‘pesiveica va nese, (weg SECRETPAIR DA ENC TENOLOGA INSTITUTO NACIONAL DE PESQUISAS ESPACIAIS INPE-5275-PRE/1689 MATERIALS FOR FRICTION DRAG REDUCTION THIN FLLMS Francisco Leme Galvao INPE SA0 José dos Campos Junho de 1991MATERIALS FOR FRICTION DRAG REDUCTION UBIN FILMS Francisco Lene Galvic Instituto Wacional de Pesquisas Espaciais - INPE c.P, S15 - CEP 12241 ~ 5, JOSE DOS CAMPOS - SP - BRASIL somary ‘This paper considers that friction azag reductions, in aersepace vehicles, could result fron the use of external surfaces thin film coverings with atomic level smocthness. Tho dynamics of the gas-solia surface interactions, the edeorption phonorena, and the effects of the surface atomic structure on adsorption foros Giscussed in order to identify th adequate materials for. euch application. It is concluded that some heavy transition metal nalides and dichalcogenides, with "van der” Waals" layered crystalline stractures, represent an adequate starting point for furthor research. 1. INTRODUCTION ‘The scattering of molecular beans praducad by single crystal surfaces in vaculim, show in most “cages, velocity distripations having zero mean values of the scattered velocity tancential component. However, in sone cases, such as He scattering by Liror pd ii “surfaces (11, most gas molecules ere elastically’ scattered resulting ina non zero nean value. Also in the last decades, the Electronic device technolocy brought with it new ‘thin film manufacturing processes (21, which permit to. extrapolate for the near future ‘the possibility of obtaining large surfaces with atomic level smoothness. These considerations suggest, _new possibilities of friction drag reductions in aerospace venicles [3], that vill be conditioned to: @) The determination of the .tnin film material properties, required ‘to. obtain the elastic scattering of air molecules. b) The development of nanufacturing ana application methods, of large atomically smooth thin film surfaces made with these materiale. Considering the benefits to be obtained 1a terms of airplane fuel econonies, and lov crbit satellite life-tines, the’ present work try to investigate the” first condition (a), ao thet with farther Research on '(b), the predicted drag reductions can be Gxperinentally verified, FRICTION DRAG in aercaynamics, tne component in tne flow Birection of the resultent of all tangential to body surface forces, that is of friction forces, is. denominated Eriction dreg. Fox elongated end “etweamlinea" bodies such ae airplanes, the friction drag can be responsible, at cricing speed _conditions, for the major part (arcund G08) of the total drag and so of the fuel consunption- Even for rather blunt bodies, like satellites, when ina low orbit, (below 700 Kn), and equipped with large surface solar panels, the friction draz_ may se Bignigicative, 10” to 208 of the drag, and fo contribute to their orbit decay. Ip continuum fluid mechanics, that is knadsen Wumbers KX <'¢ 1, tha friction forces can be computed fOr various flows regimes, using the "Boundary Layer theory" [a], which assunes that the relative fluid velocity goes to zero at the surface (zero §lip condition, see Fig. 1) and the known "Newton" equation: teu. avy where « is the tangential shear stress, aV/dy the velocity gradient at the surface normal direction and u is the viscosity. tb) a) rero slip tiow b) flow with siip Fig. 1 ~ Boundary Layer Profiles.For rarefied fluid dynanics (KW > 1), the friction torces are computed as function of the estinated tangential nonentun acconmodation coefricient values, which in Satellite aerodynamics practical drag estinations [5] are usually assumed ag unity, that ist a complete accommodation. Tt must be emphasized that in both cases, either for the boundary layer zero slip Gr for the unit accomodation coefficient, 2 diffuse reflection of the fiuia by th Surface, with sero. mean value for the tangential reflected velocities of the Fluid molocules, is implicitly ssouned. Any possible deviation of towards a more elastic lobslar or "specular" reflection (see Pig. 2) will modify thess assumptions and result in lower friction force valve: thie condition a) diffuse reflection bh) Lobular reflection Fig. 2 = Nolecular Beam Scattering 3, DYNAMIC INTERACTIONS . 