Academia.eduAcademia.edu

Aerospace Engineering

122,213 papers
595,830 followers
AI Powered
Aerospace Engineering is a branch of engineering that focuses on the design, development, testing, and production of aircraft and spacecraft. It encompasses various disciplines, including aerodynamics, materials science, structural analysis, propulsion, and avionics, to ensure the safe and efficient operation of vehicles in the atmosphere and outer space.
arrow_forwardView all topics
Active Vibration Suppression of Nonlinear Beams Using Optimal Dynamic Inversion
by S F Ali

2009

Euler–Bernoulli beams are distributed parameter systems that are governed by a non-linear partial differential equation (PDE) of motion. This paper presents a vibration control approach for such beams that directly utilizes the non-linear... more
Fig. 1 Non-linear Euler-Bernoulli beam model
In the continuous control design, equation (13) is substituted into equation (16) and then the result is substituted into equation (17). After simplification it is possible to write
For convenience, the following definition is made
Fig. 4 Uncontrolled displacement time history A beam with controllers was simulated with a weighing matrix Q (see equation (16)) of
Fig. 5 Controlled displacement profile Proc. IMechE Vol. 223 Part I: J. Systems and Control Engineering
For the simulations with the discrete controller, the same control gain (k= 10) and weighing matrix Q were taken. The tolerance value was set to tol=5x10 *. After switching the control gain, K=diag(k,, ..., ky) was used, and values of
Fig. 7 Comparison of the displacement norm Simulations were also run with a linear beam model and a linear controller, a non-linear beam with one actuator situated at the centre of the beam, and an on-linear beam with three equally spaced actuators. The controller for the linear model was designed using the techniques presented in section 3. Figures 7 and 8 are comparisons of the displace- ment and velocity norms obtained in the simulation studies. single ac reduce t worse t t is observed that non-linear control with a tuator situated at the centre of the beam can he beam vibration but the performance is han the other studied approaches. This
Fig. 8 Comparison of the velocity norm
Figure 9 shows that the control goal (with three actuators) is achieved within 6s of the initial disturbance with a little overshoot. It is evident from Fig. 10 that there is a smooth transition between the controllers. The switch between controllers occurs near every zero crossing of the displacement profile of the beam. Selecting a sufficiently low value for tol, within achievable limits of the control magnitude, the jump can be minimized. Moreover, this beha- viour can be further reduced by increasing the number of actuators and taking the actuator dy- Fig. 9 Evolution of the displacement (state) profile using three actuators
Fig. 11 Comparison of the velocity norm along space Figure 11 shows a comparison of the velocity profile of the beam for uncontrolled and controlled cases (and continuous controller, discrete controller with three actuators, and discrete controller with six actuators). In this figure the norm of the velocity is plotted with respect to the time tf which is a measure of the kinetic energy in the system. It can be seen that whereas the velocity state for the continuous controller reaches zero, it does not exactly go to zero for discontinuous controllers. This is because the final controller is a point controller and it is designed
Fig. 10 Control time history when using three actua- tors
Fig. 12 Integral error time history
Fig. 13. Uncontrolled displacement time history (ran- dom profile) The random initial displacement the second mode of vibra dynamics). The wavy n tion of the ature of profile resembles beam (linearized he uncontrolled displacement (Fig. 13) is greater and therefore an ATD approach is hard to imp ement (in this approach high-frequency dynamics are neglected). However, the controlled shows the efficiency of t he optima displacement plot (Fig. 14) dynamic inver- sion approach. Four actuators were taken to simu-
Fig. 15 Control time history (random profile)
Fig. 14 Controlled displacement profile (random pro- file)
Fig. 16 Evolution of the displacement (state) profile with four actuators (random profile) late with the discrete actuators. It is evident from Fig. 16 that the displacement state has reached its goal. Figure 17 shows the control force required to minimize the beam vibration. Simulations were also run for the three-actuators case and similar conclu-
Fig. 17 Control time history for four actuators (ran- dom profile)
Fig. 18 Comparison of the norm of the velocity along space (random profile) The first initial condition corresponds to the first mode of vibration of the beam and the second random initial condition represents the second mode of vibration of the beam. Both of these modes are effectively suppressed and reduced to a zero value within a short time interval (without demand- ing an infeasible amount of control energy). There- fore, it can be concluded that the used technique (optimal dynamic inversion) can effectively be used in civil engineering problems to suppress structural vibrations.
Table 1 Parameter values the beam execute the condition of a suddenly released load. Therefore, in this case the initial conditions are
descriptionView Paperarrow_downwardDownload
Testing and Modeling of MR Damper and its Application to SDOF Systems using Integral Backstepping Technique
by S F Ali and 
1 more
A. Ramaswamy

2009

Magnetorheological dampers are intrinsically nonlinear devices, which make the model- ing and design of a suitable control algorithm an interesting and challenging task. To evaluate the potential of magnetorheological (MR) dampers in... more
descriptionView Paperarrow_downwardDownload
Turbine blade tip heat transfer in low speed and high speed flows
by Andrew Wheeler

2009

In this paper, high and low speed tip flows are investigated for a high-pressure turbine blade. Previous experimental data are used to validate a CFD code, which is then used to study the tip heat transfer in high and low speed cascades.... more
descriptionView Paperarrow_downwardDownload
A framework to Calibrate a MEMS Sensor Network.
by Julien Bourgeois

2009

descriptionView Paperarrow_downwardDownload
Hot streak and vane coolant migration in a downstream rotor (American Society of Mechanical Engineers, ASME, IGTI Turbomachinery Best Paper Award 2008)
by Rob Miller

2009

descriptionView Paperarrow_downwardDownload
Optimisation of non-axisymmetric endwalls in compressor s-shaped ducts
by Rob Miller

2009

Fig. 1 Schematic of the duct pressure fields
Fig. 3 Net duct loss for unstrutted and strutted ducts
Fig. 4 Measured strut pressure distribution compared with the inviscid static pressure distribution for 74% length duct
Fig. 5 Flow visualization on the surface of the strut in the 74% length duct experimental test a eee... <n The exact size and location of the strut- hub corner separation were investigated using surface flow visualization, see Fig. 5. The strut-hub corner separation occupies a significant fraction of the strut surface. The separation occurs just aft of maximum strut thickness, at approximately 40% of chord. The radial pressure gradient generated by the second bend in the duct drives the low
Fig. 8 Computed exit total pressure coefficient for the base- line design Fig. 7 Area distribution of Duct74, baseline duct, and area- ruled duct with blockage of strut removed
Fig.9 Computed exit total pressure coefficient for the axisym- metric profiled design
Fig. 10 Computed near wall streamlines on the baseline strut
Fig. 11 Computed near wall streamlines on the axisymmetric profiled design
The optimization targets were to minimize the S-shaped duct total pressure loss, defined by Eq. (2), and the pitchwise static pressure distortion at inlet to the duct, defined by Eq. (6). A pres- sure variation equal to zero indicates a uniform static pressure field at the inlet to the duct and therefore reduced forcing on upstream components. Limits were applied to the mass-averaged radial flow angle (8) at the inlet and exit of the duct. The root mean square pressure variation (RMSPV) (inlet static pressure distortion) is given by where P is the mean static pressure at the height.
Fig. 12 Axial location of control points
Fig. 13 Predicted Pareto solutions for the fifth update to re- sponse surface
Fig. 14 Computed C, distributions at the strut-hub interface
Fig. 15 Predicted optimal hub endwall design. Contours of perturbation height as a fraction of hj,. “The hub endwall of the hub alone 2 duct geometry is shown in Fig. 15. The contours in Fig. 15 highlight the extent of profiling
Fig. 17 Computed optimum hub C, distribution on the hub endwall and lines of perturbation height applied. Negative contours are regions of reduced endwall height and positive are regions of increased endwall height. The profile endwall can be seen to lower the wall in the front portion of the duct and to raise it in the rear portion. To see the effect this has on the static pressure distribution on the hub endwall, Figs. 16 and 17 show contours of C,, with and without endwall profiling. Figure 17 has the contours of profiled endwall’s perturbation height su- perimposed and the labels indicate the additional concavity and convexness due to the profiling. The presence of the profiling can be seen to reduce the effect that the strut has on the hub surface static pressure. Close to the strut leading edge the profiling in- creases the convex curvature of the streamlines and so lowers the static pressure. This compensates for the raised static pressure caused by the front of the strut. Close to maximum strut thickness, the profiling increases the concave curvature of the streamlines, thereby raising the static pressure. This compensates for the re-
Fig. 16 Computed baseline C, distribution on the hub endwall
Fig. 19 Computed exit total pressure coefficient of the optimal design The experimental setup for the profile endwall tests was iden- tical to that used for the baseline tests except that the hub wall was replaced with the new profiled wall. It should be noted that the same instrumented strut was used. This meant that the pressure tappings on the strut are in an identical location to those in the baseline test. However, the presence of the endwall profiling means that they are no longer at, for example, 10% span. In fact, the rise in the hub wall at the rear of the strut meant that a number of the rear tappings were covered and are thus not used. A com-
Fig. 21 Flow visualization on the surface of the strut for the nonaxisymmetric experimental test Fig. 22 Measured exit contours of total pressure coefficient for nonaxisymmetric test
Fig. 20 Measured C, at 10% of span with and without endwall profiling
The flow field at 0.3h;, downstream of the duct’s exit is pre- sented in Fig. 22 in terms of the total pressure coefficient. In the hub region, the presence of endwall profiling can be seen to have reduced the high loss region associated with the hub separation. It
descriptionView Paperarrow_downwardDownload
Compressor leading edge spikes: A new performance criterion
by Rob Miller

