Reusable Launch Vehicles (RLV)
With NASA that has set to itself the top priority of drastically reducing--by
a factor of 10--the cost of putting 1 lb of payload in orbit, more than
ever before are innovative, even revolutionary, designs of launch vehicles
being seriously considered. In particular, the Single Stage To Orbit (SSTO)
concept appears within reach thanks to such high efficiency engines as
air breathing scramjet engines and, more specifically, aerospikes rocket
engines. These two emerging technologies are embodied in the following
two concepts, which have been our benchmark examples over the past couple
of years:
-
The "Winged
Cone:" It consists of a conical forebody frustum, a cylindrical
scramjet engine nacelle all around the basic conic structure, and a pair
of delta wings.
-
The X-33, the reduced scale demonstration vehicle of the new space
shuttle, the VentureStar: This is a "lifting body" technology vehicle
equipped with aerospike rocket engines. The basic aerospike design is to
eject the gas resulting from an explosive mixture of oxydant and oxydizer
at the root of a "spike" profile, instead of ejecting them inside a gimballed
engine bell. The accrued efficiency stems from the fact that, in the aerospike
engines, the hot gas remain stuck to the "spike" profile for all flight
conditions, whereas for engine bell design the explosion is not optimally
located relative to the bell profile for all flight conditions. The engine
of the X-33 are more specifically of the "linear aerospike" type in the
sense that the engine consists of modules assembled in a linear array with
its section only retaining the shape of a spike. Differential thrusting
of the upper and lower modules allows for longitudinal control while differential
thrusting of right and left modules allows for lateral control.
The winged cone has served as testbed for several concepts gravitating
around the overall idea that, during the hypersonic part of the launch,
good longitudinal control requires a certain amount of decoupling between
phugoid and short periodic modes. This problem was initially viewed as
a Shapiro eigenstructure assignment, which was later optimized relative
to an H-infinity criterion using genetic algorithms. More recently, the
same problem was reformulated as a controllability subspace problem and,
most recently, as an algebraic geometry problem. In this latest formulation,
decoupling is viewed as the problem of maximizing the distance between
two points subject to linear constraints on the G(2,4) Grassmannian of
2-D subspaces in 4-D space. As a warm up excercise for the X-33, an input
representing differential thrust between upper and lower scramjet engine
modules was introduced, and pitch control by a combination of elevon and
propulsion was investigated. In particular, while most of our attention
in the PCA problem focused on an H-infinity matching of the elevon response
by compensated differential thrust, here the matching has been viewed as
a geometric problem. The geometric design on the winged cone has proved
feasible.
In the X-33, because of its very limited aerodynamic control mandated
by the avoidance of hot spots during reentry and its aerospike engine technology
that does not allow for gimballing, the necessity to control the attitude,
at least in part, by propulsion is more heavily felt. Some preliminary
studies of propulsion control have been conducted. However, a more substantial
part of our effort has been devoted to reconfigurable control using
Linear
Set Valued Dynamically Varying (LSVDV) concepts inspired from our research
on controlling nonlinear dynamics. The problem is basically viewed as that
of tracking the nominal launch trajectory. The tracking error is linearized
around every point of the nominal launch trajectory, resulting in the tracking
error modeled as a linear system with its parameters generated by the nominal
(dynamically varying) trajectory. If failure is allowed, then the nominal
trajectory is no longer certain, but its dynamical evolution can nevertheless
be narrowed down to within some uncertainty set, resulting in the
Set
Valued Dynamically Varying aspect of the linear system describing the
error. Early simulation results have shown significantly improved performance
over the more conventional Linear Parametrically Varying (LPV) concept.
Selected publications
-
E. Jonckheere, "From Sioux-City to the X-33," in Proceedings of the
Conference on System Structure and Control, International Federation
of Automatic Control (IFAC), Ecole Centrale, Nantes, France, July 08--10,
1998, pp. 609--620.
-
E. Jonckheere, P. Lohsoonthorn, and S. Bohacek, "From Sioux City to the
X-33," Annual Reviews in Control, Elsevier/Pergamon, vol. 23, pp.
91-108, 1999.
-
P. Lohsoonthorn, S. Dalzell, and E. Jonckheere, "Eigenstructure versus
constrained H-infinity design for hypersonic winged cone, " Journal
of Guidance, Control, and Dynamics, American Institute of Aeronautics
and Astronautics (AIAA), Volume 24, Number 4, pp. 648-658, July-August
2001.
-
E. A. Jonckheere and P. Lohsoonthorn, "A geometric approach to model matching
reconfigurable propulsion control," American Control Conference (ACC'2000),
Chicago, IL, June 28-30, 2000, Session TM-14, pp. 2388-2392.
For more information about this project, please, contact Stephan K.
Bohacek at bohacek@eecis.udel.edu
and/or Poonsuk "Matt" Lohsoonthorn at lohsoont@hotmail.com.
This research was supported by the National Aeronautics and Space Administration
(NASA) through Marshall Space Flight Center contract NAS8-97292.
Recent development
Unfortunately, after a rocky development plagued by control and fuel tanks
problems, the X-33 project was canceled by NASA.
Other sites of interest
-
Sea Launch
-
The Energia Space Corporation
of the legendary Sergei Korolev
-
International Launch Services
-
The Russian Proton
rocket
-
The Buran shuttle of the
Molniya Corporation