CAD Malarky

 

by Alfred Scott

This article appeared in the December 1990 Falco Builders Letter.

The surest way to spot a phony in the aircraft business these days is if the promoter is spouting a lot of CAD/CAM stuff at you. The hucksterism is so obvious that it's surprising that people, particularly the media, are taken in by this nonsense.

In a recent article, the erstwhile kit promoter, Ken Wheeler, described the design of the Wheeler Express. "I just laid out the curves on my CAD system and then sat back and let the CAD system finish the design." What arrogant nonsense. That's roughly equivalent to saying that you wrote the opening paragraph and then let your word processor finish a novel. And what enormous stupidity of the magazine that printed such trash.

The Prescott Pusher was a classic piece of CAD/CAM flimflammery. All the emphasis was on the McDonnell-Douglas CAD/CAM system, the ability to rotate the airplane on the screen, the ability to precisely determine the angles of the cuts of the steel tubing, the ability to take sections through the fuselage at any angle.... They were embracing technology, and everybody got warm fuzzies just thinking about it. And somewhere in all of that, people lost sight that Tom Prescott sat at the keyboard of his computer and designed one of the really terrible airplanes of modern times.

He used square tubing when he should have used round, and then had the audacity to claim that square tubing was superior. The whole thing was assembled with pop rivets, something no competent engineer would do. The airplane was so short-coupled that part of the vertical tail actually overlapped the wing. And the landing gear was so far aft that one writer told me he got it stuck on the ground on takeoff roll and couldn't even pull the nose up.

So let's clear up some of the nonsense about CAD. I can give you a relatively good perspective on the subject. I'm a self-taught draftsman, and I've done all of the drawings for the Falco. In the process, I've spent years at the drawing board, and I take pride and some pleasure in turning out a really nice drawing. It's hard work, and it takes years to develop the skills and techniques to produce highly detailed isometric drawings with things going every-which-way.

In working on the Falco, by necessity I've had to learn how to design things, and I'm familiar with the mental processes that you go through. I know that drafting is only an incidental part of a process that consists primarily of problem-solving and decision-making.

I'm also familiar with CAD systems. I've started looking at them in 1984, and have used graphics software of various types for our construction manual. I've done a fair amount of 'beta' testing of new graphics software, and I use a CAD system that is extremely powerful yet which is light-years ahead of the competition in ease of use. I've been a beta tester on that software and am in frequent touch with the head of the software company and with the programmers. I've made a lot of suggestions that have gone into the program and have written a number of macros which add functions to the program.

I like CAD. It's one of the marvels of modern technology, but you need to understand what CAD is and what it is not. What it is good for and what it can't do. When to use it and when to use a pencil.

I think the biggest misconception is that because something is designed on CAD, it is somehow better. I have people come into my office and when they hear that I have a CAD system, they smile warmly, nod and approve. Then they point to the intricate engine/baffling/cowling layout drawings on my wall and feel reassured that I am embracing modern technology. That's when I explain that the drawings were done by hand.

CAD is just a drawing tool. What is important is the brainpower that goes into designing something, and it makes no difference if the designer used a pencil or a computer. And I've found that with CAD, you get lost in the details almost from the beginning. You lose a sense of the overall perspective of the design and a sense of proportion.

When I am designing, I like to sit with chin-in-hand and gaze for hours at the drawing, think, measure and scribble. That process of sitting with elbows on the drawing board and gazing over the drawing is where it all happens. Your eyes move over the design, and the mind wanders as you consider alternative methods of doing things. During this process, you find yourself in something of a trance. Routine daily activities and conversations become barely-noticed interruptions of a thinking process that will consume you completely. You struggle to appear attentive to conversations when your mind is really on the thing you're designing.

At some stage, the design starts to please you, and you pin it on the wall, and gaze at it while you work on other things. The rate at which sparks go off in your head slows down, and you begin to feel comfortable with the design. At some point you find that you can no longer find ways to improve the thing-it's as simple and elegant as you can make it -- and you consider the design finished. Your mind rarely returns to the part. It's finished. It's done. It's right.

The best designers have coldly, and often cruely, analytical minds. Their job is to weigh the relative merits of two alternatives, and it's imperative that you never give the slightest favoritism to your own ideas. You must be willing to throw weeks of work in the trash when a better idea comes along. The worst designers cling to their own ideas and are more fascinated by the tools of their trade then by the things that they are designing.

I asked Dave Thurston about his experience. Dave has been doing some work for a company in Houston and while Dave cranks out finished drawings with his pencil, they have wasted $150,000 playing with a 'powerful' CAD system and have literally nothing to show for it. Dave Thurston is completely wedded to his pencil and has no desire to ever own such a system.

