AIAA 2000-0504 Development of a Flush Airdata Sensing System on a Sharp-Nosed Vehicle for Flight at Mach 3 to 8 Mark C. Davis, Joseph W. Pahle, John Terry White, and Laurie A. Marshall NASA Dryden Flight Research Center Edwards, California Michael J. Mashburn Micro Craft, Inc. Tullahoma, Tennessee Rick Franks Sverdrup Corp. Arnold Air Force Base, Tennessee 38th Aerospace Sciences Meeting and Exhibit 1013 January 2000 / Reno, NV
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics, For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 1801 S.W., Washington, D.C. L'Enfant Promenade,Alexander Bell Drive, Suite 500, Reston, Virginia 22091.
DEVELOPMENT OF A FLUSH AIRDATA SENSING SYSTEM ON A SHARP-NOSED VEHICLE FOR FLIGHT AT MACH 3 TO 8
Mark C. Davis,* Joseph W. Pahle, John Terry White, and Laurie A. Marshall NASA Dryden Flight Research Center Edwards, California Michael J. Mashburn Micro Craft, Inc. Tullahoma, Tennessee Rick Franks# Sverdrup Corp. Arnold Air Force Base, Tennessee

Abstract

NASA Dryden Flight Research Center has developed a ush airdata sensing (FADS) system on a sharp-nosed, wedge-shaped vehicle. This paper details the design and calibration of a real-time angle-of-attack estimation scheme developed to meet the onboard airdata measurement requirements for a research vehicle equipped with a supersonic-combustion ramjet engine. The FADS system has been designed to perform in ights at Mach 38 and at 612 angle of attack. The description of the FADS architecture includes port layout, pneumatic design, and hardware integration. Predictive models of static and dynamic performance are compared with wind-tunnel results across the Mach and angle-of-attack range. Results indicate that static angle-of-attack accuracy and pneumatic lag can be adequately characterized and incorporated into a realtime algorithm. Acronyms FADS INS PPT SCRamjet Symbols C PPT D FADS FADS FADS g
*Aerospace Engineer. Aerospace Engineer, Senior Member. Engineering Consultant, Senior Member. Aerospace Engineer. Instrumentation and Controls Engineer. # Instrumentation Engineer. Copyright 2000 by the American Institute of Aeronautics and Astronautics, Inc. No copyright is asserted in the United States under Title 17, U.S. Code. The U.S. Government has a royalty-free license to exercise all rights under the copyright claimed herein for Governmental purposes. All other rights are reserved by the copyright owner.

Nomenclature

ush airdata sensing inertial navigation system precision pressure transducer supersonic-combustion ramjet
transducer calibration as a function of Mach number diameter of lines from port to transducer, in.
angle of attack at front of vehicle, deg angle of attack at rear of vehicle, deg angle of attack for pseudodifferential transducers, deg
acceleration caused by gravity, ft/sec2 arbitrary integer lag constant pneumatic line length, ft measured pressure at transducer, lbf/ft2 measured pressure at port, lbf/ft2 dynamic pressure, lbf/ft2

i k L L PPPT Pport q 1

American Institute of Aeronautics and Astronautics
q1 s Ve X/L est FADS INS true 3
weighting function Laplace frequency variable effective volume of the measurement system, ft3 body axis location angle of attack, deg bias angle of attack, deg reference angle of attack derived from the FADS algorithm, deg angle of attack derived from the inertial navigation system, deg angle of attack obtained from wind tunnel, deg forward angle-of-attack estimate, deg rear angle-of-attack estimate, deg pseudodifferential angle-of-attack estimate, deg angle of sideslip, deg difference dynamic viscosity of the air in the line, lbm/(ft/sec) time constant
The FADS concept uses a matrix of ush surface ports to infer airdata. The FADS system has been successfully applied to a variety of blunt forebodies,37 and one feasibility study1 has been conducted for a sharp-nosed, hypersonic conguration. To be a viable system, the FADS system must measure angle of attack to within 0.5 (because of the criticality of incidence angle of the engine inlet); measure dynamic pressure to within 5 percent for postight analysis; and survive the intense thermal environment in which a hypersonic vehicle ies.8 This paper presents the architecture, estimation algorithms, and wind-tunnel calibration of a FADS system intended for a sharp-nosed, SCRamjet test vehicle. Note that use of trade names or names of manufacturers in this document does not constitute an ofcial endorsement of such products or manufacturers, either expressed or implied, by the National Aeronautics and Space Administration.
Flush Airdata Sensing System Architecture Overview
This section describes the pneumatic architecture of the FADS system. The port layout, the sensing transducer characteristics, and the pneumatic layout of the pressure sensing system are described. The sensing components that comprise the real-time airdata system are distinguished from those used for the postight algorithm. Pressure Port Layout

Introduction

The National Aeronautics and Space Administration and aerospace community are developing air-breathing propulsion systems capable of ight at hypersonic speeds. One promising concept is the supersoniccombustion ramjet (SCRamjet) engine.1 The current design of SCRamjets allows supersonic combustion to occur only in a narrow operating range. Dynamic pressure ( q ) and angle of attack ( ) are two of the critical parameters that determine the ow into the engine inlet. Accurate measurement of these parameters is desired for real-time control and is required for postight analysis. Accurately estimating angle of attack from the inertial navigation system (INS) alone is difcult because of atmospheric variations and sensor installation and performance.2 This requirement led to the development of a nonintrusive system, the ush airdata sensing (FADS) system, that has the ability to measure angle of attack in real time and allow the remainder of the airdata parameters to be reconstructed postight. A matrix of nine pressure ports is used to sense the airdata parameters. Figure 1 shows the locations of these ports on the vehicle forebody. Four ports (indicated by the highlighted symbols along the centerline of the forebody in gure 1) are used to indirectly sense the angle of attack. The remaining ve pressure ports (indicated by the open symbols in gure 1) are used for postight evaluation of the remaining airdata parameters. To save real-time bandwidth and ensure a high data throughput, the system architecture decouples the angle-of-attack estimation from the remainder of the postight algorithm. Only the pressure data from the angle-of-attack ports is used in real time and combined with the inertial angle of attack to estimate a high-delity, vehicle angle of attack. Pressure Transducers The nine pressures are sensed using a combination of absolute and differential precision pressure transducers
(PPTs). Figure 2 shows the pneumatic layout of these sensors. Differences between the pairs of upper and lower ramp surface pressures (ports 2 and 4; ports 3 and 5) are sensed by differential pressure transducers to provide high accuracy and a high-resolution measurement for use by the real-time algorithm. Each differential pair is also teed to an absolute pressure transducer that allows the absolute pressure level at each port to be sensed or calculated. The forebody side ports (ports 6 and 8; ports 7 and 9), although not used by the real-time algorithm, are sensed in a similar manner. The single stagnation pressure (port 1) is sensed using an absolute sensor. All of the pressure transducers have serial digital outputs, which are connected through an individually addressable, multidrop RS-485 bus. The sensors also provide an optional analog output. The PPT digital output is the primary signal used in the real-time and postight algorithms. The analog signal is recorded only for postight analysis and provides data redundancy if the digital signal fails. The pressure transducers use a piezo-resistive bridge technology and have a built-in digital temperature compensation over a range from 40 to 80 C. The manufacturers specied accuracy for the sensor output for both digital and analog is 0.05 percent of full scale.9 Laboratory tests conducted at the NASA Dryden Flight Research Center (Edwards, California) have shown the sensors to be accurate to within 0.025 percent of the full-scale value. Table 1 shows the types of sensors used in this design and the sensor full-scale

