Benefits of the Star Grain Configuration for a Sounding

Rocket
Steven D. Stein*
United States Air Force Academy Department of Astronautics, USAFA, CO, 80841

Given a particular propellant formulation, the performance of a solid rocket motor is

due largely to two factors: first, its volumetric efficiency and second, its average pressure
over the course of the burn. The problem with the internal burning tube grain design is not
enough surface area exists at the beginning of the burn. This causes the pressure to rise to
very high levels towards the end of the burn and creates a potential flow separation problem
at start-up. Additionally, an internal burning tube design creates a progressive burn; which
is exactly opposite of the ambient pressure curve for a sounding rocket trajectory,
decreasing efficiency. Using a star design to add more initial burn area allows for greater
start-up pressure and a more neutral-regressive burn; which follows the ambient pressure
curve more precisely. Additionally, the star design increases volumetric efficiency;
increasing total impulse. The additional initial burn area also keeps the pressure in the
chamber at a constant level, which allows for a higher average pressure during the burn
which increases efficiency. The star grain configuration will also increase the range of
possible payloads a sounding rocket can carry. This discussion will present the advantages
of using a star grain configuration to improve the Air Force Academy’s sounding rocket’s
altitude performance.

Nomenclature
L1 = star parameter describing depth of star
L2 = star parameter describing width of star
Ls = length of the star along the main propellant grain
W = web (amount of propellant burned in the perpendicular direction)
m = slope of a line
b = y-intercept of a line
“n” = the line designated “n” (line “1”, “2”, etc.)
Өn = angle between line “n” and y-axis
R = radius of the grain (the point where the star attached itself)
ALS = area along the length of the star (present along the main propellant grain)
Atri = area of the triangle portion of the aft end of the star pattern
Aellipse = area of the ellipse portion of the aft end of the star pattern
Ө = arc length representing the area lost from the addition of a star pattern
%gone = percentage of area lost from the addition of a star pattern
La = length of the aft section of a double taper grain design
Ra = radius of the aft section of a double taper grain design
Rcase = radius of the inside of the case
Isp = specific impulse

I. Introduction

T HE main goal of a star grain configuration is to add more initial surface area. An increase in initial surface area
is highly desired for a variety of reasons. Some reasons are safety (flow separation), performance (Isp), and
payload carrying capacity (max g’s). However, increasing the initial burn area usually requires sacrificing
propellant to expose more surface area. The star grain configuration will increase the initial burn area without
sacrificing volumetric efficiency; and in some cases will actually increase the volumetric efficiency. The star design

*
Cadet, Astronautical Engineering Department, PO Box 2626, USAF Academy, CO, 80841, AIAA Student Member
1
American Institute of Aeronautics and Astronautics
can also produce a neutral-regressive burn profile. As the star burns, its surface area decreases as the outer portion
of the grain expands, exposing more surface area. This trade off results in a neutral burn profile. Combined with a
double taper grain design, the later portion of the burn becomes regressive. The neutral-regressive burn is much
more desirable when compared to a completely regressive burn profile. In the end, the star grain configuration will
out-perform all grain designs not containing a star pattern.

II. Effects of Increasing Initial Burn Area

Low initial burn area can create many problems which can lead to total mission failure. A great example of a
grain with low initial burn area is the internal burning tube grain.

A. Internal Burning Tube Grain Design

When using an internal burning tube design, there is little surface area available to produce pressure inside the
case. This can lead to ignition problems and flow separation in the nozzle. For certain propellants, a minimum
start-up pressure is required for a clean ignition. The propellant used in the FalconLAUNCH V program requires a
start-up pressure of 3.1 MPa (450 psi). If an internal burning tube design is used with the FalconLAUNCH V
propellant, the throat of the nozzle would have to be reduced significantly to account for this, causing pressure
related problems. Additionally, there are nozzle concerns due to heating and throat erosion. Knowing the initial
start-up pressure is fixed, the resulting maximum pressure is very high due to the progressive nature of the internal
burning tube. Fig. 1 provides the pressure data 20
for an internal burning tube design.
Pressure (MPa)

