Additive
Manufacturing
DECREASING POINTING INACCURACIES IN STAR TRACKERS
An Alliance for Innovation
ESAF, which comes from the German "Entwicklung einer integrativen Satellitenstruktur durch Einsatz Additiver Fertigungsverfahren" and translates to "Development of an integrative satellite structure through the use of additive manufacturing processes," was a project whose goal was to design and print a structure with integrated sensors to decrease the pointing inaccuracies of a satellite-mounted star tracker.
ESAF was selected by the AMBER Koordination und Innovationsbegleitung (a part of the Berlin Partner für Wirtschaft und Technologie GmbH) to receive funding provided by the European Union from the Operational Program of the European Regional Development Fund (EFRE) and from state grants. The ESAF project is co-financed by the European Union.
Generative Design
Generative design allows for the design of complex parts that can be optimized for structural performance and/or thermal behavior with less mass, helping us achieve our goal of high-performance customized solutions.
In general, satellite components are typically based on what has worked in the past. But now we can use increasingly sophisticated computer-aided design, modeling, simulation, and test result prediction to increase performance while reducing weight. Generative design works by feeding a computer boundary conditions (such as typical structural loads, fixture placement, and thermal environment) to optimize parts for highly specific tasks and environments.
Additive Manufacturing
Decreasing the manufacturing time of structural components is also achieved through additive manufacturing (3D printing).
Additive manufacturing is, at its core, the creation of structures via the application of multiple layers of material. Through the application of many layers, a complete structure is formed.
On its own, manufacturing parts via 3D printing has the benefit of allowing for ever more complex structures compared to conventional manufacturing. While ordering parts in large volumes is usually cheaper, additive manufacturing allows for the rapid manufacture of cost-efficient custom parts.
Marrying the Principles
Additive manufacturing truly shines when utilized in conjunction with generative design. With their strengths combined, we can realize the optimized parts created by generative design at significantly reduced manufacturing lead times and at lower cost.
Neither of these approaches are new in engineering generally – but they have yet to find general application in space engineering. This is due to several factors that include prohibitive cost, concerns about quality, and a lack of proven flight heritage. Given recent improvements in quality and drop in cost, the benefits with regards to custom satellites are not to be ignored.
Our Approach: Embedding Sensors
When designing satellites, Reflex takes advantage of both measures as well as their use together. But we don’t stop there.
Sensors are an integral part of every complex machine. A layperson may think of the payload as the only source of satellite data, things like images of our planet or data on the atmosphere. However, enabling these data are a bevy of sensors measuring the conditions regarding the satellite itself.
By teaming up with Fraunhofer ILT, Reflex employs the novel technology of integrating sensors into satellite components; measuring not the environment, but the conditions of the satellite’s structure.
This is made possible with the use of 3D printing, which allows the inclusion of sensors into the partially printed parts by stopping the print at specified locations in the proper environment.
The resulting part does not show any kind of visible interruption. The integrated sensors are able to determine the conditions more precisely, such as temperature inside the part without being influenced by outside factors like direct exposition to sunlight.
Using Sensor Data for Increased Performance
More sensor data is always nice, but on its own, it’s just a bunch of numbers on a screen. The way it is applied is the important part.
Reflex has developed an approach to correlate these data points with the deformation of the structure using machine learning. This enables us to estimate the deformation of the satellite structure due to the temperature differences inside the satellite in space.
Deformation in a satellite’s structure, even at a minuscule scale, can have massive deleterious effects; deformation significantly impacts the precision of the satellite, as it results in misalignment of extremely sensitive sensors such as star trackers, telescopes, and directional antennas.
Now come the embedded sensors. Armed with the knowledge of the structure’s temperature, and therefore measuring its deformation, Reflex can determine the error induced by the thermal deformation inside the satellite. This calculated error can then be used to again increase the now-reduced precision, returning the satellite to a state of high performance.
The Benefits for Customers
By combining generative design, 3D printing, sensor integration, machine learning and error correction, Reflex is able to increase the precision of satellites significantly. As performance requirements and the need for rapid replacement of technology increase, Reflex’s ability to produce custom, high-performance, reliable satellites at a low lead time is a tool in our belt keeping our customers prepared to rapidly innovate in space.
Our continuing work in this field is looking into other applications of generative AI in satellite manufacturing, including the use of diffusion models and large language models.