Aerospace Engineering

International Review of

Aerospace Engineering (IREASE)

Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved

Contents Corrugated Limiting Tab for Supersonic Jet Mixing by Anuj Bajpai, Ethirajan Rathakrishnan

180

Influence of Simulated Space Hazards on Polyimide ArtilonTM Type Used in Space Applications by A. M. Anwar, M. M. Osman

195

Ascent Phase Trajectory Optimization of Launch Vehicle Using Theta-Particle Swarm Optimization with Different Thrust Scenarios by Dileep M. V., Surekha Kamath, Vishnu G. Nair

200

Two Dimensional Numerical Study of Aerodynamic Characteristic for Rotating Cylinder at High Reynolds Number by Alias M. S., Mohd Rafie A. S, Wiriadidjaja S.

208

Design of High Temperature Six-Phase Starter-Generator Embedded in Aerospace Engine by Flur Ismagilov, Vavilov Vyacheslav, Lubov' Roginskaya, Semen Shapiro, Denis Gusakov

216

Integrated Electrical Machines with Permanent Magnets for Aerospace Industry by Ismagilov F., Roginskaya L., Shapiro S., Vavilov V., Karimov R., Ayguzina V.

226

International Review of Aerospace Engineering (I.RE.AS.E), Vol. 9, N. 6 ISSN 1973-7459 December 2016

Corrugated Limiting Tab for Supersonic Jet Mixing Anuj Bajpai, Ethirajan Rathakrishnan

Abstract – The mixing efficiency of limiting tabs of three geometries with triangular corrugations, in promoting the mixing of a Mach 2 elliptic jet has been studied. Limiting tabs of rectangular, triangular and circular-arc geometries with 5% blockage are placed along major and minor axis at the nozzle exit are tested for nozzle pressure ratio from 4 to 8 in steps of one. All the tabs are found to be efficient mixing promoter at all tested NPRs, when placed along major axis at the nozzle exit. However at NPR 4 and 5 circular-arc tab along minor axis is found to retard the near field jet mixing. But the corrugated rectangular tab along minor axis is found to be the best mixing promoter among tested geometries. It causes a maximum reduction of 89% in core length at NPR 6. The corresponding core length reduction for corrugated triangular and circulararc tab is 78% and 84% respectively. The maximum reduction in core length obtained for corrugated triangular and circular arc tab is found when placed along major axis at the exit of elliptical nozzle. The shadowgraph pictures of the jet reveal that the waves prevailing in the elliptic jet controlled with corrugated limiting tabs along major axis are significantly weaker than those in uncontrolled and controlled (with tabs along minor axis) elliptic jet. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Supersonic, Limiting Tab, Elliptic Jet, Tab Geometry, Expansion Level

Jet control techniques may broadly be classified into active and passive. Micro jets [5]-[8], fluid tabs [9]-[16], [17]-[21] used for jet mixing modification are active controls. Unlike active control no auxiliary power is required for passive control. Passive control relies on geometrical modifications to control the flow and noise characteristics. Tabs placed at the exit of axisymmetric and asymmetric nozzle, to achieve desired alteration in flow properties has been studied by many researchers in past. These methods focus mainly on altering the boundary layer at the nozzle exit, thereby modifying the shear layer growth and flow behavior. Tab is a solid strip kept normal to the flow at the nozzle exit. A tab generates pairs of counter rotating small-scale vortices all along its length. These span-wise vortices become stream-wise, soon after leaving the tab. Bradbury and Khadem [22] were the first to study the effect of tab on low speed jet mixing. Tabs were found to increase the jet mixing, leading to rapid centerline decay. Ahuja and brown [23] studied low supersonic jet control with two tabs at the exit. The tabs resulted in the reduction of core length and low-frequency noise of the jet up to 5 or 6 dB. The mixing enhancement due to tabs reduced the centerline temperature also. It was observed that the jet development is not influenced by parameters such as turbulence level and boundary layer thickness. Samimy et al. [24], [25] studied the evolution of axisymmetric jet control with tabs, over the Mach number range from 0.3 to 1.81. Entrainment caused by the streamwise rotating

Nomenclature NPR D pt p0 pe pb pa X Y Z

Nozzle pressure ratio Equivalent diameter of the nozzle exit Pitot pressure Settling chamber pressure Nozzle exit pressure Back pressure Ambient pressure Coordinate along the jet axis Coordinate along the minor axis Coordinate along the major axis

I.

Introduction

Enhancement high-speed jet mixing is important in many engineering applications. Increased rate of mixing of fuel and oxidizer is essential in combustion devices. To address this issue of mixing, free jet mixing has been studied with passive and active controls have been studied. Among these the passive control in the form of tabs has gained momentum owing to its simple nature. Ramjet is a typical example demanding rapid mixing of fuel and air for its efficiency. To minimize the size of combustion chamber entire mixing has to be accomplished in a very small distance [1]-[2]. Jet control aims at modifying the aerodynamic mixing and aero acoustic characteristics. Gutmark [1], and Grinstein [2], Seiner et al. [3], and, Knowles and Saddington [4], carried out a detailed study on techniques used for enhancement of mixing in jets.


DOI: 10.15866/irease.v9i6.10364

180

Anuj Bajpai, Ethirajan Rathakrishnan

vortices was visualized by Zaman et al. [26]. It was found that a tab shed pairs of counter rotating streamwise vortices, all along its sides [27]-[29]. The sense of rotation of the vortex pair is such that fluid from the center of the tab base to its tip. It was envisaged that the combined effect of the pressure-hill formed at the tab face and the presence of wall produces the pair of counter rotating vortices [29], [30]. The vortices shed by a tab have their axis parallel to the tab edge and rendered stream-wise by the inertia of the flow. Thus, if the tab is tilted downstream, vorticity from primary and secondary sources add together, improving the tab effectiveness. Zaman [29], Bohl and Foss [27], reported two regions of concentrated vorticity in the flow on each side of the tab, which is due to the pressure-hill. Stefan et al. [31] made a comparison between experimental and computational results and found good agreement between the two in terms of the vorticity field and entrainment. Further, postulations had been made on basic flow dynamics that manipulation of the size of vortices shed by the tab plays a dominant role in the mixing of free jets [29], [30]. These mass entraining large-scale eddies formed at the jet boundary and the mass transporting small-scale eddies shed from the tab should be in proper proportion for an efficient mixing of the jet. The findings of Reeder and Samimy [32], regarding the optimal tab placement and tab shape have shown that the nozzle exit in the best location for tab. Also, the tab height should not exceed the boundary layer thickness for generation of streamwise vortices effectively. Navin Kumar and Rathakrishnan [33] reported that width of the tab more effective than the length for enhancing jet mixing. Apart from influence of tab geometry on mixing efficiency they also reported that the presence of streamwise vortices right up to the jet centerline might help in enhancing mixing to the extent of about 80%. Based on these postulations Sreejith and Rathakrishnan [34] studied the effectiveness of a wire running across diameter (cross wire) as passive control on jet mixing. They found that the mixing promoting vortices shed by the wire, all along a diameter at the nozzle exit, is efficient in mixing promotion. The waves in the jet core were weakened considerably by the use of crosswire at the nozzle exit. Further, the cross wire is found to be effective at all levels of expansion. Core length reduction of about 50% is found for Mach 1.79 jet at NPR 5.66. This work demonstrated that the limit for tab length is nozzle exit radius and not the boundary layer thickness. This limit of tab length is called Rathakrishnan limit [35]. Many studies revealed that limiting tabs is effective in enhancing mixing [36], [37]. But the cross wires studies reported in literature is only with circular geometry. Therefore, to gain an understanding about the effectiveness of limiting tab of non-circular geometry, limiting tabs of noncircular shapes placed along at the exit of a Mach 2 elliptical nozzle is studied in this work.

From literature it is evident that vortex size manipulation plays a dominant role in controlling the mixing of the fluid entrained at the jet boundary with the jet fluid. From vortex theory, it is known that small-scale vortices are efficient mixing promoters. That too if the small-scale vortices are of mixed size the environment they establish for mixing promotion becomes better. In order to establish this desired environment of generating mixing promoting small-scale vortices of mixed size, corrugations could be provided along the tab edges. With this aim, limiting tabs of rectangular, triangular and circular-arc geometries, with 5% blockage, placed along major and minor axis, at the exit, of a Mach 2 elliptic nozzle are tested in the presence of different levels of expansion corresponding to nozzle pressure ratios from 4 to 8, in steps of one. Pitot pressure variation along the jet centerline, pressure profiles in the directions along and normal to the tabs, at different axial locations were taken. The waves prevailing in the jet core were visualized with shadowgraph technique. The results of the controlled elliptic jet are compared with that of uncontrolled elliptic jet [38].

II.

Experimental Details and Procedure

The experiments of this study have been conducted in the jet test facility, shown in Fig. 1. Compressed air from the storage tank was supplied to the settling chamber through a control valve. The settling chamber total pressure, p0, was maintained constant during a run, by controlling the pressureregulating valve. The settling chamber Temperature, T0, was the same as the ambient temperature, Ta, and the backpressure, pb, was the pressure of the ambient, pa, to which the jets were discharged. II.1.

Experimental Model

A Mach 2 elliptical convergent-divergent nozzle made of brass is used in this work. The equivalent throat and exit diameters of the nozzle are 10 mm and 13 mm, respectively. The Reynolds number of the Mach 2 jet, at NPR 4 and 8, based on nozzle exit diameter, are 7.12 × 105 and 16.6 × 105, respectively. The pitot pressure measured at different points over the nozzle exit plane was reduced to Mach number, using the normal shock relation, which relates the ratio of pitot pressure ahead of and behind the shock at the nose of the pitot probe. The Mach number over the nozzle exit plane is found to be uniform within ±2%. Corrugated rectangular and triangular limiting tabs made of brass strip and circular-arc tab made from hypodermic needle were used in this study. The geometries of tabs used are given in Fig. 2(b), and the dimensions of converging diverging elliptical nozzle are given in Fig. 2(c).


International Review of Aerospace Engineering, Vol. 9, N. 6

181


Fig. 1. Schematic diagram of jet facility

Fig. 2(a). Views of the elliptical nozzle [38]

Corrugated rectangular tab for minor axis

Corrugated rectangular tab for major axis

Corrugated triangular tab for minor axis

Corrugated arc tab for minor axis

Corrugated arc tab for major axis

Corrugated triangular tab for major axis

Fig. 2(b). Schematic of tab geometries (not to scale)



182

Anuj Bajpai, Ethirajan Rathak Rathakrishnan rishnan

Fig. 2(c). Nozzle dimensions (in mm) [38]

II.2.

Instrumentation

all the NPRs studied. The movement of the pitot probe mounted on the traverse had a resolution of ±0.1 mm in the linear translation. The repeatability of the pressure measurements was within ±3%.

Pitot pressure variation along the jet centerline, and in the direction along and normal to the tab, at specified axial locations were measured with a 16 channel Pressure Systems, Inc. 9010 transducer with a range of 0 - 2.1 MPa. The accuracy of the transducer (after re zero calibration) is spe specified cified to be be ±0.15% full scale. The pressure measurements in the jet flow field was done using a pitot tube of 0.4 mm inner diameter and 0.6 mm outer diameter, mounted on a rigid three threedimensional traverse; with a resolution of 0.1 mm in linear translation. In all measurements, measurements, the pitot probe stem was kept normal to the jet axis with its sensing hole facing the flow. T The he pressure measured is the mean pitot pressure, which is the ave average rage of 250 samples per second. The waves prevailing in the supersonic jet core were vvisualized isualized using a shadowgraph system with a helium spark arc light source in conjunction with a concave mirror. The shadowgraph images were recorded using a still camera. II.3. II

III. Results and Discussion The measured data consists of the pitot pressure variation along the jet axis (X (X-direction) direction) and in the directions along major (Z (Z-direction) direction) and minor axis (Y (Y-direction) of the elliptic convergent convergent--divergent divergent nozzle of aspect ratio 2. The measured pressures are presented as such, in the nonnon-dimensional dimensional form, because calculating Mach number from the measured pressure is not possible because of the compression and expansion wa wave ve prevailing in the jet field. One of the serious difficulties associated with the analysis of the experimental data of supersonic jets to quantify the core length, characteristic decay and far far-field field decay is that the measured pitot pressure cannot be deduced to Mach number (or) velocity. This is because a supersonic jet core is wave dominated, even at correctly expanded condition. Because of the presence of compression and expansion waves of varying strength, neither the pitot nor the static pressure is constant across the jet in the core region.

Data Accuracy

The settling chamber pressure during the experiments of present investigation investigation was maintained within ±2%, for



183


In addition to this, because of the crossing of waves from the edges of periphery from one end to another end, the flow encounters mixed Mach numbers made up of subsonic and supersonic values. Thus, use of isentropic pressure-Mach number relation or normal shock relation or Rayleigh pitot formula, for calculating the Mach number distribution in the jet flow field is ruled out. Therefore, it is a usual practice to use the measured pitot pressure, which is essentially the total pressure behind the normal shock at the nose of the pitot probe in non-dimensional form, by dividing with the settling chamber pressure with which the jet is run, to discern the details of the core length, rate of characteristic decay, and the location where the jet becomes fully developed.

found to be the best among the corrugated geometries tested, causing a core length reduction of as high as 89%. Centerline decay results for NPR 7, which is a marginally overexpanded state for the present jet, are shown in Fig. 3(d). It is seen that, with decrease in level of adverse pressure gradient, the waves in uncontrolled jet becomes stronger and shock cells become longer. Also, the uncontrolled jet tends to become fully developed only beyond 17D. The core length extends up to 7.5D. For this NPR, tab along minor axis promotes the near field mixing better than tab along major axis, which is different from the results for lower NPRs. Tab along major axis reduces the core form 7.5D to 1.15D which amounts to 84% reduction in core length. For tab along major axis the jet becomes fully developed as early as about 9D. Whereas for tab along minor axis though the near field mixing is significant, the far field mixing is found to be inferior that of tab along major axis. For tab along minor axis the jet becomes fully developed only beyond 15D. Thus at NPR 7 the tab along major axis is a superior mixing promoter than the tab along minor axis both in near and far field. The centerline decay results for NPR 8, which is a marginally underexpanded case with a mild favorable pressure gradient at the nozzle exit, are shown in Figure 3(e). It is seen that the waves in jet core is stronger than at NPR 7. But the core length is comparable to that of NPR 7. At NPR 8 the near field mixing caused by the tab along minor axis is found to be considerably higher than tab along major axis, but in far field the tab along major axis is found to be superior mixing promoter that the tab along the minor axis. The core length for uncontrolled jet, jet controlled with tab along major and minor axis are 8D, 1.3D and 0.9D, respectively. Thus in the presence of marginal favorable pressure gradient, the tab along minor axis causes a reduction of 88% whereas core length reduction caused by tab along major axis is only 84%. The core length variation for the controlled and uncontrolled jet as function of NPR is shown in Fig. 3(f). For uncontrolled jet the core increases monotonically with NPR. However, this increase in core length is steep for NPR 4 to 6 and becomes shallower for NPR range 6 to 8. For tab along minor axis the core decreases appreciably for NPR range from 4 to 6 and increases from NPR 6 to 7 and once again comes down for NPR 7 to 8. For tab along major axis the effect of NPR on the core length reduction is found to be marginal. The centerline decay results for circular-arc tab with concave face facing the jet, at NPRs 4, 7 and 8 are shown in Figs. 4(a)-(c). In the presence of an adverse pressure gradient of about 0.51% (NPR 4), the arc tab along minor axis is found to be an excellent mixing promoter leading to the diffusion of waves in jet core. But the core for this orientation is slightly longer than the core for tab along major axis. The far field mixing caused by tab along minor axis is better than tab along major axis. Also, for tab along minor axis jet becomes fully developed at 5D itself, whereas for tab along major axis jet becomes fully developed only beyond 10D.

III.1. Centerline Pressure Decay (CPD) The measured pitot pressure pt variation along the jet axis (X-direction) is non-dimensionalised with the settling chamber pressure p0 and plotted as a function of non-dimensional axial distance X/D. Figs. 3(a)-3(e) compare the centerline pressure decay for corrugated rectangular limiting tab at nozzle the nozzle exit, for pressure ratios from 4 to 8, in steps of one. In Fig. 3(a) centerline pressure decay of uncontrolled and controlled jet, at NPR 4, which is an overexpanded state, with an overexpansion level of 0.511, are shown. Core length of uncontrolled jet extends up to 2.5D. It is seen that the tab at both orientations enhances jet mixing. The mixing enhancement caused by the tab is seen as reduction in core length. The extent of mixing for the two orientations of the tab is found to be different. Tab along major axis reduces the core from 2.5D to 0.92D, which amounts to a reduction of 63%. Whereas for tab along minor axis the core is 2.15D, thus this tab causes just 14% reduction in core length. Also, it is seen that the effect of the tab prevails only up to 15D and beyond 15D the decay of controlled and uncontrolled jets are almost the same, as seen in Fig. 3(a). Centerline decay results for NPR 5 are shown in Fig. 3(b). NPR 5 is also an overexpanded state, but with a reduced level of adverse pressure gradient than NPR4. At this reduced level of overexpansion, both orientations of the tab offer better mixing compared to NPR 4. For tab along major axis the jet core comes down from 4.3D to 0.84D, which amount to a reduction of about 80%. Whereas for tab along minor axis the core reduces from 4.3D to 1.07D this amounts to about 75% reduction in core length. Controlled jets become fully developed at 10D itself. This clearly shows that the near field mixing caused by the tabs at NPR 5 is superior to NPR 4. The centerline decay of uncontrolled jet and jet controlled with corrugated rectangular tab at overexpanded state of NPR 6 are shown in Fig. 3(c). Tab along major axis reduce the core from 6.75D to 1.15D, which is about 82% reduction in core length. At NPR 6 corrugated rectangular tab along minor axis is



184


(a) At NPR 4

(b) At NPR 5

(c) At NPR 6

(d) At NPR 7

(e) At NPR 8

(f)Core length variation with for corrugated rectangular tab

Figs. 3. Centerline pressure decay for corrugated rectangular tab

At NPR 7 the adverse pressure gradien gradientt at the nozzle exit is considerably lower than NPR 4. In the presence of reduced adverse pressure gradient, even though the near field mixing caused by tab along minor axis continues to be superior to tab along major axis, the waves in the proximity of no nozzle zzle exit are of considerable strength. Further, the far field mixing caused by tab along major axis is found to be superior to tab along minor axis. At NPR 7, the core length reduction caused by the tab along major and minor axis is 83% and 29%, respectiv ely. The results for NPR 8, with is a case of a respectively. marginal favorable pressure gradient of about 1.14%, corresponding to NPR 8, are shown in Fig. 4(c). The tab along minor axis is found to promote better near field mixing than tab along major axis. But in the far field the mixing caused by tab along major axis seems to be superior to tab along minor axis. Core length variation

of the uncontrolled jet, jet controlled with arc tab along major and minor axis, as a function of NPR, are shown in Fig. 5. It is seen that, for arc tab along major axis core length variation with NPR is only marginal, whereas for tab along minor axis core length shows an increase with NPR in the range from 4 to 6 and decreases with NPR in the range from 6 to 8. Centerline decay results ffor or the case of limiting tab of triangular cross section with the vertex facing the jet are shown in Figs. Fig 6(a)-(c). 6(a) (c). It is essential to note that the mixing caused by triangular tab along the major and minor axis are directly opposite to that of rectangular and circular circular-arc arc tabs. Mixing caused by tab along major axis is superior to tab along minor axis in the near as well as far field. Also, the mixing promoting superiority of triangular tab along major axis is found to be independent of the pressure gradient gradien t at the nozzle exit.



185


(a) At NPR 4

(b) At NPR 7

(c) At NPR 8 Figs. 4. Centerline pressure decay for corrugated arc tab

pressure gradient also the tab along major axis is found to be a better mixing promoter than the tab along minor axis. Furthermore, mixing promoting capability of tab along minor axis is found to decrease in the presence of favorable pressure gradient, though only marginal. Thus it may be stated that, irrespective of the pressure gradient at the nozzle exit, for the case of triangular limiting tab, the tab orientation along major axis is a superior mixing promoter than tab along minor axis. This nature is totally different from rectangular and circular arc tab geometry. Core length variation with NPR, for the case of triangular tab, is shown in Fig. 7. It is seen that tab along major axis is a superior mixing promoter at all levels pressure gradient. Compared to tab along minor axis it is interesting to note that the insensitive nature of rectangular and circular-arc geometry (along major axis) has become sensitive to NPR for the case of triangular geometry. This may be because the pressure-hill at the face of triangular tab is considerably different from that at the face of rectangular, circular-arc and triangular (with vertex facing flow) tabs. The pressure results demonstrate that the pressure-hill at the face of triangular tab could be able to divert the flow, leading to shedding of mixing promoting vortices of higher vorticity content than the verities shed by the rectangular and circular-arc tab.

Fig. 5. Core length variation with NPR for corrugated arc tab

At NPR 6, as seen in Fig. 6(a), the core length for uncontrolled jet is about 6.5D. Tab along minor axis brings down the core length to about 2.5D, whereas the tab along major axis reduces the core length to about 1.5D. It is also seen that, the mixing caused by tab along major axis is superior in all the three zones; core, characteristics decay zone and far field, compared to the tab along minor axis. For tab along major axis the jet becomes fully developed at about 6D, whereas for tab along minor axis jet becomes fully developed only beyond 15D, which is identical to that of uncontrolled jet. The centerline decay results for NPR 7 are shown in Fig. 6(b). At NPR 7 also, tab along major axis is a superior mixing promoter in all the three zones of jet than tab along minor axis. Centerline decay results for NPR 8 are shown in Fig. 6(c). In presence of favorable

III.2. Pressure Profile The pitot pressure (pt) distribution, measured along the major and minor axis directions of controlled jets, is nondimensionalised by dividing with the settling chamber pressure p0.



186


(a) At NPR 6

(b) At NPR 7

(c) At NPR 8 Figs. 6. Centerline pressure decay for corrugated triangular tab

Pressure variation in the near field is asymmetric about the jet axis. This may be because of the vortex domination in the jet field, since different vortex will have different frequency and amplitude. At X/D = 1, Fig. 8(a), the maximum pressure ratio pt/p0is in parity with the pressure profile at X/D = 0.5. In this profile, minimum pressure ratio is slightly to the left of origin and pt/p0 is found to be less than 0.2. Because of varying size, frequency and amplitude of vortices in present in the flow field, the pressure profile is asymmetric along Z/D = 0. In the characteristic zone and far field region, say at X/D = 4, 8 and 10, the maximum pt/p0 attained is less than that in the core region. The expanse of pressure profiles at X/D = 8 and 10 spans from Z/D = −1.5 to +1.5. This shows that the flow is fully developed and self-similar. Pressure profiles along Y direction, for rectangular limiting corrugated tab along major axis, are shown in Fig. 8(b). It is seen that at X/D = 0.5 and 1 pt/p0 exhibits a maximum of 0.7. At X/D = 1 the pressure maximum is to the left of origin Y/D = 0.0. In the near field the pressure profiles are oscillating due to shock structure inside the core region. Pressure profile at X/D = 0.4 seems to be symmetric about the jet axis and attains a maximum of 0.42 on both sides of origin. The expanse of pressure profiles in Y direction in the far field is between Y/D = −2.5 to +2.5, which is higher than the expanse for pressure profile in Z direction in far field. In Figs. 8(c) and (d), pressure profiles for rectangular corrugated limiting tab fixed along minor axis at the nozzle exit are presented for NPR 6. From centerline decay results the best mixing caused by the tab was found to be at NPR 6, resulting in a core length reduction of about 89%.

Fig. 7. Core length variation with NPR for corrugated triangular tab

The distances in the directions of major axis (Zdirection) and minor axis (Y-direction) are made nondimensional by dividing them with the equivalent diameter of the nozzle exit (D). Pressure profiles in Z and Y direction, for the case of rectangular corrugated limiting tab fixed along major and minor axis are shown in Figs. 8(a)-8(d). Pressure profiles for rectangular corrugated limiting tab along major axis at axial distances of 0.5D, 1D, 4D, 8D and 10D, at NPR 8, are shown in Fig. 8(a). At X/D = 0.5, which lies in the near field, pressure starts rising from pt/p0 = 0.12, at the left extreme, attains a maximum value of pt/p0 = 0.56 at Z/D = − 0.5, drops to pt/p0 = 0.32 and again rises to a value of pt/p0 = 0.54 at slightly left of Z/D = 0. Pressure attains a local minimum to left of Z/D = 0.



