Adminis a ion (FAA)
egis e ed he i s comme cial d one pe mi . The mo i a ions o begin he ci il applica ions
o d ones we e hei po en ial applica ions. Ini ially, hese mo i a ions we e hei use in
su eillance, inspec ion, and helping in na u al disas e s.
1

2Chap e 1. In oduc ion
(a)
Telikino i s adio con ol sys em
Leona do To es Que edo 1901[4].
(b)
Hewi -Spe y Au oma ic Ai plane in
1918 [5].
Figu e 1.1 Ea ly d ones.
A ha momen , he classi ica ion o UAVs con empla es h ee ypes o d ones: heli-
cop e s, mul i- o o s, and ixed-wings.
•
Helicop e s:(see Figu e1.2a). They ha e a high payload and can ho e . Also, hei
ligh is s able and e icien .
•
Mul i- o o s:(see Figu e1.2b). As helicop e s, hey can ho e . Howe e , hese
pla o ms ha e highe maneu e abili y han helicop e s. In addi ion, hese ae ial
sys ems can lose one o hei p opelle s con inuing wi h he ligh . Tha makes i
possible o ope a e wi h hese pla o ms in smalle a eas. Ne e heless, hey a e less
ene gy e icien han ixed wings.
•
Fixed wings:(see Figu e1.2c) These pla o ms can glide. This capabili y makes i
possible o co e highe a eas was ing less ene gy. Howe e , hei high ligh speed
makes using his pla o m in con ined spaces di icul .
(a) Hellicop e [6]. (b) Mul icop e [7]. (c) Fixed Wing [8].
Figu e 1.2 D ones clasi ica ion.
Since 2006, UAVs ci ilian applica ions ha e expe ienced exponen ial g ow h. The
i s one was he use UAVs o ideo eco ding. This applica ion is one o he mos
s aigh o wa d. I in ol es being capable o collec ing images in he ae ial pla o m while
con olling he ajec o y o he pla o m. The i s me hod o know he cou se o he
d ones was he Global Posi ioning Sys em (GPS).
A ha momen , ci ilian d ones only use GPS o con ol hei posi ion. Bu , la e , he
minia u iza ion o he sys ems makes i possible o use mul iple senso s o imp o e he
ecogni ion o he en i onmen . Howe e , hese da a ha e o be ea ed o be use ul. Da a
usion is one o hese p ocesses.
1.1 Mo i a ion and Objec i es 3
Da a Fusion combines he in o ma ion gi en by di e en senso s o unde s and he
en i onmen a ound he ae ial pla o m. The came a is one o he p ima y senso s o
ecognize he en i onmen a ound he UAV. The p ocess me ges his da a wi h he da a o
he ligh senso s (al ime e s, gy oscopes, Ine ial Measu emen Uni (IMU), and GPS). I
makes i possible o imp o e ecogni ion o he en i onmen a ound he pla o m. Su eys
like [
9
] highligh he impo ance o da a usion p ocesses. No ably, he au ho s ocus on
using came as and he in luence o compu e ision in Fligh Con ol, Na iga ion, and
Guidance.
The ollowing lines summa ize examples o he leading ision echniques used in
da a usion p ocesses. Rega ding na iga ion, he isual se ing me hod [
10
] gi es he
posi ion and o ien a ion o he ae ial pla o m as a unc ion o he a ge . [
11
,
12
,
13
]
shows examples o isual se oing hecniques. SLAM [
14
] is one o he main me hods
in na iga ion using isual odome y and mapping [
15
,
16
,
17
]. Obs acle a oidance [
18
]
combines ligh con ol wi h na iga ion and objec ecogni ion [
19
,
20
]. Ta ge acking
[
21
] is ano he na iga ion echnique by he ecogni ion o he objec i e[
22
,
23
]. These
echniques we e de eloped be o e hei use on d ones. Howe e , hei use o d ones makes
hese me hods gain p ominence. These echniques a e c ucial o he use o d ones in
unknown si ua ions. They make i possible o eac o unce ain ies in he en i onmen .
Figu e 1.3 Da a usion p ocess.
Da a usion enabled me ging he in o ma ion o ecognize he en i onmen a ound he
ae ial pla o m and imp o e he con ol skills o he UAVs. The e o e, i makes achie able
he in e ac ion wi h he en i onmen . Fi s ly, his opic ies only o analyze he con ac . The
aim was o in es iga e he con ac and he pe u ba ion gene a ed by he en i onmen in he
en i e pla o m. Then, howe e , a new e a in ae ial obo ics was bo n:Ae ial Manipula ion
(AM).
A he beginning o his e a, he manipula ion was es ed using helicop e s ha ca y
some cus omized end e ec o s [
24
,
25
](Figu e 1.4a). The capabili ies o helicop e s o
4Chap e 1. In oduc ion
ca y ou highe loads and maneu e abili y made hem he i s candida e o de elop
hese asks. Then, esea che s used adi ional indus ial a ms o o e come he limi ed
capabili ies o he i s cus omized a ms in [
26
,
27
]. Howe e , he hea y weigh o hese
a ms was limi ing he con ol capabili ies o he pla o ms. Mo eo e , he helicop e s’
ligh pe o mance was limi ing hese wo ks. Thei big o o s made i challenging o wo k
in con ined a eas. Also, he signi ican in luence o he en i onmen in he ligh o hese
was causing ailu es.
Then, he objec i e was o de elop manipula ion asks on smalle mul i- o o s. This
pla o m has edundan p opelle s ha make he ligh sa e . Also, hese p opelle s a e
smalle , educing he in luence o he en i onmen on he ligh . Howe e , hei payload is
smalle . Then, esea che s ocus on coope a ion o ca y ou highe loads. The aim was o
dis ibu e he weigh among he pla o ms in ol ed. The wo k p esen ed in [
28
] shows
one o he i s examples o helicop e coope a ion o ca y ou highe loads ou doo s
unde windy condi ions. The au ho s in [
29
] show an example o hese applica ions using
mul i o o s indoo s.
(a)
Yale ae ial manipula o
om [25].
(b)
Helicop e using a 7 De-
g ees o F eedom a m
om [27].
(c)
Coope a i e mul i UAV
anspo a ion om [29].
Figu e 1.4 Ea ly manipula o s.
The esea ch wo ld hen ex ended collabo a i e solu ions o ca y ou a highe load.
Howe e , manipula ion is no only o ca y ou loads. I is also o g asp he objec s and
pe o m manipula ion wi h hem. In ha way, he necessi y o de eloping adap ed a ms
and cus omized end-e ec o s o ae ial obo ics con inued.
Size and weigh we e he main p oblems o he ea ly manipula o s. Thei size caused
dis u bances h oughou he ligh , and he weigh made using hem on ligh e pla o ms
impossible. These cha ac e is ics p e en he wo k on smalle spaces. Thus, he s udy o
smalle cus omized manipula o s s a s. [
30
] de eloped a manipula o o a mini UAV.
Wo ks such as [
31
] analyzed he in luence o he manipula o in he ligh . These wo ks
made i possible o each new ho izons in ae ial manipula ion. Subsequen ly, wo ks such
as [
32
] combined he d one’s con ol, knowing he pe u bances gene a ed by a obo ic a m
wi h se en Deg ees o F eedom (DoF) obo ic a m. In [
33
], he au ho s s udied he use o
complian obo ics a ms o educe he dis u bances gene a ed by he weigh and possible
impac s. Figu e 1.5 shows some examples o manipula o s conside ing he en i onmen ’s
in luence on he pla o m.
1.1 Mo i a ion and Objec i es 5
(a)
7 DoF manipula o om
[32].
(b) Complian manipula o om [33].
Figu e 1.5 Manipula o s’ publica ions o educe hei in luence on he ligh .
The basis o AM was se a ha ime. Wo ks like [
34
,
35
,
36
] made a conside able e o
o egis e hese e olu ions.
Howe e , ae ial obo ics con inues i s e olu ion. The e o e, new equi emen s appea
in e ac ion be ween obo s and humans. T adi ionally, obo ics was isola ed om humans;
an excellen example o his isola ion is using obo ic a ms in a ac o y. They used o be
sepa a ed in o cells o a oid damage o humans (see Figu e 1.6a). Howe e , i is impossible
o isola e an ae ial obo . So hen, a end s a ed de eloping sa e ae ial obo s ha made
in e ac ions wi h humans o animals possible, a oiding isk (see Figu e 1.6b). Figu e 1.6
shows an example o adi ional obo ics sys ems agains he u u e o ae ial manipula ion.
(a) Ac ual obo ics a ms in a ac o y [37]. (b)
Fu u e obo ics so ,
bioinspi a ion and
in e ac ion[38].
Figu e 1.6 T adi ional manipula ion and u u e ae ial manipula ion.
Then, esea che s p esen ed wo ways o educing he isk du ing he UAVs ope a ion.
The i s solu ion comp ises he ae ial pla o m. Mul i- o o s we e, a ha ime, he mos
used pla o ms o coope a ion. They ha e sha p p opelle s ha b eak he ai o gene a e
Th us and main ain he ligh . Howe e , hese sha p p opelle s can easily damage he
coope a o s. The e o e, esea che s looked o some solu ion o make assumable he isk
in coope a ion wi h he pla o m.
This solu ion comp ises he isola ion o ha isk. The idea was o main ain he ligh
pe o mance and isola e he p opelle s. In ha way, se e al esea che s ied o de elop
12 Chap e 1. In oduc ion
(a) Bioinspi a ion. (b) Coope a ion.
(c) So -Flap.
Figu e 1.12 Thesis objec i es.
1.2 Thesis F amewo k
This esea ch wo k has been unded unde he amewo k o Gene al complian ae ial
Robo ic manipula ion sys em In eg a ing Fixed and Flapping wings o INc ease ange
and sa e y (GRIFFIN) [
70
] and AERIAL COgni i e in eg a ed mul i- ask Robo ic sys em
wi h Ex ended ope a ion ange and sa e y (AERIAL-CORE) [
71
] p ojec s. GRIFFIN is an
Eu opean Resea ch Council (ERC) Ad anced G an hos ed by he Uni e si y o Se ille
wi h g an ag eemen ID 788247. This p ojec ies o o e come he capabili ies o he
o ni hop e s, gi ing hem manipula ion capabili ies and imp o ing ligh abili ies.
AERIAL-CORE is a Ho izon 2020 (H2020) p ojec coo dina ed by he Uni e si y
o Se ille wi h g an ag eemen ID 871479. This p ojec in ol ed a big conso ium o
di e en Eu opean uni e si ies. AERIAL-CORE p ojec aims o imp o e he condi ions
o inspec ing and main aining powe lines. The p ojec p oposes using ae ial ehicles o
mi iga e he isks o heigh wo ks in powe lines.
The au ho ca ied ou an s ay a Labo a o y o In elligen Sys ems (LIS) École Poly-
echnique Fédé ale de Lausanne (EPFL) o s udy no el y so elec o-adhesion ac ua o s o
con ibu e o he manipula ion. These ac ua o s aim o con ol he adhesion o he su ace
ac i ely, imp o ing he g ip gene a ed by he manipula o . This s udy is c ucial o imp o e
he balance du ing he pe ching. Howe e , hei applica ion in agg essi e en i onmen s is
s ill unde de elopmen .

1.3 Thesis Ou line 13
1.3 Thesis Ou line
The ollowing lis p esen s he hesis s uc u e summa izing each chap e ’s con en .
•
Chap e 2: This chap e es ablishes he ounda ion o he de elopmen s in his
hesis. Fi s , i ocuses on he a ian capabili ies aimed o mimic in he ae ial obo .
Then con inues wi h analyzing he lying condi ions o he ae ial pla o ms used.
Finally, he chap e desc ibes he so obo ics echnologies used, ocusing mainly
on ac ua ion echnologies and he wo ks ound in he s a e o he a .
•
Chap e 3: The chap e ocuses on de eloping sys ems o p o ide manipula ion
o o ni hop e s. Fi s , i s a s desc ibing, analyzing, and cha ac e izing he so
ac ua ion echnologies used in he manipula o . I hen discusses he design p ocess
and e olu ion o he manipula o h ough a ious e sions. Con inuously, he chap e
p esen s he in eg a ion o he ac ua ion sys ems o each he desi ed speci ica ions.
Finally, a he end o he chap e , he in eg a ion o he manipula o in o he ae ial
pla o m is shown.
•
Chap e 4: The ligh capabili ies o o ni hop e s limi he es o he sys ems used.
In ha way, he chap e ocuses on de eloping so bioinspi ed ae odynamic su aces
o imp o e he ac ual ligh capabili ies o o ni hop e s. The chap e di ides he
de elopmen s in o he wings and ail. Conce ning he ail, an inno a i e solu ion
is p esen ed using mo phing so ac ua o s. This sec ion will desc ibe he design
p ocess, ac ua o s used, and in eg a ion in he pla o m. On i s side, he wings p esen
a no el y geome y a ian-inspi ed, capable o gene a ing ac ua ion on he ligh and
imp o ing he gliding capabili ies o adi ional o ni hop e s. The sec ion desc ibes
he geome ical analysis, designing p ocess, manu ac u ing, and in eg a ion o he
wings in he ae ial pla o m.
•
Chap e 5: This chap e p esen s he expe imen al alida ion o he sys ems de-
eloped. The expe imen al alida ion will assess he ligh beha io , manipula ion
capabili ies, and sa e y added o he pla o m by he so bio-inspi ed de ices de el-
oped in his hesis.
•Chap e 6: P esen s he au ho ’s conclusions and u u e esea ch line o he wo k.
Figu e 1.13 p esen s he in e ela ionship be ween he di e en chap e s.
All he chap e s con ibu e o de eloping he so coope a i e o ni hop e concep
highligh ed in he cen e o he igu e. Chap e 2 se s he objec i es wan ed o eplica e
om na u e and se s he limi a ion and ini ial condi ions o he de elopmen s. Chap e 3
and 4 o e comes he limi a ions ound in Chap e 2. These chap e s de elop and design he
de ices o imp o e he ligh s age o he pla o m and p o ide manipula ion capabili ies.
Chap e 5 e alua es he sys em’s beha io de eloped o each he desi ed objec i es. Finally,
Chap e 6 p esen s he au ho ’s conclusions abou he wo k and de elopmen s pe o med
du ing he in es iga ion. A he end o his chap e , he au ho will p esen he u u e
esea ch lines ha will imp o e he solu ions ca ied ou .
14 Chap e 1. In oduc ion
Figu e 1.13 Thesis Ou line.
1.4 Con ibu ions
Fi e esea ch wo ks ha e been published in my Ph.D. The ollowing lis summa izes he
main con ibu ions o his disse a ion.
•
De elopmen o so adap able manipula o s capable o in e ac ing wi h humans
and pe ching on a ious places.
•
De elopmen o so bioinspi ed mo phing su aces o imp o e he ligh capabili ies
o o ni hop e s.
•S udy o low-weigh so echnologies o be used in ae ial obo ics.
•S udy o he o ni hop e s pe ching o gene a e a ian-inspi ed solu ions.
The ollowing lis o ganizes he publica ions ch onologically s a ing om he las one
and showing hei i le and abs ac s.
1.4 Con ibu ions 15
[
72
] High-Pe o mance Mo phing Wing o La ge-Scale Bio-Inspi ed Unmanned
Ae ial Vehicles.
IEEE Robo ics and Au oma ion Le e s (June 2022).
E. Sa as ano, V. Pe ez-Sanchez, B. A ue, and A. Olle o.
This le e p oposes a no el bio-inspi ed wing design o imp o e some cha ac e is ics o
Flapping Wing Unmanned Vehicles (FWUV) ela ed o hei po en ial applica ions such
as payload capabili y, maneu e abili y, low inju y isk, and ene gy imp o emen . The
sugges ed solu ion akes ad an age o a b oadly based a ian esea ch, ocusing on he
in eg a ion o ad anced bi d-like ea u es anging om bi d-like high-complian ai oil
o ac i e mo phing s a egies o mimic he ways bi ds na u ally manage ligh s. The
p oposed concep ual design is suppo ed by Uns eady Vo ex La ex Me hod simula ions
p o ided by a no el lapping-o ien ed open-sou ce sol e . P o o ype alida ion elies
on bo h es -bench and eal- ligh s esul s. A mo ion cap u e sys em is used o alida e
he wing ae o-elas ic cha ac e is ics. The esul ed imp o emen s a e highligh ed by a
comp ehensi e compa ison wi h di e en simila -size p o o ypes.
[
73
] Analysis o Fo ces In ol ed in he Pe ching Maneu e o Flapping-Wing Ae ial
Sys ems and De elopmen o an Ul a-Ligh weigh Pe ching Sys em
2021 In e na ional Con e ence on Unmanned Ai c a Sys ems (ICUAS) (July 2021).
V. Pé ez-Sánchez, A. Gómez-Tamm, F. Ga cía-Rubiales, B. A ue, and A. Olle o
T ying o op imize he design o ae ial obo ics sys ems, his wo k p esen s an op imized
low-weigh landing sys em o lapping-wing ae ial obo s. The design, based on he use
o low-sized neodymium magne s, in ends o p o ide ha hese ae ial obo s ha e he
capabili y o landing in es ic ed a eas by using he p esen ed solu ion. This capaci y will
inc ease he applica ion ange o hese obo s. A s udy o his si ua ion has been done o
analyze he pe ching maneu e o ces and e alua e he sys em. The solu ion p esen ed
is low-weigh , low-sized, and also ela i ely inexpensi e. The e o e, his solu ion may
apply o mos o ni hop e obo s. Design, analysis o he implied o ces, de elopmen
and expe imen al alida ion o he idea a e p esen ed in his wo k, demons a ing ha
he de eloped solu ion can o e come he o ni hop e ’s payload limi a ion p o iding an
e icien and eliable solu ion.
[
74
] Bio-inspi ed mo phing ail o lapping-wings ae ial obo s using mac o ibe
composi es
Applied Sciences (Ma ch 2021).
V. Pe ez-Sanchez, A. E. Gomez-Tamm, E. Sa as ano, B. C. A ue, and A. Olle o
The aim o his wo k is o p esen he de elopmen o a bio-inspi ed app oach o a
obo ic ail using Mac o Fibe Composi es (MFC) as ac ua o s. The use o his echnology
16 Chap e 1. In oduc ion
will allow achie ing close o he na u e app oach o he ail, aiming o mimic a bi d ail
beha io . The ail will change i s shape, pe o ming mo phing, p o iding a new ype
o ac ua ion me hodology in lapping con ol sys ems. The wo k is in ended as a i s
s ep o demons a ing he po en ial o hese echnologies o being applied in o he pa s
o he ae ials obo ics sys ems. When compa ed wi h adi ional ac ua ion app oaches,
one key ad an age ha is gi en by he use o MFC is hei abili y o adap o di e en
ligh condi ions ia geome ic ailo ing, imi a ing wha bi ds do in na u e. Theo e ical
explana ions, design, and expe imen al alida ion o he de eloped concep using di e en
me hodologies will be p esen ed in his pape .
[
75
] SMA ac ua ed low-weigh bio-inspi ed claws o g asping and pe ching using
lapping wing ae ial sys ems
2020 IEEE/RSJ In e na ional Con e ence on In elligen Robo s and Sys ems (IROS) (Feb u-
a y 2021).
A. E. Gomez-Tamm, V. Pe ez-Sanchez, B. C. A ue, and A. Olle o,
Taking inspi a ion om na u e, he wo k p esen ed in his pape aims o de elop bio-
inspi ed claws o be used o g asping and pe ching in lapping-wing ae ial sys ems. These
claws can be 3D p in ed ou o wo di e en ma e ials and will be capable o adap o
any shape. Also, hey will be so o a oiding undesi ed damages on he objec s when
pe o ming manipula ion. These claws will be ac ua ed by shape memo y alloys (SMA)
sp ings o ge id o he weigh o adi ional se os. The design o all he componen s will
be explained in his wo k. Also, he challenges o being able o con ol SMA using only a
LiPo ba e y on an ae ial ehicle will be exposed. The solu ions applied and elec onics
used will be also desc ibed. Las ly, expe imen s made bo h in es bench as on ligh will
be summa ized.
[
76
] A linea ized model o an o ni hop e in gliding ligh : Expe imen s and simula-
ions
2020 IEEE In e na ional Con e ence on Robo ics and Au oma ion (ICRA) (Sep embe
2020)
R. Lopez-Lopez, V. Pe ez-Sanchez e al.
This wo k s udies he accu acy o a simple bu e ec i e analy ical model o a lapping-
wings UAV in longi udinal gliding ligh con igu a ion compa ing i wi h expe imen al
esul s o a eal o ni hop e . The ae odynamic o ces a e modeled ollowing he linea ized
po en ial heo y o a la pla e in gliding con igu a ion, ex ended o lapping-wing episodes
modeled also by he (now uns eady) linea po en ial heo y, which a e s udied nume ically.
In he gliding con igu a ion, he model eaches a s eady-s a e descen a gi en e minal
eloci y and pi ching and gliding angles, go e ned by he wings and ail posi ion. In he
lapping-wing con igu a ion, i is no iced ha he ehicle can inc ease i s ligh eloci y
1.4 Con ibu ions 17
and pe o m climbing episodes. A ealis ic simula ion ool based on Un eal Engine 4
was de eloped o isualize he e ec o he ail posi ion and lapping equencies and
ampli udes on he o ni hop e ligh in eal ime. The pape also includes he expe imen al
alida ion o he gliding ligh and he da a has been eleased o he communi y.

