CRANFIELD UNIVERSITY
ORIOL LÁZARO CASTELL
THE IMPACT OF ADDITIVE MANUFACTURING ON MICRO-
TURBINE PARTS
SCHOOL OF ENERGY AND POWER
MSc in Ad anced Mechanical Enginee ing
MASTER OF SCIENCE
Academic Yea : 2016 - 1017
Supe iso s: P o . John Oakey and D . Nigel Simms
Sep embe 2017
CRANFIELD UNIVERSITY
SCHOOL OF ENERGY AND POWER
MSc in Ad anced Mechanical Enginee ing
MASTER OF SCIENCE
Academic Yea 2016 - 2017
ORIOL LÁZARO CASTELL
THE IMPACT OF ADDITIVE MANUFACTURING ON MICRO-
TURBINE PARTS
Supe iso : P o . John Oakey and D . Nigel Simms
Sep embe 2017
© C an ield Uni e si y 2017. All igh s ese ed. No pa o his
publica ion may be ep oduced wi hou he w i en pe mission o he
copy igh owne .
i
ABSTRACT
In he coming yea s, Addi i e Manu ac u ing will play an inc easingly ole in mic o-
u bine p oduc ion, he e o e he impo ance o know he impac o his ype o
manu ac u ing a e c ucial o he op imum design and pe o mance o he de ice
and he igh de ini ion o he s ess a igue gene a ed. Is o his eason ha he
aim o his p ojec is o s udy he impac ha Addi i e Manu ac u ing pu s on he
design and he pe o mance o a mic o- u bine wheel, ega ding he s esses
gene a ed on i .
To s udy he inal s esses p ope ly, a model o he mic o- u bine has been
de eloped wi h ANSYS Wo kbench. I includes an ini ial CFD s udy wi h ANSYS
CFX o ind he empe a u es and p essu es unde which he mic o- u bine wo ks.
Then, hese empe a u es a e impo ed o ANSYS The mal o calcula e he
empe a u e dis ibu ion in he geome y and a e , CFX and ANSYS The mal
esul s a e impo ed o ANSYS S uc u al o ca y ou a s ess analysis.
The esul s o he p ojec a e aken om he execu ion o his model wi h he
p ope ies o Inconel 625, manu ac u ed by Selec i e Lase Mel ing and
manu ac u ed by Cas ing, and a deep s udy o he s ess gene a ed a he
junc ion be ween he blade and he hub is ca ied ou a he end.
Keywo ds:
Inconel 625, Mic o- u bine, Selec i e Lase Mel ing, Cas ing.
ii
ACKNOWLEDGEMENTS
I would like o hank he sponso o his p ojec , HIETA Technologies. Wi hou i s
suppo and in o ma ion p o ided, he wo k p esen ed he e would no be possible.
I am ho oughly g a e ul o my academic supe iso P o . John Oakey o all his
suppo , help and guidance and h oughou his MSc hesis, I could no ha e done
his wi hou him. I would also like o hank D . Nigel Simms o his help as a
second supe iso .
Finally, hank you my amily who has suppo ed me uncondi ionally no only
du ing my MSc hesis, bu also my uni e si y yea s, wi h unde s anding and
pa ience.
iii
TABLE OF CONTENTS
ABSTRACT ......................................................................................................... i
ACKNOWLEDGEMENTS.................................................................................... ii
LIST OF FIGURES .............................................................................................
LIST OF TABLES ............................................................................................. iii
LIST OF EQUATIONS ........................................................................................ ix
LIST OF ABBREVIATIONS ................................................................................ x
1 INTRODUCTION ............................................................................................. 1
2 LITERATURE REVIEW ................................................................................... 3
2.1 Gas u bine ............................................................................................... 3
2.2 Mic o- u bine ............................................................................................. 3
2.2.1 B eakdown ......................................................................................... 4
2.2.2 Wo king cycle ..................................................................................... 5
2.2.3 Tu bine wheel wo king condi ions ...................................................... 6
2.2.4 Pe o mance ....................................................................................... 7
2.2.5 Tu bine wheel ma e ials ..................................................................... 8
2.2.6 Tu bine wheel manu ac u e .............................................................. 10
2.2.7 Mic os uc u e cha ac e is ics o Inconel 625 ................................... 17
3 METHODOLOGY .......................................................................................... 20
3.1 Geome y ................................................................................................ 20
3.2 Ma e ial p ope ies .................................................................................. 21
3.2.1 Con en ional manu ac u ing: Cas ing............................................... 21
3.2.2 Addi i e manu ac u ing: Selec i e Lase Mel ing .............................. 23
3.3 CFX ......................................................................................................... 24
3.3.1 Fluid domain ..................................................................................... 25
3.3.2 CFX Mesh ........................................................................................ 25
3.3.3 Bounda y condi ions ......................................................................... 28
3.4 The mal analysis ..................................................................................... 31
3.4.1 ANSYS The mal mesh ..................................................................... 31
3.4.2 Bounda y Condi ions ........................................................................ 34
3.5 S uc u al analysis................................................................................... 35
3.5.1 ANSYS S uc u al mesh ................................................................... 36
3.5.2 Bounda y Condi ions ........................................................................ 37
4 RESULTS AND ANALYSIS ........................................................................... 40
4.1 CFX ......................................................................................................... 40
4.2 The mal analysis ..................................................................................... 45
4.3 S uc u al analysis................................................................................... 46
4.3.1 S uc u al analysis esul s o a o a ion speed o 90,000 pm ........... 48
4.3.2 S uc u al analysis esul s o a o a ion speed o 100,000 pm ......... 51
4.3.3 S uc u al analysis esul s o a o a ion speed o 110,000 pm ......... 53
4.3.4 Compa ison ...................................................................................... 55

i
5 DISCUSSION ................................................................................................ 59
6 CONCLUSIONS AND FUTURE RESEARCH ............................................... 61
REFERENCES ................................................................................................. 62
APPENDICES .................................................................................................. 65
Appendix A : Da a shee o Inconel 625 manu ac u ed by SLM om CRP
Meccanica ..................................................................................................... 65
Appendix B : P ope ies o Inconel 625 manu ac u ed by cas ing om
Ma Web ........................................................................................................ 67
Appendix C : Inconel 625 p ope ies ............................................................. 68
Appendix D Mesh sensi i i y esul s ............................................................. 69
Appendix E : CFX esul s .............................................................................. 70
Appendix F : S uc u al da a esul s o 90,000 pm om ANSYS S uc u al
(example) ...................................................................................................... 71
LIST OF FIGURES
Figu e 1: Sec ional iew o a mic o gas u bine. ................................................. 4
Figu e 2: Mic o- u bine wo king cycle. ................................................................ 5
Figu e 3: The elec ical e iciency o he compe i i e o e ings in he mic o- u bine
size ange. ................................................................................................... 7
Figu e 4: In es men cas ing s eps. .................................................................. 10
Figu e 5: Con en ional manu ac u e pos -p ocessing. ..................................... 11
Figu e 6: Addi i e manu ac u ing es ing cycle. ................................................ 12
Figu e 7: SLM p ocess. .................................................................................... 13
Figu e 8: Pos -p ocessing................................................................................. 13
Figu e 9: The s ess-s ain cu es o SLM p oduced IN625 (1), MWCNT-IN625
(2) and hea - ea ed MWCNT-IN625 (3) (le ) and hei mechanical p ope ies
( igh ). ........................................................................................................ 15
Figu e 10: HIETA Technologies Wheel u bine. ............................................... 15
Figu e 11: EBM p ocess. .................................................................................. 16
Figu e 12: EBM pos -p ocessing. ..................................................................... 17
Figu e 13: Cas ing a 1289°C/30 (le ) minu es and a 1288°C /24 minu es ( igh ).
.................................................................................................................. 18
Figu e 14: Bidi ec ional scan mode o he lase o SLM p ocess. .................... 19
Figu e 15: Mic os uc u e o he op su ace o an Inconel 625 sample
manu ac u ed wi h SLM (le ) and “scales” shape o YZ di ec ion o an
Inconel 625 sample manu ac u ed wi h SLM ( igh ). .................................. 19
Figu e 16: ANSYS p ojec in e ace. ................................................................. 20
Figu e 17: Isome ic (le ) and on al ( igh ) iew o he adial u bine analysed.
