×

Manufacturing

When Surfaces and Solids Collide

31 Dec, 2003 By: Don LaCourse

---

Hybrid modelers overcome the limitations of other modelers.

Welcome to my updated column, Modeling Methods. Each month I'll dig deeper into the methods that today's mechanical CAD applications use to help product designers be more creative and productive in what they do best.

Figure 1. We'll use this fitness massager to show how hybrid modelers surpass other modelers.

This month, I'm pleased to have Brad Redmond working with me. Brad is director of industrial design marketing with think3 in Cincinnati, Ohio ( www.think3.com). As you'll see, think3's thinkiD application is a great example to use for this month's focus.

Surface vs. Solid Techniques

When it comes to the mechanical CAD techniques used to develop new ideas, product designers can be categorized into two groups: those who use surface modeling and those who use feature-based solid modeling. You can further define them as conceptual designers who prefer surface modeling for creativity and detail designers and engineers who prefer solid modeling for precision. Both solutions have tremendous benefits and significant drawbacks.

Surface modeling gives you the freedom to design as you please. You can create complex shapes with a simple click and drag, and you can tweak the shapes until you end up with what you want. The drawback is that many surface-only modeling applications don't record the history of how you achieved the design. Therefore, simple changes such as increasing the radius of a fillet or the diameter of a hole can be arduous and time consuming. You must re-create everything from the point of change forward.

Solid modeling provides great repeatability. Shapes aren't described as shapes but rather by semantics such as the diameter and depth of a hole. A solid modeling application records every action so you can make updates, such as changing fillet radii or hole diameters, by modifying the semantic to the new desired value. The computer handles the rest, rebuilding the entire shape according to the recorded history. The problem is that it's often difficult to describe a shape by semantics alone, so you end up compromising to complete the design.

Figure 2. It would be difficult to quantify the crown with solid modeling techniques, but with surface modeling you can shape the top surface to conform to style and comfort.

Realizing that designers want the benefits of each type of software, mechanical CAD developers introduced hybrid modelers that contain the best of both worlds. These programs combine the ease of complex shape creation in surface modeling with the semantics and repeatability of solid modeling. You can change a surface with a click and drag, and the shape updates. Modify the value of a semantic, and again the shape updates. The obvious benefits are the ease of shape creation by tweaking to your heart's content with the precision of feature detailing via semantics.

In the Real World

For an example of how hybrid modeling is applied, we'll use think3's thinkiD and its GSM (global shape modeling) techniques (figure 1). The mostly blue handle of this fitness massager is stylistic and designed to match human ergonomics. It's a great illustration of why you can't always classify shapes with semantics. How would you quantify the crown at the top of the handle (figure 2)? It's easier to shape the top surface to conform to style and comfort. On the other hand, the lower red housing has a specific mechanical function that must be defined using precise design semantics to ensure that it functions properly. You can easily change mechanical features, such as the countersunk holes for the fasteners, by changing the semantic value and letting the computer do the rest. However, this comes at the expense of a blander style. It's essentially a rectangular block with rounded edges. The ability to merge style with semantics is critical for the design of this product. Adding some shape to the highlighted surfaces (in orange) uses techniques from both surface and feature-based solid modeling (figure 3).

Figure 3. It's critical to be able to merge style with semantics. You can add some shape to the orange highlighted surfaces using surface and feature-based solid modeling techniques together.

You can define the target shape with points, freeform curves, sketches, surfaces, and solids-whatever you need to get to the solution. In hybrid modeling, you're not as limited by the application. In this example, the target shape is defined by the purple and yellow curves.

Roll Back History

Rolling back the history of a solid model lets you apply shape modification prior to some of the critical features. Almost every designer is constrained by some facet of the design-cost, size, weight, or some other criterion that must be met or preserved. Preservation is key. Let your mind wander, but make sure you stay in the realm of possibilities. In this case, we want to preserve the sharp edge surrounding the pink faces in figure 4. Depending on the geometry type, you can preserve projection, plane, position, tangency, curvature, and any valid combination of these characteristics.

Another important factor is the amount of control you have, not just the type. With think3's GSM functionality, you can define as many preservation and deformation conditions as you need. In figure 4, I want to preserve the sharp edge and deform the shape from the green vertical to the bent purple curve and the horizontal blue to the crowned yellow curve.

Figure 4. You can use the history tree to change a key element and yet preserve another, depending on your geometry.

Apply the Change

When you apply the change, the result deforms precisely to your desired shape. You can add a little curvature to the side of the component to soften the mechanical feel while preserving the intent of the embossed shape (figure 5). Applying this shape modification would be time consuming, if not impossible, using only solid modeling techniques. Although you can achieve the rough shape quickly with a surface-only modeling application, the hybrid modeler preserves associated modifications to all of the affected geometry and captures the design intent in the part's history just like a conventional solid modeler.

You can leverage GSM for this shape modification and reapply all of the subsequent features in the part's history. This type of change remains a simple click and drag operation while preserving the precision of the design.

Mix It Up

In this design, the chosen modeling method is to perform global shape modifications on a surface that appears early in the part's history. A surface-modeling technique is used on a basic semantic-driven solid shape. Hybrid modeling is one of the best solutions for product design where shape is a priority.

Figure 5. When you use a hybrid modeler, you apply the change and the result deforms to the shape you want.

Many feature-based solid modelers today offer improved surface modeling capabilities. Though this design challenge is within their scope, they're not yet at the point of seamless integration you find in hybrid modelers. They haven't completely eliminated the suffocating effects that semantics can place on creativity. On the other hand, surface-only modeling applications haven't taken significant steps to adopt the repeatability and precision of solid modelers. Hybrid modeling is an environment where the best of both collide to benefit product design.