Basic realities of a virtual world
January 1, 2008
Before a company purchases a CAD software package, the management team should start with a list of priorities that it expects the software to address. Of course, these priorities should be tailored to the company's mission. It's also important to keep in mind some of the basics of the virtual design world before the actual purchase takes place.
My fantasy is to have my thoughts telepathically linked to a computer—resulting in smarter and faster designs. It would be even more helpful if the computer would dream up solutions for me. Now that would be computer-aided design!
However, mind reading is a little beyond what even the best CAD software can do. In most cases, the computer's aid is limited to improving the way the concept is visualized, analyzed, and documented. The human talent for invention and vision is still a central contribution to the design process; the computer is just an assistant that takes care of some of the routine chores.
In the last paragraph, I used the term the "best CAD software," which can be construed as a loaded phrase. It is easy to fall into that trap.
"Fully featured" is not necessarily the same as "best." The best tool is the one that is productive—the more efficient the better. Even though a system may be suitable for a massive project like building the shuttle to Mars, it might not be the best system for a narrow project like manufacturing an enclosure for an electronic device.
My focus is on job shop manufacturing. The criteria I use for evaluating CAD software includes compatibility with machine controls (CNC programming); acceptance among my customers and suppliers (business network); and support for quality control (documentation standards and revision control).
I take it for granted that the software will let me display the design dynamically from any angle, and I also expect that the software's operation will be consistent with other office computer products.
If you're considering an investment in a CAD system, I recommend that you start with a similar list of priorities tailored to your business mission. It's also important to keep in mind some of the basics of this virtual design world.
One of the most important concepts bandied about among CAD experts is "design intent." So what does the idea of design intent encompass? Here are a few considerations:
That last point is probably the most important aspect of design intent. As the design progresses, it will evolve through stages of initial concept, optimization of function, improvements in manufacturability, and incorporation of suggestions from people who have used the product. If the CAD work is done properly, making revisions to the 3-D model can be efficient and painless.
Another fact that is common to nearly all design tools is that there are generally several ways to do what appears to be the same thing. A circle might be drawn as a single arc that spans 360 degrees, as a collection of arc segments, or as a polygon with many short sides. Which CAD modeling method is best? That depends on the design intent (seeFigures 1a and 1b).
People who use CAD tools frequently master only a subset of all of the software functions available. They find a path through the maze of menus and icons that gets the job done. It may not be the perfect way, but given the pressures of time, getting it done is really the first priority. As the CAD operator's skill set expands, the ability to comply with the full scope of the design intent will improve.
In an ideal world, all of the decisions made during the CAD modeling process are made in accordance with the design intent. The reality is that things are done often by brute force (or by whatever force is available). Further, it is often difficult to imagine how the design will change in the future. Nevertheless, an outline of the design intent is a good place to start when launching a CAD project.
Delving further into CAD jargon, let's review what is meant by "2-D." A flat piece of paper offers a two-dimensional surface (horizontal and vertical). Such a physical understanding of 2-D is common not just to CAD people, but to normal humans as well.
No matter what is drawn on that surface, it is physically a 2-D surface. An artist can draw an image on that 2-D surface that looks like a realistic 3-D object. That "3-D" image on the paper can be viewed only from the single angle that the artist chose. The observer can interpret the 2-D image as a 3-D representation, but in reality it is still just a 2-D piece of paper.
Computers extend the capability of a piece of paper by allowing observers to select the point of view instead of having the artist choose it for them. The image can be rotated with the mouse to look at it from any angle. Even though the observer is interpreting the image as 3-D, it is still physically a 2-D projection on the computer screen.
In CAD terminology, a distinction exists between the physical 2-D display and the 2-D or 3-D model. "2-D" implies drafting, and "3-D" relates more to visualizing.
Let's talk about drafting for a moment. In the tradition of producing mechanical drawings, the artist (also known as the draftsman) selects points of view that are in accordance with a drafting standard. These standard views also are referred to as projections. In the U.S., the common standard calls for third-angle projection, which defines the relationship between views on a drawing.
Third-angle projection is sometimes explained using the idea of the object sitting in a hemispheric punch bowl (seeFigure 2). As you look straight down into the bowl, you see the top of the object. If you slide the object up to the rim of the bowl without twisting it, you'll see the object from its side. Sliding the object to different spots on the rim reveals other sides (or projections) of the object.
In other parts of the world, first-angle projection is the standard used for drafting view projections (seeFigure 3). First-angle projection flips the part as you would naturally when examining it if it were resting on a table.
