Vehicle and mission design of a future small payload
launcher
Christie Alisa Maddock
∗
, Federico Toso
†
, Lorenzo Ricciardi
†
, Alessandro Mogavero
†
University of Strathclyde, Glasgow, G1 1XJ, United Kingdom
Kin Hing Lo
†
, Sriram Rengarajan
†
, Konstantinos Kontis
‡
University of Glasgow, G12 8QQ, United Kingdom
Andy Milne
§
, Jim Merrifield
¶
, and David Evans
k
Fluid Gravity Engineering, Hants, PO10 7DX, United Kingdom
Michael West
∗∗
BAE Systems Regional Aircraft, Prestwick, KA9 2RW, United Kingdom
Stuart McIntyre
††
Orbital Access Limited, Prestwick, KA9 2RW, United Kingdom
This paper presents the conceptual design and performance analysis of a partially reusa-
ble space launch vehicle for small payloads. The system uses a multi-stage vehicle with
rocket engines, with a reusable first stage capable of glided or powered flight, and expen-
dable upper stage(s) to inject a 500 kg payload in different low Earth orbits. The space
access vehicle is designed to be air-launched from a modified aircraft carrier. The aim of
the system design is to develop a commercially viable launch system for near-term opera-
tion, thus emphasis is placed on the efficient use of high TRL technologies. The vehicle
design are analysed using a multi-disciplinary design optimisation approach to evaluate the
performance, operational capabilities and design trade-offs.
I. Introduction
M
any of the forecast studies looking at the future of the satellite market predict a period of unprecedented
growth over a range of differing satellite sizes and types. The emergence of mega-constellations or
smaller single purpose satellite constellations exploit the ever progressing miniaturisation of powerful sensors
capable of capturing data or providing services that were hitherto only feasibly carried on large government
agency developed satellites. Combined with the inexorable development of high powered computing and high
bandwidth communications, small satellites can now deliver space based applications with levels of fidelity
and performance never previously accessible by broad based commercial users. The investment by many
countries, including the UK, in satellite application development and entrepreneurial growth in space based
service development is a major contributing factor into the forecasted growth in small satellite demand.
With the increased demand, there is increased pressure for new accessible, responsive and cost effective
small satellite launch capacity. Yet the austere period following the termination of the Shuttle programme
has not yielded many proven small satellite launch technologies. The focus for many space agencies, in
∗
Lecturer of Aerospace Engineering, christie.maddock@strath.ac.uk, AIAA Member
†
Research Associate, AIAA Member/Student Member
‡
Mechan Chair of Engineering, Professor of Aerospace Engineering, AIAA Fellow
§
Senior Scientist
¶
Director, Civil Space
k
Project Manager
∗∗
Chief Aerodynamicist and Head of Technology
††
Chief Executive Officer
1 of 21
American Institute of Aeronautics and Astronautics
particular NASA, on enabling human spaceflight through programmes such as the X Prize in the early 2000s
meant that entrepreneurial investment has mainly been focused on human spaceflight capabilities and in
replacing the ISS resupply capacity lost with the termination of the Shuttle programme. With the recent
shift towards fully commercial options for space launch, coupled with the current and predicted growth of
the market, there has been a resurgence to develop and operate a small payload launch system. A survey
from 2015
1
lists 20 launch vehicles under development around the world designed to launch small satellite
payloads weighing up to 1000 kg. Most are in development by small to medium sized businesses, with a
target first flight by 2020.
The following paper details the conceptual design and analysis of a future commercial launch system for
small payloads (up to 500 kg). The system is a multi-stage vehicle using rocket propulsion systems that will
be air-launched from a carrier aircraft. The main vehicle is a spaceplane design that will allow for glided
re-entry/return flight. The second stage is stored within the main body of the spaceplane, among other
benefits this allows for better control of the moments induced by the movement of the centre of gravity
though introduces complexity and release issues. An optional small upper stage is also investigated to
increase the range of possible orbits that can be reached. With the main operational spaceport located on
the western coast of Scotland, the air-launch increases the type of orbits that can be reached, and improves
the flexibility of the system by allowing the transport and recovery of the first stage.