2 Bio-Inspi ed So Flapping Wing
Ae ial Robo s(FWAR)
This
chap e summa izes he main concep s, limi a ions, and echnologies ha make i
possible he de elopmen o he so bioinspi ed coope a i e Flapping Wing Ae ial
Robo s (FWAR). This chap e aims o se he esea ch base and he de elopmen s’ ini ial
condi ions. The chap e s a s by analyzing he capabili ies o bi ds in na u e. The objec i e
is o s udy he skills o he bi ds ha he o ni hop e has o eplica e. Then he chap e
con inues analyzing he con ou condi ions o he de elopmen gene a ed by he ae ial
pla o m. This pa o he chap e is c ucial o examine he limi a ions o he sys ems
de eloped. In he end, he chap e p esen s he so obo ics echnologies ha can be used
conside ing he pla o m’s es ic ions.
2.1 Bioinspi a ion
Na u e is he wo ld’s leading science, and e olu ion is i s esul . E ol ing o e millions o
yea s, he sys ems shown in na u e a e almos pe ec . This sec ion ocuses on he analysis
o he solu ion achie ed by na u e ega ding bi ds. Speci ically, I e alua es he in e ac ion
capabili ies o bi ds wi h he en i onmen and o he animals and he ligh beha io ha
makes i possible. This sec ion aims o de elop a concep abou he a ian-inspi ed sys ems
ha o ni hop e s wan o eplica e.
Causes e idence ha bi d inspi es o ni hop e s. Howe e , o ni hop e capabili ies a e
s ill a om bi ds’ ones. Ne e heless, his hesis aims o de elop a pla o m as close as
possible o bi ds sol ing he cu en obo ics p oblems. Fo una ely, he bi ds’ capabili ies
in na u e and he cu en obo ics p oblems a e simila . They exhibi excellen in e ac ion
capabili ies and dex e ous manipula ion skills. Fu he mo e, hey a e he wo ld’s mos
excellen ligh expe s. This hesis aims o mimic hese pe o mance capabili ies while
eplica ing he bi d sys ems o pe o m hei s.
E ol ing om dinosau s, bi ds show an excep ional adap a ion o he en i onmen . In
he beginning, bi ds’ ances o s used o walk on he g ound. A ha ime, hey we e much
la ge han nowadays. Howe e , hey we e smalle han hei compe i o s, and hey used o
be he p ey o gian ca ni o ous hun e s. E olu ion ga e hem ea he s and ligh bones ha
we e unuse ul ini ially. O e he yea s, bi ds imp o ed hei lying abili y. Consequen ly,
hei a bo eal habi s ook p ominence o a oid hei g ound ulne abili y. 2.1 highligh s
he e olu ion o bi ds om he g ound o he ees. Figu e 2.1a shows he mos ancien
19
20 Chap e 2. Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR)
(a) Deinonychus [78]. (b) A chaeop e yx [79].
Figu e 2.1 Bi ds’ ances o s.
non-a ian ances o and Figu e 2.1b he i s bi d om he La e Ju assic A chaeop e yx a
igh [77].
The Figu e shows he beginning o he e olu ion o some o he cu en cha ac e is ics o
he bi ds in hei ances o s. As shown in he igu es, bi ds ha e ea he s om he beginning
unuse ul o ligh o Deinonychus and ga e limi ed lying capabili ies o A chaeop e yx.
Ano he pa o hei bodies ha ealize hei ligh abili y is he ail. Howe e , hei ail is
no ae odynamic a all. They used he ail o balance and no o con olling hei ligh .
Consequen ly, hei ligh was sho a low speeds.
Figu e 2.1 also shows he i s p esence o a p ominen beak. The bi ds used hei
beaks o ea he ough lesh o hei p ey. E e y hing su ounding hem was agg essi e.
Consequen ly, he Figu e shows huge claws p epa ed o wo k wi h hei beaks o su i e
in ha en i onmen .
In he beginning, he bi d’s claws we e he mo e p ominen sys ems. The bi ds used hei
claws o locomo ion and hun ing. Howe e , e olu ion made hei ligh abili y inc ease
s ep by s ep. Then, bi ds abandon he g ound, and hei a bo eal habi s inc emen ed.
Nowadays, bi ds li e in a less agg essi e en i onmen . In he cu en wo ld, e e y hing
is smalle han in he ime o he p esen ed bi ds’ ances o s. Thei p eys we e smalle han
be o e hei eeding me hod also changed. A he same ime, hei ligh abili ies inc eased.
Consequen ly, bi ds also educed hei size, and hei wings and ail gained p ominence on
hei bodies.
The cu en ligh abili ies made he bi ds he e e ence o all he ae ial sys ems de el-
oped. Bi ds show dex e ous ligh abili ies. The combina ion o ail and wings made i
possible o pe o m dex e ous maneu e s in he ai . Also, hey inc emen he in e ac ion
wi h he en i onmen making i possible o in e ac wi h hem.
Howe e , he a ie y o species o bi ds in he wo ld made i challenging o e alua e he
specie ha is wan ed o eplica e. In ha way, his sec ion analyzes he di e en species and
he in e ac ion capabili ies o each one. The aim is o se a base o ou de elopmen s and
eplica e hem. The ex di ides he analysis in o h ee. The i s one is he manipula ion
capabili ies. Depending on he specie, hey demons a ed di e en adap a ions o human
objec s o a e used o li ing in di e en en i onmen s.
2.1 Bioinspi a ion 21
(a)
Golden eagle hun ing a abbi
[80].
(b) Osp eys hun ing a ish [81]. (c)
Pa o holding a
pencil [82].
(d) C ow ea ing [83]. (e) Ba n owl ea ing. ( ) Sec e a y bi d hun ing.
Figu e 2.2 Bi ds’ manipula ion skills.
2.1.1 Bi ds Manipula ion Capabili ies
Bi ds show dex e ous manipula ion capabili ies. E olu ion imp o es he adap abili y o
he bi ds’ claws, enhancing hei manipula ion skills. Thei main use in na u e is o
eeding. Hun ing animals o helping ea he lesh wi h hei beak is he p ima y u ili y.
Ne e heless, he p esence o humans in he en i onmen is elaxing hese skills. This
p esence educed he di icul ies o inding ood. Mainly, ege a ian species ha e easie
access o ood om he expansion o ag icul u e. Thei skills in he manipula ion o objec s
come om hei cap i e uses. They do no use o ind hese objec s in na u e. Howe e ,
humans o ced his si ua ion o e he yea s, imp o ing hei abili ies. Figu e 2.2 shows
di e en examples o manipula ion conce ning a wide ange o bi d species and claws
mo phologies.
The analysis begins wi h he g oup o he Accipi idae. Figu es 2.2a and 2.2b show wo
examples o he manipula ion skills o hese bi ds. The special y o hese bi ds is he hun .
They can ca y p ey e en la ge han hemsel es. They used he hocked shapes o hei
claws o go inside he skin o hei a ge s. I makes o educe he adap abili y o hei
claws. Howe e , hey dis ibu e he s ess along he oo and legs, aking ad an age o
hei nails and he longi udinal s i ness o hei oes. To abso b hese s esses, he mass o
he manipula ion body akes almos 40% o he en i e body mass. They combine hese
manipula ion skills wi h hei lying p owess. They can ly a high eloci ies wi h high
maneu e abili y o each hei p ey. They also show high-gliding abili ies o con ol hei
a ge in he su ounding en i onmen . Then, he li and h us gene a ed by hei wings
can anspo hei p ey o a place whe e hey can ea sa ely.
The second g oup analyzed is he Psi acidae. Figu e2.2c show an example o hei
28 Chap e 2. Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR)
beha io , and capabili ies. In ha sense, Table 2.1 summa izes he main cha ac e is ics o
he pla o m ligh .
Table 2.1 E-Flap[68] main cha ac e is ic.
Weigh
(Kg)
Maximum
Payload (kg)
Maximum
Speed (m/s)
Flapping
F equency (Hz)
b
(m)
Wing
Sec ion
0.51 0.520 6 5.5 1.5 Fla Pla e / 1
4c
This pla o m mimics he bi ds eplica ing he lapping mo ion and using he ail o
con ol he ajec o y. Howe e , hei ae odynamic cha ac e is ics limi hei capabili ies
o lying.
Payload is one o he mos limi ed ligh capabili ies. The ini ial o ni hop e can ca y
520 g in his con igu a ion, including he sys em needed o a basic ligh . This educed load
limi s possible applica ions and componen s used in he ligh . Howe e , he de elopmen
o his disse a ion imp o es his limi a ion.
Focusing on he ligh beha io , he a e age ligh speed o he pla o m is 6 m/s a a
lapping equency o 6 Hz. These speeds and lapping equencies a e high o sho -
dis ance asks. Fu he mo e, obse ing na u e, his beha io is ypical o small bi ds.
Howe e , he o ni hop e has a 1.5 m wingspan. The ae odynamic condi ions o he wing
cause his beha io . The wing has a la -pla e ai oil wi h a sho cho d. This ai oil
inc eases he equency equi ed o ly, educing li and h us gene a ion. Ne e heless,
hey main ain s abili y and minimize lapping ine ia.
On he o he hand, he ai oil’s la shape educes he pla o m’s glide capabili y. As
a esul , he glide phase wi h hese ai oils is limi ed, and hey canno gene a e li by
a ying he ai low. Consequen ly, i is no easy o use he pla o m o long- ange ligh s.
In addi ion, he glide wi h his p o ile in ol es high speeds losing mo e al i ude.
The pla o m con ollabili y wi h his se up is also limi ed. The only con ol su ace on
i is he ail. The ail has li le in luence on he ligh due o he dis u bances gene a ed
in he ai low by he lu e . Ins ead, i inc eases he e ec in he gliding phase. Howe e ,
mo e ac ua ion is needed o ha e he ajec o ies pe ec ly con olled.
The ex abo e analyses he cu en s a e o he pla o m and i s beha io . The sec ion
now ocuses on he in luence o hese cha ac e is ics in he p oposed de elopmen s. This
hesis o e comes he p oblem om wo di e en sides. The i s is de eloping a sys em ha
allows manipula ing and pe ching he o ni hop e in a sa e in e ac i e way. The second
is o inc ease he ligh capabili ies conside ing he coope a ion and con ibu ing o he
manipula ion.
The ligh eloci y and weigh ook special ele ance in he de elopmen o he manipu-
la o . Usually, in Robo ics, pe ching and manipula ion a e wo p oblems well dis inguished.
The adi ional solu ion in Robo ics is o isola e hem and gene a e wo di e en ac ua o s.
Howe e , his hesis ies o achie e hem wi h he same manipula o and con olling
sys em.
Bi ds can pe o m pe ching and manipula ion asks wi h he same ac ua o . Thei
pe ching is smoo h: hey s op in he ai and p epa e he whole body o pe o m he
pe ching, as shown in Figu e 2.3 .

2.2 O ni hop he s Flapping Wings Ae ial Robo s(FWAR) 29
Howe e , o ni hop e s can no pe o m a con olled s op in he ai be o e pe ching.
Conce ning o ni hop e s, pe ching is an impac on he manipula o in his case. The
ae odynamic cha ac e is ics o he wing se he eloci y o he pe ching. The lowe
speed a which he cu en se up s ill has con ol o he ligh is a ound 3 m/s. This impac
gene a es conside able ene gy ha he manipula o has o adso b. Addi ionally, he eac ion
ime is ex emely sho . The combina ion o he high impac wi h he educed eac ion
ime makes he claw sys em e y es ic i e. This is imp o ed by he de elopmen in his
hesis o a new wing. In he u u e, conside ing he wing eloci y could also educe he
impac as bi ds do.
A e he con ac , he manipula o mus gua an ee a igh holding gene a ing enough
ic ion o main ain he balance. Usually, his p oblem is sol ed in adi ional obo ics using
sha p spikes. Howe e , he p ima y objec i e o he sys ems de eloped is o coope a e
wi h humans. So in ha way, hese kinds o sha ped solu ions a e no app op ia e. He e
akes pa icula ele ance o he so obo ics concep s.
So obo ics is close o he limi a ions o manipula ion and coope a ion wi h humans.
Howe e , he ene gy in ol ed in he pe ching makes i challenging o use a so sys em.
Conce ning manipula ion, he pla o m’s main limi s a e weigh and ae odynamic in luence.
The sys em de eloped also has o conside he load’s posi ion du ing he ligh . A change
in he balance on he pla o m causes a dis u bance on he ligh . This dis u bance has o
be co ec ed by he con ol ac ua o s. The e o e, educing his in luence on he ligh will
inc ease he pe o mance.
Conce ning he weigh o he manipula o , he ae ial obo ics pla o m’s payload limi
is a ound hal a kilog am wi h he ac ual se up. The payload es ic ions a e ough o
achie e he desi ed esul s in he p ojec p o iding manipula ion, isual, and con ol
capabili ies. A e analyzing he equi emen o he di e en pa s, he emaining load o
he manipula ion is 150 g, less han 15% o he weigh o he ae ial pla o m. In Sec ion
2.1, he manipula ion and pe ching sys em e alua ion in high-size bi ds is a ound 30%.
Then he sys em has o o e come he solu ions ound in na u e.
The educed payload o he manipula ion sys em makes i necessa y o e alua e no el y
ac ua ion me hods. These me hods use sma no el y ma e ials o inc ease he ac ua ion
capabili ies while lowe ing he sys em’s weigh . Conc e ely, i uses Shape Memo y Alloys
(SMA) o mimic he muscles o bi ds supplying ha beha io .
Howe e , s udying bioinspi ed ae odynamic su aces o he pla o m elaxes he payload
es ic ion. The wings can gene a e a g ea e h us du ing he lapping, imp o ing he
gliding phase. The aim is o de elop bi d-shaped wings and abandon he concep s o la
wings adi ionally used o o ni hop e s.
The wings de eloped in his disse a ion also ha e a so s uc u e imp o ing he in-
e ac ion skills o he pla o m. The design mimics he ea he s o he bi ds. The wings
also con ibu e o he con ol o he o ni hop e . T adi ionally, o ni hop e s only use hei
wings o li and h us gene a ion. Howe e , he new design can modi y hei su ace as
bi ds do. This de elopmen gene a es ac ua ion on he ajec o y. Adding his con ol in
he wing makes i possible o manage he payload added.
The o he con ol su ace is he ail. This ae odynamic ac ua o is he only one in he
ini ial se up. This ail uses he concep s lea ned om ixed wings. Howe e , he beha io
o he ligh demons a es ha i s ac ua ion in Flapping Wing Ae ial Robo s (FWAR) is
30 Chap e 2. Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR)
limi ed unde ce ain condi ions. The e o e, his hesis analyzes o include bioinspi ed
concep s in he ail.
Howe e , he ail and i s posi ion a e es ic i e. The dis ance be ween he ail and he
Cen e o G a i y (CoG) o he pla o m is long. This longi ude gene a es a momen ha
inc eases wi h he weigh o he ail. This hesis analyses no el y mo phing ac ua o s o
o e come his p oblem. These ac ua o s will mimic he ea he while educing he o al
weigh o he ail.
The ollowing lis summa izes he main equi emen s gene a ed by he ae ial pla o m
o he p oposed de elopmen s.
•
Payload: This may be he bigges d awback o using a FWAR. The weigh ha an
o ni hop e can ca y ou is smalle han o he adi ional ae ial pla o ms. The
cu en solu ion is o use a la pla e wing sec ion o make he payload smalle .
This cha ac e is ic limi s he payload and, as a consequence, he applica ions o he
pla o m. This hesis o e comes his p oblem by designing bioinspi ed su aces
capable o loading mo e weigh .
•
Con ollabili y: The o ni hop e ’s cu en s a e shows limi ed ajec o y con ol. The
dis u bances caused du ing he lapping limi he pla o m’s maneu e abili y. The
p oposed solu ion is o add mo e con ol deg ees o he pla o m abandoning s an-
da ds solu ions. The aim is o mimic na u e in all he con ol su aces, conside ing
he adi ional ae odynamics bu no eplica ing hese sys ems.
•
Pe ching beha io : The s a e o he pla o m a he beginning o he hesis causes
es ic i e pe ching condi ions. The con ac eloci y is bigge han he one ound
in he pe ching o bi ds. The e o e, I causes he de eloped pe ching mechanism
o abandon some bioinspi ed concep s o ocus on he con ou condi ions o he
pla o m.
•
Coope a ion beha io : O ni hop e shows excellen human coope a ion skills. Fo
example, hey do no ha e sha p edges like p opelle s in adi ional pla o ms. How-
e e , he p oposed mechanisms imp o e i s pe o mance. They make he pla o m
sa e , making i possible o in e ac physically wi h humans. The manipula ion skill
was added o manage he con ac . Mo eo e , he so e su aces also educe he
damages caused in case o ailu e.
Ul ima ely, his in es iga ion goes h ough he cu en s a e o he o ni hop e s, con-
side ing he pla o m’s limi a ions. The aim is o ob ain a pla o m sa e o in e ac ion
and capable o coope a ing wi h humans. In ha way, bioinspi a ion is c ucial in all
he de elopmen p esen ed. Howe e , he pla o m ligh con ou condi ions limi he
bioinspi a ion o he de eloped sys ems. To each comple e bioinspi a ion, he knowledge
abou he con ollabili y o he o ni hop e mus inc ease. Ne e heless, he de elopmen
p oposed made i possible o each new ho izons. They gi e o ni hop e s manipula ions
skills, be e coope a ion skills, and ex ended con ol capabili ies.
2.3 So Robo ics Technologies 31
2.3 So Robo ics Technologies
This wo k has been de eloped unde he amewo k o he ERC Ad anced G an GRIFFIN
[
65
]. The g an aims o p o ide o ni hop e s wi h dex e ous manipula ion skills and
imp o ed ligh capabili ies. E-Flap [
68
] is one o he pla o ms de eloped in his Ad anced
G an . This hesis aims o ans o m E-Flap in o a mo e in e ac i e pla o m ollowing
so obo ics and bioinspi a ion p inciples. Ul ima ely, his pla o m is called So -Flap.
So -Flap has manipula ion skills and imp o ed ligh beha io .
So -Flap ollows he cu en ends in obo ics, ying o de elop sa e in e ac ion
mechanisms adap ed o ae ial obo ics. The goal is o mimic he beha io o bi ds in
na u e. The au ho s in [
43
] unde line he necessi y o using so obo ics echnologies o
eplica e na u al sys ems. In sho , he aim is o mimic na u e using so obo ics as he
obo ic echnology ha makes his possible. Figu e 2.6 is an excellen exponen o he aim
o his wo k.
Figu e 2.6 A Ha is hawk abou o land on a b anch [98].
Figu e 2.6 shows he pe ching app oach maneu e o a hawk. The en i e body o he
hawk is wo king o each he pe ching spo s. The body could be di ided in o ou pa s:
body, claws, ail, and wings. The ail and wings a e he con ol su aces. The bi d has hem
ully opened o s op he ai low. The powe ul legs and claws a e p epa ing o he con ac .
All he sys ems shown in he igu e a e expe imen ing a conside able de o ma ion. The
ea he s b eak he ai low, and he legs abso b he impac . The de o ma ion expe ienced
by hese sys ems du ing he pe ching highligh s hei so cha ac e is ics,
This hesis s udies he applica ion o so echnologies in wo kinds o sys ems, he
wings and ail, as mo phing ac ua o s and in he claws and legs sys em. Howe e , applying
so echnologies in hese applica ions is challenging due o he o ce needed o exe .
The e o e, he sys em ind an equilib ium be ween o ces and de o ma ions o each he
desi ed ac ua ion.
This disse a ion s udies and adap s so echnologies o ae ial pla o ms. Nowadays,
so obo ics is well-known in-g ound obo ics. Howe e , he e a e ew examples o hese
echnologies in ae ial obo ics. Ne e heless, some so echnologies o e low-weigh ed
de elopmen s wi h highe capabili ies o ae ial obo ics. In ha sense, his wo k analyses
he use o so echnologies in ae ial pla o ms o imp o e hei in e ac ion capabili ies.
Tha opens a new ision o so obo ics in ae ial pla o ms wi h in e ac ion capabili ies.
32 Chap e 2. Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR)
Rega ding manipula ion, he goal is o de elop an a ian-inspi ed manipula ion sys em
o o ni hop e s. So echnologies inc ease claw in e ac ion capabili ies o e aking he
bi d ones. One o he leading so ac ua ion echnologies s udied du ing he de elopmen o
his hesis is endon-d i en ac ua ion. This echnology eplica es he connec ion be ween
he muscles and endons in a ian claws. Howe e , adi ional ac ua ion me hods canno
eplica e he o ce exe ed by animal muscles. In ha way, using no el so ma e ials
make eplica ing a ian muscles’ beha io possible. The claw de eloped use SMAs as
muscles and endon-d i en ac ua ion o deli e he o ce o he oes. The claw keeps he
so cha ac e is ics also in i s s uc u e o enable adap able manipula ion and esol e he
p oblem o pe ching.
The e a e no eco d examples o sys ems combining SMAs ma e ials o ac ua ion and
endon-d i en ac ua ion in a g ippe . This lack o de elopmen s is because o he ecen ly
disco e ed applica ions SMAs. In ha way, he analysis p esen s bo h solu ions sepa a ely.
The e alua ion s a s wi h analyzing he oldes me hod. The i s eco d use endon-
d i en ac ua ion o a so ac ua o was in 1978 in [
44
]. This concep ual wo k lays he
ounda ion o endon-d i en ac ua ion. Howe e , he ac ua o ’s capabili ies we e limi ed.
In [
99
], he au ho s inc eased he complexi y o he p o o ype by eplica ing a human hand
wi h high payload capabili ies. Howe e , he de elopmen did no ollow so obo ics basis.
The au ho used ha d ma e ial in he con ac and a non-con inuous s uc u e. In [
100
], he
au ho s de eloped con inuous join s using deposi ion manu ac u ing and analyzed hei
beha io compa ed wi h adi ional join s. This one could be one o he i s examples o
so obo ics as we know i oday. This wo k analyzed con inuous base s uc u es. This
concep is la e ollowed by so obo ics. Figu e 2.7a shows his g ippe .
The ollowing is one o he i s examples o unc ional so obo ics g ippe s using
endon-d i en ac ua ion as we know i s oday. In [
101
], he au ho s de eloped an an-
h opomo phic hand using lexible join s co e ed wi h Polydime hylsiloxane (PDMS). In
addi ion, he au ho used capaci i e senso s o measu e he ac ua ion. Figu e 2.7b shows
his de elopmen . In [
102
], he au ho s de eloped a g ippe using a con inuous s uc u e o
a iable s i ness PDMS ma e ials. This g ippe showed dex e ous adap a ion capabili ies
and high ic ion gene a ion. Howe e , I s payload capabili ies a e low due o i s highly
de o mable s uc u e.
The examples p esen ed abo e comp ise g ound obo ics mechanisms. The e a e no
s ic ly so de elopmen s used in ae ial obo ics. The absence o p o o ypes in ae ial
obo ics is because o he unce ain ies gene a ed by he so ac ua ion. Thei ac ua ion is
complex o model and mus be e ised o de elop high-load demanding sys ems.
Howe e , he e a e g ippe s in ae ial obo ics using endon-d i en ac ua ion wi hou so
obo ics cha ac e is ics. One o he leading examples is [103] shown in Figu e 2.7c. The
au ho s p esen a endon-d i en sys em o pe ch using a quad o o . This wo k is simila o
he one analyzed in his hesis. Howe e , some di e ences inc ease he complexi y o he
wo k de eloped in his hesis. The i s one is he use o a di e en ae ial pla o m. The
use o o ni hop e s gene a es a p oblem comple ely di e en , as Sec ion 2.2 p esen s. The
second one is he use o a so claw. The claw shown in [
103
] elaxed equi emen s o he
manipula ion. This claw is mo e o ien ed owa d smoo h pe ching han manipula ion.
In his Ph.D. hesis, SMAs eplica es he e ec o he muscles in he legs o he bi ds.
SMAs ha e been widely s udied since he 90s. SMAs a e an alloy o Nickel and Ti anium.
2.3 So Robo ics Technologies 33
(a)
G ippe de eloped in
[
100
] by he s udy o
addi i e manu ac u ing o
lexible join s.
(b)
An ophomo ophic
so g ippe de eloped
in [101].
(c)
Bi d inspi ed pe ching
mechanism de eloped in
[103].
Figu e 2.7 Cable d i en g ippe s examples.
They can change hei in e nal s uc u e wi h he empe a u e. The e o e, They can eco e
he ini ial memo ized shape. The mal ea men s pe o med in he alloy unde ce ain
s ess condi ions s o e his shape.
T adi ionally, SMAs a e an alloy o Nickel and Ti anium. Howe e , o he ma e ials can
be added in a mino composi ion o each some speci ic p ope ies. Two ypes o SMAs
can be de eloped depending on he he mal ea men pe o med, one-way SMAs and
wo-way SMAs. The one-way SMAs ha e one shape memo ized, and hey eco e i by
inc easing he empe a u e. The wo-way SMAs ha e wo memo ized o ms. They can
change be ween hem depending on he ansi ion empe a u es.
The use SMAs in so g ippe s can a y depending on he applica ion. The i s is
SMAs wo king as o ce gene a o s inside he s uc u e, subs i u ing he endons. In wo ks
like [
49
], he au ho s de eloped a ca e pilla -based obo using SMAs sp ings as in e nal
endons and muscles. In [
50
], he au ho s e alua e he use o SMAs as bending ac ua o s
o so obo ics applica ions. In his case, he ac ua o has he shape o a shee ha educes
he adio o med depending on he empe a u e. In [
51
], he au ho s de elop a mic o
g ippe using SMA wi es.
The wo ks abo e show h ee di e en ways o using SMAs. The i s is in a sp ing shape,
he second is in he shape o a shee , and he hi d is in a wi e shape. This Ph.D. hesis
uses SMAs wi h sp ings shape. The main eason is he linea o ce hey can gene a e in
a educed space wi h a low weigh . These capabili ies make SMAs sp ing an excellen
candida e o eplica ing a ian muscles on o ni hop e s.
The o he p oposal using so echnologies in his disse a ion is hei use o mimic
bi ds’ ligh mechanisms. The goal is o de elop so mo phing ae odynamic su aces
o imp o e con ol capabili ies and educe isky si ua ions in case o aul . [
49
] shows
he so mo phing concep applied o g ound locomo ion. In his case, he au ho s used
SMAs o a y he body o a ca e pilla , gene a ing locomo ion on he g ound. In his wo k,
he en i onmen is he ai . So he necessi y o c ea e ae odynamic su aces o con ol he
ajec o y.
The li e a u e shows examples o gene a ing so ae odynamic con ol su aces in ixed
wings pla o ms. The idea is o adap his echnology o a lapping wing de ice. Pa icula ly,
he in e es o his Ph.D. hesis is o use MFCs. MFCs we e de eloped by NASA in he 90s.