ranges. The sensors used as a part of the real-time system architecture are indicated with an asterisk. Pneumatic Layout Table 2 shows the line lengths, tubing diameters, and entrapped volumes for the various pneumatic components (g. 2). The effective volume in table 2 also includes the entrapped volume of the pneumatic ttings and the transducer volumes. Results from preliminary oblique shock theory10 and engineering judgment were used in the placement of the pressure ports on the vehicle. The pressure port size on the upper and lower ramp surfaces was 0.04-in. diameter. The pressure port size on the leading edge and sides of the vehicle had a diameter of 0.02 in. to limit stagnation heating effects. All ports were drilled normal to the surface. The teed pneumatic lines required to obtain the absolute pressure levels for ports 3 and 5 and 6 and 8 are a cause for concern because of latencies that may be introduced into the sensed pressure signals. These latencies are especially critical for the real-time sensing system. The effects of these latencies will be analyzed in detail in the Results and Discussion section.
Table 2. Pneumatic layout characteristics. Line number Line length, in. 35 Tube diameter, in. 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 0.063 Volume, in3 0.3502 0.2381 0.2049 0.2340 0.1176 0.1550 0.1509 0.2007 0.2090 0.2049 0.2007 0.1509 0.1966
Table 1. Sensor type at each port location. Sensor Parameter Port identication sensed PPT 1 PPT 2 * PPT 3 * PPT 4 * PPT 5 * PPT 6 PPT 7 PPT 8 PPT 2, 3, 6, 7, Total pressure Sensor type Absolute Differential Absolute Differential Absolute Differential Absolute Differential Absolute Range, lbf/in5 015
L1A L2A L2B L3A L4A L4B L5A L6A L6B L7A L8A L8B L9A
Wind-Tunnel Facilities, Equipment, Test Conditions, and Procedures
This section describes the facilities, procedures, equipment, and tests conditions for a series of wind-tunnel experiments conducted to evaluate the FADS system. The basic measurement systems were evaluated over a broad range of Mach numbers, and a data set allowing a preliminary airdata calibration was obtained. Facilities All wind-tunnel testing occurred at the Arnold Engineering Development Center (Arnold Air Force Base, Tennessee) Von Karman Facility in tunnels A and B. Tunnel A is a 40- by 40-in., continuous, closed-circuit, variable-density, supersonic wind tunnel with a Mach number range of 1.5 to 5.5. The tunnel is served by a main compressor system that provides a wide range of mass ow and stagnation pressures to a maximum of 195 lbf/in2 absolute.11 Tunnel B is a continuous, closed-circuit, hypersonic wind tunnel with a 50-in.diameter test section. Tunnel B uses two axisymmetric, contoured nozzles that provide two xed Mach numbers of 6 and 8 with an operating pressure range of 20 to 300 lbf/in2 absolute at Mach 6 and 50 to 900 lbf/in2 absolute at Mach 8.11 Wind-Tunnel Test Equipment Figure 3 shows the internal layout of the test article with nine PPTs and one inclinometer. The sensors were enclosed in cooling jackets to ensure that the sensor operating limits were not exceeded during the test. An inclinometer measured the model incidence angle over a range of 14.5 with an accuracy of 0.02-percent full scale. The model used in the test was an 80-percent scale model of the SCRamjet test vehicle forebody. The model was designed for hypersonic testing for extended periods. The model was milled from solid bar stock of heat-treated and solution-annealed 316 stainless steel.12 The model had a boundary-layer trip strip installed just aft of pressure port 4 (g. 1). The wind-tunnel pneumatic system was designed to duplicate the ight hardware. Analog and digital outputs from the PPTs were sensed during the wind-tunnel tests. Digital data were polled from all PPTs at a rate of 48.8 samples/sec. Analog data were obtained using a 16-bit analog-todigital converter unit controlled by the wind-tunnel 4

computer. Figure 4 shows a schematic of the data acquisition system used for the wind-tunnel tests. Figures 5 and 6 show the model as mounted in tunnels A and B for testing. Wind-Tunnel Test Procedures and Conditions Wind-tunnel data were taken during constant angles of attack and sideslip and during pitch-pause runs with sweeps in angles of attack and sideslip. Data were obtained over a Mach number range of 3 to 8, an angleof-attack range of 6 to 12, and an angle-of-sideslip range of 3. In the pitch-pause maneuvers, data were obtained in 1-deg increments. Angle-of-sideslip data were obtained in 0.5-deg increments. The dwell time at each pitch-pause data point was approximately 15 sec. Table 3 shows the wind-tunnel conditions.
Real-Time Angle-of-Attack Estimation Algorithm
The primary function of the real-time angle-of-attack estimation algorithm is to provide a pneumaticallybased measurement estimate of the bias in the INSderived angle of attack. The real-time FADS algorithm is composed of two basic routines, FADS calibration and signal selection. These algorithms require Mach number, which is provided by the INS. At relatively high velocities, inertial Mach number is sufciently accurate when used with a representative atmospheric model. For the sensor conguration shown in gure 2, only three unique angle-of-attack estimates are available, although four pressure ports and four pressure sensors are designated for real-time angle-of-attack estimation. The individual angle-of-attack measurements are as follows: ( P PPT 2 ) 1 = ----------------------- C PPT 2 q ( P PPT 4 ) 2 = ----------------------- C PPT 4 q ( P PPT 3 P PPT 5 ) 3 = --------------------------------------------- C PPT 53 q where 1 is the forward angle-of-attack estimate, 2 is the rear angle-of-attack estimate, and 3 is the pseudodifferential angle-of-attack estimate. , (1)
Table 3. Wind-tunnel test summary. Test condition sweep at 6 sweep at 4 sweep at 2 sweep at 0 sweep at 2 sweep at 4 sweep at 6 sweep at 8 sweep at 10 sweep at 12 sweep at 0 sweep at 3 sweep at 0 sweep at 3 sweep at 0 sweep at 3 Mach number 34568 Basic Basic Basic Basic Basic Basic Basic Basic Basic Basic Hysteresis/ Lag effects Hysteresis/ Lag effects Reynolds number effects Reynolds number effects Reynolds number effects Reynolds number effects Reynolds number, mil/ft 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 1.80

Remark

attack for each Mach number is required. These steadystate calibration curves were initially predicted using engineering methods, then rened with wind-tunnel data. The block diagram in gure 7 shows these calibration curves implemented as two-dimensional table lookups. The sensor selection routine is used to determine outof-range or failed FADS sensors. Because the ight control system is single-string, the INS angle of attack is assumed to be an unfailed but biased estimate of true angle of attack. The INS angle of attack is passed through a rst-order lag lter corresponding to each FADS angle-of-attack pneumatic lag model derived from wind-tunnel data. (This model will be described in the Results and Discussion section.) These lagged INS angle-of-attack signals are then compared to the three corresponding FADS angle-of-attack signals. A FADS angle-of-attack signal is considered failed if this comparison exceeds a threshold for a xed length of time. The threshold is a function of Mach number and is dependent on the amount of lag that can be tolerated by the system. The nal FADS angle of attack is the average of the unfailed signals. This nal FADS angle of attack is then used to bias the INS angle of attack through a rstorder lter as shown in gure 7. If all FADS sensors are declared failed, the bias will fade to 0 and the uncompensated INS angle of attack is used in the ight control laws. Other signicant airdata parameters sensed by the FADS system are derived from postight data using nonlinear regression algorithms. Reference 1 details how these postight airdata estimation algorithms are developed.

