SU2 Foundation wants to overcome pitfalls of open source

5 June 2019

Developing and spreading the use of new software to help engineers solve complex fluid dynamics and multidisciplinary problems. This is the goal of the SU2Foundation, a collaboration between universities, research institutes and companies, which was incorporated as a non-profit in the United States in May. The design optimisation carried out with the software increases the efficiency and lowers operational costs of for instance aircraft engines or wind turbines, says Piero Colonna, Professor of Propulsion and Power at Delft University of Technology and a founding member of the SU2 Foundation.

The organisation currently consists of teams from institutions all around the world, such as Stanford University, Technical University Kaiserslautern, Imperial College, Politecnico di Milano, the University of Michigan, Argonne National Laboratories, and the US National Institute of Aerospace. In the Netherlands, in addition to TU Delft, the University of Twente and TNO/ECN are contributors. Individuals affiliated with several companies, including Bosch, Intel, and aerospace conglomerate Embraer are also involved.

The open-source multi-physics SU2 software is a collection of programmes which was released for the first time in 2012. Like other programmes, SU2 can simulate the flow of liquids and gases, the so-called Computational Fluid Dynamics and helps engineers optimise certain features of a device. ‘The efficiency of the turbine of an aircraft engine is determined mostly by the shape of its blades. This software can simulate the functioning of these blades, and with the help of algorithms, a sort of AI, continuously deform these shapes until it finds the geometry that provides the highest efficiency. The same can be done with the shape of an entire aircraft, wind turbine blade or car’, Colonna explains. The code is used in the aircraft, automotive, ship, and renewables industries, amongst others.

Environmental benefits

By finding the optimal design, the software can help lower the costs of equipment, and increase its efficiency, which leads to a shorter payback period, the professor says. And as a consequence of the higher efficiency, there are always environmental benefits. Colonna: ‘For instance, we are now working together with a Dutch company on a new concept for the Environmental Control System of aircraft, which regulates matters such as the pressure, temperature and quality of air in the cabin. Next to the engine, this is the system with the largest energy consumption. By optimizing the design of the compressor with the software, the aircraft will use less fuel.’

The group of Colonna is regarded as one of the leading academic teams performing research on waste heat recovery or energy harvesting by means of organic Rankine cycle technology. This research is about the question how to effectively use thermal energy that is otherwise discarded to the atmosphere. NWO Domain Applied and Engineering Sciences funded several of his projects in collaboration with global companies. Fluid mechanics and optimal design of turbomachinery play an important role in this research. In this way, Colonna and fellow researchers have made and are making major contributions to the development of the SU2 code.

Quality of the code

SU2 was born as open source software, therefore any developer can contribute code. The SU2 Foundation is there to manage the code in order to guarantee its quality, says Colonna: ‘One of the pitfalls of an open-source code is that often the development is not properly managed. The code is a living thing. It is constantly growing. With bad management, this can lead to what we call spaghetti code, and the programme becomes increasingly difficult to develop and to use’.

Developing a Word processor every time you want to write a text is not a good idea

This is why the organisation is structured in groups of constituents (academia and research organisations, companies, developers, users), and is led by a board that makes decisions about important aspects of the organisation, such as priorities within the development of the code. Input is collected from all the members. The Foundation also plans to ask companies benefiting from the code for donations, which could then be used to hire professional software engineers for structural development and maintenance of the software.

Working with a community code will also greatly improve the efficiency and pace of knowledge development, Colonna thinks. ‘Worldwide, a large number of researchers develop more or less the same code in order to study new problems in a following step. We want to provide the underlying layer to all, so that every researcher can build directly on top of it, instead of reinventing the wheel.’ This calls for more cooperation, according to the professor. ‘Developing a Word processor every time you want to write a text is not a good idea. Make sure an excellent processor is already there, and everyone can create original work using it.’ According to Colonna, this connects to the broader trends of open science and open technology. This method is also beneficial for companies. ‘High-risk/high-reward research becomes less risky when carried out in a global team.’

More information

Source: NWO