.................................................................................................................. 21
Figu e 18: App oxima ed cu e o he Young Modulus o Inconel 625
manu ac u ed by cas ing e sus he Tempe a u e. .................................... 22
Figu e 19: The mal conduc i i y o Inconel 625 e sus he Tempe a u e. ........ 23
Figu e 20 Young Modulus o Inconel 625 manu ac u ed by SLM e sus he
Tempe a u e. ............................................................................................. 24
Figu e 21: CFX luid domain o he u bine wheel geome y (le ) and sec ional
iew o he u bine and he luid domain ( igh ). ......................................... 25
Figu e 22: CFX luid domain mesh o he u bine wheel geome y. ................. 27
i
Figu e 23: Mesh sensi i i y o he CFX model. ................................................ 28
Figu e 24: Inle and Ou le domain and he di ec ion o o a ion o he u bine. 29
Figu e 25: In low o he u bine. ........................................................................ 30
Figu e 26: The mal analysis mesh o he u bine geome y (le ) and e inemen
zone ( igh ). ............................................................................................... 32
Figu e 27: The mal analysis con olled mesh o he sha zone (le ) and o he
op zone ( igh ). ......................................................................................... 32
Figu e 28: Con ol poin s o he mesh sensi i i y. ............................................ 33
Figu e 29: Maximum empe a u e o he con ol poin s o he mesh sensi i i y. 34
Figu e 30: Minimum empe a u e o he con ol poin s o he mesh sensi i i y. 34
Figu e 31: Hub zone (le ) and blades zone ( igh ). .......................................... 35
Figu e 32: Impo ed empe a u e o he Hub zone (le ) and he Blades zone
( igh ). ........................................................................................................ 35
Figu e 33: Mic o- u bine mesh o he S uc u al analysis. ............................... 36
Figu e 34: Mesh sensi i i y o he S uc u al analysis. ...................................... 37
Figu e 35: Impo ed p essu e o he hub (le ) and he blades ( igh ). .............. 38
Figu e 36: Impo ed empe a u e om The mal analysis. ................................. 38
Figu e 37: Subjec ion poin o he u bine. ........................................................ 39
Figu e 38: E iciency e sus Inle p essu e o a o a ion speed o 90,000 pm. 40
Figu e 39: Powe e sus Inle p essu e o a o a ion speed o 90,000 pm. ..... 41
Figu e 40: E iciency e sus Inle p essu e o a o a ion speed o 100,000 pm.
.................................................................................................................. 42
Figu e 41: Powe e sus Inle p essu e o a o a ion speed o 100,000 pm. ... 42
Figu e 42: E iciency e sus Inle p essu e o a o a ion speed o 110,000 pm.
.................................................................................................................. 43
Figu e 43: Powe e sus Inle p essu e o a o a ion speed o 110,000 pm. ... 43
Figu e 44: Veloci y low wi hou o ex gene a ion unde ope a ing condi ions o
Pin = 3ba , Tin = 650°C and o a ion speed 110,000 pm. ........................... 44
Figu e 45: Veloci y low wi h o ex gene a ion unde ope a ing condi ions o Pin
= 2ba , Tin = 550°C and o a ion speed 110,000 pm. ................................. 44
Figu e 46: Flow empe a u e dis ibu ion on he u bine su ace unde ope a ing
condi ions o Pin = 3ba , Tin = 650°C and o a ion speed 110,000 pm. ...... 45
ii
Figu e 47: Tu bine body dis ibu ion unde he ope a ing condi ions o Pin = 3ba ,
Tin = 650°C and o a ion speed 110,000 pm. ............................................. 46
Figu e 48: S ess dis ibu ion on a u bine manu ac u ed wi h SLM and unde he
ope a ing condi ions o Pin = 3ba , Tin = 650°C and o a ion speed
110,000 pm. .............................................................................................. 47
Figu e 49: S ess dis ibu ion on a u bine manu ac u ed wi h Cas ing and unde
he ope a ing condi ions o Pin = 3ba , Tin = 650°C and o a ion speed
110,000 pm. .............................................................................................. 47
Figu e 50: Pa h o calcula e he s esses gene a ed. ....................................... 48
Figu e 51: S ess s X unde di e en ope a ing condi ions, 90,000 pm and
Inconel 625 (Cas ing)................................................................................. 50
Figu e 52: S ess s X unde di e en ope a ing condi ions, 90,000 pm and
Inconel 625 (SLM). .................................................................................... 50
Figu e 53: Di e ence be ween he alues o Cas ing and SLM e sus he inle
p essu e o 90,000 pm. ............................................................................ 51
Figu e 54: S ess s X unde di e en ope a ing condi ions, 100,000 pm and
Inconel 625 (Cas ing)................................................................................. 52
Figu e 55: S ess s X unde di e en ope a ing condi ions, 100,000 pm and
Inconel 625 (SLM). .................................................................................... 52
Figu e 56: Di e ence be ween he alues o Cas ing and SLM e sus he
p essu e wi h 100,000 pm. ........................................................................ 53
Figu e 57: S ess s X unde di e en ope a ing condi ions, 110,000 pm and
Inconel 625 (Cas ing)................................................................................. 54
Figu e 58: S ess s X unde di e en ope a ing condi ions, 110,000 pm and
Inconel 625 (SLM). .................................................................................... 54
Figu e 59: Di e ence be ween he alues o Cas ing and SLM e sus he
p essu e wi h 100,000 pm. ........................................................................ 55
Figu e 60: S ess e sus o a ion speed o σmax (middle zone). ...................... 56
Figu e 61: S ess e sus o a ion speed o σmin. .............................................. 57
Figu e 62: S ess e sus o a ion speed o σmax. ............................................. 57
Figu e 63: Di e ence be ween cas ing and SLM esul s e sus Inle p essu e
compa ison. ............................................................................................... 58
4
Conce ning he en i onmen al impac , mic o- u bines ha e a high noise le els
and equi e a speci ic acous ic sys em. Howe e , he ai cooling bea ings a oid
lub ican s con amina ion by combus ion p oduc s and p olongs he equipmen
use ul li e. Mo eo e , mic o- u bine manu ac u e s usually uses hea exchange s
o ake p o i o he combus ion gases exhaus and p ehea he ai in ake o he
combus ion chambe , achie ing a he mal e iciency o 30% (Rosa Do
Nascimen o e al., 2013).
The comme cial cus ome equi emen s o mic o- u bines a e ha hey need o
be e y clean, wi h low NOx, CO and unbu ned hyd oca bons, ha e high
e iciency and low main enance, e y low o ced ou age a e and low ins alla ion
cos o p o ide as payback o he owne s (Goli, Kondi and Timmanpalli, 2015).
2.2.1 B eakdown
This ype o u bines below ew hund eds o kilowa s in size usually uses
cen i ugal machine y (Eps ein, 2004), and consis s o a supe sonic adial low
comp esso and u bine, connec ed by he same sha . Radial- low
u bomachine y, can handle he e y small olume ic lows o ai and combus ion
p oduc s, wi h highe componen e iciency and wi h simple cons uc ion han
axial- low componen s (Soa es, 2007). The elec ical gene a o is also moun ed
on he same sha . Hence he e a e only one o a ing pa , a oiding he needed
o a gea box and he p oblems
due o conside able numbe o
mo ing pa s. This simplici y,
makes his pa ela i ely easy
o manu ac u e and main ain,
and p esen s a g ea po en ial
o inexpensi e and la ge scale
manu ac u ing (Rosa Do
Nascimen o e al., 2013).
Al hough, he low mechanical
ine ia o he sha , makes mic o- u bines di icul o con ol. The main pa s a e
showed in Figu e 1 (T a e so, Calzola i and Massa do, 2017).
Figu e 1: Sec ional iew o a mic o gas u bine.

5
The inne bea ing holding he sha is hyd odynamic bea ing, which con ains
g oo es o lub ica ion, making he bea ing smalle , cheape and mo e e icien ,
and p o iding he bea ing wi h low main enance and high se ice li e (Kingsbu y,
n.d.). The ou e bea ing u ilizes a ce amic ball ace. The hea exchange has
impo ance o inc ease he cycle e iciency o he mic o u bine (Goli, Kondi and
Timmanpalli, 2015).
2.2.2 Wo king cycle
The exhaus gases coming om he combus o wi h high eloci y and
empe a u e o a e he u bine wheel. The u bine wheel, he comp esso and he
gene a o a e moun ed on he same sha , in consequence, he gene a o o a es
wi h he same speed as he u bine and gene a es elec ici y. This elec ici y is
condi ioned h ough powe
elec onics de ices and
supplied o he equi ed
a eas. The uel is injec ed
o he combus o in he
gaseous o m and esh
and comp essed ai pass
h ough he hea exchange
and is also in oduced. The
schema ic p ocess is
showed in Figu e 2.
The ole o he hea exchange is o ake p o i om he ho u bine exhaus gases,
wi h a empe a u e abou 500 - 600ºC, and inc eases he empe a u e o he ai
coming om he comp esso ( ypically a ound 150ºC), inc easing he o e all
e iciency o he mic o gas u bine (U.S. En i omend P o ec ion Agency, 2015).
This allows he cycle e iciency o ake alues as much as 30% while he a e age
ne e iciency o un eco e ed mic o- u bines is 17 % (Rosa Do Nascimen o e al.,
2013).