2009

Fig. 1 Schematic of surface pressure distribution with en- largement of spike A second consequence of the small size of compressor leading edges is that they often deviate geometrically from the design intent, due either to manufacture variation or erosion. The spike height therefore varies between blades. It would be both prohibi- tively expensive and technically difficult to ensure that all com- pressor blades in service had no spikes and it would therefore be of use to understand what spike height could be practically toler- ated and thus develop a criterion to define this.
Fig. 3 Optimisation result in transformed space The leading edge radius to max thickness is 0.04. This is sharper than any reported by Carter [3]. He found that the lowest, 0.08, gave the greatest incidence range. The reason for the in- crease in incidence tolerance is that the spike is removed over much of the range. This is demonstrated in Fig. 5, where the spike is shown in experimental and calculated surface pressure distribu-
Fig. 5 Leading edge suction surface pressure distributions, CFD, and experimental showing Dapixe 3.2 Experimental Methods. Measurements of the boundary layer were made using a hotwire probe. Data were taken at 16 suction surface locations, indicated in Fig. 7. At each location, 100 phase locked signals were taken at a frequency of 140 times higher than blade passing. A micrometer screw gauge with 10 wm
Fig. 4 Comparison of incidence range, CFD, and surface pres- sure distributions, CFD, and experiment, i,,..=-2.9 deg
Fig. 8 Measured, time resolved, inlet incidence, and turbu- lence intensity The leading edge of only one stator blade was altered. A row of five stators, with the test blade at the center, had unsealed hubs with a hub gap of 1% chord. This suppressed the hub separation and allowed two-dimensional flow to be maintained at midheight.
Fig. 6 Schematic of compressor working section
Fig. 9 Space-time diagram of suction surface intermittency and contours of calmed region, jinjet=-2.9 deg
Fig. 11. Effect of inlet incidence on mean transition location (y=0.5) within wakes and between wakes
Fig. 13 Measured trailing edge energy thickness, jiniet= -2.9 deg
Fig. 12 Measured energy thickness at peak suction, jiniet= -2.9 deg
Fig. 14 Effect of inlet incidence on the time averaged energy thickness at the trailing edge
Fig. 15 Space-time diagram of suction surface intermittency and contours of calmed region with a missing wake and ellip- tical leading edge, jinjep=-2.9 deg
Fig. 16 Energy thickness at s/s)=0.67 with a missing wake, elliptical leading edge, jinje=-2.9 deg. Comparison of mea- sured experimental data and CFD.
Fig. 18 Effect of spike diffusion factor on trailing edge energy thickness, j,j¢=-2.9 deg
Fig. 19 Comparison of trailing edge energy thickness between blades with ellipse ratios 1.2 and 1.4 from Fig. 18, jintet= -2.9 deg
Fig. 21 Critical leading edges with different design philoso- phies, geometry, and leading edge surface pressure coeffi- cient, finiet=0.6 deg
descriptionView Paperarrow_downwardDownload
Multi-Input/Multi-Output Adaptive Output Feedback Control Design for Aeroelastic Vibration Suppression
by Jian Chen

2009

descriptionView Paperarrow_downwardDownload
Improved Demonstration of Internal Charging Hazards Using the Realistic Electron Environment Facility (REEF)
by Rich Bielby

2009

Experimental ground testing has been carried out to determine whether digital latches can spontaneously change state in severe (space weather induced) outer-belt environments via the mechanism of internal spacecraft charging/discharging.... more
Figure 1. Integral electron spectra in REEF at the maximum and minimum source- sample distances compared with predicted worst-case and average GEO environments. The electron flux (usually quoted in terms of current density, pA/cm”) at the sample c: be altered by variation of source-sample distance. Integral spectra for the maximum and minimum source-sample distances possible within REEF are shown in Fig. 1 together wit the average GEO electron environment from the NASA AE8 average model [9], the NAS Spacecraft Charging Handbook severe-case spectrum [10] and a ‘worst-case’ GEO environment predicted by the FLUMIC model [11]. FLUMIC is used in ESA's DICTAT internal charging analysis tool [12]. It can be seen that the REEF capability brackets the predicted GEO levels and provides a good spectral match up to approximately 2MeV.
Figure 4. PCB with mounted 74HC74 chip.
There are an infinite number of possible track layouts and they strongly influence the likelihood of ESD coupling into the circuit. The board layout was relatively arbitrary except that the clock lines were made long and one clock line was routed in a wide loop whereas the other was kept to a more limited area of the board. Clock lines were pulled low via 500kQ resistors. Another feature on the board was an island of copper, which could be left isolated if required in order to investigate the hazard that it presented. The mounting bolts could be isolated or grounded.
Figure 7. Test history. The top plate shows the two output states from the chip during testing. Changes in these two states represent the occurrence of anomalies (excepting tests of the normal functioning of the circuit - marked with squares below the axis). The lower plate shows the irradiation current and the temperature during the test.
Figure 8. Typical waveform detected through the PCB ground plane (the waveform was recorded on 6" May and the corresponding event is marked by w/f in Fig. 7). The measured signal corresponds to a current from the ground plane of 0.2A/V.
Figure 9. Typical waveform detected on the aerial (the waveform was recorded simultaneously with the ground plane waveform in Fig. 8).
‘igure 10. Test history of second test. The top plate shows the two output states from the chip during testing. Changes in these two states represent the occurrence of anomalies (excepting tests of the normal functioning of the circuit - marked with squares below the axis). The lower plate shows the irradiation current and the temperature during the test.
In order to examine repeatability of the results, the irradiation was repeated with a previously un-irradiated component and circuit board. All initial conditions were identical to the original test except that the initial temperature was set to -10°C so it was expected that anomalies would occur readily. The recorded latch outputs (Q1 and Q2) and the temperature and current histories are shown in Fig. 10. Markers below the trace again denote test flips (non-anomalous events) in the latch trace.
Table 3. Summary of anomalies in second test
descriptionView Paperarrow_downwardDownload
The development and test of a hydrogen peroxide monopropellant microrocket engine using MEMS technology
by Rich Bielby

2009

Given the present, relatively limited deployment of low cost and mass space missions, there are clear opportunities for the application of small-scale propulsion systems in further enabling these small satellite missions. With this... more
Figure 1: A system block diagram for a pressure-fed monopropellant micro rocket engine, readily transferable to a bipropellant concept as part of a future development programme The selection of a propellant well suited to meeting the defined requirement specification was accomplished through a_ semi-
Fig. 2. Preliminary design concept oy olllilo. The reason to investigate micro turbines as feed system is based on the widely reported efforts of the MIT’s Gas Turbine Laboratory with their micro gas turbine engine development programmes [4]. MIT have reportedly developed and tested a device which combines a centrifugal pump and a turbine on a single rotor. To date a flow
Figure 3: (Top) Layout of a single catalyst bed wafer. (Bottom) A sample catalyst bed wafer from an early fabrication trial
The MEMS-fabricated motor assembly is fixed between two precision-machined titanium clamp rings, the upper of which also fixes the nozzle (manufactured from boron nitride) to the integrated assembly. Figure 5: Basic geometry of the motor assembly in the monopropellant concent micromachined motor package assembly. The motor package consists of the catalyst wafer pack (8 bonded, silver coated catalyst bed wafers), a top sealing wafer, a bottom sealing wafer and a fluid-handling wafer. The catalyst bed wafers consist of a central honeycomb structure in which hydrogen peroxide decomposition takes place, whilst the surrounding structure is designed to minimise heat losses from the reaction chamber. The design incorporates perforation of the support beams anda50 m trench on each wafer around the honeycomb catalyst bed ( ). All wafers are etched using a DRIE process.
Figure 4: Identification of the key components of the updated DDM concept
Fig. 6.: test equipment for wafer testing
Fig.7. Test equipment designed for wafer and waferstack testing
descriptionView Paperarrow_downwardDownload
A Modelling and Vibration Analysis of Spinning Disks
by Natalie Baddour