The maximum pressure for the internal

15
burning tube design rises to very high levels.
If the pressure is too high, the pressure vessel
10
could crack or break resulting in total mission
failure. Thus, the internal burning tube was
not considered a viable option for the 5
FalconLAUNCH program.
Another safety concern is flow separation 0
1
in the nozzle. In 1954, Summerfield 0 1 2 3 4 5 6 7 8
performed a series of experiments analyzing Time (sec)
flow separation in over expanded supersonic Figure 1. Internal Burning Tube Chamber Pressure Profile.
nozzles. Summerfield’s experiments showed Start-up: 3.1 MPa; Max pressure: 18.9 MPa
that as the ambient pressure increased with
respect to the exit pressure, shockwaves present outside the nozzle would begin to move inside the nozzle. This
does two things: the effective exit diameter becomes smaller and the shockwaves inside the nozzle could cause
damage. In 1984, Oates, by analyzing Summerfield’s experiments, discovered flow separation in the nozzle occurs
when the ambient pressure is 2.5 - 3 times as large as the exit pressure. A good “rule of thumb” is to use 2.75 for
this pressure ratio between ambient and exit pressure. This means that for a sea level launch (ambient pressure =
101.7 kPa) the exit pressure must be above 37 kPa to avoid flow separation. If a nozzle were designed to maximize
the total impulse for a sea level launch using the internal burning tube design, the expansion ratio is about 13.4.
Keeping in mind the start-up chamber pressure for this design is 3.1 MPa, the start-up exit pressure becomes 25.8
kPa. Because this number is below the 37 kPa threshold, flow separation will occur in the nozzle for the internal
burning tube design. To avoid this problem, either the exit diameter would have to be decreased (decreasing
efficiency), the start-up pressure would need to increase (increasing the already large maximum chamber pressure),
and/or the launch needs to occur at a location higher than sea level (limiting the choice of launch locations).

B. Increasing Initial Burn Area

An increase in initial burn area facilitates the minimum startup pressure requirement. However, increasing the
initial burn area without sacrificing volumetric efficiency is quite a challenge. Getting the burn area to stay nearly
constant or decrease slightly over the course of the burn would provide optimum performance. The
FalconLAUNCH program realized these problems. The solution proposed is a double taper grain design to allow
for more initial burn area.

Figure 2. Side view of Double Taper Design

2
American Institute of Aeronautics and Astronautics
 The aft section of the grain is enlarged with 14
respect to the rest of the grain. This design
achieves two important objectives: one, increasing 12

the start-up pressure and, two, decreasing

10
maximum pressure due to the higher initial burn

Pressure (MPa)
area and lower maximum burn area. This 8
eliminates both the flow separation and the
maximum pressure problems present with the 6

internal burning tube design. Fig. 3 shows the

4
pressure profile for a double taper grain design
using the same nozzle throat size as the internal 2
burning tube grain design produced in Fig. 1.
The double taper design sacrifices total 0
0 1 2 3 4 5 6 7 8 9
impulse and efficiency. The volumetric efficiency
Time (sec)
of a grain design corresponds to the total impulse
produced by the rocket. The double taper gains an Figure 3. Double Taper Chamber Pressure Profile.
initial burn area increase by sacrificing propellant Start-up pressure: 5.25 MPa; max pressure: 13.0 MPa
in the aft section. The loss in propellant results in
a loss in total impulse. Additionally, the double 120
taper design produces less average chamber
pressure. Average chamber pressure is directly 100

related to specific impulse (Isp). Accounting for

volume and efficiency losses, the loss in total 80
Pressure (kPa)

