- Buyers Guide
The Meshing of Two EM Giants
ANSYS Inc. develops, markets, and supports engineering simulation software used to predict how product designs will operate and how manufacturing processes will behave in real-world environments. The company was founded in 1970 and, for nearly 40 years, has pioneered the development and application of computer-based simulation methods to solve the most challenging engineering problems. ANSYS employs approximately 1,700 employees, many of whom are engineers with advanced degrees and extensive training in fields such as finite element analysis, computational fluid dynamics and design optimization.
ANSYS advances simulation solutions by developing or acquiring the very best technology; then integrating it into a unified and customizable simulation platform that allows engineers to efficiently perform complex simulations involving the interaction of multiple physics; and, finally, providing system services to manage simulation processes and data — all so engineers and product developers can spend more time designing and improving products and less time manipulating software and searching for data. In 2008, ANSYS acquired Ansoft Corporation and expanded its domain into the RF and microwave world.
Jim Cashman, president and chief executive officer of ANSYS, Inc., has helped the company advance its leadership position in the engineering simulation field by tackling the toughest of problems with advanced technologies. Zol Cendes is chief technology officer and general manager of Ansoft.
MWJ: (to Jim Cashman) Could you give an overview for our readers of the range of technology from ANSYS and the industries you serve?
JC: ANSYS is a leader in advanced engineering simulation. Our technology is used by engineers to design and simulate products on the computer without resorting to physical prototypes. Our vision is to provide OEMs with a single platform in which all of their physics can be analyzed — from electromagnetics to thermal to structural to mechanical, whether singly or interactively. We have a very strong presence in the automotive, aerospace, chemical processing, power, and semiconductor industries.
MWJ: What was it about Ansoft that made you interested in acquiring them?
JC: The ANSYS corporate vision and strategy is based on providing unequaled technological breadth and depth. Breadth refers to a wide range of physics simulation capabilities, and depth refers to the ability to consider the entire complexity of a product, leading to an accurate solution. These capabilities can be either internally developed or acquired. The Ansoft acquisition allowed us to acquire best-in-class electromagnetic simulation capabilities for high-performance electronics and electromechanical design, shaving considerable time from a development standpoint.
MWJ: How do the technologies of both companies complement each other?
JC: ANSYS and Ansoft are both engineering simulation companies, and both have been pursuing similar corporate visions. At ANSYS, our vision is called Simulation Driven Product DevelopmentTM. The idea is to create tools that can simulate full systems in their natural operating environment early on in the process, where simulation can be used to drive new solutions rather than merely to verify existing ones. This allows for detailed “what-if” examinations of potential concepts before design problems are locked in and expensive to correct. It also allows for new innovations to be fully vetted and defects to be avoided at the most economical time. For example, 10 years ago, state-of-the art capabilities would have involved the simulation of air flow around an airfoil. Today, however, automotive engineers aren’t simply designing rear-end spoilers in a vacuum, but rather vehicles that must navigate banked tracks while crossing oil slicks at high speeds. The vision, then, is to capture the car, track, and oil in a simulation.
Individual companies have been developing technologies that will allow them to simulate these higher-order systems. The difference has been each firm’s respective starting point. In order to solve more complex electromagnetic system-level behavior, Ansoft realized it needed thermal and structural capabilities. For example, electronic package densities have become so high that electrical engineers are spending more time considering thermal issues in their package designs. On the other hand, ANSYS realized that it needed robust electromagnetic capabilities in order to solve more complex mechanical system-level designs.
Clearly, the two companies were on a convergent path. Most important is that our customers are supporting this vision. Take the automotive industry, for example: More than 40 percent of the total value of the next generation of automobiles will come from onboard electronic and electromechanical components.
MWJ: How integrated will the two companies become? Will R&D be merged?
JC: Organizational concerns are really the means to accomplish our vision. Our combined teams of technologists will be the primary determiners of that, and the details of how to achieve the vision may change from year to year, based on their guidance. The most important goal is providing value to our vast customer base, and this almost always involves sharing our technology among all product teams. This concept has guided other significant ANSYS integration activities.
MWJ: Do the combined companies make ANSYS a unique solution provider and if so, what kinds of problems will the combined company be able to address that no one else will be able to?
JC: Without question, ANSYS and Ansoft together have a very unique value proposition for Microwave Journal readers as well as our entire customer base. Advanced microwave electronics systems have reached a complexity in which traditional RF techniques alone are no longer sufficient to create optimal designs. For example, the design of an RF power amplifier must now consider how the device’s thermal performance will impact the system’s overall frequency response. We believe optimal designs can be accomplished only when both issues — electromagnetic and thermal performance — are considered in an interdependent manner. Together with Ansoft, ANSYS tools will solve problems like this. Moreover, today’s microwave electronics systems are incorporated into demanding system-level environments. A power amplifier may be part of a phased array antenna that is mounted in a UAV that is expected to circumnavigate the globe. Vibration analysis is a critical component to this integration. What impact will in situ vibrations have on the antenna’s beam steering capabilities? Only an integrated “multi-physics” solution can provide the answer. Tools from ANSYS are built around this vision. In virtually every industry, the mechanical and electrical worlds are converging. If most products are moving to combinations of mechanical and electronic interdependencies, it would seem to make sense that the verification of these products with coupled simulation would be of great benefit.
MWJ: What changes might Ansoft customers experience with the new organization?
JC: When the joint ANSYS and Ansoft vision is executed properly, our customers can expect to solve their hardest, most challenging engineering problems before fabrication ever begins. This should translate into shorter design cycles and faster times to market. In addition, Simulation Driven Product Development can be a vehicle used to lower business risk, such as reduced warranty exposure, by capturing engineering problems before they reach the manufacturing floor.