Results and Discussion

This section discusses the data obtained in the wind-tunnel test. Results are compared with both predicted static pressure and simulated pneumatic lag results. Steady-State Pressure The calibration curves used to derive angle of attack from the pressure data were initially developed using engineering analysis. Newtonian ow theory was used to obtain stagnation pressure. Oblique shock theory or Prandtl-Meyer expansion methods,10 depending on angle-of-attack and ow conditions on the wedge, were used to solve for surface pressures on the wedge itself. A
Figure 7 shows the angle-of-attack estimation algorithm in block diagram form. For PPT 2, PPT 4, and the difference between PPT 3 and PPT 5, a calibration curve of differential pressure as a function of angle of
wind-tunnel test was then conducted to validate the initial pressure model of the FADS system. Figures 8(a)(d) show the comparison of the predicted pressure model and the wind-tunnel data as a function of angle of attack for Mach numbers of 3, 4, and 8. The wind-tunnel data are shown as open symbols and the predicted data are shown as solid symbols. Results for the other Mach numbers listed in table 3 are similar to the Mach 8 results. Results from port 2 (g. 8(a)) indicate that the predicted pressures compare very well with the wind-tunnel pressures. Results from port 4 (g. 8(c)) indicate similar results; the exception is the Mach 8 case in which pressure for the high angles of attack was underpredicted. This slight underprediction may be caused by ow separation at the high Mach numbers. The two rear ports (ports 3 and 5) show large differences, especially on the lower ramp port 5 (g. 1). The results for port 3 (g. 8(b)) indicate good agreement, except for the Mach 3 case in which a small slope change appears in the wind-tunnel results. The cause for this difference is unknown, but may be because of data acquisition errors. The results for port 5 (g. 8(d)) indicate an overprediction of the pressures at the high angles of attack at Mach 3 and 4 and an underprediction at Mach 8. This difference is most likely caused by the presence of the boundary-layer trip strip located in front of port 5. The simple prediction models used for the wind-tunnel comparisons could not include a boundary-layer trip strip. The boundary-layer trip strip was installed on the model in a manner similar to that planned for the ight vehicle. Overall, the predicted pressures compared well with the wind-tunnel pressures for ports located forward of the boundarylayer trip strip. Additional corrections for boundarylayer trip strip effects could be developed for port 5 to reduce the errors even further. Pressure data obtained from the wind-tunnel test were used as input to the angle-of-attack estimation routines previously described (g. 7). True angle of attack and tunnel dynamic pressure were used as inputs instead of the INS parameters that will be used in the ight software. Figure 9 shows angle-of-attack error ( true FADS ) across the angle-of-attack range for the same Mach numbers as shown in gure 8. Figure 9(a) shows the angle-of-attack error for the forward pair of ports, and gure 9(b) shows the angle-of-attack error for the aft pair of ports.

The angle-of-attack error shown for the forward ports generally is less than 0.2 at less than 6 angle of attack, and is less than 0.5 across the entire angle-of-attack envelope. These excellent results are consistent with the pressure results shown in gures 8(a) and 8(c). The results for the aft pair of ports show large angle-ofattack errors, especially at the high angles of attack. The trends in angle-of-attack error are consistent with the errors in predicted pressures shown in gures 8(b) and 8(d). The results show the viability of the real-time angle-of-attack estimation method. Pneumatic Lag Because the FADS system is pneumatically-based, pressure lags must be taken into account. For the current angle-of-attack estimation design, the pneumatic lag models are used in the sensor selection routine to determine out-of-range or failed FADS sensors. A pressure lag model was developed for each port (or pair of ports) in the system because tubes of different length were used for each sensor (gs. 12 and table 2). The pressure lag from each port to sensor was modeled as a rst-order lag:13 P PPT k ------------ = ---------- , s+k P port (2)
where k is a nonlinear function of the measurement geometry and the input pressure Pport. The lag constant, k, can be represented by the following form: P port g D k = -- = ------------ ------------------- 128L Ve
where D is the diameter of the tube, L is the tube length, and Ve is the effective volume. Equations (2) and (3) characterize the lag from a single port to an absolute pressure measurement. Figure 10 shows wind-tunnel data from a Mach 6, dynamic, pitch-pause angle-of-attack sweep (6 to 12). As seen in the absolute pressure measurements (PPT 3 and PPT 5), the lag characteristics change signicantly over the pressure range as predicted by equation (3). In contrast, the lag characteristic of the differential pressure transducer from a pair of ports (PPT 2) remains relatively constant across the pressure range. This empirical observation allows the differential pressure lags to be adequately characterized by equation (2) with a constant lag factor across the measurement
range. In other words, the lag model for a pair of ports to a differential pressure measurement is only a function of Mach and not a function of the input pressure, thus greatly simplifying the lag characterization. An analog matching technique was used to estimate the lag constant for the differential pressure signals across the angle-of-attack measurement range. True angle of attack was converted to unlagged pressure by using the inverse angle-of-attack estimation algorithm shown in gure 7. The resulting pressure was then used as an input to a constant rst-order lag model to obtain the simulated pressure at the PPT. The lag constant was varied in order to minimize the error between the lagged results and the actual differential pressure signal, and thus to obtain the best t over the entire range. Figure 11 shows a typical result for one of the pitchpause angle-of-attack sweeps (at Mach 6). A time history of scaled true angle of attack is shown with the actual and simulated differential pressure for the forward pair of ports. Three sections of the time history are magnied to show the very good agreement between the simulated and actual signals, especially in the low-angle-of-attack range. These results show that the pneumatic lags can be characterized by a rst-order lag, where the lag constant is only a function of Mach. Figure 12 is a summary of the lag characterization for all three FADS angle-of-attack signals across the tested Mach number range. These lags are accounted for in the real-time algorithm as previously described in the RealTime Angle-of-Attack Estimation Algorithm section.

Based on dynamic wind-tunnel results, characterizing the lag from a pair of ports to a differential transducer as a constant rst-order lag is possible. The pneumatic lag models are used to determine out-of-range or failed FADS sensors in the real-time angle-of-attack algorithm. Wind-tunnel results for static and dynamic pressure data validate the prediction models and the FADS architecture. The wind-tunnel results show that the performance of a FADS system for a sharp-nosed, wedge-shaped vehicle can be designed to meet the requirements for accurate measurement of angle of attack for real-time control and for postight analysis.

References

Stephen A. and Timothy R. Moes, Measurement Uncertainty and Feasibility Study of a Flush Airdata System for a Hypersonic Flight Experiment, NASA TM-4627, 1994. Dale F., Dan J. Bugajski, John Carter, and Bob Antoniewicz, Multi-Application Controls: Robust Nonlinear Mulivariable Aerospace Controls Applications, Fourth High Alpha Conference, CP-10143, vol. 2, 1994. John P. and Earl R. Keener, Flight Evaluation of the X-15 Ball-Nose Flow-Direction Sensor as an Air-Data System, NASA TN-D-2923, 1965. Paul M. III, Martin W. Henry, and James B. Eades, Jr., Shuttle Entry Air Data System (SEADS) Advanced Air Data System Results: Air Data Across the Entry Speed Range, Orbiter Experiments (OEX) Aerothermodynamics Symposium, CP-3248, Part 1, Apr. 1995, pp. 4978. Terry J., Stephen A. Whitmore, L. J. Ehernberger, J. Blair Johnson, and Paul M. Siemers III, Qualitative Evaluation of a Flush Air Data System at Transonic Speeds and High Angles of Attack, NASA TP-2716, 1987. Stephen A., Timothy R. Moes, and Terry J. Larson, Preliminary Results From a Subsonic High Angle-of-Attack Flush Air Data Sensing (HI-FADS) System: Design, Calibration, and Flight Test Evaluation, NASA TM-101713, 1990. Stephen A., Brent R. Cobleigh, and Edward A. Haering, Design and Calibration of the X-33 Flush Airdata Sensing (FADS) System, NASA TM-1998-206540, 1998.
7Whitmore, 6Whitmore, 5Larson, 4Siemers, 3Cary, 2Enns, 1Whitmore,

Concluding Remarks

The design of a ush airdata sensing (FADS) system for a sharp-nosed, wedge-shaped vehicle has been described. Real-time angle-of-attack estimation from the FADS system can be used to bias an inertial navigation system angle of attack. Wind-tunnel tests were conducted to validate the predicted static and dynamic characteristics of the FADS system. The predicted static pressures for a matrix of ports compared well with the wind-tunnel results. Calibration curves were developed to convert differential pressures to angle of attack. Using ports forward of the boundary-layer trip strip results in angle-of-attack errors less than 0.2 at less than 6 angle of attack, and less than 0.5 for the entire angle-ofattack range.

