Solid Rocket Boosters Separation System Development for the ILR-33 Amber 2K Rocket
, et | 13 sept. 2023
Article Category: Research Article
Publié en ligne: 13 sept. 2023
Pages: 16 - 27
Reçu: 26 sept. 2022
Accepté: 29 mai 2023
DOI: https://doi.org/10.2478/tar-2023-0014
Mots clés
© 2023 Jan Kierski et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Suborbital rockets are being widely used since the beginning of space exploration as a cheap and effective way of carrying out atmosphere sounding, suborbital experiments and various scientific research activities, as well as technology maturing and validation platforms [1,2]. In terms of design, we can divide them into single stage and multistage constructions. Two main multistage designs have been developed: serial staging, with stages mounted one on top of another, and parallel staging, where stages are attached alongside each other. The use of parallel staging in small suborbital rockets is not common, but has a number of advantages over serial staging, such as a significant length reduction and a greater acceleration, which leads to a reduction of wind sensibility at launch.
ILR-33 AMBER is a two-stage suborbital rocket, developed by Łukasiewicz Research Network, Institute of Aviation, and has been in deployment since 2014. Its propulsion system consists of a hybrid rocket motor that uses high-test peroxide (98%) as an oxidiser and two strap-on solid rocket boosters (SRBs) [3,4]. Since 2019, a new 2K version of the rocket is under development, with a series of modifications, including larger SRBs [5,6]. An improved SRBs separation system also had to be implemented, to fulfil design and safety requirements. The purpose of the separation system is to transmit loads between the main core and the SRBs and to safely discard the strap-on boosters after their burnout. A comparison of both versions is presented in Figure 1.
Figure 1.
Comparison of both versions of the ILR-33 AMBER rocket: top – original design; bottom – 2K version.
Design and simulation of separation of two bodies in relative motion in the atmosphere is a task of high complexity. Investigations on the aerodynamics of separation were conducted for several years, especially by Meakin [7] for the space shuttle. Dynamic models of separation for different designs were also analysed by Lochan et al. [8,9] and Seongjin et al. [10]. They conducted numerical analyses of separation dynamics with an integrated unsteady Navier–Stokes aerodynamics solver. Some commercially accessible solutions, such as STRLNCH from AMA [11], also allow the coupling between aerodynamics and separation dynamics, but these are dedicated mostly towards application in store separation from aircraft.
Although these investigations are a valuable source of information, they cannot be fully applicable to the case of the ILR-33 AMBER 2K rocket. Due to the aerodynamic shape, smaller size of the separated elements, quicker separation dynamics and requirements of separation on a vast range of dynamic pressures, a less conventional mechanical design had to be used, compared to the previous version of the rocket [3].
This meant that a dedicated approach to the problem was needed, in order to validate the system design and to qualify it for flight tests. It consisted of:
a set of aerodynamics numerical analyses, based on computational fluid dynamics (CFD) and the wind tunnel investigation described and deepened herein;
dynamics simulation analysis with a new, in-house developed 2D-3DOF software; and
a series of ground tests of key components and of the whole system, designed to verify the mathematical models’ correctness and components’ functional requirements.
The aforementioned topics constitute the subject of this paper.
Each SRB separation system consists of three main subsystems, as shown in Figure 2: two attachment nodes and a repulsing system. The upper node is located in the nosecone of the boosters, right above the core stage’s oxidiser tank (towards the main core nose). It is responsible for the release and carries side loads generated on the boosters. A pyro-mechanical release system is used in the upper node in order to release the fixation point. Its activation allows the repulsing system to push the boosters outwards.
Figure 2.
Components of the separation system.
In order to assure proper separation, a separate repulsing system was also introduced. It consists of a gas spring mounted in the nosecone of each booster, which, after the upper mount pyrotechnic release, pushes the SRB outwards. Once the gas spring is extended, exposing boosters on positive deflection angles,
The lower node connects aft bodies of the main core and the SRBs and is a purely mechanical release system. It assures the separation at a specific angle between the boosters and the core. A ball joint system assures that all the longitudinal and side loads are transferred.
Boosters are kept attached by the lower nodes until the specific angle of their deflection is exceeded. The angle margin is foreseen in order to avoid the possibility of the SRBs’ collision with the core of the rocket right after their separation, which is explained below. Above this angle, having an outward motion and due to the longitudinal aerodynamic forces, the boosters slip out and fall off, evading the rest of the rocket.
After burnout, the boosters are detached in the manner indicated in Figure 3. SRBs are expendable, which means that IoA does not attempt to recover them for reuse in subsequent flights.