2009

This thesis considers a modelling and vibration analysis of thin spinning disks. Both in-plane and transverse vibrations are considered. Kirchhoff and von Karman theory of plates along with Hamilton's principle are used to derive linear... more
kinetic energy and KE, and KE; are given by for each of the in-plane and out of plane vibrations. we have restricted the system tc
inertia of the disk is also ignored. The justification given for this is that the in-plane In other words, for the stationary problem, ignoring the in-plane inertia is a good ap-
Note that the full nonlinear representation of the spinning disk problem requires the
The equations of motion arising from the use of linear (Kirchhoff) strain-displacement
Furthermore. note that corresponding to an in-plane purely radial displacement u(r). use
There are a few points that are worth mentioning. First. the above boundary term was nonlinearly coupled. Had the linear (Kirchhoff) strain-displacement relations been used.
‘edge reaction’ [41] be set to zero. For stationary plates (neglecting rotary inertia). the and transverse vibrations can be derived. Kirchhoff shear is given by
If the linear stress-strain relations are used along with the nonlinear (von Karman} easily as the linear boundary conditions. Note that the nonlinearity in the boundary
Table 3.1: Solutions of the frequency equation for a stationary disk without rotary inertia. =0. frequencies of vibration. However, comparison with existing work makes it evident that
shown in Chapter 2 to lead to the following equations of motion : for a thin disk plane stress conditions can be assumed. The preceding assumptions were stress-strain and linear strain-displacement relationships can be assumed. Furthermore
The displacements u, and wg are thus given by equations for 4, and B, given by
Thus for j # k and distinct prj, Pag we obtain
To find the coefficients c,,, premultiply both sides of the above equation by @f, r anc of the eigenfunctions. Consider expanding an arbitrary function G(r) as integrate from r=0 tor=a. Given the above orthogonality relationships, arbitrary functions can be expanded in terms
nj = —NQ+w it can be shown that
Figure 4.1: First Dimensionless Natural Frequencies po; and Backward Frequencies po: — i vs dimensionless angular velocity @ for v = 1/4. Wren and Fawcett [21]. These results are also in-line with those obtained by Burdess. sign (hence the backward designation) and its magnitude also tends to decrease wit]
Figure 4.2: Second Dimensionless Natural Frequencies pop2 and Backward Frequencies Po2 — -- vs dimensionless angular velocity @ for vy = 1/4. &22
Figure 4.3: Third Dimensionless Natural Frequencies po3 and Backward Frequencies ppo3 — —. vs dimensionless angular velocity @ for vy = 1/4.
Figure 4.4: First Dimensionless Natural Frequencies po, and Backward Frequencies po; — — ys dimensionless angular velocity & for vy = 1/2. fer
Figure 4.6: Third Dimensionless Natural Frequencies po3 and Backward Frequencies po3 — _ vs dimensionless angular velocity & for vy = 1/2. Ow
The above calculations indicate that Qjg = 0 for m # 2n. This point is worth some
In other words, w, is precisely the frequency of vibration of the in-plane mode that was chosen. This is a naturally intuitive result, which the calculations justify. Let us now consider how to calculate w,. Recall that w, is defined by
functions of the linear strain model of the spinning disk, then
The motivation for doing this is that H) can be recognized as the Hamiltonian of two the approach that will be undertaken here. complicated problem, it will help in the construction of an approximate solution. This is term. The fourth order term in H"!) does not couple the two oscillators. Rather. i has the effect of turning the linear 7 oscillator into a nonlinear duffing oscillator. The k;,
eh=a,+a2 isaconstant. [f equation (5.107) is substituted into equation (5.106) solution S(c,T, De, Dr, t) of the Hamilton-Jacobi equation associated with H\)(c. r. p.. pr).
The required transformation can now be found from
Note from this definition of H!) that there is a difference in its form depending on Hamiltonian’ H“) can be expressed as = 2w,. Although it may not be obvious yet, if w. = 2w,, then tl
integrated to give the first order canonical perturbation solutions for a1, a2, 3;..
first order canonical perturbation solution for the internal resonance case : Thus this procedure gives a first order solution for both the internal resonance case and
Note that for this case, since a? , a8, ° A are all constants, the solution for c and 17 For the case when there is no internal resonance, that is where w, # 2w,, then the
proceeds as follows. The equation for a, is written as roots of the discriminant : quation (5.185) is itself a quadratic in A and as such has two possible solutions. Further of real y;.. Solving for A; yields
Z, and Z> being proportional, clearly an impossibility since one has real roots while the that for w2 to be a real quantity that 42 > 1. It can be shown that A» = 1 results
Figure 5.7: The numerical solution of 7 for the 1 DOF and 2 DOF models. Many of the comments made regarding the previous case can also be made here. Note
Figure 5.10: Comparison of c with numerical solution, first order canonical! perturbatior solution and time-averaged canonical perturbation solution. Tau for Exact, First and Second Order Canonical Perturbation solutions
Figure 5.11: Comparison of + with numerical solution, first and second order canonical perturbation solution.
Figure 5.12: Comparison of c with numerical solution, first and second order canonical perturbation solution. The graphs of 7 and c for the exact solutions of the 1 DOF and 2 DOF model are
Figure 5.13: The numerical solution of r for the 1 DOF and 2 DOF models. perturbation solutions are all in perfect agreement.
Figure 5.16: Comparison of c with numerical solution, first order canonical perturbation solution and time-averaged canonical perturbation solution.
Figure 5.17: Comparison of r with numerical solution, first and second order canonical perturbation solution.
Figure 5.18: Comparison of c with numerical solution, first and second order canonical perturbation solution.
Figure 5.23: The numerical solution of r for the 1 DOF and 2 DOF models. approximate solutions and the exact solutions agree quite well.
Figure 5.24: The numerical solution of c for the 1 DOF and 2 DOF models.
Figure 5.25: Comparison of 7 with numerical solution, first order canonical perturbation solution and time-averaged canonical perturbation solution.
Figure 5.26: Comparison of c with numerical solution, first order canonical perturbation solution and time-averaged canonical perturbation solution.
Tau for Exact, First and Second Order Canonical Perturbation solutions
Figure 5.28: Comparison of c with numerical solution, first and second order canonical perturbation solution.
Note that u,, ug and u, are the displacements of the disk in the r, 0 and z directions respectively.
Conjecture 2 The eigenfunctions corresponding to the linear strain model of a spinning
descriptionView Paperarrow_downwardDownload
Teaching about the Solar System - a Multiple Intelligences Approach
by Jackie Alan Giuliano Ph.D.

2009

The study of planets lends itself to the multiple intelligences approach in such a way that lessons can be designed seamlessly, moving through each of the seven intelligences, each modality resonating and enriching the others.
descriptionView Paperarrow_downwardDownload
• Effect of Metallurgical Parameters on the Machinability of Heat-Treated 356 and 319 Aluminum Alloys, M.M Tash, F.H. Samuel, F. Mucciardi, H.W. Doty, Materials Science and Engineering A, Vol.434, 2006, pp.207-217.
by Mahmoud Tash

2009

The present study was undertaken to investigate the effect of metallurgical parameters on the drilling performance of the heat treated 356 and 319 alloys. The most important metallurgical factors undertaken in the present study that... more
descriptionView Paperarrow_downwardDownload
Impact of supercritical CO2 drying on roughness of hydrogen silsesquioxane e-beam resist
by Yordan M Georgiev and 
1 more
Max Lemme

2009

descriptionView Paperarrow_downwardDownload
Surface roughness of hydrogen silsesquioxane as a negative tone electron beam resist
by Yordan M Georgiev

2009

Fig. 1. Comparison of the contrast curves for various TMAH concentrations at the two extreme baking temperatures, (a) 90 and (b) 220°C and beam energy of 50 keV.
Fig. 2. Dose dependences of HSQ surface roughness for the lowest and highest developer concentrations, 2.5% and 25%, respectively, at the two extreme baking temperatures, 90 and 220°C. For a better comparison of the different processes, a normalized dose Dy = D/Dy, is used, where Dj, is the respective mean value between the onset dose Dp and the dose at 80% of the resist
Fig. 3. AFM images of HSQ surfaces for: (a) Tg = 90°C and (b) Tg = 220°C. The respective minimum exposure dose and the weakest developer, Cp = 2.5%, as well as for the maximum exposure dose and the strongest developer, Cp = 25%, again at: (c) Tp = 90°C and; (d) Tg = 220°C.
descriptionView Paperarrow_downwardDownload
Fabrication of 12 nm electrically variable shallow junction MOSFETs on SOI
by Yordan M Georgiev

2009

descriptionView Paperarrow_downwardDownload
Study of a high contrast process for hydrogen silsesquioxane as a negative tone electron beam resist
by Yordan M Georgiev

2009

descriptionView Paperarrow_downwardDownload
Automatic Flight Control System: Classical approach and modern control perspective
by Agus Budiyono

