Simplifying Lightweighting
Lightweighting has been a big concern for automotive, aerospace, and medical application for a while now. Plastics are consistently replacing metals as they can achieve weight reduction with same strength properties.
Properties like high moldability ensures design flexibility and makes it suitable for intricate applications. Additionally, plastics are cheaper and faster to manufacture, can be safely operated and have a high recyclability.
AirGo’s co-founder and CTO, Alireza, is a materials science PhD dropout.
Alireza Yaghoubi and team at AirGo Design are facilitating the shift from metals to plastic composites for large scale applications. The Singapore based startup has developed an AI assisted software that accelerates metal replacement by predicting anisotropic properties (material with different strength and stiffness in different directions) of complex reinforced thermoplastic components. In our conversation with Alireza (Co-founder of AirGo Design), we discussed the benefits of metal replacement and how their offering is driving the change.
CARBON is a revolutionary product in an industry that has been using aluminum for over half a century.
Performance under stress
“AirGo was founded nine years ago with the initial idea of developing a lightweight aircraft seat. We successfully developed the world’s first full composite aircraft seat using injection molding. The new lightweight seat was affordable and had virtually the same pricing structure as a typical aluminum seat. Building onto this pilot product, we decided to diversity an extend our business model beyond aerospace to developing parts using our technology for other OEMs,” began Alireza.
Unlike metals, the properties of reinforced thermoplastics can drastically change with geometry and processing parameters. In other words, each part has its own unique fingerprint and as such, tests performed on standardized coupons do not offer a realistic indication of how the component might behave in action. AirGo’s software bridges the gap between computational fluid dynamics and structural analysis tools. While commercial software use the computationally expensive “representative volume elements” or RVEs, AirGo uses AI to generate 12 unique fingerprints for each components in less than a second.
AirGo is working on very diverse applications. A composite escalator step that they are developing for a major industry player offers 95% lower CO2 footprint as compared to traditional steps made from die cast aluminum.
Metal replacement is energy and resource lean
“Everyone associates plastics with waste and environmental liability, mainly due to single use plastic and how we use it. There aren’t many inherently bad properties of plastics to begin with. People have a misconception that aluminum is a green metal and is widely recycled. But that’s not the case as it is quite energy intensive and generates a lot of carbon during production,” explained Alireza.
For every 1 kg of aluminum produced, we generate on average 14.5 kgs of CO2. A huge supply chain disparity exists as China produces 60% of the world’s aluminum and 85% of magnesium (another lightweight metal that has a carbon footprint of 22.5 kg on average). During unforeseen events such as a pandemic, the supply chain gets heavily affected, leading to overdependence and disruption. Metals tend to be quite heavy and consumes significantly more energy for dynamic applications (vehicles, aircrafts, or industrial machines). Discontinuous fiber composites, have the same strength properties as metals with lower weight, reduced energy consumption and easy production process.
“The challenging part of this process is when companies want a direct replacement of metal with composites based on a metallic design. The inherent properties of composites are anisotropic (different properties in different direction) and have a completely different production process (injection molding instead of die casting). Most of the time one to one replacement isn’t possible or it isn’t the most optimal way of doing it,” clarified Alireza.
The failure behavior of metals is based on Von Mises model. On the other hand, for injection molded composites, failure mechanics is quite complicated. Currently, the gold standard in the industry is the strain-based Tsai-Hill criteria which relies on allowables that have to be determined using expensive and time-consuming tests. In contrast, the allowables in AirGo’s proprietary Yaghoubi Failure (or Y-Failure for short) are 100% generated by the AI. Moreover, commercial software based on mean-field homogenization have to run along with the structural solver making them extremely inefficient. A typical component, depending on its and complexity, can take days or even weeks to simulate.
To capture the behavior of this composite armrest, AirGo’s Atlas generates more than 1.8 million data points in a mere 0.5 second!
Fast-tracking replacement
AirGo’s trade secret takes advantage of advanced numerical models to develop a self-sufficient input based on flow data and the structural mesh. At its core, AirGo’s AI assisted software named Atlas uses micromechanics made accurate by machine learning tools. A set of 12 functions, describing component-level behavior, are subsequently mapped onto the finite element input deck which is passed down to the structural solver with no need for further communication between the two software. As a result, a fully anisotropic analysis using Atlas does not take any longer than a typical analysis for an isotropic material.
The software learns and extrapolates many different points on the mesh to give an accurate analysis model. “In mean-field homogenization, you are essentially simulating individual fibers and the matrix and consolidating them into a representative volume element. This is a very slow process that cannot be scaled up with more CPU cores. On the other hand, our approach is very fast, fully automated, and has its own damage model which is quite accurate,” elaborated Alireza.
Fundamentally, artificial intelligence is based on data. AirGo has performed numerous tests in the past 9+ years to develop and validate Atlas.
The initial investment for creating functional prototypes with real molds is quite high and time consuming. Development cycles can take years until they are realized into full series production. When a company has had operated in a metal domain, they must hire a team of specialists to work on advanced simulation software, an engineering team and buy expensive software licenses. “What we do at AirGo is offer complete metal replacement solution; from material selection to mass production. In our approach, we tend to promote more sustainable materials with recycled content, bio-based polymers, and reclaimed fibers,” Alireza highlighted.
Diverse applications
The scope for metal replacement is quite broad, which allows AirGo to operate in several industries. They have ongoing projects in traditional industries such as automotive, aerospace, industrial equipment manufacturing, packaging, food processing, industrial automation, pharmaceuticals, and upcoming segments such as drones and e-mobility.
“With our intelligent software, we have the capability of developing material concepts that meet design and performance requirement of the customers, while reducing weight and saving costs compared to metals,” concluded Alireza.