Figure 3.
Phases of the separation process – 1. Initial flight configuration. 2. Upper node pyrotechnic release, the gas springs are pushing the boosters outwards. 3. Maximum deflection angle of the SRBs while being attached to the core, lower node release. 4. SRBs’ separation and their ballistic flight (no recovery systems for SrB foreseen). SRB, solid rocket booster.
CFD simulations were performed to determine aerodynamic characteristics of the rocket during the separation process for the following conditions:
Mach numbers: from 0.4 to 2.3;
Angles of attack for the whole rocket:
SRB deflection angles with respect to the main core:
Figure 4.
Velocity map contours (ANSYS Fluent) for a supersonic case with coordinate system and angles definitions. 3D symmetry was applied in all cases.
The model was symmetric, and each numerical case considered that both the boosters are deflected by the same
| • | Normal force coefficient (–) | |
| • | Axial force coefficient (–) | |
| • | Centre of pressure location (m) |
Next, forces acting on the SRB were calculated:
where
It was demonstrated that aerodynamic coefficients and centres of pressure on the SRB are highly dependent on the deflection angle. Due to the nose cone shape and orientation with respect to the core at zero angle of attack, the nose generates negative coefficients
As seen in Figure 5, the SRB requires positive deflection angles (higher than nominal zero angle) in order to detach from the rocket. Depending on the rocket’s angle of attack
Figure 5.
CFD results for SRB normal force behaviour at supersonic velocity (nominal separation conditions). Positive values support the separation, whereas a negative one pushes the SRB towards the core. CFD, computational fluid dynamics; SRB, solid rocket booster.
A physical model has been created to determine:
Internal loads acting on each mounting node depending on the external conditions;
Predicted SRBs trajectories (angular deflection and linear displacement in time) during the separation process, relative to the main core.
External loads from the aerodynamics, thrust and inertia forces were obtained from calculations described in Section 2.2 and from flight ballistics analysis of the rocket executed using the in-house developed code. The mass properties of the SRBs, their moment of inertia and geometrical dimensions were taken from the 3D CAD model and then verified through measurements on the prototypes.
2D 3DOF separation simulations were then conducted, with the use of basic Newton’s law equations and geometrical constrains of the system. It is assumed that during separation the loads on both boosters are symmetrical and no other disturbances occur, and accordingly the separation process does not affect the lateral movement of the main core. Three flight cases were analysed: a low-altitude 10 km flight, a medium-altitude 35 km flight and a high-altitude 100 km flight. Optimal repulsing system spring force
Too low a spring force can cause a lack of sufficient deflection angle and a lack of SRBs separation;
Too high a spring force can cause local deformation (compression) of the oxidiser tank;
Too low a release angle can cause a collision between the SRBs and the main core after the separation;
Too high a release angle can cause excessive (destructive) side forces on the SRB and a trajectory disturbance.
Further design optimisation led to a unification, permitting the use of the same spring and the same release angle for all of the flight cases. Based on the rocket flight analyses of the 10 km and 35 km altitude flights, a separation time delay of several seconds was also introduced to reduce the dynamic pressure during separation and optimise the flight trajectory (results shown in Figure 6).
Figure 6.
Deflection angle as function of time since separation initiation for chosen flight cases.
A series of tests were conducted to verify and validate the physical model of the separation.
The test was designed to determine the force profile generated by the node in relation to its linear displacement. Although this force is relatively small, compared with the one generated by the repulsing system, it has a significant effect on the separation parameters and has to be considered. A schematic principle of operation is shown in Figure 7. In addition to it, the upper node has a locking mechanism, preventing it from unwanted separation due to external forces.
Figure 7.
Upper node principle of operation.
The test stand (shown in Figure 8) was used to determine crucial parameters of the release mechanism. It consisted of a freely moving plate of a known mass, mounted on rails with linear bearings, a laser distance sensor and a high-speed camera. The following parameters were determined from the gathered data (exemplary results are shown in Figure 9):
Figure 8.
Upper node release mechanism functional test stand.
Figure 9.
Upper node test results.
| • | Linear displacement | |
| • | Time of operation | |
| • | Initial force | |
| • | End force | |
| • | Work |
During the data analysis, major oscillations of the measured force were observed. With the use of data from the high-speed camera, it was determined that they were caused by vibrations of the plate, onto which the force sensor was mounted. After filtration and correlation of data from the laser sensor and the camera, valuable data were extracted.