2009

The document was used as a lecture note for the graduate course on flight control system at the Malaysian Institute of Aviation Technology (MIAT), Kuala Lumpur, Malaysia. Some parts of the document represent an enhanced version of the... more
Classical approach and modern control perspective Automatic Flight Control System
Figure 1.1: Autopilot Control System Diagram: Example of Pitch Channel This autopilot system does not work continuously but only at a certain pe- riod of time. The AP system can be activated by turning the AP switch and deactivated by over-riding the system through the movement of the control ma- nipulator. Therefore the characteristic of the AP system is of limited authority. Fig. 1.1 shows the functional diagram of the autopilot system used in the pitch longitudinal channel of an aircraft.
Figure 1.2: Stability Augmentation System Diagram. Example of Pitch Channel The type of automatic flight control system that adds stability to the aircraft is usually called the Stability Augmentation System or SAS. This type of auto- matic flight control system improves the stability of an aircraft at certain flight configurations and conditions within the flight envelope. For conventional air- crafts, the stability augmentation will be needed during the flight at low speed and low altitude for instance during landing or approach. The control opti- mization of typical aircrafts are conducted only at a certain flight configuration such as cruise configuration. This makes the aircraft stability at other flight configurations namely approach, landing or other special configurations tend to deteriorate. The stability augmentation is therefore necessary for those configu- rations. The stability augmentation can be achieved by increasing the damping ratio of the existing aerodynamic damping ratio through the application of feed- back control system.
Figure 1.3: Command Augmentation System Diagram. Example of Pitch Chan- nel
Figure 1.4: The Stability Provider and Control Power Optimizer Control System Diagram
Figure 1.6: The aicraft cockpit as the interface between controller and controlled systems
Another very critical element of the controller is the information processor sys- tem. In the manual controller (human controller), this system is represented by the pilot’s brain supported by the basic information processing computers to speed up the decision making process. In the automatic controller, the infor- mation processing element is represented by a Flight Control Computer (FCC). The FCC works continuously in real time depending on the authority level of the implemented automatic control system. The software inside the FCC that manipulates the input of the FCC to be converted to the desired control signal by the control system is called the control law. The control law can be in the form of simple instructions which typically used by the autopilot. Some of the examples of control law are:
Figure 1.7: Mechanical Control System: example of longitudinal channel
Figure 1.8: Hydraulically power assisted control system: example of longitudinal channel
Figure 1.9: Hydromechanical Control System
Figure 1.10: Electrohydromechanical Control System
Figure 1.11: Electrohydraulic Control System
Figure 2.1: SAS as an inner loop of aircraft autopilot system
Figure 2.4: The location of auto pilot system components in the standard control diagram variable selected as the reference. On condition that the AP system is working, the pilot can release his hand and feet from the control manipulators, wheels, and controller stick or pedal. All he needs to do is occasionaly checking up on the aircraft trim condition as displayed by the Attitude-Direction-Indicator (ADI). To disengage the AP system, the pilot would just need to grasp controller stick/wheel/pedal and to displace it a little bit. In this case, the AP control loop will be automatically disengaged and the pilot regains the full control of the aircraft. With this characteristic, the autopilot is often referred to as a low authority system.
Figure 2.5: Functional diagram of pitch attitude hold system
with the static sensitivity coefficient of The longitudinal mode transfer function of the N250-100 aircraft during cruise at the altitude of h = 15000 ft and the speed of V, = 250 kts is given as the following (for the elevator control channel): As a case study, the pitch attitude hold system of the N250-100 aircraft— prototype 2 Krincing Wesi— manufactured by the Indonesian Aerospace Inc. is presented. See Fig.2.7. In the case of a conventional aircraft, Goz(s) will have two real zeros and five poles consisting of two pairs of complex poles of the aircraft and one real pole associated with the autopilot control mechanism servo. Observing the closed loop characteristic polynomial, Eq.2.7, note that K., (servo gain) is the only gain that can still be changed or varied. The vertical gyro sensitivity S,g and servo lag factor — are typically of a constant value.
Figure 2.8: Root locus of Pitch attitude hold system for N250-100 PA-2 aircraft
Figure 2.10: Time response of 6(t) due to step 0,¢f = 5° with and without pitch attitude hold system. It is apparent that for the case with the pitch attitude hold, the value o! 9 can be quickly maintained in less than 15 time units, while when the pitch attitude hold is not present, Kgs, = 0, the time it takes to reach the steady leve of Ore y = 0 is more than 500 time unit or 35 times as along as that of the syste with the pitch attitude hold. Also note that when the pitch attitude hold is off. there exists an offset angle of about 15° whereas with the pitch attitude hold the offset angle is only approximately 0.25°. Fig.2.11 shows the time respons« of velocity, u(t), and angle of attack, a(t). The bleed-off speed to hold the pitck angle @ = 5° is about —35 kts while the angle of attack increases by 0.5°.
Figure 2.11: Time response for u(t) and a(t) for K.~ = —8.969!
Figure 2.12: Speed hold system functional diagram the value of the reference speed te ¢(t). In this condition the signal of speed difference €,, from the computer will be zero. When the autopilot is working, the pilot can release the power lever and let the engine control manipulator moves by itself following the closed loop process of the speed hold autopilot system. Hence this type of flight speed holding system is commonly called an Auto Throttle system.
Figure 2.13: Mathematical diagram of speed hold system
Substituting Eqs.2.24b—2.28 to Eqs.2.22—2.23a, the following equation is ob- tained The following figure shows the root locus diagram of the speed hold system of the N250-100 aircraft for positive K,,5, gain. The root locus shown in the figure is varied as a function of the gain K = KpcK-. Notice that the pitch oscillation mode does not change at all since it is constrained by the two complex zero of Ns._.(s) = 0. Nonetheless, the damping ratio of the phugoid mode increases up to its maximum value before moving to the unstable region.
Figure 2.15: Root locus diagram for the speed hold system of N250-100 PA-2- zoomed around the pitch oscillation and phugoid modes
Figure 2.16: Time response of u(t) to maintain u,-rs = 1 with and without the speed hold system, K* = 17.3838
Figure 2.17: Time response of a(t) and @(t) to maintain u;-er = 1, with the speed hold gain K* = 17.3838
The equation of the rate of change of flight altitude can then be obtained
Figure 2.19: Outer-wing air brake (Airbus A-320/319/321) In the Laplace domain, the flight altitude can then be obtained as:
Figure 2.20: Functional diagram of altitude hold system
Figure 2.21: Kinematic diagram of aircraft rate of climb A(s) hres (8) Based on Fig.2.22, the closed loop transfer function can be written as follows:
Figure 2.22: Mathematical diagram of the altitude hold system and the denumerator polynomial is:
Figure 2.23: Root locus for the altitude hold system for the N250-100 with h —> 6 feedback with gain Ky <0
Figure 2.24: Root locus for the altitude hold system for the N250-100 with h — 6, feedback with gain K.. < 0-—zoomed to show the phugoid mode Note that Sp is the conversion gain from signal €), to signal @;¢¢ which is a constant. From the elaboration given in Sec.2.3.1, a good pitch attitude hold control loop has been obtained with Kz, = —8.9695. The corresponding roots of the closed loop polynomial have been given in Eq.2.15. Thus the inner closed loop characteristics polynomial can be expressed as:
Figure 2.25: Altitude hold system with an attitude hold system as the inner loop From the above diagram, the equation for h(s) can be expressed as:
Figure 2.26: Altitude hold system with inner loop having gain Kk? If the gain is chosen such that: the following characteristic roots are obtained:
Figure 2.27: Root locus diagram of the outer loop of the altitude hold system for N250-100 PA-2 aircraft
Figure 2.28: Root locus diagram of the outer loop of the altitude hold system for N250-100 PA-2 aircraft-zommed to show the phugoid mode Using the model of control mechanism and elevator servo as given in Eq.2.4 with the value of +, = 0.1 sec., the polynomial of the inner control loop using the forward acceleration feedback can be expressed as follows:
Figure 2.29: Time response of h(t) for an input of h,;-, = 1 from a system with and without altitude hold for N250-100
Figure 2.30: Altitude hold system with forward acceleration a, as the inner loop feedback The outer closed loop transfer function relating h(s) and hres(s) can then be derived as follows:
Figure 2.31: Altitude hold system with inner loop using forward acceleration feedback az
Figure 2.32: Root locus diagram of inner control loop 1 —> 6¢ f altitude hold system
Figure 2.33: Root locus diagram of inner control loop u — 6, for N250-1 altitude hold system—zoomed around the phugoid mode
Figure 2.34: Root locus diagram of outer control loop % — 6. for N250-1( altitude hold system
Figure 2.35: Time response of h(t) for N250-100 altitude hold system with the inner loop of a, —> 6, feedback 3. Comparator in the autopilot computer transfer function:
Figure 2.36: Functional diagram of Bank angle hold system namely fly by wire or fly by light. The block of the AP servo and control mechanism can then be modeled as a system with time response of T, representing the first order lag factor as follows:
For a conventional aircraft, Gox(s) typically has a pair of complex zeros and five poles consisting of a pair of complex poles and three real poles where one of the poles is associated with the control mechanism and servo system.
To have a more elaborate understanding of the work principle of the bank angle hold system, a case study will be taken for the Indonesian Aerospace N250- 100 PA2 ”Krincing Wesi” at the cruise flight configuration with a low speed of V = 180 KEAS and at the altitude of h = 15000 ft. The lateral directional transfer function for aileron input for the aircraft is given as follows:
Figure 2.38: Root locus diagram of the bank hold system of N250-100 aircraft for a number of 7, values
Figure 2.39: Time response of y(t) for a = 1/7, = 10,5,2 due to an impulse function input
Figure 2.40: Time response of y(t) for a = 1/7, = 10,5, 2 due to a step function input
Figure 2.41: Forces equilibrium during the turn maneuver
Figure 2.42: Functional diagram of heading hold system specified by appropriately selecting the gain ks Sa;, Hence the transfer functior of the inner loop can be expressed as:
Figure 2.43: Mathematical diagram of heading hold system
Figure 2.44: Heading hold system with bank angle hold as an inner loop In this case, the characteristic polynomial for each value of a can be given as the following:
Figure 2.45: Root locus diagram of the heading hold system of N250-100 aircraft for a number of a = 1/7, values
The time response is presented in Fig.2.46. It is clearly indicated that the smaller the value of 7, (or the larger the value of a) the faster the time response of the actuator which in turns causes the higher the damping ratio of the Dutch roll mode. For a fairly large value of 7, (namely for a = 2), the damping ratio can only be obtained for a small value of gain ky. For a small value of 7., the high overshoot peak is due to the associated high gain.
Figure 2.46: Time response of y(t) and y(t) of the heading hold system of N250-100 aircraft
Figure 2.47: Effect of wind to the aircraft flight path The actual position of the aircraft has an offset angle A with respect to the main radial line and a radial distance of R with respect of VOR station. The aircraft position has a bearing y with respect to magnetic north and a distance d with respect to the VOR main bearing. From the geometry of the VOR guidance path the following relations can be readily obtained:
Figure 2.48: VOR guidance path geometry
Figure 2.49: The functional diagram of VOR-hold guidance-control system
Figure 2.50: The mathematical diagram of the VOR-hold guidance-control sys- tem
Figure 2.51: Navigation and guidance system: VOR hold The root locus of the above characteristic polynomial is illustrated in Fig.2.5: for the gain value of ky, < 0. From the root locus, if the following working point associated with ky = —1.1152 is taken, the following characteristic polynomia is obtained:
Figure 2.52: Root locus diagram for the VOR offset A(s) with respect to the bearing V,.¢ of the N250-100 aircraft
Figure 3.1: The functional diagram of the pitch damper system
Figure 3.3: Root locus of the inner control loop: pitch damper system of N250- 100 PA-2 at cruise condition with V = 250 KIAS and h = 15000’
Figure 3.4: Root locus of the inner control loop: pitch damper system of N250- 100 PA-2 at cruise condition with V = 250 KIAS and h = 15000’— enlarged around the phugoid mode
Figure 3.5: Time response of 0(t) and q(t) with and without pitch damper of N250-100 aircraft
Figure 3.6: The mathematical diagram of the pitch attitude hold with the pitch damper as the inner loop the outer characteristic equation of the phugoid damper is given by:
Figure 3.7: Root locus of outer control loop: phugoid damper of N250-100 PA2 at cruise condition
Figure 3.8: Root locus of outer control loop: phugoid damper of N250-100 PA2 at cruise condition—enlarged around phugoid mode
Figure 3.9: Time response of 6(t) of the phugoid damper with Kj = 3.7979 and K* = —0.2432
The components of the yaw damper system transfer function are: elimination in the steady state is to make the yaw damper not in opposition to the pilot input during the the stationary turning maneuver. If the r(t) still exists, it will work against the pilot and will make the coordinated turn difficult to achieve. Therefore, this yaw damper is designed only to enhance the dutch-roll damping in the transient condition such as the period of entering the gust or turbulence or the landing phase with ever decreasing speed.
Figure 3.11: The mathematical diagram of the yaw damper system with r — 6 feedback
where, the static sensitivity coefficient is given by: The transfer functions of the N250-100 due to the rudder input 6, is given as follows: As a case study, the application of the yaw damper to the N250-100 PA-2 Krincing Wesi aircraft will be presented. The aircraft is equipped with the yaw damper system to increase the ride quality particularly during the flight in turbulence and low-speed landing phase.
as a first order system as: The polynomial characteristic of the yaw damper system can be written as:
and the characteristic polynomial is:
Figure 3.13: Root locus of the yaw damper system with r —> 6, feedbac N250-100 aircraft—enlarged around the dutch-roll mode
It is evident from the root locus that the damping of dutch roll mode is improved while the time responses for the roll subsidence and spiral become faster. For sufficiently large gain of ky,, the dutch roll mode can be overdamped to break into two dutch roll subsidence modes.
by the following relations: The results of the calculated is presented in the following figure where three cases are compared. Note from the side slip angle response that, the application of the yaw damper enable a very effective suppresion of the response. In the yaw response, it is evident that for the damper system without the wash-out, che steady state y(t) is sustained constant at w,,(t) = 1 rad. Note, however chat, in the no yaw damper case, the yaw angle 7(t) can freely increase. The nclusion of the wash-out circuit enable the yaw damper to work only during che transient period. After the f,, signal is phased-out, the yaw angle w(t) can drift and increase without being inhibited by the yaw damper system. It can be concluded that by installing the wash-out circuit in the yaw damper, che maneuvers executed by the pilot during the steady state turn will not be »pposed by the yaw damper system. Fig.3.16 shows the phase-portrait of the ‘oll rate with respect to the roll angle. The effectivity of the yaw damper system n dampening down the dutch roll mode and the wash-out circuit in transmitting he yaw rate transiently is evident.
Figure 4.1: Minimum of a cost function J(x) at a stationary point
Table 3.1: Feedback for improving longitudinal stability servo has a task to receive command from the SAS computer only. In a num- ber of large size aircrafts, the AP and SAS servos are usually separated. The mathematical diagram of the pitch damper system is illustrated in Fig.3.2. If the SAS servo can be assumed to have a fast time response, it can be modeled as a constant gain:
descriptionView Paperarrow_downwardDownload
The effect of work processes on the casing heat transfer of a transonic turbine (American Society of Mechanical Engineers, ASME, IGTI Heat Transfer Committee Best Paper Award 2005)
by Rob Miller