Figu e 2: Mic o- u bine wo king cycle.
6
2.2.3 Tu bine wheel wo king condi ions
This hesis ocuses only on he u bine wheel, hen he ope a ing condi ions a e
explained only o his pa . The ollowing in o ma ion is ex ac ed om an
expe imen al es o he ope a ing condi ions o a mic o gas u bine de ice
(Robe o Capa a, 2015) which shows he wo king condi ions o he u bine wheel
om a G aupne /Je Ca u bo-p op engine ( his in o ma ion can be ex apola ed
as an ini ial es ima ion o he pe o mance o a mic o gas u bine). As well as
hese alues a e con as ed wi h he expe imen a ion o a Tu bec T100 adial
mic o- u bine (Hohloch e al., 2017). The expe imen shows ha he inle
empe a u e ange o he u bine is om 500 o 800ºC and o he ou le
empe a u e is om 500 o 600ºC (Cado in e al., 2012).
The u bine p essu e a io is he a io o u bine p essu e ou le o inle p essu e
and dec eases wi h highe u bine speed.
𝛽𝑡𝑢𝑟𝑏𝑖𝑛𝑒 =𝑃𝑜𝑢𝑡𝑝𝑢𝑡
𝑃𝑖𝑛𝑝𝑢𝑡
Equa ion 1
The ange o p essu e a io on a mic o- u bine is om 0.5 o low u bine speed
o 0.25 o high u bine speed, and he ou le p essu e goes om 0.95 ba o 1
ba espec i ely, gi ing an inle p essu e ange om 2 o 4ba app oxima ely. The
inc ease in p essu e wi h highe u bine speed is due o he ising p essu e losses
caused by he subsequen componen s (Hohloch e al., 2017).
Conce ning he inle u bine low, he alue a ies om 0.15 kg/s a low u bine
speeds o 0.8 kg/s a high speeds. The inle u bine low is he esul o ai inle
plus inle uel mass low (Cado in e al., 2012).
𝑚𝑡𝑢𝑟𝑏𝑖𝑛𝑒,𝑖𝑛𝑙𝑒𝑡 =𝑚𝑎𝑖𝑟,𝑖𝑛𝑙𝑒𝑡 +𝑚𝑓𝑢𝑒𝑙,𝑖𝑛𝑙𝑒𝑡
Equa ion 2
The uel inle low goes om 0.001 kg/s o 0.010 kg/s demons a ing ha mos o
he inle u bine low is composed by ai .
Ano he in e es ing pa ame e is he o a ional speed o he u bine wheel, which
delimi he speed o he comp esso and he gene a o (excep some
manu ac u e s who include a gea box be ween he u bine-comp esso sha and
7
he gene a o sha o educe he speed on he gene a o ). The o a ional speed
goes om 30,000 o 120,000 pm (Rosa Do Nascimen o e al., 2013).
2.2.4 Pe o mance
Comme cial mic o u bines om 25 o 500kW used o powe gene a ion p oduce
bo h hea and elec ici y on a ela i ely small scale. Un- ecupe a ed mic o-
u bines ha e an elec ical e iciency a ound 15%. I hey ha e ecupe a o s, he
elec ical e iciency will be a ound 20 - 30%, and i i has hea eco e y, he
e iciency will a ise up o 85% (Goli, Kondi and Timmanpalli, 2015).
The pe o mance o mic o gas u bines is s ongly a ec ed by empe a u e
condi ions. The decline obse ed a highe empe a u es is explained by he lowe
ai densi y and consequen ly lowe mass low a e h ough he powe uni
(Ca esana e al., 2010).
The elec ical e iciency o he cu en mic o- u bines in he ma ke a e shown in
Figu e 3 (Gille e, 2010).
Figu e 3: The elec ical e iciency o he compe i i e o e ings in he mic o- u bine
size ange.
8
2.2.5 Tu bine wheel ma e ials
Cu en mic o- u bine u bine wheels a e manu ac u ed wi h nickel-based alloys,
p o iding hem wi h ligh weigh and good esis ance. Howe e , as he u bine
wheel is exposed o high inle empe a u es, no mal ligh weigh ed ma e ials
canno be used. These u bines a e subjec ed o a long- e m exposu e o high
empe a u e up o 800ºC, so hea esis ance is an essen ial p e equisi e.
The e o e, ligh common ma e ials like aluminium o i an-based alloys canno be
used.
The mal c eep, which is a ime-dependen de o ma ion unde high le els o s ess
below he Yield S eng h o he ma e ial, is also a p oblem when choosing he
ma e ial o he u bine. This phenomenon is consequence o he high empe a u e
o he u bine, and o a oid i , ma e ials wi h high mel ing empe a u e mus be
used, as well as c eep es da a needs o be consul ed du ing he ma e ial
selec ion.
Some inno a i e solu ions o he empe a u e p oblems o he u bine ha e
appea ed du ing he las yea s and a e s ill being es ed on mic o- u bines. One
o hem a e he The mal Ba ie Coa ings (TBC) wi h a ce amic ma e ial. This
ype o coa ing should be selec ed so ha is e ac o y enough o esis he high
empe a u es a he su ace and has a low bulk he mal conduc i i y o minimize
hea ans e o he me allic zone. Mo eo e , he he mal expansion o he
selec ed coa ing should closely ma ch he me allic one o minimize po en ial
s esses. Nowadays, he mos widely used ce amic coa ing ma e ial in u bines
indus y is Y ia s abilized zi conia (YSZ). Ano he conside a ion is ha his
coa ing mus ha e a g ain and po e s uc u e o minimize he mal conduc ion o
he me al-ce amic in e ace. The coa ing should ha e enough po osi y, so i
educes he he mal conduc i i y while simul aneously adhe ing o he me al
u bine bond-coa laye (Es ada, 2007). The cu en echniques used o TBC a e
Elec on beam physical apo deposi ion (EBPVD), Ai plasma sp ay (APS), High
eloci y oxygen uel (HVOF), Elec os a ic sp ay-assis ed apou deposi ion
(ESAVD) and Di ec apo deposi ion (DVD).
9
The las yea s SLM echniques a e inc easing sha ply in he u bine
manu ac u ing. Ma e ials like Inconel o Has elloy X a e examples o high
empe a u e and co osion esis an nickel-based alloys which can be
manu ac u ed wi h his me hod. In mos cases, hese alloys con ain ch ome, i on,
niobium and molybdenum and o he alloy componen s, and hey a e o en known
as supe alloys. Nickel-based alloys wi hs and highe empe a u es han s eels
and hey a e also highly weldable. Thei esis ance o empe a u e is achie ed
h ough a mix u e o dispe sion ha dening, p ecipi a ion ha dening and solid
solu ion s eng hening. Nickel-based alloys exhibi good mechanical
cha ac e is ic alues such as high ensile s eng h and good endu ance s eng h.
Inconel can be used a empe a u es o up o 700°C. Has elloy X can e en be
used a empe a u es o up o 1200°C. This makes hese alloys ideally sui ed o
ae ospace echnologies and o u bine p oduc ion. Mo eo e , h ough pos
p ocessing, such as ha dening, hea ea men o ho isos a ic p essing (HIP), he
componen s’ p ope ies can be adap ed o mee speci ic equi emen s (SLM
Solu ions, 2017).
Con inued imp o emen s in powde syn hesis, p ocessing, and densi ica ion
ha e esul ed in he de elopmen o a cu en gene a ion o silicon ni ide
ce amics ha ing con olled mic os uc u es consis ing o elonga ed g ains. These
so-called sel - ein o ced o in-si u oughened ma e ials exhibi supe io
pe o mance as e lec ed by inc eased s eng h, highe ac u e oughness, and
enhanced esis ance o c eep up u e (Soa es, 2007). A silicon ni ide o o o a
adial- low mic o gas u bine wi h a capaci y o 30KWel was de eloped by
F aunho e . The ce amic o o exhibi s long- e m s abili y up o 1200ºC a
maximum ope a ing loads and can be mass-p oduced. Depending on chemical
composi ion, sin e ing and pos p ocessing, speci ic p ope ies can be imp o ed.
The pos p ocessing led o high s eng h as well as high oxida ion esis ance and
a igue s eng h up o 1200ºC. The inal da a can be seen in Table 1 (S ockmann
e al., n.d.).

10
Table 1: Silicon ni ide da a.
Ope a ing empe a u e (ºC)
1200
F ac u e oughness (MPa·m1/2)
6.8
Yield S eng h (MPa)
1000
Fa igue s eng h a 1200ºC (MPa)
500
2.2.6 Tu bine wheel manu ac u e
The u bine is he pa o he mic o- u bine machine which gi es he mo ion o he
sha , he e o e i is he mos impo an pa o he de ice and needs a
sophis ica ed me hod o manu ac u e, in o de o accomplish he ha d
speci ica ions equi ed due o he s ong condi ions o s ess and empe a u e.