John D., Jr., Hypersonic and High Temperature Gas Dynamics, McGraw-Hill Book Company, New York, 1989. State Electronics Center, Precision Pressure Transducer PPT and PPT-R: Users Manual Version 2.4, Honeywell, Inc., 1996. Research Staff, Equations, Tables, and Charts for Compressible Flow, Report 1135, 1953. A. H., Performance and Operational Characteristics of AEDC/VKF Tunnels A, B, and C, AEDC-TR-80-48, July 1981.
11Boudreau, 10Ames 9Solid

8Anderson,

Eugene A. and Theodore Baumeister III, eds., Marks Standard Handbook for Mechanical Engineers, 10th ed., McGraw-Hill, Boston, Massachusetts, 1996. J. P., Jr., The Inuence of Geometry Parameters Upon Lag Error in Airborne Pressure Measuring Systems, WADC TR-57-351, July 1957.

13Lamb,

12Avallone,

X/L = 0 Port 9 Port 8

X/L = 0.31

X/L = 0.67

Port 4 Port 5
Port 2 Port 3 Port 1 Port 1

(Port 2) Port 4

Boundarylayer trip strip

(Port 3)

Port 5
Port 6 Port 7 Looking aft ( ) Indicates port is located on test article upper fuselage [ ] Indicates port is located on test article right side of the fuselage X/L = 0
Bottom planform X/L = 0.40 [Port 9] Port 6 X/L = 0.67 [Port 8] Port 7
Boundary-layer trip strip Left side view

990448

Figure 1. Test article pressure port locations.
Absolute PPT Differential PPT

Upper surface

Port 2 L2A L2B
Port 3 PPT 3 L3A L4A L4B PPT 4 PPT 2 Port 5 L5A PPT 5

Lower surface

Port 4 Boundary-layer trip strip Side internal view Port 8

Port 9 Right side

PPT 1 PPT 9 PPT 8 PPT 6 L7A PPT 7

Port 1

L8A L8B

Left side

Port 6 Top internal view Port 7

990449

Figure 2. Wind-tunnel model pressure transducer connectivities.

Cooling jacket

Inclinometer

990450

Figure 3. Internal layout of wind-tunnel model.

Ground

Directcurrent power
Analog signal acquisition system with signalconditioning boards and analog-to-digital conversion boards

Control Room

Computer To facility computers
RS-485 to RS-232 converters Inclinometer PPT PPT PPT PPT PPT PPT PPT PPT PPT Test article

990451

Figure 4. Wind-tunnel test setup.
Photograph courtesy of U. S. Air Force, AEDC, 98-104210.
Figure 5. Test article in Tunnel A test section.
Photograph courtesy of U. S. Air Force, AEDC, 98-106729.
Figure 6. Test article in Tunnel B test section.

x PPT 2/q PPT 2 calibration FADS

FADS calibration

limiter [10, 20]

Signal selection

1 PPT Mach
FADS good q1 = FADS bad q1 = 0 1
2 FADS 2 [10, 20] 3 FADS 3

Good channel average

x PPT 4/ q PPT 4 calibration (PPT 5PPT3)/ q

qi FADS

2 q2 FADS
2 PPT Mach 6 q FADS calibration
Validity model, FADS good q = 2 lag model, and FADS bad q2 = 0 diagnostics 2

qi 3 good q3 = 1

FADS bad q3 = 0 3

limiter

3 PPT 3 x pressure 4 PPT 5 +
7 Mach PPT 53 calibration
Figure 7. FADS and S estimation and integration block diagram.

Bias filter

1 s+1 + +

990452

Wind-tunnel measured Predicted

Pressure

Mach 3 Mach 4 Mach 8

990453

(a) Pressure as a function of angle of attack for port 2.

990454

(b) Pressure as a function of angle of attack for port 3. Figure 8. Comparison of empirical and wind-tunnel data for the FADS.

Mach 3

Mach 4 Mach 8 Pressure

990455

(c) Pressure as a function of angle of attack for port 4.

990456

(d) Pressure as a function of angle of attack for port 5. Figure 8. Concluded.

FADS,.5 deg

FADS, deg

true, deg

990457

990458

(a) Forward ports ( FADS ).
(b) Aft ports ( FADS and FADS ). 2 3
Figure 9. Angle-of-attack error ( true FADS ) as a function of true angle of attack.
PPT 3 (absolute forward, upper port) PPT 5 (absolute aft, lower port)
Measured pressures, lbf/ft2
PPT 2 (differential forward ports) PPT 2 PPT 3 PPT 150 Time, sec 300

990459

Figure 10. Typical wind-tunnel pitch-pause angle-of-attack sweep for Mach 6.
Scaled PPT 2 actual PPT 2 simulated PPT 2 (actual and measured), lbf/ft2

150 Time, sec

step from

5 to 4

3 to 4

9 to 10

PPT 2 (actual and measured), lbf/ft2

36 Time, sec

152 Time, sec

248 Time, sec

990460
Figure 11. Simulated and actual pressure lag characteristics for Mach 6 pitch-pause angle-of-attack sweep.
FADS (PPT 2) 1 FADS (PPT 4) 2 FADS (PPT 5 PPT 3) 3
9 Lag constant, rad/sec 7

5 Mach number

990461
Figure 12. Summary of lag characteristics for all three angle-of-attack estimations as a function of Mach number.
REPORT DOCUMENTATION PAGE
Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Ofce of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.
1. AGENCY USE ONLY (Leave blank)

2. REPORT DATE

3. REPORT TYPE AND DATES COVERED

January 2000

4. TITLE AND SUBTITLE

Conference Paper

5. FUNDING NUMBERS
Development of a Flush Airdata Sensing System on a Sharp-Nosed Vehicle for Flight at Mach 3 to 8

6. AUTHOR(S)

WU 522-51-54-00-50-00-X43
Mark C. Davis, Joseph W. Pahle, John Terry White, Laurie A. Marshall, Michael J. Mashburn, and Rick Franks
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER
NASA Dryden Flight Research Center P.O. Box 273 Edwards, California 93523-0273
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

H-2390

10. SPONSORING/MONITORING AGENCY REPORT NUMBER
National Aeronautics and Space Administration Washington, DC 20546-0001

11. SUPPLEMENTARY NOTES

AIAA 2000-0504
Paper presented at 38th AIAA Aerospace Sciences Meeting and Exhibit, 10-13 January 2000, Reno, NV, AIAA 20000504. M. Davis, J. Pahle, J. White and L. Marshall of NASA Dryden Flight Research Center, Edwards, CA. M. Mashburn of Micro Craft, Inc., Tullahoma, TN. Rick Franks of Sverdrup Corp., Arnold AFB, TN.
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
UnclassiedUnlimited Subject Category 06 This report is available at http://www.dfrc.nasa.gov/DTRS/

13. ABSTRACT (Maximum 200 words)
NASA Dryden Flight Research Center has developed a ush airdata sensing (FADS) system on a sharp-nosed, wedge-shaped vehicle. This paper details the design and calibration of a real-time angle-of-attack estimation scheme developed to meet the onboard airdata measurement requirements for a research vehicle equipped with a supersonic-combustion ramjet engine. The FADS system has been designed to perform in ights at Mach and at 612 angle of attack. The description of the FADS architecture includes port layout, pneumatic design, and hardware integration. Predictive models of static and dynamic performance are compared with wind-tunnel results across the Mach and angle-of-attack range. Results indicate that static angle-of-attack accuracy and pneumatic lag can be adequately characterized and incorporated into a real-time algorithm.