34 Chap e 2. Bio-Inspi ed So Flapping Wing Ae ial Robo s(FWAR)
(a)
Mic og ippe de eloped in [
51
] us-
ing SMAs.
(b)
An ophomo ophic g ippe
using SMAs bending ac ua-
o s de eloped in [50].
Figu e 2.8 SMAs g ippe s examples.
Ini ially, hey we e concei ed as senso s o measu e small s ains and displacemen . This
use e lec s i s capabili y o change i s in e nal esis ance depending on i s de o ma ion.
This beha io gene a es wo ks o s uc u al applica ions like [
104
]. The au ho s de eloped
a low-cos s uc u al sensing ac ua o wi h low weigh and p ecision in measu ing small
displacemen s and ib a ions.
Howe e , esea che s became awa e o o he p ope ies ha MFCs ha e. The i s
ea u e ha ocused he esea che s’ in e es o esea che s in hese ac ua o s was hei
blocking o ce du ing he ac ua ion. This beha io gene a ed he i s applica ions o MFCs
in so obo ics. In wo ks such as [
56
,
57
,
58
,
59
,
62
], he au ho s e alua ed he use o
MFC in unde wa e obo s. They de eloped con ol su aces o obo s unde wa e by
exploi ing he isola ion p ope ies o he MFCs and blocking o ces. These wo ks poin
ou he po en ial applica ions o hese ac ua o s o de elop con ol su aces.
Conce ning ae ial applica ions o MFCs, some esea che s s a ed using hem as ae ody-
namic con ol gene a ion eplacing adi ional mechanisms. In wo ks like [
105
,
106
,
107
],
he au ho s de eloped aile ons o a ixed-wing using MFCs. These wo ks aimed o con ol
he ai c a ajec o y wi h he de eloped ac ua o . They show he po en ial applica ion o
MFCs in ae ial ehicles. Ne e heless, hey did no ollow he so obo ics basis, and he
pla o m had di e en ligh condi ions.
(a)
Ac i e wing and ail using
MFCs ac ua o s [106].
(b)
Ac i e wing using MFCs a y-
ing he dihed al angle om
[107].
Figu e 2.9 MFCs con ol su aces gene a ion.
2.3 So Robo ics Technologies 35
The esul s o his hesis a e e lec ed in he pla o m So -Flap.So -Flap is an o -
ni hop e comple ely a ian-inspi ed. Thei wings a e lexible and capable o ac ua ing in
he ajec o y as bi ds do. The wings o So -Flap abandon he classical concep o he
wing o o ni hop e s shown in s a e o he a . Thei ae odynamic p o ile is a ian-inspi ed,
lea ing he la p o iles p esen ed usually in he o ni hop e s. I inc eases he capabili ies
o gliding o ac ual o ni hop e s. This p o ile mimics he de o ma ion su e ed in he
ea he s du ing he lapping o inc ease he li and h us gene a ion. The de o ma ion
o he wing ollows he so obo ics concep . The geome y o hei wing is a iable o
gene a e ac ua ion in hei ajec o y. The wings ha e a lexible ip made o elas ome
ma e ial ab ic, gene a ing a di e en ial ac ua ion o e he ROLL o he pla o m. These
wings inc ease he con ol capabili ies, payload, and sa e y du ing he ope a ion om
adi ional o ni hop e s.
So -Flap ha e an a ian-inspi ed ail. The ail is made o MFCs ollowing he concep
lea ned om he li e a u e. Howe e , his ail has o o e come hese de elopmen s pe -
o ming majo de o ma ions and being capable o suppo ing he lapping dis u bances.
The ail mimics he beha io o he ea he s o he a ian ails. This inspi a ion inc eases
he con ol capabili ies o he cu en o ni hop e s.
The claw o So -Flap p o ides manipula ion skills. They a e a ian-inspi ed, ollow-
ing he concep o so obo ics. These claw aims o gene a e manipula ion skills o
o ni hop e s. The a ian inspi a ion is c ucial o hei de elopmen as well as he con ou
condi ions o he manipula ion and pe ching o he pla o m. Ul ima ely, hese claws mus
o e come he a ian claws inc easing hei capabili ies o in e ac wi h humans.
So -Flap is a sa e o ni hop e wi h imp o ed con ol and payload capabili ies. Bioin-
spi a ion and so obo ics a e c ucial o de eloping his concep o pla o m.
3 Bio-Inspi ed Manipula o s o
O ni hop e s
This
chap e ocuses on he de elopmen o So Bio-Inspi ed manipula ion sys ems o
o ni hop e s. This de elopmen is mo i a ed o p o ide manipula ion capabili ies
o he o ni hop e p esen ed in Chap e 2.2. This kind o ae ial pla o m is es ic i e
ega ding payload and ae odynamic in luence. The e o e, his de elopmen will conside
hese equi emen s o enhance ligh capabili ies.
"So " and "Bioinspi a ion" a e he wo main cha ac e is ics o his wo k. A ew yea s
ago, obo ics’s main p oblem was con olling he ae ial pla o m. Nowadays, coope a ion
asks a e a he edge o obo ics. T adi ionally, sys ems in obo ics we e s i and adap ed
o a speci ic applica ion. I is challenging o coope a e be ween humans and obo s using
igid ac ua o s due o he high inju y isk o humans. In ha way, he solu ion is o de elop
sys ems adap able o mul iple shapes and sa e o in e ac ion. The sys em de eloped
ollows he concep o ha ing an ae ial pla o m adap ed o he en i onmen and sa e o
coope a ion.
Bioinspi a ion is he o he main keywo d o his wo k. The sys em se s aims o mimic
he beha io and capabili ies ha he bi ds ha e wi h hei claws. In ha way, his wo k
analyses di e en bi ds’ mo phology while eaching he equi emen s o he ae ial Pla o m
[
68
] o achie e be e esul s. Finally, a e de eloping he hesis, he pla o m will be ully
inspi ed by bi ds wi h so capabili ies o enhance hei in e ac ion skills.
The sys em de eloped aims o con ibu e by adding capabili ies o di e en asks. The
ollowing lis summa izes hese asks:
•
Human-Coope a ion: The claw de eloped has o be sa e o human coope a ion
ollowing he equi emen s o he ERC Ad anced G an GRIFFIN [
68
]. O ni hop e s
a e one o he mos sa e ae ial pla o ms o human in e ac ion. I s low ligh
eloci ies and he absence o sha p edges like p opelle s make i a good op ion.
The e o e, he manipula ion sys em mus p ope ly main ain hese capabili ies o
enhance in e ac ion wi h he en i onmen and humans. In ha way, de eloping a
so ac ua o suppo s hese capabili ies making i possible o in e ac wi h humans
using he claws.
•
Manipula ion Capabili ies: Adap abili y and so sys ems usually a e in e ela ed.
The aim is o each a sys em capable o manipula ing mul iple shapes, sizes, and
objec s.
37
44 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.3 Ac ua ion cycle pa ame e s.
is he shea modulus o he ma ensi ic phase. The equa ion uses he e ec i e diame e
due o he educ ion o he sp ing diame e du ing he mo ion. In Equa ion 3.5,
GA
is he
shea modulus on he aus eni ic phase.
Finally, The equa ion se s he displacemen in e ms o he o ce applied. In his wo k,
he in e es is o analyze he ension gene a ed by an ac ua o wi h a speci ic de o ma ion.
Then, he aim is o con ol he sys em by using he equilib ium o wo sp ings in opposi e
di ec ions. The main p oblem wi h hese ac ua o s is ha he p o ide doesn’ speci y he
p ope ies o he ac ua o , so in ha way, he s udy o hese p ope ies se s he beha io o
he alloy. Assuming all he sp ings p o ided ha e he same p ope ies and he alloy is he
same. The s udies o he ma e ial pe o med be o e con ibu e o de e mining ha in he
case o he same p ope ies and shape. The o ce gene a ed depends only on a cons an .
Tha simpli ies he model, making i possible o achie e mo e accu a e esul s. The es
ocus on ob aining his cons an .
3.1.2 Modeling SMA Muscles
Muscles dynamic model
The i s expe imen pe o med e alua es wo di e en sp ings. These es s aim o selec
one ega ding maximizing he o ce/weigh a io. Table 3.1 shows he geome ical physic
p ope ies o hese sp ings.
The es analyzed he sp ing’s beha io , making i s de o ma ion s able and a ying i s
empe a u e. Figu e 3.4 shows he esul s.
By analyzing he g aphic, Sp ing 1 can exe o ces up o 24 N wi h an elonga ion o
350% a 200
◦
C. The Sp ing 2 muscle eaches up o 95 N wi h simila condi ions. The
weigh o he second muscle is wo imes bigge han he i s . Howe e , he o ce eached
is almos ou imes bigge , ha ing a be e Fo ce/Weigh a io. As de e mined be o e, he
s eng h exe ed is c ucial o eplica e he bi ds’ muscles’ beha io . Also, he alloy should
o e come he equi emen s o Sec ion 3.3.

3.1 Shape Memo y Alloy(SMA) A i icial Muscles 45
Table 3.1
Geome ical pa ame e s physic o he Ni inol analyzed sp ings. The able shows
he p ope ies o he sp ing, ocusing on hei geome y and mass.
Pa ame e s Sp ing 1 Sp ing 2
Unde o med leng h (cm) 2 1.7
Sp ing Diame e (mm) 6 10
Wi e Diame e (mm) 0.75 1.5
Weigh (g) 3 6
(a) Sp ing 1: o ce in e ms o he empe a u e and elonga ion.
(b) Sp ing 2: o ce in e ms o he empe a u e and elonga ion.
Figu e 3.4
Sp ings analysis. (a) shows he o ce ha Sp ing 1 can exe in e ms o he
empe a u e(
◦
C) and elonga ion(
△L
). (b) shows he o ce ha Sp ing 2 can
exe in e ms o he empe a u e(
◦
C) and elonga ion(
△L
). The legend shows
in bo h g aphics shows he de o ma ion o each cu e.
A e pe o ming se e al es s, he sp ing expe imen s show some pe manen de o ma-
ion a e o e passing he empe a u e o he 99% o he aus eni ic phase. E alua ing his
empe a u e p e en s he loss o p ope ies du ing he ac ua ion. By es ing ou alloy, he
46 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
empe a u e is a ound 150
◦
C, so du ing he expe imen , he muscle had o wo k be ween
ambien empe a u e and 140
◦
C. This condi ion a oids pe manen de o ma ions, which
is one o he majo p oblems o his alloy. The empe a u e also co esponds wi h he
g adien s abiliza ion shown in Figu e 3.4.
Rega ding he s udy o he beha io o he SMA Alloy p esen ed in [
120
], he ans-
o ma ion be ween he ma ensi e phase o he aus eni e is linea in e ms o empe a u e.
Then he linea pa o he g aphic desc ibes he con e sion om ma ensi e o aus eni e.
Looking a Figu e 3.4, he alloy analyzed in his wo k s a s a 22
◦
C wi h 1% o aus eni e,
and he aus eni e eaches 99% a 140
◦
C. The conclusion is ha he beha io o he SMA
is linea be ween 20
◦
C and 145
◦
C. The e o e, he o ce exe ed by he SMA is linea
oo in his ange. The ollowing equa ion shows he de eloped model aking in o accoun
hese cha ac e is ics.
F(T,x) = FPD +K(x−x0)(T−T0)(3.6)
whe e
F
is he o ce exe ed,
T
is he empe a u e o he sp ing,
x−x0
is he elonga ion
o he sp ing.
FPD
is he o ce needed o pe o m a plas ic de o ma ion on he ma ensi e
phase,
K
is a cons an iden i ied om he expe imen s ha se he ela ionship be ween
empe a u e inc emen and he elonga ion, and
T0
is he empe a u e ha se s he ma ensi e
composi ion a 99%.
The empe a u e and elonga ion de ine he beha io o he muscles. This model desc ibes
he beha io o he ac ua o be ween 22
◦
C and 145
◦
C. The analysis o he expe imen
iden i ies he pa ame e
K
o he model. These expe imen s y o analyze he o ce exe ed
by he sp ing a a speci ic empe a u e and elonga ion. Fi s ly he s udy e alua es he
g adien in he cu e be ween empe a u e/ o ce a a pa icula de o ma ion. Then, his
g adien is ela ed o he cu e de o ma ion/ o ce a a s a iona y empe a u e. The accu acy
o he model inc eases wi h he amoun o da a.
FPD
is he cons an alue ob ained by eaching a pe manen de o ma ion on he ma en-
si ic phase. The model se i as he alue ob ained by applying a con inuous de o ma ion
agains a load cell in he ma ensi ic phase.
FPD
is easie o ge and e lec s he alloy’s
s a e.
A e he analysis o he muscle is essen ial o de ine he ac ua ion me hod used on he
ae ial pla o m o hea he sp ings. The elec ical model de e mines he powe consumed
by he sp ing, and he ene gy model analyzes he exchanges wi h he ai o e alua e he
empe a u e o he sp ing.
Muscles elec ic and ene gy model
The sp ing’s powe consump ion models he ene gy exchanges wi h he en i onmen . The
elec ical model se he powe consump ion o he sp ing. The e o e, he i s s ep is o
analyze how o plug he muscles in o he powe supply o de elop he elec ical model.
Due o he di icul ies o solde ing his alloy, his is a challenging ask. The alloy doesn’
admi adi ional solde ing and high empe a u es cause a igue in he ma e ial.
The s udy o he h ee me hods se he bes one educing he ene gy consump ion in he
connec ion and gua an eeing conduc i i y. The i s is o make a kno wi h he cable a ound
3.1 Shape Memo y Alloy(SMA) A i icial Muscles 47
he SMA. Al hough, un o una ely, his one is no eliable due o he own mo ion o he
sp ing, he kno loosens a couple o es s. The ollowing me hod es ed main ains he kno
using a in solde . This one shows a be e pe o mance han he p e ious one. Howe e , he
kno loses some o he in due o he ac ua ion empe a u e and he low mel ing empe a u e.
Ul ima ely, a sil e -based epoxy p o ides all he eques ed equi emen s, high elec ical
and hea conduc ion, high- empe a u e esis ance, and s abili y du ing he ac ua ion.
Figu e 3.5 p esen s he elec ical analysis o he SMA. The igu e shows he esul s
o a wide ange o expe imen s. The p ocu emen o he es s was: Fi s , a high-cu en
powe supply powe s he sp ing. The P 100 in eg a ed in o he kno hen measu es he
empe a u e, and he powe sys em so wa e s o es he cu en supplied o he muscle.
Figu e 3.5
SMA Elec ical analysis. The igu e on he le side shows he empe a u e(
◦
C)
in e ms o he elec ical cu en (A) and ime(s). The igh shows applied
Powe (W) du ing he expe imen s. The legends show he applied elec ic
cu en in di e en colo s..
This g aphic shows signi ican esul s in using he SMA ac ua ed by elec ici y. The
ollowing lis is going o summa ize he conclusion o he expe imen s.
•
Dependence on elec ic cu en : The g aphic on he le shows he empe a u e
SMA can each in e ms o he cu en and he ime. This empe a u e has a g ea
dependency on he cu en supply. Looking a he g aphic, he slope o he cu e
inc eases wi h he cu en eaching highe empe a u es in less ime. The e o e,
wi h he cu en inc ease, he ime o each he empe a u e is sho e , bu he ins an
consumed powe is highe han lowe cu en s. In ha way, he sys em has o be
analyzed o e alua e which si ua ion is mo e a o able ega ding ene gy consump ion
o each a speci ic empe a u e.
48 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
•
Tempe a u e s abiliza ion: The cu es show a s abilized empe a u e alue a e a
ime ha inc eases wi h he elec ic cu en . This esul is in e es ing because he e
is a balance in he ene gy exchange wi h he en i onmen . I makes i possible o
es ima e he ins an ene gy deli e ed o he en i onmen . The model mus conside
all he sys ems connec ed o he sp ing du ing he ac ua ion because hey exchange
ene gy wi h he en i onmen . This ene gy is no negligible.
This analysis aims o ob ain he elec ical model o he muscle, including he sys ems
in ol ed in he ac ua ion. The g aphic on he igh shows he sys em’s powe consump ion
o analyze he sys em elec ically. One o he main cha ac e is ics o his g aphic is ha
he powe is s able wi h he empe a u e. So he sys em could be modeled as a cons an
esis o . A e se e al es s using di e en empe a u es and cu en s, he mean alue o
he esis o is 0.23 Ω.
The elec ical analysis p esen ed in Figu e 3.5 opens wo possibili ies. The i s is o
use less cu en , ha ing mo e ime o hea he sp ing bu was ing less ins an ene gy. The
second one is o use a highe cu en , implying less ime o hea he sp ing. O cou se,
educing he eac ion ime is c ucial in ou sys em. In addi ion, howe e , he sys em mus
sa e as much ene gy as possible o minimize he ba e y weigh . The ba e y is one o he
hea ies sys ems in he o ni hop e , so educing ene gy consump ion is one o he ways
o educe he weigh . The ene gy model analyzes he way o ac ua ion o educe ene gy
consump ion.
The i s s ep o de eloping he ene gy model is o analyze he balance in he ene gy
exchange ound in Figu e 3.5. Figu e 3.6 shows he ins an ene gy los du ing he hea ing
p ocess o he sp ing a 0 m/s o wind eloci y.
Figu e 3.6
Powe abso bed by he en i onmen . This g aphic ep esen s in blue poin s he
ene gy deli e ed o he en i onmen a each empe a u e. The ed line shows
he heo e ical app oach o hese ene gy losses.
3.1 Shape Memo y Alloy(SMA) A i icial Muscles 49
The analysis o empe a u e s abiliza ion du ing powe supply concludes ha he en i on-
men ecei es all he ene gy supplied o he sp ing du ing his s abiliza ion. The e o e, he
g aphic p esen ed in Figu e 3.6 shows he powe ansmi ed o he hea ese oi ega ding
empe a u e. The li e a u e models o hese sys ems as elec ical esis o s exchanging wi h
he ai . The ollowing equa ion p esen s he model:
PL=hA(T−T0),(3.7)
whe e
PL
is he powe deli e ed o he ai ,
h
is he hea ans e coe icien ,
A
is he con ac
a ea wi h he ai ,
T
is he empe a u e o he sp ing and
T0
is he en i omen empe a u e.
In his equa ion, he e m
hA
is a cons an de e mined analy ically om Figu e 3.6
analysis. The hea exchange a ea is di icul o es ima e due o he numbe o sys ems
in ol ed in he ac ua ion. The componen s in ol ed a e he longi ude o he cables, he
con ol sys em, and he sp ing su ace. In ha way, he analy ical app oach is be e han
he heo e ical one. Fu he mo e,
h
depends on he hea exchange condi ion and he luid
ha w aps he sys em. In he end,
h
changes ega ding he kind o hea exchange. Howe e ,
he ela ionship o his e m wi h o he condi ions is well-known.
A e ha ing he ene gy consump ion o he en i onmen ela ed o he empe a u e and
ha ing he ime and powe consump ion o each his empe a u e, he ollowing equa ion
shows how o calcula e he ene gy.
ETo al =Z (T)
(To)
PTo ald
PTo al =I2R
ELos =Z (T)
(To)
PLd
ESp ing =ETo al −Elos
(3.8)
whe e
E o al
is he o al ene gy consumed in Joules,
(To)
is he s a ing poin o he
calcula ion,
(T)
is he ime when he sp ing eaches
T
Tempe a u e,
Elos
is he ene gy
los du ing he wa ming p ocess, and
ESp ing
is he use ul ene gy abso bed by he sp ing o
inc ease i s empe a u e.
Figu e 3.7 e alua es he ene gy losses depending on he empe a u e and elec ical
cu en applied. This igu e shows six cu es, i e co esponding o he ene gy consumed
using di e en elec ical powe s, and he dashed line is he ene gy abso bed by he sp ing.
The blue-shaded a ea is he use ul ene gy consumed by he muscle.
Analyzing he g aphic in Figu e 3.7 is in e es ing o ocus on he ene gy losses a a
speci ic empe a u e by applying di e en cu en s, o example, a 80
◦
C. The blue cu e
shows he ene gy abso bed by he sp ing o inc ease i s empe a u e. The g een cu e
co esponds o he was e o powe a 9 A. The cu e close o he blue one is he 9 A cu en
supplied; i s ene gy loss is less han he es . So he sys em loses less ene gy applying
highe cu en s. Also, he sys em has o each he ope a ional limi s o wo king be ween

