r / f •' f. /. -- > • /
?1arch 1985
DEVELOPMENT OF AN ADJUSTABLE BUOYANCY
BALLON TRACER OF ATMOSPHERIC MOTION
Phase I. Systems Design and Demonstration of Feasibility
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH Aim DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DEVELOPMENT OF AN ADJUSTABLE BUOYANCY
BALLON TRACER OF ATMOSPHERIC MOTION
Phase I. Systems Design and Demonstration of Feasibility
by
B. D. Zak, H. W. Church, A. L. Jensen,
G. T. Gay, and M. D. Ivey
Sandia National Laboraties
Alburquerque, New Mexico 87185
Interagency Agreement DW930214
Project Officers
J. S. Irwin and R. G. Lamb
Meteorology and Assessment Division
Atmospheric Sciences Research Laboratory
Research Triangle Park, NC
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
NOTICE
The information in this document has been funded by the United States
Environmental Protection Agency under Interagency Agreement DW930214 to
Sandia National Laboratories. It has been subject to the Agency's peer
and administrative review, and it has been approved for publiction as an
EPA document. Mention of trade names or commerical products does not
constitue endorsement or recommendaton for use.
-------
ABSTRACT
An Adjustable Buoyancy Balloon Tracer of Atmospheric Motion
is a research tool which allows one to follow atmospheric flows
in both the horizontal and the vertical, including the weak.
sustained vertical motion associated with meso- and synoptic-
scale atmospheric disturbances. The design goals for the
Balloon Tracer to be developed here specify a lifetime >^ 3 days.
tracking range >. 1000 km. a ceiling altitude >_ 500 mb (5.5 km).
and the capability to respond to mean vertical flows as low as
1 cm/s. The balloon tracer is also to measure and telemeter
selected meteorological variables, to be sufficiently
inexpensive to permit use in significant numbers, and to be
serviced by a ground system capable of handling several balloon
tracers at a time. While the balloon tracer has applications
throughout the atmospheric sciences, the immediate motivation
for this effort is to meet the need to evaluate the accuracies
of existing air pollution transport models, to establish source-
receptor relationships to distances of order 1000 km. and to
assess the inherent limits on the predictability of source
impact at long distances. The authors have proposed a generic
design for such a system. They also have subjected the proposed
design to theoretical analysis, have constructed a prototype.
and have conducted a series of tests with the prototype to
evaluate the concept. They conclude without reservation that a
system meeting the design goals is feasible, and are proceeding
to build that system in Phase II of this project.
111
-------
CONTENTS
Abstract iii
Figures vii
Acronyms x
Acknowledgements xi
1. Summary 1
2. Introduction. . 11
3. Concept and Theoretical Analysis 14
a. Buoyancy Control 14
b. Control Strategies 21
4. Systems Design 27
a. Tracking and Data Handling 28
b. Balloon Envelope 32
c. Payload 33
d. Ground Support Station 44
e. Control Algorithm 47
5. Demonstration of Technical Feasibility:
The Testbed Prototype 48
a. Approach 48
b. Balloon Envelope 49
c. Payload 50
d. Vertical Anemometer 60
e. Prototype Ground System 63
6. Experimental Results 66
a. Laboratory 66
b. Tower 73
c. Ambient Atmosphere 90
7. Future Work 93
8. Conclusions and Discussion 97
References 102
-------
Appendices:
A. Federal Aviation Regulations. Part 101 .... 105
B. Pumpdown Speed in a Standard Atmosphere. . . . no
C. Superpressure as a Function of Equilibrium
Altitude 112
D. Effect of Temperature Swing on Balloon
Pressure 115
E. Energy Required for Pumpdown 116
F. Lift and Maximum Altitude 118
G. Trajectories: Isentropic and Actual 121
H. The ARGOS Satellite-Based Data Collection and
Platform Location System 125
I. Pumpdown Speed with an Arbitrary
Lapse Rate 140
J. Diffusive Spread of a Marked Air Parcel and
Its Implications for Lagrangian Experiments. . 142
VI
-------
FIGURES
Number Page
1-1 Prototype Adjustable Buoyancy Balloon Tracer being
readied for flight 3
3-1 Schematic Diagram of the Adjustable Buoyancy Balloon 16
3-2 Sulfur Hexafluoride Concentration versus time on
board the LAMP (Lagrangian Measurement Platform)
during the Tennessee Plume Study 25
3-3 Trajectory of LAMP during the TPS 26
4-1 Number of ground stations required for 1000 km
square area, and VHF radio range as a function of
balloon altitude 29
4-2 Buoyancy control plumbing 34
4-3 NCAR microprocessor data system for
balloon application 38
4-4 ARGOS PTT (Platform Transmitter Terminal) for
balloon application 40
4-5 NCAR ARGOS Antenna 42
4-6 Paper lithium battery 43
4-7 Diagram of ground support station for an
operational tracer system 46
5-1 Block diagram of testbed prototype payload 52
5-2 Airsonde and tethersonde circuit boards side by side 55
5-3 Preferred sensor placement relative to the balloon . 56
5-4 Gilian pump. Klippard valve, and Japan Remote
Control RC receiver and servo 58
5-5 Assembled testbed prototype payload 59
5-6 High sensitivity vertical anemometer for
relative vertical air motion measurement 61
5-7 Vertical air motion observed with MacCready
anemometer on DaVinci II balloon flight 62
vii
-------
5-8 Block diagram of prototype ground station 64
5-9 Prototype ground station 65
6-1 Combined speed of two Gilian pumps at 6 VDC
versus backpressure 69
6-2 Combined current consumption of two Gilian
pumps at 6 VDC versus backpressure 71
6-3 Vent valve flow as a function of superpressure ... 72
6-4 Tower of Solar Central Receiver Test Facility.
the interior of which was used to test the
testbed prototype PLT 74
6-5 Interior of the Solar Tower with reference
levels indicated 75
6-6 Dimensions and reference levels of testbed
prototype PLT for Solar Tower tests 76
6-7 Valve-up number one 78
6-8 Pumpdown number one 79
6-9 Temperature vs altitude on pumpdown number one ... 80
6-10 Internal temperature, ambient temperature, and
superpressure on pumpdown number one 81
6-11 Valve-up number two 83
6-12 Pumpdown number two 85
6-13 Temperature vs. altitude on pumpdown
number two 86
6-14 Testbed prototype PLT during an "open air
tower" test 91
6-15 Testbed prototype PLT during a slack-tether
flight test 92
7-1 Lagrangian experiment instrumented aircraft
measurement pattern . 96
G-l Comparison of twelve-hour isobaric and
isentropic trajectories . 124
Vlll
-------
J-l Dimensionless diffusion parameter D versus
dimensionless parameter T proportional to time . . . 145
J-2 e1/3 versus time for the DaVinci II flight 147
J-3 "e-meter" used to make turbulence
measurements on DaVinci II 148
IX
-------
ACRONYMS
ADAS
AGL
AIR
ONES
CVB
DME
DOE
EPA
FAA
GHOST
HF
LAMP
LORAN
LUT
MSL
NASA
NCAR
NOAA
PGS
PLT
PTT
RC
RV
SNL
TP
TPS
UHF
VOR
VHF
WMO
Atmospheric Data Acquisition System
Above Ground Level
Atmospheric Instrumentation Research (Inc.)
Centre National d'Etudes Spatiales (France)
Constant Volume Balloon
Distance-Measuring Equipment
U.S. Department of Energy
U.S. Environmental Protection Agency
Federal Aviation Administration
Global Horizontal Sounding Technique
High Frequency
Lagrangian Measurement Platform
Long Range (radio) Navigation
Local User Terminal (ARGOS)
(Above) Mean Sea Level
National Aeronautics and Space Administration
National Center for Atmospheric Research
National Oceanic and Atmospheric Administration
Prototype Ground Statidn
Physical Lagrangian Tracer
Platform Transmitter Terminal (ARGOS)
Radio Control
Recreational Vehicle
Sandia National Laboratories
Testbed Prototype
Tennessee Plume Study
Ultra High Frequency
VHF Omni-directional Range
Very High Frequency
World Meteorological Organization
-------
ACKNOWLEDGMENTS
The authors wish to thank R. G. Lamb and J. S. Irwin of the
Atmospheric Sciences Research Laboratory, U.S. Environmental
Protection Agency, whose interest in broadening and deepening
understanding of atmospheric processes important to air
pollution sparked this project. We gratefully acknowledge the
substantial contributions of V. E. Lally. E. Lichfield, S.
Stenlund, and N. Carlson of the National Center for Atmospheric
Research for generously sharing with us their wealth of
experience with superpressure balloons and supporting systems.
R. Enderson and T. Markhart of Raven Industries, D. Call of
Atmospheric Instrumentation Research, Inc.. and A. L. Morris
and D. Street of Ambient Analysis, Inc.. all went far beyond
the call of duty in assisting us with this effort. We also
wish to thank J. Otts of Sandia National Laboratories for
making the solar tower available for system testing, S. Sawyer.
also of Sandia. for her dedication and skill in preparing this
report, and P. S. Homann. for assuring that it all came
together properly. The following individuals graciously
consented to review and comment on this document: J. S. Irwin
and R. G. Lamb of the U.S. Environmental Protection Agency. E.
F. Danielson of NASA Ames Research Center, V. E. Lally and E.
Lichfield of the National Center for Atmospheric Research, J.
K. Angell and C. R. Dickson of the NOAA Air Resources
Laboratory, P. B. MacCready of AeroVironment, Abdul Alkezweeny
of Pacific Northwest Laboratory, and D. S. Ballantine of the
DOE Office of Health and Environmental Research. We are
particularly indebted to A. L. Morris of Ambient Analysis, who
not only reviewed this report, but also checked the equations
describing balloon behavior. This research was funded by the
U.S. Environmental Protection Agency in part through the
National Acid Precipitation Assessment Program.
XI
-------
-------
1. SUMMARY
An Adjustable Buoyancy Balloon tracer of atmospheric
motion is a physical Lagrangian tracer, an airborne
instrumentation system that follows the flow of air, and that
can be tracked electronically (Figure 1-1). Such a system has
been desired for decades by researchers in the atmospheric
sciences to aid in understanding the dynamics of the
atmosphere, and to cast light on long range air pollution.
The present effort, however, is motivated primarily by the
more immediate need to establish source-receptor relationships
to distances of order 1000 km. to evaluate the accuracies of
existing air pollution transport models, and to assess the
inherent limits on the predictability of source impacts at
long distances.
The Adjustable Buoyancy Balloon system must operate under
Federal Aviation Regulations Part 101, which covers unmanned
free balloons. FAR 101 divides such balloons into two
classes. Those which offer little hazard to aircraft because
of their limited size, weight, and density are explicitly
exempted from most of the other stipulations of the
regulation. So-called "weather balloons" (radiosondes) fall
in this category. Hundreds of such balloons are launched
twice a day from sites all over the US and around the world to
provide data on meteorological conditions aloft. Balloons not
meeting the conditions contained in the exemption clauses of
FAR 101 are subject to strict regulation, and are treated much
like other aircraft. It is highly desirable for the balloon
tracer to operate under the exemption clauses, in that certain
other provisions of FAR 101 would severely limit the
usefulness of a balloon tracer which was not exempt. Even
though the Adjustable Buoyancy Balloon system will be exempt.
it will nevertheless carry a radar reflector and an FAA
-------
transponder so the FAA can keep track of its location.
In addition to meeting the exemption conditions of FAR
101. the design goals for the balloon tracer are:
• Lifetime :> 3 days
• Tracking range > 1000 km
• Telemetry of selected meteorological parameters
• Ground system capable of handling several PLTs at a time
• Ceiling altitude >. 500 mb (5.5 km)
• Follows mean vertical flows as low as 1 cm/s
• Sufficiently inexpensive to permit use in significant
numbers.
The project is divided into two phases:
Phase I: Systems Design and Demonstration of Feasibility.
Phase II: Development of an Operational Prototype.
This report covers work on Phase I. Phase II is now
proceeding.
The design of the Adjustable Buoyancy Balloon Tracer is
based upon an idea put forward by V. Lally of the National
Center for Atmospheric Research in 1967. Here the outer skin
of a spherical balloon is made of a high modulus of elasticity
material which expands very little as pressure in the balloon
increases. Hence, the volume of the balloon is very nearly
constant as long as the pressure of the gas inside is greater
than the ambient pressure. A thin polyethylene bag. or
"ballonet". separates the interior into two compartments. One
of these compartments is filled with helium, the lift gas.
The other is filled with air. The air serves as ballast. A
pump and valve permit additional air to be taken into the
balloon, or to be released. When the balloon is at its
-------
Figure 1-1. Prototype Adjustable Buoyancy Balloon Tracer
being readied for flight. Payload weighs 2.36 kg (5.2 Ibs)
including batteries, and is constructed of styrofoam covered
with 0.6 oz fiberglass. It meets the exemption clauses of
FAR 101.
-------
equilibrium altitude, and more air is pumped in. it becomes
heavier, and sinks to a lower altitude. When air is released
through the valve, the balloon becomes lighter, and rises to a
higher altitude.
Expressions have been derived in the appendices which
describe the rate at which pumping and valving change the
equilibrium altitude, the behavior of the excess of internal
over ambient pressure (superpressure) as a function of
equilibrium altitude, the effect of temperature changes on
balloon pressure, the energy required for pumpdown. how the
ceiling altitude is determined by system parameters, and a
procedure for properly filling the balloon to obtain the
desired characteristics. All of these calculations confirm
that a properly-designed balloon system of the type proposed
by Lally can meet the design goals.
Given this means of adjusting the buoyancy of a constant
volume balloon, the balloon will become a tracer for
atmospheric motion if the buoyancy is periodically adjusted so
that the balloon follows the vertical motion of the air. The
nature of balloons is such that they naturally follow
horizontal air motions. Hence, if a system is constructed to
also follow the vertical motions, that system will follow the
overall flow.
There are two basic approaches to the altitude control
problem. The first is to continuously measure the vertical
velocity of the air relative to the balloon, and to adjust the
buoyancy so that, on average, the relative velocity of the air
is zero -- that is, so that on average, the balloon and the
air move together. The second approach is to take advantage
of the very nearly adiabatic nature of atmospheric flows.
When flows are adiabatic. the potential temperature is
-------
constant along each air parcel trajectory. In this approach.
the buoyancy of the balloon is adjusted so that the potential
temperature is kept constant. As long as this condition is
met. the balloon will move along with the air surrounding it.
The approach to altitude control based on relative
vertical air motion is most direct, but if it were to be used
continuously for three days, the air motion measurements would
have to be extraordinarily accurate. Under most atmospheric
conditions, the approach based on potential temperature is
quite satisfactory, but under certain conditions — when the
system is in a layer of air in which active convective mixing
is talcing place -- potential temperature does not offer an
adequate guide for altitude control. Under convective mixing
conditions, the air surrounding the balloon consists of
turbulent air flows moving both up and down. The mixing makes
the potential temperature uniform with altitude within the
mixed layer.
When convective mixing engulfs a "parcel" of air. the main
effect is to disperse it. and to spread it out in the
vertical, mixing it with air from all the surrounding
parcels. If a balloon tracer is embedded in an air parcel
which is subjected to convective mixing, as long as the
balloon remains in the mixed layer, it is within the confines
of the now greatly-expanded "parcel". Following the expanding
parcel during convective mixing makes less stringent demands
on the buoyancy control system than does following a parcel in
the absence of convective activity. Hence, a number of
different control strategies are satisfactory during these
periods.
Thus it appears that a hybrid control approach will yield
best results. Different control strategies will be employed
-------
under different meteorological conditions. On the basis of
the data from the onboard sensors, an onboard microcomputer
will determine which control strategy will be implemented at
any given time.
Having confirmed on paper that the Adjustable Buoyancy
Balloon Tracer is viable in principle, we proceeded to
formulate a preliminary design for an operational system. The
major elements of the design are the tracking and data
handling system, the balloon envelope itself, the balloon
payload, and the ground support station.
Tracking and data reception will be handled by the ARGOS
satellite-based data collection and platform location system.
ARGOS is a joint undertaking of NASA, NOAA, and Centre
National d'Etudes Spatiales (CNES, France). It makes use of
NOAA satellites, and both US and French ground support
facilities, to service fixed and moving platforms collecting
environmental data. ARGOS has the advantages that it is
well-proven, has a high data recovery rate, that lightweight
hardware designed for use on balloons is commercially
available, and that it provides world-wide coverage.
The balloon envelope design was undertaken in consultation
with NCAR and in collaboration with Raven Industries, the
maker of NCAR's high altitude constant volume balloons. The
initial design adopted is for a spherical balloon of 2.9 m
diameter. 12.5 m volume, made of 3 mil bilaminated
polyester (mylar) film with a 1 mil polyethylene ballonet
inside. The balloon was designed to carry up to a 4.5 kg (10
Ib) payload to 600 mb. and to have an operational
superpressure limit in excess of 80 mb. With a lighter
payload. the ceiling altitude will exceed the design goal of
500 mb (5.5 km; 18.000 ft).
-------
The payload consists of a buoyancy adjustment subsystem.
sensors, a microprocessor or microcomputer, a telemetry
subsystem, a radio command subsystem, tracking aids, and
batteries. The buoyancy adjustment subsystem is just the
pumps, valves and associated plumbing mentioned earlier. The
initial list of sensors numbers 12. and includes those
necessary to follow either a zero relative vertical velocity.
or a constant potential temperature altitude control
strategy. The microprocessor or microcomputer processes all
data, formats it for ARGOS transmission, and uses it in a
control algorithm to determine what altitude control measures
should be taken to follow the mean vertical air flow -- no
action, vent air, or pump more air in. The telemetry
subsystem is, in ARGOS parlance, a platform transmitter
terminal (PTT). The radio command subsystem is an HF radio
receiver and command decoder enabling the user to override the
onboard computer control. The tracking aids consist of an FAA
transponder, a radar reflector, and a strobe to aid in visual
tracking. The batteries are state of the art flexible
("paper") lithium batteries with high power-to-weight ratio.
The Ground Support Station (GSS) consists of an ARGOS
local user terminal (LUT). an ARGOS uplink receiver, a radio-
theodolite or LORAN tracking system, a command transmitter,
and a desktop computer with associated peripherals. The LUT
allows one to receive data from the balloon tracer in real
time via re-transmission from the satellite whenever the
satellite is within range of the tracer and within range of
the LUT. The ARGOS uplink receiver allows one to listen
directly to the data stream being transmitted by the balloon
tracer when it is within radio range of the ground station.
The radiotheodolite or LORAN system provides for local
tracking of the tracer when it is being used within radio
range. The command transmitter is a multi-band transmitter
-------
capable of transmitting commands to the balloon tracer over
long distances. The computer receives, formats, archives, and
displays the location and meteorological data as desired. It
also includes a modem to provide for data reception by phone.
The GSS accommodates three modes of use: (a) satellite/
worldwide; (b) hybrid/regional; and (c) ground-based/local.
In the satellite/worldwide mode, location and meteorological
data are received from the tracer via the satellite either
through the LUT or through NOAA facilities accessed by phone.
The LUT gives one the data in real time, whereas the data are
available from NOAA approximately 6 hours later. Data are
only received by the satellite when it is within radio range
of the balloon tracer — about 10 minutes every 2-4 hours.
Depending upon the design of the balloon tracer payload, the
data transmitted may be only the current values of the
measurements being made by the sensors, or it may be all the
values recorded over the previous several hours.
In the hybrid/regional mode, the uplink receiver and other
remotely-located uplink receivers are strategically located so
that the balloon tracer is within radio range of at least one
everywhere within the region of interest. Consequently,
continuous real-time data reception and archiving are
available over the region covered by the uplink receiver
network. Tracking is still accomplished by satellite.
In the ground-based/local mode, the satellite link is not
used at all. The balloon tracer is locally tracked, and the
data acquired directly by the uplink receiver. This mode of
use is limited by radio range.
Demonstration of technical feasibility was accomplished by
fabrication and evaluation of a "testbed prototype" (TP)
-------
balloon tracer. The TP is sufficiently similar to the flight
system proposed in the preliminary systems design so as to
establish feasibility, but does not meet all of the design
goals itself. The major differences between the TP and an
operational balloon tracer are that the TP is designed for
local use only, and that it incorporates elements which make
changing the control algorithm easy.
The TP consists of a balloon much like those for the
operational tracer system, and a payload consisting of a
buoyancy adjustment subsystem, an AIR (Atmospheric
Instrumentation Research. Inc.) airsonde circuit board located
inside the balloon, an AIR tethersonde circuit -board located
externally, a radio control command receiver, batteries and a
strobe.