3.1 Mass Effects: Although the dynamic interactions between atoms and nolecules and solid” surfaces have been extensively studied (6] this is @ rather complex subject, which explains the ditterent theoretical models end some conflicting experimental results found. The classical mechanics models, such as “hard cubes", "soft cubes", etc, suppose & zero tangential monentun transfer tnd. & Sinpie and accurate model giving the Ueorecical conpstation of the tangential monentum accomodation, seen not to be avaliable, Eleborate numerical simulations [7] and Gissipative quentun phenomena (8), such as electron hole pair, ete., are to be considered to obtain quantitative results. A simple, but crude approach is the use of the classic “Baule" rigic soheres model 1], which allow the evaluation of the enetgy accommodation coefficient as, function of gas temperature and of the gas and solid atomic masses. to any tangential to normai, and nelazional to” rotational, somenvun transfers during ges molecules “surface collisions; « low energy acconmedation do not implies ina iow tangontial accomodation, but nevertheless, if is a secescary condition for it. ‘The “Banle" that in order to increase the elastic reflections, the solid surface atoms must ba much’ more heavier then the gac atone, and molecule: and it explains elastic scattering reculte obtained in vacuum, for the keliun on Gifferent single exyetal curfaces. model shows ‘The effects of the atonic Layers below the surface ic nore evolved and is related to the "phonon" shavacteriaties of the solid substratum. An extension of the previous nechanical reasoning, indicates that heavier inner xtons, in’ special if strongly bonded to Surface atone, vill, contribute to the reduction of phonon” excitations and to reduce the energy acconmodation. ‘The conclusion is that for aerodynamic @rag reduction purposes, only materials with heavy surfece atons and high specific Gonaities shall be investigated, 3.2 Contamination Effects Another very important aspect. which cones from the study of the dynamic interactions is the effect of the presence of adsorsed species on the surface. In, scattering experinents with elastic reflections, the contamination of the surface, in¢luding by the adsorption of jow energy gas atoms and nolecules, lead to the reduction in the intensity of’ these elastic reflections. In all practical applications at normal atmospheric conditions the surface contamination by the air constituents, including water molecules and by solid dust particle cf all sizes and shapes, is unavoidable. A partial solution to this problem, seems to be the reduction of the surface adsorption forces. ihe reasoning is thatu t he recucticn of adsorption forces and of heir tangential component, will reduce Ne contamination level, and also result n loosely adsorbed species that could esily "slip" along the surface, with less momentum exchange. 4 z e ADSORPTION FORCES +1 Surface Energy SURFACE POTENTIAL x MELTING POINT weTaLIC ELEMENTS. ny MELTING POINT 20 (1000 deg ) 30 Pig. 3 ~ Surface potential and melting point correlation. HeP (oon) Fee (oon) sectoon Pig. 4 ~ Analyzed adsorption sites surfsce geonetri Most research in which the surface enercy is a key property, is usually related to technological processes such as catalysis, painting, bonding, etc. and have opposed Dbjectives to that of the present vork,, that is, the search how to increase the adsorption. forces. A strong correlation exists between the Surface energy and the melting heat, and Between the surface potential ard! the nelzing point of materials, az shown in Fig 3, for the rface potential of the metallic elements [2]. ome layered orystsline materials have Saly "an der Waals" bonds acting between their layers, and for which surfaces, the furface energy levels are much lower than those of otherwise oriented surfaces. ‘These "van der Walls" bonds being function of atons and molecules polari: fre weaker for non polar and low polarizability substrata. 