Spike Contour Algorithm
Spike Contour Algorithm
Model Summary
The algorithm analyzes the flow through the supersonic section of the aerospike nozzle. It examines a
series of characteristics in the expansion fan which accelerates the flow. It finds the spike contour
required to maintain an equal mass flow though each characteristic. The underlying thermodynamic
model assumes that the ratio of specific heats 𝛾 and the molar mass of the working fluid are constant
through the nozzle, and that the expansion is adiabatic and isentropic. These assumptions greatly
simplify the calculations but introduce some error. The approach of the model is based on a paper
written by C.C. Lee for NASA’s Marshal Space Flight Center [1].
Relevant Requirements
List all system requirements which drive your design decisions here.
Algorithm Details
Definitions
The aerospike nozzle described here is cylindrically symmetric. The shroud is the outer solid boundary of
the nozzle. The spike is the inner solid boundary of the nozzle. The throat is formed by the corner of the
shroud (called the lip) and the spike. The exit plane of the nozzle contains the tip of the spike and is
orthogonal to the centerline of the nozzle [Figure 2].
Shroud Shroud
Throat
Throat Plane
𝑀=1
Exit Plane
Exit
𝑀 = 𝑀𝑒 > 1
Figure 2: Important features in the aerospike nozzle Figure 1: The expansion fan in the aerospike nozzle
Algorithm Inputs
The following parameters are inputs to the algorithm:
Alternatively, if the desired exit pressure is given as an input to the algorithm, the required area
expansion ratio can be found by Equation 3-25 in [3]:
1 1 𝛾−1
𝐴𝑒 𝛾 + 1 𝛾−1 𝑝𝑒 𝛾 𝛾 + 1 𝑝𝑒 𝛾
𝜖= =( ) ( ) √( )( 1 − ( ) )
𝐴𝑡 2 𝑝𝑐 𝛾−1 𝑝𝑐
𝛾−1
2 𝑝𝑐 𝛾
𝑀𝑒 = √ (( ) − 1)
𝛾 − 1 𝑝𝑒
Shroud Geometry
The shroud makes and angle 𝛿 with the horizontal. At the throat (𝑀 = 1), the velocity of the flow (𝑣⃗𝑡 )
is along the shroud surface and also make an angle δ with the horizontal [Figure 3: Shroud ].
𝛿
ℎ𝑡 𝜈𝑒
𝛿
𝑅𝑒 𝑣⃗𝑡
𝑣⃗𝑒
Figure 4: Turning angle through
the expansion fan
After the expansion fan has accelerated the flow to 𝑀 = 𝑀𝑒 , we want the velocity of the flow (𝑣⃗𝑒 ) to be
parallel to the centerline of the nozzle. As the initial and final Mach numbers of the expansion fan are
known (1, 𝑀𝑒 ), the angle through which the flow will be turned by the expansion fan 𝜈𝑒 can be found by
the Prandtl-Meyer relation:
𝛾+1 𝛾−1 2
𝜈𝑒 = √ tan−1 (√ (𝑀 − 1) ) − tan−1 (√𝑀𝑒2 − 1)
𝛾−1 𝛾+1 𝑒
The sum of the expansion fan turning angle and the shroud slope angle must be a right angle in order for
the flow to be parallel to the centerline after the expansion fan. Thus:
𝜋
𝜈𝑒 + 𝛿 =
2
𝜋
𝛿 = 𝜈𝑒 −
2
Now the ratio of the throat gap ℎ𝑡 to the shroud radius 𝑅𝑒 can be found, using equation 12 from [1]:
ℎ𝑡 𝜖 − √𝜖(𝜖 − sin(𝛿))
=
𝑅𝐸 𝜖 sin(𝛿)
Throat Conditions
As the throat Mach number is known to be 1, the fluid temperature, pressure, density and velocity can
be found at the throat:
𝛾 − 1 −1
𝑇𝑡 = 𝑇𝑐 (1 + )
2
−𝛾
𝛾 − 1 𝛾−1
𝑝𝑡 = 𝑝𝑐 (1 + )
2
𝑝𝑡
𝜌𝑡 =
𝑅 𝑇𝑡
𝑣𝑡 = √𝛾 𝑅 𝑇𝑡
𝑅𝑢𝑛𝑖𝑣𝑒𝑟𝑠𝑎𝑙
Where 𝑅 = 𝑚𝑚𝑜𝑙𝑎𝑟
is the specific gas constant of the fluid.