Du ing he las yea s, ad anced manu ac u e plays a key ole in manu ac u e,
and i in luences posi i ely o he ab ica ion o mic o- u bines.
2.2.6.1 Con en ional manu ac u e o mic o- u bines: In es men cas ing
P oducing a u bine wheel wi h such speci ic demands o pe o mance, is a
echnique de eloped o e many yea s, since he mic o- u bines appea ed. The
u bine wheel is usually manu ac u e wi h nickel-based supe alloys, a high
s eng h in es men cas ing. These ypes o alloys a e designed o wi hs and high
empe a u es and i is a ade-o in machinabili y. The e o e, an in es men
cas ing p ocess has been de eloped du ing many yea s o achie e he bes
esul s, and minimize he machining on i (Ramamu hy and Thenna a ajan,
2008).
The con en ional p ocess o in es men cas ing has he ollowing s eps, showed
in Figu e 4 (James, 2015):
Figu e 4: In es men cas ing s eps.
3D
model Wax Shell Cas ing
11
The i s s ep o he in es men cas ing is o manu ac u e he wax pa e n o he
p ocess om a 3D model. The wax is used o he ease o mel ou and he
possibili y o euse i . Se e al wax pa e ns a e assembled on a ee and ypically
he pa e n is des oyed du ing he p ocess, so a new one is equi ed o each
cas ing. Once he ee and he pa e ns a e assembled, me al cas ing pa e n is
hen dipped in a e ac o y slu y whose composi ion includes ex emely ine-
g ained silica, wa e and binde s. A ce amic coa ing is buil o e he wax su ace.
The pa e n is hen epea edly dipped in o he slu y o inc ease he hickness o
he ce amic coa . Once he ce amic coa has he oughness equi ed, i is d ied
in ai o ha den i . Then he wax is mel ed a 90 – 175°C and i lows ou he shell
(The lib a y o Manu ac u ing, n.d.).
The ce amic shell is hen hea ed o 550 - 1100°C o s eng hen u he he mould
and elimina e any es o wax, wa e o con aminan s. The me al cas ing is pou ed
while he mould is s ill ho allowing he liquid me al o low easily h ough he
mould ca i y and sh inking oge he as hey cool, which gi es a be e
dimensional accu acy. A e he cas ing is solidi ied, he ce amic shell is b oken
om he piece and all he pa s a e cu away om he ee (The lib a y o
Manu ac u ing, n.d.).
When he cas ing p ocess inish, g inding, milling and elec ical discha ge
machining (EDM) p ocesses a e equi ed (Figu e 5).
Figu e 5: Con en ional manu ac u e pos -p ocessing.
To inish he mic o- u bine manu ac u e, a inal coa ing is ca ied on, inc easing
he ma e ial p ope ies.
2.2.6.2 Ad anced Manu ac u ing o mic o- u bines
The mos widely addi i e manu ac u ing echnique used o mic o- u bine
manu ac u ing is Selec i e Lase Mel ing (SLM). This ype o manu ac u ing, as
a di e ence o con en ional cas ing, allows an ini ial as es ing and alida ion
p ocess be o e o s a he inal in eg a ed de elopmen .
G inding Milling EDM Coa ing
12
2.2.6.2.1 Selec i e Lase Mel ing
Conce ning he ad anced manu ac u ing echniques on mic o- u bines
manu ac u e, es ing and alida ion wi h SLM
p ocess can be seen in Figu e 6 (James,
2015). This p ocess is comple ely di e en
om con en ional manu ac u ing, allowing
e iciency and p ecision o he p ocess.
Mo eo e , his ype o addi i e manu ac u ing
allows he ab ica ion o nickel-based alloy
mic o- u bines wheels. The i s s age s a s
wi h he design o he model and i s simula ion
o app o e he model be o e o s a he eal
p ocess wi h Selec ed Lase Mel ing (SLM). SLM is a apid p o o yping, 3D
p in ing o addi i e manu ac u ing p ac ise which uses a high-powe densi y lase
o mel and use me allic powde oge he . This p ocess has a limi ing ac o , he
high su ace oughness (Ra ≥ 5μm). Is o his eason ha a pos -p ocessing
me hod o mee he equi emen s o su ace oughness (Ra ≥ 0.8μm) is needed.
The e o e, o achie e his small oughness, mic o machining p ocess is equi ed
(Such and Meine s, 2015).
A e he pos -p ocessing, du ing ins umen a ion s age, all he measu emen s
and c i ical dimensions a e equi ed o be checked o see i he u bine mee s he
ini ial equi emen s. To inish he cycle, a es is ca ied ou showing he quali y
o he p oduc h ough an objec i e way. When he es ing and alida ion cycle is
comple ed, ad anced manu ac u ing allows in eg a ed de elopmen wi h i e a i e
and as cycles.
The p ocess abo e has many ad an ages, i allows pa allel and in eg a ed
de elopmen p ocess, adical de elopmen app oaches, ambi ious and sho
de elopmen goals, and sho i e a i e cycles (James, 2015).
Once he es ing and e i ying cycle and he eal pe o mance o he mic o- u bine
p o o ype is accep ed, he p ocess chain o SLM p oduc ion s a s (Figu e 7).
The aw ma e ial o his p ocess a e he powde (wi h diame e lowe han 50μm),
3D design and
simula ion
SLM P ocessing
Pos p ocessing
Ins umen a ion
Tes ing
In eg a ed
de elopmen
Figu e 6: Addi i e manu ac u ing
es ing cycle.
13
hen he machine is condi ioned and he p ocess designed. The con ol
pa ame e s o he machining a e implemen ed on he SLM machine and he
u bine p ocessing s a s (James, 2015).
Figu e 7: SLM p ocess.
The pos -p ocessing s a s when he mic o- u bine is eady a e he SLM
p ocess, hen a quali y con ol de ec s i de ec s appea ed du ing he addi i e
manu ac u e p ocess. I he SLM p ocess pass he quali y con ol, he mic o-
u bine ecei e a hea ea men , o imp o e he s eng h o he u bine ma e ial,
bu his p ocess can howe e esul in undesi able esidual s esses, which may
lead o c acking and dis o ion o he componen ea ed (F y, n.d.). Ano he
quali y con ol is ca ied ou o check i he hea ea men is igh and he u bine
goes o he machining pa o inish he u bine (Figu e 8).
Figu e 8: Pos -p ocessing.
A he inal pa o he mic o- u bine machining he piece is submi ed o a inal
quali y con ol.
Ad anced manu ac u ing p ocess educes cos s and imp o es he esponse ime
o cus ome s du ing he mic o- u bine design p ocess o o small p oduc ion
quan i ies, as well as, p o ides a good oppo uni y o imp o ed e iciency wi hin
he epai and e u bishmen business. Howe e , he unde lying cha ac e is ics
and p ope ies o many me al powde s (especially high- empe a u e supe -alloys)
a e no ye well-unde s ood and es ed o he indus y's exac ing s anda ds (Alle ,
2016).
I is expec ed ha p ope ies o lase buil pa s will be di e en om adi ional
manu ac u ing. One o he p oblems o his echnique is he la ge esidual s ess
ound on he ab ica ed pieces because he apid empe a u e cycles and s eep
empe a u e g adien s occu in he scanned laye s. This esidual s esses in
Powde
supply
SLM machine
condi ion
P ocess
design
Machine se -
up
SLM
p ocessing
Quali y con ol Hea
ea men Quali y con ol Machining
Finishing and
inal quali y
con ol
20
3 METHODOLOGY
The me hodology used o he calcula ion o he s esses consis s wi h h ee main
pa s. A he beginning, a simula ion o he low on he u bine wi h CFX was done
o know he empe a u e and p essu e gene a ed on he mic o- u bine su ace.
Once hese alues we e known i was possible o impo hem o a The mal-
s eady analysis and know how he empe a u e was dis ibu ed in he u bine
geome y. A he end, he p essu e om CFX and he empe a u e om The mal-
s eady we e impo ed o a S uc u al-s eady analysis in o de o ge he s ess
gene a ed in he mic o- u bine.
The s udy was ca ied ou wi h he ma e ial p ope ies o Inconel 625
manu ac u ed by cas ing and wi h Inconel 625 manu ac u ed by SLM.
All he p ocedu e can be seen in Figu e 16, which shows he Wo kbench in e ace
o he p ojec , and all he pa s c ea ed o calcula e he inal s ess dis ibu ion o
he mic o- u bine.
Figu e 16: ANSYS p ojec in e ace.
