Simulation Software

Bruce Jenkins | Ora Research

“Air Force Awards Contract for GURU to Put the Simple in Simulation” was the headline of a news release issued by software developer MSBAI announcing it was awarded an AFWERX Small Business Innovation (SBIR) Phase 1 contract to examine integrating its GURU technology with U.S. Air Force applications.

“Engineers use computer simulations for anything from airflow over wings to thermal analysis of the hot section in gas turbines,” the company notes, citing typical defense-related uses. “The problem is the simulation software is too complicated to learn so they’re not getting the most out of it”—a familiar complaint heard from highly skilled engineers and discipline leads who are not however specialists trained in the arcana of CAE.

 

MSBAI is a privately held small business located in Los Angeles, CA, focused on development and deployment of its GURU cognitive AI assistant for engineering, and is now an Air Force Techstars 2020 company. The Air Force says this program, formally known as Air Force Accelerator Powered by Techstars, “focuses on the next generation of technologies for unmanned systems, human-machine interfaces, and immersive training.”

AFWERX describes itself as a “community of Air Force innovators who strive to connect Airmen to solutions across the force: whether that be funding, collaborating with industry, or simply receiving guidance on a project. We were established in 2017 by the Secretary of the Air Force, report to the Vice Chief of Staff of the Air Force, and are comprised of active duty, Air National Guard, Air Force Reserve, Air Force Civilian Service, and contractor personnel.”

GURU: “AI-driven assistant that learns to run the complicated software itself so you don’t have to”

MSBAI characterizes GURU as an “AI-driven assistant that learns to run the complicated software itself so you don’t have to—minimizing the human workload needed to translate engineering questions into computational workflows. With cloud systems already offering the compute power of government supercomputers from not long ago, it takes more time to set up a structural/thermal/fluid/trajectory analysis than it does for the computers to run them. The newest exascale and coming quantum systems will require this kind of AI layer for humans to be able to keep up.”

According to MSBAI, “there are numerous dual-use applications that come from enabling more engineers to use the best design and analysis software and deploy it at high-performance computing scale: manufacturers stand to gain a 500-to-1 return on investment, and the DoD will save billions of dollars in aircraft sustainment and gain advantages in rapid reaction.” Also, especially relevant in context of the COVID-19 pandemic, “GURU’s commercial deployment is SaaS B2B, and it will be a game-changer for remote work.”

Techstars director KATZ: “Will enable engineers throughout the DoD…to make many more trials per day and enable many more engineers to use these impossibly complex tools”

Warren Katz, managing director of the Air Force Accelerator Powered by Techstars program, remarked, “As an engineer who struggled with these overly complicated simulation software packages myself, I felt the pain that GURU relieves. The award of this Phase 1 SBIR to MSBAI will ultimately enable engineers throughout the DoD that are working on our toughest problems in hypersonics, quantum computing, heat transfer, optics, electromagnetics, fluid mechanics, etc. to make many more trials per day and enable many more engineers to use these impossibly complex tools.”

AFRL and AFWERX have partnered to streamline the Small Business Innovation Research process in an attempt to speed up the experience, broaden the pool of potential applicants, and decrease bureaucratic overhead. Beginning in SBIR 18.2, and now in 20.1, the Air Force has begun offering “Special” SBIR topics that are faster, leaner and open to a broader range of innovations than before.

MSBAI

Air Force Accelerator Powered by Techstars

AFWERX

Bruce Jenkins | Ora Research

Doug Neill, a 22-year MSC Software veteran whose career includes 11 years as VP of the company’s R&D group, has founded a new company providing software and services for ICME—integrated computational materials engineering. Computational Engineering Software, LLC’s mission is to help engineering and manufacturing organizations “optimize your composite design through advanced simulation.”

CES details the heretofore unsolved problems it is attacking: “Current composite design is nearly 100% empirical. In aerospace, the building-block process requires thousands (or hundreds of thousands) of tests of increasing complexity, from simple laminate coupons to structural elements, to details to components to full-scale test, to develop the design space. Design is constricted by what laminates have been tested. Exploring new design concepts or materials isn’t feasible because of the testing required.”