The paper describes the overall approach with design objectives and mission requirements, then details
the system models developed for use within a specialised integrated design platform for space access vehicles.
The optimisation used within the system performance analysis is described, with results presented examining
the trade-off in performance of altering key design variables in the configuration, specifically the engine and
wing sizing (aerodynamic efficiency). The nominal mission is to deploy a 500 kg payload into 600 km altitude
circular orbit at an inclination of 88.2 deg, with an option for an upper stage to raise the orbital altitude to
deliver a 150 kg payload to 1200 km.
II. Approach
A specialised integrated design platform was used that was developed to analyse the performance and
optimise the mission design for transatmospheric flight vehicles. The software has been used to evaluate
different space launch systems,
2, 3
from single stage to orbit vehicles
4
to expendable vertically launched
rockets.
5
Computationally fast engineering models were developed for the different subsystems to allow the
performance of the system to be evaluated using a multi-disciplinary design optimisation approach. Different
design criteria were selected as inputs, with the models relating the impact of changes on those variables
on the system. For example, the wing reference area is a design input that affects the aerodynamic lift and
drag forces, and the vehicle dry mass. A full mission is simulated, optimising different criteria with various
system and operational constraints. Target orbits and payloads were determined through a market demand
study.
6, 7
III. System models
In this section, mathematical models are presented for the vehicle design and operation. The models
are divided by discipline: vehicle mass and configuration, aerodynamics, aerothermodynamics, propulsion,
environment models for Earth including geometry, gravitational field and atmospheric model, and the flight
dynamics and control.
III.A. Vehicle configuration
The basic concept for the launch system was for operations from horizontal take-off spaceports. Early in the
concept analysis it was decided that for maximum operational flexibility the concept of an air launched system
would be investigated. This drove the basic configuration which consisted of a carrier aircraft, converted
from a large commercial airliner, a winged recoverable booster and an expendable upper stage. Two launch
vehicle configuration concepts were considered: a more conventional, horizontal stack configuration where
the upper stage and payload are attached to the front of the winged booster stage, and a larger winged
booster with the upper stage and payload housed in an internal payload bay.
Through the course of the study the concept of the winged recoverable booster evolved from a rocket
2 of 21
American Institute of Aeronautics and Astronautics
with wings and a payload bay, to an integrated aerospace plane with multiple parallel propellant tanks and
a novel payload cartridge system (see Fig. 1).
Figure 1. Conceptual drawings showing the evolution of the configuration
For the concept studied, the launch vehicle was mounted below the carrier aircraft on the fuselage
centreline. This layout was chosen primarily for flight safety reasons, as release of the launch vehicle in
a power off glide is most readily achieved by a conventional drop manoeuvre, in a similar fashion to the
Pegasus/L-1011 combination operated by Orbital ATK. This drove space envelope and mass constraints on
the launch vehicle.
The initial vehicle layout employed a 60
◦
sweep delta wing. Initial sizing was based on wing loading
requirements for manoeuvre capability at maximum mass after release and for good low speed handling
at the empty mass for approach and landing. Previous data for winged orbital re-entry vehicles, research
aircraft, air dropped vehicles, high performance military aircraft and high performance civil aircraft were
used to inform the wing sizing and concept layout models were produced in CATIA. These models were
subsequently used for the engineering analysis work.
During the study the mission requirements evolved and the vehicle concept was developed into a more
complex configuration to meet these requirements. The final configuration had a straight tapered wing of
45
◦
leading edge sweep featuring an 80
◦
sweep inboard leading edge extension. To accommodate the payload
cartridge system the fuselage width was approximately doubled with respect to the original concept. This
allowed the propellant tanks to be distributed in a manner conducive to good control of the centre of gravity
during the powered ascent phase of flight. It also gave a lifting fuselage shape.