14. SUBJECT TERMS

15. NUMBER OF PAGES
Airdata calibration, FADS, Flush airdata sensing system, Hypersonics, Wedge forebody, Wind tunnel test
17. SECURITY CLASSIFICATION OF REPORT 18. SECURITY CLASSIFICATION OF THIS PAGE 19. SECURITY CLASSIFICATION OF ABSTRACT
16. PRICE CODE 20. LIMITATION OF ABSTRACT

Unclassied

NSN 7540-01-280-5500

Unlimited

Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. Z39-18 298-102

Sharp Corporation: Developing Next-Generation Products1 Haruo Tsuji, president of Sharp Corporation, budgeted $1 billion for LCD-related product developments between 1993 and 1995. Tsuji explained Sharps strategy:
At Sharp we intend to channel all our efforts into creating next-generation products by developing and exploiting key devices. For example, it would have been impossible to create the LCD Hi 8 ViewCam without new types of liquid crystal display. We will press forward with creation of the next-generation of outstanding new products to attract customers and stimulate new consumer demand -- products such as the new ViewCam that can transmit still-video images over ordinary phone lines, and personal information tools with built-in fax transmission capabilities.
Sharp's strategy of developing key devices for use in future products had resulted in a continuing introduction of new products. According to senior executive vice president Wada:
We may have a large number of products that are being revised or sold in a given year, but the major developments are not so many in number. Those products are primarily divided into high, middle and low-end products. They each have a different set of features that are required. If we have three new product in one season, it is not too difficult to manage. In the whole line-up of new products, there may be as many as 20 new products in a given year. There are usually 3 to 5 new products in a quarter.
Of the 8 best selling electric appliances for 1993, Nikkei Research Center found that four were Sharp products. Since the introduction of Sharps VIP refrigerator in February, 1993, it had shipped 80,000 units by September, raising Sharp's market share in refrigerators 1.1%. Since the introduction of its wide screen TV, it had increased market share 0.9%. Nikko Research Center commented on Sharps new products:
In 1992, Sharp introduced the LCD-ViewCam. The new video camera was based on the development of a low reflection LCD display that was not easy for competition to duplicate. The Awash, introduced in 1992, was a new type washing machine based on sensor controlled positioning technology that was not easy for competition to duplicate. The VIP refrigerator reduced the outside size without reducing internal dimensions by using new silicon dioxide pellets in a vacuum environment to reduce insulation thickness. Competitors would be able to follow with moderate difficulty. The wide-TV set was introduced at the lowest price in the market by applying IC technology that was considered relatively easy to duplicate.
The company's new product development activity was orchestrated by Sharp's senior vice president for product development, Atsushi Asada. Sharp's "kinkyu" or "emergency program" system was applied to develop promising new products. The VIP refrigerator was one example of an emergency project to save space without reducing inside capacity, and to
Copyright 1994 by William R. Boulton, Olan Mills Professor of Strategic Management, Auburn University, and Kosei Furukawa, Professor of Management, Keio University. This case was developed with support from Keio University's Graduate School of Business Administration.
develop materials to cut heat and to allow for low cost production. The team members examined over 100 materials before coming up with silicon dioxide as a filler in a vacuum which reduced overall space without reducing capacity. The Awash was inspired after president Tsuji watched an elderly woman having to use a step to reach into the washing machine and use a hook to pull out clothes. The new washing machine is lower in height. According to Tsuji,

My company must put itself in the customers shoes as it works to create next-generation products. We can neither look smugly down on users nor fawn upon them if we intend to turn out products they will regard as valuable. What we must do is build a kind of creation-intensive company capable of creating value.
Tsuji constantly walked into stores to see what was selling and was the promoter of product development from the users point of view. According to Tetsuo Tani, an executive director of Sharp:
The president spends a lot of time looking for innovations for new products. He is interested in understanding the consumer's mind. Finance is secondary in a consumer oriented company like Sharp. The president is always looking for technological innovations and trying to understand the mind of the consumer which will affect future demand. He is always suggesting that we do something, providing real leadership to the company. That is why Sharp is an innovative company.
According to Sumy Otani, general manager of corporate public relations, Tsuji even inspirated the ViewCams development:
The vision for the ViewCam came from president Tsuji. Mr. Tsuji came out of the TV and video systems group, so he knows about these technologies and he knows how to organize and manage these development activities. Since the people know that he understands the business well, the top down approach works well here. He also encourages people to make suggestions from the bottom up. With regards to the development of the ViewCam, the person in change of the TV and video systems group asked for help from the research and development group.
Sharp was a strong competitor is its markets as shown in Table 1. Examples of Sharps market share positions include:
Color TVs VCRs Camcorders Home phones Electronic calculators Japanese word processors Plain paper copiers Microwave ovens Refrigerators 15.5% 12.0% 15.2% 22.0% 36.0% 20.2% 6.9% 20.9% 13.0%
Sharp held over 20 percent shares in markets like microwave ovens, word processors, electronic calculators, and home telephones. Camcorders were expected to reach such levels in 1994.
Sharp's Product Development System Sharp was clearly recognized as one of Japan's most creative companies. Sharp's continued innovations derived from the company's ability to commercialize its technologies. The company's ability to rapidly develop and introduce new products began as a result of the 1970s calculator battles. As a late entrant, Casio had dramatically cut prices and forced Sharp to discard market research, demand forecasting and organizational planning in order to respond quickly. According to Asada "The current product development system grew out of the calculator development problem. Out of that battle came the view that we needed to emphasize the consumer's point of view." Sharps kinkyu system had its origin in 1972 with the development of the LCD for electronic calculators. It had a completion goal of April, 1973, and took about one year to accomplish. Ichiro Fujimoto, senior executive director in charge of corporate R&D, explained:

Kinkyu projects are funded by headquarters when the results are short term, the organizational impact is significant, and the potential for sales is large. Each proposal is assigned points by a review team that I chair. If the points are high enough, then the project is submitted to the monthly engineering meeting. We make judgments about the needed resources. Each proposal has its own proposed budget, but I will ultimately decide on the budget. Typically, three years after the project is completed, the business is asked to repay half of the investment of the kinkyu project back to headquarters.
The development of kinkyu projects typically took from one to onn and a half years to complete. Senior vice president Wada provided another view:
We dont like corporate bureaucracy. We do everything we can to lessen or destroy bureaucracy. When Mr. Tsuji is reviewing a products development, for example, and there is some need for help from another department, he will ask the general manager from that department what he is doing to help in this products development. If he is not helping, Mr. Tsuji will tell him to get someone to help immediately. This way we break down any organizational boundaries. We dont like organizational barriers.
Longer term projects could be the responsibility of a business division's R&D group, and could get funding from the corporate R&D group. According to Fujimoto, "When we get involved, it means that it is a corporate-wide effort. For long term, basic research projects, we use our own judgment. If a project took over two years to complete, was not product-oriented, or didn't require multi-division involvement, then it would not qualify as a kinkyu project. Fujimoto further described the kinkyu projects:
Since each organization has its own boundary, the kinkyu project creates a driving force to make something happen. About 80 percent of the projects are product development projects. There are some component projects, with small unit prices, but sales are expected to be high. I have responsibility for these projects. The remaining 20 percent are for software development or service development for new businesses outside of my area.

The product development committee will decide on priorities for new product development. The general managers will give their input to the development needs, but when the discussion gets intense, the young engineers are brought in front of the president. He then discusses the details with them. He takes such a keen interest that he knows the details very well. Sometimes we will discuss
the product concept, sometimes we will review the prototype. He then follows up by finding out how the product actually sells. So we are very product oriented.
Figure 1: Sharp's Kinkyu Project Planning System
Corporate R&D Group Executive Directors Meeting Manufacturing Groups
Mid-Long Term R&D Strategy
Corporate Technical Strategy Meeting
Mid-Long Term OperationStrategy

R&D Planning Meeting

Operations Technology Development Strategy Laboratory Directors Meeting Operations Technology Developpment Strategy Meeting

R&D Strategy Meeting

Coordination (Request, Transfer of Technology & Technology Exchange)
Research Laboratories Corporate R&D Group
Gold Badge Kinkyu Projects Teams
Research Laboratories Manufacturing Groups/Divisions
Figure 2: Sharp Corporation's Recent Product Releases
RELEASE DATE: PRODUCT INTRODUCTION:
January, 1987 April, 1988 July, 1988 November, 1988 January, 1989 April, 1989 June, 1989 September, 1989 November, 1989 June, 1990 September, 1990 October, 1990 December, 1990 January, 1991 March, 1991 April, 1991 May, 1991
Electronic Organizer PA-7000: First kanji-based organizer. Limited-Range Cordless Telephone CJ-S100: Included intercom. Super-compact Word Processor WV-500: Used A4 batteries. Fully Automated Hot Water Washing Machine ES-V458: Integrated water heater. Dual-Swing Door Refrigerator SJ-38WB: First bi-directional door. Upright Vacuum Cleaner EC-S35: Integrated telescopic hose. LCD Projector XV-100Z: Flexible screen size. Cordless Telephone & Answering Machine CJ-A300: Accessible from extension phone. Lightweight Headphone Stereo Cassette Player JC-K99: World's lightest (99g) by using carbon glass fiber case. Notebook Computer "All in Note": Integral hard disk. Notebook Word Processor WV-700: Thinnest (30mm) and lightest (1.4kg). High-Brilliance LCD Projectors XV-H1: High luminance projector. Home Fax UX-1: World's thinnest (39mm) space saving fax. Hyper Electronic Organizer PA-9500: Touch screen display. Full-Color Facsimile JX-5000: World's first color fax. Cordless Telephone with Answering Machine CJ-A350: First integrated scrambler. HC LCD Video Projector XH-L100: First LCD projector for HDTV. Stylish Electronic Organizer PA-X1: Lightest organizer (99g) for women. High-Resolution LCD Video Projector XV-S1Z: 650,000 pixel resolution. Japanese Word Processor "Shoin" WD-SD70: Largest screen (17 inch).

