Rocket Engine digital twin – Modeling and simulation benefits

Siemens Visionary Siemens Visionary
Siemens Visionary

 

The space launch, a highly competitive market

  1. Engines performance holds the key to launcher success.
  2. Benefits of an engine transient simulation model

Mitsubishi Heavy Industries case study: the H3 as Japan’s next flagship launch vehicle.

  1. Second stage engine of the H3 launcher: LE-5B-2 transient simulation model
  2. Modern control design for the first stage engine LE-9
  3. LE-9 Control design flow using Simcenter Amesim.

Conclusion

Bibliography

 

The space launch, a highly competitive market

In 2018 a total of 114 orbital launches were performed globally, reaching the 100 launches mark for the first time since 1990. 2019 is shaping up to be of the same kind.

Underlying this trend is the ever-increasing number of payloads to launch. On top of that, new applications are emerging in the space economy, such as space tourism or mass deployment of small satellites to provide worldwide internet access.

In this context emerging actors from the private sector are driving the launch costs down thanks to a new way of doing business. One notable trend is the development of small launchers targeting lighter payloads and suitable for high launch cadence. Another focus of research is the reusability of some part. Since 2010 it has been estimated that launch cost decreased from 10 to 15%.

 

Engines performance holds the key to launcher successengine_1.png

A key enabler of these innovations is the engine performance. Modern rocket engines are designed to be shut down and restarted multiple times during the flight. The task is challenging for the designers because they must deal with unique constraints linked to propellant management, and thermal preconditioning of feed lines and engine components. For example, before an engine restarts, pumps are cooled down using the fuel onboard to avoid cavitation at start-up. This is done through a complex control sequences of valves and actuators.

Compared to older designs, the engine controller must be improved as it plays a major role in the engine capability to deal with the severe conditions related to multiple start-up and shut down.

 

Benefits of an engine transient simulation model

Fine-tuning the start/stop sequence usually requires many trials during ground testing. As an example, 3171 hot-fire tests were performed on the NASA’s Space Shuttle engine. It represents more than 1,095,677 seconds of operation.  This approach is costly and does not allow to easily extrapolate the engine operation in flight conditions. During tests the safety of the engine and the test equipment is a concern, especially at early stages when the physical knowledge of the engine behavior is not fully understood.

Transient simulation of models allows to frontload the analysis of the start/stop sequence and implement required design changes earlier in the development. They contribute to the cost and risk reduction of the engine development program by:

  • evaluating the design performance in a virtual environment
  • simulating failure cases and predicting the operating redlines (tests safety)
  • complementing the test results in preparation of the real flight conditions

When coupled with a model of the controller, transient simulation can be used to support the development and the validation of control chains.

 

Mitsubishi Heavy Industries case study: the H3 as Japan’s next flagship launch vehicleH3-launcher.jpg

The H3 launch vehicle development started in 2014 to “compete and survive in the global commercial market”, as stated by Mitsubishi Heavy Industries Ltd. (MHI) in a technical review in December 2017. Like ArianeGroup’s Ariane 6 or SpaceX’s Falcon9, the design of the H3 launcher focuses on reducing both launch and operational costs, while keeping the high reliability of the current versions H-IIA/H-IIB.

MHI is the primary contractor for the vehicle development, including the development of the engine system. The program is currently in a detailed design phase and progresses towards a first flight in 2020.

 

Second stage engine of the H3 launcher: LE-5B-2 transient simulation model

As the LE-5B-2 development started, two pain points were identified by MHI’s engineers:

  • One of the MHI’s older engine design required one hundred tests and three years duration to fix the start/stop sequence.
  • As experienced workers leave the organization, they take with them accumulated knowledge and skills. Making it difficult to maintain and upgrade legacy simulation tools.

Simcenter Amesim was selected to develop the transient simulation model of the LE-5B-2 engine. When using Simcenter Amesim the engineers benefit from the latest development of Simcenter Amesim liquid propulsion solution (and the associated libraries, Two-Phase flow and Gas Mixture).amesim_Liquid_prop_library.png

[This YouTube video will walk you through one of the rocket engine model available in the demonstrator database of Simcenter Amesim.]comparison-inhouse-amesim.pngBuilding engine models in the Simcenter Amesim environment ensures that simulation results will remain reusable and accessible to anyone in the design team.

The LE-5B-2 engine model results were compared with existing results, showing a good agreement with both the in-house simulation tool results and firing tests.

 

Modern control design for the first stage engine LE-9

The engines currently in operation use Proportional-IntegraI control as a baseline. It has limitations.

  • Interactions between fuel-oxidizer Mixture Ratio (MR) control loops and the thrust generated are observed in firing tests. It requires additional work to add decoupling and cancel them.
  • There is no control during flight because some quantities, such as the MR, are difficult to measure.
  • The engine must accommodate a wide range of operating points. This is due to variation in the launcher’s acceleration, which changes the engine operating point, and run-to-run variation.

A more robust control strategy, the Linear Quadratic Servo Control (LQG/LTR), is selected. With this approach it is easier to decouple interactions and it provides a MR estimator during the flight.

LE-9 Control design flow using Simcenter Amesim

Developing a LQG/LTR controller for the LE-9 engine requires:

  • a simulation environment able to generate a state-space representation of the engine
  • a co-simulation capability with the controller development environment

Simcenter Amesim provides these capabilities. The state-space representation can be obtained automatically thanks to the Linear Analysis tools. And the Simcenter Amesim - MATLAB®/Simulink® standard interface allows the co-simulation with models of the controller, in a Model-in-the-Loop (MiL) configuration. MiL allows fast development as you can make changes to the control model and immediately test the system.co-simulation-illustration5.png

 

Conclusions

Siemens, MHI and Churyo Engineering have developed in close collaboration a framework for rocket engine transient analysis and control development. The methodology is now mature enough to be applied on current developments and should support MHI objective of a launch service price reduced by half compared to HII family and aims to be on par with SpaceX’s Falcon 9.

The Liquid Propulsion library of Simcenter Amesim benefited from this work, contact me at sylvain.pluchart@siemens.com for more information.

 

Bibliography

About H3 Launch Vehicle Official website

As rocket companies proliferate, new enabling tech emerges as the next wave in the pace race

The New Rockets Racing to Make Space Affordable

SpaceX Falcon 9 vs ISRO’s Reusable Launch Vehicle

Why 2019 is shaping up to be a stellar year for space exploration

Space economy: an overview