Flexible machines built as a single unit will make today's rigid parts assemblies look like antiques
In 1995, one rainy day, I was driving in the Ann Arbor area of Michigan and was thinking about my windshield wipers. I was then a professor of mechanical engineering at the University of Michigan. In the years before that, I did some research on what is known in the industry as "content for assembly". The goal of my research was to reduce the number of parts in any given machine, thereby reducing production and assembly costs. While working, I began to wonder what would happen if we went so far in the design for assembly to an extreme logical thought: designing products that do not require assembly at all. Can we plan this?
While sitting behind the wheel, I suddenly realized that windshield wipers are a colossal waste of engineering investment. The wiper frame, which holds the replaceable rubber blade, must be very flexible. She must hold the blade pressed against the glass as it passes back and forth over a surface of variable curvature. Moreover, it must do this in several car models, each of which has a different windshield geometry. And what is our answer to this need for flexibility? A complicated system of rods, connections and rigid hinges.
At that time I became interested in another subject: elastic design or compliant design, which deals with the construction of flexible and strong machines from as few parts as possible. My colleagues and I have already managed to build machines from a single unit of material. For example, in 1993 I built with my students, J. K. Anentsuresh and Laxman Segre, a matching elastic staple clamp without assembly. But the windshield wipers seemed like a perfect example to me. A "monoform" wiper, in one piece, will completely eliminate the need for assembly. If such a project succeeds, it will be more than an exercise in engineering minimalism. Most of the cost in the production of wipers is covered by assembly. No one will surely be surprised that the production of such complex products, literally, was transferred a long time ago to countries where labor wages are low.
My colleagues and I did not immediately arrive at the design of a mop in one unit. For the past two decades, most of my research has focused on general principles of elastic design: developing the theoretical tools that engineers need to design and build compliant devices. But in the end we designed the mop. In fact, we used elastic design to build tiny monoform motion boosters, flexible airplane wings, robotic snakes, and other machines, each of which was an expression of a new engineering paradigm that was maturing.
living machines
In fact, compatible machines are more familiar to us than it seems at first glance. The earliest and most elegant example is the bow: when the string is stretched, elastic energy slowly builds up and is quickly released to shoot the arrow. This strong and flexible mechanism is able to operate many times accurately and without failures. A more modern example is a cap from a shampoo container: this is a monoform device that combines an easy-to-open cap with a screw-on seal, without a separate mechanical hinge. The medical forceps used in hospitals are another example, precise enough for an operating room, yet cheap enough to be disposable.
The most successful elastic designs were created by nature. I began to understand this in 1995, when I read the books of Stephen Vogel, the renowned biologist from Duke University. In books like Life's Devices and Cat's Paws and Catapults, he clearly explains how natural mechanisms work while making parallels with engineering devices. Tree branches, bird wings, crab legs and elephant hooves are all flexible and strong. Their components grow out of each other, or are connected by strong ties capable of regeneration. They are able to bend, twist and stretch thanks to their inherent elasticity, unlike systems of gears, springs and moving components.
Humans have accumulated thousands of years of experience in designing strong and rigid structures, such as bridges and buildings. Usually, we do this with strong and rigid materials, and if the pressures are too strong, we simply add more material to strengthen the structure or distribute the load. In such methods, rigidity is a good thing and flexibility is a bad thing. In rigid structures, the tendency to deform or change shape under stress is only desirable when designing for earthquake resistance.
Compliant design, on the other hand, happily embraces shapeshifting. If the stress on a certain bending point is too high, we make it thinner instead of thickening it, because the purpose of the compliant structure is to utilize the flexibility as a mechanical or kinematic means.
In the case of the caps in the shampoo containers, the pressure is focused on the thin polymer section that connects the cap to the base. The disposable forceps are based on a very similar design. When the stresses are channeled into a thin and well-defined area, the bending is called "accumulated compliance". Scientists have studied the cumulative compatibility since the 50s. Recently, excellent work has been done on the subject by Ashok Midha of the Missouri University of Science and Technology, Larry Howell of the Brigham Young University, Shoria Avter of the University of Michigan, and Martin L. Culpepper of the Massachusetts Institute of Technology (MIT), demonstrating applications of incremental compatibility in precision machinery and positioning devices. particles at the nano level.