50 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.7
Sp ing powe consump ion. The igu e shows in con inuous lines he elec ical
powe supply o he muscle in e ms o he elec ical cu en , and in dashed
line, he ene gy used o wa m he sp ing .
20 and 145
◦
C. These limi s se he minimal powe o 8.5 A in wind-s a iona y condi ions.
So ou lowe limi is 8.5 A.
The ene gy losses depend di ec ly on he sp ing’s exposi ion ime and he ene gy
exchange wi h he ai . As highe he exposi ion ime, mo e ene gy goes in o he ai ,
and less ene gy is used o wa m he sp ing. In addi ion, he o ced en ila ion he ligh
gene a es inc eases he powe deli e ed o he en i onmen .
Wi h hese equa ions, he ac ua ion is modeled on Simulink o analyze he beha io
o di e en muscles con igu a ion. Rega ding his hesis, he simula ion e alua es he
an agonis muscles con igu a ion closing he con ol loop.
The main conclusions o his sec ion a e ha he sp ing could be modeled elec ically as
a esis o o 0.23 Ω. Conce ning he ene gy model, a highe cu en du ing he ac ua ion
gene a es lowe ene gy losses educing he hea ing ime.
3.1.3 De eloping SMA Muscles
The pandemic si ua ion c ea es di icul ies in ob aining hese ypes o sma ma e ials. The
usual supplie s an ou o s ock o his alloy, especially in sp ing shape. Howe e , hey
do ha e SMA wi es. The e o e, we had o de elop a ab ica ion me hod o s udy how o
communica e he sp ing shape o non-de o med SMA.
The aim was o de elop a cus om-made SMA sp ing wi h he same p ope ies as he
one s udied be o e. The sp ing has he geome y p e iously s udied, and he p o ide
gua an ees ha he wi e’s alloy is he same. The main p oblem is communica ing he
shape and adequa ely pe o ming he he mal ea men .
3.2 Analysing he Fo ces In ol ed in he Pe ching 51
Fi s ly, he manu ac u ing ocuses on de eloping he sp ing shape wi h he wi e. Conse-
quen ly, he alloy has o each ma ensi ic plas i ica ion in he desi ed shape. Howe e , due
o he high elas ic limi o NiTinol, i is impossible o pe o m his ope a ion in he lab.
The e o e, he solu ion was o send he ma e ial o a sp ing ac o y in which hey de o m
he wi e o a sp ing shape.
The nex s ep was o analyze he he mal ea men . This ea men se s he unde o med
shape o he muscle. Howe e , i he empe a u e o he p ocess su passes he limi , he
alloy will be damaged, ob aining inclusions inside as a esul . The e o e, he s a e-o - he-
a con empla es less agg essi e ea men main aining mo e ime he empe a u e. The
he mal ea men ollows his op ion.
Due o he knowledge ex ac ed, he ini ial es s a ed wi h a he mal ea men a 250
◦
C
o 2 hou s. Un o una ely, he esul s didn’ achie e he expec ed ones. Then, an inc emen
o he empe a u e in a slope o 50
◦
C un il i eaches 450
◦
C was es ed. Howe e , he
ma e ial doesn’ ha e he same p ope ies. The limi ound in he li e a u e was 500
◦
C, so
he he mal ea men can no each his empe a u e.
A e he ea men , he ma e ial showed wo main p oblems: The elas ic limi on he
ma ensi e phase was smalle han he one e alua ed in he p e ious sec ion. Also, he
colo o he alloy was blue. Conce ning Ti anium alloys, he colo is c ucial; i indica es
he oxida ion s a e. The solu ion was o pe o m a ha dening p ocess a e he he mal
ea men . Ins ead o cooling down he alloy wi h he ambien empe a u e, he sp ing
was wi h wa e gene a ing a he mal shock. The analysis o he ma ensi e phase p o ides
simila esul s o he p e ious muscle. Then he muscle es in he aus eni e phase p o ided
he same esul s as he p e ious alloy.
3.2 Analysing he Fo ces In ol ed in he Pe ching
Be o e s a ing wi h he design pa , his sec ion is based on a wo k de eloped du ing his
hesis and p esen ed o In e na ional Con e ence on Unmanned Ai c a Sys ems (ICUAS)
2021 [
73
]. Fi s , his wo k p esen s a heo e ical analysis o he o ces in ol ed in he
pe ching. Then, he pape p esen s a simple pe ching mechanism using magne ic o ces.
De eloping his wo k is a good s a ing poin o analyzing he pe ching in o ni hop e s.
The objec i e o his sec ion is o ha e a i s app oach o he o ces in ol ed in he
pe ching. The e o e, he i s s ep is o analyze he con ou condi ions o he pe ching;
Figu e 3.8 e alua es he ligh s ages in ol ed in he maneu e .
The igu e highligh s h ee phases be o e he pe ching:
•
No mal ligh : In his phase, he o ni hop e is on a s able ligh . The ligh eloci ies
a e om 3 m/s up o 6 m/s[
68
]. The con ol ligh sys em can con ol and change
he ajec o y o he ligh .
•
App oach ligh : In his s age, he con ac is close, and he pla o m p epa es he
con ac . The o ni hop e eloci y is s ill enough o pe o m smoo h changes in he
ajec o y. This phase aims o educe he pla o m’s speed o he lowe con ol limi ,
a ound 3 m/s.
52 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.8
Fligh phases. The igu e shows he iden i ied ligh s ages be o e he pe ching.
•
Con ac phase: The con ac is imminen in his phase. The in en ion is o each
he pe ching spo wi h he lowe eloci y possible. Howe e , he in luence o he
ae odynamic su aces on he ligh du ing his s age is limi ed due o he educed
speed.
A e analyzing he s ages o he ligh be o e he pe ching condi ions de ining i s
eloci ies, he s udy se s he con ou condi ions. Nex , he s udy e alua es he wo s inpu s
o a conse a i e app oxima ion o he maneu e . In his case, he weigh conside ed is
1 kg. This is he usual lying mass o E-Flap. The impac eloci y is he lowe limi in
which he o ni hop e s ill has con ol ligh capabili ies. Finally, he pe ching mechanism
conside ed has no abso p i e pa s. All hese condi ions a e un ealis ic. Howe e , hey se
a good s a ing poin o he design.
The ex abo e e alua es he pa ame e s ha cause in luence in he p oblem. Now,
he s udy analyzes he pe ching sys em as an impac using he one-deg ee-o - eedom
ib a ion equa ion o an undamped sys em. [
121
,
122
] is he base o hese calcula ions.
Equa ion 3.9 ep esen s ha equa ion.
uc( ) = (Acosωn )(3.9)
whe e
ωn=pk/m
,
m
is he mass sys em,
k
is he s i ness,
u
is he displacemen and
A
is
he ini ial displacemen u0.
The Equa ion 3.9 assumes ha he pe ching sys em can’ abso b any impac and is
a ached o he body o he o ni hop e . Then, he s udy se s he collision be ween he
pla o m and he landing spo . In his case, I assumes an almos pe ec elas ic collision
wi h a coe icien o es i u ion o 0.9. This condi ion means he ae ial pla o m, claws,
and pe ching place could abso b some ene gy du ing he impac . Howe e , i is un ealis ic
o assume ha he collision is pu ely elas ic, meaning ha he s i ness o he ma e ials is
in ini e. Finally, he ollowing equa ions calcula e he o ces implied in he pe ching.
3.2 Analysing he Fo ces In ol ed in he Pe ching 53







mo·( +
o− −
o) = Ic
mpp ·( +
pp − −
pp) = −Ic
e=− +
o− +
pp
−
o− −
pp
(3.10)
whe e
mo
is he mass o he o ni hop e and he pe ching mechanism.
+
o
is he eloci y o
he o ni hop e a e he impac and
−
o
be o e he impac .
Ic
is he momen um.
mpp
is he
mass o he pe ching place. In his case, his mass is assumed o be much la ge han he
mass o he ae ial pla o m.
+
p
is he eloci y o he pe ching place a e he impac and
−
pp be o e he collision. Finally, eis he es i u ion coe icien .
The s udy assumes ha he mass o he pe ching spo is much bigge han he mass o
he ae ial pla o m. This condi ion makes he pe ching spo s eady du ing he pe ching
maneu e . So, he o ni hop e and he claws will manage all he ene gy gene a ed du ing
he impac . The sys em e alua es he o ni hop e ’s eloci ies a e and be o e he collision.
The las s ep is calcula ing app oxima ely he sys em’s s i ness o ob ain a magni ude
o de . A e de ining he design, he ma e ials used se he s i ness o he mechanism.
Howe e , he pe ching mechanism was unknown a his momen . The sys em was supposed
o be a Ca bon Fibe (CF) ba . The ollowing equa ion calcula es i s s i ness.
k=E·A
L(3.11)
whe e E is he Young module, A is he a ea o a ibe ca bon ube, and L is he leng h.
Finally, by balancing he sys em du ing he collision, he s udy ob ains he o ces he
mechanism abso bs. Figu e 3.9 shows he p elimina y p essu es he de eloped sys em has
o suppo .
Figu e 3.9
Impac esul s. The igu e shows he o ce gene a ed in case o pe ching wi h
an undamped sys em.
60 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
be ween he phalanxes and he join mus be ixed and s able. Howe e , he di e en
elas ici y o he ma e ial makes i di icul o each his equi emen . A ha poin , he
manu ac u ing me hod con empla ed gene a ing welding be ween he phalanx and he
join s using he mel ing poin o he plas ic. Howe e , i modi ies he p ope ies o he
join , making i impossible o analyze wha was happening inside.
In he second case shown in Figu e 3.13b, he s uc u e is con inuous o he same
ma e ial ha ing and s uc u al join be ween he phalanx and he join s. The absence o
ex e nal manu ac u ing me hods ha modi ies he ma e ial’s p ope ies makes he sys em
and he manu ac u ing mo e s able. The s udy o his con igu a ion con empla es using a
can ile e app oach o model he lexion o he join s.
Figu e 3.14 shows he pa ame e s o each join o model he sys em. The dimensional
p oblem con empla es he a ia ion o hese pa ame e s o achie e he design condi ions.
The dimensional pa ame e s a e he join hickness e, he join longi ude L, and he phalanx
hickness l. Each join has di e en dimensions con igu a ions o pe o m he app op ia e
holding mo emen as p oposed in Figu e 3.12.
Figu e 3.14
Join pa ame e s ep esen a ion. The igu e ep esen s he las join dimen-
sioning o he join s.
To analyze he dimensional p oblem o he join s, Figu e 3.15 shows he s ess s udy
pe o med on each join . The p ocedu e is o examine each join as a can ile e beam.
P ep esen s he inge ’s weigh om he join o he ip o he inge . T ep esen s he
o ce he endons deli e o each join . This s udy analyzed he unde o med shape. The
objec i e is o main ain he o m while he endons ac ua ion is no applied and s a he
mo emen when he ension is applied. In ha way, he s udy analyzes he join ’s c i ical
lexion.
Figu e 3.15 highligh wo scena ios:
•
Scena io 1( ep esen ed in ed in Figu e 3.15): In his case o s udy, he endons do
no apply o ce o he claw, only wi h he in luence o i s weigh . Psymbolizes he
weigh in he diag am. The join has o main ain i s shape du ing his scena io.
•
Scena io 2( ep esen ed in g een in he Figu e 3.15): In he second s a e, he endon’s
ension Tac ua es he claw. The objec i e is o educe he ension needed o close
he claw. In ha way, he s udy analyzes he momen um diag am looking o he
c i ical lexion poin .

3.3 Claws Mo phology 61
Figu e 3.15
Join S ess Diag am. The igu e shows he s udy pe o med in he inge join s
o esol e he dimensional p oblem. I highligh s wo scena ios. The i s one
wi h no ension applied and he second one in which he endon deli e s he
o ce o he inge s, ep esen ed in ed and g een, espec i ely. In he end, i
shows he s ess diag am in bo h scena ios.
The endon’s ension is a couple o magni ude o de s bigge han he weigh o he es
o he claw. Then by looking a he s ess analysis in 3.15, he ension is he main o ce.
The applica ion o he ension gene a es he mo ion o he inge s. The objec i e is o
educe he ension needed o close he claw applying, deli e ing a signi ican pa o i o
he g asping. As shown in he igu e, he ension o he endons gene a es an axial e o .
The dis ance om he cen e o he join o he poin o applica ion o he ension causes a
momen on he join . To maximize he o ce applied o he objec s g asped, his dis ance d
has o be maximized. Howe e , g ea e dis ance means mo e ma e ial and g ea e weigh .
The o mula ion p esen ed in Equa ions 3.13 analyzes he p e ious p oblem. The
s uc u e has o suppo Scena io 1 and collapse in Scena io 2. The ollowing equa ions
analyze he c i ical poin o he shea modulus and he momen um. Speci ically, he c i ical
shea s eng h o he phalanx mus be g ea e han he weigh i has o suppo (
Vc >P
),
and he momen esis ance mus be less han he o ce exe ed (Mc <Td).
62 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
I=1
12bh3=1
12Le3
Wz=Iz
yz
=1
6e3
Nc (e,L) = π2EI
(Lc )2=π2E
β212
e3
L
Vc (e,l) =
A
y
√3
γMo
=
el E
√3
γMo
Mc (e,l) = C1
π2EIz
L2sIwL2GI
Izπ2EIz
(3.13)
whe e
I
is he sec ion ine ia,
L
is he dis ance be ween he phalanx,
e
hickness o he TPU
join ,
l
is he hickness o he phalanx,
Wz
is he sec ion modulus,
Nc
is he ul ima e axial
load limi , Lc is he leng h in he bending plane conside ed, βis he bending coe icien ,
Vc
is he ul ima e shea s ess,
A
is he c oss sec ion a ea,
γMo
is he calcula ion sa e y
ac o ,
E
is Young’s modulus,
Mc
is he ul ima e momen suppo ed by he s uc u e,
C1
is
he coe icien ha egula e he bending condi ions,
G
is he shea modulus o he ma e ial
whe e
G=E
2(1+ν)
and
ν
is he Poisson’s a io,
I
is he o sional cons an and
Iw
is he
wa ping cons an .
Finally, he analysis shows ha he leng h be ween phalanges is di ec ly ela ed o
bending condi ions. The e o e, he s uc u e has o suppo he shea o ce
V
gene a ed by
he weigh di ec ly ela ed o he c oss-sec ion a ea. In he end, i has o collapse whi he
ension o he endons.
The ma e ial p ope ies ake pa icula ele ance a his momen o he analysis. As
explained be o e, he claw s uc u e is en i ely 3D p in ed. In his case, he phalanx and
he join s a e a con inuous s uc u e made o TPU. The p ope ies o his ma e ial a e
E=26 MPa, G=78.7 MPA, and elonga ion be o e b eak 580%. The heo e ical s udy se s
a p elimina y esul wi h hese p ope ies. Howe e , analy ical esea ch selec s he inal
geome y o each he desi ed esul s. In addi ion, he manu ac u ing p ocess se s he
p in ing in ill o 100%, which means he join is solid wi hou ai inside. This con igu a ion
se s he make ha he join p ope ies a e he same as he ma e ial. I also educes he
complexi y o he s udy o he ma e ial.
The ollowing ex analyzes he beha io o he ways o gene a e he closing mo emen s.
The i s one is o use he same dimension in all he join s. Then, he endon’s inse ion
poin s cause a di e en ial ac ua ion o pe o m he bioinspi ed closing mo ion. Na u e also
inspi ed his con igu a ion. The di e ence is ha in na u e, he bi d has a muscle capable
o exe ing g ea o ces managing each oe. Howe e , in ou design, only one ac ua o
d i es all he endons and oes.
A phalanx wi h wo join s was analyzed o e alua e he beha io o he di e en ial
momen gene a ion me hod o he closing. The echnique aims o gene a e a mo e
signi ican momen a ound he i s join han he second. In ha way, i models he
inse ion poin s o endons in he bi ds’ bones. Figu e 3.16 shows his me hod o ac ua ion.
3.3 Claws Mo phology 63
Figu e 3.16
Di e en ial momen gene a ion ac ua ion me hod. The igu e ep esen s he
di e en ial momen gene a ion ac ua ion me hod. The pink colo ep esen s
he posi ion o he endon inside he oe. The igu e also shows he pa ame e s
o he ac ua ion me hod design.
The ollowing equa ion calcula es he momen gene a ed a ound he i s and he second
join depending on he angle o he endon and he inse ion poin using he pa ame e s
shown in Figu e 3.16.
M1=cos(φ1)·T·D1
M2=cos(φ2)·T·D2(3.14)
whe e
M
is he momen gene a ed a ound he join ,
φ
is he angle o di ec ion o he enso ,
T
is he ension o he enso , and
D
is he dis ance be ween he inse ion poin and he
inse ion poin o he endon and he cen e o he join .
Analyzing he equa ions and he mo ion’s equi emen s,
M1
has o be bigge han
M2
.
This means ha
cos(φ1)·D1>cos(φ2)·D2
. Analyzing his esul o ake he majo
ad an age o he ension o ce
φ1
has o be as smalle as possible and D1 as bigge as
possible. The limi o he angle o
φ
is he ic ion gene a ed. I
φ1
is e y small, he
ic ion be ween he endon and he phalanx inc eases, abso bing much o he o ce.
Rega ding he dis ance be ween he cen e o he join and he inse ion poin o he
enso , he limi is he dis ance ha makes
φ=0
. Figu e 3.17 shows se e al es esul s o
analyze he join s’ cha ac e is ics, longi ude, and hickness.
Figu e 3.17 shows he o ce needed o close he claw using di e en join con igu a ions.
The longi ude Land he join e hickness ha e been analyzed. I is impo an o emembe
ha he sys em has o p o ide he desi ed di e en ial ac ua ion o main ain he bes
con igu a ion o he holding mo emen s.
The expe imen al esul s show ha he inge s achie ed he objec i e o ha ing a di e -
en ial ac ua ion be ween he phalanx. The p ima y dependence o close he join is he
hickness o he join , as shown in Equa ions 3.13. Howe e , he o ce needed o close he
inge s using his ac ua ion is a ound 1 kg/ inge , meaning he muscle had o pe o m 4 kg
o close he claw. Analyzing he sys em, he majo disad an age o his kind o ac ua ion
is he ic ion gene a ed be ween he endons and he walls.
A signi ican pa o he o ce is los on hese walls inc easing he ension needed o
close he inge . Howe e , he s udy con inues pe o ming addi ional es s analyzing he
s eng h o open he inge a e he holding mo ion. The objec i e is o e alua e i he
64 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.17
Analisis o he join s’ cha ac e is ics Con igu a ion 1. The igu e analyzes he
endon ension needed o close he inge s depending on he con igu a ion o
he join s. The blue colo ep esen s he o ce o close he i s phalanx, and
he o ange he second phalanx.
ic ion ha causes di icul ies in closing he claw con ibu es posi i ely, making i mo e
di icul o discha ge he mechanism con ibu ing o he holding.
In ha way, he oe endons we e p e-s essed a 2.5 kg o ension. Then, a dynamome e
measu es he esul an ex e nal o ce needed o open each phalanx. Figu e 3.18 shows he
esul s o hese expe imen s.
The esul s conclude ha ic ion posi i ely a ec s he o ce o open he claw. Also,
in hese esul s, a dependence o his o ce on he dimensions o he join s was ound,
eaching a maximum on he esis ance o ce a 1.5 mm o join hickness. Howe e , he
endon ension needed o close all he inge s o he claw is oo big. This con igu a ion
complica es he ac ua ion needing up o 4 kg o close he claw’s inge s en i ely. The e o e,
his design equi emen disca ds his ac ua ion con igu a ion.
The second con igu a ion o ac ua e he inge s is a longi udinal endon pa allel o he
cen al axis o he inge . This enso goes as nea he palm as possible o gene a e he
maximum possible momen . Then o gene a e he desi ed mo emen , he dimensions o
he join a e modi ied. The idea was o inc ease esis ance om he i s join o he claw’s
ip. The esis ance o he join s inc eases wi h he hickness o i s. Figu e 3.19 ep esen s
his ac ua ion me hod.
This con igu a ion aims o gene a e he maximum possible momen in all he join s.
Then by modi ying he esis ance o he join , ob ain he mo ion wan ed. Acco ding o
Equa ions 3.13, o se he join s’ desi ed esis ance, he pa ame e s a e: e(join hickness),
L(join longi ude), and lphalanx hickness. The decision was o a y he hickness,
main aining he dimensions o he es o he pa ame e s. As shown in Figu e 3.19, he
hickness o he i s join is smalle han he es o he join s. Then Equa ions 3.13 we e
used o calcula e he minimum hickness in Scena io 1 desc ibed abo e (no ension in he
endon). The analysis o he esul s se s a join 0.7 mm in hickness and 3 mm in longi ude
was selec ed. Finally, he design es ablishes a i s join o 1 mm o abso b he ine ia
3.3 Claws Mo phology 65
Figu e 3.18
Ex e nal o ce needed o open he join s in Con igu a ion 1. The o ange colo
ep esen s he o ce esul an in kg needed o open he i s join , and in blue,
he second join .
Figu e 3.19
Finge longi udinal enso ac ua ion me hod. The igu e ep esen s in pink he
posi ion o he endon inside he inge . I highligh s he di e en ial hickness
o each join .
gene a ed by he lapping and mass. Then he es o he join s inc eased linea ly by 0.25
mm o each join .
A e he join s be ween he phalanxes a e analyzed, he claw’s s uc u e s udy s a s.
As p esen ed be o e, looking a na u e, he idea is o de elop an anisodac yl claw s uc u e
inspi ed by eagles and alcons due o hei sizes and capabili ies. The nex s ep is o c ea e
he inge s uc u e. Se e al app oaches e alua e he inge s uc u e o selec he size
numbe and he size o he phalanxes. The goal is o de elop a claw s uc u e simila o
ha o eagles, imp o ing he adap abili y o g asping objec s inspi ed by pa o s. Figu e
3.20 shows an image o he s uc u e ha inspi ed he de elopmen o he claw.
Figu e 3.20 shows he bone s uc u e o he eagles’ claws. The Hallux, Digi 1, o back
inge , is he p incipal one gene a ing he o ce du ing he g asping. This one, combined
wi h he on cen al one, p o ides he majo pa o he o ce when holding objec s.