187


The pressure peak pt/p0 at X/D = 0.5 is around 0.62, but the peak value of pt/p0 at X/D = 1.0 is higher as it lies out of supersonic core. The pressure profiles in the characteristic decay and far field regions exhibit continuous decrease of pressure in the jet filed. Figures 9(a) and (b) show the pressure profiles for corrugated limiting triangular tab along major axis at NPR 6. It was seen from centerline decay plots that the maximum mixing for this tab is at NPR 6. From these pressure plots it is seen that maximum value of pressure peak in the Z direction is lower than that in Y direction (Fig. 9(b)). Pressure profiles close to nozzle exit are highly asymmetric because they fall inside the supersonic core. Also the expanse of pressure profiles in Y direction in is far field is more than the expanse in Z direction. In Figures 9(c) and (d), pressure profiles for corrugated limiting triangular tab fixed along minor axis at the nozzle exit, at NPR 6, are shown. At NPR 8 corrugated limiting triangular tab, along minor axis causes only a slight reduction of 7.2% in core length. Pressure profiles for corrugated limiting arc tab, along major axis are shown, at NPR 4, are shown in Figure 10(a). Plots are presented at nozzle pressure ratio of 4. From The centerline pressure decay for this case (Fig. 7) shows a core length reduction of 63%. At X/D = 0.5, there is a dip in pressure ratio and pt/p0 = 0.28, which shows that at jet centerline, velocity is at a local maximum. At X/D = 1.0, lying just outside the core, a dip in pressure ratio, at the center is found where pt/p0 becomes around 0.48, which is lesser than the dip at X/D = 0.5. Pressure profiles in Y direction are shown, the corrugated limiting arc tab along major axis are shown in Figure 10(b). Profiles at X/D = 0.5, 1 and 4 show that the pressure peak assumes maximum on the left of origin (Y/D = 0.0) and lies between 0.68 and 0.58. A slight dip in pressure is found at the origin for X/D = 0.5 and 4.0, while pt/p0dip at X/D = 0 and 1.0 is about pt/p0 = 0.28. Pressure profiles in far the field are symmetric and to peak in pressure value is more than that of other tabs tested. Pressure profiles in Z and Y directions for corrugated limiting arc tab fixed along minor axis are presented in Figs. 10(c) and 10(d). The pressure profiles are found to be oscillatory in the near field and far field for Z direction profiles. As seen from Fig. 7, the arc corrugated limiting tab at NPR 4 and 5 protects the core length. The peak in the pressure profiles in the near field at X/D = 0 and 1 is about 0.96, which implies that the tab at nozzle exit reduces the jet mixing. The pressure profiles are asymmetric about the jet axis. The profile at X/D = 4.0 falls inside the core and shows a pressure peak around 0.67, to the left of origin. Peak in pressure profile at X/D = 4.0, for this case is higher than to the pressure peaks at this location for other types of tabs studied in this work. At The pressure profiles at X/D = 8 and 10 reveal that the flow is still developing and seem to be in characteristic decay region. The behavior of pressure profiles in Figure 10(d), in Y direction is same as behavior in Figure 10(c).

III.3. Shadowgraph Pictures The shadowgraph pictures of the jet controlled with rectangular tab with triangular corrugations, at NPR 4, 6 and 8, are shown in Figures 11. In the pictures viewed along the tab the jet spread is found to be considerably larger than that in the pictures viewed normal to the tab. Also, additional waves due to the tab presence are seen in the pictures taken by viewing along the tab. The higher spread in the direction normal to the tab may be because of bifurcation of the nozzle exit cross section by the tab running across. It is seen that with increase of NPR the waves in the core become stronger, as expected, and the shock cells in the plane normal to the tab become longer and wider. For the case of corrugated tabs, as seen in Figs. 12, the jet bifurcation caused by the tab presence is explicitly seen. Indeed the tab could able to split the jet in to two different jets on either side of it, as seen from Figs. 12(a), (c) and (e). This kind of bifurcation would offer larger circumferential area for the jet to entrain the surround fluid, at the outer periphery, and higher differential shear at inner periphery. These two environments might be the reason for the higher mixing caused by the arc tab compared to the rectangular tab. Figs. 13 show shadowgraph of the jet controlled with corrugated rectangular tab fixed along minor axis at NPR 4, 6 and 8. At NPR 4 as seen from Fig. 13(a), jet is bifurcated. Whereas at NPR 6, which is found to be the best mixing promoter for this tab orientation the jet is bifurcated with longer cells than NPR 4 (Fig. 13(b)). Though two jets signature are visible (Fig. 13(c)), the jet continues to propagate as a single jet after the first shell in which there are complex waves. Also, at the end of second cell there is a Mach disk type of wave formation. This Mach disk is found at NPR 8 (Fig. 13(e)). But the Mach disk at NPR 6 appears to be stronger than that at NPR 8. Due to this beyond the Mach disk (Fig. 13(c)), the cells are found to possess weaker waves than NPR 8. This strong Mach disk at second cell may be the reason for the best mixing encountered at NPR 8. Shadowgraphs for arc tab along minor axis are shown in Fig. 14(a) to 14(f), for NORs 4, 6 and 8. The jet bifurcation caused the tab is clearly seen in Figs. 14(a), 14(c) and 14(e) that the jet is bifurcated by the wire. This is one of the causes for the efficient mixing promotion cased by the corrugated circular-arc tab along minor axis. Shadowgraph pictures for corrugated limiting triangular tab along major and minor axis are shown in Figs. 15 and 16, respectively. For tab along major axis jet does not show any tendency to bifurcate. Whereas for tab along minor axis, at NPR 4, the jet shows a tendency to bifurcate. In the presence of a marginal favorable pressure gradient corresponding to NPR 8 (Fig. 16(c)), the tendency for bifurcation comes down, but there are two Mach disk type of structures in the first cell.



188


(a) At NPR 8 tab along major axis

(b) At NPR 8 tab along major axis

(c) At NPR 6 tab along minor axis

(d) NPR 6 tab along minor axis

Figs. 8. Pressure profiles in Z and Y directions, for corrugated rectangular tab

(a) At NPR 6 along major axis



(d) At NPR 8 tab along minor axis

Figs. 9. Pressure profiles in Z and Y direction, for corrugated triangular tab



189


(a) At NPR 4 tab along major axis



(d) At NPR 4 tab along minor axis

Figs. 10. Pressure profiles in Z and Y-direction, for corrugated arc tab

(a) At NPR 4 along tab

(b) At NPR 4 normal to tab

(c) At NPR 6 along tab

(d) At NPR 6 normal to tab

(e) At NPR 8 along tab

(f) At NPR 8 normal to tab

Figs. 11. Shadowgraph pictures for corrugated rectangular tab along major axis



190


(a) At NPR 4 along the tab

(b) At NPR 4 normal to tab

(c) At NPR 6 along the tab

(d) At NPR 6 normal to tab

(e) At NPR 8 along the tab

(f) At NPR 8 normal to tab

Figs. 12. Shadowgraph pictures for corrugated arc tab along major axis


(b) At NPR 4 normal to the tab

(c) At NPR along the tab

(d) At NPR 6 normal to the tab


(f) At NPR 8 normal to the tab

Figs. 13. Shadowgraph pictures for corrugated rectangular tab along minor axis



191


(a) At NPR along the tab





(f) At NPR 8 normal to the tab

Figs. 14. Shadowgraph pictures for corrugated arc tab along minor axis





Figs. 15. Shadowgraph pictures for corrugated triangular tab along major axis



192






Figs. 16. Shadowgraph pictures for corrugated triangular tab along minor axis

IV.

[7]

Conclusion

The present study shows that all the limiting tabs studied result in core length reduction when located along major axis. But when placed along minor axis only triangular and rectangular tabs promote mixing at all NPRs. The arc tab at NPR 4 and 5 results in reduced mixing and core length of the controlled jet becomes longer by 40% and 14%, respectively, compared to uncontrolled jet. Among the limiting tab geometries and NPR studied, the best mixing is for rectangular tab along minor axis at NPR 6; resulting in maximum reduction in core length of 89%. When placed along major axis both rectangular and arc tab are found to be equally efficient in mixing promotion, at NPR 8,causing about 85% reduction in jet core length. Arc tab along minor axis also reduces core length at NPR 6, 7 and 8, but this reduction is less than 40%. This is significantly less than the reduction obtained for same tab and NPR, placed along major axis.

[8]

[9] [10]

[11]

[12]

[13]

[14]

[15]

References [1]

[2] [3]

[4]

[5]

[6]

[16]

Gutmark, E.J., Schadow, K.C. and Yu, K.H. Mixing enhancement in supersonic free shear flows, Annual Review of Fluid Mechanics, 1995, 27, (1), pp. 375-417. Gutmark, E.J. and Grinstein, F.F. Flow Control with non-circular jets, Annual Review of Fluid Mechanics, 1999, 31, pp. 239-272. Seiner, J .M., Dash, S.M. and Kenzakowski, D.C. Historical survey on enhanced mixing in Scramjet engines, Journal of Propulsion and power, 2001, 17, (6), pp. 1273-1286. Knowles, K. and Saddington, A.J. A review of jet mixing enhancement for aircraft propulsion applications, Proceedings of the IMechE, Part G: J: Aerospace Engineering, 2006, 220, (2), pp. 103-127. Ibrahim, M.K., Kunimura, R. and Nakamura, Y. Mixing enhancement of compressible jets by using unsteady micro jets as actuators, AIAA J, 2002, 40, (4), pp. 681-688. Arakeri, V.H., Krothapalli, A., Siddavaram., V., Alkislar, M.B. and Lourenco, L.M. On the use of microjets to suppress turbulence in a Mach 0.9 axis-symmetric jet, Journal of Fluid Mechanics, 2003, 490, pp. 75-98.

[17] [18]

[19]

[20]

[21] [22]

[23]


Thomas, C., Bra, J-C. andSunyach, M. Noise reduction of a Mach 0.7-0.9 jet by impinging microjets, Computer RendusMcanique, 2006, 334, (2), pp. 98-104. Thomas, C., Sunyach, M., Juv, D. and Bera, J-C. Jet noise reduction by impinging microjets: An acoustic investigation testing micro jet parameters, AIAA J, 2008, 46, (5), pp. 10811087. Davis, M.R. Variable control jet decay, AIAA J, 2008, 46, (5), pp. 1081-1087. Chauvet, N., Deck, S. and Jacquin, L. Numerical study of mixing enhancement in supersonic round jet, AIAA J, 2007, 45, (7), pp. 1675-1687. Chauvet, N., Deck, S. and Jacquin, L. Shock patterns in an slightly underexpanded sonic jet controlled by radial injections, Physics of Fluids, 2007, 19, (4). Behrouzi, P., Feng, T and Mcguirk, J.J. Active flow control of jet mixing using steady and pulsed fluid tabs, Proc. IMechE, Part I: J. Sytems and control Engineering, 2008, 222, (5), pp. 381-389. Kamran, M.A. and Mcguirk, J.J. Subsonoc jet mixing via active control using steady and pulsed control jets, AIAA J, 2011, 49, (4), pp. 712-724. Wan, C. and Yu, S.C.M. Numerical investigation of the air tabs technique in jet flow, J Propulsion and Power, 2013, 19, (1), pp. 42-49. Wan, C. and Yu, S.C.M. Investigation of the air tabs effect in supersonic jet, J Propulsion and Power, 2011, 27, (1), pp. 11571160. Yu, S.C.M., Lim., K.S. Chao, W. and Goh, X.P. Mixing enhancement in subsonic jet flow using air-tab technique, AIAA J, 2008, 46, (11), pp. 2966-2969. Viets, H. Flip-flop jet nozzle, AIAA J, 1975, 13, (10), pp. 13751379. Tam, C.K.W. and Morris, P.J. Tone excited jets, Part V: A theoretical model and comparison with experiment, J Sound and Vibration, 1985, 102, (10), pp. 119-151. Wiltze, J.M. and Glezer, A. Direct excitation of small scale motions in free shear flow, Physics of Fluids, 1998, 10, (8), pp. 2026-2036. Chanaud, R.C. Effects of geometry on the resonance frequency of Helmoltz resonator, J Sound and Vibration, 1994, 3, (178), pp. 337-348. Panton, R.L. Effect of orifice geometry on Helmoltz resonator excitation by grazing flow, AIAA J, 1990, 28, (1), pp. 60-65. Bradbury, L.J.S, and Khadem, A. H., “The Distortion of a Jet by Tabs,” Journal of Fluid Mechanics, Vol. 70, No. 4, 1975, pp. 801813. Ahuja, K. K., and Brown, W. H., “Shear Flow Control by Mechanical Tabs,” AIAA Paper 89-0994, 1989.


193


[24] Samimy, M., Reeder, M. F. and Zaman, K.B.M.Q. Supersonic jet mixing enhancement by vortex generations, AIAA JJ,, 9191-2263, 2263, 1991. [25] Samimy, M., Zaman, K.B.M.Q. and Reeder, M. F. Effec Effectt of the tabs on the flow and noise field of an axisymmetric jet, AIAA J, J 1993, 31, (4), pp. 609 609-619. 619. [26] Zaman, K.B.M.Q., Reeder, M. F. and Samimy, M. Supersonic jet mixing enhancement by delta tabs, AIAA Paper 92-3548, 3548, 1992. [27] Bohl, D., and Foss, J. F., “Enhan “Enhancement cement of Passive Mixing Tabs by the Addition of Secondary Tabs,” AIAA Paper 96-054, 96 054, 1996. [28] Wishart, D. P., Krothapalli. A., and Mungal, M. G., “Supersonic Jet Control Disturbances inside the Nozzle,” AIAA JJournal, ournal, Vol. 31, No. 7, 1993, pp. 1340 1340-1341. 1341. [29] Zama Zaman, n, K. B. M. Q., Reeder, M. F., and Samimy, M., “Control of an Axisymmetric Jet Using Vortex Generators, ““Physics Physics of Fluids, Vol. 6, 1994, pp. 778 Fluids, 778-793 793 [30] Zaman, K. B. M. Q., “Streamwise Vorticity Generation and Mixing Enhancement in Free Jets by Delta Delta-Tabs”, Tabs”, AIAA IAA Paper 93 93-3253, 3253, 1993. [31] Stefen, C. J., Reddy. D.R. and Zaman, K. B. M. Q., Numerical modeling of jet entrainment for nozzles fitted with delta tabs, AIAA paper 97-0709, 97 0709, 1997. [32] Reeder, M. F., and Samimy, M., “The Evolution of a Jet with Vortex Generating T VortexTabs abs : Real Real-Time Time Visualization and Quantitative Measurements,” Journal of Fluid Mechanics, Mechanics, Vol. 311, 1996, pp. 72 72-118. 118. [33] Navin Kumar, S., and Rathakrishnan, E., “Sonic jet control with tabs”, Journal of Turbo and Jet Engines Engines,, Vol. 19, Nos. 11-2, 2, 2002, pp. 107 107-118. 118. [34] Sreejith, R. B., and Rathakrishnan, E., “Crosswire as passive device for supersonic jet control,” AIAA paper 2002 2002--4052, 4052, 2002. [35] Rathakrishnan, E., “Experimental Studies on the Limiting Tab,” AIAA J. J. Vol. 47 (10), pp. 2475 2475--2485. 2485. [36] Lovaraju, P., Paparao, K. P. V., and Rathakrishnan, E., “Shifted cross wire for supersonic jet control paper 2004 2004--4080, 4080, 2004. cross-wire [37] Mrinal, K., Pankaj, S. T., and Rathakrishnan, E., “Studies on the effect of notches on circular sonic jet mixing,” Journal of Propulsion Power, Vol. 22, No. No. 1, 2006, pp. 211 211-214. 214. [38] S. M. Aravindh Kumar and Ethirajan Rathakrishnan, “Characteristics of Controlled Mach 2 Elliptic Jet”, Journal of Propulsion and Power Power,, Vol. 32, 2016, pp. 121 121-133, 133, DOI: 10.2514/1.B35877. [39] P. Arun Kumar and Rathakrishnan E., “Triangul “Triangular ar tabs for supersonic jet mixing enhancement,” The Aeronautical Journal Journal, Vol. 118, No. 1209, November 2014, pp. 1245 1245-1278. 1278.

Authors’ information nformation Indian Institute of technology Kanpur, India India.. Anuj Bajpa Bajpai is Ph. D. Scholar of Aerospace Engineering at Indian Indian Institute of Technology Kanpur, India. E--mail: mail: [email protected]

Ethirajan Rathakrishnan is professor of Aerospace Engineering at the Indian Institute of Technology Kanpur, India. He is well known internationally ffor or his research in the area of high speed jets. The limit for the passive control high-speed of jets, called the Rathakrishnan Limit Limit,, is his contribution to the field of jet research, and the concept of breathing blunt nose (BBN), which simultaneously reduces the pos positive itive pressure at the nose and increases the low pressure at the is his contribution to drag reduction at hypersonic speeds. Positioning the twin twin--vortex vortex Reynolds number at around 5000, by changing the geometry from cylinder, for which the maximum limit for the Reynolds number for positioning the twin twin-vortex was found to be around 160, by von Karman, to flat plate, is his addition to vortex flow theory. He has published a large number of research articles in many reputed international journals. He is a Fellow of many professional societies including the Royal Aeronautical Society. Rathakrishnan serves as the Editor Editor-in in-Chief Chief of the International International Review of Aerospace Engineering (IREASE) and International Review of Mechanical Engineering (IREME) journals. He has authored 12 textbooks. E-mail: E mail: [email protected]



194


Influence of Simulated Space Hazards on Polyimide ArtilonTM Type Used in Space Applications A. M. Anwar1, M. M. Osman2 Abstract – The polyimide performance was monitored and evaluated after exposure to ionized and particulate radiation as a space hazard. The material was irradiated with three different doses 500, 750 and 1000kGy in CO60 source in the presence of air. Both non-irradiated and irradiated materials were characterized by tensile test, Thermo-Gravimetric Analysis (TGA) and Fourier Transform Infra-Red (FTIR). Quantum modeling was executed by using "Gaussian 5" software program for chemical structure verification. The non-irradiated material showed a super ductility but revealed a brittle behavior when irradiated with 500 and 1000 kGy gamma doses. Nevertheless, the material attained moderate ductility when exposed to 750kGy. Results by the stated characterization tools matched with the evaluated behavior and confirmed by the quantum modeling. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Polyimide, Gamma Dose, Charged Particles, Ionized Radiation, Quantum Modeling, FTIR, TGA

I.

The proton beam irradiation resulted in a finer structure of some absorption bands indicating the generation of structures that are more varied. These structures enhance the embrittlement of the irradiated material with the PB. In case of EB irradiation a remarkable increase in both the modulus of elasticity and ultimate strength whereas the ductility showed slight increase. Effect of different type of radiation, ionized radiation, was evaluated on two polymeric materials namely polypropylene (PP) and polyimide (PI). Mathakaria et al [3] irradiated the stated two materials with gamma radiation CO60 source up to 250kGy simultaneously. Their mechanical and thermal properties were evaluated by tensile testing and TGA respectively. The PI showed superior properties compared with those of PP. This was explained by the more complex structure of the later material and the low concentration of the broken bond after irradiation. Also, gamma radiation was used to irradiate different types of polyimide by Danie1 Chun-Hung et al [4] but for prolonged times and consequently higher doses reached 28 MGy. Additional testing techniques such as; DSC, and viscosity & gel permeation chromatography (GPC) were used to evaluate the irradiation effect with different doses. All the tested polyimide samples showed higher tensile strength and more ductility with relatively lower doses up to 3.50 MGy. In the contrary, the mechanical properties and thermal stability of the investigated polyamides showed sever degradation when irradiated with doses above 14 MGy. Moreover, the GPC showed direct proportional molecule weights reduction with increasing the irradiation doses.

Introduction

Since the Soviet Union launched its first satellite "Sputnk-1" in 1957, intensive work has been paid to develop a high strength to weight ratio materials to be used in the heavy weight structure(s). Different high strength to weight ratio materials has been investigated thoroughly to characterize their performance as a function of the applied space hazard for prolonged intervals. These materials cover different material categories having different strength levels ranging from the ductile polyimide, composites and even high strength non-ferrous alloys. Since this date, the polymeric materials, known to be used in spacecraft applications such as mechanical fittings and thermal insulators, have been characterized to evaluate their degradation with the space environment. For the conditions of the low earth orbit, Courtney P. et al [1] studied the interaction of the particulate radiation with one of the polymeric materials namely polyimide (PI). The author traced the Chemical variations of the investigated PI by FTIR, mass spectrometry and (Ultraviolet-visual) spectrometry. The test conditions confirmed the material degradation as a result of CO, CO2 and H2 formation. Mária Porubská et al [2] studied the effect of the same irradiation type on both the mechanical and chemical properties of polyamide material. Irradiated polyamide-6 (PA-6) by a 500 kGy electron beam (EB) dose or by 500 and 1000 kGy proton beam (PB) doses was examined by FTIR spectroscopy, cross-linked portion determination, Differential Scanning calorimetry (DSC) and tensile testing. The FTIR spectra of the irradiated specimens with the PB showed major structure changes compared with those irradiated with EB.


DOI: 10.15866/irease.v9i6.10041

195

A. M. Anwar, M. M. Osman

Cota S. S. et al [5] studied the effect of gamma radiation and its rate on the high-density polyethylene (HDPE). The author concerned the influence of these parameters on HDPE mechanical properties as a result of the oxidative degradation or cross-linking. The experimental results showed that the strength of the irradiated HDPE was improved as the irradiation dose increased. Repeatable results were obtained when irradiation took place with slower rates where the interaction with the irradiated materials was more efficient. However, in this work, the mechanical properties, thermal stability and chemical structure variation of polyimide ArtilionTM type would be investigated. Quantum modeling of the PI structure will be used to verify the obtained results.

II.

Experimental Work II.1.

Material

Fig. 1. Gamma Irradiation Cell (Co60)

TM

A polyimide Artilon type is used to be investigated by a simulated space hazards, mainly the charged particles. Table I summarizes both the mechanical and physical properties of the selected material before irradiation.

II.4.

The mechanical properties of both the non-irradiated and irradiated tensile specimens were evaluated using the universal tensile testing machine GALDABINI "Quasar 100". A constant strain rate of 5 mm/min was adapted with all the tested samples. All the tests have been performed at room temperature (23±3°C) and relative humidity of (50±10 %) according to the ASTM D3039. To explain and relate the variation of the mechanical properties of the irradiated material to the chemical structure, a Fourier Transform Infra-Red (FTIR) spectrometer JASCO 4100, shown in Fig. 2, was used. A thin round disk of 7 mm in diameter was analyzed after the following preparation steps: to grind the polyimide material, to mix the polyimide powder with KBr powder in a ratio not more than 2% for 3 to 5 minutes to reduce the particle size and finally to press the mixture in-between the two threaded bolts for 2 minutes to form a pellet. All the FTIR measurements were performed with the attenuate total reflection (ATR) method in the range of the available wavelength [4004000 cm-1] with an average resolution of (4 cm-1).

TABLE I MECHANICAL PROPERTIES OF THE USED MATERIAL Material

Trade name

polyimide

ArtilonTM

II.2.

Tensile Strength (MPa) 83

Tensile Modulus (GPa) 3.37

Failure Strain (%) 10

Density (g/cm3) 1.14

Specimen Preparation

The 3mm thick polyimide ArtilonTM type was striped 18 mm in width, 200 mm in length that fits the technical requirements established by the ASTM-D-3039 standard [6]. II.3.

Evaluation Methods

Radiation Source

The stripped specimens were exposed to the gamma radiation using the Cobalt (Co60) radioactive source, shown in Fig. 1, at the Egyptian Atomic Energy Authority "National Center of Radiation" in Cairo. The exposing process has been done in the presence of air and with a dose rate of 2.08 Gy/h at room temperature. Doses of 500, 750 and 1000 kGy were selected to simulate the effect of the charged particles in low earth orbit for predetermined lifetime of 5, 7.5 and 10 years. These doses were calculated by using Space Environment Information System "SPENVIS" software to correlate the irradiation doses to the chosen lifetime [7]. To calibrate the source irradiation dose, three dosimeters were used at three different levels namely 30, 40 and 45kGy. Three tensile specimens were irradiated with each prescribed gamma dose to confirm the results.

Fig. 2. JASCO 4100 FTIR spectroscopy



196


A run for the background scan was done to eliminate the spectrum noise, to sharpen the peaks profile and consequently improve the measurement resolution. Values of the chemical bonds energy depending on the geometrical shape of the polyimide structure were calculated by using Density Functional Theory (DFT) at B3LYP/3-21G method by "Gaussian 5" software. To trace the polyimide thermal stability within a temperature range, covering the working temperature, a thermo-gravitational analysis was applied using TGA model "Select-Horn 2000367". With a heating rate of 10°C/min the temperature of the analyzed samples was raised from room temperature to about 1050°C in the presence of air burg. The sample mass was traced during the test with a balance accuracy of 1 µg.

Increasing the dose to a value of 562 kGy helps the broken 5-membered ring to reconstruct a more stable 6membered ring with lower inter-molecular energy, Fig. 5. The reconstructed ring has polarizable bonds end with electronegative atoms, which have the ability to be broken again with more energy to form the original 5membered ring.

Fig. 4. Effect of gamma radiation on the chemical structure of the Polyimide at 500 kGy

III. Results and Discussions Radiation source calibration by the dosimeter showed a 25% deviation from the theoretical calculated values. Therefore, the predetermined irradiation doses namely 500, 750 and 1000 kGy assigned for orbital lifetime of 5, 7.5 and 10 years should be reconsidered as 375, 562 and 750 kGy corresponding to 3.75, 6.52 and 7.5 years respectively. These doses affect the mechanical properties of the irradiated PI compared with the nonirradiated one as shown in Fig. 3.