2009

descriptionView Paperarrow_downwardDownload
The effect of leading-edge geometry on wake interactions in compressors (American Society of Mechanical Engineers IGTI Turbomachinery Best Paper Award 2007)
by Rob Miller

2009

Fig. 2 Leading-edge geometries for Blades A and B
Fig. 1 The Deverson rig and working section
Fig. 3 The stator-blade pressure distribution for Blades A and B at midspan
Fig. 4 Time-average boundary-layer profile at 8.68% Sy) for Blade A (6/c=0.004)
Fig. 5 The time-average leading-edge pressure distributions for Blades A and B at midspan
Fig. 7 The leading-edge time-average boundary-layer integra parameters for Blade B suction surface (s/r=56s/ Sp)
Fig. 6 The leading-edge time-average boundary-layer integral parameters for Blade A suction surface (s/r=44s/S))
Fig. 8 Variation in shape factor (H) with time at 8.68% S, for Blades A and B
Fig. 10 Raw traces of quasi-shear-stress (7,,) close to the leading edge for Blade A To investigate whether turbulent spots formed on the early suc- tion surface of Blade A, quasi-surface-shear-stress (7,,,) measure- ments were performed with an array of surface-mounted hotfilm sensors. The sensor array was mounted around the leading edge of the blade. Figure 10 shows the raw time traces measured by each sensor. For each trace, the time mean has been subtracted, and the traces are plotted at each sensor location. The traces were logged simultaneously so that the propagation of individual events could be tracked. Celerity lines at the freestream velocity V, 0.5V and 0.3V, are plotted where these were determined from the measured pressure distribution. The celerity lines are plotted so that they meet at the location where the ensemble-averaged shear stress (not shown here) indicated the first appearance of turbulent dis- turbances (s/Sj=0.07). The variation in 6 at 8.68% Sp is alsc replotted for comparison. As expected, the figure shows that tur- bulent disturbances identified as locally raised 7,, (e.g., marked A)
Fig.9 Variation in momentum thickness (6) with time at 8.68% S, for Blades A and B
Fig. 12 Raw traces of quasi-shear-stress (7,,) close to the leading edge for Blade B 4.2 Blade B: The Circular Leading Edge. For Blade B, Fig. 8 shows that the value of H for Blade B varies from H=2.3 to H=1.8 due to the wake interaction. Figure 9 shows that in the absence of the wake, the momentum thickness for Blade B is 5%
Fig. 11 Ensemble-averaged quasi-shear-stress (7,,) close to the leading edge for Blade B
The previous work of Ref. [2] showed that regions of high 6 at the leading edge propagated over the blade surface as a thickened boundary-layer structure, the front of which moved at 0.7V and the back at 0.6V. This structure was shown to be responsible for turbulent transition. To investigate the propagation of these struc- tures, measurement of the late suction-surface boundary layer on both Blades A and B was performed. The measurements involved a series of hotwire boundary-layer traverses at s/S )=0.63—0.97. The results from these measurements are plotted in Figs. 13 and 14. These figures are space-time diagrams of the normalized mo- mentum thickness. The variations in @ at 8.68% Sp are also plot- ted. Celerity lines at the freestream velocity V, 0.7V and 0.6V, are plotted. The 0.7V and 0.6V lines start at the front and back of the
Fig. 13 S-t plot of normalized momentum thickness ((@ - Om)! Om) for Blade A
Fig. 16 Total-pressure wake profiles for Blades A and B at 0.3 axial chord downstream of the trailing edge
Fig. 15 Variation in energy thickness (€) and momentum thickness (0) at 97% S, for Blades A and B
descriptionView Paperarrow_downwardDownload
The effect of wake induced coherent structures on compressor boundary layers
by Rob Miller

2009

descriptionView Paperarrow_downwardDownload
Improving the performance of a valveless pulse combustor using unsteady fuel injection
by Rob Miller

2009

Figure 1. (a) Schematic of Keller et al. valved pulse jet, (b) Spatially integrated chemiluminescence. Dashed line represents typical combustion chamber pressure.
eee EE NNT Experiments were performed on a non-premixed, ethylene-fired valveless pulse combustor. A photograph of the arrangement is shown in figure 2. The total combustor length was 1250mm and the combustion chamber interna diameter was 75mm. The combustion chamber did not contain and bluff body or other intemal features on which the flame could stabilize. Two combustion chambers were used, one manufactured from stainless steel and the othe from quartz glass, thereby offering unrestricted optical access to the flame surface.
The fuel supply rate was measured and controlled using a thermal mass meter. All measurements were taken after several minutes of operation after which time it was felt that the combustor had reached a state of thermal equilibrium. A fast-response pressure transducers was mounted on the pulse combustor inlet pipe in order to resolve the unsteady pressure field. Due to the high gas temperatures it was not possible to flush-mount the transducers on the combustor; instead an ‘infinite-line’. This method is standard practice in industry and has been shown to introduce only very small errors providing they are correctly calibrated. The transducers were flush-mounted on 80mm stand- off pipes which extended radially outwards from the combustor centreline. A 30m coil of nylon tubing was attached to each stand-off pipe, approximating an anechoic termination.
combustion heat release. The intensifier was phase-locked to the unsteady pressure in the combustion chamber. To obtain an acceptable signal-to-noise ratio it was necessary to cumulatively collect light emissions for short phase angles (~5°) over approximately 1000 combustion cycles. This method resulted in heavily ensemble-averaged images in which cycle-to-cycle variations in heat release are not captured.
Figure 5. Experimentally measured pressure and OH’ and numerically predicted inlet pipe velocity (negative outwards) for the valveless pulse combustor reported in this paper. Time at which snapshots in figure 3 were taken are marked A, B, C and D. section these measurements are used to interpret the structure of the flow and neat-release with the combustor. Figure 5 shows the experimental measured pressure and phase-locked spatially integrated OH chemiluminescence within the combustor. The computationally predicted velocity in the inlet pipe has bee superimposed to aid interpretation. The measured peak in chemiluminescence leads the peak pressure by 22°. Th width of the peak in chemiluminescence in this valveless pulse jet is broader than in the valved pulse jet shown i figure 1. In agreement with [6-8], the luminosity of the chemiluminescence never reaches zero. The measure spatial variation in OH’ chemiluminescence at four points during one pulse width are shown on the left of figure ( To aid interpretation schematics of the structure of the flow are given on the right of figure 6. The schematics wer derived from the measured heat release and numerical velocity field in figure 5. The OH* chemiluminescence time snapshots were taken at 185°, 275°, 365° and 455° phase relative to the positive crossing of the pressure signe (marked in figure 5). It should be noted that the exhaust pipe of the combustor is not plotted in figure 6 (shown i figure 2) and that the fuel injection nozzles were choked throughout the cycle.
Figure 6. (left) Ensemble averaged OH’ Chemiluminescence, inverse Abel transform applied. (right) schematic of flow in the combustor. Time of four snapshots marked in figure 5. widthically as the flame emerged from the front of the combustor. From the velocity trace in figure 2 this is likely to occur between approximately 180° and 240° phase.
Figure 7. Amplitude of pressure oscillations over operating range (bar). ee The results show that the vortex structure, caused by the rolling up of the inward jet, is responsible for the mixing of fuel, air and hot- products, and thus controls the ignition delay time and the duration of heat release. The structure of the vortex could be altered by changing the aspect ratio of the inlet pipe or the expansion ratio (inlet pipe diameter to combustor chamber diameter). Because of the coupled nature of the vortex roll-up and the wave dynamics within the combustor the designer should be aware that any alterations to the geometry will affect both However, a compromise may be achievable which further improves the combustion delay time. Figure. 3 shows that only a small part of the total heat release occurs in the region of peak heat release (close to the chamber wall), while the remainder is spread throughout the chamber. This is because the fuel is injected continuously throughou the cycle and thus only a small part of it ends up located close to the combustion chamber wall. If the phase anc pulse width of fuel injection was optimised to ensure that more of the fuel ended in this region at the correct time then it is likely that the Rayleigh efficiency would be increased.
Figure 8. Frequency of pressure oscillations over operating range
Figure 9. Effect of pulse width and fuel mass flow rate on pressure amplitude eS ea 6 ES EE CS I WO Nea The results show that the optimal pulse width of fuel injection is these tests was 80°. This was found to result in the optimal amplitude of pressure gain at all fuel mass flow rates. However, its benefit over steady and other pulse widths of fuel injection was found to decrease as fuel mass flow was raised. It should be noted that the opening and closing times of the present injectors meant that 80° was close to the limit on injector performance.
magnitude is limited. As the fuel mass flow is raised the mixture in this region is likely to become richer. This implies that there is either insufficient oxygen or mixing for complete combustion of the fuel in this region to occur. Figure 10. Effect of pulse width and fuel mass flow rate on operating frequency
It has been shown that the pulsed fuel injection can significantly increase the amplitude of pressure oscillations in the combustor for a given fuel mass flow. It is also interesting to gauge the size of the effect that pulsing the fuel has on the fuel conversion efficiency defined in the introduction. The performance improvement that can be achieved by pulsing the fuel over that achievable with steady fuel, can be measured by the ratio of the net mechanical energy flux out with pulsed fuel relative to that with steady fuel. The net mechanical flux out is given by integrating the kinetic energy flux out of the combustor (pv"/2). If waves traveling in the pulse combustor exit and inlet pipes can be considered one-dimensional acoustic waves were v cc p/p where p is the pressure and pp is the mean density. This means to a first order the ratio of fuel conversion efficiencies at a fixed fuel mass flow is given by Accurate determination of this equation requires knowledge of the time-resolved pressure and density. The time-resolved pressure is known but the time-resolved density is not. The time-resolved and time mean density are, however, primarily a function of the gas temperature. If all the fuel is assumed bumt then the temperature in the combustor is a function of the fuel mass flow rate, which is the same for the two cases, and the mean mass flow of air through the combustor. At the inlet of the combustor the density in the top term is likely to be similar to that in the bottom term. At the exit the temperatures are much higher and the density in the top term will in practice be larger than the density in the bottom term leading to a reduction in ratio of fuel conversion efficiency. This is because the mean mass flow rate of air through the combustor rises as the fuel is pulsed. Recent tests have shown that pulsing fuel causes between a 10% and 15% rise in mean air mass flow rate. For simplicity in this analysis the
descriptionView Paperarrow_downwardDownload
The time-resolved flow within an unsteady ejector (American Institute of Aeronautics and Astronautics, AIAA, Air Breathing Propulsion Best Paper Award 2008)
by Rob Miller

