During the past couple of decades, design engineers have benefited from spectacular improvements in the tools on which they depend. In the 1980s there was the development and rapid acceptance of 2D CAD systems. Subsequently, these were largely supplanted by 3D wire frame systems and, later still, by systems that were capable of solids modeling.

But while these developments unquestionably were advances in mechanical design, they did not represent a fundamental shift in the nature of the design process itself. Regardless of whether it is carried out on a drafting table or in front of a computer display screen, the process essentially consists of engineers documenting designs that they conceive in their heads.

Today, however, the demands on designers are growing, as manufacturers try to outpace competitors by being first to market with new products that are quickly supplanted with upgrades to continually differentiate their products. In addition, customers' desires for products tailored precisely to their individual needs mandate that designs be highly flexible and subject to many modifications. Clearly, designers need tools that can take them to the next level in design, advancing the nature of the process itself.

PTC, which has been a pioneer in mechanical design and engineering since the mid-1990s, has created this breakthrough technology in the form of Behavioral Modeling, part of the Pro/Engineer suite of software. Rather than just serving as a drawing board for the engineer's concepts, Behavioral Modeling lets the system create the design automatically based upon the knowledge it gleans from designers and engineers and the objectives they set for a particular design.

Because all of the essential behavioral data defines the model, engineers are freed from repetitive, time-intensive tasks necessary to achieve ideal performance design. As a result, they are able to focus more of their attention on issues, making better use of their creative talents.

For example, Loral Space Systems recently had to determine how the distance from a new satellite's center of gravity to the thrust vector changed as a panel was extended. This changing distance determined how the amount of thrust should vary. Typically, it can take as long as four weeks to generate the problem and more than 80 hours to solve it, said Scott Hersberg, manager of Loral's propulsion group. Using Behavioral Modeling, however, Loral was able to set up the problem in 30 minutes and calculate the solution in a mere five minutes.

Similarly, Nismo -- the racing arm of Nissan -- sought to make three separate pipes of equal length without violating a minimum bend radius. Its design engineers used Behavioral Modeling's optimization capabilities to find the geometry that satisfied this requirement. Previously, an engineer would have spent an entire day tweaking curve geometry manually and painstakingly analyzing the results. By automating the problem-solving process through the use of Behavioral Modeling, Nismo had its solution in six minutes.

Solving these particular problems leads to continued benefits from Behavioral Modeling, since the application was able to capture knowledge related to solving this particular problem. Other designers can avail themselves of that knowledge on future design problems.

Behavioral Modeling involves:

• Defining behavioral features. These features can be used to drive the design itself. For example, behavioral features capturing a desired weight or volume can be used to drive the size of a lawnmower gas tank. Or, an angle of reflectivity from a surface can be measured and captured within a behavioral feature that can be used to drive the surface's curvature. Behavioral features that easily capture these complex or custom measurements also can be grouped and stored in libraries for reuse or access by other members of a design team.


• Assess model feasibility, sensitivity or optimization and understand the effects of change on design objectives. Behavioral Modeling enables engineers to assess the behavior of designs through design studies that provide insight into how changes affect the model. It also determines whether the desired changes are feasible. This data is conveyed through real-time design updates and easy-to-read graphical results, including graphs and colored fringe plots. Engineers also can study the dynamic results of animating their assemblies through interactive dragging and user-defined motions. Furthermore, behavioral features capturing scattered measures can be used in optimizations. The resulting design takes more environmental factors into account.


• Design Exploration. Once design models are generated, Behavioral Modeling drives the designs based on specific goals and criteria applied to those models. In other words, Behavior Modeling automatically tries out all of the design iterations engineers would have had to produce manually. For example, to achieve optimal engine performance, engineers can assign constant flow-rate through an intake manifold as an objective on which Behavioral Modeling will create a set of designs automatically that explore the design space.

Click on the link to learn more about Behavioral Modeling.