The reconstructed structure accomplishes proportional increase in ductility and toughness with increasing the number of the polarizable bonds. The irradiated polyimide with this dose attained higher strength and ductility compared with the non-irradiated or even the irradiated material with the other doses because of the lower inter-molecular energy of the restored material. Irradiation with a higher dose value, namely 750 kGy, causes the reconstructed 6-membered ring to be changed back to the main 5-membered ring, Fig. 6. This newly formed structure with the higher inter-molecular energy attained the same mechanical properties recorded with the material irradiated by the lower dose, 375 kGy. The FTIR spectrums of either the irradiated or even the non-irradiated polyimide are shown in Fig. 7.

Fig. 3. Stress-Strain Curves for Irradiated and Non-irradiated Polyimide

In this figure, the non- irradiated polyimide showed a completely ductile behavior which reflects the high flowability. During loading, parallel bright deformation bands, similar to those Luder's bands in metallic materials, were observed along the gauge length. Reaching the ultimate strength value, a necking was nucleated and developed from the intersection of one of these bands with the sample edge. Beyond this point, the stress started decreasing till fracture. The large area under the shown curve indicates a high toughness value compared with the irradiated samples. An embrittlement behavior was clearly observed with all the irradiated samples where the superior ductility was reduced. Quantum modeling showed that exposing the investigated material to 375 kGy was sufficient to impose a value of energy exceeding the threshold necessary to break the 5membered ring (A) and forms a radical fragment (B), Fig. 4, which causes the structure instability and reduces the ductility dramatically.




197


Wave number: 3449.06 Bond (N-H) stretching Wave number: 3378.67 Bond (O-H) stretching

1 Wave number: 2919.7, 2849.3 Bond (C-H) stretching

2

5, 6

3 -

Wave number: 1630.5, 1590.02 Bond (N-H) bending, (C-C) stretching

8, 9 Wave number: 1383.6, 1350.89 Bond (C-N) stretching, (C-H) rock Wave number: 719.318 Bond (C-H) deformation

11 -

Non-Irradiated

500 kGy

750 kGy

1000 kGy

4 Wave number: 1702.84 Bond C=O stretching Ar

7 Wave number: 1463.71 Bond (C=C) aromatic

Fig. 7. FTIR Spectrometer of the Polyimide

All the characteristic peaks and their corresponding wavenumbers (wavelengths) are summarized in Table II. Increasing the irradiation dose causes the peak representing the (NH) group, at a wavenumber of 3449.06 cm-1, to be shifted toward a lower value. This slight shift indicates the phenomena of hydrogen loss and chain scission [8]. Moreover, the intensity of the observed peak 1590.02 cm-1 representing the (C-C) bond was continuously decreasing with increasing the gamma doses. This reduction is a proof for the bond breaking. Formation of new (C=O) bond was represented by the observed peak at 1702.84 cm-1 where its intensity is directly proportional with increasing the irradiated dose. This created (CO) group may have either aliphatic or non-conjugated bond, which is responsible for oxidation and confirms the principle of ring opening and breaking. This matches with what was suggested by the quantum modeling explained previously. Moreover, bonds of the ether group were broken and its characteristic peak at wavenumber 1038.48 cm-1 was disappeared. This enhances the formation of the more stable (C-H) bonds at 2919.7 cm-1 and 2849.31cm-1 wavenumber [9]-[11]. Spectrum of the irradiated PI with 562 kGy showed a different behavior than those exposed to 375 and 750 kGy. For the latterly stated doses, intensities of the corresponding characteristic peaks are almost in the same level whereas those for the formerly

stated one are different. Fig. 8, presents TGA behavior of both the non-irradiated and irradiated PI with different doses. In this figure, the non-irradiated PI showed a 3 stages profile for the weight loss with temperature. In the range of 150°C to 300°C, a little weight loss was detected because of the gasses evolving namely; H2O, -C6H4-NH2, CO2 and CO [1], [12]. In a temperature range 300°C to 600°C, decomposition of the material took place with a decomposition rate of 0.5 % per degree. Total mass loss was observed at temperature higher than 600°C. Irradiation with all the specified doses, even with the lowest dose, enhanced the thermal stability of the polyimide.

Fig. 8. TGA thermo-grams of Polyimide before and after irradiation



198


No 1 2 3 4 5 6 7 8 9 10 11

Bond (N-H) stretching [C-H] stretching (al. CH2 group) [C-H] stretching (al. CH2 group) C=O stretching (N–H) bending (C-C) stretching aromatic (C=C) aromatic (C-N) stretching (C-H) rock [-C-O-C-] stretching ether (C-H) deformation

TABLE II POLYIMIDE FTIR SPECTRUM RESULTS Peaks for Non-Irradiated Peaks for Dose 375 kGy Peaks for Dose 562 kGy Peaks for Dose 750 kGy Wave number Wave number Wave number Wave number Abs Abs Abs Abs [cm-1] [cm-1] [cm-1] [cm-1] 3449.06 1.13909 3432.67 0.873988 3431.71 0.369705 3440.39 0.60289 2919.7 0.454328 2918.73 0.75119 2919.7 0.523025 2849.31 0.403582 2848.35 0.675577 2849.31 0.474246 1702.84 0.186216 1712.48 0.312436 1711.51 0.209337 1630.52 0.570395 1631.48 0.471524 1628.59 0.356861 1630.52 0.36861 1590.02 0.643605 1589.06 0.524949 1596.77 0.365456 1590.02 0.433914 1463.71 0.118504 1469.49 0.310095 1467.56 0.579889 1468.53 0.380064 1383.68 0.218183 1383.68 0.188541 1374.03 0.250496 1381.75 0.190957 1350.89 0.234924 1350.89 0.213561 1352.82 0.269261 1350.89 0.218988 1038.48 0.08927 719.318 0.228001 718.354 0.339661 721.247 0.529586 719.318 0.408917 Various Polyimides." Journal of the Chinese Chemical Society vol.47 (2000): 583-588. [5] S. S. Cota, V. Vasconcelos, M. Senne Jr., L. L. Carvalho, D. B. Rezende and R. F. Côrrea. "Changes In Mechanical Properties Due To Gamma Irradiation Of High-Density Polyethylene (HDPE)." Brazilian Journal of Chemical Engineering vol. 24 part 2 (2007): 259 - 265. [6] Amarican Socity for Testing and Materials. "Tensile Properties of Polymer Matrix Composite Materials." Annual Book of ASTM standards October 1995: 99-109. [7] M. Kruglanski, N. Messios, E. De Donder, E. Gamby, L. Hetey, S. Calders. ―Space Environment Information System‖ (SPENVIS). ESA, Europian Space Agancy. . [8] "Effect Of Ionizing Radiation On Different Polymers and possible Use Of Polymers In Radioactive (Nuclear) Waste Management." School Of Natural And Applied Sciences Of Middle East Technical University, 2006. [9] Joao Carlos Miguez Suarez, Elisabeth Ermel da Costa Monteiro and Eloisa Biasotto Mano. "Study of the effect of gamma irradiation on polyolefins—low-density polyethylene." Polymer Degradation and Stability vol.75 (2002): 143–151. [10] Bashir Ahmed, S K Raghuvanshi, Siddhartha, A K Srivastava, J B Krishna and M A Wahab. "Optical and Structural Study of Aromatic Polymers Irradiated by Gamma Radiation." Indian Journal of Pure& Applied Physics vol. 50 (2012): 892-898. [11] Xiao Hao, Lu Yonggen, Wang Mouhua, Qin Xianying, Zhao Weizhe, Luan Jian. "Effect of gamma-irradiation on the mechanical properties of polyacrylonitrile-based carbon fiber." Carbon, vol.52 (2013):427-439 [12] Ahmad Anwar, Dalia Elkiky, Gamal Hassan, Marta Albona and Mario Marchetti. "Outgassing Effect on Spacecraft Structure Materials." International Journal of Astronomy, Astrophysics and Space Science vol.2 part.4 (2015): 34-38.

Evolving of gasses became more difficult even with temperatures higher than 300°C whereas, the decomposition started at 450°C with a steeper decomposition rate reached 1% per degree. This accelerated rate was expected to be a result of the early scissioning of the main polyimide chain exerted by the gamma radiation. The profiles are identical for the irradiated PI by the other doses. Thermal profiles of the irradiated materials are almost the same whereas, those irradiated with 750 and 1000 kGy are completely identical.

IV.

Conclusion

The gamma radiation was used as a simulation of the charged particles and ionized radiation presented in the low earth orbit (LEO). Radiation doses of 375, 562 and 750 kGy were used to simulate one of the space hazards for a corresponding orbital lifetime of 3.75, 5.62 and 7.5 years as calculated by SPENVIS. The predetermined does decrease the ductility of the PI dramatically and causes its embrittlement. The irradiated PI reveals a higher strength and ductility with the intermediate dose (562 kGy). Reversible mechanical behavior with the dose variation was explained successfully by the FTIR. Mechanical properties variation with the different doses was confirmed by modeling the PI geometrical shape, calculating its thermal structure stability, its dipole moment and finally its bond energy. Irradiation enhances the PI thermal stability where it starts decomposition at 450°C rather than 300°C.

Authors’ information Ahmad Mohamed Anwar, born in Alexandria, Egypt in 1970. Bacholar of engineering in 1993 at mechanical power and energy department, master degree in mecahnical design at 2007 from Alexandria university and PhD in mecahnical engineering at 2016. Working in the materials selection and design since starting the master degree, form 2009 till now works as a space materials designer in the space technology center.

References [1]

[2]

[3]

[4]

Kaiser, Courtney , Ennis and Ralf. "Mechanistical studies on the electron-induced degradation of polymethylmethacrylate and Kapton." Phys. Chem. Chem. Phys., 2010,12, 14902-14915. Mária Porubská, Ondrej Szöllos, Alena Kónová, Ivica Janigová , Miloslava Jasková , Klaudia Jomová and Ivan Chodák. "FTIR spectroscopy study of polyamide-6 irradiated by electron and proton beams." Polymer Degradation and Stability vol.97 (2012): 523-531. N.L. Mathakaria, V.N. Bhoraskar and S.D. Dhole. "Stress–strain and thermo-gravimetric analysis of Co60 gamma-irradiated polypropylene and polyimide." Radiation Effects & Defects in Solids vol.169.part.4 (2014): 334–343. Liu, Danie1 Chun-Hung. "Effects of Gamma Radiation on

Mahmoud Maher Osman, born in Cairo, Egypt, in 1969. Bacholar of engineering in 1991 at production department, master degree in material scince at 1997 from Cairo university and PhD in Material science at 2002 from Stathclyde universty, Glasgow, Scotland. Working as a lectueral since 2003.



199


Ascent Phase Trajectory Optimization of Launch Vehicle Using Theta-Particle Swarm Optimization with Different Thrust Scenarios Dileep M. V., Surekha Kamath, Vishnu G. Nair Abstract – Launch vehicle trajectory optimization has gained enormous significance in the recent past. Constraints handling and accuracy of launch vehicle system, are challenging factors, on account of their high degree of non-linearity. This paper brings in the application of thetaparticle swarm optimization (TH-PSO), which is a recently emerged variant of particle swarm optimization (PSO), for launch vehicle trajectory optimization, which can efficiently handle the constraints and drive the system towards optimality. TH–PSO approach is implemented on a multistage liquid propellant rocket, taking angle of attack as the control parameter. Single and dual thrust cases were solved using TH-PSO technique, and a comparative study was made with classical PSO in terms of terminal error, IE consistency of solutions. Based on the statistics, it can be confirmed that in both single and dual thrust cases TH-PSO outperformed, classical PSO. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Trajectory Optimization, PSO, THETA-PSO, Launch Vehicle, Different Thrust Scenarios

All space missions require optimal allocation of resources so as to maximize the efficiency and to minimize the risk and operation cost. The process of trajectory optimization starts with initial trajectory generation stage and it has become an integral component of space mission design .Trajectory optimization is a rapidly growing field in the area of optimization since a well-designed trajectory which satisfies mission objectives, dynamics of the launch vehicle system, the operational scenario, will play a crucial role in the overall mission performance and effectiveness. In order to reduce the operating cost and structural load, effective and optimal ascent phase guidance is a major concern. Trajectory optimization of such systems mainly concentrates on achieving certain performance parameters within specific boundary conditions [1], [2]. The launch vehicle guidance generally involves two or three phases [3]. Atmospheric phase, which constitutes the initial phase will be in open loop and during this phase, the vehicle is guided using a non-optimal piecewise linear attitude control architecture. Closed loop guidance is implemented in the second and third phases which constitutes exo-atmospheric phase of the trajectory. These phases are more reliable and efficient due to the application of closed loop technique, whereas in the case of first phase, the variation in wind profile from the mean value, which is used in the computation of attitude control program results in launch delays, which in turn increases the operation cost of the system. Succeeding literature review will bring about the afore mentioned characteristics.

Nomenclature c1 ,c2 g m re r R1 ,R2

Acceleration constant Acceleration due to gravity Instantaneous mass of the vehicle Radius of earth Radial distance from the centre of earth Random number with normal distribution

t0

Initial time

tf

T u v xf

Final time Thrust Control variable Instantaneous velocity Terminal values for state variables

xt0

State variables at initial time

xt f

Error occurred at terminal values

yi

Particles in PSO algorithm

zi

Velocity of particles in PSO algorithm

x f

Tolerance allowed for terminal values

xt f

Error occurred at terminal values

 t

Increment in phase angle

t 

Phase angle of the particle Inertia weight

I.

Introduction

Optimal trajectory generation is a dynamic research area in military as well as space technology. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved

DOI: 10.15866/irease.v9i6.10521

200

Dileep M. V., Surekha Kamath, Vishnu G. Nair

Betts, in his review [4], on major classification of optimization techniques – direct and indirect method, brought out the advantages and disadvantages of various approaches. Indirect approach, which is the most preferred method, due to its high accuracy, followed by stating the problem in terms of differential equations, maximum principle and boundary conditions associated with them [5]. This approach helps in formulating the problem as a multipoint boundary-value problem (MPBVP). The generated trajectory satisfies the differential equations within the available boundary constraints. On the other hand, direct method followed by representing the dynamic variables in parametric form and further adjusting the variables to optimize the objective function directly. Based on numerical methods for optimization, Gradient and evolutionary (stochastic) based methods are significant [6]. Among these methods, the gradient based approach is preferred over its counterpart on account of its computational efficiency. But when it comes to complex nonlinear systems, such as launch vehicle, gradient based methods exhibit ambiguities, among which trapping and locking of solution to a local minima is the most common and significant shortcoming. It also requires suitable identification of initial solution which is a problem dependent entity and is unknown initially. In order to circumvent the shortcomings of gradient based approaches, evolutionary methods- inspired by natural phenomenon in optimization problem which can handle nonlinearities effectively. Most prevalent evolutionary based methods includes, genetic algorithms (GA) [4], [36][37], ant colony optimization (ACO) [5], particle swarm optimization (PSO) etc[7]. Kennedy et al. [8], [9] proposed PSO algorithm, inspired by natural behaviour involved in fish schooling, bird flocking etc.. Since it is based on social behaviour the results obtained were more efficient than other stochastic techniques on account of the fact that it is taking the best of all the individuals in the group [10]. The characteristics of the algorithm such as computational simplicity and easiness of implementation, made it suitable for real-life nonlinear complex problems like trajectory optimization. Later Kennedy et al. [11] and Trelea [12] proved that the performance of PSO algorithm is parameter dependent such as inertia weight and acceleration coefficient. Among the variants of PSO developed over the years, the most recent and efficient among them is TH-PSO by Zhong et al. [13]. Position of particles are adjusted by velocity in classical PSO, where in the case of TH-PSO, mapping of phase angles are used for updating the positions. Effectiveness of TH-PSO lies in, handling nonlinear system and same is having explored in this paper. The problem considered in this paper is to find an optimal trajectory for a multi stage liquid propellant launch vehicle, by minimizing the terminal error under certain path constraints and boundary conditions with angle of attack as the control variable. Here we consider two thrust scenario- single and dual and relatively a new

variant of PSO, TH-PSO to solve the problem. Further for validation, results obtained using TH-PSO were compared with those obtained by PSO algorithm. Organization of this paper is as follows. Section 2, consolidates the system description and problem formulation. Section 3 describes the methodology, comprising trajectory optimization problem discretization and solution methods. Further Simulation results and numerical validations are given in Section 4. Section 5 concludes the work.

II.

System Dynamics and Problem Statement II.1.

System Dynamics

Advanced launch vehicle system model (ALS), developed by NASA, is being considered for testing the efficiency of TH-PSO, in the scenario of trajectory optimization, since the same was used for studying the behaviour of optimal and suboptimal control and guidance strategies[14], [15]. The system comprises of the core vehicle with liquid rocket boosters having three engines, which produces identical thrust. Combined effect activity of boosters and core vehicle marks the atmospheric phase, where boosters get exhausted and jettisoned. The flight of the core vehicle alone will comprise the latter phase. Forces acting upon launch vehicle during ascent phase is depicted in Fig. 1.

Fig. 1. Schematic representation of the forces acting in launch vehicle

Various assumptions made for formulating the launch vehicle model are; earth devoid of rotational motion, launch vehicle being a point mass object, the total thrust is aligned with the body longitudinal axis or slanted at a constant angle and effect due to gimballing of thrust vector is neglected. Following are the equations governing the dynamics of launch vehicle motion:


r  V sin 

(1)

1 V  T cos   D  mg sin   mV

(2)


201


 

1 V g T cos   L      cos  mV r V x  V cos 

i th

(3)

D th

particle in

dimension is denoted as,

T

Yi   yi1 , yi 2 ,..., yiD  . The velocity of each particles are T

represented by, Z i   zi1 ,zi 2 ,...,ziD  and each particle is

(4)

having individual personal best positions and is denoted where L and D are lift and drag forces respectively:

D  1  Sv 2 cD  2  1 L  Sv 2 cL  2

T

as, Pi   pi1 , pi 2 ,..., piD  . Also the entire swarm is having its best value in each iteration and it is known as global best position. It is indexed as g in the equation. The global version PSO, the swarm updates are carried out based on the given equations and the superscripts denote the iteration number [16], [17]:

(5)

where,  and S represents the aerodynamic parameters of the system under consideration. ‘g’ is a variable, which varies with altitude and its effect is considerable at altitude, upto 50 km above sea level. g is defined as:

r  g  g0  e  r II.2.





 c2 R2 pgt  d   yit  d 

2

(6)

(8)

current position of the particle respectively. R1 and R2 represents two uniformly distributed random variables. The acceleration coefficients are given by c1 and c2 which controls the exploration and its rate. Later a new parameter,  is introduced by Shi et al.[18] in classical PSO, which is used to enhance the performance. In classical PSO approach the inertia weight parameter is a fixed value in all the iterations done which results in local minima locking. To alleviate these shortcomings Eberhart et al. [19], modified the inertia weight to a linearly decreasing quantity, defined by Eq. (9), for further improvement of search capability:

variable chosen is  along with r,v and  as state variables. The problem can be stated as: 2

x f  xt f

(7)

where zit  d  and yit  d  represents the velocity and

The objective is to find an optimal angle of attack  for a constant thrust T, which transfer the system under consideration from an initial to final state within fixed time t f , by minimizing the terminal error .The control

 SX



yit 1  d   yit  d   zit 1  d 

Problem Formulation

Minimizing J 



zit 1  d    zit  d   c1 R1 pit  d   yit  d  

2

subjected to following constraints: Path constraint: x  f  x,u,t  Boundary conditions: xt f  x f  xt f , xt0  x0 State constraints: xmin  x  xmax

   min   t 1  max   max  iter  itermax 

III. Methodology Theory and methodology involved in classical and TH-PSO algorithms brought out in this section.

(9)

Various efforts were made by researchers like Clerc and Kennedy [16], Bergh and Engelbrecht [20] etc to improve the efficiency of PSO by analysing various performance parameters. Kennedy [21], [22] proposed various topological models. Many variants of PSO are proposed by Bergh and Engelbrecht [20], Chen et al.[23], Liang et al.[24] and so on. Ho et al.[25] presented a novel orthogonal method for PSO enabling it to solve task assignment problem. An adaptive variant of PSO, which has a three level learning mechanism is proposed by Li and Yang [26], [27].

III.1. Classical PSO A new algorithm based on the behaviour of social systems such as fish schooling, bird flocking etc is proposed by Kennedy and Eberhart in 1995. Since it is based on social behaviour the results obtained are more efficient than other stochastic techniques due to the fact that it is taking the best of all the individuals in the group. It has emerged as one of the most popular evolutionary method because of its speed and accuracy. Classical PSO as described by Kennedy and Eberhart is discussed in brief. Particles are the basic elements in algorithm. Let i is the total number of particles and D is the degree of freedom or dimension associate with each particle.

III.2. TH- PSO Instead of taking velocity as a major parameter, as in standard PSO, TH-PSO considers phase angle and positions are adjusted by phase angle mappings rather



202


than velocity [13], [28]. Mathematically θ-PSO can be described using the following equations:





t it 1  d   it  d   c1 R1  pi  d   it  d  



 c2 R2  git  d   it  d 



Application of the above explained algorithm to trajectory optimization problem is discussed. The first step in implementing evolutionary optimization techniques is to discretize the variables so as to convert t  t0 ,t f  to    1,1 , a process known

(10)

as transcription. The mapping function used for achieving this is:

it 1  d   it  d   it 1  d 

(11)

 xit

 f

  it

t f  t0

(12)



2

t f  t0

(14)

2

The objective is to minimize the cost function: where

it

and

it

are the phase angle and increment in



phase angle respectively. yit is the particle position.

it

  min , max 

it

phase angle of particle i at time t;

it

of particle i’s phase angle at time t;

d  t  pi

 

 L  x   ,u   ,  d



the phase

(15)

1

dX t f  t0  f L  x   ,u   ,  d 2

  x  1 ,t0 ,x 1 ,t f   0

angle of the personal best solution of particle i at time t;  git  d  the phase angle of the global best solution at time t, f

2

Subject to the following dynamic boundary conditions and path constraints:

the increment

d 

1

t f  t0

  min , max  , i=1, 2,…, s and j=1,

2,…, n. Here, c1 , c2 ,  , R1 , R2 and yit are the same as in standard PSO already explained in PSO algorithm. n represents the dimension of the problem; it  d  the

it



J   X 1 ,t f 

C ( X ( ), U ( ), ; t0 , t f )  0

constraints,



(16)

(17)

(18)

is the monotonic mapping function, which is To approximate the left hand side of the dynamic equations, global interpolating polynomials are differentiated and approximation of state and control variables are performed using N+1 Lagrange interpolating polynomials at the LG points:

expressed as:

 

f it 

xmax  xmin x  xmin sin   max 2 2

(13)

Several constraint optimization problem are solved by this method. Shayeghi et al. [29] and Amin safari[30] referred TH-PSO algorithm to solve the damping issue of two separate electrical system. Mehdi et al. [31] Proposes an optimal load shedding policy using TH-PSO. Vahid et al. [32] solved a constrained economic load dispatch (ELD) problem of thermal plants using TH-PSO and validated the effectiveness using varying degree of complexities.

N

u    U   

U  i  Li  

(19)

i 0 N

x    X   

 X  i  Li  

(20)

i 0

where Li   are Lagrange polynomials: N

Li   



  j

j 0 , j i  i

 j

(21)

To handle the dynamic constraints, dynamic objective method proposed by Haiyan Lu et al.[33] is considered. The particles which are outside the feasible region has to be bought into the region and those which are inside are used to find the optimal solution. In doing so, the initial COP (constrained optimization problem) can be converted into a bi-objective unconstrained optimization problem. One of the objective is to enter the feasible region which is expressed as the sum of constraint violation:

Fig. 2. Particle distribution in TH-PSO and APSO



203


J  y 

q



q



i 1



i 1



Table (thrust conditions) 1. Single thrust:

 max  0,gi  y    max  0, h j  y     (22)

T  15069270 N ; 0  t  370 where the second objective is explained in the previous section.

IV.

2. Dual thrust:

T  25813000 N ; 0  t  150

Results and Discussions

T  7744000 N ; 150  t  370

In this paper, the problem of trajectory optimization of launch vehicle using relatively more robust and efficient variant of PSO- theta PSO is brought out and the same has been compared with APSO, which is the most commonly used evolutionary technique, in terms of consistency and accuracy. The parameter under considerations is terminal error and engine burning time-which is kept constant to convert the given problem to a fixed time optimal control approach, angle of attack- attack is forced to a maximum allowable value so as not to increase the structural load above the maximum allowable limit during the implementation. The parameters of both PSO variants are selected from reference [34] and 50 iterations are performed. The population size is chosen to be 20. Two set of PSO parameters (inertia weight and acceleration coefficients) are recommended in this study which are also used by Trelea [12] and Clerc [11]. Tables I and II give the initial and terminal conditions of the vehicle for single thrust and dual thrust scenarios respectively. First two rows of the table represents the initial and terminal conditions of the launch vehicle Table III describes the allowable limits of the state variables. Tables IV and V show the solutions obtained from both TH-PSO and APSO for single thrust scenario.