2009

Figure 13 mass averaged velocity from the ejector compared with the simple model, L/D = 6. Firstly, if overall thrust augmentation is known, for example, by experiment then the exit velocity from the ejector can be calculated. A comparison is shown in figure 13. There is some good agreement between the prediction from the model and that predicted by CFD. The slope of the velocity decay is almost correct. However, the peak velocity is too high. The initial acceleration of the slug of fluid is clearly not an impulsive process but happens over a larger part of the cycle, this time depends on the L/D ratio of the primary jet.
C. Implications of the Model
Figure 1 Schematic Representation of an Ejector Ejectors are passive devices used to increase the thrust produced by a jet. When placed downstream of a jet, ambient fluid is entrained into the ejector, producing an aerodynamic force on the ejector. This force is directed towards the jet; conservation of momentum states that the total momentum of the jet is increased by an amount equal to the force on the ejector. Ejectors can be used with steady jets (steady ejectors) or unsteady jets (unsteady ejectors). This paper concentrates on those driven by unsteady ejectors. Thrust augmentation is the ratio of the total thrust produced by the jet and ejector to the thrust produced by the jet alone as defined in equation 1.
Figure 2 Diagram of the experimental arrangement
Figure 4 Time resolved force on the ejector predicted computationally
distance along the ejector as the vortex ring dissipates. Two phenomena are observed, a pressure wave and a convecting vortex ring.
Figure 5 Cp and Entropy plots of the flow field at 6 instances throughout the cycle L/D=7.
Fig 6 Time-Space diagram of stagnation pressure along the centerline of the ejector with mass averaged velocity at the ejector exit plane As vortex ring approaches the ejector it produces a drag force on the ejector and it accelerates the fluid inside the ejector. The low pressure vortex core then moves over the nose of the ejector and creates a positive thrust. The vortex then moves through the ejector entraining fluid behind it.
Figure 7 A Comparison of the thrust augmentation measured to that predicted by the CFD as a function of L/D A. Thrust augmentation as a function of L/D a function of L/D Figure 7 shows the thrust augmentation as a function of the jet L/D ratio. The points from the CFD calculations are in general in good agreement with those from experiment. Thrust augmentation increases almost linearly with L/D to a value of approximately 5 after which it begins to level off before peaking at a value of 7. This is in good agreement with Mason and Miller, 2006.
Figure 8. Time resolved force on the ejector at L/D=4, predicted computationally B. Flow field Figure 9 shows the Cp field and the entropy field around the ejector at a number of time instances throughout th cycle with L/D=4. This can be compared with figure 5 which shows the same flow features but at non necessarily ¢ the same point in the cycle, for L/D=7. Although the same overall flow field is evident there are two mail differences:
Figure 12 Control volume used in calculation During the portion of the cycle in which the fluid in the ejector is slowing and entraining ambient fluid, it possible to use some simple control volume analysis to calculate the force on the ejector. The control volume shown in figure 12. Assuming the flow is quasi steady, the steady flow momentum equation, (7) is applied to tl control volume; where A is area P, pressure and V is velocity. Bernoulli’s equation (6) is used to estimate tl velocity at the throat. It is noted in equation (8) that the force on the ejector, Fx, is the pressure acting on the nose the ejector. From these results equation (9) is derived; this shows that half of the momentum crossing the thro plane of the ejector is due to a force generated by low pressure around the nose of the ejector. For a divergence-fr ejector the momentum entering the plane of the throat will be the same as that leaving it, thus half the momentu leaving the ejector is supplied by an aerodynamic force on the nose of the ejector. The thrust augmentation is 2 fot divergence free ejector. This ratio of thrust generation is also true for an ideally mixed steady thrust augmenti1 ejector with no divergence and an infinite area ratio and for a fan driving a flow within an ejector like du However, for an ejector this implies that the thrust augmentation is always 2.
Figure 11 Fluid in ejector slows as fluid is entrained
Figure 10 Vortex ring accelerates fluid in ejector
Figure 15 Unsteady Force on the ejector, non dimesionalised in different ways a Figure 15 shows the force on the ejector for 4 different values of L/D. When the force on the ejector is normalized with the circulation of the vortex ring, the thrust collapses so that the peaks have approximately the same value. The difference is that the fraction of the period that these peaks, associated with the vortex ring, occupy is reduced as L/D rises. The increased separation between vortex rings at larger L/D ratios enhances their ability to entrain fluid and thus increases thrust augmentation.
Secondly, if it is assumed that the impulse from the slug on the ejector is independent of length then it is predicted that the total thrust augmentation is independent of length. Mason 2006 showed that thrust augmentation is insensitive to length once a critical value has been exceeded. In general it should be expected that viscous forces acting on the ejector will reduce thrust augmentation, therefore the thrust augmentation of long ejectors will be reduced slightly. Although the model does not directly predict the thrust augmentation of a process it does predict the exit velocity from the ejector and how this should vary with length. The change in gradient predicted by the model is similar to that observed in figure 14.
Figure 16 Efficiency of the unsteady ejector as a function of L/D Figure 16 shows the efficiencies of the various ejector simulations. It is shown that the efficiency increases with L/D though the efficiency is still quite low. Although it seems that some of the energy transfer is via pressure forces the efficiency is still relatively low. This is because unsteady jets have a low momentum content for a given energy when compared with a steady jet. As a jet becomes unsteady, it will produce no thrust for part of the cycle, to achieve the same average thrust; the velocity must be increased for the remainder of the cycle. If the velocity is increased then the energy in the jet will be increased by the power of 3.
descriptionView Paperarrow_downwardDownload
A boundary integral method for the planar external potential flow around airfoils
by Mohamed M. S. Nasser

2009

Authors: A.H.M. Murid, Mohamed M.S. Nasser & N.S. Amin This paper presents a boundary integral equation for the external potential flow problem around airfoils without cusped trailing edge angle. The derivation of the integral equation... more
descriptionView Paperarrow_downwardDownload
A boundary integral equation for the 2D external potential flow
by Mohamed M. S. Nasser

2009

Authors: Mohamed M.S. Nasser, A.H.M. Murid & N.S. Amin Based on the recently discovered second kind Fredholm integral equation for the exterior Riemann problem, a boundary integral equation is developed in this paper for the... more
descriptionView Paperarrow_downwardDownload
The archaeology of orbital space
by Alice Gorman

2008, Australian Space Science Conference

More than 10 000 objects larger than 10 cm are currently in earth orbit. These objects include satellites, launch vehicle upper stages, mission-related debris and “space junk”. The amount of debris constitutes a threat for the successful... more
Other objects come from deployments and separations of spacecraft. These manoeuvres typically involve the release of items such as separation bolts, lens caps, flywheels, nuclear reactor cores, clamp bands, auxiliary motors, launch vehicle fairings, and adapter shrouds [8]. Solid rocket motors used to boost satellite orbits have produced various debris items, including motor casings, Orbital debris has been defined as “all man-made [sic] objects injected into orbit which do not now, nor will in the foreseeable future, serve a useful purpose” [5 p128]. The European Space Operations Centre (ESOC) calculates that after approximately 4200 launches since 1957, there are 10 000 objects larger than 10 cm left in Earth orbit. Of these, 7% are operational spacecraft, 52% are decommissioned satellites, upper stages and mission related objects, and 41% are debris from the fragmentation of orbital objects (Figure 1). This has produced an estimated 100 to 150 000 pieces of debris larger than 1 cm [6] with a cumulative mass in excess of 2 000 000 kg [7]. The primary source of orbital debris is the fragmentation of satellites and launch vehicle upper stages, both of which tend to be left in orbit after the end of their mission.
Disposal of human remains in space is likely to become more popular in the future, with suppliers already offering this service on the Internet [10]. Figure 3: Slag from solid rocket fuel Figure 4: Satellite breakup
The Vanguard I satellite, launched successfully on March 17, 1958, is now the oldest manufactured object in orbit (Figure 7). It is no longer transmitting, but is in a highly stable LEO orbit with every prospect of remaining there for perhaps another 600 years. It is a physical testimony to the momentous period when humans first ventured beyond the atmosphere. Despite its failure to be first in the ‘Space Race’, Project Vanguard is acknowledged as ‘the progenitor of all American space exploration today’ [16]. For example, the Minitrack network, set up for Vanguard, became the backbone of the NASA Satellite Tracking and Data Acquisition Network used to track all the early generation of satellites [16].
Only six years after the first satellite, Syncom already shows how satellite design has moved past the early templates of the baby moon (Sputnik and Vanguard) and the rocket (Explorer 1). The Syncom series were the first spin-stabilised satellites. The basic design is still in use, for example, in the Aussat and Optus B series.
Figure 9: Artist’s impression of FedSat in orbit
Figure 10: debris tracked by NASA’s Orbital Debris Observatory
Penetration through an aluminized Mylar foil
Table 1: Commonly used orbital configurations
Table 2. Mean time between impacts on a satellite with a cross-section area of 10 square metres [3].
descriptionView Paperarrow_downwardDownload
2003 - Instability and unsteadiness of aircraft wake vortices
by Vassilis Theofilis