The TP is flown under the control of a prototype ground
station which consists of an AIR ADAS (Atmospheric Data
Acquisition System) unit, an HP85 desktop computer, an HP3421
data acquisition and control unit, and a radio command
transmitter. The ADAS receives the data from the airsonde and
tethersonde which give data on conditions inside the balloon
and in the ambient atmosphere, respectively. The HP85
processes, archives, and analyzes the data. The control
algorithm is resident in the HP85. Altitude control actions
are transmitted back to the TP via the HP3421 and the command
transmitter. This arrangement allows the control program to
be written in a high-level language, and to be altered on the
ground with a few keystrokes, even when the TP is in flight.
Almost exclusive use of minimally-modified commercially-
available elements in the TP design made demonstration of
feasibility possible within project time and resource
constraints.
-------
The Phase I experimental program was limited to the
minimum necessary to demonstrate that the concept of the
Adjustable Buoyancy Balloon Tracer is viable. Initially.
measurements were made in the laboratory on individual
components to determine if their performance was
satisfactory. Next, the testbed prototype underwent tests in
an enclosed tower. Finally, testing began in the ambient
atmosphere.
The most telling results were obtained in the tower. The
tower is part of the Solar Central Receiver Test Facility at
Sandia National Laboratories. It offers an enclosed volume
roughly 10 m square by 52 m high. Since it is enclosed, it
provides a more controlled environment than does the ambient
atmosphere, making it easier to interpret test results.
Measurements were made on several pumpdown and valve-up
cycles in the tower. The results made clear that the theory
developed does indeed describe the behavior of the balloon
tracer. They also made clear that the tracer's behavior is
more complex than is obvious from the expressions derived
under the assumption of dynamic equilibrium. The equilibrium
theory may be thought of as describing the behavior of the
equilibrium altitude of the balloon tracer, rather than its
actual instantaneous position as a function of time. The
balloon tracer oscillates around its equilibrium altitude, as
its other parameters oscillate around their equilibrium
values. The dynamic effects influence the details, but not
the gross features of tracer balloon behavior.
In the final section of the report, the authors re-examine
the design goals in light of the theoretical analysis, their
experience in designing and building the testbed prototype
balloon tracer, and the experimental results. They conclude
10
-------
without reservation that an Adjustable Buoyancy Balloon Tracer
of atmospheric motion meeting the design goals is feasible.
They are proceeding in Phase II to turn this conviction into
operational hardware.
2. INTRODUCTION
An Adjustable Buoyancy Balloon Tracer of Atmospheric
Motion is a physical Lagrangian tracer (PLT) — that is. an
airborne instrumentation system that follows the flow of air
in its vicinity, and that can be tracked electronically.
For decades, such a system has been desired by researchers to
aid in understanding the dynamics of the atmosphere, and more
recently, to cast light on long-range air pollution: acid
deposition, regional haze and oxidant episodes, and associated
problems of international diplomacy. The present effort is
primarily motivated by the need to establish source-receptor
relationships to distances of order 1000 km, to evaluate the
accuracies of available air pollution transport models, and to
assess the inherent limits on the predictability of source
impacts at that distance. In a more basic sense, however, it
addresses an underlying broad need in atmospheric science for
a convenient means of following atmospheric flows.
The need for a physical Lagrangian tracer has led to
extensive work with constant volume balloons (CVBs) reviewed
by Tatom and King (1977) and more recently by Zak (1983).
CVBs follow the horizontal motions of the volume of air in
which they are embedded, but not the vertical motions.
Coupled with wind shear, this characteristic limits their
For a detailed discussion of the meaning of the term
"Lagrangian" and of how the concept depends upon the
spatial scale of application, see Zak (1983).
11
-------
usefulness — hence the search for a better tracer of
atmospheric motion (Danielson. 1961; see also Appendix G).
PLTs are necessarily balloon-borne systems, and
consequently must operate under Federal Aviation Regulations
Part 101: Moored Balloons, Kites, Unmanned Rockets, and
Unmanned Free Balloons (Appendix A). The purpose of this
regulation is to strictly limit the hazard to air navigation
which such systems might otherwise represent. According to
FAR 101. its provisions apply to any unmanned free balloon
that:
(i) Carries a payload package that weighs more than four
pounds and has a weight/size ratio of more than
three ounces per square inch on any surface of the
package, determined by dividing the total weight in
ounces of the payload package by the area in inches
of its smallest surface;
(ii) Carries a payload package that weighs more than 6
pounds;
(iii) Carries a payload of two or more packages, weighing
more than 12 pounds; or
(iv) Uses a rope or other device for suspension of the
payload that requires an impact force of more than
50 pounds to separate the suspended payload from the
balloon.
Consequently, to be exempt from the other provisions of
FAR 101. an unmanned free balloon must either:
(a) Carry a payload package weighing less than four
pounds;
(b) Carry a payload package weighing less than six
pounds, and have an areal density of less than three
ounces per square inch;
12
-------
(c) Carry multiple payload packages each one of which
satisfies (a) or (b) above, the total weight of which
does not exceed 12 pounds, and which are connected by
a suspension with a breaking strength of less than 50
pounds.
The philosophy is that, to be exempt from the more
restrictive provisions of FAR 101. an unmanned free balloon
should represent no more hazard to air navigation than a large
bird in flight. Under these regulations, the National Weather
Service, together with the Weather Services of other nations
around the world through WMO. routinely launch hundreds of
radiosonde balloons without incident twice daily. Radiosondes
measure the meteorological conditions aloft from the surface
to beyond 60,000 feet above many major airports and certain
other selected sites.
It is highly desirable for a PLT to operate under the
exemption clauses of FAR 101. Certain other provisions of FAR
101 would severely limit the usefulness of a PLT which was not
exempt. Even though there is no FAA requirement to do so on
an exempt balloon, the PLT will also carry a radar reflector
and an FAA transponder. In this way the FAA will be able to
keep track of its altitude and location.
The intended uses also provide other PLT design goals:
• Lifetime >. 3 days.
• Tracking range > 1000 km in the northeastern
quadrant of the United States.
• Telemetry of relative vertical air motion, pressure,
temperature, and humidity.
• Ground system capable of handling several PLTs at a
time.
• Capable of establishing specified ascent and descent
rates under radio command.
13
-------
Capable of reaching altitudes up to 500 mb (5.5 km).
Capable of following mean vertical flows as low as
1 cm/s with "acceptable" fidelity.
Sufficiently inexpensive to permit use in
significant numbers on an expendable basis.
The project as currently funded is divided into two phases:
Phase I: Systems Design and Demonstration
of Feasibility
Phase II: Development of an Operational
Prototype.
Phase I involves concept development, theoretical
analysis. design, fabrication, and limited testing of a
testbed prototype PLT -- a system capable of establishing
feasibility, but which is not intended to meet all of the
design goals for an operational PLT. Note that the
operational PLT to be developed in Phase II is still
designated as a prototype. This is because for subsequent
use. it is probable that a manufacturer of flight
instrumentation would be called upon to fabricate the PLTs.
These could differ in detail from the Phase II prototype.
This report covers work on Phase I.
3. CONCEPT AND THEORETICAL ANALYSIS
a. Buoyancy Control
Balloon systems obey Archimedes' Principle: A body
immersed in a fluid is buoyed up by a force equal to the
weight of the fluid displaced. This implies that a balloon
system will be in equilibrium when the weight of the air it
displaces is equal to the weight of the system.
14
-------
Constant volume balloon systems have a unique
characteristic. The equilibrium condition is met only at one
well-defined altitude, and the CVB seeks that altitude. If it
should find itself above the equilibrium altitude, the CVB
will experience a net downward force due to gravity because
the ambient air is less dense at the higher altitude, and the
volume of air displaced is fixed. Likewise, if it should be
below its equilibrium altitude, it will experience a net
upward force, since then the buoyancy force exceeds the
gravitational force. Thus, CVBs tend to oscillate around
their equilibrium altitude. In atmospheric flows which have
zero average vertical velocity, CVBs follow the horizontal
flow at their equilibrium altitude. However, in flows in
which the vertical component is significant, a CVB will not
adequately follow the vertical flow. Consequently, wind shear
can result in large horizontal as well as vertical displace-
ments between a CVB and the volume of air it was intended to
follow -- as much as 1300 km in 12 hours (Danielson, 1961;
Appendix G). For long range applications. CVBs also suffer
from the problem that they have no mechanism for dealing with
high terrain -- they need to be flown above the altitude of
the highest terrain likely to be encountered.
Nevertheless, CVBs offer the best approximation to a PLT
of any passive balloon system. To improve on the CVB. it is
necessary to go to an active system. If. for instance, a CVB
is designed to sense its deviation from the vertical flow, and
to adjust its buoyancy to keep its average vertical motion
relative to the air near zero, then one has a system which
follows both horizontal and vertical flows. This is the
principle of the adjustable buoyancy balloon tracer. To
create a PLT, one must develop a means of adjusting the
buoyancy of a CVB, and a means of sensing deviation from the
mean vertical flow.
15
-------
Figure 3-1. Schematic Diagram of the Adjustable Buoyancy
Balloon. The outer skin of the balloon is made of a material
which expands very little as the internal pressure increases.
A thin polyethylene inner balloon, or "ballonet", keeps the
helium lift gas separate from the air ballast. A pump (P)
permits more ballast air to the taken on. A valve (V) permits
ballast air to be vented. The total volume remains nearly
constant, so pumping or valving changes the average density of
the system, and thus its altitude.
16
-------
The first problem was solved in principle by Lally (1967)
of the National Center for Atmospheric Research. He proposed
a CVB with an inner bladder, or ballonet. to contain the lift
gas (helium). The remainder of the volume of the CVB was
filled with air (Figure 3-1). A system of pumps and valves
allowed air to be pumped in or to be released, respectively
increasing the mean density of the balloon and thereby
decreasing the equilibrium altitude, or decreasing the mean
density and thus increasing the equilibrium altitude. Lally's
concept forms the basis of this project. It was tried by
the French on stratospheric balloons with some success, but
was not developed further (Blamont et al. 1974).
We examine the behavior of this proposed PLT system in a
standard atmosphere. In Appendix B, an expression is derived
for the rate at which the equilibrium altitude can be changed
by pumping or valving in a standard atmosphere. This rate
v has the dimensions of a velocity, and for convenience, we
choose to call it the "pumpdown speed". In the absence of
inertial effects and drag forces, it would be the actual
vertical speed with which the balloon would move under pumping
or valving. It is given by:
V = if = -1.042 X 104 | (1 - 2.256 x 10 5Z)
P (3-1)
Here z is the equilibrium altitude in metres MSL. S is the
speed (volume flow) of the pump or valve in cubic metres per
second, and V is the volume of the balloon in cubic metres.
Consider a system with volume 12.5 m , and pumping speed
-4 3
7.0 1/min (1.2 x 10 m /s). These are the approximate
parameters for the testbed prototype PLT system. The pumpdown
speed at sea level in a standard atmosphere given by (3-1) is
17
-------
10 cm/s. Sustained average vertical flows in the atmosphere
are typically less than 10 cm/s. Of course, higher pumping
speeds are feasible, and one can release air through a valve
far more rapidly, so a PLT capable of responding to higher
vertical velocities can be built. However, for the long range
applications presently in mind, where the interest is in
following macro-scale, rather than micro-scale flows, this
pumpdown speed appears to be adequate.
Another relationship important to PLT behavior is the rate
at which the superpressure in the balloon varies with
equilibrium altitude. This rate influences the altitude range
over which a PLT system could be used. In Appendix C, an
expression is derived which gives superpressure. P , as a
s
function of equilibrium altitude in a standard atmosphere.
Note that superpressure is the difference between the absolute
pressure inside and outside the balloon. It is given by:
P = 2.396 X 104 C (1 - 2.256 X 10~5z)
s v (•*-*)
Here superpressure is given in millibars, and C is a constant
satisfying the above equation for the initial superpressure
P specified at some altitude, z . For our testbed PLT,
so o
assuming a superpressure of 35 mb at 1000 m MSL,
C = 1.868 x 10~2
(3-3)
Since
dPs = -.541 C
dz V (3-4)
_4
we find dP /dz = -8.1 x 10 mb/m. This is a remarkable
S
result. As long as the PLT is maintained at neutral buoyancy
by inhaling or exhaling air. a change in altitude of 10 km
18
-------
results in a change in superpressure of only 8.1 mb. Thus, our
prototype PLT with a design maximum superpressure of 80 mb
could easily function anywhere between sea level and 500 mb
(5.5 km) without difficulty from over- or under-pressurization.
It should be recognized that all preceding calculations
have implicitly assumed that the gas in the balloon is at the
same temperature as the ambient air at the balloon altitude.
This is not necessarily the case. A diurnal temperature swing
due to radiative effects as great as ±6°C at 900 mb. and
-10°C to +15°C at 500 mb, is to be expected (Lally. 1985).
This is based on experience in the GHOST program.
The effect of balloon temperature differing from ambient
temperature is calculated in Appendix D. We find:
4.255
AP, = AT [3.52 (1 - 2.256 X 10~5z) + 83.24 ^]
(3-5)
where AT is the temperature difference between the balloon
gas and the ambient temperature, and AP, is the pressure
difference induced by the temperature difference. For our
prototype system, a AT of 6°C (-6° at night, +6° during the
day) results in a superpressure difference of ±20 mb at 900
mb. The expected temperature swing at 500 mb results in a
superpressure swing of from -21 to +32 mb at 500 mb. Somewhat
smaller pressure swings can be expected when t>he balloon is
shadowed by cloud on an otherwise sunny day. Thus, if the
superpressure is 35 mb when the system is at temperature
equilibrium at 1 km, the superpressure would fall to below 15
mb at night, and rise at midday to above 55 mb at 1 km. or 65
mb at 5.5 km.
Another challenge to the control system is condensation
19
-------
and liquid water adhesion to the skin of the balloon. If it
should rain, should the dew point depression become zero, or
should nighttime cooling of the balloon exceed the dew point
depression, water on the skin of the balloon would add weight
which would have to be counteracted by the control system.
According to Lally (1967). a properly treated balloon surface
2
will retain only 5 g/m of water. An untreated surface may
retain up to ten times as much. The relationship between
surface area and volume for a sphere is
(3-6)
2
Our prototype PLT has a surface area of 26 m . which implies
130 g of water to carry if the surface is properly treated.
This is the equivalent of about 110 liters of air ballast at
sea level. So for a 12.5 m balloon, about 9 mb of
superpressure would be required which could later be valved
off to compensate for the added weight.
A more serious challenge to altitude control is ice
formation in clouds above the freezing level. The hope here
is that surface treatment will ameliorate the problem. In
addition, use of a partially-aluminized balloon could raise
the balloon temperature day and night (Lally, 1975). If these
strategies are not sufficiently effective, use of the PLT
under ice-forming conditions may not be feasible.
Power requirements are another concern. All power needs
must be met by on-board power sources. Furthermore, because
of the weight constraints imposed by FAR 101, it is not
feasible to expand the battery complement without limit.
Whereas the sensors, microprocessor, and other electronics
onboard can be reduced to very low power consumption levels,
the minimum pumpdown energy expenditure is fixed by the
physics of the situation, and by the efficiency of the pump.
20
-------
In Appendix E, we have calculated the energy required per
metre (dW/dz) to pump the balloon down, assuming a perfectly
efficient pump. The result is
= 2.30 x 102 C
(3-7)
For our prototype system, with 35 mb overpressure at 1 km, we
find dW/dz is 4.3 joules/meter. So, to pump the system down
1 km requires 4300 joules, or 1.2 watt hours. This is an
encouraging result. Lithium batteries have energy densities
between 100 and 200 watt-hours/pound, depending upon the type
of battery. Even allowing for pump inefficiency, long
pumpdowns should be feasible with modest battery complements.
Having confirmed that the adjustable buoyancy balloon
system is likely to work, the practical questions arise as to
how to go about filling the system in a way designed to give
proper lift, and what the maximum altitude is that the system
can attain. Both of these questions are addressed in Appendix
F. For the testbed prototype, with approximate balloon
envelope weight of 4.5 kg. and payload weight of <2.7 kg (6
pounds), and with a desired superpressure of 35 mb at 1000 m
MSL. the "required lift" at that altitude as defined in
Appendix F is 7.75 kg, assuming a surface temperature of
282° K. This would be provided by 0.31 kilogram-moles of
helium (6.95 m at STP) . Properly filled, the prototype PLT
would have a service ceiling of 6.2 km MSL, well above the 500
mb level (5.5 km) .
b. Control Strategies
We turn now to the second problem to be overcome in
creating a PLT, developing a means of sensing deviation of the
system from the mean vertical air flow. There are two
21
-------
approaches. The first is to measure the relative vertical air
velocity, and to integrate that velocity with time to obtain
relative vertical displacement. The second is to take
advantage of the near-adiabatic nature of atmospheric flows.
and to use potential temperature, or equivalent potential
temperature, as the control parameter. Both approaches have
advantages and disadvantages.
The first approach is very direct. It yields the desired
information with few if any assumptions. On the other hand, it
places very stringent demands upon the relative vertical
velocity measurement. If the measurement involves an average
systematic bias of only 1 cm per second consistently in the
same direction, the control system will create a relative
vertical displacement which grows linearly with time at the
rate of 36 m/hour. On the other hand, if the error is entirely
random, the uncertainty a in the total relative
displacement is given by
oz = /N~0vAt (3-8)
where cr is the uncertainty in each relative vertical
velocity measurement. At is the time required for each
measurement, and N is the total number of measurements made.
If a =1 cm/s. At = 60 s. and N = 4320 (for a 72-hour
flight), then cr is only 39.5 m -- quite acceptable.
However, measurement error is always a mix of both random and
systematic error, with the proportions of the mix dependent
upon the details of the measurement technique.
Use of the near-adiabatic character of atmospheric flows
poses less of a measurement problem. In the absence of liquid
water and of diabatic heating or cooling, the potential
temperature 9 is conserved -- that is, trajectories are
22
-------
isentropic (Holton, 1979). Potential temperature 0 is given
by:
9 = T(1000/P)Ra/Cp (3-9)
Here R is the gas constant for unsaturaged air. c is
the specific heat at constant pressure for air, and
R /c = .286. Differentiating and evaluating the
a p
expression for potential temperature, one finds that at the
surface, in a standard atmosphere,
3A _-j
f2 - 3.3 x 10 °/m
32 (3-10)
So, if one controls 6 to ±0.1°, z is controlled to ±31 m
-- quite adequate for our purposes.
However, diabatic effects do occur. In the region of
interest below 500 mb. they are strongest near the surface,
and decrease with height above ground. Throughout most of the
troposphere, diabatic heating and cooling is of order l°/day
(Wallace and Hobbs, 1977). Depending upon meteorological
conditions, this may be net heating, net cooling, or interim
variation with no net gain or loss. In the absence of a means
of taking these effects into account, they would limit the
accuracy with which an isentropic trajectory reflects air
motion.
In the presence of liquid water, condensation and
evaporation occur with the result that potential temperature
is no longer conserved even if the flow is adiabatic.
However, "equivalent potential temperature" (Holton, 1979;
Wallace and Hobbs, 1977) is conserved in both wet and dry
processes. It could be used for altitude control in the
presence of liquid water. The equivalent potential
temperature of saturated air is given by:
23
-------
a 0 exp (Lqg/cpT)
(3-11)
where L is the latent heat of condensation, and q is the
s
saturation mixing ratio of water vapor in air.
It would be quite convenient to use potential temperature
(or equivalent potential temperature when the relative
humidity is 100%) as the buoyancy control parameter for a
PLT. However, in a well-mixed layer, 89/8z goes to zero.
Hence, 0 is not useful in this situation. Thus, for
pollutant transport studies in the daytime mixed layer, some
other control strategy would have to be used -- like relative
vertical displacement. On the other hand, during daytime
mixing, pollutants rapidly become uniformly distributed
throughout the layer. Hence, precisely where the PLT is in
the layer matters little (see Appendix J). Over flat terrain
in the mixed layer, the PLT could be flown as a passive CVB,
or programmed to maintain constant pressure altitude.
There is good evidence from previous experiments that this
is the case. In the DaVinci III experiment, a balloon flew
parallel to the plume from a large power plant for 14 hours in
the daytime mixed layer. This was confirmed by periodic
instrumented-aircraft plume cross sections. Statistically
significant divergence between the plume centerline and the
balloon was observed, if at all. not until the evening hours
(Zak et al, 1981). Also, in the Tennessee Plume Study (Gay et
al, 1981; Zak, 1981), a balloon was launched into the plume
from a power plant the effluent of which had been spiked with
sulfur hexafluoride. The balloon stayed in the plume during
the mixing period for the subsequent 8 hours for which it
remained airborne. This was confirmed by observing the
concentrations of SF,. on board the balloon as a function of
D
24
-------
SULFUR HEXAFLUORIDE TRACER
AUGUST 20, 1978 LAMP FLIGHT
z
g
K 103
as
UJ
a.
fe 102
UJ
a
1 10
u.