A bow that shoots arrows, on the other hand, does not have a focused bending zone: it implements "distributed compliance" over its entire length. Distributed compatibility is essential for building flexible machines that must withstand large loads, for example, wings that must hold an airplane in the air, or engines that must make millions of revolutions continuously. When I started my research in the field, I could not find any theoretical basis or general methods for designing machines with distributed compatibility. Naturally, I focused my efforts on this topic, and it is the one that keeps me busy to this day.
start small
I started working on flexible single-unit machines not because they seemed like new and intriguing gadgets to me, but because in some applications, assembly-free design is necessary. My career began with learning large mechanical systems, such as vehicle transmissions. However, in the early 90s I found myself designing really tiny machines: micro electromechanical systems (MEMS). What led to this were the circumstances of the time: communication companies began to develop optical switches for the fiber optic networks, and used tiny motors to quickly change the angle of mirrors that route optical signals in one direction or another. Shortly after I started reading Vogel's books and researching elastic design, I started working on a project in collaboration with Stephen Rogers and his team in the Microsystems Department of Sandia National Laboratories, in which monoform design seemed like a perfect solution.
Sandia sought to build a linear motor with a displacement output of at least ten microns, but the manufacturing constraints of electrostatic motors limited the motion to two microns. I knew you couldn't just minimize a transmission of gears, for example. Even if we were to find a person with hands steady enough to assemble wheels, axles and racks to the size of a micron or two, the resulting product would still not be accurate enough for modern engineering. At the MEMS scale, machines with a tenth of a micron of freedom are about as useful as children's toys. Also, similar to integrated circuits, MEMS devices are manufactured in batches of tens of thousands in an area the size of a fingernail. Therefore, I designed a monoform motion amplifier, which produced a 20 micron motion when combined with the electrostatic motor.
In 1998, the engine and amplifier worked well. I clearly remember standing in the lab and marveling at the tiny device. It operated continuously for more than ten billion cycles with no signs of fatigue, but to my mind, the most impressive thing was that the entire motion amplifier, for its complexity and flexibility, was made of a single piece of polysilicon.
flexible flying
In my opinion, of all the reasons why I chose to engage in matching design, the most attractive is the shape adaptation, or "shape transformation". The ability to change the geometry of a structure in real time allows nature's machines to operate with maximum efficiency. Compare this adaptability to the rigid geometries of the engineering world: drive transmissions in cars, airplane wings, engines, compressors, fans, and the like. All of them, and in fact all other machines that are designed using conventional methods, operate at maximum efficiency only under very certain conditions, and the rest of the time their operation is less than optimal. An airplane, for example, experiences a variety of flight conditions during the transition from point A to B: changes in altitude, winds and even weight depending on fuel consumption. That is, the plane operates almost all the time at a lower efficiency than it is capable of. Birds, on the other hand, can take off, land, hover and dive by effortlessly adjusting the configuration or shape of their wings as needed.
In the mid 90's I wondered if anyone had ever tried to change the camber of a wing in flight to improve its performance. I was amazed to learn that the Wright brothers, in their first airplane, used a different type of shapeshifting: the twisting of the wings. Then I discovered that changing the curvature of the wing according to the different flight conditions remained, for decades, an elusive goal. That's why I sat down one night at the table in the dining area and started working on a new plan.
After a few months of research, I came across a small news item in a newspaper about flexible wing research conducted in the late 80s at the US Air Force's Wright-Patterson Base in Ohio. The engineers there called the goal of their research a "mission adaptive wing" (MAW). I didn't know what the results of that study were, but I realized that a shape-shifting wing was not a far-fetched idea, and I contacted the researchers to ask if they would be interested in testing my design. Their response was stunning.