The company cites typical complaints from users of traditional tools and methods:

• “We can’t explore new design concepts. We have no reliable way to quickly evaluate them without testing, and there’s no time for that.”
• “We’ve done 30,000 coupon tests and we still don’t know why our matrix is failing.”
• “For our chopped-fiber parts, we have to proof-test every design change, and one in every eight parts in production. Why can’t we have an analysis method that will give us a reliable margin of safety, like we do for metals?”

What CES proposes is to “allow you to break out of the restrictive building-block process by using analysis to examine the constituent materials (i.e., fiber and matrix) independently, and find where the onset of failure in the material occurs—hence the name Onset Analysis. When does my matrix start to become critical? When does my fiber become critical? When (and where) is a delamination likely to start? When will the fiber overcome shear-lag and start to unload? Those are the critical questions in design that Onset Analysis can answer, for any layup, any geometry, without a mountain of test data.

“Onset Analysis also enables you to look at your design under extreme environments, like elevated temperatures, moisture, cold dry, or even the extreme cold of space applications.”

How this differs from older composite simulation methods

CES says that previous failure criteria such as Hashin, Tsai-Hill, Tsai-Wu and others “assume that a ply is a homogeneous material, and ignore the fact that there are fibers in a matrix and out-of-plane loads. They’re engineering approximations—curve fits of specific ply failure tests that create a failure envelope.”

By contrast, the company describes Onset Analysis as “actually a throwback to a classical way of doing things. It’s essentially the von Mises approach, evolved and adapted for composites. In Onset, we separate the strains in the fiber and matrix, looking at the full 3D state of strain, and evaluate them independently.”

Not “damage progression”

CES emphasizes that Onset Analysis is not the traditional “damage progression” approach to composite failure analysis. “From a design standpoint,” it notes, “if your material has failed prior to its target load, the design has failed. Why go further?”

The Onset Analysis methodology instead “looks for the beginnings, the onset, of failure in the material as design criteria instead of catastrophic failure. Of course in composite design, it would be impractical (and unnecessarily limiting) to design to the first microscopic critical value, so our analysis can also look for the initiation of higher-order events, like fiber unloading, initiation of delaminations or even estimates of laminate unloading. All of this without damage progression.”

Onset Analysis can be used to evaluate static strength of composite details including bolted joints, bearing, bearing-bypass, and especially matrix-dominant problems such as bonded joints and bonded repairs. “It is the only current method that can accurately assess matrix issues,” the company declares. The method can also be used to look at problems such as compression after impact (CAI) and fatigue.

Offerings

Composite simulation—“Use critical properties of the fiber and matrix to predict critical matrix and fiber failures and compute margin of safety, for any layup or geometry, without laminate testing.”

Consultation—“We bring industry-leading expertise in composite performance to your design and analysis challenges.”

Software development—”Our experts in software development process, frameworks and execution can help set up your team for success.”

People

CES’s key people introduce themselves:

Doug Neill, CEO and founder—“I am an innovation-focused, action-oriented, transformational software executive with decades of experience in the computer-aided engineering (CAE) space. At CES, we believe that the ICME materials revolution is stagnating due to a lack of pragmatic, easy-to-use and fast methods for validation of designs comprised of ICME materials. I have spent my entire technical career focused on automated design and design engineering simulation tools. Our company brings this focus and passion to our customers so that they can innovate new structural concepts and effectively and efficiently unlock the benefits of engineered materials. We have a great team of software and engineering experts to assist you.

“I spent 22 years at MSC Software Corporation and was Vice President of the R&D group for 11 of those years. Prior to that, I helped a startup as the Business Development head to position new FE-based design software. At the beginning of my career, I spent 15 years working on the development and application of directed-search automated multidisciplinary optimization software in aerospace and later the CAE industry.”

See our interview with Neill in his VP role at MSC Software published in this blog in 2017.

Jon Gosse, Ph.D.—“I believe that simulation, done right, can radically transform industry. In my 33 years at The Boeing Co., I worked with many talented engineers and scientists to develop practical approaches to evaluating composite design, and applied those methods to solve intractable problems on some of the most important programs in the company. My goal is to bring that knowledge to the world and revolutionize how industry designs with composites and other advanced materials.”