impulse is quite drastic. Using the examples

60
above, the total impulse of the double taper design
is 6.5% less than the internal burning tube design.
40
However, the double taper design is a more
feasible option because of the significant drop in
20
maximum chamber pressure due to pressure vessel
restrictions. 0
An increase in initial burn area will prevent 0 2 4 6 8 10
flow separation. Using the example found in Fig. Time (sec)
3, a nozzle optimized for maximum total impulse Figure 4. Ambient Pressure Curve for a
has an expansion ratio of 8.9. The initial exit Sounding Rocket Trajectory
pressure using an initial chamber pressure of 3.1
MPa is 44.6 kPa. The exit pressure is above the 37
kPa threshold needed to avoid flow separation in
the nozzle.
Aside from flow separation prevention and
maximum pressure decreases, the initial burn area
increase adds many important benefits. First, the
range of payloads the sounding rocket can carry
increases. The maximum pressure in the case
corresponds to maximum thrust and subsequently
maximum acceleration. Certain payloads can only
experience so many g’s to function properly. By
reducing the maximum acceleration, a larger range
of payloads can be carried by the sounding rocket.
Second, the burn profile becomes more
neutral/regressive in nature. The ambient pressure
for a sounding rocket trajectory is regressive Figure 5. Desired Change in Chamber Pressure Profile
(shown in Fig. 4). For maximum efficiency and
altitude, the thrust profile must match the shape of the ambient pressure curve as precisely as possible; another
reason to avoid the internal burning tube design since its thrust profile is exactly opposite of Fig. 4. Third, and most
importantly, the average pressure in the case increases. Increasing the start-up pressure in the chamber, without
affecting the maximum pressure, will increase the average pressure in the case. Since average pressure corresponds
3
American Institute of Aeronautics and Astronautics
to Isp, increasing the initial burn area will increase the specific impulse of a rocket. Fig. 5 shows the desired effect
of increasing the burn area.
The addition of a star design will achieve the desired pressure profile
without sacrificing propellant. Fig. 6 provides an example of what a star design
will look like when looking down the aft-end of the grain. Notice how the star
protrudes inward while adding extra surface area. The protrusion of the star
adds propellant to the case rather than sacrificing it like the double taper. As the
burns progresses the points of the star will burn away while the grain itself
expands. This results in a neutral burn. Combined with the double taper design,
a neutral-regressive burn can be produced to maximize the efficiency and
improve the altitude of a sounding rocket. It will be demonstrated later that the
star grain configuration not only adds more initial surface area than the double
taper, but also out-performs the internal burning tube grain design while not Figure 6. Star Example
exceeding pressure limits.

III. Star Propagation

To successfully create a model of the star grain configuration, the star must be propagated through time.
Because the exact burn rate of a particular propellant is not constant, the star design must propagate as a function of
web. Once the star has been successfully propagated, the burn area can be calculated.

A. Propagate the Star Through web

First, set up a definition system for each point of the star. Fig.
7 shows a four pointed star example.
Each star point is comprised of 3 points and two lines. The
points are designated by number (1, 2, 3, etc.) and each line is
designated with quotation marks (“1”, “2”, etc.). The reason lines
are used is for an easy propagation scheme (described later). Next
is a key definition scheme: the L1 and L2 parameters. L1 defines
how deep the point drives into the grain and L2 defines the width
of the star point. These two parameters will define the size and
shape of each point of the star and will also be used later to
determine the burn area of the star. The W Fig. 7 is the web
distance burned. Note the web always burns perpendicular to any
surface on the grain. The outside of the circle is the grain to
which the star is attached. Figure 7. 4-pointed Star with definitions.
The L1 and L2 parameters will be used to determine the initial
parameters of the star. Each line will be defined based on a set coordinate system and propagated through web.
Once the location of each point has been determined, from the location of each line, the L1 and L2 parameters will
be re-calculated to calculate a burn area. This will continue until the star design has completely burned away.
For this illustration, only one star point will be propagated. Once
modeled, the burn area is just the remainder of star points multiplied by the
area discussed here. This star point is the point comprised of points 1, 2,
and 3 and lines “1” and “2”. Before any math can be involved, an axis
system is introduced. The origin is at the center of the star or down the
center of case as shown in Fig. 8.
The propagation scheme will utilize the two lines moving together and
the outside circle moving outward as a function of web. The intersection of
each line with the edge of the circle will denote points 1 and 3 and the
intersection of the two lines will denote point 2. To accomplish this, three
Figure 8. One point with axis system
parameters are needed. The first two are the slope and y-intercept of a line
and the third is the radius, R, of the circle. Each line is defined by its slope, m, and its y-intercept, b. Actual
equations for each line’s respective m and b will come later.
Each point is determined using a system of equations and solving for their respective x and y coordinates. First,
the location of points 1 and 3: these points are located at the intersection of a line and the circle. Eq. (1) represents
the equation of a line and Eq. (2) is the circle with radius R.