In that standard tolerance allowed to each state variables and best, worst, mean solutions are specified, which will help to determine the consistency of the solution. Since the standard deviation of each terminal state variables are less in TH-PSO method we can conclude TH-PSO is producing more consistent solutions than APSO. Tables VI and VII show the best, worst solutions obtained for dual thrust scenario solved by both TH-PSO and APSO algorithms. By analyzing the tables it is interpreted that TH-PSO is producing more consistent solutions. TABLE IV NUMERICAL RESULTS OF SINGLE THRUST SCENARIO USING A PSO State variables

x f

r v 

20 10 10

xt

xt

f

xt

f

xt

f

f

 best   worst 

(mean)

 st  d 

8.4585 16.9845 0.3032 8.5667 0.05 5.09

13.4689 4.5847 2.7327

2.3936 2.3933 1.6591

Units m m/s deg

TABLE V NUMERICAL RESULTS OF SINGLE THRUST SCENARIO USING TH-PSO

TABLE I INITIAL AND FINAL CONDITIONS FOR SINGLE THRUST (INTEGRATED) Parameters Initial conditions Final conditions Units r 6377353 6525113 m v 65 7855.121 m/s  89.5 0 deg m 1523400 152052 kg t 0 370 s

State variables

x f

r v 

20 10 10

xt

f

xt

f

xt

f

xt

f

 best 

 worst 

(mean)

 st  d 

4.4768 0.0309 4.07

10.6785 3.9755 5.32

6.9403 1.9804 4.7553

1.9938 1.0511 0.4038

Units m m/s deg

Figs. 3-6 explain the parameter variations for single thrust case of the launch vehicle. Fig. 3 describes the variation of radial distance from earth centre. It is clear that in both algorithms, the terminal points are achieving successfully.

TABLE II INITIAL AND FINAL CONDITIONS FOR DUAL THRUST (SEPARATE) 1st stage 2nd stage Parameters Initial Final Initial Final Units conditions conditions conditions conditions r 6377353 6438553 6438553 6525113 m v 65 2630 2630 7855.121 m/s  89.5 --0 deg m 1523400 546,600 546,600 152052 kg t 0 150 150 370 s TABLE III BOUNDARY CONDITION FOR THE VEHICLE Parameters Max Min Units v 7860 65 m/s  120 0 deg Fig. 3. Altitude variation with time for single thrust condition



204


TABLE VII

NUMERICAL RESULTS OF DUAL THRUST SCENARIO USING TH-PSO State variables

x f

r v 

20 10 10

xt f

xt f

xt f

xt f

(best )

( worst )

(mean)

( st.d )

1.0856 0.7254 0.05

5.6254 4.4673 2.73

3.4583 2.3367 1.4963

1.2139 1.2542 0.8078

Units m m/s deg

Figs. 7-10 explain the parameter variations for single thrust case of the launch vehicle. From Fig. 7 it is clear that in both algorithms, the terminal points are achieving successfully. Similarly Figs. 9, 10 describe the variation of vehicle velocity and flightpath angle. From these figures we can conclude the state variables are achieving their targets with in tolerance level.

Fig. 4. Velocity variation with time for single thrust condition

Similarly Figs. 5, 6 describe the variation of vehicle velocity and flightpath angle. It is shown that the state variables are achieving their targets within allowable tolerance band. Fig. 4 describes the control variation. Since the accuracy of the state variables are more, the control action required is very high.

Fig. 7. Altitude variation with time for dual thrust condition

Fig. 5. Flightpath angle variation with time for single thrust condition

Fig. 8. Velocity variation with time for dual thrust condition

Fig. 6. Angle of attack variation with time for single thrust condition TABLE VI

NUMERICAL RESULTS OF DUAL THRUST SCENARIO USING APSO State variables r v 

x f

20 10 10

xt f

xt f

xt f

xt f

(best )

( worst )

(mean)

( st.d )

8.3276 0.2034 4.76

15.8461 7.4283 8.64

12.2620 4.8362 6.5010

2.2766 2.0801 1.2466

Units m m/s deg

Fig. 9. Flightpath angle variation with time for dual thrust condition



205


[3]

[4]

[5]

[6]

[7]

[8]

Fig. 10. Angle of attack variation with time for single thrust condition

Fig. 8 describes the control variation and it is evident that since the accuracy of the state variables are more, the control action required is very high. This is the only drawback which we can find in the TH-PSO, numerically it is measured as 5% more. From the graphical representation of the variables in both algorithms it is concluded that the smoothness in control variable is more in the case of theta-PSO. From the above results it is evident that theta-PSO is having high consistency rate than APSO.

[9]

[10]

[11]

[12]

V.

Conclusion

In this paper, application of TH-PSO in trajectory optimization for an initial stage of a multi stage liquid propellant launch vehicle and it’s comparison with APSO has been brought out. The novelty lies in the fact that the use of TH-PSO in trajectory optimization is first of its kind and comparison with already existing algorithm- APSO opens up new dimensions in exploring other similar techniques which may yield better and consistent estimate. From the result presented in the previous section strongly supports the use of TH-PSO over APSO, since afore mention method yielded accurate and consistent results with almost equal level of computations involved. Time taken for the simulation is almost same for all algorithms. The proposed algorithm is also tested in a two thrust scenario and from the obtained results it is concluded that TH PSO outperformed its APSO counterpart.

[13]

[14]

[15]

[16]

[17]

[18]

Acknowledgements

[19]

We would like to thank Manipal University for providing all kind of support for fulfilling this research.

[20]

[21]

References [1]

[2]

Mehta, R., Aerodynamic Design of Payload Fairing of Satellite Launch Vehicle, (2015) International Review of Aerospace Engineering (IREASE), 8 (5), pp. 167-173. Remesh, N., Ramanan, R., Lalithambika, V., Fuel Optimum Lunar Soft Landing Trajectory Design Using Different Solution Schemes, (2016) International Review of Aerospace Engineering (IREASE), 9 (5), pp. 131-143.

[22]

[23]


Martin S. K. Leung and Anthony J. Calise., Hybrid approach to near-optimal launch vehicle guidance, ,Journal of Guidance, Control, and Dynamics, Vol. 17, No. 5 (1994), pp. 881-888. John T. Betts., Survey of Numerical Methods for Trajectory Optimization. Journal Of Guidance, Control, And Dynamics Vol. 21, No. 2, AIAA 1998. Bruce A. Conway., A Survey of Methods Available for the Numerical Optimization of Continuous Dynamic Systems. Journal Optimum Theory Application (2012) 152:271–306 Shan, Jinjun, Yuan Ren. Low-thrust trajectory design with constrained particle swarm optimization. Aerospace Science and Technology 36 (2014): 114-124. Eberhart, R. C, Shi,Y., Comparison Between Genetic Algorithms and Particle Swarm Optimization, Evolutionary Programming VII, Lecture Notes in Computer Science, Vol. 1447, Springer, New York, 1998, pp. 611–616. Kennedy J, and Eberhart, R., Particle Swarm Optimization, Proceedings of the IEEE International Conference on Neural Networks, Inst. of Electrical and Electronics Engineers, Piscataway, NJ, 1995, pp. 1942–1948. Eberhart, R., Kennedy J., A New Optimizer Using Particle Swarm Theory, Proceedings of the Sixth International Symposium on Micromachine and Human Science, Nagoya, Japan, 1995, pp. 39– 43. Kalivarapu, V., Winer, E., Implementation of Digital Pheromones in Particle Swarm Optimization for Constrained Optimization Problems, 49th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, Schaumburg, IL, AIAA Paper 2008-1974, 2008. M. Clerc, J. Kennedy, The particle swarm-explosion, stability, and convergence in a multidimensional complex space, IEEE Transactions on Evolutionary Computation 6 (2002) 58–73. Trelea, I.C., The particle swarm optimization algorithm: convergence analysis and parameter selection. 2003 Inf. Process. Lett., 85(6):317-325. doi:10.1016/S0020-0190(02) 00447-7] W. Zhong, S. Li, F. Qian, θ-PSO: A New Strategy of Particle Swarm Optimization, Zhong et al. / J Zhejiang Univ Sci A 2008 9(6):786-790. Feeley, T. S., Speyer, J. L., Techniques for Developing Approximate Optimal Advanced Launch System Guidance, Journal of Guidance, Control, and Dynamics, Vol. 17, No. 5, 1994, pp. 889-896. Ping Lu., Nonlinear trajectory tracking guidance with application to a launch vehicle, Journal of Guidance, Control, and Dynamics, Vol. 19, No. 1 (1996), pp. 99-106. doi: 10.2514/3.21585 M. Clerc, J. Kennedy, The particle swarm-explosion, stability, and convergence in a multidimensional complex space, IEEE Transactions on Evolutionary Computation 6 (2002) 58–73. Ramya, S., Rajesh, N., Viswanathan, B., Vigneswari, B., Particle Swarm Optimization (PSO) based optimum Distributed Generation (DG) location and sizing for Voltage Stability and Loadability Enhancement in Radial Distribution System, (2014) International Review of Automatic Control (IREACO), 7 (3), pp. 288-293. Y. Shi, R.C. Eberhart, A modified particle swarm optimizer, in: Proceedings of the Congress Evoluationary Computer, 1998, pp. 69–73. R. C. Eberhart and Y. Shi, Tracking and optimizing dynamic systems with particle swarms, in Proc. Congr. Evol. Comput., 2001, pp. 94–100. F. van den Bergh, A.P. Engelbrecht, A cooperative approach to particle swarm optimization, IEEE Transactions on Evolutionary Computation 8 (3) (2004) 225–239 J. Kennedy, Small worlds and mega-minds: effects of neighborhood topology on particle swarm performance, in: Proceedings of IEEE Congress on Evolutionary Computation, 1999, pp. 1391–1938 J. Kennedy, R.C. Eberhart, Particle swarm optimization, in: Proceedings of the IEEE International Conference on Neural Networks, 1995, pp. 1942–1948. Y.P. Chen, W.C. Peng, M.C. Jian, Particle swarm optimization with recombination and dynamic linkage discovery, IEEE


206


[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

Transactions on Systems, Man and Cybernetics – Part B: Cybernetics 37 (6) (2007) 1460–1470. J. J. Liang, A. K. Qin, P. N. Suganthan, S. Baskar, Comprehensive learning particle swarm optimizer for global optimization of multimodal functions, IEEE Transactions on Evolutionary Computation 10 (2006) 281–295. S. Y. Ho, H. S. Lin, W. H. Liauh, S.H. Ho, OPSO: orthogonal particle swarm optimization and its application to task assignment problems, IEEE Transactionsion Systems, Man and Cybernetics – Part A: Systems and Humans 38 (2) (2008) 288–298. C. Li, S. Yang, An adaptive learning particle swarm optimizer for function optimization, in: Proceedings of the Congress Evoluationary Computer, 2009, pp. 381–388. C. Li, S. Yang, T.T. Nguyen, A self-learning particle swarm optimizer for global optimization problems, IEEE Transactions on Systems, Man, and Cybernetics – Part B: Cybernetics 42 (3) (2012) 627–646. Zhong, Weimin, Jianliang Xing, Hongwei Ge, and Feng Qian. An Improved θ-PSO Algorithm with Mutation. In Symposium of 2007 International Conference on Intelligent Systems and Knowledge Engineering, pp. 1450-1455. Chengdu, 2007. H. Shayeghi, H. A. Shayanfar, A. Safari, Damping controller design for TCSC using theta-particle swarm optimization., Journal of Applied Science 11(16): 2924-2931, 2011. Amin Safari, θ - PSO Algorithm for UPFC Based Output Feedback Damping Controller., International Journal of Control and Automation Vol. 6, No. 1, February, 2013. Mehdi Derafshian Maram, Nima Amjady, Abbas Rezaey An Optimal Load Cut Policy with Event-Driven Design against Voltage Instability Using Theta-Particle Swarm Optimization, J. Basic. Appl. Sci. Res., 3(3)91-100, 2013. Vahid Hosseinnezhad, Ebrahim Babaei., Economic load dispatch using θ - PSO., Electrical Power and Energy Systems 49 (2013) 160–169. Haiyan Lu, Weiqi Chen Dynamic-objective particle swarm optimization for constrained optimization problems., Journal of Combinatorial Optimization December 2006, Volume 12, Issue 4, pp 409-419. Mengqi Hu, Teresa Wu, and Jeffery D. Weir Adaptive Particle Swarm Optimization With Multiple Adaptive Methods., IEEE Transactions On Evolutionary Computation, Vol. 17, No. 5, October 2013. Ammar, Y., Boudghene Stambouli, A., Bekhti, M., Design and Optimization of Microsatellite Power System, (2015) International Review of Aerospace Engineering (IREASE), 8 (4), pp. 141-150. Omar, H., Developing Geno-Fuzzy Controller for Satellite Stabilization with Gravity Gradient, (2014) International Review of Aerospace Engineering (IREASE), 7 (1), pp. 8-16. Kassem, A., El-Bayoumi, G., Habib, T., Kamalaldin, K., Improving Satellite Orbit Estimation Using Commercial Cameras, (2015) International Review of Aerospace Engineering (IREASE), 8 (5), pp. 174-178. Bousson, K., Gameiro, T., A Quintic Spline Approach to 4D Trajectory Generation for Unmanned Aerial Vehicles, (2015) International Review of Aerospace Engineering (IREASE), 8 (1), pp. 1-9.

Authors’ information Dileep M. V. received a bachelor of technology(B. Tech) in electronics and communication Engineering from Kerala university, Kerala, India in 2010, a Master of technology in astronomy and space engineering from Manipal university, Karnataka, India in 2012 and now pursuing Ph. D in Department of instrumentation and control Engineering, Manipal Institute of Technology, Manipal University, Karnataka, India. His area of interest is control systems and optimization algorithms with a focus on aerospace applications. Surekha Kamath is currently working as an Associate Professor in the department of Instrumentation and Control Engineering at Manipal Institute of Technology, Manipal University, Manipal. She has obtained her B.E in Electrical and Electronics from Mysore University and her M.Tech in Biomedical Engineering from M.I.T MANIPAL. She has completed her Ph.D in the year 2010 from Manipal University. She has got more than 23 years of teaching experience in various institutes. Vishnu G. Nair is currently working as an Assistant Professor in the Department of Aeronautical & Automobile Engineering Institute of Technology, Manipal University, Manipal. He received a bachelor of technology (B. Tech) in electronics and communication Engineering from Kerala university, Kerala, India in 2010, a Master of technology in astronomy and space engineering from Manipal university, Karnataka, India in 2012 and now pursuing Ph. D in Department of mechanical engineering, National institute of technology, Surathkal, karnataka, india.



207


Two Dimensional Numerical Study of Aerodynamic Characteristic for Rotating Cylinder at High Reynolds Number Alias M. S., Mohd Rafie A. S, Wiriadidjaja S. Abstract – Efforts in this century for Unmanned Aerial Vehicle, UAV aerodynamic technology led to a broad of applications. Currently, UAV users are demanding of small, unprepared field or even no field for the aircraft to take off and landing operation. Aligned for the needs, several studies revealed the feasibility of rotating cylinders produced lifting which will impact the improvement of on lift coefficient. Magnus effect on rotating cylinder has the potential as a good lift generator. The studies have discovered the limitation on implementation discovered caused by induced and parasite drag occurrences. Accordingly, rotational rate, α, and Reynold number, Re, are the highlight in this study. The previous experimental and numerical data were used as a basis to compare the results. The methodological approach used for this research in order to prove the presence of Magnus effect, Finite Element Numerical Analysis method in form of 2D numerical is chosen and the simulation done by using ANSYS FLUENT R15.0 to examine the coefficient of lift, drag and understand the aerodynamic characteristics of the rotating cylinder surfaced body. Previous experimental studies carried out by Elliott G. Reid simulated on-design in 2D numerical analysis for validation. The results obtained showed 90.6% accuracy for the validation where the cylinder size to be tested was smaller compared to on-design size. The cylinder size of 30mm as adapted to J. Seifert studies on Magnus effect is used to compare the original size of 114.3mm where the Reynold number tested at the range of 1.17×103 ≤ Re ≤ 1.69×105 with rotational rate ranging from 0 ≤ α ≤ 4.32 determined by air velocity range within 5 ms-1 ≤ U ≤ 15 ms-1. Lift coefficient, CL and drag coefficient, CD determined in every stage analysis were recorded. The results obtained showed that the lift coefficient is slightly lower compared to the original size of cylinder at U are 5ms-1, 7ms-1, 10 ms-1 and 15 ms-1. However, the drag coefficient showed higher value U of 15 ms-1 and 10 ms-1 but lower at U of 5 ms-1 and 7 ms-1. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: UAV, Magnus Effect, Rotating Cylinder, Reynolds Number, Rotational Rate, Coefficient of Lift, Coefficient of Drag, 2D Numerical Simulation

ν μ u* Cf τw Uτ U

Nomenclature A CL CD CT CP d R L n Re UC τ α ρ ω S q y γ

Aspect ratio Lift coefficient Drag coefficient Torque coefficient Specific heat capacity Cylinder diameter Cylinder radius Lift Power law index Reynolds number Airspeed Torque Rotational rate Air density Angular frequency Platforms area Fluid dynamic pressure Distance to the nearest wall Adiabatic coefficient

Kinematic viscosity Dynamic viscosity Friction velocity at the nearest wall Skin friction coefficient Wall shear stress Frictional velocity Inlet velocity

I.

Introduction

During World War 1, the Unmanned Aerial Vehicles (UAVs) or Drones were developed. UAVs have the capability to transmit data onto real-time intelligence in the battlefield and also processed data information such as surveillance and reconnaissance. While the combat type UAVs can perform communication relays, assets neutralized target designation, attacking by its inboard munition as well overviews battle information about by risking any aircrew [1]. In 2003, the quantity of UAV used in military did not have much impact.


DOI: 10.15866/irease.v9i6.10774

208

Alias M. S., Mohd Rafie A. S, Wiriadidjaja S.

However, it increased rapidly in reconnaissance operation with 32% unmanned aerial vehicles compared to 68% manned aircrafts [2]. Irizarry, Javier mentioned in his research that Unmanned Aerial System (UAS) which control centers of a few UAVs operations can perform tasks which are similar to those that can be done by manned vehicles, often faster and safer at lower cost either at the initial stage of development [3] or on its application whereby in investigation tasks in border patrol, search and rescue, damage investigations into or after natural disasters, locating forest fires or farmland frost conditions, monitoring criminal activities, mining activities, advertising, scientific surveys, and in securing pipelines and offshore oil platforms [4]. Back in 2006, the Coast Guard acquired Bell Helicopter Textron‟s Eagle Eye UAV as part of Deepwater Modernization program. The cost was about 3 million USD; Eagle Eye takes off like an helicopter but then it tilts up its rotor to fly like a plane. Since it is capable to patrol the U.S. coastline for drug smugglers, refugees and ships in distress, it also transmits video and infrared images to the cutter and command centers ashore [5] but it is still costly in this era of economic crisis for most countries to be implemented. The task of extending the surveillance capability of cutters led the UAV fly up to 113.178ms-1 and 482.803km operate radius [2]. Helicopter implies rotor blades of vertical take-off but it has low efficiency for cruising to high speed and unable to fly for high altitude flight. It is fully proven on vertical take-off and landing operation aircraft. However, it gave several disadvantages because of the mechanical design simplicity, noise emission, stall potential whenever speed are concerned, lower cruising speed, unsafe for high altitude manure for power back operation and costly for production and maintenance [5]. Rotarywing configuration influenced its auto-gyro which attempts to dispense with the transmission system of the helicopter whenever reducing complexity is not possible and suffers from hovering and considerably fly slower compared to fixed-wing type aircraft. Front-line aircraft such as UAV currently demands on small airport, unprepared field or even un-field for having the technology [6]. Many improvements have been made concerning a few major factors affecting lift, such as thrust loading, wing loading and lift coefficient at takeoff state that can be considered as primary components of climbing up however affected the thrust to weight ratio, lift to drag ratio as avionic systems embedded directly gained the gross weight of the aircraft [7]. Previous researchers discovered the potential for S/VTOL for flight operation by introducing Magnus effect as lifting device. However, technology limited to lower efficiency than existing propulsion caused by inherent loss regions, the so-called drag, whether induced or parasite, lower cruising speed, heavy in total mass and difficulties in landing [8]. Magnus Effect is the effect of moving airstream to the spinning ball or cylinder where the potential flow theory predicts zero drag force (CD = 0) and a lift force given by Kutta-Joukowski theorem;

(1) Hence, the potential flow theory predetermined lift coefficient given by; (2) where R is the cylinder radius, ω is the angular frequency of the cylinder and UC is the air speed. This theoretical value would give higher results than the experimental ones reached by S. Carstensen, a discrepancy that is primarily due to viscosity [9]. However, the measured variation on CL with respect to the speed ratio is to some extent in accordance with the theoretically predetermined linear variation at least over a certain range of speed ratios and we concluded that, the lower the aspect ratio the lower the lift coefficient for the same flow conditions and spinning speed, as for M. Principi [10]. The velocity distribution of the cylinder wakes at the back of the rotating cylinder, recirculation length follows the rate of rotation of fluid at the area and the separation angle from laminar flow transiting to turbulent flow [11]. Stagnation point also is one of the considerations whereby vorticity is a necessity to move the stagnation point obtained by irrotational flow calculation to the correct trailing edge point which depends on the wind angle of incidence. N. Rostamy carried out a research for local flow field and rear wake of a surface-mounted finite circular cylinder in low speed wind tunnel whereby Re was set at 4.2×104 with different aspect ratio by using Particle Image Velocimetry (PIV). The result is that higher aspect ratio may increase higher turbulent and higher up-wash formed with flow separation occurred at the leading edge free-end surface and a mean recirculation formation [12]. Separation points by A. Roy were identified on the basis of the highest pressure coefficient on the cylinder surfaces in their research [13].

II.

Previous Work

An aircraft designed with short field take-off and landing capabilities required to have the ability to fly slow [6]. A potential lift generator is the Magnus effect of a spinning cylinder with constant rotational rate while a turbulent flow flew through it [14]. It has been proven by several inventions of Magnus rotor wing aircrafts in the early 1920s but developments have stopped due to high costs [15]. However, researchers continue to develop fan wing aircraft which increase the lift coefficient but have difficulties for maneuvering in low speed, caused by hard landing [8] and instabilities associated with short period and phugoid modes compared to aerodynamic vectored type aircrafts [16], [22]-[25]. In 2012 to 2014, inventors designed small scale size aircrafts named as rotor wing aircrafts which used the concept of Magnus rotor wing in order to prove flight feasibility by using the Magnus effect principle, the result was promising for take-off and even landing in short distance. Apparently, a gap still



209


exists on the research which focused on spinning cylinder as good lift generators. The research on rotating cylinder gave benefits to many industries and even countries for the effect caused by the cylinder rotation through fluid, at specific spinning velocities with suitable Reynolds number, Re whereby achieving broader insight to maneuverability controllability and stability in lateral motion and longitudinal control for submarines and ships. As an inference, the rate of lifting is nearly independent due to the angle of attack and the angle of incidence, UAV maneuverability, controllability and stability which using rotating cylinder can be manipulated by the study of angular velocity, ω as well by involving Reynold number, Re. In this work, the limitation of the current Magnus Effect concept in aircraft flight operation can be improved by analysing the optimum parameters focusing on incompressible unsteady airflow. J. Seifert successfully experimented the effect of Magnus Rotor in Micro Air Vehicle. It should be 0.05 kg of payloads with a span size for rotating cylinder surfaces and about 580 mm where cylinder diameter of 30 mm which gave the ratio of 50:0.58 for coefficient of forces (see Fig. 1). Therefore, the size of cylinder and payload is an important factor contributing to MAV lifting [17].

provide aerodynamic force when it interacts with a moving stream of air. Objective was addressed according to current studies to analyze on-design 2D numerical simulation on rotating cylinder size effect compared to the experimental data.