They explained that most, if not all, past attempts to create a shape-shifting wing used rigid structures: heavy and complex mechanisms with dozens of powerful motors to change the geometry of the wing structure. In one case, engineers installed flexible panels in the wings of an F-111 fighter jet. Their adaptive wing was aerodynamically promising, but the overall structure was too heavy and complicated for practical needs.
This did not surprise me. Designing a practical wing with variable geometry requires meeting many conflicting requirements. The wing should be light, strong enough to withstand an air load of thousands of kilograms, reliable enough to operate for hundreds of thousands of hours, simple to manufacture and maintain, and resistant to exposure to chemicals, ultraviolet radiation and large changes in temperature. The software tools and the thinking tools at that time were not suited to the design of monoform machines, and certainly not ones capable of meeting so many competing demands.
The flexible wing design I submitted to Wright-Patterson took advantage of the elasticity of the experimental materials, which were completely normal materials in the aviation world. The wing had an internal structure designed to change shape easily when a compact internal engine applied power, and still remain rigid when strong forces were applied externally in a wind tunnel test. The senior engineers at Bright-Patterson were excited about the new design, and so was I. In fact, I got so excited that in December 2000 I founded a company called FlexSys to develop practical applications for compliant design.
Six years later, following a lot of development work and several successful experiments in a wind tunnel, we were able to arrange to attach a prototype of the flexible wing to the underside of a Scaled Composites White Knight aircraft, for test flights in the Mojave Desert. The wing is placed under the body of the jet plane, and equipped with all the necessary devices to measure lift and drag. The lift coefficient ranges from 0.1 to 1.1 without increasing drag, and this means increased fuel efficiency of up to 12% in wings that will be designed to take full advantage of the new flexible racks. Such shelves that will be attached to existing wings will result in an improvement of about 4%. Considering that US aircraft consume about 60 billion gallons of jet fuel each year, even these seemingly small percentages can be important. The wing is also simpler, with no moving parts in the shape-shifting mechanism. As a result, it is more reliable, and its weight-to-power ratio is good. more.
The real test for adaptively shaped aircraft wings will be when flexible control surfaces completely replace conventional racks. We close the corners of such a product: FlexSys, in cooperation with the research laboratories of the US Air Force, designed and built a continuous surface that bends (arches) and curves along the wing spar to significantly improve aerodynamic performance, instead of the usual rear shelves that create drag. We replaced the regular wings of a Gulfstream Aerospace GIII model with FlexFoil surfaces with controlled variable geometry, and apart from the significant fuel savings, our design should also reduce the noise of the aircraft. According to NASA, most of the noise during takeoff and landing of airplanes originates from vortices created at the sharp edges and gaps between the racks and between the fixed parts of the wing. Our design has transition surfaces that eliminate these gaps. Test flights are scheduled to begin in July 2014, at the Neil A. Armstrong Flight Research Center at NASA.
crawling and twisting
In recent years, my students Joshua Bishop-Moser and Girish Krishnan and I have embarked on research into elastic design inspired by the world's most flexible natural machines: skeletal animals. The strangest of these life forms, such as annelids and nematodes (two types of worms), behave in ways we are only beginning to understand. More familiar examples, such as octopuses, are an ideal aspiration in the eyes of elasticity engineers.
Soft-bodied animals, such as worms or octopuses, have no visible skeletal structure, yet are able to move with vigor and grace. Most of the time, they do this with the help of what is known as "elastoploidy" (elastic flow). In engineering terms, their body is a hydrostat, that is, composed of an arrangement of connective tissue fibers and muscles, which surround a space filled with compressed fluid. Anatomical studies of such animals usually reveal a criss-cross coiled arrangement of fibers and muscle surrounding the internal organs in the fluid core. These fibers resist the fluid pressure created by muscle contraction, and the direction of the fibers determines the range of motion.
In the animal world there are many versions of hydrostatic skeletons. The arms of the octopus are muscular hydrostatics, and in the trunk of the elephant, compressed muscle fibers are activated around a hydrostatic body. The eel's skin, which is reinforced with fibers, is like an external tendon that allows the eel to create a strong driving force for swimming.
