HBK delivers sensors, measurement electronics and software for data acquisition and analytics tools as products or end-to-end solutions into the Test & Measurement world. These help to assess structural integrity, noise and vibration, efficiency, and performance and many more aspects of the structure under test in lab, bench, or in-vehicle testing.
Looking at major trends in technology and industry, such as lightweight structural designs and their validation, HBK plays a major role in offering the optimal eco system for lifetime simulation and data acquisition and analysis in physical testing, helping to validate modern designs and constructions on their way to mass production.
To give you some insights and experience into the challenging field of lightweight constructions and their validation, we have interviewed our HBK Experts on Lightweight Structures: Gianmarco Sironi, Lance Steinmacher, Dr. Andrew Halfpenny, Michelle Hill, Manuel Schultheiss and Sandro Di Natale.
Gianmarco Sironi and Lance Steinmacher
This is a nice question to start with. From a carbon footprint perspective, saving weight is crucial to minimize the impact on climate change. We see aircraft manufacturers using CAE/CAD to look at different composite layups to control strength or flexibility where desired. Examples of this are the 787, 777X wings where reduced weight means less fuel burn. For rotorcraft, you can control the composite layups to provide stiffness in one direction and desired flexibility in another.
However, we also need to talk about downsides. Recycling or disposal of composites is much more complex or even impossible while traditional metals can be readily recycled. Also with composites, some of the forming techniques are not the most environmentally friendly.
Manuel Schultheiss and Sandro Di Natale
Lightweight designs are typically seen in the aircraft industry, automotive industry, and sport equipment:
To add some numbers, here is a small example:
In the aviation and space industries, lightweight designs simply pay off. Reducing weight by 1 kg saves about 0.02 – 0.03 kg of kerosene or around 2 – 3 cents per 1000 km. With an aircraft like the 777 flying more than 50 million miles in its lifetime this would lead to approximately 80,000,000 km x 2.5 cents/1,000 km = €2,000 per kg less in weight for an aircraft lifetime. Think about an aircraft 100 kg lighter by design!
Therefore, aircraft and spacecraft are already in the second generation of lightweight designs. Carbon Nano-reinforced Polymers (CNRP) replace parts currently made with other types of composites, as they are stronger and offer 30% less weight. Additive manufacturing is widely used in the newest generation of aircraft, enabling new designs at lower weight, for example cabin brackets.
Michelle Hill and Dr. Andrew Halfpenny
The categorization of lightweight materials is a big topic. You can look at it from different angles:
Despite the advantages of lightweight construction, there is the aspect of cost efficiency. Not in terms of the material, but in the cost of energy consumption. With a customer base that is mostly from the ground vehicle, automotive, trucks, trains, or aerospace sectors, you will always have the problem that extra mass means extra required force and consequently more energy. Lightweight constructions can reduce the required tractive force and thus not only save costs but also generate environmental friendliness.
In addition to long-fiber composites, you also must look at polymers (colloquially plastic). You can see a lot of structural use of polymers in automotive components. Again, it is all about the trick of having the material exactly where you want it. That is where additive manufacturing (AM) comes to its advantage. With AM you can have a more exotic geometry. So, it saves a lot of time and money because you do not have to chop away all the redundant material.
Lance Steinmacher
When creating some of these new materials, a company will protect its intellectual property to give them a competitive advantage. By doing this, some of the composite material properties are either patented or proprietary and hidden behind a non-disclosure agreement (NDA). This goes for the actual material layups (chopped fiber, fiber, directional, unidirectional, etc.), adhesives or epoxies used, and the manufacturing process. As a result of that, only limited information regarding composites is open to the public.
Gianmarco Sironi and Lance Steinmacher
Many consider the advantage of composites being the lightweight construction, but it is not the only one: in some applications, composites have been introduced not for weight saving but for their great performance in fatigue life.
A helicopter main rotor blade made in composite material is not so 'lightweight' compared to its old aluminum alloy counterpart, but it is a lot more durable: its fatigue life in flight hours is an order of magnitude superior. That is the great leap forward, but it is not the only one. Metal rotor blades were a maintenance nightmare and needed a lot of non-destructive inspection (NDI) as they developed cracks fast and without notice, causing a significant number of accidents in the past. Thanks to composite blades, significant improvements have been possible, especially for medium and heavy helicopters. So, it is not just about weight, it is also about safety of flight and easier and more efficient maintenance.