MWJ: Many of our readers are in the defense market. Is this an area where ANSYS had a strong presence even before the Ansoft acquisition?
JC: Yes, before the acquisition, aerospace and defense represented more than 15 percent of our broad-based business. With Ansoft, our support of and presence in the aerospace and defense industries will expand.
MWJ: Given the current economic climate, where does ANSYS see the best opportunities?
JC: Right now, we’re staying focused on the basics — on blocking and tackling, to use a football metaphor. In an uncertain macroeconomic environment, the key is to remain focused on the critical issues. We have a plan in place to address these issues, and we intend to deliver on that plan. And, there are reasons for cautious optimism. Yes, a few of our customers are experiencing some financial distress. Yet, the way out for these customers is to produce innovative, market-winning products. Our simulation solutions are a key part of that answer.
MWJ: (to Zol Cendes) Is there a particular technology in the ANSYS arsenal that could be incorporated into Ansoft products to enhance them?
ZC: There are great synergies between our technologies. Both companies have finite element solvers that require 3-D modeling, meshing, coupled matrix solutions, and a design environment with design management and post-processing. In this regard we can share common and disparate technologies to deliver the best technological solutions to our customers. Ansoft pioneered such techniques as tangential vector finite elements and automatic adaptive meshing. Likewise, ANSYS has delivered advancements in solver technology and meshing. Our R&D teams will work together to identify the best of both, and we will deliver those methods across multiple physics including electromagnetics. Two areas where our customers can gain some immediate benefit are direct bi-directional links to third-party CAD modeling and mesh morphing. ANSYS has built the ANSYS Workbench platform with a remarkable link to tools such as AutoCAD and SolidWorks. The parametric 3-D CAD geometry data remains within the native CAD environment and links to the ANSYS solvers via procedural interfaces. In the future, Ansoft users will be able to take advantage of this design automation capability. Models can be simulated for electromagnetic behavior and optimized for performance within the simulation environment with immediate connection to the original model. On mesh morphing, parametric geometry changes can be made, and the mesh can stretch to follow the geometric changes. This is a remarkable ANSYS technology that we intend to deliver to our electromagnetic customers in the future.
MWJ: Will we be seeing links between ANSYS products and Ansoft’s EM products?
ZC: Yes. We already have performed linkage between some of the ANSYS solvers and utilities in our R&D department. Your long-time readers may remember the cover feature that the Microwave Journal published on HFSS version 1.0 back in 1990. One of the geometries shown on that cover was a coax-to-multistep waveguide adaptor. We parameterized that model and re-simulated it using the optimization tool called ANSYS DesignExplorer to produce not just the minimal return loss but also a complete response surface using Design of Experiments to illustrate the most sensitive design point. Although this is a simple example, we nevertheless were able to show linkage between ANSYS and Ansoft products for robust design. Of course, there are also much deeper levels of integration. We are very excited about the multiphysics linkage between ANSYS and Ansoft, such as the tie between electromagnetics and thermal. There are two ways to solve these linked multiphysics problems: a multi-field approach and a direct coupled-field approach. The multi-field approach has two separate solvers linked by data interfaces with a separate, sequential solution and, possibly, a separate mesh. This is useful when the physics are dependent upon one another but not strongly coupled. An example is a high-power microwave component that warms up. The direct coupled approach is used for physics that are much more tightly coupled. In this case, direct coupling at the finite element level is used on a single model and common mesh. Think of that same high-power microwave component where the dielectric properties change dramatically as the temperature rises.
MWJ: Can you provide a preview of what customers might expect to see in the next version of HFSS?
ZC: HFSS version 12 will be released in 2009. Of course, we continue to concentrate on our flagship product in addition to the integration with ANSYS because we know that our customers are relying on us. Version 12 has new features that address the scale of modern electronic and RF capabilities that our customers have asked of us. We are not quite ready to announce all of the new features, but your readers should know that the next version will allow much larger simulations on traditional multiprocessor computers. Version 12 will take advantage of the trends in multicore and compute clusters that enable our customers to perform parallel computations on large problems. We have developed significant advances in finite element meshing that provide even more robust meshing of imported 3-D geometries. I think our customers are going to be pleased with the numerous significant advances in HFSS version 12.
MWJ: Both companies are located near Pittsburgh, which is a bit unusual for an engineering software company. What are the origins of the two companies?
ZC: As your readers may know, I was a professor at Carnegie Mellon University here in Pittsburgh. My research group produced the inventions that made finite elements reliable for full-wave electromagnetics. We founded Ansoft in 1984, released our first 2-D products soon after, and delivered HFSS in 1990. We have since made significant investments in circuit simulation so we can solve today’s system-level problems. ANSYS was founded in 1970 by Dr. John Swanson, who previously worked at Westinghouse Astronuclear Labs in Pittsburgh, where he developed stress analysis simulation of components in nuclear reactor rockets. Since that time, ANSYS has grown to more than 1,700 employees with over 60 sales and support offices on three continents and 21 major development centers worldwide.
MWJ: Will any products from either company be incorporated into the other company’s design environment, dramatically changing the user interface?
ZC: There are no immediate plans to significantly alter the user interface on any products for the sake of product integration. Our plan is to focus on our customers and the engineering problems they wish to solve. Abruptly changing the interface is neither in our interest nor our customers’ interest. Enabling new capability in terms of links to 3-D CAD, meshing, design optimization, and multiple physics is very much in our customers’ interest. We intend to pursue these capabilities first. That is, we will focus on the engineering and the challenging problems to solve. As we go forward in the coming years, it is expected that we will unify the multiphysics solution into a common environment, namely, the ANSYS Workbench. This will enable a much broader and comprehensive solution for our customers.