Improving aircraft environmental control system performance

Features - Software

Airbus reduces turnaround time by 90% by optimizing static air mixer using Siemens PLM software.

A cabin airflow simulation for the Airbus A320 aircraft.

How often do we marvel at the fact that humans are flying around in climate-controlled metallic tubes, 30,000ft above the ground in an extreme environment? Providing a comfortable cabin atmosphere when the external conditions are around -40°F to -60°F is a complex task for the air conditioning and cabin pressurization system. Even more challenging are demands for pre-conditioned air to the avionics and ice-protection systems. Moreover, stricter environmental regulations necessitate an efficient system that improves fuel efficiency and reduces emissions.

Environmental control systems (ECS) typically refer to systems and equipment that provide a comfortable atmosphere to the aircraft payload, including people, avionics, and other onboard systems. Environmental protection systems (EPS) protect against external conditions – extreme temperature and pressure, ice buildup, etc. Well-designed ECS and EPS lead to a comfortable flight for passengers and improve fuel economy, since ECS is the second biggest power consumer from engines after thrust.

The ECS group at Airbus Operations GmbH follows the mantra of innovation and pioneering technical solutions, constantly looking for better ways to design the best performing ECS for Airbus aircraft. Taking advantage of improvements in hardware resources and numerical modeling, the ECS group has deployed numerical simulation to understand and improve systems and sub-components faster than through expensive physical testing. Simulation, particularly computational fluid dynamics (CFD) tools, has been beneficial in cockpit design, avionics cooling, mixing and pressure loss in ducting, cabin thermal comfort, and other areas. Recently, the ECS group has been leveraging design-space exploration in a production environment to improve bleed-air systems in future aircraft.

Figure 1: Bleed air system of an Airbus A320neo aircraft.

Static air mixer

Bleed air from the engine fan and compressor sections pressurizes and heats the main cabin and supplies hot air to the ice-protection system. A typical bleed air system of an Airbus 320neo (Figure 1) is located in the pylon under the wing. The system delivers preconditioned air at a temperature and pressure for cabin air conditioning and wing ice protection. Hot bleed air from the engine compressor and cool air from the engine fan pass through a plenum into a heat exchanger. The bleed air is cooled to roughly 200°C in the heat exchanger. Uniform and thorough mixing of the core and bypass air is achieved by a static mixer and is critical for maintaining subsystem performance. The temperature of the mixed air from the static mixer is monitored via temperature sensors before being delivered to various systems. Airbus set a goal to improve the current static mixer design and achieve total pressure loss of less than 5kPa between pre-cooler and temperature sensor and a temperature range of less than 12 K at the sensor location.

Initial optimization

Figure 2: Baseline design and a sample of design iterations from Optimate+.

Finding the optimum design and the best compromise between the conflicting objectives is worth its weight in gold, if the optimum design can be identified as part of the design cycle. This saves time and money and leads to the best possible design for production. The real challenge is to identify and deploy an efficient optimization process that fits within production constraints. Various steps are involved to go from an initial geometry to the optimum design – experiment design, 3D geometry modeling, low-fidelity CFD modeling, surrogate modeling, pareto front identification, and postprocessing. The first optimization workflow evaluated a mixture of commercial and in-house tools to perform these tasks. A helical mixer concept was optimized to evaluate the viability of this process. Three design parameters were used in the optimization: radius of curvature of the helical channel, depth of cut into the duct, and the width of the helical cut.

Due to infrastructure and communication issues between the different tools, only two design parameters were used in optimization to meet time constraints. The optimization study identified a design offering a pressure drop of 5,000Pa and a temperature difference of 45 K at the sensor location. The total turnaround time for the initial project was six months. The final design improved the temperature difference but was not close enough to the design objective. The question remained: how do we meet the design objective while reducing the turnaround time to make this process viable in a production environment? To resolve these concerns, Airbus decided to collaborate with Siemens PLM Software to find a new process.

STAR-CCM+ optimization

Computational mesh on the baseline design.

STAR-CCM+ software, a multidisciplinary tool for engineering simulation from Siemens PLM Software, facilitates the analysis and exploration of real-world problems by accurately predicting product performance. The integrated user-interface includes tools for geometry handling, meshing, analysis, and postprocessing to automate workflow and reduce turnaround time. The Optimate+ add-on provides design exploration and optimization from within STAR-CCM+, powered by the Sherpa search algorithm of HEEDS software, a multidisciplinary design optimization tool also from Siemens. The Sherpa algorithm adjusts search strategies using hybrid, adaptive search techniques. This process eliminates communicating between different tools and brings the entire design optimization process within a single environment.

To meet the design objectives, a concentric mixing blade design was chosen instead of the helical mixer. This design involved two concentric rows of mixing blades. A total of 11 design parameters were chosen for the optimization study, including blade number (3 to 13), blade angle (0° to 75°), blade curvature (a complex set of semi-dependent parameters), and the ratio of duct occupied by inner blades (0.3 to 0.7). The design concept was chosen to provide maximum flexibility in the design space, and the direct optimization technique ensured that no simplification of the design space was needed. The geometry, modeled in Dassault Systèmes’ CATIA, was modified by Optimate+ based on the design parameters (Figure 2).

With the automated workflow of STAR-CCM+, Optimate+ evaluated 211 design variants by varying the independent parameters. For each simulation, a polyhedral mesh (Figure 3a & 3b) was automatically generated with a cell count between 2.4 million and 3.5 million cells. The segregated flow solver was used with constant density air and the SST k-turbulence model. All simulations started from a converged solution for the baseline design. Convergence was achieved when the average pressure drop was within 50Pa and maximum temperature difference was within 4 K throughout 1,000 iterations. Designs that didn’t converge after 15,000 iterations were rejected.

Figure 4: Plot showing influence of different design parameters.

Optimized in two weeks

The baseline design achieved a pressure drop of 996Pa and temperature difference of 153 K. Optimate+ used the Sherpa algorithm to find the best compromise between the conflicting design objectives from 211 design iterations. The best performing design achieved a pressure drop of 4,775Pa and a temperature difference of 16.4 K. A pareto optimization plot of the results showed the best performing designs had a small number of blades and smaller curvature. Armed with this information, 57 additional design iterations were performed from the optimum design to minimize the temperature difference further (Figure 4). The red lines show all design iterations and the blue lines show the designs that achieved the required pressure drop objective. The best designs are shown in green, achieving both objectives. From the plot, the influences of various design parameters on the best designs can be found. The best designs had a higher number of blades on the outside, fewer blades on the inside, higher outside curvature, and lower inside curvature. The diameter ratio for the best designs was around 0.5.

The best design from the optimization had nine outer blades at 62° curvature, moderate curvature at 25% chord, three inner blades at 5° curvature and mild curvature at 48% chord, achieving a pressure drop of 4,961Pa and temperature difference of 13.6 K. The total turnaround time to achieve the best design was two weeks (Figure 5a & 5b).

Better design, faster

Figure 5a/5b: Baseline (top) to best design bottom).

The optimization process with STAR-CCM+ and Optimate+ achieved a 90% reduction in turnaround time (Table 1, above). This method will be deployed in the design cycle for future designs, allowing Airbus to find better ECS designs faster. The result is an improved ECS system with better mixing, reduced time-to-market, reduced testing, better energy efficiency, and better customer experience.

Airbus Operations GmbH

Siemens PLM Software

About the authors: Andreas Ruch is a systems engineer at Airbus Operations Gmb. Contact Ruch at Prashanth S. Shankara is a senior technical marketing engineer with Siemens PLM Software. Shankara can be reached at