The density r of a bar or rod is the mass per unit length. But if the bar is not uniform (perhaps not uniform thickness), the density will not be constant, so to really have a good idea of the density of the rod at a given point x along its length, we have to take the infinitesimal ratio of mass per length at x . If you think of the mass of the rod up to the point x as m(x) , then the density is the rate of change of m(x) , r = dm/dx .
Later on, in other courses, you'll get more interpretations of this notion of density. Clearly, really, density should be mass per unit volume, but that requires, in its full generality, techniques of third-semester calculus. So, for the time being, we'll stick to a one-dimensional model.
Example 3 Assume that the thickness of a wire increases slightly along the length, so that the mass up to x meters is m(x) : = 3x+x2/10 grams. Find the density three inches from the beginning of the wire.
r(x) = dm/dx = 3+x/5 , so r(3) = 3.6 g/m .
The total cost of producing x items is called (surprisingly) the cost function, C(x) . The derivative, C¢(x) , is the marginal cost, and represents (approximately) the cost of producing the next item. The word ``marginal'' in economics refers to a rate of change, so that your ``marginal tax rate'' is the tax percentage on the next dollar you earn.
Typically, the costs involved in producing x items include a certain amount of fixed costs, which would include the cost of the factory, machinery, and other things you need no matter how many of the widgets you produce. In addition, materials and labor costs typically are proportional to the number of items produced, but there might be some costs that grow more quickly as the number of items increase, such as warehousing costs. So, a typical cost function might look like C(x) = 300+20x+x2/10 .
Example 4 Presume that the cost function to produce x widgets is C(x) = 300+20x+x2/10 . Then, find the marginal cost at the production level of 100 widgets.
The marginal cost is C¢(x) = 20+x/5 , so the marginal cost at a production level of 100 widgets is C¢(100) = 40 .
The Leibniz notation for derivatives, df/dx , isn't supposed to mean a
fraction, really. But, in reality it does have some meaning like that. Since
the derivative is the limit of fractions, if h is near
a , then (f(a+h)-f(a))/h will be near f¢(a) . That
is, replacing a+h by the more generic x and h by x-a :
|
or, solving for f(x) ,
|
Another way to look at this is that the equation of the tangent line L to the curve y = f(x) at (a,f(a)) is y = f(a)+f¢(a)(x-a) . Since the tangent line is near the graph, the y will be almost the same as f(x) .
The point here is to approximate f(x) , which might be a very complicated function, by a simpler function - a linear one, which is why it is called linear approximation.
We sometimes write Dx for x-a (or h ), to indicate
the change in x . In that notation:
|
|
Example 5
Find an approximate value of
___
Ö4.2
. Solution:
Example 6 Use differentials to find the relative error (the percentage of error) in the computation of the volume of a cube, if the measurements of the sides might be off by as much as 1%.
Related Rates problems are one of the classic types of ``word problems'' in calculus classes. They provide a first introduction to mathematical modeling: taking a problem, distilling from it the mathematical essentials, then, using standard mathematical techniques which can be applied to any problem of the same general type, solve the resulting mathematical problem.
For the problems we'll be looking at now, the underlying mathematics is quite simple. Let's say we have two functions, f and g . We can make them both functions of t to keep in the spirit of many of these applications. Then, if we know a relationship between f and g (some equation involving them), we can differentiate both sides of the equation to get a relationship between their derivatives. In applications, we usually know something about one of the derivatives, and we want to figure out the other.
For example, if
|
|
an equation involving the derivatives of f and g (the rates of the functions).
Example 7
A boater has his boat tied to the pier by a long rope. He feeds the pier-end
of the rope through a winch and pulls his boat in towards the pier. The tide
is out, so the bow of the boat is 6 feet below the level of the pier. If the
winch is pulling in the line at a constant rate of 2 feet per second, how fast
is the boat moving towards the pier when there is 10 feet of rope between the
boat and the winch?
[Animation: Have a pair of arrows to run the winch; one arrow pulls it in,
the other out. Also put in a stop button. The winch always runs at a fixed speed
of 2 feet/sec, and the idea is to see the boat moving faster or slower as the
distance to the dock changes. It's be nice if (1) we came up with a rationale
for why the boat moves outward when the rope is played out (offshore wind?),
(2) if the boat is allowed to run up to the dock, it should crash into the dock,
damaging the dock some, and the boat should sink...]
Copyright (c) 2000 by David L. Johnson.