2008

This paper presents a review of theoretical and experimental results on stability and other unsteady properties of aircraft wakes. The basic mechanisms responsible for the propagation and the amplification of perturbation along vortices,... more
descriptionView Paperarrow_downwardDownload
Long-term variability of the low latitude mesospheric SAO and QBO and their relation with stratospheric QBO
by kishore kumar

2008

1] Long term variations of low-latitude mesospheric semi-annual oscillation (MSAO) and quasi-biennial oscillation (MQBO) in the zonal wind are studied using observations available from 1977 to 2006. The MQBO showed large inter-annual... more
descriptionView Paperarrow_downwardDownload
2004 - Issues in active flow control: theory, control, simulation and experiments
by Avraham Seifert and 
1 more
Vassilis Theofilis

2008

The goal of this paper is to provide a perspective on the current status and future directions for active flow-control technology with particular emphasis on oscillatory control. This is not a comprehensive review of the literature;... more
Fig. 1. Left: spatial structure of the chordwise component of the most unstable stationary BiGlobal eigenmode. Right: a schematic representation of vortex shedding from separation bubbles, on account of BiGlobal instability [52]. The recovery of unstable eigenvalues in the problem at hand must be seen in the perspective of the artificial (from a physical point of view) homogeneous Dirichlet inflow boundary conditions. In practice, TS instability in the flow under consideration is orders of magnitude
Fig. 2. Three dimensionalization of the reattachment zone on account of amplification of a BiGlobal eigenmode, in the absence of KH/TS instability [52].
Fig. 3. Regular three-dimensional structures in the wake of a circular cylinder in transitional flow [55]. Experimental findings in support of this conjecture are the following. On a qualitative level, the two- dimensional closed separation bubble on the unswept cylinder first becomes spanwise unsteady with increasing Reynolds number value; the related spanwise periodic vortical cellular structure appearing as a result is shown in Fig. 4. Associated is the formation of Karman-like vortices, having their axes parallel to the generator of the cylinder, while up to this point the same qualitative scenario is followed on the airfoil. The topological description of one of the spanwise periodic cells at the upper side of the cylinder is shown in the lower part of Fig. 4; a pair of streamwise vortices separates from one
Fig. 4. Magnification of Fig. 3 and sketch of the topological characteristics of an isolated structure.
Fig. 5. Upper: asymmetric structures in transitional flow over an airfoil. Lower: topological characteristics of these structures [58].
Fig. 6. Schematic of the generic flap configuration using in the experiments of Nishri [209].
Fig. 7. Typical closed-loop block diagram. environmental disturbances or uncertainties in the models used to of environmenta (e.g. freestream design the controller. The prevalence disturbances in flow control systems turbulence, sound, and vibrations) combined with the fact that most of the models we use to predict fluid flows are approximate, makes the issue of robust contro design of particular importance.
Fig. 8. Modern-control block diagram. and controller yields the compensator as shown in Fig. 8. Ideally, one would like the compensator to perform well even when the system is subjected to external distur- bances, shown as w(f) in Fig. 8. One of the most useful results of modern control theory for linear systems is that for a given model (estimator) one can synthesize a controller that has desirable physical properties (e.g. mi- nimizes drag or unsteadiness) even in the presence of external disturbances, w, that excite the closed-loop system.
Fig. 9. (a),(b) Drag and turbulent kinetic energy (TKE) histories for LES-based terminal turbulent kinetic energy control for different optimization windows T* at R, = 100 [126].
Fig. 10. Schematic of optimal DNS control.
Fig. 11. (a),(b) Drag and TKE histories for LES, DNS, and hybrid LES/DNS-based terminal TKE control for optimization interval T* = 30 at R, = 100 [126].
Fig. 13. Effect of Reynolds number on power savings ratios: — , actual ration of power saved to power input; —-—, conservative estimate of power saved to power input. The ---- shows the average of the two power savings ratios [107]. Fig. 12. Effect of sensing plane location on drag at different Reynolds numbers [107].
Fig. 15. Effect of control on the spatial eigenstructure for the dominant roll mode: (a) no control, (b) opposition control, and (c) optimal control. The solid horizontal line denotes the location of the virtual wall [151].
Fig. 16. Peak exit velocity with lip thickness [186].
Fig. 17. Leading edge and upper surface notch.
Fig. 18. Lift coefficients versus angle of attack (computed using full simulation pressure distributions). [188].
Fig. 20. Sketch of NACA0015/flap configuration [189]. Upper plot shows flap shoulder slot design. Fig. 19. Lift coefficients with angle of attack, where simulation C, was computed using the same pressure tap locations as used in the experiment. [188].
Fig. 22. Residual and lift coefficient (C;) versus the number of iterations (~ = 0°).
Fig. 21. Single-block structured grid. For the initial analysis, the NACA/flap configuration was studied in a free stream environment using a single- block grid that extended 10 chord lengths around the airfoil. Fig. 21 shows the total and near-airfoil region of the grid. For this C-grid, two grids were used to assess grid invariant solutions. First, 197 nodes around the airfoil and 81 nodes in the normal direction were used for the computations. Then, a 385 x 161 grid was used. The coarse grid was insufficient to resolve the flow and only the fine-grid results are presented here. It is recognized that the trailing edge and actuator region may be under-resolved; however, part of the goal of this study is to assess minimum grid resolution requirements for both the baseline (no control) and flow control cases.
Fig. 23. Pressure coefficient with chordwise location (« = —4°). Next, a multi-block grid treatment is used with local enrichment where high-gradient flow conditions exist and the configuration is placed in the tunnel as in the experiment. Inviscid conditions are enforced on the walls of the tunnel and constant enthalpy and entropy are forced as inflow conditions. Fig. 24 shows the multiple blocks configured in the tunnel with attention paid to the key geometric features (actuator, hinge, and trailing edge) and Table 2 lists the grid density in each block.
Fig. 24. Multi-block grid of NACA0015/flap configuration in wind tunnel.
Fig. 25. Residual and lift coefficient (C;) with number of iterations (R = 17 x 10° and 28 x 10° with « = 0°). Note that the curves at the two different Reynolds numbers are nearly identical.
Fig. 26. Streamwise (1) velocity contours at R = 8 x 10° and « = 0°.
Fig. 27. Pressure coefficient with chordwise location (« = 0°; R= 28 x 10°).
Fig. 28. Unstructured mesh around airfoil/flap configuration with notch discontinuity.
Fig. 29. Streamwise velocity contours (R = 8 x 10°; a = 0°).
Fig. 31. Streamwise velocity contours (R = 8 x 10°; a = 0°).
Fig. 33. Generic active flow control process. Once the setup (experimental or numerical) is ready, and the baseline flow is known, the actuator can be turned on. The choice of the frequency, amplitude, phase, waveform and 3D distribution of these para- meters will be discussed later. Many types of actuators exist, and for any type, it is crucial to evaluate the actuator’s performance without the presence of the baseline flow and later to quantify the effect of the external/core flow on the actuator’s performance.
Fig. 34. Computed actuator output in still air (left) and interacting with boundary layer (right) showing the effect on both, (R = 0 and 1200 where Reynolds number is based on the displacement thickness of the incoming boundary layers), courtesy Utturkar et al. [203]). Fig. 36. A cross section (upper part) and a top view (lower part) of the Piezo-bender actuators on the Pr8-SE airfoil [204].
Fig. 35. The IAI Pr8-SE airfoil with ten Piezo-bender actuators (only two are shown schematically) across the 609 mm span of the airfoil [204]. Surface mounted, mechanical actuators, rely on the interaction with the incoming boundary layer to
Fig. 37. RMS velocity fluctuations measured close to the tip of a Piezo-bender actuator in the turbulent boundary layer of the IAI Pr8-SE [204]. where P is the power provided to the actuators (see Fig. 39). A value of F.M.>1 indicates that it is more energy efficient to introduce power to the actuator than to the powerplant. Besides the important applicability
Fig. 38. Lift to Drag ratio of the IAI Pr8-SE plotted against lift due to the activation of the segmented Piezo benders at 2D and 3D modes [204].
Fig. 39. Figure of merit (F.M. see Eq. (10)) of the IAI Pr8-SE due to the activation of the segmented Piezo benders at 2D and 3D modes [204].
Fig. 42. The effect of the excitation amplitude on the frequency response of the separated flow about the “generic flap”’ [6] using zero mass-flux excitation.
Fig. 41. The effect of the excitation amplitude on the frequency response of separated flow [209].
Fig. 43. The combined influence of unsteady and steady excitation for a turbulent boundary layer separation. (a) Sectional side view of the “Hump” model with an excitation slot at x/c=0.64 [211]. (b) Mean and fluctuating model pressures showing the effect of superimposed weak steady mass flux superimposed on the excitation signal. (c) The effect of the added mass flux on the fundamental excitation frequency and its harmonic downstream of the slot. Unsteady C, = 0.23%, R, = 16 x 10°, M =0.25, F* = 1.15.
Fig. 44. Effect of forcing frequency for unsteady excitation of a turbulent boundary layer separation. (a) Mean and fluctuating pressures showing the effect of the excitation frequency, and (b) the evolution of the fundamental excitation frequency down- stream of the x/c = 0.64 slot (vertical dashed line). C, =0.13%, R. = 16 x 10°, M = 0.25 [211].
Fig. 46. Baseline jet, Seifert, 1996 unpublished. Fig. 45. A schematic description of the jet vectoring experi- mental set-up, side view (a) and front view (b) with the diffuser removed [215].
Fig. 47. Deflected jet, excited upper i circumference.
Fig. 48. Jet deflection angle as a function of excitation momentum coefficient [97,98].
Fig. 49. Diffuser and actuator cavity mean pressures due to enhanced cavity pressure fluctuations level causing enhanced jet deflection [97,98].
Fig. 51. Comparison of oscillatory and steady blowing for post-stall lift-recovery of a NACA 0015 at, R, = 300,000 [7].
The goal of this paper is to provide a perspective on the current status and future directions for active flow-control technology with particular emphasis on oscillatory control. This is not a comprehensive review of the literature; rather, certain issues that are often neglected in studies are highlighted showing their importance or impact on the reported observations and targeted outcomes. Feasible routes using flow instability as an efficiency enhancement tool are discussed as an emerging means to explain the physical phenomena of active flow-control and as a tool for control law design and development. Traditional and more recent theoretical approaches to control design are discussed and recommendations are made relevant to numerical complications on the route to design oscillatory flow-control systems. A generic flow control process is put forward and illustrated using experimental examples. © 2004 Elsevier Ltd. All rights reserved. ‘Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy under contract DE-AC04-94AL85000. 0376-0421/$-see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.paerosci.2004.06.001
Status of computational aerodynamics [167] Table 1
Multi-block in-tunnel grids For «=0°, residual history and lift coefficient convergence are shown in Fig. 25 for R= 17x 10° and R = 28 x 10°. The lift is principally independent of Reynolds number in the simulations and a steady flow field is observed. Although the experiments found such Reynolds number independence, the experimental mea- surements found that the flow was highly unsteady. In retrospect, perhaps viscous boundary-conditions are required on the wind tunnel walls, or numerical dissipation suppressed the unsteadiness, or the inflow/ outflow boundary conditions may not properly capture the unsteady flow phenomena for this compressible flow.
descriptionView Paperarrow_downwardDownload
2004 - On instability characteristics of isolated vortices and models of trailing-vortex systems
by Vassilis Theofilis