Spike Contour
A series of 𝑁 Mach numbers in the set [1, 𝑀𝑒 ) are examined. An expansion fan expands the flow
through an infinite number of Mach lines, which meet at the corner of the shroud. Each Mach number
𝑀𝑥 defines a particular Mach line which makes an angle 𝜙𝑥 = 𝜈𝑒 − 𝜈𝑥 + 𝜇𝑥 with the centerline [Figure
5]. 𝜈𝑥 is the Prandtl-Meyer angle at the current Mach number:
𝛾+1 𝛾−1 2
𝜈𝑥 = √ tan−1 (√ (𝑀 − 1) ) − tan−1 (√𝑀𝑥2 − 1)
𝛾−1 𝛾+1 𝑥
𝛾+1 𝛾−1 2
𝜈𝑒 = √ tan−1 (√ (𝑀 − 1) ) − tan−1 (√𝑀𝑒2 − 1)
𝛾−1 𝛾+1 𝑒
And 𝜇𝑥 is the Mach angle at the current Mach number 𝑀𝑥 . It is the angle between the flow velocity
direction and the local Mach line when the flow has expanded to 𝑀𝑥 .
1
𝜇𝑥 = sin−1 ( )
𝑀𝑥
Figure 5: Using a characteristic to define a point on the surface of the spike contour. Figure taken from [1]. Note that the
orientation of this figure is rotated 90° form the orientation of the other figures in this report.
The expansion fan is cylindrically symmetric about the centerline of the nozzle, so each Mach line in the
expansion fan can be revolved about the centerline to form a frustum. The flow is traveling at Mach
number 𝑀𝑥 when it passes through this frustum, and the side of the frustum makes an angle 𝜙𝑥 with
the centerline. The top of the frustum is set at the lip of the shroud and has a radius 𝑅𝑒 . The base of the
frustum is where the Mach line meets the contour of the spike, at a radius 𝑅𝑥 . The height of the frustum
is the axial distance from the shroud lip to the intersection of the Mach line and the spike contour, and
is 𝑋𝑥 . The side area of the frustum is [Equation 20 from [1]]:
𝜋 (𝑅𝑒2 − 𝑅𝑥2 )
𝐴𝑥 =
sin(𝜙𝑥 )
If we can find this 𝐴𝑥 , we can solve for 𝑅𝑋 , 𝑋𝑥 and fix a point on the spike contour. Note that mass is
conserved through the expansion fan. Therefore, during steady-state operation of the nozzle, the mass
flow through the throat 𝑚̇𝑡 must equal the mass flow though each Mach-line frustum, 𝑚̇𝑥 .
𝑚̇𝑡 = 𝑚̇𝑥
𝜌𝑡 𝑣𝑡 𝐴𝑡 = 𝜌𝑥 𝑣𝑥 𝐴𝑥
𝜌𝑡
𝜌𝑐 𝐴𝑡
𝐴𝑥 = 𝜌 𝑣
𝑥 𝑥
𝜌𝑐 𝑣𝑡 sin(𝜇𝑥 )
By equating the two equations for 𝐴𝑥 , 𝑅𝑥 can be found as a function of 𝑀𝑥 [Equation 26 from [1],
derivation in Appendix 1: Derivation of 𝑹𝒙 𝑴𝒙 Relation]:
𝛾+1
2 𝛾−1 2(𝛾−1)
𝑅𝑥 √ [(𝛾 + 1) (1 + 2 𝑀𝑥2 )] sin(𝜈𝑒 − 𝜈𝑥 + 𝜇𝑥 )
= 1−
𝑅𝑒 𝜖
𝑅𝑒 − 𝑅𝑥
𝑋𝑥 =
tan(𝜙𝑥 )
Using these equations, each Mach number defines a unique point on the contour of the spike. This point
is where the Mach line associated with that Mach number must intersect the spike in order to have
steady mass flow though the surface defined by that Mach line.
By examining a range of Mach numbers in the set [1, 𝑀𝑒 ), a set of points on the contour of the spike is
found. A spline is then fit through these points to define the spike contour.
It is computationally feasible to calculate enough points that the spike contour is defined to a precision
greater than that of available fabrication tools. For example, with a small engine (𝑅𝑒 = 15.0mm) and a
lathe with 0.0005in = 0.013mm precision, approximately 1200 points are required for the radial
precision of the model to exceed the tolerance to which the physical part can be manufactured.