4
American Institute of Aeronautics and Astronautics
 y = mx + b (1)

y = R2 − x2 (2)

Set (1) and (2) equal to each other and solve for x. With the aide of the quadratic equation and simplifying
techniques, x becomes:

− 2mb ± m 2 R 2 − b 2 + R 2
x= (3)
2m 2 + 2

Eq. (3) is a very important equation since it defines the most complex point to determine. The +/- sign can be a
little tricky. To determine when each sign should be used, look at Fig. 8. “-“ is used for point 1 and “+” is used for
point 3. y is determined by using Eq. (1) now that x is known.
Point 2 is determined by solving both Eqs. (4) and (5) (intersection of lines “1” and 2”):

y = m1 x + b1 (4)

y = m2 x + b2 (5)

Solving for the x component yields:

b2 − b1 (6)
x=
m1 − m2

-y is found using either Eqs.(4) or (5) by subbing in the x value found in Eq. (6).
The slope (m) for each line is fairly easy to determine through inspection. The equations are as follows:

L1
m1 = − (7)
L2

L1
m2 = (8)
L2
The intercept (b) is slightly more complicated. This requires both
inspection and mathematical manipulation. Use the equation of each
respective line, Eq. (4) and Eq. (5), and circle, Eq. (2), to compute. Using
inspection, x is equal to –L2 for point 1 +L2 for point 3. Solve the
equations simultaneously to find each line’s respective b value.

b1 = R 2 − L 2 2 − m1 (− L 2) (9)

b 2 = R 2 − L 2 2 − m 2 ( L 2) (10)

The next step is to propagate each line as a function of web. To

accomplish this, it is assumed the line will move along a line normal to
itself. The slope of the line will remain the same while the y-intercept will
change. Fig. 9 gives a depiction of how this change in b, ∆b, is determined
using geometry.
The web distance burned, W, forms a right triangle with the initial
“1” line, the future “1” line, and the y-axis. Geometry can be used to Figure 9. Determination of ∆b
5
American Institute of Aeronautics and Astronautics
determine the distance between the old b and the new b (∆b). Since the slopes of both lines are the same, the angles
symbolized as 1 are exactly the same.

θ1 = tan −1 m1 (11)

 L2 
θ1 = tan −1   (12)
 L1 

Solving for ∆b:

W (13)
∆b =
sin(θ1 )

Eq. (13) can be used to propagate both lines “1” and “2”. The equation
makes sense theoretically since the y-intercept of both points will move at
exactly the same rate and move up (positive y-direction). The star decreases
in size over time and will eventually disappear as it moves closer to the main
grain. In addition, the intersection of the two lines should not deviate left or
right, but should move straight up the y-axis as web is increased.
The simulation of the points is complete. All the initial m’s and b’s have
been determined as well their rate of change. By using web, the new slope
and y-intercepts of lines “1” and “2” can be determined. The new R is
updated using a separate simulation used to propagate the main propellant
grain since the radius, R, will change at a different rate depending on what
type of grain geometry is used. Once the new m, b, and R have been Figure 10. First Star Point
determined, the points 1, 2, and 3 can be found by using Eqs. (3) and (1) to Propagation
find the x and y coordinates, respectfully. Through the use of Microsoft
Excel, the following simulation of points 1, 2 and 3 propagated through web is shown in Fig. 10.
The point is propagated exactly as expected. The outsides of the point (points 1 and 3) converge on the y-axis
at exactly the same rate and point 2 flows up the y-axis. The whole star point diminishes in size while burning along
with the main propellant grain.
Now that a working simulation of the star points is verified, the new L1 and L2 points can be determined.
Looking to Fig. 8, the L1 and L2 points can be found by inspection.

L1 = y1 − y 2 (14)

L 2 = x3 (15)

The subscripts of each x or y correspond to its respective point (i.e. y1 = the y component of point 1) on Fig. 8.
This same process can be used to determine the propagation scheme for the remaining three star points, but each
point will propagate exactly the same since the same L1 and L2 parameters will be used for each star point.

B. Burn Area Calculations

There are two areas to be calculated in this section. The first is an area calculation for a double taper grain
design and the second is for an internal burning tube design. The double taper design is the current grain design for
the FalconLAUNCH V motor. Thus, the star design was integrated into the double taper design for comparison
purposes. Additionally, an internal burning tube star pattern was simulated as another design option.
There are three main components to compute to determine the total area of the burn for a star design. First is
the area along the side of star which exists along the main propellant grain. Second is the aft-end section of the star,
assuming one aft end is burning along with the main section. Third is an area loss since the star will be taking the
place of a once present section on the main propellant grain (this is the area between points 1 and 3 in Fig. 8).
There are three key parameters needed to determine the area of the burn the star design creates. These are the
L1 and L2 parameters as well as the length of the star which runs along the main propellant grain, denoted as Ls.
6
American Institute of Aeronautics and Astronautics
For the double taper design, the star will only be present in the aft section of the double taper since the forward
section does not have enough open space for a star design to really add any performance. Also, the L/D for the
forward section is already close to 20, which is a rule of thumb to mitigate the risk of erosive burning. The addition
of a star pattern in the forward section could push the L/D over 20, causing erosive burning. Fig. 11 gives a visual
of where the star exists in the double taper grain design.

Figure 11. Double Taper with addition of star.

Location of star indicated by red shading

For the internal burning tube grain design, the star will be present for the entire length of the grain. Fig. 12
gives a depiction of the placement of the star for the internal burning tube design.

Figure 12. Internal Burning tube with addition of star.

Location of star indicated by red shading

Since L1 and L2 parameters have already been determined by using the propagation scheme mentioned earlier,
the only remaining parameter needed is Ls. The Ls for the internal burning tube design is simple: it is the same as
the length of the grain (substitute the length of the grain from the main simulation for Ls). The Ls for the double
taper design is much more complicated. As the star burns, the Ls with respect to the aft length of the main grain
becomes smaller. Compare Fig. 13 to Fig. 11 to visualize this.

Figure 13. Double Taper with star addition after a certain burn duration.
hypothetical, not actual burn simulation

The Ls term here is significantly less than the aft section of the grain. To find this new Ls term, another
determination of points method similar to the one used in the star propagation will be used. The line that is Ls will
move towards the grain as the burn progresses. Therefore, the point where the line that is Ls intersects with the main
propellant grain is the point used to determine Ls. This point is defined as yn2 to be distinguished from y2 mentioned
in the star propagation. The same process used earlier will be used again, except this time the x-coordinate is
already known. The y2 coordinate from the star propagation simulation is the x-coordinate used in Eq. (16). The m
and b have already been determined in the main propellant simulation; they are the m and b which describe the line
of the aft section of the grain for the double taper design.

y n 2 = m a y 2 + ba (16)

-Where: ma => slope of the aft section line

ba => y-intercept of the aft section line
y2 => y-coordinate of point 2 from Fig. 2
Once this point is determined, subtract it from the aft end of the main propellant grain to find the Ls term.

Ls = y n 2 − y a (17)

-Where: ya => location of furthest aft section of the grain

7
American Institute of Aeronautics and Astronautics
 The first section to be calculated is the long section of the star which runs along the side of the main propellant
grain. This is the section of the star visible in Fig.’s 11, 12 and 13. First, the calculation for the double taper design.
By looking at Fig. 11, the area along the long section of the grain forms a triangle. The short end of this triangle is
line “1”mentioned in the star propagation section (reference Fig. 8). The long section of the triangle is Ls. Because
of the choice to use the parameters L1 and L2 to describe the shape of the star, the area calculations become
simplified. The length of line “1” in Fig. 2 is shown in Eq. (18). Knowing the area of a triangle is ½bh, the area of
the long section of the star is Eq. (19).

"1" = L12 + L 2 2 (18)

1
ALS = L12 + L 2 2 Ls (19)
2

This equation describes the burn area for only one section of one star point. There are two sections of burn area
per star point. Therefore, Eq. (19) must be corrected to factor in the additional star points and sections. The term
cuts will be used to designate how many star points exist in the star pattern.

ALS = cuts * L12 + L 2 2 Ls (20)

-Where: cuts => number of star points in the star pattern

The internal burning tube design is the same, except the area is a rectangle eliminating the ½ term in Eq. (19).

ALS = 2 * cuts * L12 + L 2 2 Ls (21)

Next is the end burning section of the star. This area calculation is
fairly intuitive looking at Fig. 14.
There are two sections to this calculation: the first is the triangle
(colored gray) and the second is the ellipse (colored red). Again, because
L1 and L2 were chosen to define the star point, the area calculations are
simple.
The area of the triangle is very simple using the L1 and L2
parameters. The area of half of the triangle section shown in Fig. 8 is
computed using Eq. (22) while the entire end section burn is in Eq. (23): Figure 14. Aft-end section of star

1
Atri = ( L1)( L 2) (22)
2

Atri = cuts * ( L1)( L2) (23)