III. Methodology The element to analyze is the fluid flow characteristic on rotating cylinder by using ANSYS Fluent 15.0. Since the work design was neglecting air flow interference from outside to inside and vice versa region, 2D case is concerned for numerical study. In this work, Elliott G. Reid [20] experimental result has been numerically simulated and validated. A well-recognized statistical analysis software, Statistical Package for the Social Sciences (SPSS Statistics), has been used. The aimed to use this is to select the best and most accurate parameters set and to have comparative results for reference. Flow structures are discussed and comparisons are made between coefficients of lift since this work focused on lift force produced. The findings from the numerical simulation tests carried out were recorded and analyzed. As per set, fluid was assumed to be an ideal gas whereby having the constant specific heats capacity. The air ideal gas at 1 ATM pressure in 15°C temperature constitutes the following properties, specific heat, CP = 1.005×103 J kg-1 K-1, adiabatic coefficient, γ: 1.404, Prandtl number: 0.717, kinematic viscosity, ν: 1.466×10-5 kg m-1 s-1, dynamic viscosity, μ: 1.789×10-5 kg m-1 s-1 and air density, ρ: 1.225 kg m-3. Flow behavior near the wall is a complicated fact and to distinguish the different regions near the wall, concept of Y+ has been formulated. Thus, Y+ is a dimensionless quantity, and is the distance from the wall measured in the terms of viscous length. The concept of Y+ and the law of wall were introduced by Theodore Von Karman in 1930 as the universal law of the turbulent wall boundary layers. It is related to turbulence modelling especially mesh generation processes and how it is going to affect the flow result of the simulation [21]. A non-dimensional wall distance between a wall bounded flows can be determined by;

Fig. 1. Remotely Piloted Aircraft Spicy (Spinning cylinder)

Previous study on supercritical flow passing a cylinder revealed the fully turbulence of boundary layer by using several Reynold numbers and rotational rates ranging 0 ≤ α ≤ 5. The study was extended due to the investigation into the rotational effects on higher rotational rates than those considered to date [18]. Investigated by previous researcher, the range that can give contribution for rotating cylinder flight were the parameters of PowerLaw Index of 1 ≥ n ≥ 0.2, Reynolds number of 0.1 ≤ Re ≤ 40 and rotational rate of 0 ≤ α ≤ 6 but no lifting was produced with positive drag. However, manipulation on parameters may result differently e.g. the increase in Re value and α for negative drag production [19]. Therefore, the identified gap between knowledge is enhancement of the aerodynamic principle aligned with the needs of the UAVs to be able to operate on front-liner consumers. Hence, optimum parameters and significant variables are the concern in order to improve current UAV operation to meet economical, efficiency and safety desires. Previous works and conventional type airfoils

(3) where is the friction velocity at the nearest wall, y is the distance to the nearest wall and is the local kinematic viscosity of the fluid. often referred to as y plus and is commonly used in boundary layer theory and in defining the law of the wall. To calculate cell height, initial calculation of Reynolds number is necessary based on the characteristics scales of geometry. The calculation of wall height starts from the Skin Friction Coefficient, Cf equation. As external flows: (4)



210


The wall shear stress τw can be calculated from skin friction coefficient whereby:

is performed. This process does not include pseudo-time terms and therefore the equations are solved implicitly with direct discretization and formulation of the algebraic system. On solving the equations in unsteady form, second-order discretization scheme is used to integrate the equations in time. This scheme is implicit and solving process is done for numeric stability. Time step is chosen based on the dimensionless step [18]. The solution method is one of the important methods for solving the analysis, pressure to velocity scheme set to „coupled‟. On spatial discretization, the gradient is „least square cell based‟, „second order pressure‟, „second order upwind‟ momentum. The turbulent kinetic energy selected „first order upwind‟ by specific dissipation rates. Finally, the calculation is run using the time stepping method set to fixed on stepping within time step size (s) = 1 with 0 number time steps. For overall calculation, maximum iterations over time step are set at 500.

(5) Therefore, the calculation of frictional velocity, Uτ:

(6)

Finally, the wall height estimation equation: (7) The flow analyses were carried out using commercial CFD package ANSYS CFX which executed 3D RANS equations based on the finite volume numerical method. The flow domain boundary condition is illustrated in Fig. 2 and Fig. 3 in which are represented the geometry model and domain discretization respectively. Solver type used is pressure based on absolute velocity formulation, time considered at transient on 2D space planar. On model, viscous air flows reflected on shear-stress transport with K omega, K-ω [26]for rotated object with defined air density, ρ of 1.225 kg m-3 and viscosity, μ is 1.789×10-5. Cylinder as a wall, rotated cylinder momentum parameter is clarified for moving wall which is relative to adjacent cell zone with set rotational speed. To prevent from any other disturbance, the shear conditioned set for no-slip condition. Inlet velocity momentum as velocity specification method of magnitude and direction with reference frame is absolute to specific velocity magnitude according to the validation values required. Since the air flow considered is turbulent flow, intensity of 5% and viscosity ratio of 10 are used. Outlet velocity while backflow direction specification method defined significantly normal to its boundary. The reference values regard for the affected area determined by πr. Algebraic equations are solved in iterative manner resulting in velocity, and pressure fields are updated together. Fully implicit coupling is achieved by implicitly discretizing the pressure gradient terms with momentum equations, including the implicit discretizing mass flux. All the procedures are linked with an algebraic multigrid method and the set of equations is solved by a point or block Gauss-Seidel technique and as for spatial discretization, the second order upwind scheme is selected for momentum, first order upwind for turbulent kinetic energy and specific dissipation rate to discretize the convective terms of the governing equations, while the second-order central scheme discretizes the diffusion terms. Accordingly, to avoid checker boarding for the implemented co-located grids, interpolation for the computation of the mass fluxes in the continuity equation

Fig. 2. Geometry model on cylinder size change from 114.3mm (original form experiment [20]) to 30mm

Fig. 3. Domain discretization of modified cylinder size from original diameter



211


IV. Results and Discussion Model

The validation model Elliott G. Reid uses is the test of rotating cylinder. It was initially a meshed model with 62345 elements (refer Table I).

dimension0 1 .952a .906 .901 .331091 a. Predictors: (Constant), Rotational Speed, Air Velocity b. Dependent Variable: Rotational Rate

TABLE I MESHING STATISTICS Nodes Elements Min Max Average Standard Deviation

TABLE II MODEL SUMMARYb ON VALIDATION R R Square Adjusted R Std. Error of the Square Estimate

Expect cumulative problem with observed cumulative problem plot (analyzed from experiment carried out by Elliott G. Reid vs 2D numerical simulation works) shown that data lie within the linear line which is accurate and data is within the range of analysis as illustrated in Fig. 6. The result showed a good agreement on the pattern of experiment and 2D simulation using Computational Fluid Dynamic software to analyze flow over the rotated cylinder.

62704 62345 0.590 0.999 0.986 2.3×10-3

The validation justified on the orthogonal quality values close to 1 which determined good quality of meshes. This validation is important in order to simulate on higher rotational rate, α. The comparison between experimental and 2D numerical simulation was carried out as referred to in Fig. 4 and Fig. 5 at inlet velocity of 15 ms-1. After the statistical analysis carried out, result from model validation gave 90.6% of confidence about the correctness of the testing referring to in Table II.

Fig. 6. P-P plot Expected Cumulative Problem vs Observed Cumulative Problem

The results obtained showed positive agreement on both experimental and 2D numerical work based on the trend and pattern generated. Coefficient of lift, CL illustrated in Fig. 7, Fig. 8, Fig. 9 and Fig. 10 showed slightly lower CL value compared with 114.3 mm and 30 mm cylinder size whereby the air velocity range from 7 ms-1 ≤ U ≤ 15 ms-1 and rotational rate, α of 0.01 to 3.07. At this stage, different size of the cylinder can be considered not affecting the CL. However, the CL started to decrease from low air velocity which 5 ms-1 applied with higher rotational rate ranging from 2.16 to 4.32.

Fig. 4. Graph of CL vs Rotational Rate of experimental and 2D numerical simulation on rotating cylinder at air velocity of 15ms-1

Fig. 5. Graph of CD vs Rotational Rate of experimental and 2D numerical simulation on rotating cylinder at air velocity of 15ms-1 Fig. 7. Graph coefficient of lift, CL at air velocity = 15 ms-1



212


Coefficient of Lift at Vinlet of 10 ms-1

Coefficient of Lift, CL

3,000 2,500 2,000 1,500 1,000 D = 0.03m D = 0.1143m

0,500 0,000

Fig. 11. Graph coefficient of drag, CD at air velocity = 15 ms-1 0,78 0,9 1,02 1,14 1,26 1,38 1,5 1,62 1,74 Rotational Rate, α

Fig. 8. Graph coefficient of lift, CL at air velocity = 10 ms-1 Coefficient of Lift at Vinlet of 7 ms-1

Coefficient of Lift, CL

6,000 5,000 4,000 3,000 Fig. 12. Graph coefficient of drag, CD at air velocity = 10 ms-1

2,000 D = 0.03m D = 0.1143m

1,000 0,000 1,54

1,79

2,05

2,3

2,56

2,82

3,07

Rotational Rate, α Fig. 9. Graph coefficient of lift, CL at air velocity = 7 ms-1

Fig. 13. Graph coefficient of drag, CD at air velocity = 7 ms-1

Fig. 10. Graph coefficient of lift, CL at velocity = 5 ms-1

Different phenomenon identified in Fig. 11, Fig. 12, Fig. 13 and Fig. 14 whereby drag coefficient showed higher value for 30 mm cylinder diameter compared to 114.3 mm for air velocity ranging from 7 ms-1 to 15 ms-1 whereabouts the set rotational of 0.01 to 1.79.

Fig. 14. Graph coefficient of drag, CD at air velocity = 5 ms-1



213


Yet the CD rapidly decreased from α of 2.05 to 3.07 for 7 ms-1 testing. At 5 ms-1 testing, the CD values have decreased similar to 7 ms-1 testing compared to original experimental value where α of 2.51 up to 4.32.

References [1]

[2]

V.

Conclusion

[3]

The main objective of the thesis is to see the effect of smaller cylinder size of high Reynolds number. It is achieved by validating previous work carried out by other researchers on the effect of rotating cylinder at high Reynolds number in two dimensional numerical analyses and experiments. These validations were crucially analyzed and selected with the best accuracy in order to re-use the method of systematic numerically investigations to understand the impact on the cylinder effect (from 114.3 mm to 30 mm). The statistical regression analysis showed 90.6% accuracy. The result showed a good agreement on simulation onto 30 mm cylinder size graph pattern compared to simulation data onto 114.3 mm cylinder size. This result showed an agreement with previous researcher studies to reveal Magnus effect feasibility on MAV using 30mm cylinder radius [17]. Hence, we concluded that higher Reynolds numbers 1.17×103 < Re < 1.36×105 with lower rotational rate 0.01 < α < 1.74 gave positive impact coefficient of lift, CL while lower Reynolds number 8.44×104 < Re < 1.69×105 with 1.54 < α < 4.32 gave negative impact expectation of lift, CL which was considered as the week point of the lift force generated. In this study, 2D numerical simulation has been introduced limited to the study between 1.17×103 ≤ Re ≤ 1.69×105, rotational rate ranging from 0 ≤ α ≤ 4.32 and air velocity range within 5 ms-1 ≤ U ≤ 15 ms-1. It is advisable to carry out a 3D numerical simulation as well experimenting the model in order to capture better flow field behavior and aerodynamic characteristic knowledge more in details.

[4]

[5] [6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14] [15]

Acknowledgements

[16]

The author would like to thank Associate Prof. Dr. Azmin Shakrine Bin Mohd Rafie for introducing me to this interesting topic. Associate Prof. Dr. -Ing. Surjatin Wiriadidjaja is gratefully acknowledged for his support with the aerodynamic characteristics and application. The door to their office was always open whenever I ran into a trouble spot or had a question about my research or writing. They consistently allowed this paper to be my own work, but steered me in the right direction whenever they thought I needed it. Finally, I must express my very profound gratitude to my beloved spouse Melissa Ang Bt. Azlan and my beloved daughter Nur Qaira Adriana Bt. Mohd Shahidi for providing me with unfailing support and continuous encouragement throughout my years of study and through the process of researching and writing of this thesis. This accomplishment would not have been possible without them.

[17]

[18]

[19]

[20] [21]

[22]

[23]


J. F. Keane and S. S. Carr, “A Brief History of Early Unmanned Aircraft,” Johns Hopkins APL Technical Digest, vol. 32, no. 3, pp. 558–571, 2013. E. Bone and C. Bolkcom, “Unmanned aerial vehicles: Background and issues for congress,” Library of Congress, 2003. Marzouk, F., Boukhdir, K., Medromi, H., Design, Modeling and Realization of a UAV Based on Multi-Agent Systems on an Embedded Platform, (2014) International Journal on Engineering Applications (IREA), 2 (6), pp. 189-194. J. Irizarry and E. N. Johnson, “Feasibility study to determine the economic and operational benefits of utilizing unmanned aerial vehicles (UAVs)”, Technical Report, Georgia Institute of Technology. School of Aerospace Engineering, 2014. T. Robinson, “Revolution in the air,” Aerosp. Int., p. 20, 2004. J. K. Wimpress, “Short Take-off and Landing for High Speed Aircraft,” Aircraft Engineering and Aerospace Technology 1966 38:6 , 14-19. Boukhdir, K., Boualam, A., Tallal, S., Medromi, H., Benhadou, S., Conception, Design and Implementation of Secured UAV Combining Multi-Agent Systems and Ubiquitous Lightweight IDPS (Intrusion Detection and Prevention System), (2015) International Journal on Engineering Applications (IREA), 3 (1), pp. 1-5. Thong Q. Dang, Peter R. Bushnell, Aerodynamics of cross-flow fans and their application to aircraft propulsion and flow control, Progress in Aerospace Sciences, Volume 45, Issues 1–3, January– April 2009, Pages 1-29. S. Carstensen, X. Mandviwalla, L. Vita, and U. S. Paulsen, “Lift of a Rotating Circular Cylinder in Unsteady Flows,” The Twentysecond International Offshore and Polar Engineering Conference, 17-22 June, Rhodes, Greece. M. Principi and S. Prince, “Feasibility Assessment of Spinning Cylinder Lift for Remote Flight in Mars and Titan Atmospheres,” ICAS2014, pp. 1–16, 1877. M. Abrahamsen Prsic, M. C. Ong, B. Pettersen, and D. Myrhaug, “Large Eddy Simulations of flow around a smooth circular cylinder in a uniform current in the subcritical flow regime,” Ocean Eng., vol. 77, pp. 61–73, 2014. N. Rostamy, D. Sumner, D. J. Bergstrom, and J. D. Bugg, “Local flow field of a surface-mounted finite circular cylinder,” J. Fluids Struct., vol. 34, pp. 105–122, 2012. A. B. Harichandan and A. Roy, “Numerical investigation of flow past single and tandem cylindrical bodies in the vicinity of a plane wall,” J. Fluids Struct., vol. 33, pp. 19–43, 2012. T. Reynolds, “Flow past a rotating cylinder,” vol. 476, pp. 303– 334, 2003. J. Seifert, “A review of the Magnus effect in aeronautics,” Prog. Aerosp. Sci., vol. 55, pp. 17–45, 2012. Manzoor, M., Maqsood, A., Hasan, A., Quadratic Optimal Control of Aerodynamic Vectored UAV at High Angle of Attack, (2016) International Review of Aerospace Engineering (IREASE), 9 (3), pp. 70-79. J. Seifert, “Micro Air Vehicle lifted by a Magnus Rotor,” 50th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, no. January, pp. 1–10, 2012. S. J. Karabelas, B. C. Koumroglou, C. D. Argyropoulos, and N. C. Markatos, “High Reynolds number turbulent flow past a rotating cylinder,” Appl. Math. Model., vol. 36, no. 1, pp. 379– 398, 2012. S. K. Panda and R. P. Chhabra, “Journal of Non-Newtonian Fluid Mechanics Laminar flow of power-law fluids past a rotating cylinder,” J. Nonnewton. Fluid Mech., vol. 165, no. 21–22, pp. 1442–1461, 2010. E. G. Reid, “Tests of rotating cylinders,” 1924. E. R. Gowree and S. A. Prince, “A Computational study of the aerodynamics of a spinning cylinder in a crossflow of high Reynolds number,” in Proceedings of the 28th ICAS International Congress of the Aeronautical Sciences, 2012. Aziz, M., Elsayed, A., CFD Investigations for UAV and MAV Low Speed Airfoils Characteristics, (2015) International Review of Aerospace Engineering (IREASE), 8 (3), pp. 95-100. Deif, T., Kassem, A., El Baioumi, G., Modeling and Attitude


214


Stabilization of Indoor Quad Rotor, (2014) International Review of Aerospace Engineering (IREASE), 7 (2), pp. 43-47. [24] Bousson, K., Gameiro, T., A Quintic Spline Approach to 4D Trajectory Generation for Unmanned Aerial Vehicles, (2015) International Review of Aerospace Engineering (IREASE), 8 (1), pp. 1-9. [25] Deif, T., Kassem, A., El Baioumi, G., Modeling, Robustness, and Attitude Stabilization of Indoor Quad Rotor Using Fuzzy Logic Control, (2014) International Review of Aerospace Engineering (IREASE), 7 (6), pp. 192-201. [26] Bekka, N., Bessaïh, R., Sellam, M., Numerical Study of Transonic Flows Using Various Turbulence Models, (2015) International Review of Aerospace Engineering (IREASE), 8 (6), pp. 216-224.

Authors’ information Alias M. S. born on 21st February 1987. Education: Bachelor in Aircraft Engineering Technology (mechanical) in University of Kuala Lumpur, Malaysian Institute of Aviation Technology (UNIKL MIAT), Selangor, Malaysia started from 2007 to 2011. Affiliations and functions: 2011 to 2014 – DRB Hicom – Industrial Engineering, Production and Process Division. 2014 to current – Management and Science University Tutor, Instructor and Lecturer of Aircraft Maintenance Technology and Mechanical Engineering. Teaching: DCAM / EASA Modules (Aircraft Hydraulic and Pneumatic System, Air-conditioning, Cabin Pressurization, Oxygen System, Landing Gear and Brakes, Aviation Legislation, Turbine and APU, Propeller) and Mechanical Engineering (Metrology and Instrumentations, Entrepreneurship and Industrial Economy). Present position: Program Coordinator in Department of Aircraft Maintenance Technology in MSU. Interest: Aerodynamics, Aircraft Maintenance Technology and Aircraft engineering Technology. Alias M. S. is the corresponding author and can be contacted at: [email protected] or [email protected]. Dr. Mohd Rafie. A. S born in 1974. Associate Professor at University Putra Malaysia, Faculty of Aerospace Engineering. Education: PhD. (Aerospace), 2007, Universiti Putra Malaysia, M. Eng. (Mechanical), Universiti Teknologi Malaysia, B. Eng. (Aeronautic), 1999, Universiti Teknologi Malaysia. Dip. Mech. Eng. (Aeronautic), 1995, Universiti Teknologi Malaysia. Affiliation and functions: Graduate Member - Board of Engineers Malaysia (BEM), Graduate Member - Institution of Engineers Malaysia (IEM), Member - American Institute of Aeronautics and Astronautics (AIAA), Member - Malaysian Society for Engineering and Technology (mSet), Member - Kelab Sukan dan Sosial Fakulti Kejuruteraan (KSSFK). Interest: Aerodynamics and aeroelasticity. Publications: numerous journal and conference papers (more than 40); present position: Head of Department, Faculty of Aerospace Engineering in UPM and Aerodynamics Laboratory Coordinator for Aerospace Department. Email address: [email protected]. Dr. Wiriadidjaja S. born in 1951. Associate Professor at University Putra Malaysia, Faculty of Aerospace Engineering. Education: Dr.-Ing. Fluid Mechanics, 1996, Technische Universität München, München – Germany, Dipl.-Ing. Aeronautical Engineering, 1979, Technische Universität Berlin, Berlin - Germany. Affiliation and functions: Member, American Institute of Aeronautics and Astronautics (AIAA). Interest: Experimental Aerodynamics and Wind Tunnel Design and Development. Publications: numerous journal and conference papers (more than 10); present position: Lecturer, Department of Aerospace Engineering, Faculty of Engineering, UPM, Consultant to the Aeronautical Research Center, Khartoum, Sudan. Email address: [email protected].



215


Design of High Temperature Six-Phase Starter-Generator Embedded in Aerospace Engine Flur Ismagilov, Vavilov Vyacheslav, Lubov' Roginskaya, Semen Shapiro, Denis Gusakov Abstract – This paper examines the problem of the installation of the electric machine to the high pressure shaft to increase the electrification of aircraft engines and create More Electrical Engine. Different ways of synchronous generator integration in the aircraft engine by worldwide aircraft engines manufacturers was discussed. A new design of the high temperature synchronous generator mounted on high pressure shaft is proposed. To evaluate the effectiveness of synchronous generator the electromagnetic, thermal and mechanical calculations are made. High temperature synchronous generator cooling system was designed and system of mechanical decoupling of the stator at short circuits is proposed. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: More Electrical Engine, Synchronous Generator, High Pressure Shaft



Nomenclature AE MEE HPS LPS SG IM PM P M n Physteresis

khysteresis

B f

Peddy currents keddy currents kexcess eddy currents r Eeddy currents

 fe T Pspecific eddy currents k specific hysteresis

Aircraft Engines More Electrical Engine High Pressure Shaft Low Pressure Shaft Synchronous Generator Induction Machine Permanent magnets Power Electromagnetic torque Rotor speed Hysteresis losses in magnetic core Coefficient characterizing the hysteresis losses in magnetic core Flux density in the magnetic core Magnetization reversal frequency Eddy current losses in the stator magnetic core Coefficient characterizing eddycurrent losses Coefficient characterizing the excess eddy current losses Cobalt alloy resistivity EMF of the eddy currents in the magnetic core Magnetic temperature coefficient of resistance material Magnetic core temperature Specific eddy current losses at 20 °C Coefficient characterizing the hysteresis losses at 20 °C

Pspecific hysteresis

 PM TPM

The temperature coefficient of the hysteresis loop characterizing reduction in its area with increasing temperature Specific hysteresis losses at 20 °C The temperature coefficient of PM resistance PM temperature

I.

Introduction

A significant increase in the cost-effectiveness of civil aircraft and the profitability of air transportation is possible by increasing the environmental friendliness, controllability and fuel efficiency of aircraft engines (AE). There is active work to increase the electrification of AE and the creation of More Electrical Engine (MEE) and to solve this problem by worldwide AE manufacturers such as Rolls-Royce, MTU Aero Engines, United Engine Corporation and PW Canada [1], [2]. Refusal of transmission of Auxiliary Gear-Box between the shaft of an AE and the electric machine is a major conceptual solution for the implementation of MEE. In other words, the machine is mounted directly to the High Pressure Shaft (HPS) or Low Pressure Shaft (LPS) of AE. At the same time environmental conditions for the installation of the machine to LPS did not create any difficulty, unlike HPS, where the ambient temperature is 300-330°C and 5 bars pressure. Therefore, this paper examines the machine mounted on the HPS. HPS rotational speed is variable and changes depending on the flight mission mode (HPS speed of the AE ranges from 9 000 to 15 000 rpm). Both these systems have certain advantages and disadvantages considered in [3] – [6]. In this case, the


DOI: 10.15866/irease.v9i6.10893

216

Flur Ismagilov et al.

use of constant frequency 400 Hz will increase the weight and overall dimensions of aircrafts. This paper examines SG, working with rectifier DC 270 V. This is due to the fact that a frequency greater than 380 Hz can provide a machine with the 6 poles or more at the 9 000 rpm minimum AE rotor speed. Sixpoles SG provide 750 Hz output voltage frequency at maximum rotor speed (15 000 rpm), but 8-poles SG will provide 1 000 Hz output voltage frequency at maximum speed, that is not included within the specified frequency range (380-800 Hz). In other words, only 6-poles machine is suitable for the 380-800 Hz power system, which will provide a lower moment than 8 or 10 pole SG in starter mode. In addition, significant limitations on the SG output voltage quality (set in MIL-STD 704F standard) also place high demands on SG design features working directly on-board network, including requirements for the harmonic composition of the current and voltage, etc. The rectifier should be minimum 12-pulse [7] – [9] in compliance with the MIL-STD 704F standard in part of the higher harmonics attributable to semiconductor converters. Therefore, in modern DC aircraft power systems, an interim inverter converting the three-phase voltage in the six-phase is installed between a threephase machine and a six-phase rectifier. This inverter has its weight and overall dimensions and cost, but the tendency of the aircraft power supply systems is the elimination of the intermediary, through the implementation of the SG with six or more phases. Moreover, multi-phase systems are used in integrated SG to improve reliability, since at the breakage of one phase the other phases remain operational, which is very important for the SG mounted on HPS. Besides the desired phases number is 6, a SG mounted on HPS must be operated at an ambient temperature of 300-350oC in the limited cooling conditions. High degree of SG integration into AE is also a problem. At SG installing on the HPS, the generator should be operated with all AE as a single system, where the emergency situation in the SG causes emergency in AE. It is necessary that the SG had increased safety and reliability in order to avoid accidents in the AE. At the same time, the SG should have, like all aircraft systems, minimal weight and overall dimensions to ensure the effectiveness of such integration and minimal losses on its active elements and the current density (increase in losses and the current density significantly complicate the SG operating temperature conditions). In other words, the development of SG mounted on HPS is a highly complex, non-trivial task. The Politecnico di Torino University developed and studied two types of machine to use in AE: six-phase induction machine (IM) [10], [11], serving as a starter and 6-phase interior PM synchronous motor (IPMSM), where the cooling air is pumped by the air gap between the stator and the rotor [12] – [14] IPMSM installed on the LPC, which allows the electrical start of an AE. Both types of electric machines can’t effectively solve

the problem of creating MEE: IM has significant weight and overall dimensions, as well as low efficiency for aircraft systems [15], [16]. IPMSM has acceptable weight and overall dimensions, but pumping air through the air gap to cool the generator leads to complication of the design, increasing in aerodynamic losses of IPMSM and consequently reducing its efficiency. In addition, the proposed place of installation does not allow the electrical start of an AE depriving the concept of MEE of one of its major benefits [17]. The University of Sheffield developed two machines integrated in an AE (High temperature embedded machine) [18] – [20]. One of them is inductor SG mounted on the HPS, and the other is machine mounted on a LPS. SG mounted on the HPS has rotational speed of 13 500 rpm and a power of 100-150 kilowatts. This machine has 4 phases, 18 poles and 24 teeth; its cooling system is not described. The rotor poles are located in a non-magnetic rim made from heat resistant alloy, Inconel 718, which is part of AE, so that it simplifies the integration of the SG in AE. SG ensures overload up to 110%. The development is made in the interests of RollsRoyce. It is known from general theory of electrical machines that inductor SG have high weight and overall dimensions, and their use as an SG mounted for HPS is probably less efficient (the University of Sheffield did not show weight and overall dimensions of the SG) than the use of the SG with PM as in the work by Politecnico di Torino. Besides, the SG does not provide significant congestion (total 110% of rated power), and a significant number of poles pairs (18, the frequency of magnetic reversal of the stator magnetic circuit will be 2 025 Hz) will increase losses in the SG. Thales AES develops SG for integration into HPS based on machine with high-coercivity PM [1]. The SG is made on the 150 kW power at a rotational speed of 9 000-13 500 rpm at generating mode and provides a 350 Nm moment at a 4 800 rpm rotational speed at the starter mode. SG has an oil cooling and duplicated coil. SG stator and rotor weight is 88 kg [1]. The disadvantages of the SG according to Thales are its significant weight and overall dimensions as well as the complexity of ensuring oil cooling SG in AE. Basic requirements for SG mounted on HPS formulated on the basis of the review and experience of the aircraft machine creation: – The possibility of reliable operation at ambient temperatures of 300-330°C and pressures up to 5 bars; – Maximum efficiency (90%); – Minimum weight and overall dimensions; – High strength at mechanical, thermal and electromagnetic loads and overloads, the possibility of operating at high vibrations; – Self-excitation; – The quality of electricity in generator mode must comply with the requirements of MIL STD 704IE; – SG materials should not sustain burning for 5 minutes to ensure fire safety of AE. To comply with these requirements in developed SG



217


integrated in AE used installation with two duplicate windings as in the SG by Thales company, or multi-phase machine where other phases remain operational during one phase loss [9] (SG Politecnico di Torino). Fire safety requirements for SG mounted on HPS, as well as an increased level of integration makes undesirable the installation of machine with PM and the multiphase winding located on one of the stator magnetic core or with duplicated winding on the HPS shaft. This is due to the danger of sparks at the interturn short circuit. Integrated SG should withstand spark and does not burn for at least 5 minutes at interturn short circuit, according to the requirements. SG has to be stopped during this time to prevent interturn short circuit. Generally, on aircrafts, two or more electric power generators are used and the system loses power during a power shutdown of SG, which eliminates the effect of using a multiphase machine, or with duplicated winding and also requires a second SG overloading to twice value. All this leads to an increase in SG weight and overall dimensions as well as lower reliability of the whole power supply system of the aircraft. Politecnico di Torino used reinforced insulation to solve the problem, but it did not provide 100% protection against short-circuit and the associated interturn effects. To solve this problem AES Thales provides the inductance of the winding, where the short-circuit current (interturn or three-phase) is slightly different from the rated current of SG. This solution does not allow achieving a possible SG over-current. That is, there are several concepts for the creation of SG for MEE, but there are still many unsolved problems.