3.1 Geome y
The geome y analysed was a adial u bine p o ided by he company HIETA
Technologies, an addi i e manu ac u ing company. This geome y has a

21
diame e o 6.2cm and 11 blades. The geome y can be seen in Figu e 17, which
is composed by he u bine wheel and he subjec ion sha .
Figu e 17: Isome ic (le ) and on al ( igh ) iew o he adial u bine analysed.
This geome y allowed o ca y ou a s udy wi h a eal geome y, which gi es
mo e ealis ic esul s and he possibili y o see he di e ence be ween he
s esses gene a ed, depending on he wo ypes o manu ac u ing.
3.2 Ma e ial p ope ies
The ma e ial used o he compa ison be ween he wo ypes o manu ac u ing
was as s a ed abo e Inconel 625. The p ope ies depending on he empe a u e,
o make e en mo e ealis ic he esul s could no be ob ained du ing his b ie
pe iod o he hesis, and his could be one o he nex s eps o he p ojec .
The e o e, o ca y ou he s udy, he cu es o he Young Modulus depending on
he empe a u e we e app oxima ed o bo h ma e ials. This app oxima ion
allowed o achie e he esul s o his hesis, bu hey should be compa ed wi h
he eal alues, and i equi es expe imen a ion.
3.2.1 Con en ional manu ac u ing: Cas ing
The mechanical p ope ies used o Inconel 625 manu ac u ed by cas ing we e
ob ained om he web page Ma Web (Appendix B) and a e shown in Table 2.
22
Table 2: Main p ope ies o Inconel 625 manu ac u ed by cas ing.
Mechanical p ope ies a 20°C
Tensile S eng h [MPa]
836.3
Yield S eng h [MPa]
390
Modulus Elas ici y [GPa]
208
Poisson Ra io
0.28
Shea Modulus [GPa]
81.4
The Young Modulus used was 208GPa a 20°C, co esponding o Inconel 625
(Annealed 1180°C + Aged 760°C). This alue was ex apola ed o make i
dependan on he empe a u e, ollowing he same slope as he na u al Inconel
625 p ope ies (Appendix C).
The inal cu e used o he Young Modulus (E) e sus he empe a u e is shown
in Figu e 18.
Figu e 18: App oxima ed cu e o he Young Modulus o Inconel 625 manu ac u ed
by cas ing e sus he Tempe a u e.
0
50
100
150
200
250
0100 200 300 400 500 600 700 800
Ecas ing [GPa]
Tempe a u e [°C]
Young Modulus Inconel 625 (Cas ing)
23
As can be seen in he plo abo e, he Young Modulus dec eases abou 50GPa
wi h an inc emen o he empe a u e o 800°C.
Fo he he mal s udy, he he mal conduc i i y was equi ed o pe o m he
analysis. Knowing ha he he mal conduc i i y is app oxima ely he same o he
Inconel 625 manu ac u ed by cas ing and by SLM, he ollowing he mal
conduc i i y, showed in Figu e 19, was aken o bo h ype o ma e ials and was
ex ac ed om he p ope ies o Inconel 625 showed in Appendix C.
Figu e 19: The mal conduc i i y o Inconel 625 e sus he Tempe a u e.
3.2.2 Addi i e manu ac u ing: Selec i e Lase Mel ing
The mechanical p ope ies used o his ma e ial we e aken om he CRP
Meccanica web page, a CNC machining company. The da a shee can be seen
in Appendix A and he main p ope ies a e s a ed in Table 3.
Table 3: Main p ope ies o Inconel 625 manu ac u ed by SLM.
Mechanical p ope ies a 20°C
Tensile S eng h in ho izon al di ec ion XY [MPa]
827
Tensile S eng h in e ical di ec ion Z [MPa]
827
Yield S eng h in ho izon al di ec ion XY [MPa]
414
0
5
10
15
20
25
0100 200 300 400 500 600 700 800 900
The mal conduc i i y [W/(m·K)]
Tempe a u e [°C]
The mal conduc i i y Inconel 625
24
Yield S eng h in e ical di ec ion Z [MPa]
414
Modulus Elas ici y in ho izon al di ec ion XY [GPa]
170
Modulus Elas ici y in ho izon al di ec ion Z [GPa]
160
Maximum ope a ing empe a u e unde load [°C]
650
The Young Modulus was also app oxima ed wi h he same c i e ia as he cas ing
manu ac u e and ollowing he slope o he Young Modulus in Appendix C, due
o he impossibili y o ind speci ic and accu a e da a, o example he σ-ε cu e
depending on he empe a u e. Following he sec ion 2.2.7.2, he di e ence o
his ype o manu ac u ing is ha he Inconel 625 manu ac u ed wi h SLM (s ess
elie ed) in oduce aniso opy o he piece and he p ope ies change depending
on he di ec ion as can be seen in Table 3 and in Figu e 20.
Figu e 20 Young Modulus o Inconel 625 manu ac u ed by SLM e sus he
Tempe a u e.
3.3 CFX
This Compu a ional Fluid Dynamics (CFD) so wa e ool deli e s eliable and
accu a e solu ions o u bomachine y applica ions. And i is ecognized o i s
ou s anding accu acy, obus ness and speed wi h o a ing machine y (ANSYS,
2017). The CFX analysis is he i s pa o he model, and gi es he condi ions o
p essu e and empe a u e unde he u bine wo ks.
100
110
120
130
140
150
160
170
180
0100 200 300 400 500 600 700 800
E [MPa]
Tempe a u e [°C]
Young Modulus SLM (s ess elie ed)
XY di ec ion
Z di ec ion
25
The main p oblem o his analysis was ha he eal condi ions o his geome y
we e no known. The e o e, i was necessa y o ca y ou an accu a e analysis in
o de o ealise which we e he app oxima e ope a ing condi ions o his u bine,
and making he simula ion mo e eliable.
3.3.1 Fluid domain
The i s s ep was o cons uc he luid domain. This domain was designed wi h
Solidwo ks, and a e impo ed o ANSYS. In ANSYS was used he ope a ion
Boolean o sub ac he u bine geome y om he luid domain, and he esul is
shown in Figu e 21.
Figu e 21: CFX luid domain o he u bine wheel geome y (le ) and sec ional
iew o he u bine and he luid domain ( igh ).
The inle heigh o he luid domain is 7.8mm and he o al leng h o he geome y
50mm.
3.3.2 CFX Mesh
This was he mos impo an pa o he CFX analysis o ge eliable and accu a e
esul s and i equi ed a special a en ion. The inal model was analysed wi hou
any simpli ica ion, because he appea ance o many di icul ies o apply ci cula
symme y and a e impo he esul s o ANSYS The mal and ANSYS S uc u al
due o he geome y gi en.

26
To ca ch he wall e ec s was necessa y o calcula e he i s bounda y laye o
he mesh o e he blades. The i s s ep was o calcula e he eloci y a he inle :
𝑣𝑖𝑛𝑙𝑒𝑡 =𝐹𝑙𝑜𝑤
𝐴𝑟𝑒𝑎 =155.57𝑚
𝑠
Equa ion 3
The dynamic iscosi y o he luid a 4ba and 800ºC is 4.579x10-5 Pa·s-1, hen:
𝑅𝑒=𝜌∙𝑣𝑖𝑛𝑙𝑒𝑡 ∙𝐷ℎ
𝜇=261,191.5
Equa ion 4
Whe e 𝜌 is he densi y o he ai , 𝑣𝑖𝑛𝑙𝑒𝑡 he eloci y a he inle , 𝐷ℎ he hyd aulic
diame e and 𝜇 he dynamic iscosi y.
To calcula e he skin ic ion coe icien , he Blasius equa ion was used:
𝐶𝑓 =0.079·𝑅𝑒−0.25 =3.495𝑥10−3
Equa ion 5
Then, i was possible o calcula e he wall shea s ess and shea eloci y (also
called ic ion eloci y):
𝜏𝑤=0.5·𝐶𝑓·𝜌·𝑣𝑖𝑛𝑙𝑒𝑡2=52.44 𝑃𝑎
Equa ion 6
𝑢𝑇=(𝜏𝑤
𝜌)0.5 =6.503𝑚
𝑠
Equa ion 7
The e o e, aking a y+ = 30, he i s laye hickness had o be:
𝛿𝑦=𝑦+·𝜇
𝜌·𝑢𝑇=30·4.579𝑥10−5 𝑃𝑎 𝑠
⁄
1.24𝑘𝑔 𝑚3
⁄·6.503𝑚𝑠
⁄=1.7𝑥104 𝑚
Equa ion 8
The inal inpu pa ame e s o he mesh gene a ion we e a Size unc ion o
P oximi y and Cu a u e, Fine Rele ance Cen e, Medium Smoo hing, Slow
T ansi ion and Fine Span angle cen e. The minimum ace and edge size was
0.015mm and he maximum 3mm. The g ow h a e was he common alue o 1.2.