Which drafting standard is best? This parallels the debate involving metric versus imperial units of measure. The "right" method is the one that your business network uses.
Most mechanical drawings that are used in manufacturing are 2-D drawings with standard projections. This tradition has served us well for more than 200 years. Only 40 years ago, nearly all mechanical drawings were generated by hand on paper. Computers began to replace rulers and pencils about 30 years ago. The software at that time essentially was aiding in drawing flat projections—the birth of 2-D CAD. The drawings were somewhat easier to generate and revise with a computer, but drafting was still a tedious chore.
The innovation of 3-D CAD modeling is that the primary effort goes into creating an object that can be seen on the computer screen from any angle (as opposed to putting all of the effort into 2-D drafting). Once the 3-D object is created, the computer system automates the generation of the projected views on the mechanical drawing. Figure 4was made with just a few mouse clicks once the 3-D model of the toy engine was complete.
The entire vocation of the draftsman has been replaced almost entirely with a mouse click. A goal of most 3-D CAD systems is to automate the drafting of 2-D mechanical drawings. This routine chore falls into something that the software programmers can handle; it is governed by fully developed standards, and very little innovation is tolerated. That makes it perfect for computer automation. It is a blessing to relieve designers and inventors of the mechanical drafting chore.
However, the mechanical drawings are being produced for the benefit of other humans. The drawings are a tool for communication of the design intent. I still see a need for involvement of designers and engineers in preparing the drawings—with most of the tedium safely delegated to the computer. That leaves the 3-D modeling of the design as the primary focus for the CAD operator.
When training new CAD operators, I recommend that they start every project with a pencil and a blank piece of paper. Doodling is an important first step in the design process. It is quick, noncommittal, and very mnemonic. By "mnemonic" I mean that a random pencil mark can represent an entire cascade of thought that is meaningful to the inventor in spite of lacking any realistic aspect. A few occasions may arise when the designer might use a compass or ruler, but such hardware subliminally restricts the free flow of ideas at the conceptualization stage.
As an employer, how do you tell whether or not your CAD staff is daydreaming or working? Sometimes the only clue will be the nature of the doodles and the frequency of agonized mumbling.
Once the seed of the invention has been planted with the pencil, it is time to grab the mouse and start drawing in a more orderly way. The CAD software does a terrific job of making straight lines and smooth curves. However, the computerized tool requires mastery of some rules. Those rules of operating the software are what I mean by "CAD overhead." The more CAD overhead, the more disruptive it is to the inventive process.
Even though the eventual goal is to create a fully detailed 3-D model, sometimes the best way to work out an idea is to draw things in quick 2-D sketches and not worry at all about the realistic details like thickness, texture, or color (see Figure 5). This plan of attack early in the design process minimizes the CAD overhead and maximizes the creative experience.
Some CAD software systems have tools to group drawing elements (lines, arcs, etc.) so that they can be moved in a coordinated way. Among other things, these "blocks" can be used to represent elements of a mechanism. It is relatively easy to sketch linkages, belts, or gears with simple rectangles and circles. With a little mastery of the CAD software, you can "animate" the sketch to visualize how the relative motions behave. By simply dragging things around with a mouse you elicit a realistic set of motions. Such a functional 2-D sketch quickly lays the foundation for a full 3-D model.
That is the question—is it better to "click and arrow" the menus for full 3-D modeling or rough out a skeletal sketch at the outset? I'm guilty of starting some projects in full 3-D mode. It is a form of doodling that sometimes helps me work through a design problem.
Over the years I've become a master of some CAD tools to the point where they are nearly as second nature as using a pencil. The choice of where to start a design project is best made on a case-by-case basis. In the end, I feel it is best to develop a full 3-D model and then use it to generate the 2-D mechanical drafting. All revisions are made to the 3-D model; updates to the 2-D drawing then follow automatically. Unless the design will never be revised, 3-D modeling confers a significant advantage in design management.
I have observed dozens of people as they progress through various stages of skill with a variety of CAD software systems. I have yet to meet a person who is an instant CAD master. The state of the art in CAD user interface is still laden with CAD overhead. Unlike a pencil and paper, most of the rules are not self-evident. I offer that sad commentary on software to point out a basic CAD fact: Training and practice are required to master CAD tools.
Gerald would love to have you send him your comments and questions. You are not alone, and the problems you face often are shared by others. Please send your questions and comments to email@example.com.
Gerald would love to have you send him your comments and questions. You are not alone, and the problems you face often are shared by others. Share the grief, and perhaps we will all share in the joy of finding answers. Please send your questions and comments to firstname.lastname@example.org.