Eddy (Joe) Sharp—“In my 31 years at The Boeing Co. I had many roles. I began my career as a stress analysis, then moved to the development of Boeing’s internal stress analysis software, engineering methods and material allowables, and finally it was my privilege to manage a group of brilliant engineers and scientists working on bringing together structural simulation and materials science to improve composite materials performance. Working with them every day was like getting an advanced degree in a wide-ranging curriculum including solid mechanics, mechanics of composite failure, polymer science and molecular dynamics modeling.

“That background led to a passion for advanced materials and composite analysis, and building practical tools for engineers so they can focus on the design of their part, and not on the quirks and nuances of modeling with esoteric CAE packages.”

Kunaseelan Kanthasamy, Ph.D.—“I embrace the Art of the Possible and am reputed for ‘taking any fresh, blue-sky project from zero to full steam ahead.’ As a visionary thought leader, I continuously move people, products and companies forward, always connecting the customer’s needs to the end product. In pioneering next-generation digital technology, I build high-performance R&D teams and devise ways to differentiate businesses.

“I thrive in igniting energy around conceiving new products from legacies, capturing new product channels to grow business, strategically anticipating future customer needs, and guiding approaches to the distinctive aspects of doing business in other cultures. My technical acumen includes engineering new digital platforms, pre-NPI, NPI, TRR; enterprise solutions, VAST, RFID, FQC Code, NFC standards/platforms; industry security frameworks, cryptography, symmetric/asymmetric, hash functions, encryption/signatures, Digital Twin and Digital Thread.

“I have filed 36 patents and own 21, and won 16 product innovation awards.”

Computational Engineering Software, LLC

Integrated computational materials engineering

von Mises yield criterion

Correctly defining boundary conditions is of vital importance when using finite element analysis (FEA). As with any simulation or calculation, if the input is rubbish, the output will also be rubbish. For structural analysis, the important boundary conditions are the fixtures preventing movement of the structure and the external forces acting on it.

Dr. Jody Muelaner, PhD CEng MIMechE

For wheels fitted with pneumatic tires, defining the boundary conditions properly can be particularly challenging. The forces acting on the rim of the wheel are caused by air pressure within the tire and by the ground reaction force transferred through the tire. Defining these forces in a realistic way requires some thought and without experience of this type of problem, the most significant force could easily be overlooked. Luckily, considerable work has been done in this area and so we can apply standard methods to correctly define the boundary conditions. This means that it’s not necessary to include the tire in the FEA and it can be represented by a few simple boundary conditions, determined by a combination of analytical and empirical methods.

Detailed models of tire-ground and tire-rim interactions
Before we get into the practical methods for applying boundary conditions to a wheel, let’s take a moment to consider the alternative. If we don’t make any assumptions about how the tire will impart forces on the rim, we would need to first model how the tire makes contact with the ground. This might be idealized as a toroidal or cylindrical face of the tire making contact with a planar ground surface. This would respectively result in a point or line contact initially, with corresponding infinite force. The tire then deforms to produce a contact patch. This type of contact analysis always requires an iterative solution but is relatively straightforward when the contact is between two linear elastic solids. However, although the compression of the air within the tire is elastic, the casing of the tire will undergo large deformations and there are significant damping effects. It is this damping that often accounts for most of the rolling resistance when a wheel rolls over a firm surface. Modeling these effects requires non-linear material models, adding significantly to the complexity of the analysis. The complexity doesn’t end there as the tire is not made up of a homogeneous isotropic solid material. Tires have anisotropic textile casings and bead wires, encased in a rubber matrix. Modeling this type of composite structure becomes extremely complex. It is only when all of this has been simulated that the deformation into a contact patch and the transfer of force from the ground into the rim can be determined.

Analysis of the forces acting on rim
A typical wheel consists of a central hub with a disk that supports the rim. The rim has flanges at each side which prevent sideways movement of the tire and bead seats which hold the tire radially. It is through the bead seats that the ground reaction force is transferred. The parts of a rim are labeled in the photo below.