𝛾 − 1 2 −1
𝑇𝑥 = 𝑇𝑐 (1 + 𝑀𝑥 )
2
Specific Impulse
This model can also be used to predict the specific impulse of the engine, given the thrust force
produced by integrating the spike pressure over the spike area down to some axial distance 𝑋𝑥 . This can
give a crude estimate of the 𝐼𝑠𝑝 loss due to truncating the spike at to a length 𝑋𝑥 . The thrust force 𝐹 is
found by:
𝑥
The first term is the momentum thrust at the throat, the second term is the pressure thrust at the
throat, and the third term is the pressure thrust along the spike from the throat to point 𝑥. Now divide
all terms by the mass flow to find the specific impulse:
𝑥
(𝑝𝑡 − 𝑝𝑎 )𝐴𝑡 sin(𝛿) (𝑝𝑥 ′ − 𝑝𝑎 )
𝐼𝑠𝑝 𝑥 𝑔 = 𝑣𝑡 sin(𝛿) + + ∫ d𝐴
𝜌𝑡 𝐴𝑡 𝑣𝑡 𝜌𝑡 𝐴𝑡 𝑣𝑡
0
Using thermodynamic relations, this can be rewritten as [Equation 30 from [1], note that 𝑝𝑒 should be
𝑝𝑐 ]:
𝑥
𝑣𝑡 sin(𝛿) 𝑝𝑎 𝑝𝑐 𝑣𝑡 𝑝𝑐 (𝑝𝑥 ′ − 𝑝𝑎 )
𝐼𝑠𝑝 𝑔 = 𝑣𝑡 sin(𝛿) + [1 − ( ) ( )] + ( ) ∫ d𝐴
𝑥 𝛾 𝑝𝑐 𝑝𝑡 𝛾 𝑝𝑡 𝑝𝑐
0
Rewriting the integral in finite difference form [Equation 31 from [1], note that 𝑝𝑒 should be 𝑝𝑐 ]:
𝛾
1 𝛾 + 1 𝛾−1 𝑝𝑎
𝐼𝑠𝑝 𝑥 𝑔 = 𝑣𝑡 sin(𝛿) [1 + (1 − ( ) ( ))]
𝛾 2 𝑝𝑐
𝛾 𝑁
𝑣𝑡 𝛾 + 1 𝛾−1 𝜖 𝑝(𝑥−1) − 𝑝𝑎 𝑝(𝑥) − 𝑝𝑎 𝑅(𝑥−1) 2 𝑅(𝑥) 2
+ ( ) ( ) ∑ [( )+( )] [( ) −( ) ]
𝛾 2 2 𝑝𝑐 𝑝𝑐 𝑅𝑒 𝑅𝑒
𝑥=1
For several sample cases, the specific impulse predictions by this method were found to be within 1% of
the predictions made by the Rocket Propulsion Analysis software tool. RPA was set to predict the specific
impulse at matched ambient pressure with the chemical composition of the fluid freezing at the throat.
𝜌𝑡
1 𝑅𝑥2 𝜌𝑐 1
(1 − 2 ) = 𝜌 𝑣 ( )
sin(𝜙𝑥 ) 𝑅𝑒 𝑥 𝑥 ) 𝜖
𝜌𝑐 𝑣𝑡 sin(𝜇𝑥
𝜌𝑡
𝑅𝑥2 𝜌𝑐 sin(𝜙𝑥 ) 1
2 = 1 − (𝜌 𝑣 ) ( )( )
𝑅𝑒 𝑥 𝑥 sin(𝜇𝑥 ) 𝜖
𝜌𝑐 𝑣𝑡
𝜌𝑡 1 1
𝜌𝑐 1 2 𝛾 − 1 2 𝛾−1 2 𝛾−1 2 2
(𝜌 𝑣 ) = [( ) (1 + 𝑀𝑥 )] [( ) (1 + 𝑀𝑥 )]
𝑥 𝑥 𝑀𝑥 𝛾 + 1 2 𝛾+1 2
𝜌𝑐 𝑣𝑡
1 1 𝛾+1
Note that 2 + 𝛾−1 = 2(𝛾−1)
:
𝜌𝑡 𝛾+1
𝜌𝑐 1 2 𝛾 − 1 2 2(𝛾−1)
(𝜌 𝑣 ) = [( ) (1 + 𝑀𝑥 )]
𝑥 𝑥 𝑀𝑥 𝛾 + 1 2
𝜌𝑐 𝑣𝑡
sin(𝜙𝑥 ) sin(𝜈𝑒 − 𝜈𝑥 + 𝜇𝑥 )
( )=
sin(𝜇𝑥 ) 1
𝑀𝑥
𝛾+1
2 𝛾−1 2(𝛾−1)
𝑅𝑥 √ [(𝛾 + 1) (1 + 2 𝑀𝑥2 )] sin(𝜈𝑒 − 𝜈𝑥 + 𝜇𝑥 )
= 1−
𝑅𝑒 𝜖
References
[1] C.C. Lee. “FORTRAN Programs for Plug Nozzle Design.” Prepared for the Advanced Propulsion
Section, Propulsion and Mechanics Branch, P&VE Division of the George C Marshall Space Flight
Center by Brown Engineering Company, Inc. March 1963.
[2] J. Majdalani and B. A. Maickie. “Explicit Inversion of Stodola’s Area-Mach Number Equation.”
Transactions of the ASME, Vol. 133, July 2011, pp. 071702-1 to 071702-7.
[3] G. Stutton. Rocket Propulsion Elements Seventh Edition, New York, NY: John Wiley & Sons, Inc.,
2001.
Subsection 2
A table: