Lukas Latzke

BEng Aeronautical & Mechanical Engineering
BEng Peirianneg Awyrennol A Mecanyddol

Numerical Design Optimisation of the Flettner Rotor by Using the Ansys CFD Software.

Optimeiddio Dylunio Niwclear y Rotor Flettner drwy Ddefnyddio Meddalwedd Ansys CFD.

Maritime transport emits 940 million tonnes of CO2 per year and is responsible for around 2.5% of global greenhouse gas (GHG) emissions. As this is a significant part of global greenhouse gas emissions, it needs to be reduced. One possibility is to support the propulsion of ships using alternative propulsion systems. An idea that was explored in the 1920s, which powered a converted sailing ship for the first time, seems to be very innovative for this purpose. Here, the so-called Flettner rotors, which act like a sail, generate thrust from the wind that impacts them. These rotors are named after their inventor Anton Flettner (1885-1961), who identified their practical application on ships. The difference to sails is that large rotating cylinders utilise an aerodynamic phenomenon called the Magnus effect, named after the German physicist Heinrich Gustav Magnus (1802-1870).
This study deals with the numerical design optimisation of the Flettner rotor using the computational fluid dynamics (CFD) software ANSYS. Initially, publications that have already examined this topic and provide experimental and numerical data that can be used to validate and verify the CFD outcome were evaluated. Furthermore, selected existing ships equipped with Flettner rotors were studied and discussed. Then, the dimensions of the rotors and the wind velocities at sea were investigated. For further simulations, a diameter that can be used for maritime applications has been researched, which requires low rotational speeds, as these are the most suitable. Therefore, investigations of how the rotational speed, wind speed and diameter of the rotor affect CL, CD and efficiency (CL/CD) at a fixed height were carried out. Once a suitable diameter has been found, it is fixed for the further simulations as the spin ratio, aspect ratio, and endplate diameter are with respect to the rotor diameter. Because an endplate on top of the rotor can lead to an increase in efficiency, the effect of this endplate on efficiency has also been researched in this study. The study's outcome is to determine the best possible geometry and speed range of the Flettner rotor for practical use at sea under good wind conditions.
The analysis of the selected vessels revealed that the height of the rotors range from 15.6 m – 27 m and that the diameters are from 2.8 m – 4 m, resulting in an aspect ratio of 4.3 - 6.8.
The wind speeds at sea were researched using wind velocity maps and the Beauford scale and were defined as 5 knots, 10 knots, 15 knots, and 20 knots, corresponding to a force of 2 - 5 on the
Lukas Latzke
ii
Beaufort scale. For the numerical setup, a publication was used that conducted a 2D study of the Flettner rotor and whose style approach was adopted and compared with the experimental data this paper obtained. The style approach of the domain size for the 3D setup was taken from another publication. For verification, a mesh dependency study was conducted, which achieved good results. To determine the best-suited turbulence models, a turbulence model dependency study was performed comparing the turbulence model ?−? -SST and ?−?. The turbulence model ?−? -SST correlated very well with the experimental values and was therefore used. The diameter of the 2D rotor for validation was 0.08 m, and the rotational speeds ranged from 0-3008 rpm. The wind velocities were 11 m/s, 13 m/s, and 15 m/s. The Reynolds number for the validation case were between ??=1.5×105 and ??=2.14×105. The 3D simulations were conducted with diameters of 2 m, 3 m, and 4 m and also wind velocities of 5 knots, 10 knots, 15 knots, and 20 knots. The height of the rotor was initially fixed to 10m. The Reynolds number here ranged from ??=3.47×105 to ??=2.78×106. It was found that high wind velocities and small rotor diameters require high rotational speeds and vice versa. The diameter of 3.145m proved to be the most suitable. Subsequently, using this diameter and a wind speed of 20 knots, simulations were carried out for the aspect ratio AR=2, AR=4, AR=6, and AR=8, and also with endplate diameters of de/d=1, de/d=2, and de/d=3. This demonstrated that the endplate and generally the aspect ratio increased the efficiency. The best setup for 20 knots was: 3.145 m diameter, speed of 187.31 rpm, the height of 25.16 m (AR=8) and an endplate diameter 9.435 m (de/d=3).