X
I 1
DC
LJL
D
• 1 1 1 1 I 1 II
- t
,_ _
: • ' .
r- . • • . —
"*
r •
» ..
— . i _
- ..••!!
1 1 1 1 1 1 1 1
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900
TIME COT
Figure 3-2. Sulfur hexafluoride concentration versus time
measured on board the LAMP (Lagrangian Measurement Platform)
balloon during the Tennessee Plume Study (Gay et al, 1981).
The balloon was launched into a power plant plume spiked with
SF6.
25
-------
H-
o
g
o
"»
s
0
s
o
o
i/>
tr
UJ
i-
LU
2
S
O
CJ
O
o
tn
u
_i
Z
UJ
i—
3
I
P
4->
cn
0)
e
0)
(0
CO
-------
time (Figures 3-2 and 3-3). as well as by instrumented
aircraft cross-sections of the plume in the vicinity of the
balloon.
It is when atmospheric stability suppresses mixing that
buoyancy control is important. Danielson (1961) has shown
that under these conditions, even in the presence of diabatic
effects, isentropic trajectories yield far better
approximations to air parcel motion than do isobaric
trajectories (Appendix G). Danielson remarks that a passive
CVB trajectory approximates an isobaric trajectory.
From the foregoing discussion, it appears that a hybrid
control algorithm will yield the best results, with different
control strategies to be used for different situations --
convective mixing, stable and dry, stable and wet. and perhaps
others. The output of onboard sensors will determine which
option is selected at a given time during a flight. In the
testbed prototype PLT, the initial altitude control scheme
makes use of potential temperature. At the same time, a
sensitive vertical anemometer has been developed which permits
exploration of control algorithms based upon measured relative
vertical displacement.
4. SYSTEMS DESIGN
A major task during Phase I was development of a
conceptual design for a complete operational system capable of
meeting the design goals listed in Section 1. This was
accomplished. The preliminary systems design, and the
rationale for each of the major decisions it embodies, are
presented in this section. Note, however, that the generic
design presented here is not rigid. As we proceed with Phase
27
-------
II. it is possible, even likely, that some features of the
design will be modified as work progresses.
a. Tracking and Data Handling
A fundamental systems design decision concerns how the
PLTs are to be tracked, and how the data acquired are to be
received. Tracking techniques considered include Omega,
LORAN. FAA VOR/DME, special DME, Automatic Direction-Finding
(ADF). radar, and satellite techniques. Each system has its
pros and cons, but all but the satellite techniques suffer
from a common problem. For all of the others, tracking and
data reception can be accomplished only when the PLTs are
within radio range of an appropriately-equipped ground or
airborne station. For the VHF frequencies, this implies
line-of-sight for reliable communications. HF frequencies
propagate to longer ranges than VHF. but other problems
militate against their use.
If one presumes tracking by ground stations rather than
specially-equipped aircraft, a large number of such stations
are required to track over the desired 1000 km x 1000 km
area. This is because line-of-sight communication is limited
by the curvature of the earth. The number of ground stations
for a specified minimum flight altitude is given in Figure
4-1. Note that for a flight altitude of 1 km. approximately
35 ground stations would be required. With this ground
network, one would sacrifice data from PLTs which are carried
below 1 km. Ground stations capable of both data reception
and tracking would likely cost $50-100 K each, exclusive of
installation and maintenance. Thus a ground network could be
very costly. The alternative of tracking and data reception
using specially-equipped aircraft involves lower capital, but
higher operating costs, and is not consistent with being able
28
-------
ALTITUDE IN HUNDREDS OF FEET
M 01 5 g § 8 8 I
^v
x
\
^
m
NUMBER OF STATIONS
'yro COVER IOOO^M
\SQUARE AREA
^
/
V
\
/
\
/
/R
Kl
s
>
/
\N
.0
S
>
/
3E
VIE
S
f
: i
:T
<
V
•J
\
s
t
/
I
/
^
X
x^
\
\
/
f
' /
/
\
.
'
10
20
50
100 200
500
RANGE OR NUMBER
Figure 4-1. Number of ground stations required to cover a 1000
km x 1000 km area, and VHP radio range as a function of
balloon altitude. Radio range d in miles is given by
where h is altitude in feet.
29
-------
to track several widely-dispersed PLTs simultaneously.
There exists, however, a satellite-based tracking and data
handling system well-suited to our needs. It is called ARGOS.
and is a joint undertaking of NASA, NOAA. and of Centre
National d'Etudes Spatiales (CNES, France). Appropriately,
the ARGOS system is primarily intended for applications
concerned with environmental data collection. An overview of
the ARGOS system is included as Appendix H, so only a brief
description of its application to our effort is given here.
The ARGOS satellite-based data collection and platform
location system makes use of NOAA satellites, a pair of which
orbit at an altitude of 850 km with periods of 101 minutes.
The PLT payload will carry an ARGOS platform transmitter
terminal (PTT). The PTT about once a minute broadcasts a
series of data frames on 401.65 MHz. If a PLT Ground Support
Station (GSS) or properly equipped aircraft is within line of
sight of the PLT, the data stream can be received directly
with an ARGOS uplink receiver. The GSS as currently
envisioned will be described in detail later.
For long range experiments, the satellite link is most
convenient. The data from the PLT will be received by the
satellite as it comes within view (within 2500 km). The
satellite, in addition to storing the data on board for later
retransmission to Service ARGOS. also retransmits the data
immediately on 136.77 or 137.77 MHz. The GSS will receive
those data along with position information via an ARGOS Local
User Terminal (LUT). This takes place only when the marker is
within view of the satellite and when the satellite is within
view of the LUT. For the northeastern US, on average, this
happens 10 times per day (minimum: 8; maximum: 12). The
satellite "sees" an area 5000 km wide on each pass. While the
30
-------
radio link exists, data can be transmitted for approximately
10 minutes. These can be both real-time and archived data
taken since the last satellite pass. Thus the LUT provides
location and other data every 2-3 hours. The location
accuracy obtained using the LUT link is about 3-5 km.
Once during each orbit, each NOAA satellite dumps its data
to a Service ARGOS ground station, from which they are
transferred to the World Meteorological Organization Service
ARGOS data processing center in Toulouse, France. After
processing, the data are transferred to a NOAA satellite data
center in Suitland. Maryland. Typically 4-6 hours after data
is broadcast from the PLT, they will be available to the GSS
via modem and telephone lines from NOAA Suitland. The
location accuracy from this delayed link is considerably
better than from the LUT — typically ±700 m. Service ARGOS
can handle more than 200 PTTs simultaneously.
It is important to note that the use of ARGOS does not
preclude continuous data reception. If either a ground
network or airborne data reception units with ARGOS uplink
receivers are provided for a given experiment, continuous data
will be obtained. However, long range experiments can be done
without such facilities. Hence, the minimum investment needed
for an ARGOS-based PLT system is much smaller than for systems
based on other tracking and data-handling options.
Other advantages of ARGOS which led to its selection are:
it already exists -- it doesn't have to be developed from
scratch; it is very well proven -- it works with close to 100%
data recovery; lightweight PTTs designed for use on balloons
are commercially available, as are uplink receivers and local
user terminals; ARGOS offers world-wide coverage -- it's not
limited to the northeastern U.S.; and finally, the resources
31
-------
available for this project do not permit the development of an
analogous system based upon some other tracking option.
b. Balloon Envelope
Balloon design was undertaken in consultation with the
National Center for Atmospheric Research, and in collaboration
with Raven Industries, the maker of NCAR's high altitude
constant volume balloons. Three different CVB shapes were
considered: cylindrical, tetrahedral. and spherical.
Cylindrical and tetrahedral balloons are simpler to fabricate.
and hence offer the potential for lower unit cost. On the
other hand, the spherical shape provides the lowest surface
area per unit enclosed volume, and the lowest skin stress for
a given superpressure and skin thickness.
Raven carried out a detailed design study on all three
shapes. They found that the increased skin thickness
necessary to accommodate the higher stresses in the first two
shapes considerably increased balloon weight. Thus, those
balloons would have to have been considerably larger in order
to carry the desired payload plus the weight of the balloon
itself. In part as a result, the cost difference between the
three designs was only about 10%. The smaller size of the
spherical balloon, its significantly lower surface area, and
the long experience of both Raven and NCAR with spherical-
design CVBs made the choice obvious.
The calculated parameters for the respective designs to
meet the specifications for the testbed prototype system are
given below:
32
-------
Sphere
Volume
Diameter
Material
Ballonet
Weight
12.47 m
2.88 m
3.3 mil polyester (bilaminated)
1.0 mil polyethylene
4.1 kg
Tetrahedron
Volume
Diameter
Material
Ballonet
Weight
19.84 m
4.24 m
4.8 mil polyester (bilaminated)
1.0 mil polyethylene
8.2 kg
Cylinder
Volume
Diameter
Length
Material
Ballonet
Weight
23.80 m
1.83 m
7.32 m
4.2 mil polyester (bilaminated)
1.0 mil polyethylene
11.8 kg
Balloons of these designs are able to carry a 4.5 kg
payload to a pressure altitude of 600 mb with a superpressure
of 80 mb. and with a maximum skin stress of 10,000 psi. A
lighter payload. of course, could be carried to higher
altitude.
c. Payload
The operational PLT payload will consist of the
following elements:
33
-------
IT
Figure 4-2. Buoyancy Control Plumbing. Pumps (P) and valves
(V) are electrically-operated. For the Testbed Prototype,
only 2 pumps are used, but for the Operational Prototype, 3
will be incorporated to gain a 50% increase in pumpdown
speed. A valve is in series with the pumps to avoid leakage
through the pumps when they are not operating.
34
-------
• Buoyancy adjustment subsystem
• Sensors
• Microprocessor or microcomputer
• Telemetry subsystem
• Radio command subsystem
• Tracking aids
• Batteries
These are described below.
The buoyancy control subsystem is just the set of pumps
and valves necessary to change the amount of air ballast
»
carried by the PLT. The buoyancy control plumbing is shown in
Figure 4-2. The pumps used in the testbed prototype are
adequate for our purposes, but the manufacturer may be able to
further optimize them. A valve is in series with the pumps
because the pumps do not prevent backflow when off.
The list of sensors which are either essential or
desirable is quite long:
• Ambient pressure
• Ambient temperature (aspirated)
• Ambient humidity (aspirated)
• Vertical anemometer
• Balloon orientation
• Rain sensor
• Internal pressure
• Internal temperature
• Skin strain
• Upward-looking radiometer
• Voltage
• Ma s s f1ow
35
-------
According to the specified design goals, ambient pressure.
temperature, and humidity are to be measured and reported.
They are standard meteorological measurements that are already
made on the testbed prototype. Minor improvements in the •
sensors will be made, but no major problems are anticipated.
The vertical anemometer needs to be an order of magnitude
more sensitive than standard meteorological instruments.
Sonic, hot wire, ionic, and propeller anemometers were
considered. A sonic anemometer would certainly be able to
make the measurement with adequate accuracy. However, weight.
cost, and power consumption considerations for commercially
.available sensors make this choice somewhat unattractive. The
hot wire anemometer suffers from the difficulties that it does
not give direction of flow, and that power consumption is
typically high. Ionic anemometers are in an early stage of
development (Barat, 1982). At the level of accuracy we
require, they suffer from drift problems, and are not
commercially available.
That leaves the propeller anemometer. It was recalled
that some years ago P. B. MacCready developed a very sensitive
vertical anemometer under contract to SNL for the DaVinci
experiments (MacCready, 1977; MacCready and Mullen, 1981).
Gill (1979) has also developed a similar device. The
MacCready design shows promise of being able to meet our needs
at low cost, weight, and power consumption, using
commercially-available components. It is being further
developed for the testbed prototype, and will be discussed in
Section 5. If it should not live up to its promise, it may be
necessary to fall back on the sonic anemometer.
If a propeller anemometer is used, rotation of the balloon
itself would be a source of error. Consequently, it would be
36
-------
necessary to keep track of the balloon orientation.
Fortunately, a convenient electronic compass is available and
has been incorporated into the testbed prototype.
It is likely that a different control algorithm would be
required in rain than under other conditions. Hence a rain
sensor is incorporated.
Pressure and temperature measurements within the balloon
will be made to assess its condition, and to aid experimenters
in understanding the dynamics and thermodynamics of the
system. Skin strain will also be measured with the same view
in mind.
In Section 3, during consideration of an isentropic
control strategy, it became clear that it would be
advantageous to be able to estimate and correct for diabatic
effects. A simple upward-looking radiometer capable of
sensing solar insolation during the day. and of measuring
radiant energy loss at night, might provide adequate
information for such an estimate.
The state of health of the payload will be reflected in
the values of selected reference voltages. Hence a
"voltmeter" will be required.
Finally, mass flow into and out of the balloon will be a
function not only of the pump and valve used, but of the
ambient pressure and the superpressure as well. Kurz
Instruments of Carmel Valley, California, has recently put on
the market a small sensor and associated electronics which
seems suitable to our needs. It could simplify the control
problem.
37
-------
Figure 4-3. NCAR microprocessor data system for balloon
application, shown approximately actual size.
38
-------
Returning now to the main list of payload elements, note
that the microprocessor or microcomputer is the heart of the
system. Its function is to analyze the sensor data in
accordance with the control algorithm, and to operate the pump
and valve appropriately. It may have other functions as well
— processing the sensor output, formatting the sensor data
for transmission, and whatever other data-handling and control
tasks may be necessary on the payload.
Microprocessor technology has advanced so rapidly that it
may be possible to incorporate an entire microcomputer on a
single printed circuit board addressable in a high level
language, rather than a microprocessor with read only memory.
If so, this would greatly increase the flexibility of the
system.
Microprocessor-based data systems have been used by
Lally's group and others for sophisticated data archiving,
manipulation, and reporting via ARGOS. A typical system is
shown in Figure 4-3. One would expect the operational
prototype PLT microprocessor or microcomputer system to be
similar.
Having made the choice of ARGOS for tracking and data
collection, the telemetry system becomes well-defined. An
existing, commercially-available ARGOS PTT designed for
balloon application is shown in Figure 4-4. It is similar to
the unit to be incorporated into the operational PLT, except
that the new unit does not require a constant temperature
cell. An ARGOS antenna is shown in Figure 4-5.
The design goals call for the capability to establish
specified ascent or descent rates upon radio command. The
only other command capability deemed necessary is to cause the
39
-------
M-l
Ol 03 03 S4
•H OJ >1 0)
CO .H 1-4 4->
M
EH XI 4-> 0)
Qj THO<
^ 03 g 03
&
03 3 ro <1)
C rH M CO
•H O 4J 3
g c
M -H CO
I -H 03
^1 4_) 4J (0
03 £5 O
0) O 03 t3
M -H 4J
3 rH CO T3
O^ Cil CH d)
•H D) O 03
[YJ fO O J3
40
-------
balloon to descend rapidly -- commonly called "cutdown".
Command transmitters, receivers, and decoders are in common
use at SNL and at NCAR, and are not anticipated to present a
problem. The system to be used would most likely provide
protection against radio interference or command
misinterpretation. This is done by automatically transmitting
the coded command twice. The command is not acted upon unless
the replicate transmissions received are confirmed to be
identical. The encoding permits selection of the balloon to
which the command is addressed if several are simultaneously
airborne, as well as selection of the action to be initiated.
It is planned to incorporate a radio command system which
operates at HF, rather than VHP or UHF frequencies to get
around the requirement for line of sight between the
transmitter and the balloon system. At the high power level
available from a ground-based transmitter, HF signals can be
received at great distances, frequently, around the world;
more about this later.
While the ARGOS system provides primary tracking data to
the experimenters, it is also desirable to provide an
independent means for FAA Air Traffic Control to identify and
track each balloon on their radar screens. This is done by
incorporating a commercially-available FAA transponder. The
transponder need not operate continuously. Rather, the
microprocessor will cycle it on and off frequently to reduce
average power consumption. To permit radar-equipped aircraft
to see the PLT. it will also carry a very light-weight radar
reflector. At night, a lightweight strobe light will operate.
Much of the above equipment requires electrical power
which can only be provided by batteries (in this application.
solar cells are not feasible). The power requirements, though
41
-------
N
O
CN
fl N
O vo
iH CO
•H x:
o en
4-1 -H
U C
2-H
ro
• >«^
in
i e
^ o
cu vo
u •
3 r-
CT
•i-l CO
42
-------
Figure 4-6. "Paper" lithium battery. The battery shown here
has a 1.2 Ahr (7.2 Whr) capacity, weighs 34 g (1.2 oz) and is
0.46 cm (0.18 in) thick.
43
-------
modest for short to intermediate operating periods, become
considerable when a three-day lifetime is considered.
Fortunately, great advances have been .made in recent years in
lithium battery technology. Lithium batteries based on
several different chemical reactions are commercially
available, all with energy densities of about 100-200 watt
hours/pound. Several of these have eliminated the overheating
hazards associated with earlier lithium batteries. Thus for a
72-hour flight, an average power consumption of 3-6 watts is
obtainable with a 1 leg battery pack. This appears to be
adequate to our needs if care is taken in electrical systems
design.
A development of particular note is that of "paper"
batteries (Margolin. 1982). These batteries are made of
paper, thin plastic materials, metal foils, and electrolytes.
They are deformable, rather than rigid. They are attractive
from the point of view of further reducing the potential
hazard to aircraft even beyond the requirements of FAR 101.
Figure 4-6 shows a typical paper lithium battery.
d. Ground Support Station
As presently conceived, the GSS will accommodate three
distinct modes of use of the PLT. The mode we have focused
upon to this point makes exclusive use of the ARGOS satellite
for tracking and data handling. This mode provides data
whenever one of the NOAA satellites is within view of the PLT
-- about ten minutes every two to three hours. If the PLT is
equipped for data archiving, one can receive, store and
transmit via ARGOS all the data acquired since the last
satellite pass.
44
-------
In some applications, however, it is desirable to have
continuous real-time data. Over a limited geographical area.
this can be accomplished by locating "up-link" receivers at
towers or other high points around the region of interest.
They will receive data from all PLTs within line-of-sight.
The data can be accessed by telephone modem. If it is not
necessary to have continuous tracking information, but only
continuous data from the on-board sensors, this can be
accomplished at modest cost -- $10-$15 K per site for the
hardware. Here tracking would still be done only by the
satellites. The GSS can accommodate this mode of use provided
that it is equipped with appropriate telephone service.
The final mode of use provides for continuous tracking and
data reception for short range experiments in which the PLT
remains within radio range of the GSS — tens of kilometers,
depending upon PLT altitude. In this case, it is possible to
continuously track the PLT with a radiotheodolite or LORAN
tracking system. If a radiotheodolite were used, it could
track only one PLT at a time, so if more than one PLT were
involved in the experiment, they would have to be observed
sequentially, or more than one radiotheodolite would be
needed. The LORAN tracking systems may be more attractive.
For experiments of a very local nature, either could be
replaced by an optical theodolite equipped with shaft encoders,
The three modes of use are designated, respectively:
(a) satellite/world-wide; (b) hybrid/regional; (c) ground-
based/local. In typical long range use. one would expect the
modes to be combined, especially if several PLTs are released
from a single site, and the GSS is located at that site. Each
PLT would be tracked locally while it remained within range,
and thereafter by satellite.
45
-------
ARGOS
LOCAL USER
TERMINAL
ARGOS
UPLINK
RECEIVER
VIDEO
DISPLAY
TERMINAL
HARD DISC
RADIO-
THEODOLITE
DESK TOP
COMPUTER
DUAL FLOPPY
DISCS
COMMAND
TRANSMITTER
PHONE
LINE
MODEM
PRINTER
PLOTTER
Figure 4-7. Block Diagram of Ground Support Station for an
operational PLT system.
46
-------
The elements of the GSS are shown in Figure 4-7. with the
exception of the microcomputer and the standard peripherals,
they have all been discussed in previous sections.
It should be noted that the problem of "skip" of HF radio
signals which could affect the reliability of the command
transmissions would be avoided by having another command
transmitter at a remote location accessible by telephone
modem, and also by having at least two different frequencies
available in different bands. It is unlikely that a PLT would
be in a "skip" zone relative to both command transmitters and
both frequencies. This arrangement still allows world-wide
command from a single fixed location.
In addition to accessing delayed ARGOS data from NOAA and
real-time data from remote up-link receivers, the modem would
also be used to retrieve any National Weather Service data
relevant to the experiment via Weather Services, Incorporated,
or some comparable supplier, to aid in conducting the
experiment as meteorological conditions change.
The Ground Support Station would be a compact facility,
easily accommodated in a small trailer or RV which could be
moved to the area where experiments are planned, or to the
home base of the institution responsible for a given set of
experiments.
e. Control Algorithm
For the long range experiments currently envisioned, the
main features of the control algorithm can now be discerned.