II.

system of the aircraft continues to operate when disconnected from the rectifier module winding, as saved operability of the second module. This allows the normal functioning of the aircraft power supply system under emergency conditions in SG, without increasing SG weight and overall dimensions. Both modules are three-phase machine with PM and external rotor, rotor back located on the rim made from Inconel 718 or titanium, which is part of the AE. The advantage of this design is that the external rotor is cooled by air, providing an acceptable operating temperature (330 oC) for PM (for example, magnets EEC 22-T450 [21], the maximum operating temperature which is not more than 450 oC), at the same time cooling system with an external blower of rotor not complicate the SG design. Stator magnetic core and winding is cooled by air that passes through the holes and slots in the stator core. At the same time, the heat shield is set in the air gap between stator and rotor, preventing the PM from heat flows generated by stator windings. And the heat shield is not electrically conductive. Tooth stator winding used to minimize length, which has minimum dimensions of frontal parts, which makes this technology almost the key for various transport systems [12].

Statement of Research Problems and the Machine Type Selection

Therefore, it was proposed the concept of using sixphase, two-module machine as SG mounted on the HPS, Figure 1. In this case, weight and overall dimensions of two-module SG should be no more than mass characteristics of Thales SG or University of Sheffield SG, at equivalent power and speed. In our solution, both rotors of each module connected to the HPS rim of AE and established with an 60 degrees offset relative to each other (to form a six-phase system), both modules of the stator winding output to a common 12-pulse rectifier with the possibility of disabling for each phase from the rectifier. Between the frontal parts of each module of the stator winding set fireproof lining. Due to the special fastening of the stator inside an AE, it provided the mechanical decoupling of the stator with rigid mounting of an AE at short circuit. Thus a stator drop in the radial direction and mechanical coupling of the stator with PM of rotor occurs due to magnetic attraction forces, as shown in Figure 2. This design will allow stopping the electromechanical conversion of energy in it due to the mechanical coupling of the stator and rotor at occurrence of interturn short circuit in one module. In this case, the power supply

Fig. 1. SG mounted on HPS

Fig. 2. Stator mechanical disconnector at interturn short circuit



218


At the same time, it is necessary to consider the disadvantages of this winding type: high losses in the PM and rotor back generated by eddy currents. Studies of machine with tooth-coil windings are presented in [22] – [27], but in this paper it was proposed a new SG design with various unique technical solutions (mechanical decoupling of stator, modular six-phase structure) which operated at high ambient temperatures (300-350 °C) and with the possibility of integration in the AE. A study of machine with the requirements for overload and reliability are not presented in the literature, which proves the originality and relevance of the research.

integrated into AE. Furthermore, this magnetic system should provide maximum flux density in the magnetic air gap and the mechanical strength of the AE rotor. A nonmagnetic rim made from heat-resistant titanium alloy or from Inconel 718, depending on the AE design. TABLE I PRELIMINARY PARAMETERS OF THE RESEARCHED SG MOUNTED ON HPS Components SG with external rotor Power in the generator mode, kW 150 Rotor speed, rpm 13 500 (9 000–15 000) Number of poles 10 Number of phases 6 Number of teeth 2×12 Rated voltage, V 200 Rotor outer diameter, mm 300–315 Active length, mm 120–150 PM thickness, mm 8–12 Stator outer diameter, mm 260–264

III. Machine Design High-temperature six-phase SG with the possibility of integration in the HPS is the subject of research work. SG has facing design with tooth-coil windings and two magnetic cores, each of which has 12 teeth and rotor has 10 poles. This teeth and pole ratio is selected in view of its effectiveness [22], Figs. 3, 4. Particular attention at SG mounted on HPS designing should be given with reference to the stator winding. Currently, industry produces several types of hightemperature wires: high-temperature wires where the electric conductor is made from chromes bronze BrHNB (Cr-0.3, Nb-0.1, Cu-0.4, 0.2 impurities, e.g. PRP 400, PRP 450 wire designed for electrical equipment of nuclear power stations). Maximum operating temperature of wires does not exceed 450 °C, and the thermal conductivity is 380 W/mK. Thus, the resistivity of wires is 0.0185 ohms/m at 20 °C and 0.0517 ohms/m at 400 °C. Another type is a wire of nickel-plated copper, with working temperature up to 600 °C (HELUTHERM 600). Wire sectional area is 1 mm2, active resistance is 0.088 ohms/m at 20 °C. Active resistance of the wire, which is 5 times greater than the resistance of copper, will increase ohmic losses in the high-temperature SG. Therefore, it is necessary to use 450 POT wire in the module SG mounted on HPS, but winding and cooling systems should be designed so that wires maximum temperature does not exceed 420-430 °C at an ambient temperature of 300-350 °C. In the future, hydrocarbon nanotubes may be a competitor for nickel and bronze high temperature wires [28]. 22 EEC-T450 used as the PM magnets (PM based on the alloy Sm2Co17 with working temperature up to 450 °C) [21]. Stator material is Vacoflux 50 (cobalt alloy) with 0.1 mm sheet thickness [29]. Preliminary geometrical dimensions, power, rotor speed and the voltage shown in Table I were selected on the basis of [1], [17], [20]. Controlled rectifier is made on power up to 110 kW, so it could work with the overload when one SG fails, Figure 4.

Halbach magnetic arrays are typically applied when using a non-magnetic rotor back [30], therefore, it was initially assumed to use Halbach array in the considered SG design. To evaluate the effectiveness of this technical solution computer modeling of these magnetic systems was carried out in the software package Ansoft Maxwell: a magnetic Halbach assembled on nonmagnetic rim and a magnetic system with radial PM on the rim made from Vacoflux 50.

Fig. 3. Six-Phase high-temperature SG HPS

Fig. 4. Architecture of the Six-Phase high temperature SG mounted on HPS

III.1. Mechanical Design The location of the PM rotor on nonmagnetic rim is the main requirement for the SG rotor magnetic system

The simulation results showed that the flux density in the magnetic air gap is 0.59 T by using the Halbach



219


magnetic array and 0.5 T by using a magnetic system with radial PM. But in this case, the flux density of SG pole decreases due to pole area reduction in the Halbach magnetic array (1/2 pole arc does not create radial component of the magnetic field and it is used for passage of flux density in a tangential direction). At 2 and 4 pole magnetic systems it is advantageous to use Halbach arrays, while at 10 pole system efficiency is lower in comparison with the magnetic system with radial PM. But the use of a magnetic system with radial PM at the non-magnetic rotor back is not effective. Therefore, the following embodiment of a hightemperature SG rotor magnetic system has been considered: PM with radial magnetization from the EEC 22-T450 located at the 10 mm thick hub from Vacoflux 50, which is later pressed into the 5-7 mm thick rim from Inconel 718 or titanium alloy. Cooling ducts are configured similar to the rim ducts formed on the AE blades. This allows providing PM cooling, while maintaining the necessary electromagnetic load. At the same time all proposed design should provide mechanical strength. The model was developed in the software package Solid Works to evaluate mechanical strength and stress calculations in the AE rim. Margin of safety of rim is 1.6 and the maximum mechanical stress is 404 MPa. This proves the efficiency of the selected rotor design. Therefore, the electromagnetic and thermal calculations of this rotor design will be used.

(a)

(b)

Figs. 5. The magnetic flux density (a) and distribution of the magnetic flux lines (b) of SG module mounted on HPS

(a)

(b)

Figs. 6. SG electromagnetic torque in generator mode at rated load (a) and at overload (b)

(a)

III.2. Electromagnetic Design The electromagnetic design of the machine is carried out with the finite element software Ansoft Maxwell. Preliminary geometrical dimensions of SG mounted on HPS module were calculated using analytical methods. Computer modeling was carried out in the generate mode at the speed of 13 500 rpm. Since both modules (electrical machines) belonging to SG are the same, then calculations were made only for one module. Preliminary module power was 75 kW. Computer modeling was carried out taking into account the SG load and overload as well as the demagnetization action of armature reaction. The calculations take into account the reduction of the energy characteristics of the PM rotor under the influence of temperature (magnet temperature was assumed to be 330 °C, and the residual flux density EEC 22-T450 was 0.83 T, and the coercive force was 400 kA/m). Two modes of electromagnetic calculations of SG were considered: the nominal mode and the overloading mode. The flux density in the SG active elements and the distribution of the magnetic flux lines in the SG at rated load are shown in Figures 5. The electromagnetic torque and phase flux linkage at rated load and overload are shown in Figs. 6, 7, respectively. Figs. 5 show that the maximum flux density in the teeth reaches 1.81-1.9 T, in the stator back 0.65 T, in the magnetic air gap 0.48-0.5 T.

(b)

(c) Figs. 7. Flux linkage at rated load (a), during overload (b) and harmonic order of the flux linkage and voltage (c)

The obtained values of flux density in active elements of the SG will use in further calculations. Analyze of the magnetic field distribution also shows that 30% of the stator back area of the magnetic field is very low. This allows performing magnetic annular cooling ducts at these parts without reducing the SG electromagnetic characteristics.



220


These electromagnetic torque curves show that there are electromagnetic torque pulsations which are characteristic of machine with tooth-coil winding [23]. The amplitude of the ripple is 20 Nm. It was also found that the SG module provides power at rated load and at specified geometric dimensions of: M  n 60 13500 P   84 kW. The SG module power 9550 9550 will be 94 kW at maximum speed (15 000 rpm), and 54 kW at a minimum speed of HPS (9 000 rpm). That is, the total power of SG in the nominal mode is from 108 to 188 kW. SG module power is 113 kW at 13 500 rpm in the overloading mode, and the SG power is 226 kW, i.e. the SG available overload is 130% with two working modules in continuous mode. If the aircraft (power consumption 334 kW) has two AE with two-module six-phase SG (4 modules on the aircraft, two modules at AE) and in case the modules (84 kW) fail, its power is evenly (33 %) divided into three operating modules. Then the power of each module at a 9 000 rpm speed is 111.5 kW and these values correspond to long-term SG overloading operation mode. In this case, the aircraft power system operates without disconnecting power consumers and the whole power system will still be working at 334 kW, which increases the reliability of the aircraft power supply system. If two modules fail then the system power will be 230 kW considering the possibility of module overloading. This requires reducing the load capacity of 104 kW, a similar situation would occur with a traditional design with two three-phase SG or SG with duplicated coil. Figs. 7 show the flux of the windings at rated speed and overload and voltage harmonic composition and flux at rated load. From Figs. 7 it’s seen that flux linkage and hence SG voltage decreases at 15-18% due to the magnetic field of the armature reaction during overload. This makes the task of increasing the SG voltage mandatory. Analysis of the flux density distribution under the SG pole for different operation modes was made for the voltage control system design, Fig. 8. These results will be used in the future to form SG voltage stabilization system algorithms.

All machines with tooth-coil winding have a voltage and flux linkage with significant third harmonic (12-14% of the basic harmonic for the voltage) and also fifth harmonic (8-10% of the basic harmonic for the voltage). The emergences of third and fifth harmonics are caused by eddy current losses in the rotor sleeve and PM, which will be discussed later. Standard methods, such as PM skewing or use of dummy slots may be used to reduce the voltage and flux linkage harmonic order [31]. A finite element analysis has also been made to evaluate the effectiveness. As a result of this analysis, it was found that these methods allow reducing the amplitude of voltage third harmonic almost by 30-35%. Table II shows the parameters and dimensions of the developed SG mounted on HPS by finite element and analytical methods. Comparisons of developed construction (A) with SG with duplicated three-phase windings (B) and SG construction by Thales (C) are shown for clarity [17]. Table II shows that the proposed SG design has a minimum weight and overall dimensions in comparison with analogues at maximum power. Designed and studied SG at 188 kW power weighs 59 kg, when the analogs weigh 75 and 82 kg at a 150 kW power, respectively, and the weight and reliability of aircraft systems is a priority criteria. The current density of the designed SG is 20% more than current density of SG with distributed, replicated three-phase winding, due to the higher power of developed SG. TABLE II RESULTS OF SG MOUNTED ON HPS CALCULATION Components A B C Maximum power in generator 188 150 150 mode, kW Rotor speed, rpm 9 000-15 000 9 0009 00015 000 15 000 Number of poles 10 8 8 Number of phases 6 3×2 3×2 Number of teeth 2×12 48 48 Rated current, A 2×257 (514) 440 476 Rated voltage,V 200 200 200 Current density, A/mm2 2,8-2,9 2,3 3,74 Moment of rotor inertia, kg*m2 0,84 0,84 0,137 Mechanical time constant, sec 2,84 2,84 0,42 Inductive resistance along the d/q 0,17/0,17 0,09/0,09 0.16/0.16 axes, Ohm Flux density in the magnetic air 0,5 0,54 0.3 gap, T The rotor outer diameter, mm 300 300 186 The stator outer diameter, mm 262 261,5 300 Non-magnetic gap size, mm 2 2,25 6 The SG mass of active elements, 58,896 75,34 82 kg Magnets weight, kg 8,34 8,34 Magnetic core weight, kg 11,7×2 (23,4) 30 Winding weight, kg 9,078×2 28 (18,156) Rotor back weight, kg 9 9 Active length, mm 2×70 (140) 122 150 Full-length, mm 210 246 250-270 The length of the frontal parts, 15 62 40-60 mm Stator teeth weight, kg 2×4,83 (9,66) 14 Stator back weight, kg 2×6,87 16 (13,74)

Fig. 8. Flux density changing in the magnetic air gap under the SG pole at different operating modes



221


These results confirm the prospects of the proposed design, as it has a minimal weight and overall dimensions, high reliability and maximum power. In addition, the proposed construction was cooled by using an external blowing, which greatly simplifies the task of integrating it into AE.

temperature coefficient of resistance material; T magnetic core temperature. Then for (3) with (4) can be written: Peddy currents  Pspecific eddy currents

III.3. Loss Analysis

(1)

Physteresis  khysteresis B 2 f

(2)

Peddy currents  keddy currents  Bf 

(5)

20 °C. Hysteresis losses are characterized by hysteresis loop area that decreases with increasing temperature. In this connection, it seems appropriate to enter the coefficient characterizing the reduction of the hysteresis loop area under the influence of temperature (by analogy with the temperature coefficient of resistance). Then for (2):

Physteresis  kspecific hysteresis 1   T  20   B 2 f   Pspecific hysteresis 1   T  20   where

k specific hysteresis

–

coefficient

(6)

characterizing

hysteresis losses at 20 °C;  – the temperature coefficient of the hysteresis loop characterizing reduction in its area with increasing temperature; Pspecific hysteresis – specific hysteresis losses at 20 °C. In this view, an empirical model that characterizes a change in the stator magnetic core losses depending on the temperature, frequency and flux density can be written as the sum of (5) and (6). Thus, it is evident that losses in the stator magnetic core are reduced with increasing temperature, hysteresis losses and eddy current losses are also reduced. Experimental studies have been made in the 0,1 mm tape of the alloy Co-Fe for a numerical estimate of losses reduction. Experimental studies were carried out on the MK-4 installation, Figure 9. This installation allows measuring the specific losses in the ring samples from a variety of ferromagnetic materials with different magnetic reversal frequencies (up to 3 000 Hz).

2

(3)

1,5

1   fe T - 20  

where Pspecific eddy currents – specific eddy current losses at

In the papers [32], [33] there is a presentation of a losses estimation technique for machines at a temperature not exceeding 25-30 °C. With increasing temperature, the losses of SG active elements changed. Moreover, winding losses (ohmic losses) increase due to increasing temperature, and the eddy current losses, hysteresis in the stator magnetic core and eddy current losses in the PM are reduced. Therefore, it seems appropriate to develop an empirical model that describes the losses in high temperature SG, confirm it experimentally and using finite element and make losses calculations in the researched SG taking into account high temperatures. Losses changing in windings under the influence of temperature is generally known and not discussed here. However, the losses change of cobalt alloys as well as PM under the influence of temperature is not described in the papers, and consequently it is of scientific interest. The model describing the losses in the magnetic core in general can be written as [32]:

P  Physteresis  Peddy currents

1

–

 kexcess eddy currents  Bf 

where Physteresis – hysteresis losses in magnetic core;

khysteresis – coefficient characterizing the hysteresis losses in magnetic core; B – flux density in the magnetic core; f – magnetization reversal frequency;

Peddy currents – eddy current losses in the stator magnetic core; keddy currents current

losses;

– coefficient characterizing eddy-

kexcess eddy currents

–

coefficient

characterizing the excess eddy current losses. Eddy current losses can be represented in the general form: 2 Eeddy currents (4) Peddy currents  r 1   fe T  20 





Fig. 9. Research of losses in the stator magnetic core

In experimental studies tested a 0,1 mm thick sample tape made from alloy Co-Fe at a temperatures of 150 °C and 20 °C at a frequency of 1 000 Hz. The experimental results are presented in Figure 10.

where r – cobalt alloy resistivity; Eeddy currents – EMF of the eddy currents in the magnetic core;  fe – magnetic



222


Therefore, as mentioned above, the SG mounted on HPS cooling system must be designed so that the winding temperature does not exceed 420-430 °C, at ambient temperature 300-350 °C.

Fig. 10. Dependence of specific losses on the temperature

The experimental results have shown correctness of analytical conclusions. Also the results allowed to determine losses in the stator magnetic core made from Co-Fe 0,1 mm thickness at a temperature of 400-450 °C (Table III). TABLE III LOSSES IN THE STATOR MAGNETIC CORE Components Stator teeth Stator back Losses, W/kg 53 7 Frequency, Hz 1 000 Flux density, T 1.8 0.6

Fig. 11. Losses in PM of SG mounted on HPS at 20 °C temperature TABLE IV LOSSES IN THE ACTIVE ELEMENTS OF 75 KW SG MODULE Components Nickel wire Bronze wire The stator winding losses, W 3 200 614 The winding resistance, Ohm 0,00386/ 0,00111/ 0,0031 0,0164 The magnets losses, W 270 270 The stator teeth losses, W 255 255 The stator back losses, W 49 49 The total losses in one module / 3774/0,95 1188/0,98 efficiency

Losses in PM of machine with concentric winding described in [34]-[36] and therefore here only makes some clarifications on the calculation of these losses at high temperatures. Losses in dependence of PM temperature can also be described as a function of the eddy current losses, depending on the temperature: PPM  Pspecific eddy current

1 1    PM TPM  20  

(7) III.4. Thermal analysis A variant with external and internal blowing has been selected when creating a cooling system. The internal air flows through ducts in the stator slot and the outer air blows HPS rim for cooling the permanent magnets. The proposed cooling system is shown in Fig. 12.

where Pspecific eddy current – specific eddy current losses at 20 °C;  PM – the temperature coefficient of PM resistance; TPM – PM temperature. Calculations were made to estimate the PM losses for these losses at PM conductivity of 1116190 S/m in the software package Ansoft Maxwell. The result (Fig. 11) shows that PM losses are about 350-400 W with the initial temperature conditions. PM losses will be about 220-270 W (in view of PM conductivity reducing at a 350 °C temperature), which is almost equivalent to the PM losses of high-speed machine with three-phase distributed winding at ordinary temperature to 100 °C. This is achieved due to the PM high temperature in SG mounted on HPS. Losses in the PM are reduced due to a significant decrease in conductivity at high ambient temperatures. This increases the efficiency of the machine with concentrated windings at high ambient temperatures. Losses in the active elements of the SG module mounted on HPS were defined as a result of made research. The results of calculations were carried out for a high temperature nickel and bronze wire, Table IV. Thus Table IV shows that the use of a nickel wire is not effective.

Fig. 12. Cooling scheme of SG mounted on HPS

To evaluate the effectiveness of the proposed cooling system the computer simulation was carried out by using Ansys software complex. The results are shown in Fig. 13. The results of thermal calculation show that the winding temperature does not exceed 400-427 °C and that the temperature of the PM is 320-350 °C corresponds to the accepted calculated data and operating conditions, and consequently proves efficiency of the proposed SG design and its cooling system.



223


[3]

[4]

[5]

[6]

[7] Fig. 13. The results of thermal calculation

IV.

[8]

Conclusion

[9]

The publications analysis devoted to the high temperature creation of the SG for MEE concept implementation, determining their weaknesses and general trends was made. As a result of analytical studies it was developed a new design of the high temperature SG mounted on HPS with minimal weight and overall dimensions and improved reliability compared to its known analogues. The materials used in this design are also justified in the paper. Electromagnetic, thermal and mechanical calculations by finite element method as well as the experimental study of losses in its active elements are made. High temperature SG cooling system has been designed and system of mechanical decoupling of the stator at short circuits is proposed. All this has allowed developing a SG weigh of 59 kg at 188 kW power when analogues weigh 75 and 82 kg, with a 150 kW power, and the weight and reliability of aircraft systems is a priority criteria. This project is now transferred to our industrial partner for the manufacture of an experimental layout. These results confirm the promising proposals design, as it has minimal weight and overall dimensions, high reliability and maximum power. In addition the proposed design is cooled by blowing air, which greatly simplifies the task of integrating it into AE.

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

Acknowledgements

[18]

The research has been supported by the research program of the Russian Science Foundation, project No. 16-19-10005. [19]

References [1]

[2]

[20]

Besnard, J.-P.,Biais, F.,Martinez, M. Electrical rotating machines and power electronics for new aircraft equipment systems, ICASSecretariat - 25th Congress of the International Council of the Aeronautical Sciences 2006 Van Der Geest M.,Polinder H.,Ferreira J.A.,Zeilstra D. Machine selection and initial design of an aerospace starter/generator, 2013 IEEE International Electric Machines and Drives Conference, IEMDC 2013; Chicago, IL; United States; 12 May 2013 through 15 May 2013; Code 98445.