2008

This paper demonstrates the applicability of a two-dimensional eigenvalue problem approach to the study of linear instability of analytically constructed and numerically calculated models of trailing-vortex systems. Chebyshev collocation... more
descriptionView Paperarrow_downwardDownload
A revisit of spinning disk models. Part I: Derivation of equations of motion
by Natalie Baddour

2008

Previous models of spinning disks have focused on modelling the disk as a spinning membrane. The effect of bending stiffness was then incorporated by adding the appropriate term to the previously derived spinning membrane equation. A pure... more
descriptionView Paperarrow_downwardDownload
A revisit of spinning disk models. Part II: Linear transverse vibrations
by Natalie Baddour

2008

In this paper, two recently derived linear models for the transverse vibrations of a spinning plate are considered. The disk is modelled as a pure plate with no membrane effects. Furthermore, the effect of the rotary inertia of the plate... more
descriptionView Paperarrow_downwardDownload
Laser-induced photothermoacoustic pressure-wave pulses in a polystyrene well and water system used for photomechanical drug delivery
by Natalie Baddour

2008

A linear time-domain thermoelastic (photothermoacoustic) theory of a composite solid-liquid geometry has been developed. The theory includes multiple interreflections at all interfaces, acoustic diffraction and viscosity effects, and... more
descriptionView Paperarrow_downwardDownload
Time-averaged canonical perturbation of nonlinear vibrations of a spinning disk
by Natalie Baddour

2008

This paper considers the dynamics of a spinning disk with the inclusion of in-plane inertia. The inclusion of the in-plane inertia couples the in-plane and transverse vibrations of the spinning disk. By considering one mode of each... more
descriptionView Paperarrow_downwardDownload
Towards a Generic Architecture for Autonomous Landing Systems
by James Goodwin

2008

This paper outlines the design of a 2D analogue for autonomous landing on an asteroid as a hardware demonstrator/development platform, within an overall requirements structure that aims to develop a generic architecture for autonomous... more
descriptionView Paperarrow_downwardDownload
(pdf)
by J D Gibbon

2008

A well known limitation with stretched vortex solutions of the 3D Navier-Stokes (and Euler) equations, such as those of Burgers type, is that they possess uni-directional vorticity which is stretched by a strain field that is decoupled... more
descriptionView Paperarrow_downwardDownload
(.pdf)
by Alfred O. Hero

2008

There has been considerable recent interest in applying maximal invariant (MI) hypothesis testing as an alternative to the generalized likelihood ratio test (GLRT). This interest has been motivated by several attractive theoretical... more
descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

This paper presents a particle filter approach to recursively estimating the joint multitarget probability density (JMPD) for the purposes of simultaneous multitarget detection and tracking. The JMPD is a conditional probability density... more
descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

This paper presents a particle filter approach to recursively estimating the joint multitarget probability density (JMPD) for the purposes of simultaneous multitarget detection and tracking. The JMPD is a conditional probability density... more
descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

This paper addresses the problem of tracking multiple moving targets by recursively estimating the joint multitarget probability density (JMPD). Estimation of the JMPD is done in a Bayesian framework and provides a method for tracking... more
descriptionView Paperarrow_downwardDownload
PDF
by Alfred O. Hero

2008

This paper addresses the problem of tracking multiple moving targets by recursively estimating the joint multitarget probability density (JMPD). Estimation of the JMPD is done in a Bayesian framework and provides a method for tracking... more
descriptionView Paperarrow_downwardDownload
Calculating the lift on a finite stack of cylindrical <b>aerofoils</b>
by Darren Crowdy

2008

The classic exact solution due to Lagally Die reibungslose strömung im aussengebiet zweier kreise. Z. Angew. Math. Mech. 9, 299-305.) for streaming flow past two cylindrical aerofoils (or obstacles) is generalized to the case of an... more
descriptionView Paperarrow_downwardDownload
Energy Saving AND Partition Bridging USING Directional Antennas IN Mobile Ad Hoc Networks
by Anna Ron

2008

In this thesis, I present the design and evaluation of new techniques for using directional antennas to save energy and to bridge network partitions in a mobile ad hoc network. This thesis advocates close but simple collaboration between... more
descriptionView Paperarrow_downwardDownload
Propagation of Shocks through Nonuniformly Heated Gases
by Sergey Nazarenko

2008

A 2D numerical investigation is presented of shock wave propagation into a gas whose density is modulated in the transverse direction across the width of a shock tube. These density modulations represent temperature distributions in which... more
descriptionView Paperarrow_downwardDownload
Fast numerical simulation using a dynamical model for a small-scale turbulent motion
by Sergey Nazarenko

2008

We present numerical simulation of 2D turbulent flow using a new model for the subgrid scales which are computed using a dynamic equation linking the subgrid scales with the resolved velocity. This equation is not postulated, but derived... more
descriptionView Paperarrow_downwardDownload
Orientational order and instabilities in suspensions of self-locomoting rods
by Michael J. Shelley

2008

The orientational order and dynamics in suspensions of self-locomoting slender rods are investigated numerically. In agreement with previous theoretical predictions, nematic suspensions of swimming particles are found to be unstable at... more
descriptionView Paperarrow_downwardDownload
Dynamics of a Deformable Body in a Fast Flowing Soap Film
by Michael J. Shelley

2008

We study the behavior of an elastic loop embedded in a flowing soap film. This deformable loop is wetted into the film and is held fixed at a single point against the oncoming flow. We interpret this system as a two-dimensional flexible... more
descriptionView Paperarrow_downwardDownload
Drag Reduction through Self-Similar Bending of a Flexible Body
by Michael J. Shelley

2008

shows small variations in film thickness. c, Free-streamline model solution for non- dimensional velocity 7 = 27. The pressure field, shown in colour, varies from its maximum value of 1 at the stagnation point on the fibre (black) to its minimum value of ( on the free streamlines (dashed lines); overlaid are the streamlines of the flow. Figure 2 Examples of the flow field and fibre shape in experiment and model. a,b, Flow around a flexible fibre of length 4.1 cm and rigidity 2.8 erg cm, at flow speeds of 69cms | (a) and 144cms | (b). Monochromatic light from a low-pressure sodium lamp reflects from the two surfaces of the film, creating an interference pattern which
Figure 3 Comparison of six experimental fibre shapes with model shapes. a, Shapes from the experiment (green and orange lines) superimposed on a manifold of numerical solutions (black lines). The numerical solutions range from 7 = 5.5 to 33, in increments of 10%. The flat plate (7 = 0) and a very bent solution (7 = 30,000) are shown for comparison. b, Comparison of measured drag on experimental fibres in a with computed drag of nearest numerical solutions, determined by matching tip curvature. c,d, Transition to self-similarity in experimental (c) and numerical solutions (d) (not shown at the same scale). The six fibres, and their nearest numerical solutions, are dilated by 77° and superimposed. The black fibres are numerical solutions at higher », and show the subsequent convergence to a universal shape, as predicted by the model. A solution is found for each value of y, giving fluid velocity and pressure fields, fibre shape, and free streamline shapes. Figure 2c (A similar expression involving the curvature arises, for example, in studying the deformations of an inextensible elastic beam in a very viscous fluid*®.) This closes the model, with the non-dimensional speed y its only parameter. The beam/FST system is solved numeri- cally (S.A., M.S. and J.Z., manuscript in preparation).
Figure 4 Comparison of drag data from experiment and model. a, Drag per unit fibre length versus flow velocity for a flexible fibre (L = 3.3.cm; red circles) and a rigid fibre (L = 2.0.cm; green squares). b, Log—log plot of drag data in a, non-dimensionalized as D = C pn”: also shown are data for two fibres with the same rigidity as the flexible fibre ir a but different lengths (pink plus signs, L = 1.8 cm; blue triangles, L = 5.2 cm). The solid line is the values of Din the model (well-fitted by power laws y @ for» < 1 and» for 7 >> 1, as shown by dash-dotted lines). The short-dashed line is the solid line shiftec by a factor of 2.8 in 7, to correspond with the shift in the shape comparison. This shift compensates for the ‘back pressure’ in the wake, as explained in the text.
Received 13 August; accepted 14 October 2002; doi:10.1038/nature01232.
descriptionView Paperarrow_downwardDownload

Related Topics

  • Aerodynamics
  • Aeronautical Engineering
  • Aircraft Design
  • Flight Dynamics and Control
  • Hypersonic Aerothermodynamics
  • Engineering
  • Mechanical Engineering
  • Spacecraft Engineering
  • LES and DNS
  • Computational Fluid Dynamics (CFD) modelling and simulation
    • 580 California St., Suite 400
    San Francisco, CA, 94104