It would very likely be of the hybrid type discussed in
Section 3. making use of potential temperature and a measure
of diabatic effects as control parameters during periods of
47
-------
stability, and relative vertical motion or even barometric
pressure during periods of instability. It should be
recognized, however, that the PLT is a generic tool which can
be put to many different uses. For each use. a somewhat
different control algorithm would likely be optimal. Hence,
control algorithm development is an open-ended task which will
continue in parallel with research employing PLTs.
5. DEMONSTRATION OF TECHNICAL FEASIBILITY: THE TESTBED
PROTOTYPE (TP)
a. Approach
The theoretical analysis in Section 3 strongly suggests
that the systems design presented in Section 4 will, in fact.
work. To demonstrate feasibility, it remains to turn those
essential elements of the design about which there could be
doubt into functioning hardware.
There is no question of feasibility about most of the
system. The ARGOS data collection and platform tracking system
has been in operation for years, and has been used to track
stratospheric balloons on flights which carried them around the
world dozens of times (Lichfield. 1981). Likewise,
microprocessor data acquisition and control systems for balloon
application, as well as balloon radio command systems, are
frequently used on other projects (see Morris, 1975; NCAR).
The question that needed to be answered in Phase I concerned
the buoyancy adjustment subsystem: Could that subsystem be
built so that it performs as required within the weight and
power constraints imposed by the balloon environment and FAR
101? The testbed prototype was designed to answer that
question.
48
-------
It was recognized early that a major task in the
development of an operational PLT will be optimization of the
control algorithm. This process will be facilitated if the
control algorithm can be easily modified, especially if it can
be modified in flight. The testbed prototype PLT accommodates
that need. The microcomputer which contains the control
algorithm is on the ground. All the necessary data are
provided by telemetry, and all pump and valve commands are
transmitted by radio back to the payload in flight.
Modifications in the algorithm can be accomplished with a few
keystrokes.
As the name implies, the testbed prototype is intended to
be of continuing utility. In Phase II. as elements of the
operational prototype emerge from the laboratory, they will be
incorporated into this prototype for evaluation.
b. Balloon Envelope
The specifications for the seven TP balloons supplied by
Raven were:
• 4.5 kg maximum payload
• 600 mb ceiling altitude
• 80 mb maximum superpressure
• 10,000 psi maximum skin stress
• Internal full volume ballonet
• Base fitting for required electrical and air
feedthroughs
• Adequate load suspension harness
• Top-mounted inflation fitting
• Handling lines
The calculated parameters of the spherical balloon to meet
these specifications were given in Section 3a. To get to 80
49
-------
mb superpressure with skin stress limited to 10,000 psi would
have required 3.3 mil polyester film. At Raven's suggestion.
we relaxed that specification by 10% to take advantage of 3
mil bilaminated polyester film which Raven had in stock. This
permitted us to avoid having the film manufactured on special
order with attendant high cost and long delay. Estimated skin
stress at failure is approximately a factor of two higher than
the stress in the 3 mil material at 80 mb. The 10,000 psi
spec would only have minimized balloon expansion at high
superpressure.
Including fittings and control wires connecting the base
with the crown, the balloons weigh 4.568±.0l8 kg. This is
about 0.5 kg above the weight calculated from the design
parameters, and is entirely acceptable.
One of the balloons was earmarked to be tested to
failure. However, after raising the pressure to 92 mb (12 mb
above the design spec). Raven chose not to continue the test
to failure because of the possiblity of damage to windows at
the site where the test was being conducted.
c. Payload
The decision to keep the buoyancy control microcomputer on
the ground in Phase I permitted the testbed payload to be put
together in a particularly convenient way. For the most part,
commercially available components were used. Figure 5-1 gives
a block diagram.
The airsonde printed circuit board used to measure
pressure and temperature inside the balloon comes from a
commercially-produced rawinsonde made by AIR (Atmospheric
50
-------
Instrumentation Research), of Boulder. Colorado. Its
specifications are given in Table 5-1.
TABLE 5-1
AS-3A AIRSONDE SPECIFICATIONS
Temperature
Range: -t-50°C to -80°C
Precision: 0.5°C for +40°C to -40°C
Resolution: 0.01°C
Humidity
Range: 10% to 100% RH
Precision: 5% RH 40°C to -40°C
Resolution: 0.1% RH
Pressure (absolute barometric)
Range: 1050 mb to 5 mb
Precision: 1 mb
Resolution: 0.1 mb
Temperature Compensation: Bead thermistor with automatic
software correction
Telemetry Range; 100 km (nominal)
The tethersonde payload is also made by AIR. Its
specifications are given in Table 5-2.
51
-------
AIRSONDE
CIRCUIT BOARD
BUOYANCY
ADJUSTMENT
SUBSYSTEM
COMMAND RECEIVER
TETHERSONDE
CIRCUIT BOARD
BATTERIES
\ STROBE /
Figure 5-1. Block Diagram of the Testbed Prototype Payload
52
-------
Temperature
Range:
Precision:
Thermistor Match:
Resolution:
RMS Noise
Equivalent:
Response Time:
Humidity
Range:
Precision:
TABLE 5-2
TS-3A-SP TETHERSONDE SPECIFICATIONS
+50°C to -70°C
0.5°C for +40°C to -40°C.
typical accuracy 0.2°C
0.1°C for +35°C to -20°C
0.01°C
0.04°C
12 S
3% to 100% RH
3% for 0°C to 50°C;
5% for -10°C to 0°C;
10% for -25°C to -10°C
Pressure (absolute barometric)
Range: 1050 mb to 600 mb
Precision: 1 mb
Resolution: 0.1 mb
Temperature Compensation: Bead thermistor with automatic
software correction
Wind Speed*
Range:
Precision:
Resolution:
Wind Direction*
Range:
Precision:
Resolution:
Spare Inputs (4)
Range:
Precision:
Resolution:
0-20 m/s
0.25 m/s
0.1 m/s
2° - 358° (4° deadband)
5°
1°
0-100%
1%
0.1%
53
-------
TABLE 5-2 (Continued)
Weight: 225 g with alkaline battery
Telemetry Range: 20 km (nominal)
*In the TP. the wind speed sensor is not used; it is replaced
by the high sensitivity vertical anemometer. The electronic
compass associated with wind direction is used with the
vertical anemometer to sense and correct for balloon rotation.
Figure 5-2 is a picture of the airsonde and tethersonde
circuit boards.
Note that the specs on the temperature sensor for the
tethersonde payload may be marginal for an operational PLT
based on potential temperature. We are informed, however, that
AIR can calibrate the sensors to the required accuracy.
In an operational PLT, the temperature, humidity, and
vertical anemometer sensors will have to be located away from
the body of the balloon itself to avoid errors caused by
balloon-induced convection currents. There are two choices.
One can suspend the sensors well below the balloon (ideally,
ten diameters), or one can mount them laterally out beyond the
balloon radius. The latter approach is more appealing. For
the expected differences between the air temperature and the
balloon skin temperature, the balloon boundary layer which
would carry the convection currents would be quite thin. Using
the results of Chiang et al (1964). we find that for a
temperature difference of 5°K, the maximum convective velocity
in the vicinity of the balloon equator is about 25 cm/s, and
this occurs at a distance from the balloon skin of about 1 cm.
At distances > 10 cm from the balloon skin, the effect is
54
-------
CD
T3
•H
CO
CD
N
O
3
O
O «-'
C co
O •
CO VO
s-i va
cu
J3 T3
4-J G
0) nj
+J
/— ^
•O N
fl O
(0
ro
<1) .
•O H
C ^
O
CO &
u
ro
CM .C
I O<
in -H
cu
-------
oc
O
z
ULl
V)
cc
UJ
X
o
Q
<
O
>
<
Q.
Z
DC
^
< m
2s
t°
QC «5
c
o
o
rtJ
ja
0)
0)
>
•H
4->
(Q
rH
-------
insignificant. The same is true of the effect on temperature.
One comes to similar conclusions when one considers the flow
patterns induced by motion of the balloon relative to the air.
that is. forced convection (Schlichting. 1979). Thus, mounting
the sensors out laterally would be feasible, and more
convenient than suspending them far below the balloon (Figure
5-3). This sensor placement will be implemented in Phase II.
The testbed buoyancy adjustment system consists of two
Gilian Model No. B22A06F48 pumps in parallel, two Klippard
model No. MJV-2 mechanically-activated valves, and a Japan
Remote Control servo. The servo actuates the valves. The
servo is driven by an RC (radio control) receiver. Both are
part of a Japan Remote Control N7C-4SM system. Figure 5-4
shows these components.
The Gilian pumps weigh 158 g each, and consist mostly of
plastic. They are stock items which are satisfactory for an
operational prototype as is. but effort will be devoted in
Phase II to work with the manufacturer to see if their pumping
speed can be increased.
The mechanically-activated valves, servo, and RC receiver
will not be a part of the operational prototype. This set of
components will be replaced by solenoid-operated valves, and
will be driven by the onboard microprocessor or microcomputer.
""
In the TP. the batteries used are conventional lithium "D
cells, capable of operating the payload for approximately 72
hours. They will be replaced with paper lithium batteries in
the operational prototype.
Figure 5-5 shows the assembled testbed prototype payload.
It weights 2.36 kg, including 0.99 kg of batteries. The main
57
-------
4J rH
G I -H •
O t3 3 , O
> 4J £!
r-l (U O CO
fl3 J3 4->
> O
rH >
T3 rH
M •«-!
m 2
Oi 03
•H > O O
rH rH -H ^
hd (13
M 13
a 0) ,
"^* fTJ JZI *Q
i
Ifl M T3 t3
P O
M *r-l (0 nl
P 0) M r-l
^^^ U flJ (*\j
•H (U 0j
CLI M O M
58
-------
Figure 5-5. Assembled Testbed Prototype Payload. Tubing is air
ballast vent line. Wires are airsonde and tethersonde
antennae. Smaller diameter cylinder contains the aspirated
temperature and humidity probes. Payload housing is styrofoam
covered with 0.6 oz fiberglass. The payload meets the exemption
conditions of FAR 101.
59
-------
cylinder is 31.5 cm (12.5 in) long, and 16.5 cm (6.5 in) in
diameter. The weight/size ratio satisfies the exemption
clauses of FAR 101. The payload structure is made of expanded
styrofoam covered with 0.6 oz fiberglass. The combination
yields a strong but light structure.
d. Vertical Anemometer
The sensor for the vertical anemometer being developed and
tested mafces use of a stock 22.9 cm (9 in) diameter expanded
styrofoam Gill propeller from R. M. Young, and a slightly-
modified Spaulding Instruments Cl rotation sensor (Figure
5-6). The Spaulding sensor is a photoelectric type which
gives both rate and direction of rotation. It makes use of
the highest quality low-friction instrument bearings available
without special order. It is modified only in that dust seals
normally supplied were removed to minimize friction.
This wind sensor has a starting air speed of under 2 cm/s.
and a measurement threshold of under 3 cm/s (MacCready. 1981)
-- that is. at 3 cm/s. a reproducible air speed measurement
can be made. This appears to be just about as well as Gill
(1979) did with a special laboratory-produced 50 cm propeller
and sensor.
It is unlikely that starting speed or very low velocity
performance will determine whether or not relative vertical
air motion measurements averaged over 1-5 minutes can be made
satisfactorily. Even at a low intensity of turbulence, the
instantaneous vertical air velocities are likely to be
considerably higher. The point is illustrated by the data
from one of the experiments in which MacCready's anemometer
was used. DaVinci II (Figure 5-7). The figure presents five
minute averages of data taken once per second. Only very
60
-------
Figure 5-6. High sensitivity vertical anemometer for relative
vertical air motion measurement.
61
-------
1
Q
Q
LU
O
QC
LU
5
LU
___
UJ
£^
U.
1
•MM
MM
O
z
;^
^
g
i i I i i I
*
*
"
*
*
._
•
. •
.
•
« *
•
.*
* • «
•
,
>
,
*
.
ii iii i
• i i i i i
—
—
_
**. ~
-
-
-*•
**
• • —
' '
° * • — «
*
-
—
•
_
. • /
* « ^
• **.
. • -
•'
• * i
. • : •
*
_
_
1 1 1 1 t t
m
CM
O
c>
CM
q
r-l
0
TT
q
i-i
o
00
0
in
q
CM
O
§OOOQOOOQOQOg'H
in o m o in m o in o in o
01 CM CM ••* ^ 'T'T'CMCMfO
' * '
39S/LUO
3
rx
00
ub
o
in
0
vj
<^
D
O
•x.
z
1.
o
z
i
2
O
cr
LU
LU
2
H-
fi 0
O rH
0) 0)
M ja x:
0) 4->
4-1 0)
0) CO O O)
0) 4-> 0) rH
OD C Jld 03
J3 <1) flJ 3
o e 2t o
0) .H
C O J3 M
O (Q 4-1 01
•H rH -iH 01
4-> cu 3
0 4->
g mo
• 4J C!
S-l 4-1 03
03 O> M
rH rH J3 3
rt3 M-l 4-1
O 03
•H £3 H3
rH .H O
03 £> O
E3 0)
[-• rH O >
1 rH O
IT) rH rrj
•H rH
j-i c ja M
P T-4 (!)
5^ ^^ Oi 03
•H m x: JQ
64 Q 4J 0
62
-------
rarely were one second values of velocities as low as 2.5
cm/s. MacCready remarks that for certain nighttime portions
of the DaVinci flight, the intensities of turbulence were
among the lowest ever measured.
Because of these encouraging results, supporting
electronics were built for the MacCready sensor to enable the
tethersonde telemetry to handle the vertical anemometer data
for the testbed prototype. Currently, the range of the
resulting anemometer is 0.3 - 300 cm/s. It makes a
measurement to better than 1 cm/s in one cycle of the
tethersonde data system, about 10 s. Taking into account
random error alone, in one minute, the estimated uncertainty
on average velocity is less than ±0.4 cm/s. and in 5
minutes. ± 0.2 cm/s.
e. Prototype Ground Station
The choices made for the testbed prototype payload largely
dictated the choices for the prototype ground station (PGS).
The block diagram of the PGS is given in Figure 5-8.
Data from both the airsonde within the balloon and the
tethersonde in the ambient air are telemetered to the AIR
ground unit, the ADAS (Atmospheric Data Acquisition System).
The ADAS contains a Z-80 microprocessor which calculates the
values of the selected variables from the raw data coming from
the sondes. and from either an optical theodolite or
radiotheodolite equipped with AIR shaft encoders. It does
this by applying calibration factors for each sonde which are
read into the ADAS from calibration punched paper tapes at the
time the program is started. Upon completing the calculations
for each sonde. the ADAS dumps a line of data to an external
63
-------
CC
I-
X
o
cc
CM
co
a.
I
in
oo
0.
CD
A -
4J in
cc! co
i-i en Qu
< 0) IE
4) O 0)
£3 O .C
6-1 CO 4-)
T3
O
• fQ 4-1
C C!
O OJ OJ
•H 4->
4-> CO OJ
OJ
C CD
P M CO
O CO
l_l ^-, (Q
C5 S co
CO oj
CU 4-) CU
CU CO
4-1 en C
O OJ
4-> a
o o •
M -H CO
a. 4-> r-i
•H OJ
O
Q oj ca
a M
m co
O O £!
O -H 4->
rH M CO
ffl 0) 4->
• co C
CO O OJ
i e
in 4-1
X3 OJ
•H
CO 4J
CO O
4-» OJ
OJ
> O
•H 4->
U 0)
OJ 4->
03
in -H
CO U
CU Cu
tXi O
OJ ^i )LJ
E-t oj 4->
ca -i-i
. oj 6
S ca
x: co C
4-> CO OJ
u p 4->
o ca
D» o U
rH i—l 03
OJ O
cu
rH JC £!
O O 4->
4-> -H jC
C 3 O>
O CO p
O O
ca i-i
co co si
.C > 4->
4-J -r-l
D> CO
CO 0)
fl X3 >
•H O rH
OJ -i-l OJ
4-> £3 >
C 3
O M
O • O
r-l
J3 CM ca
CJ ^< Cu
•H n g
J3 cu p
5 EC Cu
64
-------
tJ
£
a
an 01
o - J3
o x: *J
w en
d, -H +J
M (0
• 4->
in 4->
G -H
65
-------
printer and to the HP85 desktop computer. Tracking
information is included.
The HP85 applies the control algorithm, displays, prints,
and writes to cassette tape selected calculated variables, and
transmits appropriate instructions to the HP3421 interface
unit to actuate the pumps and valves. The HP3421 effects
switch closures which cause the attached RC unit to transmit
the radio commands to the RC receiver in the payload to
operate the pumps, operate the valves, or do nothing.
whichever the control algorithm dictates. The PGS is shown in
Figure 5-9.
6. EXPERIMENTAL RESULTS
In Phase I, time and resources limited the experimental
program to the minimum necessary to demonstrate that the
concept for the PLT is viable. Initially, measurements were
made on individual components in the laboratory to determine
if their performance was acceptable. Next, the testbed
prototype system underwent tests in a tower. Finally, testing
began in the ambient atmosphere. Phase I did not include true
"free flight" testing. Such tests are manpower intensive and.
for flights of any duration, reguire a tracking capability
which will not be available until Phase II. Thus, the focus
was on testing in the controlled atmosphere of the tower.
a. Laboratory
It has already been mentioned that Raven conducted a
superpressure test on one of the balloons to assure that they
are capable of withstanding the 80 mb specified maximum
superpressure. SNL will carry out a test during Phase II to
66
-------
determine the superpressure at which balloon failure occurs.
It is important to establish the margin of safety between
operational superpressure levels, and superpressure at
failure. For the operational system, a device will be
incorporated which would initiate a controlled cutdown if the
superpressure should exceed safe levels.
Superpressure balloons, or CVBs. have only approximately
constant volume. The material from which they are
constructed, polyester film (mylar), has high but finite
modulus of elasticity. Consequently, the balloon will expand
to some degree as the superpressure rises.
A test was conducted to measure the change in volume of
the balloon with superpressure. The volume was determined by
measuring the polar and equatorial circumferences of the
balloon. This was necessary because the balloon is slightly
ellipsoidal. A 30 mb change in superpressure (from 20 to 50
mb). such as one might expect during the day-night transition.
results in a 1.3% change in volume. In the absence of a
corrective response by the control system, this would result
in a 135 m change in altitude at 1000 m in a standard
atmosphere.
After the balloon itself, the element which involved the
most uncertainty was the pump. At the outset, it was not
certain that a pump could be found which met our requirements
for flow at high backpressure, low weight, and high
efficiency. Four different pumps were examined. They were
each weighed, and then operated at their rated voltages
against backpressures of 34.5 and 69 mb (0.5 and 1 psi). The
pumping speeds and power consumptions were measured at these
backpressures. The results are given in Table 6-1.
67
-------
0.
X
oo
vo co
• •
CM r—
in in
• •
en o
co CM
r«- CM
CO CM
i—CM
•— CO
r- CM
•* O
>a- in
co co
• •
00
in
*
•— •*
CO CO
O.r—
GO *—
CO CO
CM «*•
O O
• •
i— co
in
in i—
• *
CM CO
in
r— «*
* •
r— CM
GO
O
CO
o
<
o;
s. <
S.—-
10 •«*•
CM CM
•—O
• •
CM CM
O 00
• •
in •*
CM o
Lf> IT)
CM O
in i—
CM CM
00 CO
cr> cr>
o o
p«. r^
o o
o
OL.
0)
CD
CM CM
CM CM
in in
LiJ
I
03
•
co
in
10 co
LO
•
Crt
a>
A: TJ
co O
0
en r-
CM O
i
>,CM
>f^
i^
i— •< CT>
•r- CM C
(O 1 »r™
S- O CO
O3 t— ^—
68
-------
III I I I/ I I I
CO
.a
S 1
LU
a:
z>
en
c/5
LU
o
CM
I I
1 I
00
«*
Mcru
0)
u
3
CO
CO
a<
(Q
CO
p
en
M
-------
The second to the last column in the table gives pumping
speed times backpressure divided by power consumption. This
is a figure of merit related to efficiency (multiply by .167
to get % efficiency). The last column gives speed times
backpressure divided by weight. This is also a figure of
merit, but it is less important because the batteries to run
the pump are likely to outweigh the pump in any case.
On the basis of published data, the larger Brey pump had
been procured for the testbed prototype. However, on the
basis of the test data, it became clear that the Gilian is
several times more efficient. Hence, Gilian pumps were
selected.
To get the reguired flow from the TP system, two Gilians
were operated in parallel. For the actual pumps used, pumping
speed and current measurements were made as a function of
backpressure. The results are given in Figures 6-1 and 6-2.