66 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.20 Eagle claw [130]. Mo phology ha inspi ed he de elopmen o he manipu-
la o .
F on -side inge s gene a e s abili y in he g ip and pe ch. The size o he inge s changes
depending on he di e en species o eagles. Some species ha e a smalle , s i e hallux o
gene a e g ea e o ce. The capabili ies his oe suppo s a e simila o he ones p o ided
by he humb inge in human hands.
The de elopmen o he claw s udy is ocused on he de elopmen o wo kinds o
inge s: he main inge and he equilib ium inge s. The main oe is he one ha can exe
he g ea es o ces. This inge supplies he capabili ies suppo ed by he Middle alon
and hallux. The equilib ium inge p o ides he la e al balance du ing he pe ching and
g asping. I mimics he abili ies o he Ou e and Inne alons.
The i s app oach p esen ed in [
75
] has he same size in all he oes o he phalanx and
he join s. The a ia ion be ween inge s was he numbe o phalanxes; he on cen al
and back inge s we e one phalanx bigge han he o he . This condi ion p o ided good
adap abili y o objec s. Howe e , i was impossible o hold hings au onomously. This
beha io is due o he absence o he desi ed closing mechanism. In addi ion, he ac ua ion
o he Middle oe and Hallux did no ma ch he beha io shown in na u e.
A e an exhaus i e analysis o he unc ionali ies and shape o he inge s, his s udy
highligh s a ious concep s abou bi ds’ claws summa ized on wo cen al ones.
The i s one is ha no one species o bi d ha e he same claw mo phology as o he
species, including species o he same amily. O cou se, he di e ences in he claw o
di e en amilies o species a e mo e signi ican . Howe e , he size o he oo a ies in
he same amily, depending on whe e he animals li e and he ype o die . Fo example,
claws p epa ed o hun ing bigge p ey show sha p nails and smalle s i hallux alons
capable o exe ing g ea e o ces. Con e sely, i hey hun smalle p ey, hei nails a e
usually smalle , which educes he inju y isk.
The s i ness o he inge s ook pa icula ele ance when i wan ed o exe g ea e
o ce. The i s de elopmen o he claw eached some limi s. One p incipal limi was he
weigh he claw could ca y ou . Du ing his las pa o de elopmen , one o he main
p ope ies ound in he ee o bi ds ha wan o ca y hea ie weigh s is he igidi y o
hei hallux, which makes his oe capable o wi hs anding g ea e o ces.
3.3 Claws Mo phology 67
Figu e 3.21
Middle and Hallux alons e olu ion. The igu e shows a scaled compa a i e o
he di e en oes de eloped o he claw’s Hallux and Middle alons posi ions.
The a ow se s he di ec ion o he e olu ion om he i s e sion o he las
e sion. The legend shows he scale o he igu e. The colo o he pa s o
he claw made o PLA is blue, and he g ay o he ones made o TPU.
The conclusion is ha he bi ds’ claws ound in na u e ha e shown an excep ional
adap a ion o whe e hey li e. Howe e , de eloping a uni e sal claw capable o holding
and pe ching on a wide ange o objec s and places is mo e challenging.
3.3.3 Claws Design E olu ion
The documen ocuses now on he i e a ions pe o med o de elop he s uc u e o he claw.
A e analyzing he na u e and se e al i e a ions, he decision was o c ea e a claw wi h wo
kinds o inge s. One can exe a g ea e o ce on he on cen e and back inge s(Hallux
and Middle Talon). The o he is mo e adap able and app op ia ely used in he on la e al
inge s(Inne and Ou e Talons). The on cen e inge and back inge a e he main
o ou claw, and he la e al inge s p o ide equilib ium o he g asping. S a ing wi h
he de elopmen o he main inge s(Hallux and Middle Talon), Figu e 3.21 shows he
e olu ion du ing he de elopmen .
The ollowing lis summa izes he e olu ion o he main de eloped obo ic oes eplica -
ing he Hallux and Middle alons.
68 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
•
Finge I: I was he i s de eloped. I comp ised h ee phalanxes and nails made
o PLA. The join s we e made o TPU, ha ing a cons an hickness o 1 mm. The
nail made o PLA eplica es he nail o a small bi d. The dis ibu ion o he endon
ension was longi udinal and pa allel o he cen al axis o he inge . The s udy o
he inge also e alua ed o a y he numbe o phalanxes. The es s concluded ha
he s i ness o he inge dec eases wi h he numbe o phalanxes. Tha inc eases
he adap abili y educing he capabili y o exe o ces wi h he oe.
•
Finge II: I was he i s inge ha ied o de elop speci ic closing mo emen s.
I used he di e en ial endon ac ua ion me hod p esen ed in Figu e 3.16. In his
inge , he s i ness inc eases, inc easing he capabili y o exe o ce. In his case,
he phalanx ma e ial is PLA, and he join ma e ial is TPU, bo h 3D p in ed. The
join s had he same hickness due o he equi emen s o he di e en ial ac ua ion.
The main p oblem ound was adap abili y. The s i ness o he inge was high,
di icul ing hei adap a ion capabili ies.
•
Finge III: The main inno a ion o his inge was he de elopmen o a con inuous
s uc u e made o TPU ha ing a s uc u al join be ween he join s and phalanx. This
inge educed he s i ness o he whole inge using he lexibili y o he TPU in he
phalanx. Howe e , using he TPU combined wi h he di e en ial momen gene a ion
sys em inc eases he ic ion be ween he enso and he phalanx. Analyzing his
scena io, he o ce needed o close he inge g ows, signi ican ly educing he
gene a ion g asping p essu e. Ano he inno a ion p esen ed on his inge is he
linea educ ion o he hickness o he phalanx om he beginning up o he nail.
Tha educes he mass o he inge , making i possible o educe he hickness o
he join s. Ano he skill p esen ed on his inge is using ha d edges in he palm
o he inge o con ibu e o ic ion gene a ion. This idea ied o emula e Ha d
Robo ics mechanisms, mimicking spikes o go inside he ma e ial. Ou idea was o
ha e ha de edges using mo e concen a ion o TPU on i . These so -ha d edges
looked o he objec ’s edges o go inside, gene a ing mo e ic ion.
•
Finge IV:Finge IV is he las inge -de eloped e sion. I includes he longi udinal
endon ac ua ion p esen ed in Figu e 3.19 and explained be o e. As shown in Figu e
3.21, he sys em has wo phalanxes and h ee join s made in a con inuous s uc u e
o TPU. The hickness o he join s inc eases om he i s one up o he nail,
pe o ming he desi ed closing mo ion. A s udy pe o med on he nail se s endon’s
posi ion o gene a e he g ea es momen . Finally, he nails used in his dis ibu ion
pe ec ly eplica e he shape o he nail o an eagle.
The de elopmen o he on la e al inge s(Inne and Ou e alons) p esen s a simila
e olu ion o he main inge s. The di e ence be ween he main ac ua ion inge s and
he la e al ones is hei s i ness, which also ansla es in o he numbe o phalanxes. In
addi ion, he la e al inge is a suppo ing inge ha is mo e adap able and used o p o ide
equilib ium o he sys em du ing he holding w apping a ound he objec s. Figu e 3.22
summa izes his de elopmen .
3.3 Claws Mo phology 69
Figu e 3.22
La e al inge s(Inne and Ou e alons) e olu ion. The igu e compa es he
oes de eloped o he claw’s Inne and Ou e alons posi ions. The a ow se s
he di ec ion o he e olu ion om he i s e sion o he las e sion. The
legend shows he scale o he igu e. The colo o he pa s o he claw made
o PLA is blue, and he g ay o he ones made o TPU.
The ollowing ex highligh he main di e ences be ween he inge e sions. The e olu-
ion is simila o he one shown on he cen al ac ua ion inge . Howe e , he equi emen s
o hese inge s a e di e en , a ying hei geome y.
•
Finge I: This oe is he i s e sion de eloped o he la e al inge s. This inge
was simila o he i s e sion o he o he ac ua ion inge s. An analysis o na u e
looking o bi ds inspi ed claws ounds one eagle specie wi h smalle la e al inge s.
The inge de eloped had TPU join s and PLA phalanxes o hese bi ds. This inge
was one phalanx smalle han he main ac ua ion inge s Ve sion I o he Middle
alon de elopmen . The ac ua ion me hod used was longi udinal enso ac ua ion.
•
Finge II: This is he second e sion o he la e al inge s. This e sion has simila
concep s o he i s one. The inge was smalle han he main one o p o ide
equilib ium o he pe ching and he g asping. The ac ua ion me hod applied was he
di e en ial endons ac ua ion.
76 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
(a)
Leg V3 in eg a ed in o GRIFFIN’s[
65
]
pla o m E-Flap[68].
(b) Leg V3 ende iza ion.
(c)
Schema ic ep esen a ion o he leg including he main pa ame e s.
Figu e 3.27
Leg hi d e sion.(a) Shows he legs in eg a ed in o GRIFFIN’s[
65
] pla o m
E-Flap[
68
]. (b) Shows he ende ing o he sys em, highligh ing he ma e ial
composi ion. (c) Shows he p incipal geome ical pa ame e s o he sys em.
The angle
β
ep esen ed in he Figu e 3.27c educes he legs’ on al su ace, main aining
he dis ance o ac ua ion o minimize he ae odynamic in luence.
The legs conduc he ene gy o he same dumpe o be e deli e he impac o ces
using an angle
α
o gua an ee he dis ance be ween hem. In his e sion, compa ed o
he p e ious e sion, he dumpe was made o TPU wi h a iangle shape o abso b he
ene gy. As a esul , his dumpe a oids p oblems o s abili y ound wi h he sp ing dumpe
p esen ed in he p e ious e sion. In addi ion, his 3D-made dumpe can easily modi y i s
in ill p ope ies by adjus ing he dumping a io o he shock o he pla o m.
Ve sion V4 p esen ed in igu e 3.28 was he mos complex leg. These legs combine all
he concep s p esen ed in he p e ious e sion and esol e mos o hem. The ollowing
lis enume a es he concep s o he de elopmen o he legs.
•
Impac abso bance: The mechanism o con ibu e o he pe ching passi ely and o
abso b he impac ene gy a e di ec ly ela ed. The idea is o use he ene gy abso bed
om he impac o con ibu e o he pe ching ans e ing his o ce o he claw. The
claws abso b he impac by using endons and he scisso mechanism.

3.4 Leg Mo phology 77
(a) Leg V4 in eg a ed in o he ae ial pla o m.. (b) Leg V4 ende iza ion.
(c)
Schema ic ep esen a ion o he leg including he main pa ame e s.
Figu e 3.28
Leg ou h e sion.(a) Shows he legs in eg a ed in o GRIFFIN’s[
65
] pla o m
E-Flap[
68
]. (b) Shows he ende ing o he sys em, highligh ing he ma e ial
composi ion. (c) Shows he p incipal geome ical pa ame e s o he sys em.
•
Passi e pe ching con ibu ion: The passi e damping mechanism de eloped wi hin
his e sion con ibu es o he pe ching. This mechanism uses he impac o con-
ibu e o he pe ching. In addi ion, his bi d-based mechanism uses he endons
inside he legs o con ibu e o he pe ching. A e he shock, his mechanism akes
ad an age o he mass o he o ni hop e o main ain he pe ching. Then, he ac i e
ac ua ion combined wi h his one gene a es enough ac ua ion o keep he pe ching.
•
Low ae odynamic in luence: As shown in Ve sion V2, he ae odynamical in luence
o he claw is impo an . This de elopmen implemen s wo ways o educe ae ody-
namic in luence. The i s one is o use he same ac ua ion con igu a ion p esen ed
in V2.V4 sol es he p oblems ound in V2 wi h his mechanism by using Ti anium
sc ews, a bea ing, and C.F. laye -composed pulleys. The second one is ela ed o he
posi ion o he claw du ing he ligh . The objec i e is o ly wi h he claw backwa d,
educing he ae odynamic pe u bances. Be o e he pe ching, he hip join mo es
he claw o wa d, o ien ed o he pe ching spo . In na u e, bi ds ly wi h hei claws
back and hen change hem o wa d o pe ching.
•
Possi ioning con ol: The e sions p esen ed be o e do no ha e he op ion o con ol
he posi ion o he claw. I is impossible o each equilib ium in he pe ching by
78 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
balancing he claws. This e sion uses a hip join ac ua ed by SMA muscles o
con ol he posi ion o he claw educing he o que needed o keep he pe ching and
manage i s posi ion du ing he ligh .
The lis p esen ed be o e enume a es he main concep s achie ed in de eloping Ve sion
4o he leg. The ex con inues ocusing on his de elopmen and he way o achie e hese
esul s.
As shown in he igu e, his leg is one o he mos complex in de elopmen . I is only
possible o unde s and his de elopmen by looking a he p e ious ones. CF is his leg’s
main componen , inc easing he ab ica ion p ocess’s complexi y.
This de elopmen shows he in eg a ion o wo cen al sys ems o he i s ime: he
an agonis muscles con igu a ion and he sys em ha abso bs he impac con ibu ing o
he pe ching. The passi e pe ching/impac abso bance and leg posi ioning con ol sys ems
a e p esen ed i s . Figu e 3.29 shows his sys em.
Figu e 3.29
Leg dumping sys em ep esen a ion. The igu e shows in yellow he endon
posi ion inside he legs and highligh s he main geome ical pa ame e s.
The sys em comp ises eigh ansmi e s in he shape o a scisso s mechanism. Then, a
p isma ic join o one deg ee o eedom connec s he leg and he hip join . The e a e wo
poin s o iew conce ning passi e pe ching, he mechanism, and he endons.
Focusing on he mechanism: Ini ially, he claws ecei e he impac shock. Then, he legs
ansmi his o ce o he mechanism. In he end, he p isma ic pa con e s his o ce o
one di ec ion. I closes he scisso mechanism e ically and expands i ho izon ally. The
endons abso b he ho izon al expansion. They connec he claw and he scisso mechanism.
The o ce gene a ed in he la e al expansion is ansmi ed h ough he endons o he claws.
The ac ha he o ce needed o close he claws inc eases p opo ionally wi h he closing
mo ion makes i possible o gene a e a dumpe . In he end, he claws hemsel es a e ou
sys em’s dumpe .
The sys em design gene a es as much o ce as possible o con ibu e wi h he SMA
ac ua o s o he holding. The e o e, he mechanism mus cause ension o close he claw
du ing impac . The geome y o he scisso mechanism models he quan i y o o ce
3.4 Leg Mo phology 79
deli e ed o he claw. Making a balance and a g aphical analysis, he equa ions o ob ain
he longi ude o he ansmi e s a e he ollowing:
4T=Fimpac ·cosα
△L endon =2·L(cos(αmin)−cos(αmax)) (3.15)
whe e
T
is he ension o he endon.
Fimpac
is he o ce o he impac ,
α
is he pa ame e
ha de ines he posi ion o he scisso mechanism,
△L endon
is he inc emen o he endon
longi ude,
L
is he longi ude o he ansmi e o he endon mechanism,
αmin
is he
pa ame e ha de ines he inal posi ion o he mechanism, and
αmax
is he pa ame e ha
de ines he ini ial posi ion o he mechanism.
Fimpac
is ob ained heo e ically in Sec ion 3.2 o e alua e hese equa ions.
△L endon
is wo cen ime e s, he longi ude necessa y o close he claw.
αmin
is se o 15
◦
due o
he geome ical limi . Finally, by esol ing he sys em, he longi ude ob ained o he
ansmi e is
L=46.2
mm. This sys em eplica es he beha io o bi ds’ legs while
pe ching. Howe e , his solu ion is adap ed o he pe ching con ou condi ions o he
o ni hop e s.
Two muscles in he an agonis con igu a ion manage he posi ioning sys em o he legs.
A pulley placed on he hip handles he ension gene a ed by he SMA, inc easing he
momen gene a ed o con ol he posi ion. Figu e 3.30 syn hesizes he hip join posi ioning
con ol sys em.
Figu e 3.30
Claw posi ioning con ol. The igu e ep esen s he main geome ical pa ame-
e s o he sys em. I also desc ibes he way o connec he an agonis muscles
con igu a ion o he join using a yellow line o highligh he posi ion o he
endons.
The sys em has o main ain he leg back du ing he ligh , educing he sys em’s ae o-
dynamic in luence, and hen o wa d o p epa e he pe ching. Du ing he pe ching, his
sys em mus main ain he o ni hop e ’s posi ion. SMA 1 ac ua o mo es he claw o wa d,
and SMA 2 mo es i backwa d. Sec ion 3.5 p esen s a con ol sys em o close he loop o
he posi ion wi h he empe a u e o he SMAs.
80 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
By applying Eule -Lag ange, he dynamics o he hip join a e analyzed.
(Izz +MeqR2
CoG)¨
θ( )−MeqRCoGgsinθ( ) = ∑FSMA( ) −c˙
θ( ),(3.16)
whe e
Meq(kg)
is he equi alen mass o he sys em in he CoG,
Izz(kg.m2)
is he momen
o ine ia o he leg,
R(m)
is he dis ance om he o a ional poin o he CoG,
g=9.81m/s2
is he g a i y cons an ,
θ( )( ad)
is he gene alized coo dina e o he sys em,
∑F(N)
is he
esul an o he o ces o he SMA sp ing in an agonis con igu a ion,
(m)
is he in e nal
adius o he pulley and c(kg.m/s)is he dynamic ic ion coe icien .
The ex ocuses now on he manu ac u ing me hod and inal p o o ype. The leg is
en i ely made o CF pla es. I is composed o 30 di e en pa s o each he desi ed
speci ica ions. Fi s , he pa has been designed laye by laye wi h a CAD p og am. Then,
a Compu e Nume ical Con ol (CNC) cu s he laye om CF pla es o 1.5 mm o 2mm in
hickness, depending on he componen . Las ly, he pa s a e glued oge he using Epoxy
esin and a acuum sys em. Figu e 3.31 de ails he in eg a ion p ocess.
(a) Laye s cu wi h a CNC. (b)
Laye ’s mechanical in eg a ion.
Figu e 3.31
Leg Laye in eg a ion p ocess. (a) shows he laye s cu om he CNC machine.
(b) shows he mechanical in eg a ion o he laye s glued wi h Epoxy and using
CF ba s o gene a e a mechanical union be ween i s.
Figu e 3.31 de ails he laye in eg a ion p ocess. Fi s , he laye s we e glued using
Epoxy; sc ews we e used o gua an ee he posi ion du ing d ying. A e ha , he sc ews
a e subs i u ed by CF ba s o make a pe manen mechanical join be ween laye s. In he
end, e e y hing is sanded, p o iding he inishing ouch.
All he concep s lea ned in Leg V4 ha e been applied o a wo-legged e sion o his
de elopmen shown in Figu e 3.32.
The igu e shows he in eg a ion o he mechanism in a wo-leg sys em. Figu e 3.32a
shows he leg posi ion o ien ed o he pe ching spo . Figu e 3.32b shows he legs’ ole
du ing he ligh . The inclusion o wo legs simpli ied he la e al equilib ium. Thei
posi ion backwa d educes he ae odynamic in luence o he manipula ion sys em as bi ds
do in na u e. Finally, Figu e 3.32c shows he in eg a ion in o he GRIFFIN pla o m.
3.4 Leg Mo phology 81
(a)
Schema ic ep esen a ion o he legs’ posi ion
be o e pe ching.
(b)
Schema ic ep esen a ion o he legs’ posi ion
du ing he ligh .
(c) Leg’s in eg a ion in o E-Flap[68].
Figu e 3.32
Two legs in eg a ion. (a) and (b) shows a schema ic ep esen a ion o he
sys em in he posi ions o pe ching and du ing he ligh , espec i ely. (c)
displays he sys em in eg a ed in o E-Flap[68] .
Figu e 3.33 shows a dynamic analysis o e alua e i he sys em can suppo he pe ching
shock.
This analysis used he S ess Analysis en i onmen o In en o
®
so wa e. The idea is
o simula e he o ces in ol ed in he pe ching, analyzing he weak poin s o he s uc u e.
Two o ces o 125 N we e applied in he cen e o he palm o each claw o emula e he
pe ching s esses. The analysis conside s he scisso mechanism s uck in he posi ion in
which i ansmi s all he endon ension o he claw. The s ess esul s conclude ha all
he de o ma ions made du ing he impac a e elas ics. We ound some weak poin s in he
scisso mechanism’s ansmi e s, bu hey can s ill suppo he s ess gene a ed. Finally,
20% o ole ance was conside ed o p e en a igue damage in he di e en pa s o he
mechanism.
The sys em de eloped can main ain he geome y o he legs du ing he pe ching con-
ibu ing o i . The o ce gene a ed by he endons aking ad an age o he ene gy o he
pe ching is enough o close he claw. I also con ibu es o he sys em’s eac ion ime, help-
ing he SMAs’ wa ming ime. This sys em makes i possible o use he claws in o ni hop e s
educing he weigh and inc easing he ope a ional a ea. I is only unde s andable o ha e
one sys em wi h he o he .

82 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Figu e 3.33
Legs s ess analysis. The igu e ep esen s he legs mapped in colo s analyzing
he s ess in each pa o he sys em. The colo coding o he map is on he le
o he igu e. The yellow a ows ep esen he o ce loca ion and di ec ion.
This sec ion shows he e olu ion o he leg concep du ing his wo k. The legs we e
aking mo e ele ance and complexi y du ing he de elopmen . Ul ima ely, leg and claw
wo k oge he , keeping he bioinspi a ion. This esul also highligh s he pe sonal e olu ion
o he esea che du ing his pe iod.
The ollowing sec ion will ocus on he ac ua ion me hod implemen ed on he o ni hop e .
The no el y me hod p esen ed will o e come he payload limi a ions.
3.5 Ac ua ion Me hod and In eg a ion
This chap e desc ibes in eg a ing he di e en ac ua ion me hods de eloped du ing his
hesis. The e o e, I discusses he physical aspec o in eg a ing he sys em’s elec onics.
A he end o his chap e , he whole manipula ion sys em is eady o wo k oge he .
3.5.1 SMAs Ac ua ion Con igu a ion
This sec ion ocuses on he solu ions o in eg a e he SMAs in o he manipula ion sys em
p esen ed be o e. I esol es wo main equi emen s: he posi ion and con igu a ion o he
SMAs o ac ua e he claw and he elec onic de eloped o con ol he muscles.
The sec ion s a s analyzing he posi ion and he con igu a ion o he ac ua ion. As
shown in he legs designs o Sec ion 3.4, wo me hods exis o place he muscles in he leg,
bo h based on he an agonis muscle con igu a ion. These me hods use wo SMA muscles.
One is used o apply he o ce di ec ly o he endon; he o he muscle is opposed o his,
educing he ension applied o he endon. Figu e 3.34 p esen s wo me hods o place he
SMA using his ac ua ion me hod.
Figu e 3.34a p esen s he linea posi ion o he an agonis muscle ac ua ion me hod.
The ac i e muscle is di ec ly connec ed o he o he using a endon. The place o bo h
3.5 Ac ua ion Me hod and In eg a ion 83
(a) Linea muscles an agonis con igu a ion. (b) Pa allel muscles an agonis con igu a ion.
Figu e 3.34
An agonis muscles posi ioning me hod. (a) Shows a schema iza ion o he
posi ioning me hod in which bo h muscles a e in he same ace o he leg. (b)
P esen s he con igu a ion ha educes he space needed o place he SMAs
in opposi es aces o he leg..
ac ua o s is he on al pa o he leg. The main ad an age o his ac ua ion me hod is
i s educed complexi y. The e o e, no elemen s be ween he muscles educe he possible
losses in he ac ua ion. The main disad an age o his ac ua ion con igu a ion is he leng h
needed o ac ua e, making he leg la ge . As p esen ed in he p e ious sec ion, a la ge
limb causes he claw o need o deli e mo e o que o compensa e o he mass in balance
du ing he pe ch.
Figu e 3.34b p esen s a con igu a ion ha educes he leng h needed o he ac ua ion. In
his solu ion, one muscle is in on o he o he . The sys em ansmi s he ension om one
o he o he using a endon o connec bo h muscles and a pulley o educe he ension loss
a he edges. The pulley wo ks by educing ic ion wi h he axle. This solu ion di ides
he leng h needed o place he muscles by wo. This con igu a ion makes i possible o
educe he size con ibu ing o he manipula ion sys em’s pe ching, mass educ ion, and
ae odynamic in luence.
The con igu a ion used in he end was he second one. This con igu a ion educes i s
nega i e aspec s by educing he ic ion in he ansmission. The compa ison did no ind
a nega i e pa o his one compa ed o he o he .
3.5.2 Tendon D i en Ac ua ion
This sec ion ocuses on how o ansmi he o ce om he SMA sp ings o he inge s o
he claw. The ollowing es analyzed he ma e ial and he in eg a ion o he endons in he
manipula ion sys em.
84 Chap e 3. Bio-Inspi ed Manipula o s o O ni hop e s
Tendons connec he claw and muscles. They ha e special equi emen s due o he
ope a ing condi ions o he SMA. They mus wi hs and high empe a u es up o 145°C
combined wi h high o ces. Some adi ional ma e ials, such as nylon, show elas ic
de o ma ions due o he applied s ess. This de o ma ion abso bs a pa o he ene gy
educing he ac ua ion. In applica ions such as ishing, his p ope y educes he impac
o ce ecei ed by he ishe man. The o he p oblem wi h connec ing nylon di ec ly o
SMAs is ha i s mel ing empe a u e is lowe han he ac ua ion empe a u e. The mel ing
empe a u e makes i impossible o use nylon o his applica ion. Howe e , nylon has
excellen ic ion p ope ies due o i s polished su ace.
The ex abo e highligh s ha nylon is no a possible solu ion o be connec ed di ec ly o
he muscles. E alua ing he possible ma e ials wi h high-s ess esis ance and empe a u e
ole ance, Ke la appea s as po en ial ma e ial. Rega ding i s applica ions, Ke la was
combined wi h CF in some ab ics o a oid sc a ches. T adi ionally Ke la was used in
he mili a y indus y o make bulle p oo es s. Looking a he second applica ion implies
empe a u e esis ance and high-s ess esis ance. Analyzing he ma ke , we selec Ke la
in h eads and co ds. Table 3.2 compa es bo h ma e ials conce ning he p ope ies wan ed
o e alua e.
Table 3.2
Tendon ma e ial compa ison. The able compa es Nylon and Ke la o hei
use as endons.
Ma e ial Tensile s eng h
(MPa)
Elonga ion be o e
b eak
(%)
Suppo ed
empe a u e
(◦C)
Nylon [136] up o 607 up o 200 up o 120
Ke la [137] up o 3450 3-4 up o 425
The able ocuses on he p ope ies o he ma e ial. The analysis ocused on h ee
comme cial p oduc s, nylon ishing lines, ke la h eads, and ke la mic o co ds. Low-
empe a u e esis ance and elonga ion be o e b eak elimina ed nylon ishing lines om
he selec ion. Conce ning Ke la , i s high ic ion wi h he claw’s plas ic pa s makes i
di icul o use in ou design. In addi ion, he ic ion and he p esence o only longi udinal
ibe s in he h ead some imes cause b eaks du ing expe imen s. The e o e, Ke la mic o
co ds we e analyzed. The main di e ence be ween he Ke la h ead and Ke la mic o
co ds is he p esence o ibe s in all di ec ions. Tha gene a es a mo e homogeneous su ace
educing he ic ion coe icien be ween he endon and he pa s o he claw. Howe e ,
he ic ion coe icien is s ill high o use inside he inge s. So, he inal design uses wo
ma e ials. Nylon in he sys em’s pa in which he ension and empe a u e equi emen s
a e elaxed, and Ke la mic o co d in he o he places. [
138
] shows he Ke la mic o co ds
used wi h a wo king empe a u e o up o 400
◦
C and a ensile S engh o 50 kg. [
139
]
shows he p ope ies o he nylon used capable o esis ing up o 14.6 kg o longi udinal
ension.
Ano he c i ical opic in his sec ion is how o deli e he o ce o he inge s. In he i s
e sion o he claw p esen ed in [
75
], he sys em has a manually adjus ed enso mechanism
ha manages a ixed elonga ion o each endon. Then, be o e he objec expe imen a ion,
3.5 Ac ua ion Me hod and In eg a ion 85
(a) Con igu a ion 1. (b) Con igu a ion 2.
Figu e 3.35
Tendon o ce dis ibu ion. The blue lines ep esen he endons, he blue
ci cles he sliding nodes, and he ex highligh he posi ion o he muscles
and oes.
we adap he dimensions o each endon. In his e sion o he claw sys em, we de elop a
di e en ial endon sys em ha au onomously manages he o ce deli e ed o each inge .
One o he main ad an ages o his sys em is i s capabili y o adap o a wide ange o
objec s. In con as , he p e ious sys em has manually adap ed o each objec ’s shape. This
sys em can upda e he endon elonga ion du ing he holding and adjus he claw o he
objec .
The sys em is based on balancing he ension o he inge s’ endons. Each inge
has a endon, and hey a e in e connec ed using his sys em. The idea is o make an
uns able equilib ium. Fo example, a inge s uck ansmi s i s endon ension o he o he
in e connec ed inge s. Tha makes i possible o con inue w apping a ound objec s wi h
he e ogeneous shapes.
This sec ion p esen s he s udy o wo endon con igu a ions. In he i s one, based
on na u e, he on cen al inge and he back inge ha e mo e o ce han he es o
he inge . The e o e, he on middle alon doubles he p essu e exe ed by he on
la e al inge s. On he o he hand, he back inge has he same o ce as he h ee on
inge s. In he second con igu a ion, he ou inge s can exe he same o ce. Figu e 3.35
summa ises hese concep s.
Equa ions 3.17 and Equa ions 3.18 e alua e he sys em’s beha io depending on he
endon dis ibu ion se .
Con igu a ion 1:
FCen al =2·FLa e al
FF on =FCen al +2·FLa e al
FBack =FF on
FSMA =FF on +FBack
(3.17)
92 Chap e 4. Bio-Inspi ed Mo phing Su aces
mo phing ac ua ion me hods like piezoelec ics a e s udied. Thei way o ac ua ion is o
de o m he hos su ace gene a ing a con olled de o ma ion on i . On he con a y, se os
and o he adi ional me hods only ansmi a punc ual o ce. Bi ds’ a e ou e e ence o
ake ad an age o yea s o e olu ion.
The objec i e o he wings is o inc ease he con ol inpu s by con olling he wing’s
su ace. A he same ime, he ma e ials used a e so e , mimicking he bi ds’ wings. So
he idea is o a y asymme ically he su ace o he wings gene a ing ac ua ion on he
ROLL angle o he ae ial obo . The co e is made o elas ome ma e ials o imp o e he
human coope a ion capabili ies o he ae ial pla o m. S udying a bi d-based wing p o ile
enhances he gliding capabili ies and main ains he lapping beha io .
This chap e is s uc u ed as ollows. Sec ion 4.1 is ocalized on he de elopmen o he
mo phing ail s udying he ac ua ion me hod, mo phology, and ways o con ol i onboa d.
Sec ion 4.2 ocuses on he de elopmen o he wings. Mainly, his sec ion ocuses on
cons uc ing i o suppo he o ces gene a ed by he lapping. Finally, Chap e 5 p esen s
he expe imen al alida ion o he sys em.
4.1 Mo phing Tail
This sec ion ocuses on he de elopmen o he ail. The objec i e is o de elop a bi d-based
ail p ope o o ni hop e s. T adi ionally, as shown in [
140
], hey use plain igid su aces
ac ua ed by se os. Howe e , bi ds’ ails in na u e a e composed o ea he s mo ing in all
space di ec ions. Fu he mo e, he ac ua ion is no linea . Ins ead, he ac ua ion desc ibes
an exponen ial unc ion ending in he ip. Figu e 4.2 shows a na u al bi d ail o explain
he concep s.
Figu e 4.2 Pe eg ine Falcon [141].
The igu e shows he size and shape o he bi d’s ail. Conc e ely, i s ocuses on he
pe eg ine alcon. The alcon is pe o ming an app oach maneu e o he a ge . The V-Tail
is widely open, inc easing he con ol ac ion o he ail. The su ace o he ail open has
he same magni ude o de as he es o he body. This ac ua ion is impossible o pe o m
wi h adi ional ac ua o s.
The weigh and inabili y o gene a e de o ma ions in an a ea discha ge se os o m
he g oup o possible ac ua o s. Ins ead o se os Sma Ma e ials like Mac o Fibe