June, 1991 July, 1991 August, 1991 September, 1991 November, 1991
January, 1991 February, 1992 April, 1992 June, 1992 June, 1992 July,,1992
August, 1992 October, 1992 January, 1993 February, 1993 March, 1993
May, 1993 June, 1993 July, 1993
August, 1993 September, 1993 November, 1993 January, 1994
Laptop Word Processor with Super-Clear Outline Fonts WD-A550: First cordless input unit. Sensor System Oven RE-SE1: One touch cooking. Wall-mount Color LCD TVs 9E-H Series: First 8.6 inch wall mount. Fully Automatic Bubble Action Washing Machine: First bubble action. Pocket Cordless Telephone/Answering Machine Combos CJ-A30/31 Notebook-Size UNIX Workstation UN-10: Most compact & lightest. Laptop Japanese Word Processor WD-A551: Post card writing. S-VHS VCR with Build-In Satellite Broadcast Tuner High-Brilliance LCD Projector XV-A1Z: New micro-lenses. 16.5 inch Color TFT LCD Panel: First wide-vision panel. Twin Lens 8mm Camcorder VL-MX7: First twin lens camcorder. 36-inch HDTV with "MUSE" Decoder 36C-SE1: 1,125 scanning lines. Personal Wordprocessor WD-A750: Pen operation system. Personal Computer Shoin PC-WD1 Series: Combined WP/PC. Twin Lens 8mm Camcorder VL-HX1: Super-wide angle, superimpose, and picture switching functions. Hyper Electronic Organizer with Superior Management Capabilities PV-F1. Limited Frequency Radio Transceiver CB-T10: Prevent eavesdropping. Wide Vision TV with Built-In Satellite Tuner 28C-WD2: Low priced with 16:9 aspect ratio. Compact and Lightweight LCD Video Projector XV-P1: One-third size and weight. LCD Camcorder with 4-inch Color LCD Monitor VL-HL1: First 4-inch Color LCD monitor. VIP (vacuum insulation panel) Refrigerator: 445-liters at 350-liter size. Headphone MiniDisc Player MD-S10: Smallest and lightest. Headphone MiniDisc Player MD-D10: Uses LCD for title display. Energy -Efficient Solar-Powered Home Heating/Cooling Air Conditioner AYC28FSL: Space saving, solar indoor unit. Color Image Scanner/ Transparency Scanner JX-325: One-third scanning time. Pen-Based Word Processor WD-A770/780: Industry's first swing top pen design. Portable CD Radio Cassette/Players for home karaoke fans. 4" Color LCD TV 4E-C3: uses high brightness, low reflection screen. 10.4" Active-Matrix Color LCD Display for multimedia video images. Three Sharp LCD ViewCam models with expanded functionality. Two 24" Widescreen TVs with Built-In Broadcast Satellite Tuners and VCR: First 24" with Built-In VCR. Portable Mini-Disc Recorder MD-M11: Editing functions with text data recording. Five-Door Refrigerator SJ-45M: Zero ozone depletion.
Sharp's Customer Orientation In 1975, the baby boomers became the target of Sharp's new life strategy. In 1984 and 1985, however, the baby boomers families began taking over and Sharp's products didn't sell well. That required Sharp to develop a new concept and a new strategy. The company set up a research project, which became the creative life planning group, to determine Sharp's future market strategy. Otani explained the results of that research:

As a result of Sharp's creative life-style research, the company developed a new strategy. Otani explained:
We identified different roles that people played. For example, we considered the most sophisticated people in a market to be professionals. The next level of consumer is the sense leader, the next level is the sense follower. At the bottom of the market is the no-sense consumer or the mass market. Matsushita and Sanyo target the mass market. Sony and JVC target the professional. Sharp wanted to target the sense leaders, those that influenced others to buy new products. In order to identify the sense leader, we would call the person who bought a product. We would ask who recommended the product and they might say that their older brother recommended the product. Then they would call the older brother and ask who had recommended the product to
them. He might say that his friend had recommended it, so we would ask for the friend's telephone. The objective was to identify the original source of the recommendation. We found that such people typically read professional magazines and had a reason to recommend a product. We would try to locate about five such people and would invite them to meet with company employees. We discussed our products and talked about their hobbies and establish relationships with "sense leaders." Finding these people was a very time consuming process. We are the professionals on the manufacturing side. Sense leaders are the professionals on the consumer side. If we can establish communications between the two sides, it is more exciting for both of us. That is how we started this process eight or so years ago. For each scene, there are different sense leaders. Back then we had about 50 such people, ten groups of five people looking at five windows. Today we have about 600 such people involved in about 120 groups. The most difficult part is in finding these people, but it is a lot cheaper than hiring Dentsu or one of the market research firms.
As the teams of sense leaders continued to grow at Sharp, they have become increasingly sophisticated. Otani continued:
The group size now depends on the product. We have a group that now manages this type of research. We started with ten people each in Osaka and Tokyo, but have grown the group since then. The ages of sense leaders range from 17 years old to 75 years old. What is important is the description of the sense leader: five aspects of the person. One relates to the person's wealth and their attitude to money. A second relates to their health. The third relates to the family history. The fourth relates to a person's style and looks. The fifth relates to their philosophy and how they view life. There is quality of a person as a whole and quality of a product as a whole. The money aspect relates to pricing of the product. Health relates to the performance of the product. The history relates to the brand image. The individual style has to do with design. Philosophy relates to the products concept. The most important part is the concept. That is what the creative life-style center works on. The pricing, performance and styling are the responsibility of the divisions. Our public relations office is responsible for the corporate image and identity. This responsibility is very clear and there is nothing like this concept in other companies.

Meetings with the sense leaders typically lasted about two hours. It included five people from the company, including the manager and his younger people. After every meeting, a memo was sent to a top officer to initiate action on the recommendations. On occasion, Mr. Tsuji sat in the meetings and listenee from the corner. Three years ago, the center became one of Sharp's headquarters functions with a vice president is in charge of the activity as shown in organization chart in Figure 3. Asada was in charge of the function and had close ties with the president. Mr. Asada was one of the inventors of the calculator and had a Ph.D. in physics. He was also the father of Sharp's LCD operations and was a key person in the company. With his leadership, the organization had an important impact within the company even though its number was relatively small. "That is why we call this the brain of the company," explained Otani. With the globalization of Sharps business activities, the needs of foreign customers had to also be met. Wada explained:
We are watching overseas markets closely. We have been successful in expanding our overseas position, especially in Asian countries other than China and India. But now I believe we are entering a new phase of global competition. Our previous focus covered a market of one billion people: Europe, the U.S. and Japan. But today, there are large new markets opened up like China with 1.2 billion people and India with 850 million people. Russia has maybe 250 million. The free market has expanded from one billion to 2.5 billion people. This growth in mostly in developing countries. Countries in Latin American and Pakistan are sending a lot of missions here with the view of inviting us to set up factories. So this is a totally different world. What they want is different from that of the U.S., Europe and Japan. These emerging markets, like China, Southeast Asia and India, are growing between six and ten percent per year. So the rules are changing and we must meet their requirements. You have to come up with the right product for them. The type of products that sell in Japan will not sell in Thailand. The real challenge is in figuring out how to meet those local needs. We want to maintain Sharps brand image and technical competence in those market, so we dont aim to introduce products at the lowest end of the market. At the same time, we have to produce products that are not as expensive as we can sell in Japan. We will set up a factory in India to produce a video cassette recorder. We found that they appreciate Sharps technology, but want products that are different from what we now produce. We have to produce a product that is suitable for people with lower incomes and different tastes than Americans. We have to use local design engineers that can make good use of components and materials sourced from local factories in order to cut costs. In fact, our factory in Malaysia uses non-Japanese components valued at 85 percent of the cost of ur locally produced VCR. That means that we outsource the majority of our components locally. That is an example of our strategy for using local suppliers to cut costs. We established the creative life group to study new trends, especially among young people who are the trend setters. We have also set up product design groups in our different market areas. They often propose products for development. We have a group in the Philippines that have been very good at developing products for that market. They like it loud in the Philippines. Our basic audio technology comes from Hiroshima where we have our communication center for the business group. We are putting emphasis on high volume and high power sound for that market.