[21] [22]


Rajashekara, K., Grieve, J., Daggett, D., Hybrid Fuel Cell Power in Aircrafrt: A feasibility study for onboard power generation using a combination of solid oxide fuel cells and gas turbines, IEEE Industry Application Magazine, vol. 14, no. 3, pp. 54–60, 2008 Xin Zhao; Guerrero, J.M.; Xiaohua Wu "Review of aircraft electric power systems and architectures", Energy Conference (ENERGYCON), 2014 IEEE International, On page(s): 949 – 953 Jones, R.I., "The More Electric Aircraft: the past and the future," Electrical Machines and Systems for the More Electric Aircraft, pp. 1/1-1/4, 1999. Quigley, R.E.J., "More Electric Aircraft", IEEE Applied Power Electronics Conference and Exposition, pp. 906-911 APEC '1993. J. Kang, Multi-Pulse Rectifier Solutions for Input Harmonics Mitigation, Yaskawa Electric America, 2005. Y. Nishida, Y. Okuma, K. Mino, Practical Evaluation of Simple 12Pulse Three-Phase-Bridge Diode Rectifier of Capacitor-InputType, International exhibition and conference for power electronics, PCIM EUROPE, 2007, Nuremberg Dieter Gerling, Mohammed Alnajjar, Six-Phase Electrically Excited Synchronous Generator for More Electric Aircraft, International Symposium on Power Electronics, Electrical Drives, Automation and Motion, 2016, pp. 7–13. Boglietti A.,Cavagnino A.,Staton D.A.,Popescu M. Experimental assessment of end region cooling arrangements in induction motor endwindings, IET Electric Power Applications.February 2011.Vol. 5. Issue 2. Pр. 203…209. R. Bojoi, Z. Li, S. A. Odhano, G. Griva and A. Tenconi, Unified direct-flux vector control of induction motor drives with maximum torque per ampere operation, Conf. Rec. IEEE ECCE 2013, pp. 3888-3895 Tosetti M.,Maggiore P.,Cavagnino A.,Vaschetto S. Conjugate heat transfer analysis of integrated brushless generators for more electric engines, 5th Annual IEEE Energy Conversion Congress and Exhibition.ECCE 2013; Denver, CO; United States; 15 September 2013through19 September 2013. Pp. 1518…1525. Bojoi, R.,Cavagnino, A.,Tenconi, A.,Vaschetto, S. Control of shaft-line-embedded multiphase starter/generator for aero-engine. IEEE Transactions on Industrial Electronics , 2016, 641 - 652 Cavagnino A.,Li Z.,Tenconi A.,Vaschetto S. Integrated generator for more electric engine: Design and testing of a scaled-size prototype, IEEE Transactions on Industry Applications.2013. Vol. 49. Issue 5.Pp. 2034…2043. C. Wenping, B. Mecrow, G. Atkinson, J. Bennet, D. Atkinson, Overview of Electric Motor Technologies Used for More Electric Aircraft, IEEE Transactions on Industrial Electronics, Vol. 59, No. 9, pp. 3523-3531, 2012. D. Ganev, High-Performance Electric Drives for Aerospace More Electric Architectures, IEEE Power Engineering Society Meeting, pp. 1-8, 2007. Ismagilov F. R., Khairullin I.,Vavilov V., Farrakhov D., Yakupov A., Bekuzin V. A high-temperature frameless startergenerator integrated into an aircraft engine, Russian Aeronautics 2016, Volume 59, Issue 1, pp 107–111 Jiabin Wang, Z. Sun, J. D. Ede, G. W. Jewell, J. J. A. Cullen, and A. J. Mitcham. Testing of a 250-Kilowatt Fault-Tolerant Permanent Magnet Power Generation System for Large Civil Aeroengines, Journal of Propulsion and Power, Vol. 24, No. 2 (2008), pp. 330-335. Wang J., Howe D. Advanced electrical machines for new and emerging applications, Nordic Seminar on ‘Advanced Magnetic Materials and their Applications’ 10th/11th October 2007.Pori, Finland. Rodrigues Leon. High temperature embedded electrical machines for aerospace turbine applications. PhD thesis, University of Sheffield. 2013. Electron Energy Corporation [Online]. Available: http://www.electronenergy.com/ [Accessed: 27-Sep-2016]. G. Dajaku, D. Gerling: Magnetic Radial Force Density of the PM Machine with 12teeth/10-poles Winding Topology, IEEE International Electric Machines and Drives Conference, IEMDC2009, Florida USA, May 3-6, 2009, pp.157-164


224


[23] G. Heins, D. Ionel, M. Thiele, Winding Factors and Magnetic Fields in Permanent Magnet Brushless Machines with Concentrated Windings and Modular Stator Cores, Energy Conversion Congress and Exposition (ECCE), pp. 5048 – 5055, 15.-19. September 2013. [24] A.M. El-Refaie, Fractional-Slot Concentrated-Windings Synchronous Permanent Magnet Machines: Opportunities and Challenges, IEEE Transactions on Industrial Electronics, Jan. 2010. [25] D. Ishak, Z. Q. Zhu: Comparison of PM Brushless Motors, Having Either All Teeth or Alternate Teeth Wound, IEEE Transactions on Energy Conversion, Vol. 21, No. 1, March 2006, pp. 95-103. [26] Magnussen F., Sadarangani Ch.: Winding factors and Joule losses of permanent magnet machines with concentrated windings. 2003 IEEE International Electric Machines & Drives Conference (IEMDC 2003), 01-04.06 Madison Wisconsin, USA. [27] Gurakuq Dajaku, Sachar Spas, Xhevat Dajaku, and Dieter Gerling, Comparison of Two FSCW PM Machines for Integrated Traction Motor/Generator, 2015 IEEE International Electric Machines & Drives Conference (IEMDC) pp. 187–194 [28] Pyrhönen, J., Montonen, J., Lindh, P., Vauterin, J., Otto, M., Replacing Copper with New Carbon Nanomaterials in Electrical Machine Windings, (2015) International Review of Electrical Engineering (IREE), 10 (1), pp. 12-21. [29] Advanced Materials [Online]. Available: http://www.vacuumschmelze.com/ [Accessed: 27-Sep-2016]. [30] Vavilov V., Ismagilov F.R., Hairullin I., Gusakov D. High Efficiency Ultra-High Speed Microgenerator Conf. Rec. IEEE IECON, 2016. [31] Nagorny A., Dravid N., Jansen R., Kenny B., “Design Aspects of a High Speed Permanent Magnet Synchronous Motor/Generator for Flywheel Applications”, NASA/TM-2005-213651, pp.1-7, 2005. [32] Bailey C., Saban D., Guedes-Pinto P. Design of High-Speed Direct-Connected Permanent-Magnet Motors and Generators for the Petrochemical Industry, IEEE Transactions on Industry Applications. – 2009. – Vol. 45. № 3. – pp. 1159–1165. [33] Aleksandar Borisavljeviс Limits, Modeling and Design of HighSpeed Permanent Magnet Machines, Printed by Wormann Print Service (Zutphen, the Netherlands, 2011) [34] Ismagilov, F., Khayrullin, I., Vavilov, V., Electromagnetic Processes in the Rotor Shroud of a High-Speed Magneto-Electric Generator Under Sudden Short-Circuit, (2014) International Review of Electrical Engineering (IREE), 9 (5), pp. 913-918. [35] K. Atallah , D. Howe , P. H. Mellor and D. A. Stone, Rotor loss in permanent-magnet brushless AC machines, IEEE Trans. Ind. Appl., vol. 36, no. 6, pp. 1612-1617, 2000 [36] H. Toda , Z. Xia , J. Wang , K. Atallah and D. Howe, Rotor eddycurrent loss in permanent magnet brushless machines, IEEE Trans. Magn., vol. 40, no. 4, pp. 2104-2106, 2004

Authors’ information Flur R. Ismagilov, professor, head of lecturer Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 1973 he graduated from the Ufa Aviation Institute Electromechanics, Ph.D on the elements and control devices.

Roginskaya Lyubov Emmanuilovna, Prof. Dept. of Electromechanics. Dipl. Electromechanical engineer. (Gorky Polytechnic Institute, 1959). Dr. of Tech. Sci. (MPEI, 1994).

Semen V. Shapiro, professor of department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 1956 he graduated from Moscow Power Engineering Institute. In 1961 he received the Ph.D degree in electrical engineering from Tomsk Polytechnic University, Tomsk, Russia. In 1970 he received the Second Ph.D degree in electrical engineering from Tomsk Polytechnic University, Tomsk, Russia. Vyacheslav Vavilov, senior lecturer, department of Electromechanics, Ufa State Aviation Technical UniversityUfa, Russia. In 2010 he graduated from the Ufa State Aviation Technical University, majoring in electrical engineering. In 2013 he defended his thesis.

Denis V. Gusakov, research assistant, department of Electromechanics, Ufa State Aviation Technical UniversityUfa, Russia. In 2011 he graduated from the Ufa State Aviation Technical University, majoring in electrical engineering. In 2016 he defended his thesis.



225


Integrated Electrical Machines with Permanent Magnets for Aerospace Industry Ismagilov F., Roginskaya L., Shapiro S., Vavilov V., Karimov R., Ayguzina V. Abstract – This paper presents the evaluation of the effectiveness of the power supply system of perspective aircrafts, in which all the electrical generators are integrated into the shafts of the main aircraft engine and into the gearless auxiliary power unit. In addition, this paper presents the development of the requirements of integrated electrical generators. The results of the calculation and analysis of the electrical generators for integration on the gearless auxiliary power unit, the high-pressure compressor shaft and the low-pressure compressor shaft of the main aircraft engine are presented. After analyzing the weight and size of the aircraft with integrated electrical generators, the obtained results and the data for Boeing 787have been compared. In the proposed concept, the total mass of electrical machines is less than the total mass of electrical machines of Boeing 787. Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved.

Keywords: Electrical Generator, Starter-Generator, Auxiliary Power Unit, More Electrical Aircraft, Main Aircraft Engine

to reduce take off weight, to increase fuel economy, to reduce operating costs of the aircraft, as well asto improve reliability and a simplification of the aircraft maintenance [10]. The effectiveness of MEA concept depends on the effectiveness of the electrical generator (EG) used to implement it. The improvement of the efficiency of the EG of the aircraft, the minimization of their weight and size and the simplification of maintenance can be achieved through the use of electrical machines with permanent magnets (PM), of the direct integration of EG with PM in the main aircraft engine (MAE) and / or of the auxiliary power unit (APU). In modern aircrafts, three major places of EG integration into the MAE and the APU have been considered [11] (see Fig. 1): - installation of the EG without a gearbox on the lowpressure shaft (LPS) of the MAE, this concept is analyzed in [12]-[14]; - installation of the starter-generator (SG) without a gearbox on the high-pressure shaft (HPS) of the MAE, which is considered in [15]-[17]; - installation of the starter-generator (SG) without a gearbox on the shaft of the APU, which is considered in [18]-[20] It should be noted that the publications about integrated electrical machines for aerospace applications [12]-[20] mainly consider the implementation of the integrated electrical machines with a total electricity power up to 500 kW-800 kW. Article [15] considers the creation of two electric machines with an installed power rating of one engine of 400 kW. Article [17] considers the design of the electrical machine with a power rating of 150 kW, article [19] considers the creation of the SG with a power rating of up to 500 kW, etc.

Abbreviations EG MEA PM APU SG HPS LPS EG MAE PWM PPDU

Electrical generators More electric aircraft Permanent magnets Auxiliary power unit Starter-generator High-pressure compressor shaft Low-pressure compressor shaft Electrical generator Main aircraft engine Pulse width modulation Primary Power Distribution Units

I.

Introduction

Electrical generators (EG) for perspective power supply systems of autonomous units are very important in modern aerospace industry [1]-[2]. One of the major trends shaping the future of aircrafts is the concept of more electric aircraft (MEA). The large-scale practical implementation of this concept will significantly improve the fuel efficiency of aircraft, will reduce the cost of aircraft operation and maintenance, and will reduce harmful emissions of the aircraft into the environment [3] – [7]. The MEA conception involves the use of electricity for all non-propulsive systems. Traditionally, these systems are driven by a combination of various secondary energy sources: hydraulic, pneumatic, mechanic and electric power sources. This reduces the reliability of aircraft systems and increases their energy consumption [8] – [9]. The transition to a single type of energy, electrical in this case, leads to several advantages for aircrafts, namely: Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved

DOI: 10.15866/irease.v9i6.10929

226

Ismagilov F. et al.

However, modern aircrafts such as the Boeing 787 already require more than 1.4 MW electrical power. This paper considers the design of the integrated electrical machines (two for each of the two aircraft engines and one for the APU), which provide an aircraft with a total of 1.4 MW of useful electrical energy.

This analysis is carried out based on the experience in the creation of EG for aircraft, as well as on experimental studies and calculation of various constructive schemes of EG and SG for aircrafts using the Finite Element Method Magnetics (FEMM). During the analysis, several promising EG has been calculated, for example, the EG with rotor speed of 50,000 rpm and a power rating of 450 kW, as well as the six-phase SG with reverse design for integration on the HPS of the MAE. After analyzing the weight and size of the aircraft with the integrated EG, the obtained results and the data for Boeing 787have been compared. This formulation of the problem is new and has not yet been considered in literature. Research suggested that the installed capacity of the power supply system is 1.4-1.5 MW, herewith SG with a capacity of 450-500 kW is integrated in the APU; two SG of 180 kW are integrated on the HPS of MAE and two EG of 300 kW are integrated on the LPS of MAE.

II.

The Starter-Generator Integrated Into the Auxiliary Power Unit

The auxiliary power unit (APU) is a single-shaft gas turbine engine of low power (maximum power produced by the APU does not exceed 500 kW (Boeing 787)). APU is mainly used for the operation of air-conditioning systems, to generate electricity for the aircraft at the airport and in the case of failure of the main power supply system. APU also provides launch of the MAE (pneumatic launch in traditional layout scheme of aircraft and electrical launch in the Boeing 787). During the flight, the APU is not commonly used. The electrical machine integrated into the APU must be capable of working in engine mode (starter-generator), as this will ensure the electrical launch of the MAE. The electrical launch of the MAE is performed as follows: the battery runs the SG of the APU and the APU unwinds before reaching the required speed according to the sequence diagram. After that, the SG of the APU switches to the generator mode and feeds the SG installed into the HPS. It is important to notice that the APU design allows efficient water-cooling of the SG. Upon cancellation of the gearbox between the APU and the SG, the speed of the APU shaft and the SG shaft is up to 50,000-60,000 rpm. Based on this, the main requirements for the SG integrated in the APU without a gearbox can be formulated as: - maximum efficiency (more than 90%); - minimum weight and overall dimensions; - high mechanical strength, thermal and electromagnetic loads and overloads, the possibility of operating at high vibrations; - rotor speed of 50,000-60,000 rpm; - self-excitation; - quality of the generated electricity in the generator mode during operation in the generating channel must conform to the requirements of MIL STD 704IE. As shown in [18], the SG with permanent magnets fully meet the specified requirements, as they are non-

Fig. 1. Power supply system of the aircraft with a fully-integrated starter-generator SG and an electrical generator EG

It is important to note that the effectiveness of power supply systems, in which all the electrical machines are integrated without a gearbox on both shafts of the MAE and on the APU shaft, is not considered in publications [10]-[20]. The effectiveness of using only one or two places for electrical machines integration in MAE or in APU is extremely limited in power. Indeed, a significant increase in the power of the integrated electrical machine will lead to increase its size, and hence to increase the space for its installation. This will lead to increase the volume of MAE or APU and to reduce their effectiveness. Thus, while for the integration it is necessary to consider not only the parameters of the electrical machine, but also the parameters of the integration object (MAE or APU), and if possible it is necessary to use the free space of the MAE or APU without increasing their dimension. In this approach, distribution and integration of two electrical machines with average power on each shaft of the engine is more efficient than the use of a single highpower electrical machine integrated in one place. Therefore, the main purposes and ideas of this paper are: - the study of the effectiveness of the power supply system of perspective aircraft, in which all the electrical generators (EG) are integrated into the shafts of the MAE without a gearbox and without increasing MAE dimensions; - the development of the requirements and analysis of the characteristics of the EG for each integration place.



227

Ismagilov F. et al.

contact and allow efficient and reliable operations at frequencies of 50,000-60,000 rpm. Thales Group developed a high-speed SG with a power rating of 60 kW and a rotor rotation frequency of 50,000 rpm to integrate into the APU [17]. For the initial evaluation of the efficiency of highspeed gearless APU, it seems appropriate to compare the gearless APU with a high-speed SG and the APU implemented on the traditional pattern scheme. Results are presented in Table I.

The cross-section of the SG shaft is a square; it allows to increase the mechanical strength and reduce fluctuations in the operation process. The magnet form is a semicircle allowing the increase of sinusoidal output voltage. Rotor magnets are made of Sm2Co17 alloy. The control system of the SG is a frequency converter, which is a two-tier reversible converter with a DC link and a PWM (pulse width modulation)control signal of power keys of the inverter-rectifier. Because the waveform of the output voltage does not meet the quality requirements of electrical energy in the on-board network of the aircraft (MIL STD 704IE), the inverter is connected to the network through a sinus filter that smooths the voltage waveform and reduces the harmonic distortion to an acceptable level. It should be noted that one of the main problems of using the SG with PM in the aerospace industry is the complexity of their control system. The SG or the EG with the PM must provide a constant voltage level when the load and the speed changes in a wide range, and this is effectively achieved at present only by applying the frequency converter, whose power corresponds to the power of the SG. There are, of course, other ways to stabilize the voltage of the SG with PM and the EG (parallel stabilization, stabilization with magnetic bias of the stator core [22] – [25]), but now their effectiveness for the aerospace industry is not fully proven. Therefore, in this paper the control system of the SG with PM refers to the frequency converter, whose power is equal to the power of the SG with the PM. This approach leads to a decrease in the efficiency of the SG with PM, but provides the aircraft with guaranteed electricity that meets all the parameters of MIL STD 704IE.

TABLE I COMPARISON OF CONVENTIONAL APU AND GEARLESS APU WITH A HIGH-SPEED ROTATION AT A FREQUENCY OF 60,000 rpm AND A CAPACITY OF 50-60 kW Gearless APU,in the TA18-100 example of the generator with Parameter starter-generator an electrodeveloped by Thales magnetic [17] excitation APU mass, kg 110 150 Generator mass, kg 19 37 Mass of control system, kg 27 5 Total mass of the system, kg 156 192

Table I shows that the total mass of the gearless APU of 60 kW with generator and control system is 36 kg (18.75%) less than the total mass of the conventional APU (with a gearbox). In addition, the specific total mass of all the gearless APU is 3.12 kg/kW, while the mass of the conventional APU is 3.2 kg/kW. For a full evaluation of the effectiveness, it seems advisable to evaluate the gain in mass when using the gearless APU for higher power. The experimental layout of the high-speed SG with PM of 120 kW and the control system have been calculated and developed with a frequency of 50,000 rpm and with the possibility of starter mode of the APU. The calculations of the SG were made analytically [21], [22] and followed by verification in the Ansys Maxwell software package. Figs. 2 show the rotor of the SG with power rating of 120 kW, the general view of the experimental layout and the stator. Basic parameters of the SG are presented in Table II.

TABLE II BASIC PARAMETERS OF THE STARTER-GENERATOR WITH PERMANENT MAGNETS OF THE GEARLESS APU (Fig. 2) Parameter Value The effective value of the starter current, A 345 Linear current load, A/m 71,000 Current density, А/mm2 6 Induction in the starter teeth, T 1.75 Induction in the starter back, T 1.52 The induction in the air gap T 0.57 Gross weight of the SG, kg 25 Efficiency, % 96

Table III presents the results of the comparison of the developed gearless APU with SG and the conventional APU with SG with electromagnetic excitation. In this case, the total mass of the system (the gas turbine of the APU, the SG and the control system) with the gearless APU is 58 kg (23.96 %) less than the total mass of the system with the conventional APU. The specific total mass of the system with the gearless APU is 1.53 kg/kW, while the specific total mass of the system with the conventional APU is 2.68 kg/kW. As shown, with power increase, the efficiency of the gearless APU also increases. It should be noticed that the analytical calculation methods used showed good convergence with experimental data.

Figs. 2. The general view of the experimental layout; b- stator and cooling jacket; c- rotor



228

Ismagilov F. et al.

TABLE III COMPARISON OF THE DEVELOPED GEARLESS APU WITH SG AND THE CONVENTIONAL APU WITH SG WITH ELECTROMAGNETIC EXCITATION (ROTOR SPEED OF 50,000 rpm, POWER RATING OF 90-120 kW) Parameter Gearless APU VGDT ТА18-200 with SG, Fig. 2 with an SG with electro-magnetic excitation APU mass, kg 130 190 Generator mass, kg 25 47 Mass of control system, 28 5 kg Total mass of the 184 242 system, kg

However, the total mass of the system with gearless APU with the developed SG of 450 kW is 50 kg less than the total mass of the APS 5000 with two SG of 225 kW connected to the APU shaft via the gearbox. According to the above method, the calculations of the APU and SG of 750 kW and1 MW were produced (the frequency of the rotor rotation was reduced to 36,00024,000 rpm due to the increase of its diameter). Comparison of the results of these calculations with the industrially produced APU have not been made, because they are absent. The results of these calculations were approximated, and dependence of power from the specific mass is presented in Fig. 3.

It is interesting to consider the possibility to create a high-speed SG with a power of 400-500 kW. To solve this task, analytical methods were used [21], [22] followed by a verification in the Ansys Maxwell software package. The calculation results are presented in Table IV. TABLE IV CALCULATION RESULTS FOR THE HIGH-SPEED STARTERGENERATOR WITH POWER RATING OF 450 kW AND ROTOR SPEED OF 50,000 rpm Parameter Value Power, kW 450 Type of PM Sm2Co17 Current, A 1443 Current density, А/mm2 9 Short-circuit current, A 1817 Number of poles 4 Number of slots 24 Current frequency, Hz 1666.67 Starter outer diameter, mm 185 The starter bore diameter, mm 90 Active length, mm 190

Fig. 3. Dependence of power from the specific mass of the of the system (the gas turbine of the APU, the SG and the control system)

The obtained dependence has an extreme because the increase in the capacity of the SG leads to an increase of power of the APU and, consequently, to an increase in the diameter of the turbine. The increase in diameter of the turbine leads to an increase in mechanical loads on it. Therefore, to ensure the mechanical strength by increasing the power of the APU, it is necessary to reduce its speed and, thus, the speed of SG, which leads to an increase in weight and size of SG and the APU. The further increase in the capacity of the gearless APU with a high-speed SG leads to increase its specific mass and to reduce the effectiveness of its application. Thus, it can be concluded that the effective APU from the point of view of minimum weight and size is an APU with a capacity of 450-500 kW with high-speed SG (50,000-60,000 rpm).

The mass of the control system for this SG is about130 kg. The total mass of the APU with the SG and the control system is 445 kg (APS 5000 gas turbine of the APU is taken as a basis), or 1 kg/kW. That is, the resulting mass is minimal compared to all the above options. Table V presents the results of the comparison of the gearless APU with the developed 450 kW SG and the APU of the Boeing 787 (APS 5000) with two SG of 225 kW connected to the APU shaft via the gearbox. Table V shows that mass of control system of the gearless APU is much more than mass of the conventional APU due to a large power generator. TABLE V COMPARISON OF CONVENTIONAL APU AND GEARLESS APU WITH A HIGH-SPEED SG (50,000 rpm, 450 kW) Parameter Gearless APU APS 5000 with gearbox and with SG, Fig. 2 two SG APU mass, kg 245 245 Mass of the – 30 gearbox, kg Generator mass, 70 Mass of one SG with a power kg output of 225 kW and mass of 90.42 kg [26], the mass of the two SG is 180.8 kg Mass of control 130 40 for two SG system, kg Total mass of the 445 495.8 system, kg

III. SG Integrated on the High-Pressure Compressor Shaft (HPS) of the Main Aircraft Engine HPS of the MAE has a rotation speed of 9,00013,000 rpm and a high operating temperature (up to 350oC). The location of the SG on this shaft allows providing electrical launch of the MAE. Furthermore, the use of this shaft is effective during takeoff, landing and taxi modes, etc. For this reason, the possibility to implement SG for installation on the shaft the HPS has been studied.



229

Ismagilov F. et al.

The University of Sheffield is developing two tw high hightemperature embedded electrical machine machines for MAE [19 9]. ]. One of them is the switched reluctance electrical machine ((like like a SG) mounted on the HPS and the other electric electrical machine is the SG mounted on the LPS. The SG of the HPS has a rotation speed of 13 13,,500 500 rpm and a power rating of 100-150 100 150 kW. k . This electric electrical al machine has 4 phases, 24 teeth and 18poles; 18poles; the cooling system is not described. The ppoles oles of the rotor are made of the nonnon magnetic rim of Inconel 718 alloy. This rim is part of the MAE; thereby integration of the SG to the MAE becomes easier easier. The SG provides overload to 110 %.Development is performed in the interests of Rolls RollsRoyce. It is known from the General theory of electrical machines that switched reluctance electrical machine have high dimensions and weight, so their use as a SG is less effective than the use of SG with PM ((the the University of Sheffield did not show their weight and size parameters). In addition, this SG does not provide significant overload (only 110% of rated power), and a significant number of poles pairs (18, frequency of magnetization reversal is 2025 Hz) which will increase the losses in the SG. Fig. 4 shows the concept of integration implemented by the University of Sheffield on the example of the Trent 500 engine.