4.1 Mo phing Tail 93
Composi e
™
(MFC) show be e capabili ies o pe o m hese asks. This ma e ial can
pe o m de o ma ion in he whole su ace o he ac ua o . Also, hey a e capable o exe ing
high blocking o ces wi h educed weigh . Howe e , be o e his wo k, hey we e used
as senso s o measu e small de o ma ions. Usually, hey a e glued o o he su aces o
gene a e de o ma ion. This wo k s udies a new way o use MFC o de elop he ail’s
s uc u e, mimicking bi ds’ ea he s.
The ail will be able o eplica e he mo emen and he ac ua ion o a na u al bi d ail.
Addi ionally, o inc ease he ac ua ion, he con ol su ace is augmen ed. The combina ion
o he bi d-inspi ed mo ion and he majo su ace makes i possible o inc ease he con ol
capabili ies o he ae ial pla o m. Ul ima ely, i will inc ease he con ol capabili ies
making i possible o load mo e weigh . In addi ion, o he applica ions like manipula ion
will ake ad an age o weigh educ ion.
The sec ion is s uc u ed as ollows. Fi s , i s a s wi h s udying he ac ua o s used
in he ail. Then, i con inues wi h he de elopmen . In he end, Chap e 5 p esen s he
expe imen al alida ion.
4.1.1 MFC
NASA de eloped MFC in he ea ly ’90s. Ini ially, hey c ea e MFC o wo k as senso s o
measu e small de o ma ions due o he sensibili y o hei in e nal esis o o he displace-
men . Then, esea che s disco e ed he high locking o ce exe ed by hese unde ce ain
condi ions. The capabili ies o MFC o pe o m ac ua ion in an a ea makes hem a good
candida e o mo phing applica ions. Figu e 4.3 ep esen s i s shape and lexibili y.
(a) MFC. (b) MFC lexibili y [142].
Figu e 4.3
MFC ac ua o . (a) Shows he geome y o he MFC. (b) Shows he lexibili y
and hickness o he ac ua o .
MFCs a e ligh weigh ed and hin ac ua o s ha canno pe o m displacemen s. The
displacemen ’s capabili ies depend mainly on he ma e ial o he hos lamina. The e o e,
he s uc u e o MFC laye s is c ucial o unde s and i s beha io . They a e a piezoelec ic
ma e ial sandwiched wi h elec odes capable o gene a ing a speci ic elec ic ield o
modi y he shape o he piezoelec ic. The es o he laye s gua an ee he isola ion o he
elec ic ield. Figu e 4.4 shows an example o he laye s uc u e o he MFC. Speci ically,
he igu e shows he ones used in his wo k.
94 Chap e 4. Bio-Inspi ed Mo phing Su aces
Figu e 4.4
MFC laye s uc u e. The igu e shows he laye s uc u e. The a ows a ound
he igu e highligh he ma e ial o each laye . The laye s a e symme ic om
he Piezoelec ic ma e ial laye .
The laye in cha ge o he ield gene a ion is he laye o coppe elec odes. In his case,
his laye ed design se he ac ua ion o wo k be ween 0V and 1500V. The isola ion laye s
a e made o ac ylic and Kap on o a oid sho ci cui s. The piezoelec ic ma e ial in he
cen al laye expands depending on he ield gene a ed.
The de o ma ion pe o med by he elec ic ield in he piezoelec ic gene a es he de o -
ma ion. The piezoelec ic ma e ial is a oam-based ma e ial ha p oduces an expansion
wi h he elec ical ield. The laye in cha ge o he ield gene a ion is he elec ode coppe
laye .
The ac ua o i sel is capable o gene a ing minimal de o ma ions in he longi udinal
di ec ion. Howe e , he o ce deli e ed in his di ec ion is high. The e o e, he objec i e
o he ail is o con e i s de o ma ion o he o hogonal di ec ion.
Di e en esea che s e alua e he wo k o hese ac ua o s. Due o i s isola ion capa-
bili ies, unde wa e applica ions like [
56
,
57
,
58
,
59
,
62
] use his ac ua o . Then a e
hese app oaches, o he esea che s s udied he use o MFC in ixed wings in wo ks like
[
105
,
106
,
107
]. The wo k de eloped in his hesis is a combina ion o bo h wo ks. The
main di e ence be ween unde wa e wo ks and he wo k de eloped in his hesis a e he
en i onmen and he locomo ion gene a ion. In his hesis, he domain is he ai , which
isn’ wan ed o gene a e locomo ion. Howe e , he displacemen needed o con ol he
ajec o y o he o ni hop e is signi ican as he one pe o med on unde wa e applica-
ions. The densi y o he wa e is also bigge han he ai , making i necessa y o gene a e
s onge mo emen s. Ul ima ely, o o e come he con ou condi ions o he pla o m, he
ail de eloped in his hesis wan s o c ea e as , long, and s ong mo ions as he wo ks o
unde wa e applica ions.
Conce ning he wo ks in ixed wings, hey use hese ac ua o s o modi y he wing’s
su ace. These publica ions show wo kinds o esul s. The i s uses he MFC ac ua o s o
al e he cambe line o he wing p o ile, and he second employs hese ac ua o s as laps.
The objec i e is o add mo e inpu o con ol he ajec o y. The displacemen s wan ed on
hese publica ions a e smalle han hose wished o pe o m.
4.1 Mo phing Tail 95
Addi ionally, hey use ha d su aces o glue he ac ua o s. Usually, hey u ilize me al
shee s and Ca bon Fibe (CF) shee s. These su aces make i possible o gene a e a highe
blocking o ce losing in he way mos o he displacemen .
Using hese ac ua o s is desi ed o de elop an en i e ail ha mimics he beha io o he
bi d’s ail. The s uc u e is he MFC combined wi h o he ma e ials o gene a e enough
ac ua ion mimicking he bi ds’ ea he s. As bi ds’ ails a e so and adap able, he one
de eloped is so and capable o inc easing sa e y du ing he in e ac ion.
De eloping a so ail made o MFC o o ni hop e s is no a di ec in eg a ion. Di e en
design equi emen s de ine he way o in eg a e hese ac ua o s in o he o ni hop e .
•
So in eg a ion: Usually, esea che s employ ha d ma e ials o glue he ac ua o
blocking he expansion o he ac ua o , eaching he displacemen a educed dis-
placemen in he o hogonal di ec ion. Con a y o his, high displacemen is wan ed.
Using hese ac ua o s is wan ed o eplica e he ea he s o he ail o a bi d. So
he expansion has o be blocked by using o he ma e ials ha main ain he so
capabili ies o he sys em. The use o mo e lexible ma e ials enables he gene a ion
o bigge de o ma ions. Ul ima ely, he ac ua o mus combine o hogonal lexibili y
and longi udinal s i ness.
•
Flapping equency: The lapping equency o he o ni hop e s goes h ough he
en i e body o he o ni hop e gene a ing a high ine ia e en on s i ma e ial. The
ac ua o ’s s i ness is c ucial o educe he pe u bances on he ail.
•
Weigh : Weigh is one o he main limi a ions o o ni hop e s. Compa ed wi h o he
sys ems on he o ni hop e , he weigh o he ail is mo e c i ical e en due o i s
dis ance o he CoG. I could gene a e a passi e momen ha makes he pla o m
uns able. The objec i e is o eplace he wo se os and he ail’s s uc u e o
an en i e mo phing su ace made o MFC, educing ha weigh and gene a ing a
bi d-inspi ed ac ua ion.
•
Bimo ph ac ua ion: The ail has o ac in he ac ua o ’s o hogonal di ec ions. MFC
can only ac ua e in one way, so he idea is o combine i o ha e ac ua ion in all he
ways.
The so MFC’s in eg a ion aims o block he expansion in he longi udinal di ec ion.
The ac ua o s ops he g ow h by using CF h eads and epoxy. This app oach also keeps
he o hogonal lexibili y. The ollowing sec ion heo e ically models he beha io o he
MFC o wo k as so mo phing ac ua o s.
Modeling MFC
To analyze he mo emen ha MFC can de elop, he Rayleigh-Ri z me hod was applied,
ollowing he de elopmen s pe o med in [143].
As was highligh ed be o e, he MFCs a e composed o laye s o di e en ma e ials.
The e o e, analyzing he laye s and hos lamina is c ucial o ob ain he p ope ies o he
ac ua o . In ha way, wo ks like [
143
] use he Classical Lamina ion Theo y (CLT) o
es ima e and compa e he p ope ies o he ma e ials wi h hose gi en by he manu ac u e .
Then, ou wo k uses he p ope ies es ima ed in his wo k.
96 Chap e 4. Bio-Inspi ed Mo phing Su aces
Howe e , i is necessa y o unde s and he mechanical p ope ies o he MFCs o analyze
he ac ua ion. Mechanically, he MFCs a e composed o wo laye s o iso opic ma e ials
and ano he wo o o ho opic ma e ials. Iso opic ma e ials ha e he same composi ion
in all di ec ions and o ien a ions, so hei mechanical p ope ies a e he exac in-plane
di ec ions. So by de ini ion, Kap on and epoxy laye s a e iso opic laye s. O ho opic
ma e ials ha e di e en p ope ies depending on he o ien a ion. In he case o MFCs,
he coope elec odes laye has a di ec ion o 90
◦
, and he piezoelec ic ma e ial has a
on age o 0
◦
. Figu e 4.3b is an excellen example o highligh hese concep s. The igu e
highligh s he o ho opic p ope ies o he whole ac ua o . I is shown he longi udinal
lexibili y compa ed wi h he o hogonal s i ness.
Be o e s a ing he calcula ion, i is necessa y o de ine he e e ence sys em(see Figu e
4.5). In his hesis, he hos lamina and he MFC ha e he exac longi ude in he XY plane.
Howe e , he hickness o he MFC laye and he hos lamina di e . The e o e,
Lx
and
Ly
de ine he longi ude in he XY plane, and
HMFC
and
HHOST
de e mine he hickness. The
igu e also ep esen s he MFC geome y in he e e ence sys em.
Figu e 4.5
Ac ua o geome y pa ame e s. The igu e ep esen s he geome y pa ame e s
o calcula ing he p ope ies o he MFC. I shows he axis o he calcula ion
de ined in black a ows wi h he name o each axis. The da k yellow ep esen s
he MFC, and he ligh yellow is he ac ua o ’s hos lamina.
Following he app oxima ion de eloped by Schul z in [
144
,
145
], he midplane’s ou -
e m displacemen equa ion mus be assumed o apply he Rayleigh-Ri z me hod. The ou
unknown coe icien s c1,c2,c3,c4de ine he equa ion.
4.1 Mo phing Tail 97
u0(x,y) = c3x−c2
1x3
6−c1c2xy2
4
0(x,y) = c4y−c2
2y3
6−c1c2x2y
4
w0(x,y) = 1
2(c1x2+c2y2)
(4.1)
whe e
u, ,w
a e he displacemen s in he midplane in he axis
X,Y,Z
espec i ely,
c1c2c3c4
a e he unknown coe icien s ha ha e o be calcula ed and
x,y
is he posi ion o he poin
in which he displacemen is calcula ed.
The equa ion o he displacemen is de ined. Then, he Rayleigh-Ri z me hod e alua es
he po en ial ene gy o he comple e sys em. The e a e wo si ua ions o calcula e he
sys em’s po en ial ene gy, he non-ac ua ed and he ac ua ed s a e. The objec i e is o
de e mine he po en ial ene gy o he MFC in e ms o he unknown coe icien s
c1,c2,c3,c4
.
Finally, by analyzing he alues o he coe icien s ha minimize he o al po en ial ene gy,
c1,c2,c3,c4
could be ob ained. A e all, he displacemen is e alua ed by plugin hese
coe icien s in o Equa ion 4.1.
The de elopmen s a s calcula ing he po en ial ene gy in he hos lamina. The hos
lamina has he same p ope ies in bo h si ua ions. So he equa ion ha e alua es he
po en ial ene gy is he same in bo h s a es. Howe e , he coe icien s change be ween he
wo s a es. The e o e, he ollowing equa ion calcula es he po en ial ene gy o he hos
lamina:
UHOST =1
2ZLx
2
−Lx
2ZLy
2
−Ly
2ZHHOST
2
−HHOST
2
[σHOST,xεHOST,x+σHOST,yεHOST,y
+σHOST,xyγHOST,xy]dzdydx
(4.2)
whe e
σ
a e he s esses in he di ec ions o he plane,
ε
and
γ
a e he s ains in he di e en
di ec ions o he plane.
The nonlinea s ain-displacemen ela ionship o mino s ains and o a ions in e ms
o he displacemen s se s he s ain in he mid-plane.
ε0
x=∂u0
∂x+1
2∂w0
∂x2
ε0
y=∂ 0
∂y+1
2∂w0
∂y2
γ0
xy =∂u0
∂y+∂ 0
∂x+∂w0
∂x∂w0
∂y
(4.3)

98 Chap e 4. Bio-Inspi ed Mo phing Su aces
The ollowing equa ion de ines he cu a u e in he midplane o he laye o ob ain he
s ain in all he sec ions:
k0
x=−∂2w0
∂2x
k0
y=−∂2w0
∂2y
k0
xy =−2∂2w0
∂x∂y
(4.4)
The coming equa ion de ines he s ains in he olume by applying he Ki cho hypo h-
esis:



εx
εy
γxy


=



ε0
x
ε0
y
γ0
xy




+Z



k0
x
k0
y
k0
xy




(4.5)
c1,c2,c3,c4
e alua es he s ain by subs i u ing he ou - e m displacemen equa ions in
all he o mulas. Then, he educed s i ness ma ix Qlinks he s ain and s ess.
σ=Q·ε(4.6)
whe e
σ
a e he s esses in all di ec ions,
Q
is he educed s i ness ma ix o he sys em
and εa e he s ains.
Rega ding he ma e ials used as hos lamina, he ollowing equa ion de ines Q:
Q=



E1
1−ν12ν21
ν12E1
1−ν12ν21 0
ν12E1
1−ν12ν21
E2
1−ν12ν21 0
0 0 G12




(4.7)
whe e
E
is Young’s modulus o he ma e ial,
ν
is Poisson’s a io,
G
is he Shea modulus.
Finally, he unknown coe icien s
c1,c2,c3,c4
de e mine he po en ial ene gy equa ion
o he hos lamina.
Conce ning he MFC, he e a e wo s a es: he non-ac ua ed s a e wi h no ol age applied
and he ac ua ed s a e when he ol age is applied.
In he non-ac ua ed s a e, he beha io o he MFCs is simila o he hos ma e ial.
Howe e , he e a e di e en conside a ions due o i s laye s uc u e. The ollowing
equa ion exp esses he po en ial ene gy o he MFCs:
4.1 Mo phing Tail 99
UMFC =1
2ZLx
2
−Lx
2ZLy
2
−Ly
2ZHHOST
2+HMFC
HHOST
2
[σMFC,xεMFC,x+σMFC,yεMFC,y
+σMFC,xyγMFC,xy]dzdydx
(4.8)
The ela ionship be ween he cu a u e and he s ains ollows he assump ions made
be o e in he hos lamina. The o mula ion assumes ha he cu a u e o he midplane o
he MFC and he hos is he same. So he s ain in all he su aces o he MFC could be
exp essed as ollows:



εx
εy
γxy


=Z−HMFC +HHOST
2



k0
x
k0
y
k0
xy




(4.9)
Finally, he po en ial ene gy o he MFC could be in he non-ac ua ed s a e. The ollowing
equa ion de ines he sys em’s po en ial ene gy in he non-ac ua ed o m.
UTOTAL =UHOST(c1,c2,c3,c4)+UMFC(c1,c2,c3,c4)(4.10)
A e all, as exposed in he Rayleigh-Ri z me hod, he coe icien s a e se by minimizing
he po en ial ene gy unc ion:
∂UTOTAL
∂ci
=0i=1,2,3,4 (4.11)
A his poin , he o mula ions de e mine he sys em unde non-ac ua ed condi ions. All
he coe icien s a e equal o ze o, highligh ing ha he sys em has no displacemen . The
esul in he coe icien s is an ob ious solu ion. Howe e , i is a good s a ing poin o
con inue analyzing he sys em.
A his poin , he o mula ion e alua es he displacemen s by applying a ol age. The e-
o e, he po en ial ene gy on he hos lamina does no a y be ween he ac ua ed and
non-ac ua ed s a es. Howe e , due o he ol age applied, he MFC changes i s po en ial
ene gy. The ollowing equa ion de ines i du ing he ac ua ion.
UMFC,Ac ua ed =1
2ZLx
2
−Lx
2ZLy
2
−Ly
2ZHHOST
2+HMFC
HHOST
2
[(σMFC,ac ua ed,x)εMFC,x
+(σMFC,ac ua ed,y)εMFC,y+(σMFC,ac ua ed,xy)γMFC,xy)]dxdydz
(4.12)
whe e
σMFC,ac ua ed
a e he s esses in he MFC wi hou he s esses induced by he piezo-
elec ic ma e ial. I could be exp essed as:
100 Chap e 4. Bio-Inspi ed Mo phing Su aces
θMFC,ac ua ed =θMFC −θPiezoelec ic (4.13)
whe e σpiezoelec ic is:
σpiezoelec ic =QMFC △V
△x


d11
d12
0

(4.14)
Finally,
c1,c2,c3,c4
sol e he po en ial ene gy equa ion, and he Rayleigh-Ri z me hod
de e mines hese coe icien s.
∂UTOTAL
∂ci
=0i=1,2,3,4 (4.15)
The equa ion 4.1 de e mines he app oxima e displacemen equa ion by eplacing he
coe icien s. Figu e 4.6 shows he displacemen esul s o he heo e ical model.
Figu e 4.6
Displacemen s heo e ical analysis. The igu e shows he displacemen s in he
Z axis o he MFC mapped in colo on i s geome y. The scale o he colo s
and he alue ha co esponds o each colo is on he igu e’s igh side.
The igu e e alua es he displacemen s o he ac ua o de eloped. The hos lamina
o he model is ano he MFC (MFC-2). The ac ua ion on he g aphic is asymme ical,
meaning ha when one MFC is ac i a ed and he o he is no . The e o e, i MFC-2 is
deac i a ed, MFC-1 pe o ms he ac ua ion. On he o he hand, when MFC-2 is ac i a ed,
4.1 Mo phing Tail 101
he displacemen o he ac ua o will be in he opposi e di ec ion.
The model is an imp essi e s a ing poin o e alua e he echnology de eloped o he
ail. Howe e , he cons uc ion and expe imen a ion o he sys em a e c ucial because he
model is only a heo e ical app oxima ion.
The ollowing sec ion ocuses on de eloping he bio-inspi ed ail using hese ac ua o s.
4.1.2 Tail de elopmen
The aim is o eplica e he beha io o a bi d’s ail using he ac ua o s p esen ed be o e.
The ollowing lis summa izes he main concep s o ail de elopmen , highligh ing he
objec i es.
•
So ac ua ion: In na u e, he bi ds’ ails a e composed o ea he s. Thei ail is
adap able o mos condi ions. In a c ash, hei so s uc u e a oids damage o he
body and he en i onmen . The objec i e is o eplica e his capabili y by de eloping
a so ail using he MFCs as a s uc u e.
•
Tu ning eloci y: Looking a he bi ds’ ligh , hey exhibi an excep ional u ning
eloci y o he ail. MFCs a e one o he as e -known ac ua o s. This eloci y
inc eases he con ol skills o he pla o m. This si ua ion is analyzed using a Mo ion
Cap u e sys em in he expe imen al pa o his de elopmen .
•
Weigh dis ibu ion: The cen al pa o he bi ds’ weigh is a ound he CoG. Rega d-
ing o ni hop e s, a usual p oblem is he weigh o he se os o con ol he ail. The
size o he se os needed o main ain i is signi ican due o he high ae odynamic
o ces gene a ed. On he o he hand, MFCs exhibi high blocking capabili ies wi h
a much lowe weigh . Howe e , he elec onics o con ol MFC a e mo e complex.
This pa is iny compa ed o adi ional se os, and he loca ion is close o he CoG
o he bi d, educing he in luence on he ajec o y.
T adi ionally o ni hop e s’ ails use se os as ac ua o s. Table 4.1 compa es se os and
MFC, ocusing on he concep s and equi emen s p e iously highligh ed.
Table 4.1 Compa ison Se omo o s and MFC Ac ua o s.
Pa ame e s Se omo o MFC
Dimensions (mm) 22.8x12x29.4 101x20x0.5
Weigh (g) 15.8 2
Blocking Fo ce (N) 58.5 152
Powe consump ion (W) 2 ∼0
Ene gy E iciency 85 100
The able shows se os a e hea ie han MFCs. Thei blocking o ce gene a ed is
lowe han MFCs. Se os pe o m only punc ual ac ua ion, and hei isola ion om he
en i onmen is lowe . The powe consump ion o se os is also bigge . The blocking
o ce gene a ed by MFCs imp o es he ac ua ion o he ail, inc easing he con ol on he
ajec o y. The ac ua ion in he su ace o he MFCs make i possible o de elop mo phing
108 Chap e 4. Bio-Inspi ed Mo phing Su aces
Table 4.2
Tails compa ison: The able shows he compa ison be ween wo V- ails and he
p o o ype de eloped in his hesis.
Pa ame e s Comme cial
V-Tail [133] V-Tail [140] V-Tail de eloped
Ac ua ion me hod Se os Se os MFCs
Ac ua o weigh (g) 15,8 9 2
Numbe o
ac ua o s 2 2 10
Ac ua o blocking
o ce (N) 58.5 38.5 152
Ac i e su ace (cm2)462 740 1190
To al weigh (g) 67.1 108 51.8
Ac i e
su ace/weigh
Ra io
6.89 6.85 22.97
Speed Medium Medium High
as he less-weigh ed se o sys em. The blocking o ce is ou imes bigge han his sys em
and h ee imes bigge han he o he . I made i possible o inc ease he su ace o he
ac ua o . In he end, he isola ion p ope ies o he MFCs made i possible o wo k in mo e
complex en i onmen s. Se os show isola ion p oblems in dus y en i onmen s.
Rega ding ou expe ience wi h hese ac ua o s, he dus comes inside he gea s damaging
hem. The impac esis ance o se os is also smalle han MFCs. The ac ha MFCs
does no ha e mechanical connec ion educes he damages gene a ed in case o a c ash.
Focusing on he whole ail, he ac ua o de eloped is less weigh ed han he o he s. This
di e ence gains p ominence by locking a he ac i e su ace. The ac i e su ace is wo
imes bigge han he es o he ails, gene a ing mo e han wo imes ac ua ion due o he
bio-inspi ed ac ua ion. The eloci y o change he ail’s posi ion is mo e signi ican han
he es o he sys ems. This ansi ion eloci y is due o he capabili ies o he MFCs o
shi he o ce gene a ed in e ms o ol age. The esis ance o he impac o he ail is
imp o ing he es ing capabili ies shown in he sys em p esen ed in he s a e o he a .
The expe imen al alida ion o he ail will be p esen ed in Chap e 5 oge he wi h he
o he sys em highligh ing he in e ela ionship be ween hem.
4.2 Mo phing Wing
T adi ionally, he wings o o ni hop e s use plain s i , ae odynamic su aces. These p o iles
a e one o he mos simple in enginee ing. Thei main bene i is he educed weigh o he
wing. The absence o ee olume makes i possible o a oid illing ma e ials. Howe e ,
he capabili ies o hese p o iles could be highe . They a e no able o ha e li gene a ion

4.2 Mo phing Wing 109
hemsel es. Ins ead, he lapping and he su ace o he wing gene a e he li . In o he
wo ds, he o ni hop e has low gliding capabili ies.
Howe e , he gliding phase o he ligh is c ucial. I makes i possible o educe
ene gy consump ion and inc ease aul ole ance. Du ing he gliding phase, he ene gy
consump ion o he o ni hop e is nea ze o, and he ligh is s able. The pla o m’s esponse
o he con ol inpu s signi ican ly educes he ac ua ion needed. The second conce n ocuses
on possible ligh aul s. Inc easing he gliding capabili ies o he o ni hop e makes he
gliding eloci y smalle . As a esul , he o ni hop e can ha e con olled ligh s a lowe
speeds. The o ni hop e can smoo hly land in he case whe e he lapping mechanism
b eaks suddenly, a oiding isky si ua ions. Conce ning he pe ching, using his kind o
p o ile also help he pe ching ask by educing he ene gy o he impac .
To imp o e he con ol o he ajec o y, bi ds ha e ac ua o s in he ail and wing.
O ni hop e s use o con ol i wi h he udde s in he back pa o he pla o m. Howe e ,
due o he educed con ol ac ua ion du ing he lapping is in e es ing o analyze o include
con ol on he wings.
This wo k p oposed looking a na u e mimicking bi ds’ mechanisms o glide and con ol
hei ajec o ies. Bi ds can change be ween gliding and lapping wi hou di icul ies.
Mig a o y bi ds a e used o lap o he app op ia e wind gus o he mal cu en . Then hey
can spend se e al hou s in he low. These bi ds can pe o m long ligh s looking o a
place o li e du ing challenging seasons.
The aim is o de elop a bi d-based wing aking ad an age o he geome ies and con ols
ways lea ned om na u e. Ne e heless, he inal objec i e o he ae ial pla o m is o
imp o e he in e ac ion skills o he ae ial sys em. Figu e 4.12 shows he concep ual design
o he wing.
Figu e 4.12a shows an eagle aking ad an age o he wind gus s o main ain he ligh .
Figu e 4.12b shows he s uc u e o he ea he s o a wing. Figu e 4.12c summa ized he
p ima y mechanism o con ol he wing’s ajec o y, dis ibu ion, and geome y. Figu e
4.12d shows a schema iza ion o he de elopmen p oposed. Finally, Figu e 4.12e shows
he wing de eloped du ing his wo k in eg a ed in o he pla o m c ea ed in he p ojec
GRIFFIN [65].
The objec i e is o eplica e bi ds’ wing ip con ol o imp o e he con ol skills, as
shown in 4.12c. They use hei ea he s and he las join o modi y he su ace o he
wings. This di e en ial ac ua ion gene a es a momen a ound he ROLL o he ae ial
pla o m, adding ano he deg ee o eedom o con ol he ajec o y.
To educe he damage in case o impac . The bi ds’ wings a e lexible and composed
o ea he s and ligh bones. The aim is o educe he s i ness o adi ional o ni hop e
wings, making i possible o main ain ligh beha io .
The bi d-based ae odynamic p o ile mimics he bi ds’ gliding beha io . Fu he mo e,
his p o ile also mimics he mo emen s pe o med by he ea he s du ing he lapping. The
hickness o he p o ile a ies be ween he leading and ailing edges. The leading edge
is much hicke han he ailing edge, making i possible o eplica e he wing ea he s
mo ion p esen ed in Figu e 4.12b.
The ollowing sec ions will ocus on he design and manu ac u ing p ocess o he wing,
and Sec ion 5 e alua es i s beha io .
110 Chap e 4. Bio-Inspi ed Mo phing Su aces
(a) Eagle lying [147]. (b) Wing ea he opology om [148].
(c) Ske ch o he mechanism. (d) Ske ch o he pu posed de elopmen .
(e) De eloped wing in eg a ed in o E-Flap[68].
Figu e 4.12
Wing concep ual design. The igu e shows he concep s lea ned om na u e on
he op (a) and (b). The middle pa o he igu e o e s he ansc ip ion o hese
concep s o a ske ch o he bi d’s wing (c) and a so olding mechanism (d).
In he end, (e) shows he in eg a ion o he so s uc u e wi h he ae odynamic
p o ile in an o ni hop e .
4.2.1 Design
This sec ion ocuses on he de elopmen o he wing. This de elopmen ’s s a ing poin
is analyzing he bi ds’ wings dimensions. The sys em de eloped should o e come he
gliding capabili ies o a adi ional o ni hop e , making he pla o m sa e o in e ac ion
wi h humans. Howe e , he lapping ligh is he main cha ac e is ic o he o ni hop e . So
he wing de eloped mus ha e an equilib ium be ween he lapping and gliding capabili ies
while imp o ing he in e ac ion and maneu e abili y skills.
By looking a he na u e and esea ch wo ks like [
149
], bi ds o p ey suppo mos o
he inpu s o he design. The de elopmen ocuses on he mo phology o a bald eagle.
These bi ds show dex e ous con ol skills ha ing high maneu e abili y o ollow hei p ey.
4.2 Mo phing Wing 111
Figu e 4.13
Wing concep ual design om [
72
]. The igu e summa ized he main cha ac-
e is ics and sys ems de eloped in he wing. The igu e ep esen s he di e en
pa s o he wing mapped wi h a pa e n. The mapping legend is on he bo om
igh side. The op le shows he s uc u e o he olding mechanism, and he
bo om le shows he ae odynamic p o ile o he wing.
Ano he o i s main cha ac e is ics is i s high payload. They can load a weigh simila o
hei mass.
A la ge wingspan and a mid- ange aspec a io cha ac e ize hei wing geome y. A
high li gene a ion a low lapping equencies depic s hese wings. Compa ed wi h o he
species, he bigge cho d size makes i possible o imp o e he gliding capabili ies. Figu e
4.13 shows he dimensions o he p oposed wing.
Figu e 4.13 syn hesizes de elopmen . I includes summa izing all he sys ems de eloped
in he wing and i s main cha ac e is ics.
The wing geome y eplica es bald eagles. The wing span is b=2 m, and he cho d c=0.33
m. Then an ac i e ip was designed o mimic he mo emen ’s p ima y and seconda y
ea he s, ha ing a maximum su ace on he un olded con igu a ion o 0.498 m
2
and a
minimum o 0.420 m
2
. The con ol o he wing’s a ea gene a es ac ua ion on he ROLL
angle o he ae ial pla o m.
These cha ac e is ics de ine he plana s uc u e o he wing. I mimics he ea u e o he
bald eagle’s wings, mainly ocused a his poin on he Aspec Ra io (AR). Theo e ically,
his s uc u e main ains hese bi ds’ lapping/gliding capabili ies.
The nex s ep o con inue wi h he design o he wing is he selec ion o he ae odynamic
p o ile. The objec i e is o use an ae odynamic p o ile ha emula es he wing’s ea he s
du ing he lapping. The e o e, AS-se ies om [
150
] was e alua ed. In his pape , he
au ho s e alua e he design o bi d-inspi ed p o iles by using he mul ipoin in e se ai oil
design. Pa icula ly in his wo k, he p o ile AS-6091 can main ain he lapping while
keeping he gliding capabili ies. Figu e 4.13.II shows his p o ile. This p o ile has a
bulbous leading edge mimicking he leading edge o he bi d’s wings wi h a hing ip
emula ing hei ea he s. The sec ion o he p o ile emula es he sec ion o an eagle wing.
One o he main no el ies o he wo k is he use o lexible ab ic. The objec i e is o
main ain he ae odynamic su ace du ing he ip mo emen as bi ds do wi h hei ea he s.
112 Chap e 4. Bio-Inspi ed Mo phing Su aces
Table 4.3 Elas ome ma e ials es ed.
P ope ies
[151]
Mixed
Viscosi y
(Pa ·s)
Speci ic
Volume
(m3/Kg)
Po Li e
(min)
Tensile
S eng h
(KPa)
Elonga ion
a b eak
(%)
D agon
Skin™10
Slow
23 9.32·10−445 3275 1000
D agon
Skin™30
Slow
20 9.28·10−445 3447 364
Eco lex™
00-30 3 9.4·10−445 1378 900
Eco lex™
00-35 Fas 3.5 9.4·10−42.5 1378 900
Eco lex™
00-50 8 9.32·10−418 2172 980
A wide ange o a iables akes ele ance in he de elopmen o he ab ic. The ab ic has
o be as hick as possible while main aining he ae odynamic su ace du ing he lapping
and gliding. Di e en elas ome p oduc s om he company Smoo h-On [
151
] we e
e alua ed, ocusing on Eco lex
™
Se ies and D agon Skin Se ies
™
. Table 4.3 shows he
main p ope ies o he ma e ial es ed o his de elopmen .
I is needed a ab ic capable o suppo ing high e o s and high de o ma ions. Con-
ce ning his concep , he alues o he D agon Skin
™
Se ies conc e ely D agon Skin
™
10 Slow show he bes alues. The p ope ies ha se his ma e ial as he bes om he
mechanical poin o iew a e Tensile S eng h and Elonga ion a B eak. This ma e ial
is in he middle o he way be ween D agon Skin
™
Se ies and Eco lex Se ies.D agon
Skin
™
10 Slow ha e alues simila o Eco lex
™
Se ies in elonga ion main aining he alues
o esis ance o de o ma ion o D agon Skin
™
Se ies. Howe e , all he ma e ials on he
able suppo he o ces gene a ed du ing he lapping and gliding phases; he need o ha e
hick low-weigh ed ab ic di icul he manu ac u ing p ocess. The ab ic is handmade due
o i s no el y de elopmen . The ab ica ion p ocess de e mines he ma e ial used. The
selec ed ma e ial will be exposed in he nex sec ion when he manu ac u ing p ocess is
de e mined.
Nylon ips op ab ic composed he su ace o he non-ac i e pa o he wing. Usually,
his kind o pla o m uses his ab ic. I s main cha ac e is ics a e i s low weigh and
esis ance o ea ing and ipping.
Designing he ame o he wings, he concep s lea ned om he li e a u e, and ou
knowledge om o he wing de elopmen s a e c ucial. The wings a e long (2 me e s
o wing span) and mus deli e high o ces due o hei inc eased payload capabili ies.
The e o e, we de eloped a e iew o he ma e ial p ope o he wing s uc u e using
4.2 Mo phing Wing 113
(a) Fou ibs analysis.
(b) Fi e ibs analysis.
Figu e 4.14
Ribs displacemen /S ess analysis unde maximum load condi ions: (a) Shows
he displacemen s in [mm] mapped in colo s on a wing wi h ou ibs. (b)
Shows he displacemen s gene a ed in a wing wi h i e ibs. Bo h igu es
ep esen he colo scale on he le side.
Ansys
™
G an a Edupack. Ligh weigh composi es ha we e s i enough ocused he
e iew. Howe e , he e we e no su p ises. The bes ma e ial o he wing ame was CF in
di e en composi ions. The main eason o selec CF is he easy way o ind his ma e ial
on he ma ke in a ious con igu a ions.
The wing has h ee main pa s o be s udied sepa a ely: he ae odynamic p o iles, he
olding mechanism, and he ansmission pa o he wing.
The s udy begins wi h he ae odynamic p o iles. The selec ed ae odynamic shapes can
mimic he mo emen s o he ea he s du ing he lapping. Howe e , he analysis mus i s
se he numbe and hickness o he p o iles o main ain hese capabili ies in he whole
wing.
The designing p ocess se s he li gene a ion o 1.4 kg. The wings abso b his o ce
du ing he lapping; each hal -wing has o abso b 0.7 kg. The e o e, each ib has o
d ink a p opo ional pa o he o ce. As a esul , he ib in eg a ed in o he olding

114 Chap e 4. Bio-Inspi ed Mo phing Su aces
Figu e 4.15
Folding mechanism s uc u al s udy. The igu e analyses he displacemen in
mm pe o med in he ip wi h a o ce o 2.5 N. The igu e shows he colo
scale o he displacemen s on he le side .
mechanism does no ha e an ae odynamic p o ile. In he end, he o al numbe o ibs is
one mo e han hose wi h an ae odynamic shape, as shown in Figu e 4.13. The analysis
e alua ed each ib’s displacemen s sepa a ely o analyze he beha io o he ibs ha se he
wing’s ae odynamic p o ile. The goal is o main ain he desi ed de o ma ions o mimic he
ea he s. To e alua e he de o ma ion In en o
®
Nas an
®
was used. Figu e 4.14 e alua es
he displacemen beha io o he p o iles unde ully loaded.
The u he analysis p esen ed in his wo k only shows he displacemen esul s. The
aim is o e alua e i he wing de eloped can pe o m he desi ed bioinspi ed mo emen s.
Howe e , he s ess analysis e alua es i he mechanism can suppo he e o s gene a ed.
The s ess analysis se s he limi s and he ma e ial used in he wing.
The e alua ion o he e o s se he hickness o he p o ile o 1.5 mm. The ma e ial
selec ed was CF pla es wi h ply o ien a ion (wo en 0/90
◦
). The con igu a ion wi h h ee
ibs was discha ged o being close o he b eak limi o he CF, gene a ing a igue in
he ma e ial. Fi e ibs se up (shown in Figu e4.14b) was dismissed o maximizing he
de o ma ion o he wing, minimizing he mass. The de o ma ion achie ed wi h ou ibs
was close o he expec ed one, ha ing 2.12 mm o de o ma ion on he ip as shown in
Figu e 4.14a. The s uc u e can suppo he s esses gene a ed a 2 Hz o lapping equency.
The secu i y ma gin in he s ains using his con igu a ion is 30 %.
The ansmission pa o he wing is closes o he lapping mechanism and mus be
s i e o ansmi he lapping equency. The wing’s olding ip mus be lexible enough
o pe o m he olding mo emen s. A CF ube is he ma e ial o he ansmi e , and he
olding mechanism uses a CF od. The od is glued inside he ansmi e o educe he
weigh o he join . Figu e 4.15 shows he s udy pe o med in In en o
®
Nas an
®
o selec
he size o he ubes and ba s.
The ba selec ed o he ip has 4 mm in diame e wi h unidi ec ional ibe s o acili a e
4.2 Mo phing Wing 115
he olding. The longi ude o he ba is 300 mm o se he olding a ea as designed be o e.
The igu e shows he displacemen pe o med wi h 2.5 N o pull. A se o placed in he
body o he o ni hop e gene a es he o ce, and endons ansmi i inside he ansmi e
ube ill he ip o he olding mechanism, as shown in Figu e 4.13.
The geome y selec ed o he ansmi e is a 6 mm in diame e ube. The longi ude
o he ba is 700 mm o gene a e he equi ed wing span. The od has 1 mm o hickness
and wo laye s o comp essed bidi ec ional CF ab ic (ply wo en 0/90
◦
). This pa o he
wing implies high-e o ansmission. In addi ion, he o ien a ion o he ab ic’s ibe s
gua an ees he ba ’s s i ness in all di ec ions mimicking he wing bones o he eagles.
The displacemen eached in he olding mechanism was 175 mm. The s udy p esen ed
in Figu e 4.15 also shows ha he ansmi e is s i enough o main ain a signi ican pa
o he de o ma ions in he ip. Fu he mo e, he s ess analysis gua an ees he sys em can
suppo he o ces gene a ed.
The nex s age gene a es he wing model wi h all he componen s. Figu e 4.16 shows
i s displacemen and beha io du ing lapping.
Fi s , he selec ed p o iles mus suppo he s esses gene a ed wi hou plas ic de o ma-
ion. The epo ed geome ies achie e his equi emen . Then he displacemen s we e
analyzed. The goal o his wing design is o mimic he de o ma ion pe o med by an eagle
wing. Du ing he ups oke, he de o ma ions o he p o iles (shown in Figu e 4.16b) make
he a ea mo e mino , educing he esis ance o pe o m he mo ion. The ups oke gene a es
o ce opposi e o he li . This a ea educ ion makes i possible o educe hese o ces
making he ligh mo e s able. Conce ning he downs oke, he de o ma ion o he p o iles
maximizes he wing’s a ea o inc ease li gene a ion. This beha io is possible due o he
p o iles selec ed. I pe ec ly eplica es he mo ion o a alcon’s wings.
Theo e ically, as shown in he s udies, he wing design can mimic he de o ma ions o
bi ds’ wings. This condi ion inc eases he beha io o a egula one. Howe e , i is only
possible i he manu ac u ing p ocess main ains he design pa ame e s. The e o e, he
ollowing sec ion ocuses on he cons uc ion p ocess.
4.2.2 Manu ac u ing P ocess
Figu e 4.17 summa izes he cons uc ion p ocess. Fo a clea explana ion, he ollowing
ex di ided he desc ip ion in o he pa s shown in he lis :
•
Ae odynamic ibs manu ac u ing: This pa is ela ed o he Figu e 4.17.1.a. I
ocuses on ob aining he ibs used o gene a e he wing p o ile.
•
Mechanical pa s in eg a ion: Shown in Figu e 4.17.1.b. I is ela ed o in eg a ing
all he pa s o he wing’s ame. In his pa , he illing ma e ial also is de eloped
and in eg a ed.
•Ris op ab ic uni ica ion: This pa o he manu ac u ing p ocess shows comp ises
he in eg a ion o he ab ic on he wing s uc u e. Figu e 4.17.1.c. shows his
p ocess.
•
Elas ome ab ic de elopmen : I is ocused on gene a ing he ab ic. Figu e 4.17.2.a
shows ha his pa o manu ac u ing is one o he mos impo an .
116 Chap e 4. Bio-Inspi ed Mo phing Su aces
(a) Downs oke.
(b) Ups oke.
Figu e 4.16
Wing s uc u al s udy. The igu es show he de o ma ion pe o med on he
wing in he wo s load cases possible. (a) Du ing downs oke. (b) Du ing
ups oke. The colo scale o he displacemen s is ep esen ed on he le side
o he igu es.
•
Elas ome ab ic in eg a ion: This pa o he p ocess ocuses on de eloping he
olding mechanism wi h he elas ome ab ic, Figu e 4.17.2.b.
4.2 Mo phing Wing 117
Figu e 4.17 Wing manu ac u ing p ocess. This igu e di ides he manu ac u ing p ocess
in o wo. Yellow dashed lines su ound he ab ica ion o he wing’s main
s uc u e, and g een dashed lines ounded he wing ip’s de elopmen . The
a ows highligh he o de o he p ocesses o de elop he wing.
The explana ion o he manu ac u ing p ocess ollows he o de se in he lis p esen ed
be o e. The ibs wi h he desi ed ae odynamic p o ile a e he i s on he lis .
A CNC machine cu he desi ed p o ile om a CF pla e. The ske ch c ea ed in In en o
®
has he geome ical da a o [
150
]. The so wa e Vca e ansla ed he ske ch o G-Code.
Finally, he pos -p ocesso WinPC-NC commands he wo k o de s o he machine. The
machine used was HIGH-Z S-1400/T CNC ROUTER om he company CNC-STEP. The
milling cu e used had 2mm in diame e , speci ic o CF milling wi h a diamond s uc u e.
The nex s ep is o in eg a e he de eloped p o iles in o hei assigned posi ion. As shown
in he p e ious s udy, hese p o iles mus suppo high e o s du ing he lapping. Di e en
me hods e alua e how o ha e a solid join be ween he p o iles and he ansmi e ube.
The i s idea was o de elop a mechanical union, discha ged due o he complexi y added.
The cylind ical shape o he od makes i di icul o place he join exac ly in he same
place. In he end, he es s conduc ed cause he loss o ma e ial. The solu ion was o
glue he p o iles o he od. Ne e heless, his union has o suppo all he p essu e o
he wing. S a ing om he adi ional epoxies and e alua ing as and slow epoxies, he
e alua ion pe o med e alua es i s capabili ies o he applica ion. In he end, Me alse
™
A4 is used. This epoxy has me al pa icles inside, making a mechanical join . Howe e ,
he main capabili y added o ou sys em by his epoxy is he possibili y o inc ease he
con ac a ea. The con ac a ea can dis ibu e he p essu e on a bigge su ace by making i
bigge , educing he punc ual e o s. I eplica es pe ec ly he cons ain s se in he s udy;
howe e , i inc eases he mass o he wing.
A cus om-made s encil gua an ees he place o he p o iles in he desi ed posi ion and
124 Chap e 5. Expe imen al Valida ion
(a) Tendon Con igu a ion 1 om Figu e 3.35a.
(b) Tendon Con igu a ion 2 om Figu e 3.35b.
Figu e 5.3
Tendon’s con igu a ions expe imen al esul s. (a) Shows he esul s o Con-
igu a ion 1 and (b) o Con igu a ion 2. Each le el o e s ou g aphs, each
applying a di e en ension o he main endon, shown a he op o he g aphs.
The g aphs classi ied he objec s in e ms o hei weigh and eliabili y le el.
The dashed lines show he bo de s be ween he laye s shown in he lis abo e.
The g aph shows he claw’s adap a ion esul s. As expec ed, adap a ion capabili y
inc eases wi h he applied ension. Howe e , he cha s also show less weigh ed objec s
wi h challenging g aspings. These objec s used o ha e polished su aces and a ying
shapes. Howe e , he claw p o es o gua an ee adap a ion o hese objec s. I showed a
igh g asp wi h all he objec s measu ed.
Conce ning he analysis o he con igu a ions, he beha io o bo h is close. Howe e ,
Con igu a ion 2, in which all he inge s ha e he same ension in he endons, shows be e
esul s o g asping objec s. The g aphs p o e his con igu a ion equi es less o ce o hold
he same hings. The cha wi h 8 kg o ension applied shows ha he g ippe can i mly
and o ce ully g asp all es ed objec s. The es o he g aphs also ag ee wi h his end,