Sharp TV & Video Systems Group Sharp's TV & video systems group includes the TV systems division, the LCD visual systems division, the video systems division, and the TV & video systems laboratory. The television business division's products are centered around color TVs and video-based media and services such as satellite and multiplexed videotext, CATV, and home computers.
From this foundation, we constantly strive for improved performance in all aspects of technologies and systems, from design and development through production and shipping. We have established rigorous inspection procedures driven by the newest and most powerful production systems and computers as well as implemented flexible manufacturing structures applicable to products from large-screen TVs to personal TVs. For the future, we will, of course, work for improved image quality and resolution, and will be improving products with even greater reliability.
Figure 3: Sharp Corporation's Organization Chart

As of October 1, 1994

Advertising Department Management Planning Board Corporate Public Relations Division Advertising Department Construction Planning Personnel Development Center Personnel Affairs Group Corporate Finance Group Corporate Accounting and Control Group Law Group Corporate Design Group Creative Life-Style Planning Group Corporate Procurement Group Reliability Control Group Production Technology Development Group Tokyo Branch Corporate Research and Development Group International Business Group Domestic Sales and Marketing Group -- Consumer Electronics Domestic Sales and Marketing Group -- Information and Communication Equipment Domestic Sales and Marketing Group -- Special Outlets International Sales and Marketing Group -- IC/Electronic Components Domestic Sales and Marketing Group -- Electronic Components TV and Video Systems Group Communications & Audio Systems Group
Legal Division Intellectual Property Center Information Systems Development Center Production Technology Laboratories Production Technology Development Center CAE Center Central Research Laboratories Functional Devices Laroratories Energy Conversion Laboratories Mechatronics Systems Laboratories Information Technology Laboratories Information Systems R&D Center Integrated Media Laboratories Image Systems Laboratories Telecommunication R&D Laboratories

President

Multimedia Systems R&D Center
Senior Executive Vice Presidents

Appliance Systems Group

Information Systems Group

Also important is a built-in speaker; thus making a private earpiece optional. You can even take the camera along on trips and use it to watch prerecorded movies on 8mm videocassettes. Conventional camcorders were mostly sold to families with small children who made videos of children growing up, or of family events, or of vacations. The ViewCam's ease of use was expected to revitalize the market. According to Kuwata:
The market size was an uncertain element when we started. In 1990, it was 1.8 million units and fell to 1.44 million units in 1991. Then it fell to 1.16 million units in 1992. The ViewCam came to the market in September, 1992. It has expanded the market to 1.25 million units in 1993. We expect the market to expand to 2.0 million units by 1995. We hope to see the market grow to 3.0 million units per year.
When the video systems division was reviewing different ideas for the new camcorder, president Tsuji sent down the word to "use liquid crystal display in a camcorder." This was a difficult challenge since the LCD screen was difficult to see in bright light or from an angle. It also had to be thin and light for ease of carrying. This required a totally new type of low reflection LCD and a new system to allow the LCD screen to flip 140 degree and
maintain an upright picture. The responsibility for the development was placed in the hands of senior executive vice president Asada. Kuwata explained:
To achieve our goal, we had to conduct basic research. So the development was far more difficult than that for a conventional product. Theoretically, we felt it possible to reach our goal in six months, so we formed a kinkyu project team to achieve commercialization in October 1991. The major difficulty at that time was thought to be the technology for reducing reflection of external light. This problem was solved by applying a special, anti-reflective coat to the surface of the liquid crystal panel itself. This lowered the previous reflection rate of 23.3 percent to an astonishing 2.6 percent, about one-tenth of what it was. This solved the problem of outdoor visibility. Next, it was necessary to make the screen itself brighter and easier to see, and to make the whole unit lighter and more compact. Using the new type of backlight system, it was possible to attain twice the brightness of conventional panels. A reduction of panel thickness of one-third that of conventional products was also achieved.
Improvements in low-reflection, high luminosity, compactness and lightness lead to the new LCD panel. While developed for the LCD Hi-8 ViewCam, it has now been applied to the high performance liquid crystal portable television. The LCD panel improvement, however, turned out to be the easier problem to solve. The most difficult problem turned out to be the design of the mechanism and software that allowed the LCD unit to flip over 140 degrees:

Through 1993, Sharp had sold 470,000 ViewCams, including 250,000 in the domestic market and 220,000 through export markets. The sales of ViewCam sales were: Domestic 120,000 250,000 Overseas 15,000 220,000

1992 1993

Competitive Responses The major competitors in the video camera market were Sony, JVC and Matsushita. Sony was currently number one, Sharp was number two, and Matsushita was number three. JVC admitted that they lost share in 1993. Kuwata was optimistic about Sharp's product leadership:
We surpassed Matsushita this year to move into second place although they have not yet admitted that they lost share. We are approaching a 25% market share and project 40 percent by the first half of 1994, depending on what our competition does.
The market was currently divided into view finder and LCD type video cameras. If you look at what is happening to market position, consumers prefer the LCD screen to the viewfinder. Sharp owned 100 percent of the LCD type segment through 1993. As a result, Sharp had moved into third place in 1993 with 15.2 percent of the market. Kuwata explained the shift:
If you set 1992 sales at 100 percent for view finders and LCD, the viewfinder segment has fallen to 70 percent while the LCD segment increased to 200 in 1993, or nearly 25% of the market. The viewfinder segment is going to be further reduced in 1994. The people that have taken pictures before are now playing with the new functions. We want to add functions to this to improve communications. If we can add new functions, we can increase the size of the market again.
Sharp continued to increase capacity of its Tochigi plant to 800,000 ViewCams in 1994 and planned one million in 1995. According to Daiwa Institute of Research's Yoshimasa Takashina, LCD-equipped camcorders will likely outpace sales of conventional camcorders. Sales of camcorder units with LCD monitors in Japan were estimated to grow from 310,000 in 1993 to 600,000 in 1994. In 1993, Sony held a 41.2 percent share of the domestic camcorder market. Until the ViewCam's introduction in 1992, Sony Corporation had held over 75 percent market share of the small format 8mm video camera market. Sony had gained its strong position with the development and application of the charge coupled device (CCD) to 8mm format cameras in 1983. Sony Corporation and Fuji Photo Film Co. Ltd. responded to Sharp's ViewCam with the introduction of their own LCD-based camcorders in February 1994. Sony's smaller, light weight (under two pounds) Handycam Snap was priced at $1,200. The Sony Snap was half as wide and a pound lighter than Sharp's latest VL-EL300 ViewCam. The smaller size of the Snap allowed for one-handed operation, unlike the larger ViewCam which is awkward to use without both hands free. Sony supplies a sun shade for the screen if you're viewing the picture in bright light, though the view finder could be used without the screen. Sharp offered a sun screen as an option. While the Sony was easier to carry, the ViewCam offered more advanced filming capabilities. Sony offered only two modes, 118mm telephoto and 38mm wide-angle, which makes a Snap noise during mode changes which is recorded. The Snap, which looks like an old boxy Kodak Brownie, used a three inch (diagonal) color LCD screen with the lower resolution Video 8 format cassettes. Matsushita Electric Industrial Co. followed in July, 1994 using a VHS video format. Japan Victor Company (JVC), of which 52.4 percent is owned by Matsushita, was the last to respond with a VHS video recording system. The lower resolution of the Video 8 is slightly better than VHS formats used by Matsushita and JVC. Sharp and Fuji Photo Film used the Hi-8 system, an 8mm tape developed by Sony. This requires an external microphone which attaches to a weak rubber connection. The Snap also offered only an A/V Out connection to attach to an external monitor. The ViewCam offered standard RCA jacks for both A/V In and Out connections. Sony's Snap provided an optical viewfinder. This allowed the LCD screen to be turned off during shooting to extend battery power. The Snap's rechargeable lithium ion battery lasted about 45 minutes with the screen on and 75 minutes with it off. The ViewCam's slightly larger nicad battery life lasts about 45 minutes. The Snap also allowed AC power to be connected with the battery being charged at the same time. The ViewCam required removal of the battery for changing or AC operation. Since the 8mm format allows three hours of programming per cassette, extra batteries are essential. Matsushita was Japan's second largest camcorder maker, using the VHS-C system. Both Matsushita and JVC offered lower priced viewfinder camcorder models which allowed their 8mm format tapes to be played in standard VHS video players. According to Kenichi Kitayama, director of Matsushita's VCR and camcorder business, "If you shoot with a VHS camcorder, you can easily play the cassette back with a VCR at home. We will appeal to the