The ellipse can be a bit difficult. The area of an ellipse is

pi*(ab) and is depicted in Fig. 15. The difficult part is finding a
and b. a is found by inspecting Fig. 8 and comparing that to the
definitions found in Fig. 2 . a is L2. b is a not as intuitive. Since
the y-coordinate of point 1 (Fig. 2) is known and the radius of the
circle, R, is known, the difference between them is b. b = R – y1.
Figure 15. Area of an ellipse definition
The area of the ellipse represented in Fig. 8 is only half of an
ellipse and each star point has one ellipse section. Therefore, the
final equation becomes:

8
American Institute of Aeronautics and Astronautics
 cuts
Aellipse = L 2( R − y1 ) (24)
2
Next, the loss of area. The goal is to find the arc length lost and convert this
into a total percentage lost along the grain. Fig. 16 provides the relation between
the arc length and the parameters L1 and L2.
To determine the arc length, the unit circle is used to find the actual theta
which defines the amount of area lost. Using a scaling factor to size radius of the
circle to 1, the following arc length is determined using L2 and R:

 L2 
θ = sin −1   (25)
 R 
Figure 16: Arc Length
There are two of these thetas per star point and there are 2*pi radians in a Determination
circle. Dividing all the thetas added together by 2*pi will yield an area loss. For
the double taper, do not forget the star is only present in the aft section of the grain. This means there is a ½ factor
which needs to be accounted for because the star dwindles down to nothing at the middle part of the grain.
Additionally, as the star burns away, the Ls term becomes smaller than the aft length of the grain. This makes the
percentage lost along the main grain even less by a factor of Ls/La (La being the aft length of the grain). Eq. (26)
describes the double taper area loss whereas Eq. (27) will describe the area loss for the internal burning tube grain
design.

cuts * θ Ls (26)
% gone = *
2π La

cuts * θ Ls (27)
% gone = *
π La

IV. Results
After the propagation and area calculations are complete it is time to add the star to a burn simulation and
analyze the results. To add the star design to the burn simulation, simply add the additional burn area a star adds to
the overall burn area. The two key parameters analyzed will be volumetric efficiency and average chamber
pressure, which is related to specific impulse. The overall goal is to maximize total impulse and subsequently
altitude.
The baseline grain design used for comparisons is the flight design for FalconLAUNCH V. The flight design is
a double taper to take advantage of the initial burn area while not exceeding maximum pressure limits. Three
different star grain configurations will be used to compare performance with this baseline grain. The first is an
internal burning tube design with a 10-point star pattern embedded throughout the length of the grain. The second
two are double taper designs with a star embedded into the aft section of the grain (reference Fig. 11). The first of
the double taper/star designs is the same as the baseline motor, but with a 10-pointed star embedded designed to
maximize total impulse. The second double taper/star design is the same as the baseline motor, but the aft radius has
been increased by .3 in. A 10-point star is then added and optimized to maximize total impulse. The three grain
designs will be compared against each other to determine which thrust profile obtains the most altitude. Fig. 17
shows the thrust profiles of the baseline motor and the three star designs. One key thing to notice is the green plot is
a neutral-regressive burn. It was mentioned earlier why a neutral-regressive burn is desired and that a star grain
configuration could achieve a neutral-regressive burn. This plot verifies the possibility of a neutral-regressive burn
using a star pattern.
The lowest performing of the three star grain configurations is the internal burning tube grain design. The two
double taper designs out-performed the internal burning tube star grain in every major category. Table 1
summarizes the performance of each grain design while Table 2 summarizes the flight predictions.

9
American Institute of Aeronautics and Astronautics
 25000

20000
Thrust (N)

15000

10000

D o u b le T a p e r B a se lin e
D o u b le T a p e r B a se lin e - 1 0 - p o in t sta r
5000
N e u tra l-R e g re ssiv e - 1 0 -p o in t sta r
In te rn a l B u rn in g T u b e - 1 0 -p o in t s ta r

0
0 2 4 6 8 10
T im e ( s e c )
Figure 17: Thrust profiles for the Four Grain Designs

Table 1. Performance Comparisons of Star Grain Configurations

Grain Design Vol Efficiency (%) Isp vac (sec) Total Impulse (N-sec)
Baseline 85.71 266.03 120054
Double Taper Baseline – 10 point star 91.65 269.44 129940
Neutral-Regressive – 10-point star 91.22 269.86 129555
Internal Burning Tube – 10-point star 88.70 269.38 127448

Table 2. Flight Predictions of Star Grain Configurations

Grain Design Max Pressure (MPa) Altitude (km) Max g’s
Baseline 7.76 64.57 26.704
Double Taper Baseline – 10 point star 9.32 89.89 28.423
Neutral-Regressive – 10-point star 9.39 91.90 25.140
Internal Burning Tube – 10-point star 8.33 67.89 71.833
The flight results were generated using the TAOS program using an input thrust profile and mdot simulation.
Fig. 18 and Fig. 19 give a depiction of what the star pattern will look like with dimensions added.