Similarly, flow vs superpressure measurements were made
through the Klippard valves and associated plumbing. The
results are given in Figure 6-3.
From these results, the tradeoffs influencing choice of
initial superpressure become apparent. At high superpressure,
pump speed decreases, and power consumption goes up, but so
does the rate at which air can be valved off. Hence the
pumpdown speed decreases, but the valve-up speed increases.
The reverse is true at low superpressure. If inadequate
superpressure is available to counteract condensation on the
balloon, then when all the available superpressure has been
valved off, the balloon will go slack while still negatively
buoyant. It will then begin descending, and will continue to
descend until it reaches the ground. On the other hand, if
the superpressure is too high, the maximum safe superpressure
70
-------
u
Q
m
CO
PI
e
e,
(0
•H
C5
O
3
s
LU
ir
LU
CCL
Q_
fl
O
•H
4->
Oi
CO
G
O
O
4->
c
0)
o
CM
D
•O
0)
G
•H
O
u
(I)
(0
CO
H
Cu
O
co
CM
CM
O
m
0) Cfl
M P
p CO
en M
•H 0)
fa ^
71
-------
§
S
nj
UJ
ca
o w>
3 uj
CE
0.
u
y
0}
CO
a)
M
Oi
u
4)
04
3
ca
M-l
O
c
o
c
3
(fl
CO
(Q
3
o
0)
(Q
4->
C
ro
I
vo
a)
U
3
D>
MOId
72
-------
may be exceeded when the PLT is swept up in a thermal, causing
the automatic cutdown system to activate to prevent the
balloon from failing in an uncontrolled manner. The
auto-cutdown provides a controlled descent, but also
decisively terminates the experiment.
b. Tower
In order to check the actual behavior of the testbed
prototype under pumping and valving against the equations
which describe its behavior, tests were conducted inside a
tower. The tower is part of the Solar Central Receiver Test
Facility at Sandia National Laboratories, Albuquerque (Figure
6-4). The tower offers an enclosed volume roughly 10 m square
by 52 m high. Since the tower is enclosed, it offers a more
controlled environment than does the ambient atmosphere,
making it easier to interpret test results.
Figure 6-5 is a diagram of the interior of the tower and
the various levels to which one has access. A monofilament
fishing line was run from the top to the bottom of the tower
at about the center of the shaft to act as a guide for the
balloon. The balloon was attached to the guideline by a
miniature carabiner which was free to move up and down the
line with very little friction. On one side of the tower, a
high door extends from ground level up to about 23 m. The
door is the width of the interior of the elevator shaft. A
significant air leak exists around the periphery of the door,
and the door itself acts as a thermal leak.
Figure 6-6 gives the approximate dimensions and reference
levels of the testbed prototype as used in these tests.
Initially, the ballonet was nearly completely filled with
helium. The mass of the payload was then increased with
73
-------
U
> H
•H J
0) CU
O
•
rH 4J
«tj O
M 4J
4-> O
G M
(1) Oi
U
-o
M 0)
O CO
CO 0)
O 0)
Jl tj
0^
3 4J
O CO
P
-------
SOLAR TOWER INTERIOR
LEVEL
LABEL
260
250-
240'
220-
200-
180-
160-
140-
120-
110-
77777 100-
80-
HEIGHT
OF
FLOOR (m)
51.82
48.77
• 42.67
• 36.58
-30.48
• 24.38
18.29
• 12.19
9.14
6.10 77777
HEIGHT
OF
REFERENCE
RAILING (m)
52.89
49.84
43.74
37.65
31.55
25.45
19.36
13.26
10.22
7.17
GROUND
LEVEL
Figure 6-5. Interior of the Solar Tower with reference levels
indicated.
75
-------
REFERENCE POINTS:
TOP-+1.44 m
MIDDLE-0
BOTTOM--1.44
PAYLOAD--2.51
Figure 6-6. Dimensions and Reference Levels on testbed
prototype PLT for Tower tests.
76
-------
ballast to the point that pressurizing the constant volume
balloon with air to about 40 mb yielded neutral buoyancy at
the floor level of the tower.
•
Subsequently, two valve-up and pumpdown cycles were
executed. Vertical position as a function of time was
recorded by recording the time when reference points on the
balloon passed reference levels in the tower. When the marker
balloon was moving slowly, it was possible to keep up with it,
moving from level to level, recording its passage. When it
was moving rapidly, it was not possible to record its passage
at every reference level. Hence the data points are unevenly
distributed. The time standard was a high-quality stopwatch.
Figure 6-7 shows valve-up number one. Note the much more
rapid rise in the upper portion of the tower than in the
lower. The balloon hit the top of the tower with considerable
excess buoyancy.
Figure 6-8 shows the results of pumpdown number one. Note
that it took nearly 500 seconds of pumping to overcome the
excess buoyancy generated during the preceding valve-up.
After the excess buoyancy had been overcome, the balloon began
to descend, slowly at first, and then more rapidly. The
apparent variations in descent rate are real. As noted on the
figure, the uncertainties are smaller than the data points.
During the pumpdown, the tethersonde system was taking
temperature data. Those data are given in Figure 6-10. Here
the altitude shown is from the tethersonde system itself,
which gets altitude by integrating the hydrostatic equation.
Note the grossly different thermal structure in the upper and
lower portions of the tower. The existence of the door which
provides a leak path for both air and heat undoubtedly
77
-------
500
TIME (s)
Figure 6-7. Valve-up number one. Uncertainties are smaller than
the data points. Valve was opened at A. From B to C. the mean
vertical velocity was 8.1 cm/s; from C to D. 24.9 cm/s.
78
-------
50
B
40
h-
I
g
HI
i
z
o
o
CD
30
20
10
500 1000
TIME(s)
1500
Figure 6-8. Pumpdown number one. Pumps were turned on at time
zero. It took the balloon system approximately 500 s to
overcome excess buoyancy, and to begin to descend (point A).
From A to B, the mean vertical velocity was -1.6 cm/s; from B
to C. -4.6 cm/s; from C to D, -2.9 cm/s; from D to E, -5.8
cm/s; and from E to F, -2.2 cm/s.
79
-------
DC
o
CO
50
LU ^ 40
CO £
LU ^
CC H 30
HO
-------
£«
CO
242
CC
a.
cc
^40
38
36
.••V
INTERNAL
TEMPERATURE
AMBIENT
TEMPERATURE
I
29.5
o
o
Ul
CC
29.0
CC
Ul
CL
S
Ul
H
28.5
500
1000
TIME (s)
1500
Figure 6-10. Internal temperature, ambient temperature, and
superpressure as a function of time on pumpdown number one.
Pumping began at t=0; descent at about 500 s.
81
-------
influences the thermal structure in the lower part of the
tower.
The internal balloon temperature, the external temperature
at the payload. and the superpressure -- the difference
between the internal and external pressures — were plotted
for pumpdown number one as a function of time. Those results
are given in Figure 6-10.
Note that when the balloon reached neutral buoyancy at the
top of the tower (came off the ceiling) the superpressure was
about 41 mb. When it reached the bottom of the tower, the
superpressure was about 43 mb. At the top of the tower, the
balloon was only 0.15° above ambient temperature, whereas at
the bottom, it was 0.9° above ambient. If the balloon were
brought to thermal equilibrium with its surroundings at the
base of the tower, the superpressure would have been very
nearly the same at the base as at the top of the tower, as
expected. The ambient pressure varies by about 5 mb over this
altitude range.
Next, valve-up number two was carried out. The results
are shown in Figure 6-11. This run was marred by the balloon
straying to one side of the tower and hitting the underside of
a gallery, in spite of the guideline (which had only minimal
tension on it, since it was a 10-pound test line). The
gallery it hit is immediately above the high door. It is
probable that a gentle flow caused by a convection cell
involving the door drew the balloon towards the door. The
tests were done at night when the external temperature on the
other side of the door was substantially cooler than the
interior temperature.
82
-------
50
40
I-
o
E30
x
-z.
o
o
<
CD
500
TIME (s)
Figure 6-11. Valve-up number two. Valve was opened at A.
From A to B, the mean vertical velocity was 6.4 cm/s. At B.
the balloon hit an overhang and bounced down. It tarried in
the vicinity of the overhang from B to C. From C to D, the
mean vertical velocity was 41.5 cm/s.
83
-------
The results from pumpdown number two are shown in Figure
6-12. Again, considerable excess buoyancy had been generated
on the way up. plastering the balloon against the ceiling of
the tower. It took nearly 700 seconds of pumping to overcome
that and start the balloon down. Again, the first few metres
were traversed rather slowly, followed by a much more rapid
descent to the vicinity of 20-22 m. Here the balloon
literally stopped. It didn't hit a solid obstruction — it
hit a thermal barrier. Once that barrier was overcome, it
dropped rapidly to the base of the shaft.
Figure 6-13 shows the results from the temperature
measurements made on the way down. The thermal barrier in the
20-22 m region is very clear.
One starts on the interpretation of the data with an
expression for the pumpdown speed in terms of the lapse rate
Y (Appendix I):
vp = - S T / [ V (Mg/R - Y) 1 (6-1)
Here S is pumping speed in m ; T is ambient temperature at
the balloon; V is the volume of the balloon; R is the
universal gas constant; M is the molecular weight of air; g is
the acceleration due to gravity; and Y is tne lapse rate
defined as Y * - dT/dz (here taken as positive if T
-4 3
decreases with altitude). We take S = 1.17 x 10 m /s
(7.0 I/rain); V = 12.5 m3; g = 9.8 m/s2. R = 8314
J/(kg-mol°K). For a standard atmosphere, this expression
agrees with that given in Equation (3-1). At T = 293°K. this
expression gives
84
-------
50 -
B
500
1000 1500
TIME (s)
2000
Figure 6-12. Pumpdown number two. Pumps were turned on at time
zero. It took the balloon system approximately 700 s to overcome
excess buoyancy and to begin to descend. From A to B, the mean
vertical velocity was -0.9 cm/s; from B to C. -9.2 cm/s; from C to
D, 0 cm/s; from D to E. -1.1 cm/s; and from E to F, -16.8 cm/s.
85
-------
g 50
CO
LLJ ^40
CO £
HI ^
D Z 3°
^ S 20
111 I
O.
5 10
LJJ
I-
0
h '-^..:-
• *\
\
•
/ *
' . • *
~ ^ X
yfc-L -
* X*.
/
/
•. •— •
• "*• __-^ ••"
— • ^* • ^^
Illll lltllllllll llti
28.0 28.5 29.0 29.5
TEMPERATURE (°C)
Figure 6-13. Temperature versus Altitude on pumpdown
number two.
86
-------
VP
(cm/s)
13.5
(adiabatic) 10 11.3
(standard) 6.5 9.9
(isothermal) 0 8.0
6.2
5.0
(observed in lower -34 4.0
part of tower on run #1)
It is important to emphasize that the pumpdown speed given
by the expression above is the speed with which the equilibrium
altitude changes, which is not necessarily the speed with which
the balloon actually moves. It would be the actual vertical
speed of the balloon in the absence of drag and dynamic effects
-- if the balloon system were massless. and if the air had 2ero
viscosity. In the presence of dynamic effects, the balloon
will not always be in equilibrium, and Newton's second law must
be used to determine its motion.
During both pumpdowns. initially the balloon falls very
slowly. This may be due in part to a dynamic effect -- it
requires some time to approach "terminal velocity" -- but it
might also be possible that at the very top of the tower, above
the region accessible to the tethersonde payload, there is a
strong inversion that accounts for the slow pumpdown.
Below the first few metres, there is a region with a
positive lapse rate during both pumpdowns. In the first
pumpdown. this region is traversed at about 4.6 cm/s, whereas
in the second, the speed is 9.2 cm/s. Both are slower than the
10-20 K/km lapse rate would lead one to expect on the basis of
the pumpdown speed expression. The thermal structure does not
seem to be able to account for the difference, although
87
-------
differences in thermal structure smaller than can be detected
by the tethersonde can make a significant difference in
pumpdown speed. Another possibility is that vertical air
flows of a few cm/s are present and are complicating the
dynamics.
In the lower part of the tower, during pumpdown number
one. the vertical velocity observed is in reasonable agreement
with the predicted pumpdown velocity. In pumpdown number two.
however, the thermal barrier introduces a delay, followed by
an accelerated fall rate -- a dynamic effect.
The dynamic effects were examined in a simplified
fashion. Ultimately, a computer model will be needed, but an
approximate analytical treatment yields considerable insight.
Newton's second law can be written:
(6-2)
_»
where F is the force on a system, m is the mass of the system.
and v is its velocity. For the balloon system, we may write
-g [V (Pb-Pa) + m] k - PaCDAD|vb- va|(vb- va)
1 dv
[pbv + m H- i pav] ^
(6-3)
Here the first term is the buoyant force, the second the drag
force, and the quantity in brackets on the right side of the
equation is the "virtual mass" of the balloon system. Here
p, is the mean density of the gas in the balloon, p
O 3
is the density of the ambient air. m is the mass of the
-X
balloon system excluding the mass of the gas in the balloon, k
88
-------
is a vertical unit vector. C_ is the drag coefficient (taken
as 0.47 for low Reynolds number); AD is the cross-sectional
area of the balloon, v. is the velocity of the balloon, v
t> a
is the velocity of the air. and the term (l/2)p V on the
3
right side of the equation accounts for an increase in
effective mass of the system due to entrainment of air.
If we choose to make the equation specific for the system
heading downwards, we can drop the vector notation:
,2 dv
-mLg + bv = mv ^
(6-4)
where here mr is the net (buoyant) mass of the system
Li
mL E V (pb - pa) + m (6-5)
(here presumed constant), b is a redefined drag coefficient.
and m is the virtual (inertial) mass of the system. Now v
is the velocity relative to the air. which is presumed to be
either quiescent or in uniform (non-accelerated) motion.
The solution to this equation, starting from rest relative
to the air, is
v = -a tanh fit (6-6)
where a can be shown to be the terminal velocity, and is
given by:
a =
\i r\
(6-7)
and
(6-8)
89
-------
This solution implies a characteristic relaxation time t
to reach (1 - 1/e) of terminal velocity a:
T ' nir g
(6-9)
For our system.
2.3 _ 5.6
\^L a (6-10)
The relaxation times can be quite significant. For a =
0.1 m/s (10 cm/s). T = 56 s. For smaller terminal
velocities, the relaxation times can be very long indeed.
When the pump or valve is operated with the system neutrally
buoyant, initially m.. =0. so T is infinite. As m
departs from zero. T becomes shorter.
The complexities of the dynamics and of the tower
environment make it difficult to confirm our understanding of
the system in detail, but the observations are consistent with
the analysis to within the uncertainties induced by the system
dynamics. A better model, taking the dynamics into account
more accurately, is clearly in order. Nonetheless, the tower
tests confirm the ability to adjust the equilibrium altitude
of the system at the required rate -- 10 cm/s -- in a standard
atmosphere.
c. Ambient Atmosphere
In Phase I. the ambient atmosphere tests were of two
types. The first is closely analogous to the tower tests, in
that the testbed prototype was constrained by a guideline.
only now the guideline is the tetherline of an
aerodynamically-shaped tethered balloon (Figure 6-14). We
90
-------
*s»
Figure 6-14. Testbed Prototype PLT during "open air tower"
tests, viewed from the solar tower. Here the TP rides up and
down on the tetherline of a tethered balloon out of the field
of view above.
91
-------
Figure 6-15. Testbed prototype PLT during a slack-tether
flight test.
92
-------
call these "open air tower" tests. This permits greater
vertical excursions, and exposure to the outdoor environment.
At very low wind speed, this arrangement approximates a
free-flight test. The balloon is free to respond to relative
vertical air motions, and to heating or cooling caused by
clouds and night sky. The great advantage is that the balloon
remains in place laterally, and testing can go on for hours
without having to chase the balloon. However, horizontal air
motion exerts forces on the balloon which are not present in
free flight. Consequently, the results of these tests need to
be interpreted with this in mind.
In the second type of test, the balloon is prevented from
floating away by a lightweight tether. The tether is kept
slack during a flight by keeping a chase vehicle under the
balloon, and by reeling out or reeling in tether as
appropriate (Figure 6-15). This is still not a perfect
approximation to free flight, but it is an improvement.
Unfortunately, the duration of the flights obtainable in this
manner are limited by where the chase vehicle can go to follow
the balloon. In Phase II. true free-flight testing will begin.
The outdoor tests further qualitatively confirmed that the
system behaves as expected. The outdoor environment is so
complex, however, that it became clear that in the absence of
long free-flight tests, quantitative understanding of system*
behavior could only come from the tower.
7. FUTURE WORK
There are two major thrusts in Phase II. The first is to
turn the design for the operational prototype flight system
93
-------
into functional hardware. The second is to continue
development and refinement of the hybrid control algorithm
through a continuing test program, at first with the testbed
prototype, but later with the operational prototype. This
involves modifying the prototype ground station to accommodate
a radiotheodolite or LORAN tracking system, an ARGOS uplink
receiver, and telephone reception of ARGOS data from NOAA. It
does not involve significantly upgrading the ground station
computer facility, or the incorporation of a Local User
Terminal to permit reception of real time data from the PLT
via satellite. In Phase II. ARGOS data can either be directly
received from a nearby PLT via the uplink receiver, or it can
be received via Service ARGOS/NOAA with a 4-6 hour delay.
Because of resource constraints, the GSS described in Section
4 will not be built during Phase II. Nonetheless, at the end
of Phase II, an operational prototype system will exist which
can be used in both long and short range field studies.
For applications of the PLT involving more than a few
single-balloon experiments, it would be cost-effective to
proceed to Phase III: Building the Ground Support Station as
described in Section 4, and carrying out systems tests with
commercially-produced PLT payloads. Routine or major use of
the PLT system will require a commercial source for PLT
payloads. and the capabilities represented by the GSS.
Parallel to the systems development work described above,
we are also tasked to conceive, develop, and propose
appropriate applications for the balloon tracer system. Only
one will be discussed here.
To date, most thought has been given to using the tracer
to investigate long range transport -- specifically, source-
receptor relationships, intrinsic limits on predictability.
94
-------
and transport model verification. These applications are
indeed important, but less thought has been given to use of
the system in the investigation of the chemical and physical
transformations relevant to long range air pollution. In such
studies, it would be extremely helpful to be able to follow
the processes in a Lagrangian frame. It has long been
recognized that such a frame is a natural choice for
investigation of pollutant behavior (Seinfeld et al, 1973).
but only a few such studies have in fact been done (Zak.
1983). This is the case because the technology has not been
available to support Lagrangian studies. By the end of Phase
II, it will be.
For example, once the operational PLT exists, it would be
quite simple to put a transponder on board which would permit
an aircraft in the area to range on the balloon -- perhaps
using an inexpensive distance-finding system such as that
manufactured by Meeda Instrumentation (Meeda. no date) for use
as a balloon radio altimeter, or alternately, using an
airborne radar.
Once having a means of measuring the distance to the
balloon, an aircraft instrumented to investigate homogenous
and heterogenous processes could execute a standard circular
path relative to it. perhaps at more than one altitude, every
few hours for up to three days (Figure 7-1).
If the balloon had been injected into the plume from a
major source -- urban or industrial -- this would provide a
means of observing the time-dependent behavior of the source
effluents over an extended period of time. Alternately,
studies could be done in areas with low background
contamination, but with an intentional release of a selected
95
-------
Figure 7-1. Lagrangian experiment instrumented-aircraft
measurement pattern. PLT marks mean flow; measurement
aircraft periodically executes pattern relative to marker at
2-4 altitudes.
96
-------
mix of pollutants around the balloon. The subsequent behavior
of those effluents could be observed with instrumented
aircraft as well, albeit for a shorter period. The PLT would
permit either type of experiment to be done in cloud, perhaps
even in rain, as well as in clear air. To some of us
concerned with acid deposition, this is an exciting prospect.
8. CONCLUSIONS AND DISCUSSION
Now, in light of our theoretical analysis, the experience
we have gained in designing and building the testbed prototype
PLT. and the experimental results we have obtained with it. we
go back and examine the design goals set forth in Section 2.
• Operates under the exemption clauses of FAR 101.
FAR 101 permits up to 12 Ibs (5.45 kg) of payload. as long as
the payload is distributed in packages weighing no more than 6
Ibs (2.73 kg) each, which meet the 3 oz/in areal density
limit. The current testbed prototype payload weighs about
2.36 kg including batteries, and meets the areal density
limits. The operational PLT payload could be twice as large.
and as long as care were exercised in its design, it would
still satisfy the exemption provisions of FAR 101.
• Lifetime >. 3 days.