Our research in elastoploidy is still in its infancy, but our hypothesis is that such components could be used as components for "soft robots" and other devices, which could operate safely with humans and the environment. The first applications will, most likely, be in the field of orthopedics. For example, a person suffering from arm cramping due to muscle stiffness, joint deformity, or a stiff joint could benefit from a flexible orthopedic device, which will gently force the arm back into a functional position for day-to-day functioning.
Compatibility is welcome
With the help of many talented students in the Compatible Systems Design Laboratory at the University of Michigan, the basic research we began in 1992 yielded a wealth of useful insights and avenues for systematic design. These students, too numerous to list here, are now working on their own elastic design research at Penn State, Illinois at Urbana-Champaign, and Chicago, Bucknell University, NASA's Jet Propulsion Laboratory, Sandia National Laboratories, the US Air Force Research Laboratory, At KLA-Tencor, at Ford Motor Company, at FlexSys, at Raytheon And Intel. Thanks to the work of the talented engineers at FlexSys, we have completed the weather resistance tests and the preparation of the monoform wiper frame, and negotiations are underway with vehicle manufacturers to integrate the product into rear wipers The monoform is made of glass-filled thermoplastic polymer, and it works well both in freezing conditions and in heat. It doesn't break or warp even when ice and snow build up on the windshield. When it hits the market, it should be much more durable and reliable, and cheaper to manufacture, than any competing device.
Technically, our flexible aircraft wings are practically ready for commercial application right now. Replacing the outer edges of existing shelves with a sub-shelf with variable geometry for a cruise flight can, alone, save 5% of jet fuel consumption. Replacing the entire rack with jointless FlexFoil will provide approximately 12% fuel savings in new aircraft. It may take a few more years before we receive approval from the American Aviation Administration, but we believe that once the industry learns to trust the flexible wings, they will completely replace the hinged racks, in all types of fixed-wing aircraft.
There are many places in the automotive, electrical, medical, and consumer industries where elastic design can dramatically reduce the number of parts in any given device. The biggest challenge is to bring the news to the industrial designers. Extensive use of our innovative products, such as the matching wiper, will strengthen the status of the elastic design. Even then, a significant challenge will still remain: there is no convenient software tool for elastic design today. Today, FlexSys develops such software under contract with the US National Science Foundation.
It will be a few years before the elastic design reaches any kind of critical mass, but in our opinion, eventually its widespread adoption is inevitable. The power, precision, versatility and efficiency that elasticity is able to provide will give engineers in many fields a new and complete toolbox - once we know how to appreciate the power that flexibility can give us.
____________________________________________________________________________________________________________________________________________________________________
About the author
Sridhar Kota is a professor of engineering at the University of Michigan, as well as the founder and CEO of FlexSys.
in brief
The flexibility of machines designed by humans originates from complicated, and often inefficient, systems of rigid parts. Often, strength and flexibility are at odds with each other.
Elastic, or conformal, design is an engineering approach that embraces flexibility and spreads loads across secondary-form devices made from as few parts as possible.
This approach may yield new machines, such as shape-shifting airplane wings or robotic snakes, as well as ways to improve the durability and efficiency of devices of all kinds.
The base
How to build a robotic snake
The lack of a rigid skeleton does not pose a problem for worms, octopuses and other molluscs. They get along thanks to elastoploidy. Their body is made up of muscular tubes interlaced with fibers, which surround spaces filled with compressed fluid. The fibers exert force against the fluid pressure created when the muscles contract, and the direction of the fibers determines the range of motion. Engineers at the University of Michigan are developing "robots" based on the same principle. The applications for this could include orthopedic devices that would assist in the movement of the limbs, and robots that would be able to manipulate delicate objects and work safely alongside humans.
More on the subject
Better Bent Than Broken. Steven Vogel in Discover, pages 62-67; May 1995.
Videos showing some of the devices
The article was published with the permission of Scientific American Israel
One response
Nice article