Stelio Frati got a CAD computer for his office a few years ago, and I was curious how he had adapted to it. Frati works in an unusual way, working out the overall dimensions on paper and then building the parts. Drawings are produced only after the parts are made and serve as a record of what they have done. Frati is a pencil-and-slide-rule man, and rarely even uses an electronic calculator. He scribbles on a piece of paper at his desk, and on occasion they use the computers at the local university for finite element analysis for the design of a complex part.

Frati's drawing are done by a number of draftsmen, led by Carlo Maccabruni, who has been with Frati since the SF.260. Maccabruni works for Frati on the weekends, and during the week works on a large CAD/CAM system at another company. When they first installed that system, Maccabruni was very excited about it, but after a few years he told Mr. Frati that he prefers to design with his own system, the "Mac-Man"-as in Maccabruni-Mano (mano is Italian for hand). Like Thurston, Frati and Maccabruni are devotees of the pencil, the drafting pen, the ruler and the circle template.

One of the problems with CAD is that much of the industry is dominated by people who know a great deal about CAD but very little about drawing. The programmers, the managers of software companies, writers, industry pundits, CAD consultants, managers of CAD departments are almost entirely composed of people who can compare specs and features ad infinitum, but who couldn't draw our isometric Falco cowling installation drawing with a pen if their life depended on it. Most have little comprehension of what it means to sit at a drawing board and draw, and very few actually understand the mental processes of designing. Instead they sell systems based on arrogant claims that border on fraud. A friend told me, "I learn something every day that my CAD system won't do."

The head of the company whose CAD system I use -- and who has spent years sitting at a drawing board -- agrees that the industry is perhaps 15 to 20 years away from creating a system that is as natural an extension of a designer's thoughts as a pencil on paper. He concentrates his efforts on the things that the system is useful for, and makes none of the usual outrageous claims.

CAD is most successful when the user is allowed to make the purchasing decisions and the decisions of when to use it. I'm a passionate devotee of both CAD and the pencil, but the important thing to remember is that CAD is just another tool-phenomenally useful in some circumstances, and a frightful impediment in others.

The key word in CAD is repetition. It generally takes longer to draw things on a computer, but if you have to use it repetitively, CAD is a truly wonderful invention. It's a benefit to you, the Falco builder, mainly in that it makes it possible for me to generate a whole series of drawings that show how the wing is assembled. There's no way I could do that manually.

The other useful thing about CAD is that you can use it as a central database for the geometry and dimensions of the design. This makes it possible for the product to be brought to the market faster, sometimes more cheaply, and with accurately designed parts. The Malibu is an example where CAD was used effectively. So was the Stealth bomber, where there never was a prototype. They just made all the parts and knew from the beginning that they would all fit together.

It's also possible to do that without CAD, indeed most of the parts I designed for the Falco were never built as prototypes -- and that includes the cowling and baffling. I knew from the beginning that they were dimensionally correct. It isn't particularly hard to do with something as relatively simple as the Falco; you just have to check everything very thoroughly.

The CAM part of CAD/CAM is simply a programming convenience for the machinist. By making parts on computer-controlled machines, as we have always done, you get an accurate, repeatable part. There's no magic in that, everybody does it that way, but I can't imagine why a customer would care if the machinist used a programming aid to make his life simpler.

I've also seen examples where well-intentioned people get lost in the details of the CAD system and lose an overall sense of a design. I think the SwiftFury is an example of that. Roy LoPresti told me he gave Stuart Millar a budget of $50,000 for computers and eventually spent $500,000. They leased for $20,000 some software from Lockheed that permitted them to move the wing and reduce the pressure drag of the airplane by 20% (pressure drag is actually a relatively small part of total drag) but then had to clamp 20 pounds of lead on the engine mount of the 180 hp SwiftFury to keep it in balance.

All of us who visited the LoPresti skunkworks were surprised by the number of people concentrated around the computers in the office and how few were out in the shop. For a tiny fraction of the man-hours that those Italians spent resurrecting an old American design, this American modernized an old Italian design. And at the same time that the LoPrestis modified the Swift, Frati's shop of equal size designed and built from scratch the Penguino and also produced a second retractable-gear version. With slide rules, pencils, and drawings with cat paw prints all over them.

Years ago, an aspiring writer sent a manuscript to Truman Capote who commented, "That's not writing, that's typing." And when I see someone working on their CAD machine, I mumble to myself, "That's not designing, that's drafting."

What really counts in a design is the brains, skill and experience of the designer, and a passion for simplicity and elegance in design. The tools mean very little. So the next time you hear someone spouting a lot of CAD/CAM talk, you can be sure he's just a promoter.

 

 

Go back to CAD & WildTools     Welcome to WildTools