The pair of Gilian pumps which yield adequate pumpdown speed
consume 2-2.5 watts total power when pumping. If the pumps
operated 50% of the time during a 72-hour experiment, they
would consume 90 watt hours of energy. At lithium battery
energy densities of 100-200 watt hrs/lb, this would take less
than one pound of batteries. Even with three pumps operating
97
-------
more than 50% of the time, the battery weight required would
likely be acceptable. Thus, as long as the remainder of the
payload is designed so that it uses no more than the pumps.
power should not be a problem for 3 day'experiments.
Tracking range > 1000 km in the northeast
quadrant of the United States.
ARGOS tracks worldwide.
• Telemetry of relative vertical air motion.
pressure, temperature, and humidity.
ARGOS data handling can easily accommodate these and the
several other parameters it would be desirable to transmit. If
continuous data are needed, a regional/hybrid system can
provide it at additional cost for a network of ground stations
equipped with ARGOS uplink receivers.
• Ground system capable of handling several
PLTs" at a time.
ARGOS can handle 200 PTTs in its field of view simultaneously.
• Capable of establishing specified ascent
and descent rates under radio command.
A 2-site. 2-band HF radio command system should offer reliable
command capability. Command technology is well developed.
• Capable of reaching altitudes up to 500 mb (5.5 km).
Our calculations indicate that would not be a problem for the
testbed prototype PLT. As long as the balloon is properly
98
-------
designed and filled for the desired ceiling altitude, and the
operational PLT payload weight, it should not be a problem.
• Capable of following mean vertical flows as
low as 1 cm/s with "acceptable" fidelity.
Under standard atmosphere conditions, we've shown theoretically
that a trajectory based on potential temperature meets that
specification (Appendix G). We've shown that the testbed
prototype buoyancy control system is capable of handling mean
vertical flows as high as 10 cm/s on average, which equals or
exceeds expected long-term average vertical air speeds, and
that higher speeds can be accommodated. However, in the mixed
layer, potential temperature is not a good control parameter.
During convective mixing, it would be necessary to go to some
other control strategy. We also put forward an argument that
in a mixed layer, because of the rapid spread of effluents
throughout the layer, the exact position of the balloon tracer
within the layer is not important. The conclusion is that this
design goal can be met most of the time, and when it cannot
with the potential temperature control strategy because of
mixing, it doesn't matter. This point, and the meaning of
"acceptable" fidelity, are discussed in detail in Appendix J.
• Sufficiently inexpensive to permit use in
significant numbers on an expendable basis.
This goal is somewhat elastic, in that it depends upon the
budgetary constraints in existence at a given time. We choose
to address it by considering the likely cost of the elements of
the operational PLT in small numbers (1-10) and in large (>50).
99
-------
Cost Elements Cost ($K)*
1-10 (ea) >50 (ea)
CVB with Ballonet 1.5 1.0
Buoyancy Adjustment Subsystem 0.5 0.3
Sensors 1.2 0.7
Microcomputer 1.0 0.7
ARGOS Telemetry 1.3 0.8
Command Receiver and Decoder 0.7 0.5
Tracking Aids, including
FAA Transponder 1.3 0.9
Batteries 0.1 0.1
Assembly Labor l.Q 0.5
$8.6K $5.5K
*CY84 dollars.
The costs in quantity are approximate and merely project
typical quantity procurement savings. Assuming these for the
near future, one can make a reasonable estimate of the cost per
use. V. E. Lally indicates that balloon payloads carrying a
simple message to return to a stated address have better than a
50% return rate in the Continental U.S. If a reward is offered
for return, an even higher return rate could be obtained. We
ignore that to allow for damage on the returned payloads. On
the other hand, the balloon itself must be considered expended
on each use. With these assumptions, in quantity, the PLTs
would cost approximately $3.2 K per use. This cost does not
seem to prohibit use in significant numbers on an "expendable"
basis. For the longer term, more dramatic cost savings are
likely to come about if continual use occurs, much as they have
for tethersonde payloads. which underwent a factor of three
reduction in cost over several years.
In summary, the authors are confident in concluding that an
adjustable buoyancy physical Lagrangian tracer meeting the
100
-------
stated design goals is both technically and economically
feasible. We are proceeding in Phase II to turn that
conviction into operational hardware.
101
-------
REFERENCES
Barat. J. (1982). "Initial Results from the Use of Ionic
Anemometers Under Stratospheric Balloons: Applications to the
High-Resolution Analysis of Stratospheric Motions." J. Appl.
Meteor. 2.1. 1489.
Blamont, J., Heinsheimer. T.. and Pommereau. J. (1974). "New
Method of Study of the Dynamics of the Stratosphere -
Principle and First Results." Academic des Sciences (Paris).
Comptes Rendus (B) 228. 249.
Boas. M. L. (1966). Mathematical Methods in the Physical
Sciences, John Wiley and Sons. New York.
Chiang. T., Ossin, A.. Tien. C. L. (1964). "Laminar Free
Convection from a Sphere." J. Heat Transfer. 537.
Danielson. E.F. (1961). "Trajectories: Isobaric. Isentropic.
and Actual," J. Meteor. 18. 479.
Gay. G.T.. Zak. B.D., Barker. B.. Holland. R.M.. and Homann.
P.S. (1981). Lagranqian Measurement Platform Flights in
Support of the Tennessee Plume Study: Field Effort and Data.
Sandia National Laboratories Report SAND79-1336. Albuquerque.
NM.
Gill. G.C. (1979). Personal communication with V.E. Lally.
Holton. J.R. (1979). An Introduction to Dynamic Meteorology.
Second Edition. Academic Press. New York. NY.
Lally. V.E. (1967). "Superpressure Balloons," in Scientific
Ballooning Handbook, A. L. Morris. Ed.. National Center for
Atmospheric Research Technical Note NCAR-TN/IA-99. Boulder, CO.
Lally. V.E. (1975). Superpressure Balloons for Horizontal
Soundings of the Atmosphere. National Center for Atmospheric
Research Technical Note NCAR-TN-28. Boulder. CO.
Lally. V. E. (1985). Personal communication.
Lamb, R. G. (1984). Air Pollution Models as Descriptors of
Cause-Effect Relationships,". Atmos. Environ. 18. 591.
Lichfield, E. (1981) "Tracking and Communication in Long
Duration Flights." COSPAR Session on Scientific Ballooning.
102
-------
MacCready, P. B., Baboolal. L. A.. Lissamen. P.B.S. (1974).
"Diffusion and Turbulence Aloft over Complex Terrain," in
preprint volume. Symposium on Atmospheric Diffusion and Air
Pollution. Santa Barbara, CA, Sept. 9-13. Amer. Meteor. Soc.
218-225.
MacCready. P.B.. and Mullen. J.B. (1977). Turbulence
Investigation on DaVinci II, AeroVironment Report No. AVFR
7141. Pasadena, CA.
MacCready. P.B. (1981). "Turbulence and the Local Flow
Field. The AV Experiment on DaVinci I", in Final Report on
Project DaVinci: A Study of Long Range Air Pollution Using A
Balloon-Borne Laqrangian Measurement Platform. Vol. 2: Reports
of Participants in DaVinci I. Sandia National Laboratories
Report SAND78-0403/2. Albuquerque. NM.
Margolin, B. (1982). "Paper Thin Lithium Battery Powers
Handheld Electronics." Electronic Products, March 26, 1982.
Meeda Scientific Instrumentation (no date). "Series 200
Balloon-Borne Radio Altimeter." Romat-Gan. Israel.
Morris. A.L. (1975). Scientific Ballooning Handbook. National
Center for Atmospheric Research Technical Note NCAR-TN/IA-99.
Boulder, CO.
Morris, A.L. and Solot, S.B. (1975). "The Atmosphere," in
Scientific Ballooning Handbook, A. L. Morris. Ed., National
Center for Atmospheric Research Technical Note NCAR-TN/IA-99.
Boulder. CO.
NCAR, Atmospheric Technology, periodical published by the
National Center for Atmospheric Research. Boulder, CO.
Reif, F. (1965). Fundamentals of Statistical and Thermal
Physics. McGraw-Hill. New York.
Schlichting, H. (1979). Boundary Layer Theory. Seventh
Edition. Translated by J. Kestin. McGraw-Hill. New York, p."
238.
Seinfeld. J.H.. Hecht. T.A.. and Roth. P.M. (1973). Existing
Needs in the Observational Study of Atmospheric Chemical
Reactions. U.S.E.P.A. Report EPA-R4-73-031.
Tatom, F.B. and King, R.L. (1977). Constant Volume Balloon
Capabilities for Aeronautical Research. NASA Contractor Report
NASA CR-2805. Huntsville. AL.
103
-------
Wallace, J.M. and Hobbs. P.V. (1977). Atmospheric Science, an
Introductory Survey. Academic Press. New York. NY.
Zak, B.D. (1981a). Final Report on Project DaVinci: A Study
of Long Range Air Pollution Using 'A Balloon-Borne Lagrangian
Measurement Platform. Vol. 1; Overview and Data Analysis.
Sandia National Laboratories Report SAND78-0403/1.
Albuquerque. NM.
Zak. B.D. (1981b). "Lagrangian Measurements of Sulfur Dioxide
to Sulfate Conversion Rates." Atmos. Environ. 15, 2583.
Zak, B.D. (1983). "Lagrangian Studies of Atmospheric
Pollutant Transformations,'" in Trace Atmospheric
Constituents: Properties. Transformations, and Fates. Volume
12 in series. Advances in Environmental Science and
Technology. S.E. Schwartz, Ed., John Wiley and Sons. New
York, NY.
104
-------
APPENDIX A
FEDERAL AVIATION REGULATIONS PART 101:
MOORED BALLOONS. KITES. UNMANNED ROCKETS,
AND UNMANNED FREE BALLOONS
The following is a reprint of the FAA regulation most
relevant to the adjustable buoyancy balloon.
105
-------
Part 101—Moored Balloons, Kites, Unmanned Rockets,
and Unmanned Free Balloons
Subpart A—General
1101.1 Applicability.
(a) This Part prescribes rules governing
the operation, in the United States, of the
following:
(1) Except as provided for in §101.7 of
this Part, any balloon that is moored to the
surface of the earth or an object thereon
and that lias a diameter of more than 6
feet or a gas capacity of more than 115
cubic feet.
(2) Except as provided for in § 101.7 of
this Part, any kite that weighs more than
5 pounds and is intended to be flown at
the end of a rope or cable.
(3) Any unmanned rocket except—
(i) Aerial firework displays; and
(ii) Model rockets—
(a) Using not more than 4 ounces
of propellant;
(b) Using a slow-burning propellant;
(
-------
MOOBED BALLOONS, KITES, UNMANNED ROCKETS, AND UNMANNED FREE BALLOONS PART 101
operating a moored balloon or kite within a
restricted area must comply only with §101.19
and with additional limitations imposed by the
using or controlling agency, as appropriate.
1101.13 Operating limitations.
(a) Except as provided in paragraph (b) of
this section, no person may operate a moored
balloon or kite—
(1) Less than 500 feet from the base of
any cloud;
(2) More than 500 feet above the surface
of the earth ;
(3) From an area where the ground
visibility is less than three miles; or
(4) Within five miles of the boundary of
any airport.
(b) Paragraph (a) of this section does not
apply to the operation of a balloon or kite
below the top of any structure and within 250
feet of it, if that shielded operation does not
obscure any lighting on the structure.
1101.15 Notice requirements.
No person may operate an unshielded moored
balloon or kite more than 150 feet above the
surface of the earth unless, at least 24 hours
before beginning the operation, he gives the
following information to the FAA ATC
facility that is nearest to the place of intended
operation:
(a) The names and addresses of the
owners and operators.
(b) The size of the balloon or the size and
weight of the kite.
(c) The location of the operation.
(d) The height above the surface of the
earth at which the balloon or kite is to be
operated.
(e) The date, time, and duration of the
operation.
1101.17 Lighting and marking requirements.
(a) No person may operate a moored bal-
loon or kite [between sunset and sunrise] un-
less the balloon or kite, and its mooring lines,
are lighted so as to give a visual warning equal
to that required for obstructions to air navi-
gation in the FAA publication "Obstruction
Marking and Lighting".
(b) No person may operate a moored bal-
loon or kite [between sunrise and sunset] un-
less its mooring lines have colored pennants or
streamers attached at not more than 50-foot
intervals beginning at 150 feet above the sur-
face of the earth and visible for at least one
mile.
1101.19 Rapid deflation device.
No person may operate a moored balloon
unless it has a device that trill automatically
and rapidly deflate the bclloon if it escapes
from its moorings. If the device does not
function properly, the operator shall im-
mediately notify the nearest ATC facility of
the location and time of the escape and the
estimated flight path of the balloon.
Subport C—Unmanned Rockets
1101.21 Applicability.
This subpart applies to the operation of un-
manned rockets. However, a person operating
an unmanned rocket within a restricted area
must comply only with subparagraph 101.23
(g) and with additional limitations imposed
by the using or controlling agency, as appro-
priate.
1101.23 Operating limitation!.
No person may operate an unmanned
rocket—
(a) In a manner that creates a collision
hazard with other aircraft;
(b) In controlled airspace;
(c) Within five miles of the boundary of
any airport;
(d) At any altitude where clouds or ob-
scuring phenomena of more than five-tenths
coverage prevails;
(e) At any altitude where the horizontal
visibility is less than five miles;
(f) Into any cloud;
(g) Within 1,500 feet of any person or
property that is not associated with the op-
erations; or
(h) [Between sunset and sunrise.]
1101.25 Notice requirements.
No person may operate an unmanned rocket
unless, within 24 to 48 hours before beginning
the operation, he gives the following :'nfo.ma-
Oi. 1 (Amdt. 101-4, M. 1/20/74)
107
-------
PXHT 101 HOOKED BALLOONS, KITES. UNMANNED ROCKETS, ANI> UNMANNED FREE BALLOONS
3
turn to the FAA ATC facility that is nearest
to the place of intended operation:
(a) The names and addresses of the
operators.
(b) The number of rockets to be operated.
(c) The size and weight of each rocket.
(d) The maximum altitude to which each
rocket will be operated.
(e) The location of the operation.
(f) The date, time, and duration of the
operation.
(g) Any other pertinent information
requested by the ATC facility.
Subpart D—Unmanned Free Balloons
1101.31 Applicability.
This subpart applies to the operation of
unmanned free balloons. However, a person
operating an unmanned free balloon within a
restricted area must comply only with § 101.33
(d) and (e) and with any additional limita-
tions that are imposed by the using or control-
ling agency, as appropriate.
I 101.33 Operating limitation*.
No person may operate an unmanned free
balloon—
(a) Unless otherwise authorized by ATC,
in a control zone below 2,000 feet above the
surface, or in an airport traffic area;
(b) At any altitude where there are
clouds or obscuring phenomena of more than
five-tenths coverage,
(c) At any altitude below 60,000 feet
standard pressure altitude where the hori-
zontal visibility is less than five miles;
(d) During the first 1,000 feet of ascent,
over a congested area of a city, town or set-
tlement or an open-air assembly of persons
not associated with the operation; or
(e) In such a manner that impact of the
balloon, or part thereof including its pay-
load, with the surface creates a hazard to
persons or property not associated with the
operation.
Oi. 1 (Am*. 101-4, til. 1/20/741
I 101.35 Equipment and marking require-
ment*.
(a) No person may operate an unmanned
free balloon unless—
(1) It is equipped with at least two pay-
load cut-down systems or devices that oper-
ate independently of each other;
(2) At least two methods, systems, de-
vices, or combinations thereof, that function
independently of each other are employed for
terminating the Sight of the balloon enve-
lope ; and
(3) The balloon envelpe is equipped
with a radar reflective device (s) or material
that will present an echo to surface radar
operating in the 200 MHz to 2700 MHz
frequency range.
The operator shall activate the appropriate
devices required by subparagraphs (1) and (2)
of this paragraph when weather conditions are
less than those prescribed for operation under
this subpart, or if a malfunction or any other
reason makes the further operation hazardous
to other air traffic or to persons and property
on the surface.
[(b) No person may operate an unmanned
free balloon below 60,000 feet standard pres-
sure altitude between sunset and sunrise (as
corrected to the altitude of operation) unless
the balloon and its attachments and payload,
whether or not they become separated during
the operation, are equipped with lights that are
visible for at least 5 miles and have a flash
frequency of at least 40, and not more than 100,
cycles per minute.]
(c) No person may operate an unmanned
free balloon that is equipped with a trailing
antenna, that requires an impact force of more
than 50 pounds to break it at any point, unless
the antenna has colored pennants or streamers
that are attached at not more than 50-foot in-
tervals and that are visible for at least one
mile.
(d) No person may operate [between sunrise
and sunset] an unmanned free balloon that is
equipped with a suspension device (other than
a highly conspicuously colored open para-
chute) more than 50 feet long, ^nle s the
108
-------
MOORED BALLOONS, KITES, UNMANNED ROCKETS, AND UNMANNED FREE BALLOONS PART 101
suspension device is colored in alternate bands
of high conspicuity colors or has colored pen-
nants or streamers attached which are visible
for at least one mile.
t 101.37 Notice requirements.
(a) Prelaunch notice. Except as provided
in paragraph (b) of this section, no person
may operate an unmanned free balloon unless,
within 6 to 24 hours before beginning the
operation, he gives the following information
to the FAA ATC facility that is nearest to the
place of intended operation:
(1) The balloon identification.
(2) The estimated date and time of
launching, amended as necessary to remain
within plus or minus 30 minutes.
(3) The location of the launching site.
(4) The cruising altitude.
(5) The forecast trajectory and esti-
mated time to cruising altitude or 60,000
feet standard pressure altitude, whichever
is lower.
(6) The length and diameter of the bal-
loon, length of the suspension device, weight
of the payload, and length of the trailing
antenna.
(7) The duration of flight.
(8) The forecast time and location of im-
pact with the surface of the earth.
(b) For solar or cosmic disturbance in-
vestigations involving a critical time element,
the information in paragraph (a) of this sec-
tion shall be given within 30 minutes to 24
hours before beginning the operation.
(c) CanceUatior. notice. If the operation
is canceled, the person who intended to con-
duct the operation shall immediately notify
the nearest FAA ATC facility.
(d) Launch notice. Each person operating
an unmanned free balloon shall notify the
nearest FAA or military ATC facility of the
launch time immediately after the balloon is
launched.
S 101.39 Balloon position reports.
(a) Each person operating an unmanned
free balloon shall—
(1) Unless ATC requires otherwise, mon-
itor the course of the balloon and record its
position at least even- two hours; and
(2) Forward any balloon position re-
ports requested by ATC.
(b) One hour before beginning descent,
each person operating an unmanned free bal-
loon shall forward to the nearest FAA ATC
facility the following information regarding
the balloon:
(1) The current geographical position.
(2) The altitude.
(3) The forecast time of penetration of
60,000 feet standard pressure altitude (if
applicable).
(4) The forecast trajectory for the bal-
ance of the flight.
(5) The forecast time and location of
impact with the surface of the earth.
(c) If a balloon position report is not
recorded for any two-hour period of flight,
the person operating an unmanned free balloon
shall immediately notify the nearest FAA
ATC facility. The notice shall include the
last recorded position and any revision of the
forecast trajectory. The nearest FAA ATC
facility shall be notified immediately when
tracking of the balloon is re-established.
(d) Each person operating an unmanned
free balloon shall notify the nearest FAA
ATC facility when the operation is ended.
* u i arawor mmM ofnct UTT. 237-709/499
Ch. 1
109
-------
by:
APPENDIX B
PUMPDOWN SPEED IN A STANDARD ATMOSPHERE
The net buoyant force F, on the balloon system is given
Fb = tVpa ~ Vpb ~ m] g (
where V = volume of balloon system (m3)
(presumed constant).
pa = density of the ambient air at the
altitude of the system (kg/m3).
Pb = 'average density of gas contained in
balloon.
m = mass of balloon system exclusive of gas
in balloon.
g = acceleration due to gravity.
At the system's equilibrium altitude, the net buoyant
force is zero, so:
Vpa = Vpb + m (B-2)
If we pump air into the superpressure balloon for time dt. we
add to the contained mass Sp dt (where S is the pumping
3
speed in m /s), and hence change its density, so
Spa
dpb • IT dt (B-3)
But if we take the differential of Equation B2, we see
dpa = dpb (B-4)
Spa
dpa = ~- dt (B-5)
110
-------
Assuming a standard atmosphere below the tropopause
(Morris. 1975):
Mg. _,
- aR
where
Po
M
a
R
Mg.
aR
(B-6)
1.2250 kg/m3.
28.96 kg/(kg - mol). average molecular
weight of air.
9.807 m/s2.
6.5 x 10~3 K/m. standard lapse rate.
8314.3 J/ K (kg - mol). universal gas
constant; if pressures expressed in mb.
then 83.143.
5.255.