To gene a e he igh bounda y laye - Figu e 22 -, he in la ion mode was
ac i a ed on he hub and blades su aces, wi h an in la ion op ion o Fi s
bounda y laye equal o 0.017mm as calcula ed abo e.
27
Figu e 22: CFX luid domain mesh o he u bine wheel geome y.
The inal mesh had 764,048 elemen s and 1,582,908 nodes. The mesh sensi i i y
explained below was ca ied ou o gene a e he op imum mesh and check he
obus ness o he model.
3.3.2.1 Mesh sensi i i y
A o al o ou meshes we e analysed o ge he op imum numbe o elemen s.
The pa ame e analysed o he mesh sensi i i y was he isen opic e iciency wi h
an ope a ing condi ions close o he inal expec ed alues. The di e ence
be ween he bigges mesh wi h 1,480,211 elemen s go a di e ence be ween he
mesh h ee wi h 764,048 elemen s, o 0.34%. This alue was lowe han 1%, so
was assumed ha he isen opic e iciency con e ged. Figu e 23 shows he
isen opic e iciency agains he numbe o elemen s, and can be seen ha a e
764,048 elemen s, he cu e became almos cons an .
28
Figu e 23: Mesh sensi i i y o he CFX model.
3.3.3 Bounda y condi ions
The luid domain mesh was impo ed o CFX in o de o s udy he ope a ing
condi ions and knowing he empe a u e and p essu e dis ibu ions on he u bine
wheel. The analysis was wi h a s eady s a e.
The basic bounda y condi ions de ined we e a he inle and he ou le o he
u bine. As can be seen in Figu e 24 he inle domain was ma ked wi h in low
a ows and he ou le domain as ou low a ows.
Once he inle and ou le we e de ined, he angula eloci y was in oduced, as
well as he di ec ion o o a ion. The alue o he o a ion speed as explained in
sec ion 2.2.3, a ies om 80,000 o 120,000 pm o he mic o- u bines, so a
o a ion speed sensi i i y was done in sec ion 3.3.4, be ween he eloci ies
90,000 and 110,000 pm. The Figu e 24 also shows he di ec ion o o a ion o
he adial u bine.
81.5
82.0
82.5
83.0
83.5
84.0
84.5
85.0
85.5
86.0
200,000 600,000 1,000,000 1,400,000
Isen opic e iciency [%]
Mesh Elemen s
CFX Mesh sensi i i y
29
Figu e 24: Inle and Ou le domain and he di ec ion o o a ion o he u bine.
The luid assigned was Ai Ideal Gas, as he e ec o he uel was conside ed
negligible, because as explained in sec ion 2.2.3, is less han 1% o he o al low.
The e e ence p essu e was se a 1ba , he hea ans e was de ined as To al
ene gy and he u bulence model assigned was he Shea s ess anspo , he
common model used o his ype o u bines.
The in low/ou low bounda y condi ions we e s udied and, as he o a ion speed,
i was seen how he a ia ion o his pa ame e s changed he inal ope a ing
condi ions. The e o e, ollowing he sec ion 2.2.3, he alue o he p essu e a
he inle we e s udied be ween 2 and 4ba , and he empe a u e be ween 550 and
650ºC. The ou le p essu e was se cons an a 1 ba .
Fo he in low di ec ion condi ion, a alue o 70 deg ees espec he adial
di ec ion was assigned, as seen in Figu e 25.
36
3.5.1 ANSYS S uc u al mesh
The mic o- u bine mesh designed o he s uc u al analysis - Figu e 33 - di e s
om he he mal analysis mesh. I equi ed mo e accu acy a zones whe e he
s ess was highe o ca ch all he e ec s. The mos impo an e inemen zone
was he join be ween he blades and he hub. This is he zone wi h highe
s esses on he u bine (excluding he sha because his hesis ocuses only on
he u bine geome y wi hou he sha ). The e o e, an elemen size o 0.8mm was
speci ied a his zone.
Figu e 33: Mic o- u bine mesh o he S uc u al analysis.
On all he co ne s be ween he junc ion o he blade-hub and he bo om pa o
he u bine, a sphe e o 0.5mm o adius and a sizing o 0.08mm was applied,
and some changes we e obse ed be o e and a e i .
Mo eo e , a sizing o 1mm and ha d beha iou on he sha was applied o a oid
small elemen s.
The Size Func ion used was P oximi y and Cu a u e, a Medium Rele ance
Cen e, a Medium Smoo hing, a Medium Span Angle Cen e and Fas ansi ion.
The minimum size o he aces and edges was 0.08mm and he maximum 1mm.

37
The o al elemen s o he op imised mesh was 960,582 as can be seen in he
mesh sensi i i y below.
3.5.1.1 Mesh sensi i i y
Since he zone o in e es is he join be ween he blades and he hub, he
maximum s ess gene a ed on he join a bo h pa s o he blade was plo ed.
Figu e 34 shows he mesh sensi i i y ca ied ou a his zone.
Figu e 34: Mesh sensi i i y o he S uc u al analysis.
The mesh sensi i i y showed ha he key poin o he mesh elemen s was
960,582. The s ess alue a his poin o elemen s di e ed om he las alue
less han 1%, hen i was conside ed ha he mesh con e ged.
3.5.2 Bounda y Condi ions
The bounda y condi ions applied o he S uc u al analysis we e he p essu es
impo ed om CFX, he empe a u es impo ed om he The mal analysis, and
he subjec ion o he u bine.
The p essu e impo ed is shown in Figu e 35, which was di ided by he hub zone
and he blades zone.
35
45
55
65
75
85
100,000 350,000 600,000 850,000 1,100,000 1,350,000
S ess [MPa]
Mesh elemen s
P1 max
P2 max
38
Figu e 35: Impo ed p essu e o he hub (le ) and he blades ( igh ).
The empe a u e impo ed is shown in Figu e 36. I shows he body empe a u e
o he u bine, which in luences he s esses gene a ed depending on he
empe a u e o he poin .
Figu e 36: Impo ed empe a u e om The mal analysis.
And he las inpu needed o speci y was he subjec ion poin o he u bine. I is
shown in Figu e 37. The exac subjec ion does no in luence he s esses on he
blade s udied, since he s esses on he sha we e no s udied.
39
Figu e 37: Subjec ion poin o he u bine.
40
4 RESULTS AND ANALYSIS
4.1 CFX
The ope a ing condi ions o his geome y we e no known. The e o e, o obse e
how he ope a ing condi ions in luence he inal s ess gene a ed on he
geome y, a sensi i i y analysis was done o he inle p essu e, he inle
empe a u e and he o a ion speed.
The inle p essu e was s udied be ween 2 and 4ba and he empe a u e a he
inle be ween 550 and 650°C ( he maximum wo king empe a u e o Inconel 625
is app oxima ely 650°C). Mo eo e , each o hese anges we e simula ed o he
o a ion speeds be ween 90,000 and 110,000 pm.
The ou le p essu e was ixed du ing all he analysis a 1ba and he in low angle
a 70° wi h espec o he adial di ec ion.
The esul s ob ained o he E iciency and Powe wi h a o a ion speed o
90,000 pm a e shown in Figu e 38 and Figu e 39.
Figu e 38: E iciency e sus Inle p essu e o a o a ion speed o 90,000 pm.
82
84
86
88
90
92
94
1.5 2 2.5 3 3.5 4 4.5
E iciency [%]
Inle p essu e [ba ]
550°C, 90000 pm
600°C, 90000 pm
650°C, 90000 pm
41
Figu e 39: Powe e sus Inle p essu e o a o a ion speed o 90,000 pm.
I is clea ha when he p essu e is 2ba , he e iciency does no change so much
wi h he empe a u e a ia ion, and he alues a e be ween 91.5 - 92.5%.
Howe e , when he p essu e inc eases, he alue o he e iciency s a s o di e
depending on he empe a u e, be ween 83 - 86%.
The powe plo shows ha he empe a u e does no ha e a conside able
in luence on he ou pu powe . Howe e , he ou pu powe inc eases conside ably
when he inle p essu e a ises (be ween 12 - 40kW). Mo eo e , wi h an inc ease
o he empe a u e he ou pu powe sligh ly inc eases, bu he e iciency
dec eases.
The same beha iou is shown in Figu e 40 and 41, whe e he o a ion speed was
ixed o 100,000 pm. Howe e , he e iciency and powe is sligh ly highe o his
o a ion speed han he e iciency and powe o 90,000 pm.
10
15
20
25
30
35
40
1.5 2 2.5 3 3.5 4 4.5
Powe [KW]
Inle p essu e [ba ]
550°C, 90000 pm
600°C, 90000 pm
650°C, 90000 pm

42
Figu e 40: E iciency e sus Inle p essu e o a o a ion speed o 100,000 pm.