4.2 Surface Structure In order to verify the effects of the Surface atomic structure on the adsorotion “rorces and on its tangential component, a simplified numerical model vas elaborated, Using substratun atoms fixed at lattice Positions (neglecting the thermal notion). A "6-12 Lenard Jones" [9] type interaction Potential vas assumed to act between these Substratum atone and a surface adsorbed atom, and the resultant force was computed by simple summation, what is acceptable for weak interaction: HCP, FCC, BCC and Thetraearal solids, vithout any surface reconstruction (Fig. 4) and “three different ratios ot tne external ("RO"), to the internal ("a") equilibrium distance, have —_been Investigated (RO/a = 0.7, 1,0 and 1.4), From the results, some of vhich are shows in Fig 5 to 6, it was concluded that, for ‘the same interaction potential parameters: has @ minor influence component. of the and soon the = The crystal type on the nommal \daorption force adsorption enersy- - The tangential component of acsorption force and so the enorsy moving the adsorbed. atom slong surface, are greatly Influenced by the irince etructure. They are lower for tome surface directions and minimized by Glosed packed (HCPOO01) arrangements, the for the—LarTice atoms ~- ADSORBED ATOMS 5 ° 5 a) RO/a = 047 >) RO/a = 1.0, c} RO/a = 164 Fig. 5 = Adsorption force normal. components. Os 10 Fig. 6 = Adsorption force tangential component and equilibriom Position above surface: da 5. concLUsroNs Xt is concluded that, for aerodynamic friction drag reduction purposes, che Rateriels to be choosed for further thin fils menufacturing research; shall. have, besides obviously a chemical stability in respect to air constituents: - High atonie and molecular weight. = Low surface enercy, that Le: low melting or cublimating | temperature, lew polarizability, and "Van der’ Waals" Ynyerea structire. = Closed packed atoms surface structure. These properties are easy to be alone, but fev materials combine all as is the case. for sone halides and di- chaleogenides of heavy transition metals. Examples are the Bismth Tri-iodide and the Tungsten Di-selenice, but the most natural candidate for research is the Tungsten Divsulfide (nolibdenite), well known by its solid lubricant properties. To extend the present single crystal epitaxial grow laboratory technologies, Using these low adherent naterials (10) U1], to large engineering surface coating processes, undoubting will represent, a fornidable nolecular encineering defy. 6, BIBLIOGRAPEY 1, ZANGWILL Ay; "Physics at surfaces"; Cambridge, G.8.; Cambridge University Press, 1588, 2. LEWIS, 8, and ANDERSON, §—_J.Cuz “Nucleation ana Grow of Thin Filme"; Wow York, U.S.A. Academic Fress, 1976. 3. GALVAO, Files "Friction brag Reduction"; “ENPE=4278-PRE/1150; _INPE, SG0 José dos Campos, BR.; INPE, 1987. 4. scunrcurimc, my "Boundary ayer Thoory"; New York U.8.A.7 Ne Graw HilL International Book, i968, 5, SCETCHER, R.D.; "Satellite xezeay- namice"; ‘Apostila do Curse Interno; 520 Joes dos Campos, BR.y INPE, 1983. "pynamice Nownvork, 7, TOLIY, J,C.2 "Simulation of Gas Surface dynamics* in: “Many Body Phenomena at Surfaces: Orlando, U.S.A.: Academic Press, 1984.WUNDQUIST, B, I.; "Electronic Damping 10, UENO, K.;_ SAIKI, K.; SINADA, 7. and Of Adsorrate Motion on Metal Surfaces” KOMA, Aj "Epitaxial Grow of Transition int Many Body Phenomena at Surfaces, Metal 'Dichalcogenides on Cleaved Orlando; U,S.A.; Academic Press, 1964, Surfaces of Mica"; Journal of Vacuum Science Tecnology, AS (1), Jan/Fev 30. 9. ASHKROFT , N.W, and MBRMIN, N.D¥7 “solid State Physics"; Holt Rineh and Wilson, 1976, HE 11. JAGER A, 6 all, "W Se? Tm: Prepared by Soft selenizavion' Solid Films, Vol 200, n° ly Nay 12936