A number of parametric mass prediction methods were used for the initial concept level mass predictions,
alongside mass data from a NASA reusable launch vehicle study. Methods were sourced from a methods
database document produced by Rohrschneider.
8
The data were applied carefully based on the quoted
sources. A comparison was made with a NASA study
9
which helped to inform certain aspects of the final
mass statement.
Full mass statements were prepared breaking the vehicle down into its major structural components (e.g.,
wing, fins, fuselage structure, propellant tanks) and the major systems (e.g., propulsion, avionics, landing
gear). To allow for resizing during the vehicle optimisation phase, parametric scaling equations of the form,
m
new
= m
ref
S
new
S
ref
b
(1)
were developed for the major components, where m is the mass, S is a reference value which is scaled, b a
scaling exponent, and the subscripts ref refer to the original value and new to the scaled value.
3 of 21
American Institute of Aeronautics and Astronautics
Knowing the mass breakdown and component layouts, the vehicle centre of gravity and its variation
with fuel burn and payload deployment was determined and assessments made of the ability to trim i.e.,
reduce the pitching moment to zero during ascent and re-entry. Following this the propellant tanks were
redistributed to give an acceptable centre of gravity range during flight. The internal layout of the final
configuration is shown in Figure 2. Note that the propulsion system shown is indicative of the size and
location but does not include any engineering details of the installation.
Figure 2. Final vehicle configuration including a possible internal layout
III.B. Aerodynamics
The aerodynamics of the vehicle configuration were analysed for Mach numbers ranging from 0.2 to 30,
angles of attack of −5
◦
to 40
◦
and for altitudes up to 100 km. The approach estimates the drag coefficient
at zero incidence C
D
0
and the normal force coefficient C
N
at different angles of attack α for each component
of the vehicle (fuselage, fairing, wings and tail). The coefficients are determined based on different theories
for each Mach number range, from subsonic to hypersonic, detailed by Mason
10
and Fleeman.
11
The lift and drag forces of each component at different Mach numbers and angles of attack are modelled
by,
C
L
= C
N
cos α − C
D
0
sin α (2a)
C
D
= C
N
sin α + C
D
0
cos α (2b)
Equation 2 is valid for small angles of attack, when the axial force is approximately equal to drag. Although
large angles of attack are considered the method would overpredict the lift at such angles, since the effect
of stall is not accounted for. However, the effect of flow separation at the base of the fuselage is considered.
The fuselage of the vehicle is approximated to have an elliptic cross section (with same area of cross section
and major axis equal to half of the maximum width of the fuselage) in order to enable the application of
theories. The lift and drag coefficients, after appropriate normalization (using the wing surface area) are
then added up to give the total lift and drag coefficient of the entire configuration. Application of linear
theory and modified Newtonion theory is used to deduce the wave drag coefficient at zero incidence over
slender circular/elliptic nose C
d0,wave,b
, wave drag coefficient at zero incidence over the delta wing (as well
as tail, which has similar form) C
d0,wave,w
, and the normal force coefficient as a function of angle of attack
for the cone-cylinder C
N,b
as well as wings C
N,w
, given by the following equations.
C
d0,wave,b
= 0 for M < 1 (3a)
C
d0,wave,b
=
3.6d
N
`
N
(M − 1) + 3
for M ≥ 1 (3b)
C
d0,wave,w
= 0 for M < 1 (4a)
C
d0,wave,w
= f (M
λLE
, γ, δ
LE
, tb/S
w
) for M ≥ 1 (4b)
|C
N,b
| =
a
N
b
N
sin(2α) cos(α/2) + 2
`
C
d
C
(5)
4 of 21
American Institute of Aeronautics and Astronautics
|C
N,w
| =
πA
2
|sin α cos α| + 2 sin
2
α for M
2
< 1 + (8/πA)
2
(6a)
|C
N,w
| =
4 |sin α cos α|
(M
2
− 1)
1/2
+ 2 sin
2
α for M
2
≥ 1 + (8/πA)
2
(6b)
where `
N
is the length of the cone nose, d
N
is the equivalent diameter with major axis a
N
and minor axis
b
N
, `
C
is the length of the cylindrical body, A is the aspect ratio of the wing, t is the wing thickness, b is
the wing width, S
w
wing reference area, δ
LE
is the wing thickness angle, γ is the specific heat ratio, α is
the angle of attack, and M is the freestream Mach number with M
λLE
the Mach number resolved in the
direction normal to the wing leading edge with a sweep angle λ
LE
.