On the other hand, materials like aluminum or titanium alloys still present the intrinsic advantage that there is plenty of literature about their fatigue behavior as they have been studied for decades. Advanced composites are relatively young compared to metals and this literature is not available yet, or if present, it is somewhat limited.
Composite materials are neither uniform nor isotropous and this makes characterizing them a real challenge. Furthermore, every time you change a single ply in the laminate (or just keep the same ply but change its orientation) you have basically created a new material. So, you have to start again with the fatigue characterization beginning from the simple test sample, that is a lot of time and money. Therefore, it is very convenient to have quite good material knowledge and simulation capabilities when you first design the laminate.
Sandro Di Natale
It is hard to answer this question generically. There are so many different technologies summarized under this term. Besides consumer-driven technologies based on filaments, there is stereolithography, binder jetting, and many more. From an industrial perspective, I believe that selective laser sintering and melting (SLS and SLM) working with metal powder are among the most promising technologies.
The components manufactured, for example, from titanium powder, behave in a similar way as cast or machined components. However, special attention in testing is needed to ensure that the characteristics are isotropic and that there is no influence of the layer structure. If this is given, the potential for weight reduction and new designs is almost infinite. Unfortunately, the powders are still quite expensive and manufacturing speed is slow. The latest and biggest machines can build at rates of a few hundred cubic centimeters per hour.
Sandro Di Natale
Whether additive manufacturing can be used in production depends on the following criteria:
The advantages become even more evident with lot size one. In the medical industry, additively manufactured prostheses and corrective supports are well-established.
Manuel Schultheiss
Bionics play a highly significant role when talking about lightweight structures, as we learned from the design of birds how to build a plane and make it fly.
We can further learn a lot more about optimizing technical designs with nature as an example, when we think about sharkskin surfaces that are applied on the surface of airplanes, or winglets to reduce turbulence at the wingtips of airplanes. The whole mechanical structure of plants and trees can be used to derive the optimal mechanical design with lowest risk of breaking and lowest stresses inside the material. Nature has learned and adapted to the best design fit for the environment.
The exciting element of new materials and manufacturing methods is that it is easier to establish customized designs. One example is additive manufacturing. With this approach you can create smoother roundings in designs with less stress concentration and create optimal mechanical frames with the lowest stresses and longest lifetimes.
Michelle Hill and Dr. Andrew Halfpenny
There are two things. The first shouldn't even be a surprise, but it was. We use many materials, and we must join them together. We only welded steel to steel in the past, but now we are looking at different types of joints, such as self-piercing rivets, bolted connections, adhesives, or a hybrid use, for example, adhesives with rivets. There are many more exotic types of joining and more accuracy is required. In the past, welds were done conservatively, and everyone was fine with that. But now, with the need to reduce weight even in weld curves, we do not want conservatives anymore.
The second point in considering qualities is the requirement for knowledge about uncertainties. Previously, engineers designed something and simply applied the so-called safety factor. This is a combination of real safety and "coefficient of ignorance". Now this is no longer acceptable, it needs to be quantified. This need is led by safety-critical organizations such as the nuclear industry, aerospace, etc. There is a requirement for a better understanding of where the safety margin comes from, and whether it really is as big as we think it is. So, it is no longer about predicting a number, it is about giving a range of values so that people can say with confidence that they accept that one in a thousand will fail, and if it does fail, it will not be catastrophic.
For data acquisition (DAQ), this means that accuracy, data precision or the requirements for it are increasing. A concrete example of this is our current work on probabilistic fatigue. Calculating the probability of component life is not new. Ten years ago, we were talking to customers about this new technology, but at the time most of them said they did not even know what the expected loads were, let alone how much variation there was in them. Now the same people come and say that with the Internet of Things, we have a pretty good idea of how high these loads are, and we even know their standard deviation. It is a big change in the last decade with all the data available, which we never had before. That is what is really driving this need for accuracy. We now have the capability, the input of the data we need, and the need in lightweight designs to bring it out. In the beginning, the gauge inaccuracy might be small, but it grows exponentially until fatigue. This you need to track.
Michelle Hill and Dr. Andrew Halfpenny
For the battery, from a non-chemistry point of view, everything needs a chassis or a support system. The interesting thing is that we use these batteries structurally in electric cars, for example, so they are part of the structure of the vehicle and the chassis must be able to transfer structural loads. Furthermore, the massive battery itself bounces up and down, creating a dynamic load. So, the designer has the complex scenario of the fusion of heavy weight being vibrated and structural loads being transferred.
In addition, there is the requirement to shield people from the high voltage inside. This means that the use of metal can be a problem. Besides reducing the weight, non-metallic joints, such as adhesives, become more important at this point.
Overall, it is quite challenging to test this. Now we are receiving a lot of questions we don't know how to answer yet, but we are very excited to be getting into the topic.
Gianmarco Sironi and Lance Steinmacher
It is somewhat like an 'efficiency mindset'. Lightweight is about putting only the material you need to achieve the required static strength and/or fatigue life. However, in some industries like aeronautics, this mindset often crushes with safety and redundancy requirements. In our opinion, a lightweight mindset is important, but it must never gain the upper hand on the safety mindset. Structural testing remains the most effective way to ensure both requisites are satisfied.
Manuel Schultheiss
Cost and time play a big role. There are quite some challenges for series production parts. When manufacturing time is not that relevant for single items, such as parts for motorsport teams or for a yacht, this is not an issue but there must be processes that allow quick and process-safe production of these parts. Especially when talking about additive manufacturing and fiber composite materials (for example, carbon-fiber-reinforced polymers (CFRPs)), this is a point that is not solved process-wise. There are a lot of innovations in this field.
Another point is the recycling of these materials. There is the claim to create products that follow the thought of a sustainable future. When materials cannot be reused because of their structure, then this is critical. When using more natural approaches of fibers and epoxies, this could solve the impediment.
Manuel Schultheiss
We should not overstretch the influence of the pandemic. I think there is no significant push by the pandemic itself to go for lightweight materials or construction. Processes are constantly accelerated and digitalized. Some companies might have used the time to “reinvent” themselves and try something new out here.
Michelle Hill and Dr. Andrew Halfpenny
When talking about physical tests, most people think of full-scale testing such as a helicopter being set in vibration or something similar. However, if you look at the triangle (Figure 1) that represents the scope of the tests, you can see that full-scale testing is only a small part of it. We divide the tests into qualification tests, including full scale testing, and parametric tests. We use our hardware and software primarily for parameter tests, to be precise, mainly coupon tests, which are currently increasing.
The main goal when doing coupon testing is to get physical parameters to derive physical models that can then be used for simulations. Before lightweighting, it used to be okay to get material properties from Google when we had massive coefficient of ignorance. Now you cannot afford to do that. It is much cheaper to spend €15 – €30 k on a test than to go out with a million overdesigned cars.
On the next level in the triangle, you see the element test. Even at the level of simple light components, these are usually structures that are casted or fabricated in some way. We test them to failure, so we determine the load versus the life. We now need to convert these to stress versus life. This involves the back-calculation of material properties through the simulation model. We therefore build a finite element model of the test element and perform a complex optimization to determine the required parameters. Here, it is important that we use the same finite element modelling guidelines as our customers do in their full-structure model, only then can we provide them with the results they need.
The level above shows the component test. Here, we want to calculate some of the more wholistic parameters. For example, with vibration, damping is super critical because it is the only thing that absorbs energy. We need to know that, and at the component testing level we try to get parameters for the simulation.
Qualification tests usually take place at the end of the cycle and allow us to do a lot of measurements to check whether our assumptions were correct. If the structure fails, we need to simulate why it failed so that we can use simulations to weed out the failure before we move on to the next round. And, because the qualification tests are at the end, it means that any change is massively expensive for clients. So parametric testing must increase to make our models more robust and accurate.
In the figure you will find a third category of tests, the reliability test. While the parametric test maps the physical model to failure, the reliability tests map the statistical model to failure. At HBK, the goal of the tests is to get information about how many of the products will fail if the customer gives, say, a 10-year warranty and how much that will cost.
Manuel Schultheiss
There are different levels of certifications. Let us look on the materials level:
Looking at the certification of structures or whole products such as aircraft, it is even more complex. These cover more aspects than simple “how to test coupons” and define material properties. These certifications include design, manufacturing, and maintenance aspects as well as the whole picture.