Figu e 41: Powe e sus Inle p essu e o a o a ion speed o 100,000 pm.
Fo a o a ion speed o 110,000 pm, he beha iou ob ained is qui e di e en ,
especially when he inle p essu e is 2ba .
The plo s a e shown in Figu e 42 and Figu e 43.
87
88
89
90
91
92
93
94
95
1.5 2 2.5 3 3.5 4 4.5
E iciency [%]
Inle p essu e [ba ]
550°C, 100000 pm
600°C, 100000 pm
650°C, 100000 pm
10
15
20
25
30
35
40
45
1.5 2 2.5 3 3.5 4 4.5
Powe [KW]
Inle p essu e [ba ]
550°C, 100000 pm
600°C, 100000 pm
650°C, 100000 pm
43
Figu e 42: E iciency e sus Inle p essu e o a o a ion speed o 110,000 pm.
Figu e 43: Powe e sus Inle p essu e o a o a ion speed o 110,000 pm.
Figu e 42 and Figu e 43 show ha he e iciency and ou pu powe when he inle
p essu e is 2ba and he o a ion speed 110,000 pm is lowe han he e iciency
and powe ob ained o 90,000 and 100,000 pm a he same p essu e. This
beha iou is due o he in low angle which is no he igh one. Compa ed wi h he
low when he in low angle is he igh one (Figu e 44), his si ua ion gene a es a
o ex (Figu e 45), he eby a dec ease o he u bine e iciency.
87
88
89
90
91
92
93
94
95
1.5 2 2.5 3 3.5 4 4.5
E iciency [%]
Inle p essu e [ba ]
550°C, 110000 pm
600°C, 110000 pm
650°C, 110000 pm
5
10
15
20
25
30
35
40
45
1.5 2 2.5 3 3.5 4 4.5
Powe [KW]
Inle p essu e [ba ]
550°C, 110000 pm
600°C, 110000 pm
650°C, 110000 pm
44
Figu e 44: Veloci y low wi hou o ex gene a ion unde ope a ing condi ions o
Pin = 3ba , Tin = 650°C and o a ion speed 110,000 pm.
Figu e 45: Veloci y low wi h o ex gene a ion unde ope a ing condi ions o Pin =
2ba , Tin = 550°C and o a ion speed 110,000 pm.
45
4.2 The mal analysis
Since his s ep is no ele an o he s udy and is only a ansi ion be ween he
inpu pa ame e s a he ini ial pa o he model and he inal s esses calcula ed
a he ou pu , in his sec ion is explained only he empe a u e dis ibu ion o a
u bine unde he ope a ing condi ions o Pin = 3ba , Tin = 650°C and a o a ion
speed o 110,000 pm, which gene a es an ou pu powe o 29.5kW and an
e iciency o 92.4%. The beha iou agains he empe a u e can be ex apola ed
o he o he ope a ing condi ions s udied in his hesis.
The esul s show ha he maximum empe a u e is loca ed a he inle poin o
he u bine, whe e he ho inle low insides. The empe a u e dec eases h ough
he blade om he inle o he ou le . Fo his case, he low empe a u e a he
inle pa o he u bine has app oxima ely 620°C as can be seen in he Figu e 46,
and his low gene a es a empe a u e on he u bine geome y o 614°C a he
inle pa (Figu e 47). Following he same c i e ia, he low empe a u e a he
ou le is 470°C, he eby a empe a u e o 440°C a he ou le pa o he u bine.
Figu e 46: Flow empe a u e dis ibu ion on he u bine su ace unde ope a ing
condi ions o Pin = 3ba , Tin = 650°C and o a ion speed 110,000 pm.
52
and o Pin = 4ba he cas ing alue o 5.97MPa is also highe han he SLM alue
o 5.82MPa.
Figu e 54: S ess s X unde di e en ope a ing condi ions, 100,000 pm and
Inconel 625 (Cas ing).
Figu e 55: S ess s X unde di e en ope a ing condi ions, 100,000 pm and
Inconel 625 (SLM).
As i is done a he p e ious sec ion 4.3.1, Figu e 56 shows how he di e ence
be ween he alues o s ess be ween he cas ing manu ac u e and he SLM
manu ac u e dec ease wi h he inc emen o he p essu e and he s ess alue.
0
1
2
3
4
5
6
7
0 5 10 15 20 25 30
S ess [MPa]
X [mm]
Cas ing
550°C, 2ba
550°C, 3ba
550°C, 4ba
600°C, 2ba
600°C, 3ba
600°C, 4ba
650°C, 2ba
650°C, 3ba
650°C, 4ba
0
1
2
3
4
5
6
7
0 5 10 15 20 25 30
S ess [MPa]
X [mm]
SLM
550°C, 2ba
550°C, 3ba
550°C, 4ba
600°C, 2ba
600°C, 3ba
600°C, 4ba
650°C, 2ba
650°C, 3ba
650°C, 4ba

53
Figu e 56: Di e ence be ween he alues o Cas ing and SLM e sus he p essu e
wi h 100,000 pm.
4.3.3 S uc u al analysis esul s o a o a ion speed o 110,000 pm
The same explana ion han sec ions 4.3.1 and 4.3.2 can be used o a o a ion
speed o 110,000 pm, he plo s in Figu es 57 and 58 shows he same beha iou
agains he inc emen o p essu e and empe a u e. Howe e , o his o a ion
speed, he p essu e when he inle p essu e is 2ba and he inle empe a u e
550°C has a di e en beha iou han he o he o a ion speeds. This is because
o he highe p essu e on he blade a he inle induced by he o ex gene a ed.
The poin s wi h maximum s ess a he middle zone o he edge o his o a ion
speed a e X = 7.32mm (Cas ing 0.66MPa) and X = 9.85mm (SLM 0.72MPa) o
Pin = 2ba . Fo Pin = 3ba , he alues a e X = 7.33mm (Cas ing 1.78MPa) and X=
9.85mm (SLM 1.93MPa), and o Pin = 4ba a e X = 7.33mm (Cas ing 3.18MPa)
and X = 9.9mm (SLM 3.32MPa).
The minimum s ess o Pin = 2ba is loca ed a X = 24.3mm (Cas ing 0.35MPa)
and X = 18.5mm (SLM 0.26MPa). When Pin = 3ba , he minimum s ess poin s
a e X = 22.1mm (Cas ing 1.05MPa) and X = 21.2mm (SLM 0.89MPa) and o Pin
= 4ba he alues a e X = 22.1mm (Cas ing 1.56MPa) and X = 21.1mm (SLM
1.38MPa).
0
5
10
15
20
25
30
1.5 2 2.5 3 3.5 4 4.5
Di e ence [%]
Inle p essu e [ba ]
100,000 pm
σmax
σmax (middle zone)
σmin
54
The maximum s ess when Tin = 650°C appea s also a he end o he pa h, o
Pin = 2ba he cas ing alue is 1.06MPa and he alue o SLM is 1.04MPa. Fo
Pin = 3ba is 3.41MPa o cas ing and 3.34MPa o SLM, and o Pin = 4ba ,
4.92MPa is he maximum s ess alue o cas ing and 4.85MPa o SLM.
Figu e 57: S ess s X unde di e en ope a ing condi ions, 110,000 pm and
Inconel 625 (Cas ing).
Figu e 58: S ess s X unde di e en ope a ing condi ions, 110,000 pm and
Inconel 625 (SLM).
0
1
2
3
4
5
6
0 5 10 15 20 25 30
S ess [MPa]
X [mm]
Cas ing
550°C, 2ba
550°C, 3ba
550°C, 4ba
600°C, 2ba
600°C, 3ba
600°C, 4ba
650°C, 2ba
650°C, 3ba
650°C, 4ba
0
1
2
3
4
5
6
0 5 10 15 20 25 30
S ess [MPa]
X [mm]
SLM
550°C, 2ba
550°C, 3ba
550°C, 4ba
600°C, 2ba
600°C, 3ba
600°C, 4ba
650°C, 2ba
650°C, 3ba
650°C, 4ba
55
In Figu e 59 can be seen he di e ences be ween he alues s udied o cas ing
and SLM ollow he same end han he alues s udied in sec ions 4.3.1 and
4.3.2, bu o 110,000 pm, he di e ence o he s ess alues a highe p essu e
a e lowe han he alues ob ained a 90,000 pm and 100,000 pm.
Figu e 59: Di e ence be ween he alues o Cas ing and SLM e sus he p essu e
wi h 100,000 pm.
4.3.4 Compa ison
The s udied pa h poin s o he σmax a he middle zone in Figu e 60 shows how
he SLM manu ac u ed u bine has always highe s ess a hese poin s, and i
inc eases wi h he p essu e. The end o he s ess a his poin is o dec ease
wi h he inc emen o o a ion speed.