5.1 Bio-Inspi ed Claws Expe imen s 125
showing be e esul s o Con igu a ion 2 han Con igu a ion 1.
I is no a good analysis o measu e he di icul ies o g asping, conside ing only i s
weigh . Also, he holding olume is signi ican and he geome y o he objec s held.
The ollowing igu es show images o expe imen s wi h di e en hings ha highligh he
capabili ies o he claw o adap o mul iple geome ies. The igu es ha e been classi ied
by he ension applied o he main endon. Figu es 5.4a,b,c,d show he esul s o 2 kg o
o ce applied, Figu e 5.4e, ,g,h o 4 kg, and Figu e 5.4i,j,k,l o 8 kg.
The igu es show he adap a ion skills o he claw. Analyzing he expe imen pe o med
using 2 kg o ension, which implies applying 0,5 kg in he endon o each inge , i s a s
o show angible esul s. Conce ning adap a ion, he image shown in Figu e 5.4d demon-
s a ed he skills o he claw o hold i mly sphe ical shapes. This shape is challenging
o keep and usually implies ha ing independen con ol o each inge . Tha means he
endons con igu a ion co ec ly balances he sys em o each a igh g asp. Figu e 5.4a
shows he adap a ion o sho cylind ical olumes. All he inge s a e in con ac wi h he
objec keeping he objec held wi h his geome y. Figu e 5.4b shows he adap abili y o
he claw using low ension on small holding olumes. In he end, o he con igu a ion
analysis using 2 kg o p essu e, Figu e 5.4b shows he adap a ion o an aluminum ba . The
edges o his ba p o ile make i possible o he nail o he claw o go inside, simpli ying
he g asping.
The analysis con inues wi h he ollowing le el o he igu es, he expe imen s pe o med
wi h 4 kg o o ce applied o he endon. The g aphs con inue showing he adap a ion o
he sys em o daily objec s. Figu e 5.4e demons a es adap abili y o plie s. This g asping
is in e es ing o analyze o coope a ion wo k wi h wo ke s deli e ing ools in di icul
access a eas. Figu e 5.4 demons a ed he abili y o he claw o hold a ying shaped
objec s wi h high masses. The bo le in he igu e has an i egula conical-shaped polished
su ace. The claw s ill has some di icul ies main aining his objec wi h 4 kg. Howe e ,
he g asping is sui able o ligh bu imp o es wi h mo e ension, as is shown la e . Figu e
5.4g is in e es ing. The objec is hin, bu he claw can main ain he holding au onomously,
adap ing o i s su ace. Ul ima ely o his ension, Figu e 5.4h demons a ed he adap a ion
o he claw o cylind ical polished high-weigh ed geome ies. The claw can main ain he
g asping i mly, gene a ing g ip enough wi h he con ac su aces.
The inal pa o he analysis comp ises applying 8 kg o ension o he endon. This
o ce is pe ec ly eachable by he muscles designed. The objec i e o analyzing his o ce
is o es he adap abili y and e alua e he possible damages he claw can exe . Figu e
5.4i shows he capabili ies o he claw wi h a c ys al bo le. The bo le has a challenging
su ace wi h many polished edges bu can s ill hold he objec au onomously. I was able
o p ese e he bo le wi hou sc a ching o damage. Figu e 5.4j analyzes hea y-weigh ed
conical shapes. The claw has demons a ed ha i could gene a e enough g ip o main ain
his kind o objec held om he middle, aking ad an age o edges o inc ease he g ip
blocking he mo ion o he objec . The g asping is a balance be ween g ip and adap a ion.
Ul ima ely Figu es 5.4k and 5.4l analyze he coope a ion wi h humans. These es s
examine he damage he claw can exe using hea y load on he endon. The adap abili y
shown in Figu e 5.4k is ema kable. Howe e , as he holding olume is small, he o ce
applied is also smalle . The es conduc ed in Figu e 5.4l wi h a human analyzes he
possibili y o damaging a human. This holding olume is majo , so he o ces applied a e
126 Chap e 5. Expe imen al Valida ion
(a) Tape 95.4 g. (b) B ush 75.6 g. (c) Squa e ba 200.5 g. (d) Tennis ball 58 g.
(e) Plie s 298.9 g. ( ) Glue bo le 281.3 g. (g)
Aluminum T-shaped
pla e 57.2 g.
(h)
Sp ay pain bo le
147.3 g.
(i)
C ys al Liquo bo le
450 g.
(j) Glue bo le 281.3 g. (k) Human Finge . (l) Human a m.
Figu e 5.4
Con igu a ion 2 analysis a ying he holding olume. (a)(b)(c)(d) Shows he
esul s using 2 kg o ension in he endon, (e)( )(g)(h) uses 4 kg, and (i)(j)(k)(l)
8 kg. Each igu e labels he weigh o he objec held a he bo om, and he
image shows i s geome y.
also signi ican . This es concludes wi h li le ma ks on he a m ha disappea 1 min a e
he ial, no e en a sc a ch o hema oma.
The se up expe imen s show excellen esul s demons a ing ha he geome y and
in e nal o ce dis ibu ions pe o m acco ding o he design speci ica ions. The ollowing
5.1 Bio-Inspi ed Claws Expe imen s 127
es s in eg a e he claw wi h he ac ua ion o he SMA muscles o u he analyze he
g ipping capabili ies.
The expe imen s o analyze he claw’s beha io wi h he de eloped a i icial muscles
e alua e he sys em holding di e en objec s. The objec i e is o eplica e he sys em
pe o mance shown in he p e ious expe imen s. The sys em se up o his es includes a
The mal/RGB came a ocused on he ac ua ion o e alua e he holding du ing he ac ua ion.
This came a was he WIRIS P o om he company Wo kswell
®
[
152
], specially designed
o he mal inspec ion in ae ial obo s. I is a adiome ic inspec ion he mal came a p ope
o d ones due o i s ligh weigh . The esolu ion o he he mog am is 1266×1010 px
eaching up o 1200
◦
C, and he esolu ion o he RGB came a is Full HD (1920 x 1080
px). The accu acy o he he mal measu emen is
±2%
. The cha ac e is ic o he came a
is enough o p o ide a isualiza ion o he sys em’s beha io . The inal ac ua ion con ol
uses P 100 senso s o educe he weigh . Figu e5.5 shows he isualiza ion o he es o
ex ac he conclusions.
The igu e ocus on wo expe imen s. The i s one, conce ning Figu es 5.5a and 5.5b
show he sequence o holding a sphe ical shape. The i s shows he RGB iew, and he
second is he he mal iew. This geome y is one o he mos challenging o main ain he
hold. O he g ippe s can no e ain i because hey need o gene a e mo e ic ion wi h
he su ace, no o slide. Howe e , ou claw can accommoda e i s geome y o he objec
gene a ing enough g ipping o ce. This beha io is due o i s ma e ial composi ion and
he endons’ ensions dis ibu ion. E e y hing is con ibu ing o his g asping. Fi s , he
nails o PLA a e s i enough o gene a e ic ion wi h he su ace and li le edges. Second,
he palms o Eco lex adap o he phalanxes de eloping mo e g ip. Thi d, he Ca bon
Fibe s uc u e main ains he longi udinal deli e y o o ces. Finally, mos impo an ,
he s uc u e o The moplas ic Polyu e hane (TPU) makes i possible o adap he shape
gene a ing he au onomous holding mo ion.
The sequences expose he e olu ion o he sys em du ing he g asping p ocess. They
show ha he claw can hold he sphe ical shape objec au onomously in less han wo
seconds. The i s igu e is he sys em’s s a e be o e he ac ua ion. The igu e in he cen e
shows when we lea e he objec in hand. The las one on he le o each sequence shows
momen s be o e he holding. Figu e 5.5b shows ha he empe a u e in he muscle is less
han 50
◦
C o main ain he objec . The maximum ac ua ion empe a u e desi ed in he
ac ua o is 120 ◦C. So i can s ill load bigge hings.
The es shown in Figu e 5.5c and Figu e 5.5d e alua e he sys em’s limi s. I consis s
o using a educed holding olume wi h a huge load. These expe imen s a e challenging
because o he so capabili ies o he claw. Usually, i is impossible o use his kind o
obo ics sys em o hea y objec s. This beha io is because hey used o be de o med by
he objec , making i impossible o main ain he g ip. Howe e , ou sys em’s s uc u e can
main ain so capabili ies while exe ing conside able o ce compa ed wi h i s weigh . The
igu e shows his beha io wi h a b ie case handle o 2 kg. The inal mean empe a u e o
he SMA o main ain he objec hold is a ound 70
◦
C. The e o e, he ac ua o is s ill on
elas ic de o ma ion, being possible o use i o hea ie objec s. Fu he mo e, he claw
s uc u e keeps he objec weigh ing only 40 g, making a signi ican ela ionship be ween
he o ce exe ed and he sys em’s weigh .
128 Chap e 5. Expe imen al Valida ion
(a) RGB iew holding a sphe ical shape o 55.4 g.
(b) The mal iew holding a sphe ical shape o 55.4 g.
(c) RGB iew holding he came a case o 2 kg.
(d) The mal iew holding he came a case o 2 kg.
Figu e 5.5
The mal iew o he ac ua ion. (a) and (b) a e he RGB and he mal came as o
he sequence holding a sphe ical shape o 95.4 g. (c) and (d) a e, espec i ely,
he RGB and he mal came a o he g asping o he sui case o 2 kg. The
scale o he he mal came a comes om 10
◦
C o 100
◦
C om cold o wa m
colo s. On he bo om side o he he mal igu es, he label shows he a e age
empe a u e o he sp ing(
TmeanS
) in (
◦
C) and he ime o he sequence ( ) in
(s) .
Ul ima ely, he load was limi ed o 2 kg o his expe imen . Howe e , he beha io
shown makes i possible o inc ease he weigh es ed, al hough he capabili ies o he
o ni hop e s would need o imp o e o be es ed ealis ically.
Figu e 5.5 shows only wo expe imen s in de ail. Howe e , hese es s wi h he ac ua ion
included a wide ange o objec s. Figu e 5.6 summa izes hese expe imen s.
The numbe o objec s held du ing he expe imen s was la ge. The igu e shows hei
geome y and weigh . I exposes he quan i y and di e ences in hei shapes, which
5.1 Bio-Inspi ed Claws Expe imen s 129
Figu e 5.6
Holding Expe imen s Summa y. This igu e shows a summa y o he expe -
imen s ha analyze he adap abili y o he claw o a wide ange o objec s
changing i s size and shape.
di e in he holding olumes. The ollowing ex highligh s he beha io o he claw in
pa icula cases. The beha io shown in he expe imen s wi h he 2 kg case highligh s
he imp o emen o he a io (Objec weigh
/
Handling sys em weigh ), bu also ha he
sys em can hold a iable shapes wi h a high weigh , such as he yellow glue bo le a he
bo om o he second one s a ing om he le . The case o he educed sc ewd i e ’s
olume on he op igh side is also impo an . Also, he cylind ical shapes on he op o
he i e igu es s a om he le wi h di e en masses and diame e s, i also in e es ing.
The igu e on he middle bo om side shows he claw holding a whi eboa d e ase in which
he g ippe looks au onomously o he edges, main aining he g ip. In he end, he claw
uses i s so cha ac e is ics o be adap able o mul iple shapes, he o ce gene a ion o he
muscles o apply enough o ce o he endons, and he longi udinal s i ness o main ain
he objec hold. This beha io is a good ep esen a ion o he bi ds’ claws.
In he end, i is di icul o measu e his beha io objec i ely. Figu e 5.7 do i by
measu ing he objec s’ weigh and es ima ing he g asping olume.
The igu e classi ied he g asping olume in o i e g oups. The explana ion ollows he
o de in he legend o he cha . The i s one is Cylind ical olumes. This olume e e s
o he case o he pain bo le shown in Figu e 5.4h. The nex is he Ci cula olumes. I
e e s, o example, o he ape shown in Figu e 5.4a. The hi d is Sphe ical olumes. This
one comp ises he cases o he di e en balls es ed. The ou h is Rec angula olumes.
In his case, he g asping olume is a pa allelog am. I is shown in Figu e 5.4a. The i h
is Va ying olumes. I e e s o g asping olumes in which he olume a ies inside he
claw, o example, Figu e 5.4j. (No e: The use o hese igu es is only o cla i ica ion; he
expe imen s shown on his cha also implied he SMA ac ua ion).
The expe imen al esul s be o e his summa y conclude ha he claw can manage a ious
shapes. Howe e , Figu e 5.7 objec i ely summa izes hese expe imen s. Fi s , i classi ies

130 Chap e 5. Expe imen al Valida ion
Figu e 5.7
Holding analysis. The igu e summa izes he objec s expe imen ed wi h in
e ms o hei holding shape, weigh , and g asping olume. The legend shows
he i e g asping olumes classi ica ion e alua ed, ep esen ed wi h di e en
ma ks. The cha uses hese ma ks o de elop he ep esen a ion.
he objec by i s g asping olume in
m3
, weigh in g, and he classi ica ion o he g asping
olume showed be o e. These pa ame e s a e he pa ame e ha ou esea ch iden i ies
as he mos impo an ones ega ding manipula ion. Then he cha shows he ange o
shapes and weigh s he claw can hold classi ied in o hose pa ame e s.
Conce ning g asping olumes, he manipula o can hold objec s a 500% bigge han
he smalle ones. The so adap abili y o he claw p o ides his beha io , bu also, he
longi udinal s i ness o he oes akes ele ance, ocalizing he p essu e o he ac ua ion
in he objec . These esul s a e no el o so obo ics manipula o s and di icul o see in
o he ac ua o s wi h he same cha ac e is ics.
Conce ning he weigh o he objec s, an ac ua o wi h less han 40 g can manage up o 2
kg. These esul s also highligh he esis ance o he ac ua o o de o ma ion. In some way,
he claws use hei palms wi h he objec s a oiding damage. The longi udinal s i ness
and he o ce ac ua ion o he de eloped a i icial muscles make his beha io possible.
This beha io is also a no el y o so ac ua o s. Usually, hey only can manage small
loads o main ain hei cha ac e is ics.
Ul ima ely, one o he mos challenging si ua ions is g asping a ying olumes. I
implies au onomously adap ing he ension in each inge ’s endons o gene a e enough
g ip. Rega ding hese esul s, s udying he endon dis ibu ion o o ces was c ucial.
The beha io demons a ed by he obo ic claw goes h ough he s a e o he a o
so obo ics manipula o s. The g ippe is capable o mimicking bi ds’ claws, imp o ing
hei adap abili y. Some imes, he bi ds’ ee a e challenging o coope a e wi h, and he
ope a o s use p o ec i e glo es o deal wi h hem. This scena io is he case o he alcon y.
5.1 Bio-Inspi ed Claws Expe imen s 131
The de elopmen o he claw akes inspi a ion om alcons and eagles. Howe e , I
imp o es he adap abili y o di e en shapes wi h a so e s uc u e main aining he na u al
cha ac e is ic.
The ollowing expe imen s analyze he claw’s beha io o objec deli e y. The objec i e
is o es he claw concep and he an agonis muscles’ con igu a ion con olling he elease
o objec s du ing he ligh .
(a) Deli e y o medical esou ces.
(b) Deli e y o a secu i y es .
(c) Deli e y o a Mic o Ae ial Vehicles (MAV).
Figu e 5.8
Deli e ing objec s analysis. The igu e shows he o ni hop e anspo ing
objec s o speci ic loca ions. The igu e’s blue a ows and ed ci cles poin ou
he hings anspo ed.
These es s we e possible wi hin he i s e sion o he claw. In he i s e sion, he
numbe o objec s was mo e limi ed. Howe e , i demons a es ha he sys em can deli e
hings in a con olled way. The las e sion o he claw can manage mo e objec s mo e
nimbly.
Figu e 5.8a shows he o ni hop e anspo ing a bandage. The es s show ha he
o ni hop e can anspo small medical esou ces o a pe son in need. In addi ion, his
pla o m inc eases he sa e y p o ided by o he ae ials pla o ms a oiding p opelle s and
educing he impac ene gy. These cha ac e is ics make o ni hop e s one o he sa e
132 Chap e 5. Expe imen al Valida ion
pla o ms o human in e ac ion.
Figu e 5.8b demons a es he abili y o suppo signi ican loads. In his igu e, he
o ni hop e is anspo ing a li e es . In his case, he objec ’s ae odynamic in luence is
mo e conside able, causing he pla o m’s con ol di icul ies. This si ua ion imp o ed by
adding he so ae odynamic con ol su aces s udied, making i possible o inc ease he
con ol capabili ies in hese si ua ions. Conce ning he g asping and elease, he claws
showed good beha io deli e ing he objec o he desi ed poin .
Figu e 5.8c shows a Mic o Ae ial Vehicles (MAV) elease o inspec speci ic places.
The objec i e is ha he educed consump ion o he o ni hop e makes i possible o each
he a hes loca ions. Then, he MAV is eleased o pe o m he inspec ion. A e inishing
he e iew, i e u ns o he claw, and bo h e u n o he base.
This expe imen al alida ion ealizes he capabili ies o he claw o manipula e and
possible applica ions o o ni hop e s. The ollowing sec ion ocus on he analysis o he
pe ching skills o he sys em.
5.1.2 Pe ching Expe imen a ion
This sec ion ocuses on he e alua ion o he sys em’s pe ching beha io . The pe ching
in ol es he claw expe imen ed in he p e ious sec ion and he legs. The design p ocess
se s he legs’ unc ionali ies. They ha e o suppo he impac s con ibu ing o he pe ching
and con ol he posi ion o he ac ua o . The sec ion s a s by de ining he expe imen a ion
se up. One o he mos c i ical pa ame e s in o ni hop e s is he weigh o he sys ems.
The e o e, Figu e 5.9 shows he weigh o he di e en pa s o i .
Figu e 5.9
Expe imen al se up weigh . The igu e shows he weigh o he di e en pa s
o he manipula ion sys em. The ep esen a ion ounds wi h a ci cle, label he
componen , and an a ow connec s he elemen s wi h hei co esponding pa
o he g aphic.
5.1 Bio-Inspi ed Claws Expe imen s 133
The igu e di ides he manipula ion sys em in o h ee main pa s. The i s one is
elec onics. This pa in ol es he muscles, con ol elec onic and ac ua ion elec onics
p esen ed in he design p ocess. I s weigh is 50 g, 27% o he sys em’s o al weigh . The
second pa is he legs. The legs a e composed o hei s uc u e and he damping sys em.
I s weigh is 44 g, 24% o he o al weigh . The las pa , and he mos impo an , is he
claws. I cons i u es 49% o he o al weigh o he sys em. Thei con ibu ion is essen ial
o keep he pe ching balanced. Finally, he o al weigh o he manipula ion sys em is
a ound 15%. When compa ed wi h bi ds, he pe cen age o he mass o his sys em is
smalle han he examples ound in na u e. In na u e, his pe cen age ounds he 30%
eaching he 40% bi d’s weigh .
The i s analysis de eloped du ing he expe imen a ion ocuses on he claw’s posi ion
con ol. This sys em manages he posi ioning o he claw du ing he ligh and keeps
he pla o m’s balance du ing he pe ching. Figu e 5.10 p esen s he se up o hese
expe imen s.
Figu e 5.10
Leg posi ioning con ol expe imen al se up. The igu e di ides he con igu-
a ion o hese expe imen s in o wo g oups: he onboa d and he o boa d
se up. The onboa d se up is on he igh side, including all he componen s
inside he o ni hop e du ing he expe imen a ion. The o boa d se up is on
he igh side and shows he ep esen a ion se up and da a used.
The igu e highligh s wo kinds o se ups o his expe imen . The i s one on he le
side is he o boa d se up. The igu e shows all he signals he p ocesso boa d ecei es
o he ep esen a ion. In his case, he boa d was A duino. This boa d analyzes he
ac ua ion da a gene a ed by he con olle and he di e en signals o he senso s onboa d.
A duino p ocesses he da a and sends i h ough he se ial po o a compu e o online
ep esen a ion. The compu e p ocesses he da a in Ma lab
®
, making i possible o analyze
he beha io online. The online expe imen al con ol is c ucial du ing he i s ials because
he sys em has high ins an ene gy consump ion, and he ba e y can su e damages wi h
inadequa e ac ua ion.