We have to think about the 21st century from here on out. What is important about young people under 20 is their philosophy. People without a concept are running towards new religions, or following people that are the current fad. They just go enmass in one direction. It is important to make products to meet the needs of these people. Finding a concept that fits these people is very complicated. Until now, what was needed in the planning group was technology and perhaps economics and literature majors. From here on, we probably need philosophy majors and psychologists, a different kind of person to think about new products. Engineers will always be needed since any concept must be backed up with technology, but additional people are needed. Software people are going to be needed for future planning.
Nikkei Sangyo's newspaper article noted that Sharp's current weakness was in communication technology. Sharp's telephone experience was not considered adequate and Sharp required new technological alliances. In 1992, the company entered the small power transmission market but depended on Nippon Musen for software for cordless systems, low powered and wireless technologies. The company did not have the know-how to cope with the processing of complex code that enables the reduction or prevention of noise in communication. As a result, Sharp was developing strategic alliances with a number of major companies. Wada explained:
We have a top down approach that is focused on building the infrastructure for product development and in making large technology investments, like image compression technology in multimedia. In the bottom up approach, we have a lot of products that can be used to approach the multimedia market, like the video cam. We can use it to transfer still pictures over the telephone up to six of seven times per minute, now. With the cooperation of AT&T, within a year we hope to increase this capability to 15 pictures per second. That is getting close to television reception. We are now developing the basic technology that will allow us to develop such capabilities. We are jointly developing these capabilities with AT&T. I expect that we may find some other areas for joint development. Flash memory has potential. We are now expanding our production and developing the second generation flash memory with Intel. This is strategic cooperation with the world famous microprocessor company. They are the best imaginable partner for us. We are always extending feelers for potential cooperation. Ill give you the news release. Wada was also applying the new technologies to its existing products: We are now trying to combine the paging function with the Zaurus, so that you can send four or five lines on the pager screen so you dont have to go to the telephone to get the message. We have also included fax capability in the Zaurus, and infrared communications. It can transfer information between computers and send Fax over the telephone lines. It is handy and portable, but is already in the market. It is sold in the U.S. as an electronic organizer called the Wizard. We dont really have a competitor in the market with the same type of product. These product has created a lot of good will in the market. We also have miniaturization technology that allows us to develop very light-weight and portable products. Our PHS, personal handy phone, is an example. We are one of many suppliers of this product OEM to NTT. Sharps is the preferred PHS. In the PHS market, there are 43 models being produced by 29 vendors. Sharps two models account for between 60 and 70 percent of the market. Our models weigh about half as much as other models at 200 grams. Because it has NTTs brand name, consumers dont realize that they prefer Sharps models. The smallest cellular phone is 95 grams. Its so small that women lose it in their purses.

Table 1: Market Share for Major Electronics Products and Components (reported in 1994) Consumer Electronics and Home Appliances (domestic units or sales, change in share in 1993) Color TVs VCRs Camcorders 1. Matsushita 21.0% 1. Matsushita 23.0% 1. Sony 41.2% 2. Sharp 15.5% 2. Sony 14.0% 2. Matsushita 25.0% 3. Toshiba 14.5% 3. JVC 14.0% 3. Sharp 15.2% 4. JVC 7.0% 4. Sony 14.0% 4. Sharp 12.0% 5. Fuji Film 4.0% 5. Hitachi 10.5% 5. Toshiba 11.0% (1,169,000 units shipped) 3.4% (8,143,000 units shipped) 1.9% (4,486,000 units shipped) 2.4% Home Phones Refrigerators Air Conditioners 1. Sharp 22.0% 1. Matsushita 20.0% 1. Matsushita 17.0% 2. Sanyo 21.5% 2. Toshiba 16.0% 2. Mitsubishi 13.5% 3. NTT 16.0% 3. Hitachi 16.0% 3. Toshiba 13.0% 4. Matsushita 11.7% 4. Sanyo 15.0% 4. Hitachi 11.0% 5. Sony 8.0% 5. Sharp 13.0% 5. Sanyo 10.0% (1,410,000,000 yen sales) 4.8% (4,127,000 units shipped) 3.6% (5,048,000 units shipped) 19.3% Information & Office Systems and LCDs Facsimiles Plain Paper Copiers 1. Ricoh 16.5% 1. Ricoh 31.5% 2. Matsushita 16.0% 2. Canon 27.8% 3. Canon 15.0% 3. Fuji Xerox 23.3% 4. Sharp 6.9% 4. NEC 10.5% 5. Konica 4.8% 5. Toshiba 8.0% (620,000 units shipped) 5.0% (1,900 million yen sales) 3.6% Office Computers Personal Computers 1. NEC 28.5% 1. NEC 52.7% 2. Apple Computer 12.2% 2. Fujitsu 27.0% 3. Toshiba 9.7% 3. Fujitsu 11.3% 4. IBM Japan 9.4% 4. IBM Japan 8.1% 5. Mitsubishi 8.4% 5. Seiko-Epson 6.7% (2,460,000 units shipped) 10.0% (163,700 units shipped) 14.0% Source: Nikkei Sangyo Shimbun, June 22, 1994.
LCD TVs 1. Casio 51.8% 2. Sharp 18.2% 3. Epson 13.6% 4. CVM 7.3% 5. Seiko 5.5% (1,100,000 units shipped) 22.2% Microwave Ovens 1. Matsushita 26.3% 2. Sharp 20.9% 3. Mitsubishi 11.5% 4. Toshiba 10.9% 5. Hitachi 10.1% (2,730,000 units shipped) 0.7%
Electronic Calculators 1. Casio 52.7% 2. Sharp 36.0% 3. Sanyo 1.9% 4. Canon 1.8% 5. Toshiba 1.8% (77,800 million yen sales) 15.7% Projection TVs 1. Sharp 54.0% 2. Pioneer 18.0% 3. Mitsubishi 12.0% 4. Hitachi 10.0% 5. Other 6.0% (50,000 units shipped) 9.1%
Japanese Wordprocessors 1. Sharp 20.0% 2. NEC 14.5% 3. Toshiba 13.8% 4. Fujitsu 13.5% 5. Casio 13.0% (2,220,000 units shipped) 4.3% LCDs 1. Sharp 39.1% 2. Toshiba 11.5% 3. Seiko Epson 9.8% 4. NEC 8.8% 5. Optrex 7.1% (480 billion yen sales) 8.3%