288.15 K.
height in m MSL. or more strictly,
"geopotential meters'.1.
we can evaluate dpa:
aR
„.„ -I
aR
dz
(B-7)
Substituting this expression into Equation B5. and solving for
dz/dt. we find:
or
dz_ = -1 1
dt (a/TQ) (Mg/aR - 1) V
(B-8)
V = dz = -1.042 X 104 J (1 - 2.256 X 10 5z)
P dt V (B-9)
111
-------
APPENDIX C
SUPERPRESSURE AS A FUNCTION
OF EQUILIBRIUM ALTITUDE
By definition, superpressure is given by
Ps ' Pb - Pa (C-V
where P, is the absolute pressure in the balloon, and P
O 3
is the ambient pressure of the air at the balloon altitude.
P,(z) is given by the gas law coupled with the
expression for temperature variation in a standard atmosphere:
Pb " M RT = V* RT - ^ RTo(1 - I Z) (C-2)
b o
where n^ = total number of fcg-moles of gas in the
balloon (air and helium).
Mfc = average molecular weight of gas in balloon.
T = temperature in K.
In a standard atmosphere.
Ma
Pa - po <1 - T0 Z)aR (C-3)
Here P = 1013 millibars or 1.013 x 105 N/m2
o
In order to evaluate P . we need n, (z).
S D
We have:
nb = na + nHe, (C-4)
112
-------
na = kg-moles of air in balloon,
nHe = kg-moles of helium in balloon.
In valving and pumping operations, only the quantity of air in
the balloon is affected, not the helium. Hence,
differentiating Equation C-4:
dn, dn
b _ a re* 51
we can obtain n from the buoyancy condition. Equation B-2,
if we note that
Vp, = n M + n It. (C-6)
b a He Tie
where
Mjje = Molecular weight of helium
(4 kg/[kg - mol]).
Then we find:
+ m (C-7)
If we differentiate this expression:
dp dn
y a _ jyr O f C— R\
Thus
dz~ =dz~ =M~dz~ (C-9)
and
n, (z) = ^ Pa(z) + C (C-10)
D Ma
where C is an integration constant to be determined from a
boundary condition. Combining Equations Cl. C2 and C3 with 10:
113
-------
—o _ a 2,fv A _ a_ _\aR L C1
v <1 TO 2>|_M "o^1 TO zj * CJ -
_ a W
TQ «)
\ C — 11 }
Expanding and collecting terms:
RT o CRT
o
The quantity in the first parenthesis is zero by the gas law.
Hence:
CRT
Ps = —° (1 - | z) (C-13)
o
To express pressures in millibar, we take R = 83.14;
then
P = 2.396 X 104 id- 2.256 X 10~5z) (C-14)
s v
and
dP
_s = -.541 C
dz V
Using the expression for PSt one can evaluate C.
VP
C =
2.396 X 104 (1 - 2.256 X 10 5 z ) ,_ ..
O (C—ID
where P is the initial superpressure at some initial
s o
height ZQ.
114
-------
APPENDIX D
EFFECT OF TEMPERATURE SWING ON BALLOON PRESSURE
We may calculate the effect of a swing in balloon
temperature caused by a change in the radiation environment
from the gas law and Equation CIO:
n.RT
Pb = -^r- (D-l)
n.R
Apb = ~v~ AT (D-2)
* §]
(0-4,
T _t;A?e;t; n ~\
AP = 3.52 (1 - 2.256 X 10 3z) + 83.14 ~ AT (D-5)
115
-------
APPENDIX E
ENERGY REQUIRED FOR PUMPDOWN
The incremental work done in bringing about a change in
volume of gas dV is:
dW = PdV (E-l)
= P ^ dt (E-2)
for our system P = PSj but Ps must be expressed in
newtons/m2, where
1 mb = 100 N/m2 (E-3)
We interpret dV/dt as the pumping speed S. Then, choosing to
keep Ps in mb, we have
dW = 100 Ps S dt (E-4)
or
dW = 100 PS dt
dz s dz
finally
= 100 (E-6)
where Vp is the pumpdown speed. Combining with Equations B9
and C14, we obtain
HW 7
= 2.30 X 10 C (E-7)
116
-------
Using Equation C14 to evaluate C. assuming some specified
superpressure P at a specified altitude z . we find
SO O
.„ 9.60 X 10~3 P V
dW = so
dZ (1 - 2.256 X 10~5ZQ) (E-8)
117
-------
APPENDIX F
LIFT AND MAXIMUM ALTITUDE
During filling, initially one puts helium into the inner
balloon. The lift L generated is
t = v. cpa - pHe) , (F_1)
where p and p are the densities of air and helium
ct rl6
respectively at the fill altitude, and V is the volume of
helium put into the balloon, and hence the (slack) volume of
the balloon. Note that
naM nHe**He
pa = ~V^ and pHe = ~ (F-2)
where n, and n.. are the number of moles of air displaced,
a tie
and the number of moles of helium occupying the volume v
respectively, and M and M^ are the molecular weights of air
rl6
and helium respectively. However, since the balloon is slack.
and the helium is presumed to be at ambient temperature.
na = nHe ' (F-3)
Thus
L = nHe (M - MHe)g (P-4)
So the lift from a slack (zero superpressure) balloon in
thermal equilibrium with the ambient air is a function only of
the number of moles of helium it contains.
One needs to put in enough helium to generate lift not
only to carry the combined mass m of the balloon and payload.
but also to carry the weight of excess air required to
generate the desired amount of superpressure P at the
s
surface where balloon fill is taking place. Thus, the
118
-------
required lift is given by
LR = (m + An M)g (F_5)
where An is the number of moles of excess air. The term
An can be obtained from application of the gas law. When
the balloon is superpressured. the volume of the rigid balloon
becomes V. and:
so.
PSV - An RT (F-6)
PsV
LR= m + M — g (F_?)
Thus, to fill the balloon properly, one first puts in
helium until the gross lift is measured to be L., as given by
K
(F-7). Then, one seals the inner balloon, and inflates the
outer balloon with air, until the desired superpressure P
s
is reached. At that point, the balloon system, after
attachment of the payload. should be approximately neutrally
buoyant.
The question next arises as to what the ceiling altitude
of this balloon system is. At maximum altitude, all of the
air ballast will have been vented, leaving the balloon
entirely filled with helium. Thus, at maximum altitude, the
equilibrium condition becomes
Vpa = VpHe + m (F-8)
but
= nHeMHe (F-9)
119
-------
Here njje can be obtained by equating lift with required
lift. From (F-4) and (F-7):
nHe (« - MHe) = m + M
RT
(F-10)
m + M
n
RT
He
(F-ll)
So (F-8) becomes
n
He
(F-12)
However, in a standard atmosphere (Equation B-6).
p = 1.2250 (1 - 2.256 X 10 5 z)4'255
3
(F-13)
Substituting (F-13) into (F-12). one obtains
, -> •>** v in" , v.
1 - 2.256 X 10 2m)
I1HeMHe'1'
1.255 V
where here z = maximum, obtainable .height.
m
(F-14)
1 - 2.256 x 10 5 z =
m
1.255 V
0.2350
(F-15)
Zm -
2.256 X 10
-5
/"He'Ste +
\ 1.255 V
0.2350"
(F-16)
This is the ceiling altitude.
120
-------
APPENDIX G
TRAJECTORIES: ISENTROPIC AND ACTUAL
This discussion is based upon a paper by Danielson
(1961). He postulates a balloon which follows an isentropic
trajectory. If such a balloon is released from some point, it
will in general trace out a somewhat different trajectory than
the centroid of the volume of air initially surrounding it.
The time-dependent positions of the balloon and the air parcel
in which it was originally embedded are given by position
vectors in three dimensions. However, the problem can be
reduced to two dimensions by considering it in the vertical
plane which contains the air parcel, the isentropic balloon.
and the difference vector connecting them. In this coordinate
system, the horizontal and vertical components, r1 and z'. of
the difference vector are given respectively by:
// [<
<« - v If
f?
(dt)
(G-l)
(W -
z'
_
az
dt
(G-2)
Here V = V(r.z.t) and W = W(r.z.t) are the horizontal and
vertical components of the air velocity within the plane, and
W« is the vertical velocity of the balloon. The
horizontal component of air velocity orthogonal to the plane
does not enter. It does not affect horizontal or vertical
separation. In (G-l), the horizontal acceleration of the air
at the air parcel has been replaced with its Taylor expansion
121
-------
about the position of the balloon (see Boas, 1966). Similarly
in (G-2), the vertical velocity of the air at the air parcel
has been replaced by its Taylor expansion about the position
of the balloon. W and all derivatives of V and W in both
equations are to be evaluated at the balloon position. Higher
order terms in the expansions are denoted by dots.
The first term in (G-l) gives the effect of wind shear on
horizontal separation; the second term, the effect on
horizontal separation of differences in the temporal behavior
of the wind at the balloon and at the air parcel. The
expression under the integral in (G-2) is equivalent to the
difference between the vertical velocity of the air at the air
parcel, and the vertical velocity of the balloon.
The dominant terms in (G-l) and (G-2) can be expressed in
terms of the diabatic heating rate. d0/dt:
(W - WQ)
av = av d_e
82 " dt (G-3)
w w _
w - we - ae dt (Q_4)
In (G-4). if d9/dt is of order l°/day (1.16 x 10"5 °/s).
and if 39/3z is taken as the standard atmosphere value of
3.3 x 10~ °/m, then W - WQ becomes 0.35 cm/s. This
compares with observed vertical velocities relative to isobars
more than an order of magnitude higher. Hence, diabatic
heating or cooling induces only small changes in vertical
velocities. Note that in the absence of diabatic effects, the
leading terms in the expression for the separation of an
isentropic balloon from the centroid of the corresponding air
parcel go to zero.
122
-------
Danielson goes on to show that isentropic trajectories
typically yield differences from actual air parcel
trajectories an order of magnitude or more smaller than
isobaric trajectories, such as those which CVBs might
execute. He also reviews trajectory calculations based on
actual meteorological data to illustrate the point. One such
case is shown in Figure G-l.
123
-------
^
\
\
-.X- •
V
r-
\
\
\
\
w
Figure G-l. Comparison of twelve-hour isobaric and
isentropic trajectories originating at 700 mb at 0300 GCT 28
March 1956. Horizontal deviation is 1300 ±200 km. Passive
constant volume balloons approximate isobaric trajectories.
whereas air parcel trajectories are nearly isentropic.
124
-------
APPENDIX H
THE ARGOS SATELLITE-BASED DATA COLLECTION
AND PLATFORM LOCATION SYSTEM
A pamphlet describing the ARGOS system is
reproduced here courtesy of Service ARGOS.
125
-------
NASA-NOAA-CNES
THE ARGOS
SATELLITE-BASED
DATA COLLECTION
AND PLATFORM
LOCATION SYSTEM
126
-------
contents
c
Ce document a ete edite
par \e Centre National
d'Etudes Spatiales
Service Argos
18, avenue Edouard-Behn
31055 Toulouse cedex
La maquette,
la mute en page et I'i/lustration
ont etc reatisees
par MM Francois So/tana
et Jean-Louis Reilles
1 bis rue Joseph Mangnac
31300 St-Martm-du-Touch Tou/ous«
La photocomposition a ete reahsee
par Aqwtatne Arts Grapniques
5, ailees de Tourrty
33000 Bordeaux
// a efe tmprtme
par /'/mpnmene du Sud
24, rue N^greneys a Touiouse
La traduction a &t& realises
par M Stephen Dyson
introduction
( general description
( space segment
user platforms
• data collection
• platform location
9 ) ( onboard data collection system
( data processing centers
11) ( applications of the Argos System
( Service Argos
f Users - Service Argos relations
127
-------
crasteSlM^K^^^ • ^
(CNES) Toulouse
Space Center
Service Argos
buiWmg in
/ore ground
128
-------
introduction
The Argos data collection and platform location system
offers capabilities for the location affixed and moving
platforms and for the collection of sensor data transmitted by
platforms located anywhere on the Earth's surface.
The Argos System is the fruit of a cooperative
project between:
• the Centre National d'Etudes Spatiales (CNES, France),
• the National Aeronautics and Space Administration
(NASA, USA), and
• the National Oceanic and Atmospheric Administration
(NOAA, USA).
CNES, NASA and NOAA are bound by
a Memorandum of Agreement signed on December
10,1974 which defines the responsabilities of each agency.
CNES, benefiting from the experience acquired in this type
of activity through the EOLE program (1970-1974), is the
prime contractor for the design, development and operation
of the system.
The first Argos onboard equipment package (also known as
the onboard data collection system or DCS) was carried by the
TIROS-N satellite. The ten operational satellites designated
NOAAA thru NOAAJ, each equipped with the Argos DCS,
will be launched at such times as to ensure that two
operational satellites are available at all times.
Continuous service started in 1979 and is excepted to
continue through until 1990 at least
The Argos System is primarily intended for applications
concerned with environmental data collection
(Le. meteorology, oceanography, hydrology, ecology, remote
sensing of earth resources, etc.). Satellite-based data
collection and platform location is thus a new research
tool... and frequently one-offering unique features.
Michel Taillade
Head of Service Argos
129
-------
general
description
The Argos System comprises:
• a space segment consisting of two
satellites in orbit at any one time, each
being equipped with an onboard
data collection system (or DCS)
ensuring platform message
reception, processing and
retransmission.
• a set of user platforms,
each being equipped with
sensors and a platform
transmitter terminal
• the ground data processing centers.
composition
of the Argos System
platforms
data
processing
center
space segment
At any given time, the space segment
of the Argos System comprises two
operational orbiting satellites.
Orbit characteristics
• Configuration: circular.
. Altitude (satellite I): 830 ± 18 km,
(satellite II): 870 ± 18 km.
• Inclination: 98° (polar orbits; during
each orbit each satellite "sees" both
the North and South Poles).
• Period: 101 minutes (time required
for each satellite to circle the Earth).
• Relative disposition: orbital plane of
satellite I is inclined 75° relative to that
of satellite II.
• Sun-synchronous orbits: the angle
between the orbital plane of each
satellite and the Sun direction is
constant
Each satellite crosses the equatorial
plane at a fixed time (local solar time)
each day. These times are:
satellite 1:14 h 30 (ascending node)
and 2 h 30 (descending node).
satellite II: 19 h 30 and 7 h 30.
This characteristic is important from
the user's viewpoint since, from one
day to the next a given platform
comes within a satellite's coverage at
the same time (local solar time)
dissemination of results
130
-------
Oh
satellite I
2H30
7h30
satellite D
19H30
14H30
The following table gives approximate
data for satellite visibility, as a function
of latitude, for a 24-hour period.
north
earth
r
latitude
± 0°
±15°
±30°
±45°
±55°
±65°
±75°
V ±90°
cumulative
visibility
24 hours
80 min.
88min.
100 min.
128 mia
170 min.
246 min.
322 mia
384 min.
number of passes
in 24 hours
S"^^5S5l£^S4
6
8
8
10
16
21
28
28
7
8
9
11
16
22
28
28
8
9
12
12
18
23
28
28
mean pass
duration
10 min.
J
Technical data concerning orbit
geometry
At any given moment, each satellite
"sees" all the platforms located within
a circle 5000 km in diameter.
As the satellite orbits, the ground track
of this circle corresponds to a swath
5000 km in width encompassing the
Earth. At each orbit this swath covers
both the North and South Poles (polar
orbits). For a given satellite, the swath
is displaced by 25° (i.e 2800 km at the
Equator) every 24 hours as a result of
the rotation of the Earth.
Data collection performance, which
is determined by orbit geometry, is
thus a function of latitude.
north
user platforms
Each platform is equipped with
a platform transmitter terminal (PIT)
providing the up-link between
platform and satellite.
Platform sensors are linked directly to
the PTT. Analog or digital data from
up to 32 sensors can be handled by
each PTT.
The overall design of the Argos System
is such that the PTTs providing up-lmks
to the orbiting satellites offer
the following features
• simplicity PTT electronics consits
essentially of a transmitter
• lightweight less than 1 5 kg
• low power consumption
approximately 200 mW on auerage
• ease of implementation
• low cost: FF11000 to FF17.000 depending
on type and options.
All Pi is transmit on the same
frequency (401.650 MHz) and at
regular intervals: every 40 to
60 seconds in the case of "location"
platforms and every 100 to 200
seconds for "data-collection-only" type
platforms Each message transmitted
includes that particular platform's
number and the sensor output values
sampled at the time of transmission
Message duration is always less than
1 second.
Photo CEIS Espace
certified PTTs
The list cf manufacturers
producing certified PTTs
(Le satisfying Argos System
requirement^) is pubfehed
regularly in Argos Newsletter
or available from Service Argos
on simpte request
131
-------
data collection
PTTs transmit their messages
periodically, on the same frequency,
independently of each other, and
without the need for satellite
interrogation.
The only communication links
between users' platforms and the
satellites are one-way platform-to-
satellite links (or up-links).
Messages from platforms within view
of a satellite appear at the input to the
onboard receiver in a random fashion:
• message separation in time is
obtained through the asynchronizafion
of transmissions and the use of
different repetition periods.
• message separation in frequency is
achieved as a result of the different
Doppler shifts in the earner frequency
transmitted by the various platforms.
Random access
In the event of a number of messages
reaching the receiver input
simultaneously, up to four can be
acquired provided they are separated
in frequency.
The probability of acquisition of a
message transmitted by a platform
during a satellite pass is 0.99 provided
all messages transmitted dunng
periods of about 10 minutes
are identical.
/ platform location
Principle of platform location
The location of each platform is
determined solely by measuring the
Doppler effect on the carrier
frequency of in-coming messages
(the transmitting frequency being
fixed and the same for all platforms).
Each measurement made by the
satellite corresponds to a field of
possible positions of the platform
under consideration. This field takes
the form of a cone with the satellite at
its apex and the velocity vector as
the axis of symmetry-
The altitude of the satellite being
assumed to be known, the
intersection of several of these cones
(each corresponding to a separate
measurement) with the altitude sphere
J yields the solution.
,'/]' Taking into account the nature of the
geometrical fields concerned (cones
with the orbits as their axes of
symmetry), the statistical processing
of the measurement data obtained by
the satellite yields two symmetrical
solutions relative to the ground track.
One of these points is the required
solution, the other its "image"
The ambiguity cannot be resolved
without additional information,
e.g. previous positions, range of
possible speeds, etc.
The slower the platform moves, the
more easily the ambiguity can be
resolved.
132
-------
Platform location performances
The following curve gives the
number of locations per day you may
expect with the Argos System as a
function of the latitude.
Number at locations per day
f performance data
f for99%oicases
location accuracy
V accuracy of speed determination
Drifting
buoys
1km
0.3 m/s
DriftingX
balloons \
3km 1
1.5 m/s/
Errors associated with platform
location
The main sources of error regarding
platform location are:
• the accuracy with which the platform
altitude is known. This is relatively
unimportant in the case of drifting
buoys, but may be very important in
the case of balloons;
• the stability of the FIT oscillator, in
particular the drift in transmitting
frequency during a satellite pass
(approximately 10 minutes);
orbit determination platforms
Orbit determination platforms
Eleven special orbit determination
platform are installed at appropnate
locations. The geodetic position
of each of these sites have been
accurately determined
These platforms are used for the
high-precision determination of each
satellite's orbit within 300 m along
the orbit and within 250 m along the
two other axes
The resultant data being m turn
employed for platform location
133
-------
Argos data flow
NOAA.
Wallops Island (USA)
telemetry station
T1ROS-N satellite
NOAA, Gibnore Creek (USA)
telemetry station
METEO, Lannion (France)
telemetry station
Users
8
134
-------
breakdown of interval between data collection and availability
collection system x
component
The onboard DCSs are equipped with
receivers that pick up messages
transmitted by platforms within the
satellite's coverage. Message reception
is on a random access basis. As each
message is acquired, the DCS records
the time and date, measures the
carrier frequency and demodulates
the platform identification number and
sensor data
Four messages can be received and
processed at any one time.
These data are then formatted and
stored by one of the onboard
magnetic tape recorders. Each time
the satellite passes over one of the
three telemetry stations, the data
recorded on tape are read out and
transmitted to ground.
capacity of the Argos DCS
The onboard data collection system
maintains nominal performance with
920 data collection only platforms
or 230 platform location only
platforms (plus minimal data
collection) respectively that falls
simultaneously within satellite
coverage.
direct transmission
in addition to recording the DCS
data, the spacecraft transmits it in
real-time on 136.770 or 137770 MHz.
The DCS data is multiplexed with
other instrument data at a low bit
rate. Users can receive sensor data
from platforms within the satellite's
coverage at the time of transmission.
transmission to
Service Argos
processing at
Service Argos
data processing
centers
Data are read out once every orbit
i.e. every 100 minutes for each
satellite.