The base wave drag on the wing C
d0,wave,w
has a complex algebraic functional form f whose expansion
is given by Fleeman.
11
The above coefficients are all normalized by their respective reference areas (and not
a common reference area).
The coast drag of the cone-cylinder body C
d0,c
is given by the following engineering correlation.
11
C
d0,c
= 0.12 + 0.13M
2
for M < 1 (7a)
C
d0,c
= 0.25/M for M ≥ 1 (7b)
The inviscid drag at zero incidence also includes drag due to nose and leading edge bluntness, which are also
estimated using the semi-empirical expressions given by Fleeman.
11
Severe peak heat transfer rates are among the important issues at hypersonic flow conditions. Blunt
noses and leading edges are particularly preferred at hypersonic speeds in order to reduce the peak heat
transfer rate. This would however increase the drag. A configuration that was initially considered for the
study consisted of sharp nose and leading edges. A preliminary analysis was done at different hypersonic
Mach numbers, introducing various bluntness radii at the nose and leading edge, predicting the peak heating
theoretically
12
at hypersonic Mach numbers, and the corresponding increase in drag due to bluntness. At
Mach 8, a bluntness radius of 180 mm (20% of the fuselage radius of the initial configuration, almost
comparable with the equivalent radius of the present configuration) is found to significantly bring down the
peak heating, by an order of magnitude, while the increase in the nose wave drag is within 10%. A wing
leading edge with a radius of 16 mm resulted in substantial decrease in stagnation heating, while increasing
the leading edge wave drag by 15%. In addition the introduction of nose bluntness also has the advantage of
pushing the nose shock away from the wing tips; the bow shock profile was computed using semi-empirical
correlations
13
in order to monitor its position relative to the wing. Clearly the inclusion of bluntness is
inevitable in order to manage the issues encountered in hypersonic flows. Accordingly, the present vehicle
configuration consists of nose and leading edges with significant bluntness radii, and the drag due to the
bluntness are a part of the total drag. A more detailed aerothermodynamic analysis is presented in Section
III.C.
The inviscid coefficients are only dependent on Mach number and angle of attack and independent of
altitude (independence with Reynolds number). However, the contribution of skin friction which is dependent
of Reynolds number, leads to altitude dependence of the force coefficients. The skin friction drag coefficient
at zero incidence for the cone-cylinder body C
D0,f,b
and for the wing C
D0,f,w
(tail too has similar functional
form) are given by the following engineering correlations.
C
d0,f,b
= 0.053
`
d
M
q`
0.2
(8a)
C
d0,f,w
=
0.0266
(qc
max
)
0.2
(8b)
In the above equations q is the dynamic pressure and c
max
is the length of mean wing chord. The skin
friction drag coefficient is added to the inviscid drag coefficients at zero incidence (for each component). The
total drag coefficient at zero incidence together with the normal force coefficients are then used to calculate
the lift and drag coefficients due to each component using Equation 2. The explicit density and velocity
dependence of dynamic pressure leads to the altitude (and Reynolds number) dependence of the skin friction
drag coefficient implicitly. The altitude dependence of the drag at zero dependence is exemplified in Fig. 3
for the initial configuration. There is an obvious and significant increase in drag, particularly at subsonic
5 of 21
American Institute of Aeronautics and Astronautics