The part of a tire which contacts with the ground is the tread. The tread generally has an approximately cylindrical outer surface, although it may be toroidal, especially for tires used on bikes, which lean into corners. In common usage, tread may only refer to the outer textured surface which makes direct contact with the ground but the term is used here to refer to the entire thickness of the horizontal part of the tire. The sides of the tire, which extend vertically to meet the tread at each size is known as the sidewall. The inner circumference of each sidewall is referred to as the bead. The bead is supported radially by the bead seat of the rim and is contained axially by the rim flange. The bead may contain wire to help it resist the radial force caused by the air pressure within the tire. The parts of a tire are shown below.

The air within the tire exerts uniform pressure on all internal faces of the tire and rim, the inflation pressure, P. The simplest boundary condition for the wheel is where any surfaces of the rim not covered by the tire are exposed to this pressure. Where the inflation pressure acts on the inside of the tread, it is contained by the tire casing and bead, causing internal hoop stresses in the tire but no reaction forces on the rim. Where the inflation pressure acts on the sidewall of the tire, it is resisted at the outer circumference of the sidewall by the tread and the inner circumference by the rim flange. The resulting reaction force, Fs, is an important boundary condition for the wheel. Note that in the below diagram, the force Fs is shown acting on the tire, the force acts on the wheel in the opposite direction.

To calculate the reaction force, Fs, we need to consider the axial component of the inflation pressure, and the area over which it acts. It acts over the area of the sidewall, projected onto the vertical plane, which is given by:

If we assume that the force on the sidewall is divided equally between the tread and the rim flange, the force is given by:

 

Ground reaction forces have three components: Radial force due to the vehicle’s weight, tangential force caused by acceleration and braking, and axial force caused by cornering. These forces are distributed according to a cosine function over a region of the bead seat and rim flange related to the tire’s contact patch. The range of this distributed load is given by an angle of loading, θ. There are no analytical methods to determine the loading angle. It depends on the shape of the rim, the shape and stiffness of the tire, the tire pressure, and the ground reaction force. The loading angle may be determined experimentally or a worst-case value may be used. To use the worst-case value, multiple simulations are carried out with different loading angles to determine how the angle affects the stress in the wheel. It can be assumed that it will not be smaller than the contact patch of the tire which can be easily observed. The ground reaction may create a cyclic loading in the wheel if the wheel contains periodic spokes. This is because the stress state when the ground reaction is centered on a single spoke is different from the stress state when the ground reaction is between two spokes. For every revolution of the wheel, each spoke will experience one cycle which should be taken into account for fatigue calculations.

There can be up to five forces acting on the rim, although axial and tangential reactions are not always present:
• Inflation pressure, P, acting uniformly on the internal faces of the rim, not in contact with the tire.
• Sidewall pressure reaction, Fs, acting on both rim flanges.
• Radial ground reaction, Fv, distributed sinusoidally over both bead seats
• Axial cornering reaction, FA, distributed sinusoidally over one of the rim flanges depending on cornering direction.
• Tangential braking or cornering reaction, FT, acts over the same region of the bead seats as the vertical ground reaction and is also sinusoidally distributed, with the tangential load transfer related to the normal force.

Practical issues of applying boundary conditions in FEA software
Before attempting to apply the tire forces to the rim, a few changes should be made to the solid model to simplify the analysis. Firstly, the faces of the bead seats must be split at the extents of the load angles. It is also a good idea to cut the wheel in half so that symmetry can be used to simplify the model. It may also be useful to de-feature the model to further reduce meshing and solution times. For example, by removing external fillets and other small features that won’t affect the stress significantly. It may even be worth extracting mid surfaces and meshing using shell elements.

In addition to the forces applied to the rim, fixtures will be needed at the hub. This is probably best achieved using a frictionless support on the inner face in contact with the hub and bolt connectors at the holes.

In this example, an inflation pressure of 0.345 N/mm2 (50 psi) was simulated. With a bead seat radius of 163 mm and a tread inner radius of 268 mm this results in a sidewall gives a reaction force of 24,525 N. The ground reaction force is 2 kN and the sinusoidal distribution can be applied using a bearing load. Because of the symmetry in the model these forces are halved.

The final von Mises stress results, following H-adaptive meshing, are shown below. The greatest stress occurs at the radius between the bead seat and the rim flange. This is a result of the bending stress caused by the sidewall reaction force. The ground reaction force has surprisingly little effect on the stress field, only increasing the peak stress by 3%. This really highlights how important it is to fully consider how the air pressure is transferred through the tire.