-

High strength under mechanical, thermal and electromagnetic loads and overloads; possibility to operate at higher vibrations; - Self Self-excitation; excitation; - Quality of the generated electricity in the generator mode during operation in the generating channel must conform to the requirements of MIL STD 704IE; - In case of emergency, the problem should be localized in the SG and should not affect the process processes in the MAE; - The SG should be made of m materials aterials that do not support combustion for 5 minutes to ensure the fire safety of the MAE. To implement these requirements, it has has been proposed to use six-phase six phase two two-module module electrical machines with PM (see Fig. 55)) as a SG for installation on HPS The weight HPS. weight and overall dimensions of SG should be less than the weight and size of the Thales SG or of the SG of the University of Sheffield Sheffield,, with equivalent power and speed values. values.

Fig. 5. The SG on the HPS and the SG on the LPS

Both rotors of each module in the proposed solution are connected with the rim of the HPS and they are installed with an offset of 60 degrees relative to each other (for forming a six six-phase phase system). The stator winding of both modules are output to a single shared 12-pulse with th the possibility that each phase can 12 pulse rectifier wi disconnect from the rectifier rectifier.. The he fire rated gasket is installed between the front parts of the stator windings of each module module.. Both modules are three-phase three phase electrical machines with external rotor with PM PM. The rotor has 10 poles. The back of the rotor is mounted on the rim of Inconel 718 or titanium, which is part of the MAE. Stator cores of each module ha have 12 teeth. The advantage of this design is the air air-cooling cooling of the external rotor. This provides an acceptable op operating erating temperature (up to 330ºC) for permanent magnets (e.g., for magnets EEC 22 22-T450 T450 [27], the maximum operating temperature is not more than 450ºC). The stator core and winding windings are also cooled by air, which passes through the slots and holes of the stator stator core. To protect the magnets from the heat fluxes generated by stator windings windings, nonnonconductive heat-insulation heat insulation screen is installed in the air

MAE Fig. 4. The concept of integration of the SG in the MAE, University of Sheffield [1 [15]]

Thales AES is developing a SG to be install installed on HPS basedd on the electrical machines with high base high--coercivity coercivity PM [17 7]. ]. Thales made a SG with 150 kW output power at 9,000-13,500 9,00013,500 rpm in generator mode. IIn n the starter mode, it provides a torque of 350 Nm at 4800 rpm. The SG is oil cooled and it has the duplicated coil oil-cooled coil.. The mass of the stator and the rotor of this SG is 88 kg [1 [17]. ]. The disadvantages of this Thales Thales’ generator are its considerable weight and overall dimensions, complexity of providing oil cooling of the SG in the MAE. Based on the review and experience by the authors of the design of the aviation electric al machines, the basic electrical requirements for the SG for installation on the HPS can be formulated as follows: follows - Reliable operation in ambient temperatures of 300ºC 300ºC - 330 ºC and pressure up to 5 bar; 330ºC - Maximum eefficiency fficiency (more than 90%); - Minimum weight and overall dimensions; Copyright © 201 2016 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Aer Aerospace ospace Engineering, Vol. 9, N N. 6

230

Ismagilov F. et al.

gap between the stator and the rotor. To minimize the active length of SG, concentrated stator windings are used. Such windings frontal parts has a minimum size, which makes it almost a key for the different transport systems [28]. Of course, it is necessary to consider the disadvantages of this type of coil: high losses in the PM and in the rotor back generated by the eddy currents. Particular attention in the design of the SG of the HPS should be given to stator windings. Currently, the industry produces several types of high-temperature wires, e.g. temperature resistant winding wires with a conductor from an alloy, which consists of Cr (30 %), Nb (10 %), Cu (40 %) and other elements (20%). The maximum operating temperature of this wire does not exceed 450ºС, and the thermal conductivity is 380 W/m K. While the resistivity of this wire is 0.0185 Ohm/m at 20oC, and 0.0517 Ohm/m at 400oC. The other type is a Nickel-plated copper wire; whose operating temperature is up to 600ºC (HELUTHERM 600). The sectional area of the wire is 1 mm2, the resistance at 20ºC is 0,088 Ohm/m. The resistance of the wire is 5 times more than the resistance of copper, which will lead to increased ohmic losses in the SG of the HPS. Therefore, in the module of the SG of the HPS, the temperature resistant winding wire should be applied. Moreover, the winding and cooling system should be designed so that the maximum temperature of the wire does not exceed 420-430ºC when the ambient temperature is 300-350ºC. Magnets EEC 22-T450 are used in the designed SG (permanent magnets based on the Sm2Co17 alloy, operating temperature is up to 450 0C). The material of the stator core is Vacoflux 50 (cobalt alloy) with a sheet thickness of 0.1 mm. The preliminary geometric dimensions, power, rotor speed and voltage are presented in Table I and they are chosen based on the work [17]. The capacity of the controlled rectifier is up to 110 kV. Therefore, in case of failure of one SG, the entire system will continue to operate, but with overload (Fig. 6 and Fig. 7).

The electromagnetic design of the SG was carried out in the Ansoft Maxwell software using the FEMM. Preliminary geometrical dimensions of the SG HPS module were calculated using analytical methods. Computer modeling was carried out in the generator mode at a speed of 13,500 rpm. Since both modules (electrical machine) included in SG are the same, the calculations considered only one module. Preliminary power was 75 kW. Computer modeling was carried out taking into account the load and overload of the SG, as well as the demagnetization of the armature reaction. The calculations take into account the decrease of the energy characteristics of the PM of the rotor under the action of temperature (the temperature of the magnets was taken equal to 330ºC, residual induction EEC 22-T450 is 0.83 T, and the coercive force is 400 kA/m). Two modes of SG were considered in electromagnetic calculations: the nominal and the overload.

Fig. 7. The architecture of the two-module six-phase SG of the HPS

Fig. 8 shows the distribution of the magnetic flux density in the active elements of the SG and the distribution of magnetic induction lines in the SG at rated load. In Fig. 9, the electromagnetic torque at nominal load and overload is presented. Fig. 9 shows that the maximum magnetic flux density in the teeth is 1.81-1.9 T, and 0.65 T in the back of the stator, and0.48-0.5 T in the air gap. The obtained values

Fig. 6. Two-module six-phase SG of the HPS



231

Ismagilov F. et al.

of magnetic induction in the active elements of the SG of the HPS are used in further calculations. Analysis of magnetic field distribution also shows that for 30 % of the back stator area, the magnetic field is very low. This allows executing the annular cooling channels in these places of the stator core without reducing the electromagnetic characteristics of the SG. The obtained curves of the electromagnetic torque are characterized by torque pulsation, which is typical for electrical machines with concentrated winding. The pulsation amplitude is 20 Nm.

working modules is 130 % in continuous mode. It is assumed that the two MAE are installed in the aircraft with power consumption of 334 kW, and that two-module six-phase SG (4 modules on the aircraft, 2 modules in the MAE) is embedded in each of the MAE. In the case of failure of one of the 84 kW modules, its power will be evenly divided among three working modules (33% for each module). The power of each module at 9,000 rpm will be equal to 111.5 kW, which corresponds to the continuous mode of the SG. The power supply system of the aircraft operates without disconnection of the power consumers, and the power of the entire power supply system remains at 334 kW, which, of course, leads to increase the reliability of the power supply system of the aircraft.

Fig. 8. The distribution of magnetic flux density in the active elements of the SG (upstairs) and distribution of magnetic induction lines in the SG (downstairs) at rated load Fig. 9. Electromagnetic torque of the SG at nominal load (upstairs) and overload (downstairs)

It is also established that for given geometrical dimensions, the SG module at rated load provides the power: P

M  n 60 13500   84 9550 9550

If two modules fail, the capacity of the power system will be 230 kW, taking into account the possible overload of the module. This requires the reduction of the power load on 104 kW. A similar situation would have occurred under the traditional scheme with two threephase SG or SG with the duplicated coil. The parameters and geometric dimensions of the SG designed using FEMM and analytical methods are presented in Table VI. For clarity, the comparison between the developed SG with the Thales’ SG is also given in Table VI. Analysis of Table VI shows that the proposed SG has a minimal weight and overall dimensions in comparison

(1)

where M – electromagnetic torque, n – rotation speed. At maximum rotation speed (15,000 rpm) of the HPS, power of the SG module is 94 kW; at minimum rotation speed (9,000 rpm) of the HPS, power of the SG is 54 kW. The total power of the SG at rated speed is about 108-188 kW. In overload mode, the power of the each module is 113 kW at 13,500 rpm, and power of the SG is 226 kW, i.e. permissible overload of the SG with two Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved


232

Ismagilov F. et al.

with its analogue at maximum power. Moreover, the mass of the developed 188 kW SG is 59 kg, and the mass of the analogue of 150 kW is 82 kg. In addition, the proposed SG is cooled by external air, which greatly simplifies its integration into the MAE. Increasing the capacity of the SG installed on the shaft of HPS may lead to a significant increase in its mass and occupy space inside the engine, which considerably reduces the efficiency of integration. Therefore, the integration of the SG on the HPS is effective in power values below 200 kW.

meets the requirements of aerospace standards as of reliability and redundancy. The mass of the converter is 88 kg. The total mass of the developed SG integrated on the HPS and the control system is 147 kg. Thus, the total mass of the developed SG is32 kg more than the mass of the SG of the Boeing 787. However, the Boeing 787 has four Primary Power Distribution Units (PPDU).The mass of each PPDU is 453 kg. The PPDU also includes the AC-DC static converters of 270 V and the AC (variable frequency of 360-800 Hz) – AC (constant frequency 400 Hz) converters. The mass of the AC- DC converters for four SG with the mass of the transformers is about 160 kg. The mass of the AC (variable frequency of 360-800 Hz) – AC (constant frequency 400 Hz) converters is 40-50 kg for four SG. Thus, the mass of one SG in the Boeing 787 with the mass of the converters is 167.92 kg. In the proposed system, the frequency converter of the SG performs the function of these converters. In addition, the mass of the developed system of SG and the control system is 20 kg (!) less than mass of system of the Boeing 787. Moreover, developed system is air-cooled, which greatly simplifies its operation.

TABLE VI THE RESULTS OF CALCULATION OF THE SG OF THE HPS Two-module SG with inner Parameter SG with rotor external rotor The maximum power output in 188 150 the generator mode, kW Rotation speed, rpm 9,000-15,000 9,000-15,000 Number of poles 10 8 Number of phases 6 3x2 Number of teeth 2×12 (24) 48 Rated current, A 2×257 (514) 476 Rated voltage, V 200 200 Current density, А/mm2 2.8-2.9 3.74 2 Moment of inertia, kg m 0.84 0.137 Mechanical time constant, s 2.84 0.42 Inductive resistance of the axes 0.17 / 0.17 0.16 / 0.16 d/q, Ohm The induction in the air gap T 0.5 0.3 Induction in the starter teeth, T 1.81 1.44 Magnetic induction in the back, T 0.65 1.45 The outer diameter of the rotor, 300 186 mm Starter outer diameter, mm 262 300 Non-magnetic gap size, mm 2 6 The mass of active elements SG, 58.896 82 kg Mass magnets kg 8.34 Mass of magnetic pipe, kg 11.7×2 (23.4) Mass of coil, kg 9.078×2 (18.156) The mass of the rotor back 9 Active length mm 2×70 (140) 150 Overall Length 210 250-270 The length of the coil parts, mm 15 40-60 Mass of starter teeth, kg 2×4,83 (9.66) Mass of starter backrest, kg 2×6,87 (13.74) -

IV.

SG Integrated on the Low-Pressure Compressor Shaft of the Main Aircraft Engine

Rotation frequency of the LPS of the MAE is 10,00012,000 rpm. The LPS is characterized by a low operating temperature (up to 150ºC). The location of the electrical machine on the LPS does not allow the electrical launch of the MAE. At Politecnico di Torino, the six-phase interior permanent magnet synchronous motor (IPMSM) is developed and investigated. For IPMSM cooling air is pumped through the air gap between the stator and the rotor [12] – [14]. In Fig. 10, the IPMSM integrated in the LPS is presented. This design is taken as a basis for the formulation of requirements with the integration of the EG on the LHS of the MAE.

To evaluate the effectiveness of the results, it seems appropriate to compare them with the systems of Boeing 787. Four SGs of 250 kVA (two on each MAE) were installed on the Boeing 787. These SG are directly connected with the gearbox of the MAE and represented the six-pole SG designed for variable frequency to provide an alternating voltage of 235 V at 360-800 Hz. The mass of the SG is 90.42 kg [26]; the mass of their control system is 20 kg for one SG. The total mass of one SG including fuel mass on cooling (not more than 5 kg) is 115.42 kg. The frequency inverter of 180-200 kW should be used to provide electricity to the power supply system of the aircraft with a quality corresponding to MIL STD 704IE, and to ensure the starter launch of the SG. Thales developed a frequency converter [17] for 200 kW aircraft electrical machines. This converter fully

Fig. 10. IPMSM integrated in the LPC

Based on the analysis of works [12] – [14], the main requirements for the EG of the LPS are the following: - Reliable operation at ambient temperatures up to150ºC; - Maximum efficiency (more than 90%);



233

Ismagilov F. et al.

Minimum weight and overall dime dimensions; nsions; High strength under mechanical, thermal and electromagnetic loads and overloads; possibility to operat at higher vibrations; operate - Self Self-excitation; excitation; - Quality of the generated electricity in the generator mode during operation in the generating channel channel must conform to the requirements of MIL STD 704IE; - In case of emergency in the SG, the emergency should be localized in the SG and it should not affect processes in the MAE; - The SG should be made of materials that do not support combustion for 5 min minutes utes to ensure fire safety of the MAE. Analysis of these requirements shows their similarity to the requirements of the SG of the HPS, except that the SG does not work in starter mode and operate operatess in less aggressive environmental conditions. It is also propose proposed pose to use the two two-module module six six-phase phase EG with the PM to be mount mounted on the LPS (see Fig. 5). According to the calculations, the low ambient temperature and the possibility to increase increas current density allows the creation of a two two-module module EG of 300 kW ((the the rotor rotor speed is 9, 9,000 00012,,000 000 rpm, the mass is 70 kg) integrated at the LPS. Based on the data from Thales, the mass of the control system of this EG will not be more than 115 kg. As a result, the total mass of the system on the LPS of 300 kW (with air cooling) will be 185 185-190 190 kg. air--cooling)

system can improve performance by another 22-5%. 5%. This will allow approaching the implementation of new generation aircraft. To improve the design approach the nnew ew algorithms, methods and techniques will be developed and used. Furthermore, tthe he subsequent subsequent experimental study of presented electrical machines (including reliability testing) will allow a full assessment of the effectiveness of these machines.

-

V.

Acknowled Acknowledgements gements This work was supported by the Russian Science Foundation (project 16– 16–19– –10005). 10005).

References [1] Ammar, Y., Boudghene Stambouli, A., Bekhti, M., Design and Optimization of Microsatellite Power System, (2015) International Review of Aerospace Engineering (IREASE), 8 (4), pp. 141 141-150. 150. [2] Vavilov, V., Ismagilov, F., Khayrullin, I., Gumerova, M., Application of Hybrid Magnetic Bearings in Aviation Starter Starter-Generators, (2014) International Review of Electrical Engineering (IREE), 9 (3), pp. 506 (IREE), 506-510. 510. [3] M. Van Der Ge Geest, est, H. Polinder, J.A. Ferreira, D. Zeilstra Zeilstra.. Machine selection and initial design of an aero aerospace space starter/generator. IEEE International Electric Machines and Drives Conference, Conference, IEMDC 2013; Chicago, IL; United States; (2013 2013);; Code 98445. [4] K. Rajashekara, Rajashekara, J. Grieve, Grieve D. Daggett Daggett. Hybrid Fuel Cell Power in Aircraft: A feasibility study for on board power generation using a combination of solid oxide fuel cells and gas turbines in IEEE Industry Application Magazine Magazine. V Vol. ol. 14, no. 3 (2008), (2008), 54--60. [5] Zh. Xin, Xin, J. M. Gu Guerrero, errero, W. Xiaohua. Review of aircraft electric power syste systems ms and architectures in IEEE Intern International, ational, Energy Conference (ENERGYCON) (2014), 949 949––953. 953. [6] R. I. Jones. The More Electric Aircraft: the past and the future future.. in Electrical Machines and Systems for tthe he More Electric Aircraft Aircraft(1999), (1999), 1/1-1 1/ 1// 4. [7] R. E. J. Quigley. More Electric Aircraft in IEEE Applied Power Electronics Conference and Exposition Exposition, APEC (1993) (1993), 906-911 906 911. [8] Vavilov, V., Ismagilov, F., Khayrullin, I., Karimov, R., Multi Multi-Disciplinary Design of Hig High--RPM RPM Electric Generator with External Rotor for Unmanned Aerial Vehicle, (2016) International Review of Aerospace Engineering (IREASE), 9 (4), pp. 123 123-130. 130. [9] Ismagilov, F., Vavilov, V., Bekuzin, V., Ayguzina, V., High High-Speed Magneto-Electric Magneto Electric Slotless Generator, Generator, Integrated into Auxiliary Power Unit: Design and Experimental Research of a Scaled-Size ScaledSize Prototype, (2016) International Review of Aerospace Engineering (IREASE), 9 (5), pp. 173 173-179. 179. [10] A. Borisavljevic, H.Polinder, H.Polinder, J. Ferreira Ferreira.On On the Speed Limits of Permanent Permanent-Magnet Magnet Machines in IEEE Transactions on Industrial Electronics Electronics.. Vol. 57, No. 1 (2010), 220 220–227. 227. [11] R. Klaass, ass, C. Della Corte.The Corte The Quest for Oil Oil-Free Free Gas Turbine Engines Enginesin SAE Technical Paper 2006-01 2006 01-3055 3055 (2006 2006). ). [12] M. Tosetti,P. Tosetti,P. Maggiore, Maggiore, A. A.Cavagnino, Cavagnino, S. Vaschetto. Vaschetto Conjugate heat transfer analysis of integrated brushless genera generators tors for more electric engines.5th engines.5th Annual Annual IEEE Energy Conversion Congress and Exhibition.ECCE. Exhibition.ECCE. Denver, CO; United States ((2013 2013), 1518 2013), 1518-1525. [13] A. Cavagnino, Z. Li, A. Tenconi, S. Vaschetto. Control of shaft shaft-line embedded multiphase starter/generator for aero line-embedded aero--engine enginein in IEEE Transactions on Industrial Electronics Electronics((2016 2016),, 641-652 641 652. [14] A. Cavagnino, Z. Li, A. Tenconi, S. Vaschetto. Vaschetto Integrated generator for more electric engine: Design and test testing ing of a scaledscaledsize prototype in IEEE Transactions on Industry Applications Applications.. Vol. 49. Issue 5 (2013), 2034 2034-2043. 2043.

Results and Conclusions

Analyzing nalyzing the results, the total mass of all electrical machines and their control systems, integrated to the shafts of the MAE and the gearless APU can be calculated. For the two MAE and the one APU on the aircraft with a total installed capacity of 1.4 MW, the aircraft mass of the system is equal to 190*2+147*2+445=1119 kg. In the Boeing 787, the mass of electrical machines is 1167 42 kg (the capacity is 1.4 MW) 1167.42 MW). Thus, in the proposed concept concept, the total mass of electric al machines on both shaft of the MAE and the electrical shaft of the APU is 48 kg less than the total mass of electrical machines in the Bo Boeeing ing concept concept. The developed electrical machines with the exception of the SG of the APU are air cooled, which greatly air-cooled, simplifies their operation. Total integration of all electrical machines reduces operating costs for the maintenance of the aircraft by 2-3 3 %. In addition, the proposed solution allows the use of DC current on the aircraft with m minimal inimal cost, which will lead to a significant decrease in weight and size parameters of all significant systems of the aircraft. aircraft These calculations and research show the effectiveness of the integration of electric electrical al machines on the aircraft. The fuel efficiency and environmental friendliness of the Boei ng 787 is greater than that of conventional Boeing aircraft 5%, therefore the use of the proposed aircrafts by 33-5%,

Copyright © 201 2016 Praise Worthy Prize S.r.l. - All rights reserved

International Review of Aer Aerospace ospace Engineering, Vol. 9, N N. 6

234

Ismagilov F. et al.

[15] Wang J., Howe D. Advanced electrical machines for new and emerging applications in Nordic Seminar on ‘Advanced Magnetic Materials and their Applications’.Pori, Finland. (2007). [16] F. R.Ismagilov, I. Kh.Khairullin, V. E. Vavilov, D. R. Farrakhov, A. M. Yakupov, V. I. Bekuzin. A high-temperature frameless starter-generator integrated into an aircraft engine inRussian Aeronautics. Vol. 59,Issue 1 (2016), 107–111. [17] Besnard, J.-P.,Biais, F.,Martinez, M. Electrical rotating machines and power electronics for new aircraft equipment systems in ICAS-Secretariat - 25th Congress of the International Council of the Aeronautical Sciences (2006). [18] E.Ganev. Selecting the Best Electric Machines for Electrical Power Generation Systems:High-performance solutions for aerospace More electric architectures in IEEE Electrication Magazine.Vol. 2. Issue 4(2014), 14-22. [19] H. C.Lahne, D.Gerling. Investigation of High-performance Materials in Design of a 50,000 rpm Highspeed Induction Generator for Use in Aircraft Applications inAircraft System Technologies (AST-2015), 1-10. [20] J. Croft. APU manufacturers Honeywell and Hamilton Sundstrand are powering up to meet the demands of next-generation aircraft in Flight International(2010). [21] F. R. Ismagilov, I. Kh. Khayrullin, A. M. Yakupov, V. E.Vavilov. Method of Designing High-Speed Generators for the Biogas Plant in International Journal of Renewable Energy Research. Vol. 6, No. 2. (2016), 447-454. [22] Ismagilov, F., Vavilov, V., Research of the Magnetic Field of High-Speed Magnetoelectric Generator, (2016) International Review of Electrical Engineering (IREE), 11 (2), pp. 136-141. [23] F. R. Ismagilov, I. Kh. Khayrullin, D. R. Farrakhov et al. Management and protection of the magneto-electric synchronous generator for autonomous objects (Program registration certificate for the computer. No. 2015613868 from 30/03/2015). [24] D. Kraehenbuehl, C.Zwyssig, J. Kolar.Half-Controlled Boost Rectifier for Low-Power High-Speed Permanent-Magnet Generators in IEEE Transactions on Industrial Electronics. Vol. 58.No. 11. (2011), 5066-5075. [25] A. A. Gerasin, F. R. Ismagilov, I. Kh. Khayrullin, V. E. Vavilov, I. I. Yamalov. A Way to Control and Stabilize the Output Voltage in a System for Generating an Alternating Current with Stable Frequency on the Base of a Magneto-Electric Generatorin Journal of Computer and Systems Sciences International, , Vol. 55, No. 5 (2016), 770-777. [26] Xiuxian Xia, Dynamic Power Distribution Management for All Electric Aircraft.MSc by Research Thesis. Cranfield University, 2011. [27] Samarium cobalt. Available at: http://www.electronenergy.com/products/materials/samariumcobalt/. [28] Dieter Gerling, Mohammed Alnajjar Six-Phase Electrically Excited Synchronous Generator for More Electric Aircraft, in International Symposium on Power Electronics, Electrical Drives, Automation and Motion(2016), 7–13.

Lyubov E. Roginskaya, professor of department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 1959 she graduated from Gorky Polytechnic Institute. In 1966 she received the Ph.D degree in electrical engineering from Gorky Polytechnic Institute, Gorky, Russia. In 1994she received the Second Ph.D degree in electrical engineering from Moscow Power Engineering Institute, Moscow, Russia. Semen V. Shapiro, professor of department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 1956 he graduated from Moscow Power Engineering Institute. In 1961 he received the Ph.D degree in electrical engineering from Tomsk Polytechnic University, Tomsk, Russia. In 1970 he received the Second Ph.D degree in electrical engineering from Tomsk Polytechnic University, Tomsk, Russia. Vyacheslav E. Vavilov, senior lecturer of department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 2010 he graduated from Ufa State Aviation Technical University, department of Electromechanics. In 2013 he received the Ph.D degree in electrical engineering from Ufa State Aviation Technical University, Ufa, Russia. Ruslan D. Karimov, assistant lecturer of department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 2010 he graduated from the Ufa State Aviation Technical University. In 2012he obtained the Master’sdegree in electrical engineering from Ufa State Aviation Technical University, Ufa, Russia. Valentina V. Ayguzina, postgraduate student, department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 2016 she graduated from Ufa State Aviation Technical University, department of Electromechanics

Authors’ information Ufa State Aviation Technical University Flur R. Ismagilov, professor, head of Department of Electromechanics, Ufa State Aviation Technical University, Ufa, Russia. In 1973 he graduated from Ufa Aviation Institute, Department of Electromechanics. In 1981 he received the Ph.D degree in electrical engineering from Ufa State Aviation Technical University, Ufa, Russia. In 1998he received the Second Ph.D degree in electrical engineering from Ufa State Aviation Technical University, Ufa, Russia.



235

1973-7440(201612)9:6;1-W 1973-7440(201408)7:4;1-Z Copyright © 2016 Praise Worthy Prize S.r.l. - All rights reserved