Once a satellite has completed
telemetry data transmission for a
particular pass, the received data are
transmitted to the NESS (National
Environmental Satellite Service)
Center at Suitland (Maryland, USA).
Data concerning the Argos System
are separated from those concerning
other satellite systems and transmitted
to the CNES Toulouse Space Center
where the Argos Data Processing
Center is located.
The processing performed at the
Center permits the determination of
platform positions and the extraction
of sensor data.
breakdown of interval
between data collection
and availability
The minimum interval between data
collection and data availability in a
nominal system can be evaluated as
follows-
• a particular message collected dunng
a single orbit may have been stored
up to 100 minutes before the onboard
tape is played back to the ground
telemetry station.
• the time required-for data
transmission between the receiving
telemetry station and the NESS
facility is estimated-at 45 minutes
while the time required to retransmit
Argos data to Service Argos
is estimated at 20 minutes
• the time required to process the data
generated by a single orbit is
40 minutes at the maximum.
The interval between the time when
a given platform message is acquired
by the satellite and the time when the
data is processed and available to
users will vary between 1 hour 45
minutes and 3 hours 25 minutes
depending on the platform's position.
These figures thus represent the
ultimate possibilities of the system.
Argos data availability
About two third of the results
are made available to users at
the Toulouse Space Center within
3 hours after the onboard recording
of the corresponding message.
135
-------
results supplied to users
In the case of data-collection-only
platforms, the results supplied to users
include:
• experiment identification code
(project n°).
• platform number (Argos ID n°).
• time of data collection.
• sensor data after conversion to pure
binary code or to physical units.
In the case of location-and-data-
collcction platforms, the results include
all the above plus:
• time of location.
• position data (latitude and longitude).
• latitudinal and longitudinal velocity
components in degrees per day.
data dissemination options
Each user can choose between
various modes of data dissemination
to achieve an optimal time/cost
trade-off for his particular
requirements and circumstances.
off-line data distribution
Results, in this case, are made
available to users from the data bank
Data stored in the bank are read out
once a week for all users and held
available for a period of three months
only.
Results are available in the
following forms once a fortnight
or once a month:
• computer pnnt-out
• magnetic tapes.
The normal mode of distribution is by
the fastest postal means.
on-line data distribution
Results, in this case, are supplied the
moment they become available, i.e.
immediately the corresponding data
flow has been processed. A suitable
high-speed means of communication
must then be chosen from among the
following'
• permanent links:
• dedicated lines hired by the user.
. dedicated network through
the French TRANSPAC and other
dedicated networks such as
EURONET and TIMENET...
• the Global Telecommunications
System (GTS) operated by the
World Meteorological Organization Service Argos data processing
(W.M Ol Center in Toulouse (France)
• links via switched networks (telex or
telephone):
• sending of telex messages generated
automatically by the Argos Data
Processing Center,
. sending of data following
a telex ori telephone call from
the user (computer results
files consulted directly).
data processing
and dissemination
10
post
user consultation
136
-------
applications
of the
Argos System
1 Meteorology - 2 Oceanography
3 Offshore-4 Ghcioiogy
5 Hydro/ogy - 6 Animal tracking
7 Sismo/ogy - 8 Vo/cono/ogy
Argos
applications
The Argos System is particularly
suitable for gathering environmental
data in various areas such as:
Meteorology
Worldwide data collection of
observations concerning the state of
the atmosphere is possible.
Meteorological applications include:
• world meteorological and
oceanographic networks (e.g. FGGE.
First Garp Global Experiment) using
mainly buoys and balloons,
• anchored meteorological buoy
networks,
• land based stations used
for meteorological measurements.
• meteorological keyboard terminals
for vessels,
• automatic mini weather stations for
merchant ship developped pursuant
to a WMO recommendation.
This kind of equipment is also
successfully used for offshore yacht
races,
• studies on surface currents using
rapid drifting buoys.
Oceanography
represents a wide variety of current
and potential applications for
the Argos System. Typical examples
include:
• regional experiments such as
theEECsCOST43,
• courses, studies, width and speed
of major currents,
• subsurface studies with
measurements of water temperature
at various depths, of drogue line
parameters, of acoustic parameters.
Such oceanographic research is also
of great use to the fishing industry.
• ocean swells and waves (formation.
spectrum, propagation, correlation
with underwater ambient noise
levels studies)
/ French Mansonde buoy
{photo Sa/are-Crouzet)
2 Mini weather station on a sail boat
(photo CE1S Espacel
3 Automatic weather station
-------
1 Hydrological station for nuer monitoring
(photo ORSTOM)
2 Norwegian Meteo-Oceanographic buoy 2
(photo 1KU)
3 Animal tracking
(photo NOAA - South West fisheries Center)
3 I
Offshore
in connection with meteorological,
oceanographic and glaciology
utilization, for prospection studies.
environmental control, oil spill
tracking, pollution studies, ocean
climate studies, operation monitoring,
mooring failure momtonng
Glaciology
to study polar currents and compare
buoys and iceberg trajectones.
Operational avalanche nsk forecast
program can be included in glaciology
application.
Biology
is used to cover animal tracking.
Experiments were conducted on
the following species to improve
hardware problems dolphins.
basking sharks, turtles, whales
and polar bears
12
Hydrology
covers a wide range of potential
applications due to the low cost
of the Argos compatible hydrologic
station management of hydrologic
networks, water survey and supply
and assessment of water ressources.
Geology
monitoring and prediction of
earthquakes and volcanic eruptions.
prediction of fault movements.
study of the thermal inertia of soil
types
The Argos System is strongly
recommanded to all WMO (World
Meteorological Organization)/
IOC (UNESCO Intergovernmental
Oceanographic Commission)
participating countnes for utilization
in future international climate
programs.
Service Argos
Supervised by the Argos
Operations Committee, Service
Argos is a bilateral Franco-
American (CNES, NOAA. NASA)
organization set up by CNES at
the Toulouse Space Center in
France Service Argos is the
contact point for all system users.
The main responsibilities of Service
Argos are:
• coordination of the operation
and monitoring of the entire Argos
System,
• promotion activities to extend
the use of the system,
• preparation and dissemenation
of documentation. "
• maintenance of the orbit
determination network
• technical certification of prototype
PTTs submitted by manufacturers.
• provision of technical assistance
services concerning the
implementation and operation
of transmission equipment for use
with Argos System satellites.
Users-Service Argos
relations
Telephone Telex
• (61) 53.11.12 • 531081F
General Documentation
• Argos Newsletter, published
penodically by Service Argos and
available on request.
• Argos System User's Guide.
Access to the Argos System
• the document entitled "Argos
application form" is available.
on request to any individual or
group in the process of preparing
a project or program.
Visits
• persons with a professional
interest in the Argos System can
arrange to visit the Argos facilities
Manufacture and
Certification of PTTs
• specification for PIT
manufacture and details concerning
the certification procedure are
available either from Service
Argos or from manufacturers
138
-------
139
-------
APPENDIX I
PUMPDOWN SPEED WITH AN ARBITRARY LAPSE RATE
Consider the density p of the ambient air in the
3
vicinity of the balloon as it is pumped down. The gas law
gives:
o
p
a RT (1-1)
where M is the average molecular weight of air.
The rate of change of ambient air density during pumpdown
is given by
dpa _ M |"l dPa dz. _ fa dT
dt ~ R |_T dz dt T2 dz
dz]
dtj
R _T dz T2 dz
dz.
dt
(1-2)
where here dz/dt is the rate of change of altitude, and
-dT/dz = T, the lapse rate. To evaluate dP /dz, note that
cl
dP = -p g dz (1-3)
d ct
so
dz RT (1-4)
140
-------
Substituting:
dp.
'a _ M
dt ~ R i T J RT
dz.
dt
d-5)
If the balloon remains in equilibrium during pumpdown. then by
definition dz/dt is the pumpdo'
equilibrium condition applies:
definition dz/dt is the pumpdown speed, v , and the
m
Taking the time derivative
dpa dp
dt ~ dT
But dp,/dt is given by Equation (B-3).
b
(1-6)
(1-7)
dt
Substituting.
V V RT
M
-M
R T RT
Solving for v :
v = - ST/[V(Mg/R
H
v_
(1-8)
d-9)
d-10)
141
-------
APPENDIX J
DIFFUSIVE SPREAD OF A MARKED AIR PARCEL AND
ITS IMPLICATIONS FOR LAGRANGIAN EXPERIMENTS
A design goal for the PLT is that it be capable of
following mean vertical flows as low as 1 cm/s with
"acceptable" fidelity. The questions to answer are. how is
the "mean flow" to be interpreted, and what is "acceptable"
fidelity?
To answer these questions, it is useful to consider a
discrete initial volume of air in which the balloon is
embedded. If one conceptually marks every molecule in that
volume, and follows each through subsequent time, one finds
that diffusion will increase the RMS separation of this set of
marked molecules from the mean in both the horizontal and the
vertical; that is. a . a. and a will grow with time. Here the
x y z
-------
dimensions of the volume one desires to follow. Clearly that
choice is experiment-dependent. In air pollution experiments.
there is a physical puff or plume one desires to follow, the
dimensions of which determine the choice. The principle is
that the volume should be "representative" of the puff or
plume. If one is interested in following the plume from some
industrial facility in complex terrain over a few kilometres.
the volume of interest might be some tens of metres in size —
perhaps comparable in scale to the plume cross section during
the transport of interest. If one is interested in following
the urban plume from Chicago, the volume might be 100 km in
horizontal dimension, and the depth of the mixing layer in the
vertical.
This interpretation of the mean flow leads to the
practical guestion of how one can in fact follow the centroid
of the volume of interest. Here, we borrow an idea from
statistical mechanics. Under certain broad conditions.
ensemble averages and time averages can be assumed to be equal
(Reif, 1965). This suggests that the time average of the
vertical velocity of the air evaluated at the PLT over an
appropriately-chosen time interval may approximate the average
vertical velocity of the air integrated over some volume
surrounding the PLT. It is certainly plausible that the time
average at the PLT is more characteristic of the motion of the
centroid than the instantaneous velocity. It is also
plausible that the longer the averaging time, the larger the
associated equivalent averaging volume.
A characteristic time associated with a volume is the time
required for molecules emitted from a central point within the
volume to diffuse throughout. For lack of a better choice, we
take this as an estimate of the appropriate averaging time for
a PLT intended to follow the centroid of the specified volume.
143
-------
To evaluate the time required to diffuse throughout the
specified volume, we take the approach of MacCready et al
(1974). From MacCready's derivation, it can be shown that:
aZ - °'88 LwD (
where
-------
5 r
4 -
o —
2 -
1 -
10
T
15
20
Figure J-l. Dimensionless diffusion parameter D versus
dimensionless parameter T proportional to time. For T < 0.5,
D = T; for T > 3. D *
145
-------
based on values of e
1/3
measured aloft on the DaVinci II
balloon flight (Zak. 1981) over a 24-hour period (Figure J-2).
During the convective portion of the day. c1 3 averaged
about 2; during the evening and early night, about 0.5; and
during the post-midnight hours, sometimes as low as 0.1. Here
the units are cm2 3s~1. A sketch of the device used to
make these measurements is shown in Figure J-3.
It can be shown that:
T.= 2.64 x 10 3e 1/3t
(J-5)
if e1/3 is left in units of cma/3s~1'
t (s)
10
30
100
300
1.000
3.000
10,000
30.000
TABLE J-l
T as a Function of ei/3 and Time
i/3
2.0
.053
.16
.53
1.58
5.28
15.8
52.8
158.4
0.5
.013
.040
.13
.40
1.32
3.96
13.2
39.6
0.1
.0026
.0079
.026
.079
.26
.79
2.64
7.92
146
-------
S
<
Q
O
UJ
O
<
o:
UJ
>
^
UJ
•
t—
z
5
^M
UJ
O
i
~
— _
O
^mm
>
^
Q
<
9 0
CO ^
ai&iapow w3p aiqiSijSaN
juaujssassv 8Ai}e}i|en5 s.jopd
* *
•v?
7^*
* "•" ~:S
--v*
~z*
»- -"^
— - ~— ~ _^_ ^ * ~~
- /- - -. ~^~^*~
~ — ~~-~~s*~~ ~
~~ *~ ~*~ _r
•*»»•*—
-T". ^-C-^"
""" -"V^S"'
— m
-, — "i**-^" -~~~ —
" ~ -**..'=* "•—
"" — * ~^
. - ~ "-^Vr » ~ ~
' -" 7 p~r ... -. -
* * ^ ^ "*•* ~ — **-^-*> « ^^
"^r" **^---^."Vt*'"* ™
- " •
A EH
OlQ
•H O
r-t
14-1 VO
in
M 00
i-» O
•H CO
O
(U
X! J3
*J 0
C
ti 3
O flj
UJ J
OJ
g •
•H X!
•(-> 0
C
ca 3
r^ n«
rJ H)
03 I-H
t-l
g
> 0
M
n (H
•v
H 03
U S-|
3
• O
CM A
1
>n C
•rH
0)
M C
3 >
•H -H
in
147
-------
To gondola
22.9 cm diameter
propeller
2.3 kg weight
Figure J-3. "e-meter" used to make turbulence measurements on
DaVinci II. The two wind sensors are identical to the one
shown in Figure 4-6. with the exception of supporting
electronics.
148
-------
Knowing T. we may use Equation J-l. <* . .= 0.88 L D
and Figure J-l to evaluate a . However, it should be
noted that for T <. 0.5, D = T. and that for T >3. D =
TABLE J-2
a (metres) as a Function of c1 and Time
1.
3.
10,
30,
t (S)
10
30
100
300
000
000
000
000
c]
2.0
4.2
13.
42.
87.
182.
315.
575.
997.
i/3 (cma/3S-1)
0.5
1.0
3.2
10.
32.
77.
158.
288.
498.
0.
.
2.
6.
21.
49.
126.
223.
1
21
63
1
3
For the convective period, a rough check is available on
the rate of diffusive spread. On the day of DaVinci II,
sunrise occurred at 5:36 am. Between sunrise and noon, the
mixed layer grew to about 2 km. A puff originating at ground
level at sunrise would have continuously grown to fill the
mixed layer in the vertical to its maximum height. We assume
that the mixed depth is at most 2 a , so that at t = 2.3 x
4 2
10 s, a = 1000 m.
149
-------
Using our technique, we predict a = 870 m for this
z
time. In view of the approximations made, this is quite
reasonable agreement.
If the reasoning presented earlier is valid, the data in
Table J-2 allow one to crudely estimate the thickness of the
volume characterized by a specified averaging time. For an
averaging time of 300 seconds (5 minutes), during a period of
convective mixing with e1 3 = 2, the volume is about 173 m
thick. During the decay of convective mixing characteristic
of evening, the same averaging time corresponds to a layer of
63 m thickness. In the hours before dawn in the presence of
minimal turbulence, a 5-minute averaging time characterizes a
layer at least 12 m thick. If under any of these conditions
the volume of interest is in fact thicker than the dimensions
given above, then either the averaging time should be
appropriately increased to attempt to follow the centroid more
faithfully, or one must accept having a "better" Lagrangian
tracer, that is. one which acts more like an individual gas
molecule and which follows motions of smaller scale than the
scale of interest, and hence simulates the motion of the
centroid of the volume less well.
Table J-2 also casts light on the question of "acceptable"
fidelity. Even with e1/3 = 0.1 cma/a s""1. which
is close to the minimum turbulence believed to be commonly
present in the atmosphere for sustained periods, a puff
spreads in the vertical from its centroid at an initial rate
of about 2.1 cm/s. Only after hours of diffusion does the
rate of spread fall to 1 cm/s (see Figure J-l). Consequently,
even in extremely light turbulence, following the mean flow to
within ± 1 cm/s will keep the marker balloon within ±a
of the centroid of the ensemble of molecules which started out
in its immediate vicinity. On the other hand, during active
150
-------
daytime convective mixing, the initial rate of diffusive
spread is about 42 cm/s. This falls to 3.3 cm/s in eight
hours. Consequently, if a marker balloon follows the mean
flow to within ±3.3 cm/s over eight hours under convective
conditions, even if the deviation is systematic, the balloon
will remain within ±a of the centroid of the point source
z
puff it started with.
It should be pointed out that the arguments presented
above are approximate, and furthermore, depend upon
relationships for the inertial subrange of turbulence. As the
scale of the volume of interest increases...say beyond a
kilometre or so...these relationships become less accurate.
However, we have been applying these arguments to motion in
the vertical. Here, the region of interest is usually limited
to only about 2 km. Thus the arguments should be reasonably
sound throughout this region. We do not consider the
horizontal because, for the horizontal, there is no mechanism
analagous to buoyancy which could lead to systematic bias in
the balloon motion.
Note also that for validating long range transport and
transformation models, even the condition that the PLT remain
within ± o , ± o , and ± o of the centroid of the
x y z
volume of interest may be stronger than is necessary. Lamb
(1984) has pointed out that the meteorological data available
do not uniquely define the atmospheric flow field present at
the time of an experiment. Rather, the data merely limit the
flow field to some ensemble of flow fields consistent both
with the data and with physical principles. What models
really seek to do is predict ensemble average pollutant
concentrations, rather than the concentrations produced by the
flow field actually present on a given day. Analogously, for
model validation, perhaps the trajectory of the PLT need only
151
-------
approximate the mean trajectory over the ensemble of flow
fields consistent with the data, rather than the trajectory of
the centroid of the volume of interest under the action of the
flow field actually present. Because the meteorological data
usually available admits of quite a range of flow fields, this
is likely to be a considerably less demanding requirement.
Consideration of Lamb's views leads to yet another
perspective of the PLT. Clearly, the PLT itself yields
information on the flow field, quasi-continuous information in
the immediate vicinity of the volume of interest. This is in
marked contrast to the widely-spaced grid of data available
every 12 hours from the winds-aloft network. The information
from the PLT can be used to narrow the ensemble of flow fields
consistent with the available meteorological data. If these
data are appropriately used, the flow field entered into a
model would likely approximate the actual flow field in the
vicinity of the volume of interest much more accurately. This
could be of considerable value in experiments which include
both a gaseous tracer species and a PLT for model validation.
The PLT would minimize uncertainties arising from the flow
field, and the gaseous tracer would allow measurements of
concentrations to be directly compared with model
predictions. In the experiment configuration shown in Figure
7-1. such experiments could be done with a single PLT, and a
single instrumented aircraft, and would not necessarily
require extensive ground sampler arrays.
152
-------
TECHNICAL REPORT DATA
(Please read huauctions an the reverse before completing)
1. REPORT NO.
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
DEVELOPMENT OF AN ADJUSTABLE BUOYANCY BALLOON
TRACER OF ATMOSPHERIC MOTION
Phase I. Systems Design and Demonstration of Feasibil
6. REPORT DATE
6. PERFORMING ORGANIZATION CODE
ty
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
B.D. Zak, H.W. Church, A.L. Jensen, G.T. Gay, and
M.D. Ivey
Sandia National Laboratories
Albuquerque. MM 87185 . '
10. PROGRAM ELEMENT NO.
CDTAID/08/3005 (FY-85)
11. CONTRACT/GRANT NO.
IAG DW930214
12. SPONSORING AGENCY NAME AND ADDRESS
Atmospheric Sciences Research Laboratory - RTP^NC
Meteorology and Assessment Division
U. S. Environmental Protection Agency
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Interim
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An Adjustable Buoyancy Balloon Tracer of Atmospheric Motion is a research tool which
allows one to follow atmospheric flows in both the horizontal and the vertical, in-
cluding the weak, sustained vertical motion associated with meso- and synoptic- scale
atmospheric disturbances. The design goals for the Balloon Tracer to be developed
here specify a lifetime _> 3 days, tracking range _> 1000 km, a ceiling altitude _> 500
mb (5.5 km), and the capability to respond to mean vertical flows as low as 1 cm/s.
The balloon tracer is also to measure and telemeter selected meteorological variables,
to be sufficiently inexpensive to permit use in significant numbers, and to be serviced
by a ground system capable of handling several balloon tracers at a time. While the
balloon tracer has applications throughout the atmospheric sciences, the immediate mo-
tivation for this effort is to meet the need to evaluate the accuracies of existing air
pollution transport models, to establish source-receptor relationships to distances of
order 1000 km, and to assess the inherent limits on the predictability of source impact
at long distances. The authors have proposed a generic design for such a system. They
also have subjected the proposed design to theoretical .analysis, have constructed a
prototype, and have conducted a series of tests with the prototype to evaluate the con-
cept. They conclude, without reservation that a system meeting the design goals is
feasible, and are proceeding to build that system in Phase II of this project.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
flCCTFTFD
21. NO. OF PAGES
20. SECURITY CLASS (ThSpage)
22. PRICE
EPA Form 2220-1 (9-73)
------- |