Mae trafnidiaeth forol yn gollwng 940 miliwn tunnell o CO2 y flwyddyn ac mae'n gyfrifol am tua 2.5% o allyriadau nwyon tŷ gwydr byd-eang. Gan fod hyn yn rhan sylweddol o allyriadau nwyon tŷ gwydr byd-eang, mae angen ei leihau. Un posibilrwydd yw cefnogi gwthio llongau gan ddefnyddio systemau gwthio amgen. Mae'n ymddangos bod syniad a archwiliwyd yn y 1920au, a oedd yn pweru llong hwylio wedi'i haddasu am y tro cyntaf, yn arloesol iawn i'r diben hwn. Yma, mae'r rotorau Flettner, fel maen nhw’n cael eu galw, sy'n gweithredu fel hwyl, yn creu gwthiad gan y gwynt sy'n effeithio arnynt. Mae’r rotorau hyn wedi cael eu henwi ar ôl eu dyfeisiwr Anton Flettner (1885-1961), a nododd eu defnydd ymarferol ar longau. Y gwahaniaeth i hwyliau yw bod silindrau cylchdro mawr yn defnyddio ffenomenon aerodynamig o'r enw effaith Magnus, a enwwyd ar ôl y ffisegydd Almaenaidd Heinrich Gustav Magnus (1802-1870). Mae'r astudiaeth hon yn ymdrin ag optimeiddio dylunio rhifiadol rotor Flettner gan ddefnyddio'r feddalwedd dynameg hylif cyfrifiadurol (CFD) ANSYS. I ddechrau, gwerthuswyd cyhoeddiadau sydd eisoes wedi archwilio'r pwnc hwn ac sy'n darparu data arbrofol a rhifiadol y gellir eu defnyddio i ddilysu a gwirio canlyniad y CFD. At hynny, astudiwyd a thrafodwyd y llongau dethol presennol gyda rotorau Flettner. Yna, ymchwiliwyd i ddimensiynau'r rotorau a chyflymder y gwynt yn y môr. Ar gyfer efelychiadau pellach, ymchwiliwyd i ddiamedr y gellir ei ddefnyddio ar gyfer defnydd morol, sy'n gofyn am gyflymder cylchdro isel, gan mai'r rhain yw'r rhai mwyaf addas. Felly, cynhaliwyd ymchwiliadau i sut mae cyflymder cylchdro, cyflymder gwynt a diamedr y rotor yn effeithio ar CL, CD ac effeithlonrwydd (CL/CD) ar uchder sefydlog. Unwaith y bydd diamedr addas wedi'i ganfod, mae'n cael ei osod ar gyfer yr efelychiadau pellach fel mae’r gymhareb sbin, y gymhareb agwedd, a diamedr y terfynblat mewn perthynas â'r diamedr pydru. Oherwydd y gall y terfynblat ar ben y rotor arwain at gynnydd mewn effeithlonrwydd, ymchwiliwyd hefyd i effaith y terfynblat hwn ar effeithlonrwydd yn yr astudiaeth hon. Canlyniad yr astudiaeth yw pennu'r geometreg a chyflymder gorau posibl ar gyfer y rotor Flettner i'w ddefnyddio'n ymarferol yn y môr o dan amodau gwynt da. Datgelodd y dadansoddiad o'r llongau a ddewiswyd fod uchder y rotorau yn amrywio o 15.6 m – 27 m a bod y diamedr rhwng 2.8 m a 4 m, gan arwain at gymhareb agwedd o 4.3 - 6.8. Ymchwiliwyd i gyflymder y gwynt yn y môr gan ddefnyddio mapiau cyflymder gwynt a graddfa Beaufort ac fe'u diffiniwyd fel 5 not, 10 not, 15 not, ac 20 not, sy'n cyfateb i rym o 2 - 5 ar raddfa Lukas Latzke ii Beaufort. Ar gyfer y gosodiad rhifiadol, defnyddiwyd cyhoeddiad a gynhaliodd astudiaeth 2D o'r rotor Flettner a mabwysiadwyd ei arddull ac fe’i cymharwyd â'r data arbrofol a gafwyd gan y papur hwn. Cafwyd yr ymagwedd arddull maint y parth ar gyfer y gosodiad 3D o gyhoeddiad arall. Er mwyn dilysu, cynhaliwyd astudiaeth ddibyniaeth rhwyll, a gyflawnodd ganlyniadau da. Er mwyn pennu'r modelau tyrfedd mwyaf addas, perfformiwyd astudiaeth ddibyniaeth model tyrfedd yn cymharu'r model tyrfedd k−ω -SST a k−ε. Roedd y model tyrfedd k−ω -SST yn cydberthyn yn dda iawn â'r gwerthoedd arbrofol ac felly fe'i defnyddiwyd. Diamedr y rotor 2D i'w ddilysu oedd 0.08 m, ac roedd y cyflymderau cylchdro yn amrywio o 0-3008 rpm. Cyflymder y gwynt oedd 11 m/s, 13 m/s, a 15 m/s. Roedd y rhif Reynolds ar gyfer yr achos dilysu rhwng ?? =1.5×105 a ??=2.14×105. Cynhaliwyd yr efelychiadau 3D gyda diamedr o 2 m, 3 m, a 4 m a hefyd cyflymder gwynt o 5 not, 10 not, 15 not, ac 20 not. Roedd uchder y rotor wedi'i osod i ddechrau i 10m. Roedd rhif Reynolds yma yn amrywio o ??=3.47×105 i ??=2.78×106. Canfuwyd bod angen cyflymderau cylchdro uchel ac i'r gwrthwyneb ar gyflymder gwynt uchel a diamedr rotor bach. Y diamedr o 3.145m oedd y mwyaf addas. Yn dilyn hynny, gan ddefnyddio'r diamedr hwn a chyflymder gwynt o 20 not, cynhaliwyd efelychiadau ar gyfer y gymhareb agwedd AR=2, AR=4, AR=6, ac AR=8, a hefyd gyda diamedr o de/d=1, de/d=2, a de/d=3. Dangosodd hyn fod y gymhareb terfynblat ac agwedd yn gyffredinol yn cynyddu effeithlonrwydd. Y gosodiad gorau ar gyfer 20 not oedd: 3.145 m diamedr, cyflymder o 187.31 rpm, uchder o 25.16 m (AR=8) a diamedr di-densiwn 9.435 m (de/d=3).