TABLE OF CONTENTS The next few pages will examine explicitly what we've already seen in some of the homework problems the question of how to get a good enough answer instead of an exact one. Although this chapter and its problems could have been broken up and distributed among the earlier chapters, the approach seems to be distinctive enough to merit a more unified presentation. To a large extent the ability to do good seat-of-the-pants engineering is an acquired skill, one that comes after having gone through the tedium of producing exact results for a great many problems. The difficulty for the student is that the whole idea of seat-of-the-pants analysis is very seldom referred to in any text even though it underlies much of the successful engineering of this world. Really good engineers rarely sit down and precisely analyze a problem at the start. Rather, they consider many options, get quick approximations, redesign, get some more approximate solutions, and then finally zero in on a final solution. Only after a particular problem has been adequately pinned down do really serious analysis tools come into play. This chapter attempts to give the student some sense of what can be accomplished by stepping back and thinking about a particular problem in a general way before bringing out all the analytical tools the preceding chapters have given us.
Seat-of-the-pants engineering refers to the ability to analyze a system reasonably well, but not exactly, in a short time.