Michelle Hill and Dr. Andrew Halfpenny
Reducing the coefficient of ignorance is a big challenge. But another point is that we are now testing components rather than materials. With joints, again you have the problem that you need to use the same finite element meshing rules as the customers, because we specify properties with respect to a specified mesh. This requires back-calculation to find the properties.
Also, the big differences between welds are a challenge. If car company A wants to test a weld, you can be sure that the weld you have to test for car company B is different. That's why sample geometries must be tested for each customer. But the differences are not only between different customers, but the customer also has to produce components that later correspond to mass production. The question here is whether this can work for fully automated mass production.
When it comes to composites, it gets even more complex. How do we define failure? With joints it's simple, they break in two. But composites do not fall apart, they can lose stiffness or strength. Also, at a macroscopic level, you cannot compare composites to alloys. Composites can fail at the macroscopic level through many different mechanisms. Debonding of fibers, cracking in the matrix or cracking of fibers could be part of the highly progressive path to failure. When we decide what is failure, we have to talk about the stress. Is it the stress in each volume, between the fiber and the matrix, or only the stress in the fiber? All these are still open questions, and we need to put more effort into answering them.
Another challenge is the standards for composite materials. While ASTM and ISO standards often come from Airbus or Boeing, they are specified for aerospace. But while aerospace sets its sights on high quality, it becomes very difficult for the automotive industry, for example, as they mainly want quickly produced and cheap composites like Chopped Strand Mat. But the modelling here is a nightmare: you have no knowledge of the fiber direction, which is essential for modelling, but you still have to satisfy the high standards. And then you still only have standards for tests, but not for how to interpret the data. Here, too, it is important to find answers together with research laboratories.
Michelle Hill and Dr. Andrew Halfpenny
The traditional way is for HBK to offer solutions for the material tests as well as for the flight profile (loads), which go into the simulation. Fatigue analysis needs three inputs: the loads, the materials, and the geometry. We can simulate the geometry and if it does not work, we just change the CAD/ FEA model until we get a good life result. Then we create a prototype, test it, and then correlate it with our analysis. If we are right, we end up needing only one prototype.
Nevertheless, it is different with composites as they have a much more complicated simulation path. Here, the material properties change due to the structure. There are no such things as composite materials, there are composite components with changing properties wherever you look. For a composite, the tasks and results become much more iterative. Right now, we are using the 'good old method' again. We need a lot more prototypes because you cannot trust simulation anymore, as this is still very new for composites, unlike metals, where simulation works quite well.
So you see, even 70 years later, the design methodology used for the DeHavilland Comet is similar to the design-it – test-it – fix-it methodology used in the Dreamliner. We need far more prototypes, at different scales, because the simulation still cannot adequately cope with composites
Michelle Hill and Dr. Andrew Halfpenny
At first, we thought there was a lack of data for traditional metal as well. The question with the given data is, is it reliable? Some of the data you get from standards are from the Sixties or maybe collected for different sectors. You need to keep a close eye on where the data is coming from. With composites, you have the additional problem that even if you have parameters, you cannot be sure that they are the same throughout the material.
Our customers are mostly active in component and full structure qualification. We test coupon material and get the information for composites from that. However, when we work with welds, we test the welds and calculate back what the properties of the coupon was by running the simulation. If it is a composite, it is a component that we test. It may look like a coupon, but it is a component.
Here, it is worth mentioning that additive manufacturing with titanium and aluminum is treated like a composite. That is because if you have two customers, who are both using the same additive manufacturing machines, you are not going to get the same material properties because of the different settings of the machines. There is a huge need for further testing here, because with AM there are also new types of defects such as porosity or lack of fusion.
Michelle Hill and Dr. Andrew Halfpenny
We use different HBK equipment, for example, load cells, strain gauges as well as QuantumX for data acquisition, but also test machines from Instron and MTS.
With the strain gauges, we had the problem that one of the new materials was so good that it outperformed the strain gauges. For HBK, this means that we have to constantly innovate our equipment to keep up with the material. Speaking of innovation, we had the opportunity to test with a laser extensometer. It measures within micrometers without touching and eliminates the problem that composites, being “explosive” when they fail, might damage the equipment.