Figure 18: Double Taper Baseline – 10-point Star Figure 19. Neutral-Regressive Design
L1 = 1.05 in; L2 = .62 in; Ra = 2 in; Rcase = 3.14 in
L1 = 1.34 in; L2 = .71 in; Ra = 2.3 in; Rcase = 3.14 in

10
American Institute of Aeronautics and Astronautics
 The neutral-regressive grain design produced the best flight performance of any grain design; proving it is the
most desirable profile to have. Not only did the neutral-regressive burn generate the most altitude, but also endured
the lowest max g load. This is due to the neutral-regressive burn giving most of its thrust in the early stages of the
burn, since the most mass exists in the rocket at the beginning of the burn. In direct opposition, the internal-burning
tube design performed poorly compared to the neutral-regressive design. The internal burning tube grain design
produced similar performance numbers when compared to the neutral-regressive. However this is a sounding rocket
going to altitude, so the results that really matter are the flight results. The neutral-progressive nature of the internal
burning tube is in exact opposition of the desired neutral-regressive profile. Not only did the internal burning tube
design fail to make any significant improvements in altitude, but the maximum acceleration felt by the rocket is
catastrophic. The internal burning tube design generated more than 70 g’s of acceleration. Not only does this mean
a payload can’t be flown, but the rocket may not even structurally survive the launch without heavier materials to
reinforce the structures. The star grain configuration will not only allow the rocket to survive the launch, but will
expand the range of possible payloads by lowering the maximum accelerations produced while maximizing altitude.
If max g’s are a significant problem, the star design produces so much initial pressure due to the initial burn
area present that the throat can be expanded to reduce the total max pressure inside the case. While this will
decrease efficiency due to average pressure loss, the maximum pressure and max g load have been decreased
significantly. Also, a
nearly neutral burn is also 12000
possible to bring the 10000
Thrust (N)

maximum pressure down 8000

even further. Fig. 20 is a
6000
plot of a low g star option.
This design only gets to 4000
61 km, but the max g load 2000
is only 18.1. The star 0
grain configuration is very 0 2 4 6 8 10 12
versatile in terms of the Time (sec)
types of thrust profiles it
can produce to achieve a Figure 20. Low g thrust profile
mission.
The star grain configuration also avoids flow separation in the nozzle. It was motioned earlier the exit pressure
must be above 37 kPa to avoid flow separation. Using the neutral-regressive design with an ideal nozzle and an
initial chamber pressure of 9.39 MPa, the initial exit pressure is 89.7 kPa. This is easily above the 37 kPa
requirement. In addition, the minimum exit pressure throughout the burn is about 50 kPa at the end of the burn. The
star grain configuration avoids flow separation in the nozzle for the duration of the burn.

V. Conclusion
The star grain configuration gains all the benefits of higher initial burn area without the drawback of sacrificing
volumetric efficiency. In most cases, the volumetric efficiency is increased when using a star grain configuration.
Also, the star pattern is able to achieve a higher Isp because of a higher average chamber pressure throughout the
burn. An increase in both volumetric efficiency and Isp leads to a very large increase in total impulse. From a
safety perspective, the star pattern prevents flow separation in the nozzle and reduces the maximum chamber
pressure to avoid pressure vessel failures. The reduction in maximum chamber pressure also reduces the maximum
acceleration of the rocket; increasing the availability of possible payloads. The final benefit of the star grain
configuration is the possibility of a neutral-regressive burn. The neutral-regressive burn profile follows the ambient
pressure curve more precisely to achieve maximum efficiency. All these benefits added together maximize the
performance and flight characteristics of a sounding rocket.

References
1
Humble, R. W., Henry, G. N., Larson, W. J., Space Propulsion Analysis and Design, 1st ed. – revised, The McGraw-Hill
Companies, Inc., New York, 1995, pp. 115-117.

11
American Institute of Aeronautics and Astronautics