A dozen tests of the mechanism in different development configurations have been conducted. All of the tested specimens worked successfully. A high degree of repeatability in terms of forces and time of operation was demonstrated. The gathered data were used to validate and upgrade the separation simulation.
In conducting the test, the aim was to demonstrate a full separation sequence in various flight cases. Owing to the high variability of the aerodynamic forces depending on the deflection angle, a direct reproduction of flight conditions on the test stand would be complex and very expensive. Accordingly, an alternative approach had been proposed earlier, namely one involving the identification of representative equivalent cases to verify selected separation phases. A test stand (shown in Figure 10) with a dedicated mechanism, consisting of a set of gas springs connected to the SRB, has been used to simulate equivalent forces for each test case. The following parameters were determined from the gathered data:
Figure 10.
Separation system functional test stand.
| • | Linear displacement | |
| • | Angular displacement | |
| • | Release angle | |
| • | Time of operation |
Separation simulations corresponding to these cases were also carried out, using the same physical model and same equations as the flight cases. Next, both experiment and simulations results were correlated, leading to verification and validation of the model (exemplary results are shown in Figure 11, and exemplary frames from the test in Figure 12).
Figure 11.
Functional test – comparison between the test and simulation data.
Figure 12.
Different separation phases – frames from the high-speed camera.
A fairly accurate correlation between the test and the simulation data was obtained. The angular motion of the SRB begins approximately at the same time and acts with similar dynamics (curve shape). The beginning of the linear motion in the simulation is shifted relative to the test data, but once started, it progresses in a similar way. The discrepancies between both results can be explained by:
Differences in the moment of inertia in both cases. The moment of inertia of the SRB used in the simulation was obtained from the 3D CAD model and may be inaccurate. An adjustment of the moment of inertia leads to an improvement in the correlation of the deflection angle curve.
Differences in the release angle (the angle at which the linear displacement begins) in both cases. The measured release angle is smaller than the simulated one, which leads to an earlier commencement of the linear displacement. A reduction of the release angle leads to an improvement in the correlation of the linear displacement curve. These discrepancies can be explained by manufacturing and assembling inaccuracies in the used prototypes.
Differences in the characteristics of the test stand elements between the model and the real setup.
Despite the demonstrated differences, the results obtained at the test stand are valuable data that prove a proper operation of the separation system in an equivalent environment and can constitute a basis for arriving at a decision regarding qualification of the system for flight tests.
The results presented above are characterised by a moderate level of accuracy and fidelity. To fully verify the effects of some phenomena, an in-flight test campaign is indispensable. A high degree of system complexity lowers the level of fidelity involved in the dynamic identification of an object. Despite this, the results show that all major effects affecting the separation have been identified and their variations are understood. This qualitative approach allowed qualification of the system for launch.
The authors of this paper are fully aware that the approach shown above is a significantly simplified one in comparison with the analyses performed in the various papers cited within the present study. Contrastingly, this paper also presents a complex approach consisting of a comparison between simulations and test stand results, as well as mathematical model validation, which is rare insofar as separation analyses involving strap-on boosters are concerned. It is caused by a relatively small size of the SRBs and the ILR-33 AMBER 2K rocket as a whole. The authors hope that the presented methodology will be useful for other small rocket designs, where the replacement of advanced computational analysis by an extended test campaign is often profitable in terms of time and schedule and allows a significant increase in the level of reliability and safety during the launch campaign.
The research activities presented in this study were part of a large IoA project, consisting of the design and development of the ILR-33 AMBER 2K rocket. Their aim was to verify the functional requirements of the SRBs separation system and to qualify the chosen design for flight tests.
This task was completed with the use of numerical analysis, including CFD calculations of the separation aerodynamics and 3DOF dynamics simulation. Next, a series of component and system level tests were performed to determine characteristics of key components and to correlate test results with numerical models. Despite the simplifications mentioned in the paper, a good level of correlation has been achieved. The approach presented in the paper has proven to be adequate for the project needs and fully sufficient for the performance of the assigned tasks. In effect, a production of flight-ready components and their acceptance for the launch campaign have been made possible.
However, the following improvements are planned in the next version of the system:
Minor mechanical design changes, which will facilitate the assembly and the preparation for flight of the subsystems;
Modifications in the simulation software, which are required to allow for an automated coupling between the separation simulation and the flight dynamics simulation. Currently the simulations are fully independent, which means that separation simulations are performed in quasi-static flight conditions.
The system has been successfully tested during the ILR-33 AMBER 2K low-altitude test flight campaign [14]. However, flight results’ correlation with the model is beyond the scope of this paper. The system is currently awaiting verification in high-altitude flights.