0
5
10
15
20
25
30
1.5 2 2.5 3 3.5 4 4.5
Di e ence [%]
Inle p essu e [ba ]
110,000 pm
σmax
σmax (middle zone)
σmin
56
Figu e 60: S ess e sus o a ion speed o σmax (middle zone).
Figu e 61 wi h σmin shows app oxima ely he same beha iou , bu wi h a
signi ican di e ence, he SLM manu ac u ed u bine has always he lowe
minimum s ess. The same beha iou can be seen in Figu e 62, whe e he
maximum s ess is plo ed o he di e en p essu es and o a ion speeds. I also
shows how he s ess o he cas ing is highe han he SLM manu ac u ed
u bine. This beha iou is due o he p oximi y o he ee su ace and he
aniso opy o he ma e ial. Mo eo e , ano he impo an hing o explain is ha he
s ess becomes also equal o bo h ypes o manu ac u es when he p essu e and
he o a ion speed inc ease.
0
0.5
1
1.5
2
2.5
3
3.5
4
85000 90000 95000 100000 105000 110000 115000
S ess [MPa]
Ro a ion speed [ pm]
σmax (middle zone)
Cas ing, 2ba
Cas ing, 3ba
Cas ing, 4ba
SLM, 2ba
SLM, 3ba
SLM, 4ba
57
Figu e 61: S ess e sus o a ion speed o σmin.
Figu e 62: S ess e sus o a ion speed o σmax.
To conclude he s uc u al analysis, sec ion 4.3, Figu e 63 shows how he
di e ence be ween he SLM manu ac u e and he cas ing manu ac u e
dec eases when he s ess inc eases. Fu he mo e, when he o a ion speed
a ises, he di e ence dec eases oge he wi h he p essu e.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
85000 90000 95000 100000 105000 110000 115000
S ess [MPa]
Ro a ion speed [ pm]
σmin
Cas ing, 2ba
Cas ing, 3ba
Cas ing, 4ba
SLM, 2ba
SLM, 3ba
SLM, 4ba
0
1
2
3
4
5
6
7
8
85000 90000 95000 100000 105000 110000 115000
S ess [MPa]
Ro a ion speed [ pm]
σmax
Cas ing, 2ba
Cas ing, 3ba
Cas ing, 4ba
SLM, 2ba
SLM, 3ba
SLM, 4ba

58
Figu e 63: Di e ence be ween cas ing and SLM esul s e sus Inle p essu e
compa ison.
0
5
10
15
20
25
30
35
1.5 2 2.5 3 3.5 4 4.5
Di e ence [%]
Inle p essu e [ba ]
σmax, 90000 pm
σmax, 100000 pm
σmax, 110000 pm
σmax (middle zone), 90000 pm
σmax (middle zone), 100000 pm
σmax (middle zone), 110000 pm
σmin, 90000 pm
σmin, 100000 pm
σmin, 110000 pm
59
5 DISCUSSION
Se e al poin s can be ex ac ed om he esul s and analysis sec ion. Fi s ly,
conce ning he o e all pe o mance o he u bine. When he inle p essu e is low
he o e all e iciency is less in luenced by he empe a u e, an inc emen o he
inle p essu e gene a es an inc ease o ou pu powe , i he empe a u e ises he
ou pu powe is highe bu he o e all e iciency dec eases. Mo eo e , when he
o a ion speed inc eases, he in low angle may change o a oid he o ex
gene a ion, because he p esence o a o ex dec eases he e iciency and he
ou pu powe .
Secondly, wi h e e ence o he ope a ing condi ions in luencing he s ess
dis ibu ion, he s ess on he blade base inc eases when he inle p essu e o he
u bine inc eases, howe e , i inc eases mo e om 2 o 3ba han om 3 o 4ba .
The inle empe a u e gene a es an inc ease o he s ess, bu i is only no iceable
a he inle pa o he u bine, whe e he empe a u es and s esses a e highe .
Fu he mo e, i a o ex appea s on he u bine blade, he s ess a he inle zone
becomes highe .
Thi dly, ega ding he s udy o he impac in oduced on a mic o- u bine due o
SLM manu ac u e. I includes he impac o he SLM manu ac u e compa ed wi h
he in es men cas ing, he adi ional manu ac u ing and widely used o mic o-
u bines manu ac u ing. The di e ence be ween bo h ypes o manu ac u ing
declines wi h he inc ease o he inle p essu e and wi h he inc ease o he
o a ion speed as i was seen in Figu e 63. The highe s ess di e ence be ween
bo h ypes o manu ac u ing is eached by σmin, ollowed by he poin wi h
maximum s ess a he cen al zone o he pa h s udied and σmax espec i ely.
Fu he mo e, he s ess a he poin wi h maximum s ess a he cen al pa h zone
is highe o he SLM manu ac u e in con adis inc ion o he σmin and σmax poin s.
This beha iou is due o he p oximi y o σmax o he ee su ace and he
aniso opy o he ma e ial. The dec ease o he σmax o SLM manu ac u e
in luences he σmin poin , gene a ing also a lowe s ess o his poin han he
cas ing manu ac u ing. The e o e, he SLM manu ac u e gene a es lowe s ess
a he inle pa o he u bine, and highe s ess a he ou le .
60
The e o e, se e al conclusions a e ex ac ed om he s udy. The join be ween
he blade and he hub is he c i ical a ea o he u bine o he wo ypes o
manu ac u ing and he u bine manu ac u ed by SLM has lowe s ess a he inle
pa han he u bine manu ac u ed by cas ing. Howe e , he s ess gene a ed a
he middle o he c i ical a ea is highe o he u bine manu ac u ed by cas ing.
This shows he necessi y o conside he ype o manu ac u ing be o e o s a he
p oduc ion, in o de o s eng hen he a eas a ec ed by he aniso opy (specially
he middle o he c i ical a ea, whe e he s ess luc ua ion is highe ). Mo eo e ,
he inc ease o he s ess in his a ea and he necessi y o ein o ce i , can c ea e
design p oblems. As well as, he eliabili y o ma e ials manu ac u ed by selec i e
lase mel ing is lowe han cas ing and he inal esul s depend a lo on he igh
s udy o he op imum condi ions ( eloci y, lase powe , wo king a mosphe e, e c.)
o he manu ac u ing p ocess. I hese condi ions a e no he op imum, he
a iabili y o he mechanical p ope ies in oduced, gene a es mo e a iabili y a
he middle o he c i ical a ea, gi ing he possibili y o appea highe s esses
whe e i was no expec ed and in consequence highe a igue.
61
6 CONCLUSIONS AND FUTURE RESEARCH
The o e all aim o his hesis was o de e mine he impac ha addi i e
manu ac u e ha e on a mic o- u bine wheel. This was assessed h ough a
Wo kbench 17.1 model, including a CFX analysis, a The mal analysis and a
S uc u al analysis o calcula e he s ess gene a ed on he u bine.
To conclude, he esul s ob ained om he Wo kbench model a e:
• The c i ical a ea is he join be ween he blade and he hub o bo h ypes
o manu ac u ing.
• I he ype o manu ac u ing is SLM, i needs o be de e mined be o e he
mic o- u bine design o a oid p oblems a he middle o he c i ical a ea
and ein o ce i .
• The s ess luc ua ion a he c i ical zone is highe o SLM and can
inc ease he s ess a igue.
• The eliabili y o ma e ials manu ac u ed by SLM is lowe allowing he
possibili y o u u e p oblems.
Howe e , he esul s o his hesis a e ob ained wi h a CFD and FEA model and
using an app oxima ion o he ma e ial p ope ies o bo h manu ac u ing
p ocesses due o he inaccessibili y o he eal da a. The e o e, a u he s udy
should be ca ied ou , s udying h ough an expe imen al me hod he eal
p ope ies and s esses gene a ed, in o de o compa e hem wi h he esul s o
he model designed in his hesis. Mo eo e , his hesis can be complemen ed by
he s udy o a ied ma e ials and se e al ypes o addi i e manu ac u ing, and
inding wha is he bes o he mic o- u bines manu ac u ing. Ano he
imp o emen can be he s udy o he ugosi y in oduced on he mic o u bine
su ace due o he addi i e manu ac u ing p ocesses, wha can in luence he
ae odynamics o he u bine.
68
Appendix C : Inconel 625 p ope ies
C.1 Mechanical p ope ies
C.2 The mal p ope ies

69
Appendix D Mesh sensi i i y esul s
D.1 CFX
D.2 The mal
D.3 S uc u al
70
Appendix E : CFX esul s
71
Appendix F : S uc u al da a esul s o 90,000 pm om
ANSYS S uc u al (example)
The able below has he alues o he s ess on he pa h a he conca e pa o
he blade ( on ), wha was s udied a he esul s sec ion, and he con ex pa o
he blade (Back).
72
73

74
75
76
77