Chapter 2 Differential Calculus 2.1 Higher order derivatives If the derivative y/ = f"(x) of a function y = f(x) is itself differentiable at ir, we can calculate its derivative, which we call the second derivative of f and denoted by y” = f"(x). Other notations used include Py dd fa pq) 24 — © 4) — Dty— Dt "= I"@) = B= BEI = Fall) = Div = DY) In general, the n'* derivative of y — f(r) is dy a dn de” y) = f(a) = f(z) = Dry = Dsf(z) and is defined to be the derivative of the (n — 1)* derivative. For n = 1,2,3, primes are still normally used: f(x) = f"(x), f(x) = f"(x). Sometimes we denote f(x) = f(x). Example 2.1.1 The velocity of a moving object is the rate of change of the position of the object with respect to time. If the object moves along the x-avis and is al position x = f(t) at time L, then ils velocity al that lime is given by dey =S-s'0 v= S70 Similarly, the acceleration of the object is the rate of change of velocity. Thus, the acceleration is the second derivative of the position: Cr any de dee 172, y" = 6, yl") = 0 and all ”, then Example 2.1.2 If y = x3, then y! = 32?,y” higher derivatives are zero. In general, if y yz) = n(n —1)(n— 2)... (n= (k= 1))a"™* -{ at poh fOSk n. For n (—3)(1 +2) * ~3'(—4)(1 + a) = 41+ 2)> (a) eeee We are now able to predict the n'* derivative from the above pattern, which is yO) = (-1)"nl(d $2) However, this does not prove that this formula is true for every n. So to complete the proof, we use mathematical induction on n. We see from the first derivative above that the formula is true for n = 1. Suppose that the formula is valid for n = k, where k is some positive integer. Consider y/*! vt Sym 2 yhaa ey = (-1)AY(—k = 1). +2) (aye + DIG +2) This is what the formula predicted for the (k +1)" derivative. Thus, if the formula is true for n = k, then it is true for n = k +1. Therefore, by induction, the formula is true for all n > 1 Example 2.1.5 Find a formula for f(x), given that f(x) = sin(ax + 6). 19Solution We again begin by calculating a few derivatives and see if we can predict the n't derivative S'(2) = acos(ax + 5) S"(z) = a? sin(ax + b) — -a? f(x) f"(x) = —a¥ cos(ax + 6) = a f"(x) f(x) = a! sin(ax +b) = a' f(x) f(x) = a® cos(ax +b) = a f'(x) We can see from the pattern above that each derivative is —a? times the second previous one. Therefore, a formula that gives all derivatives is (k=0,1,2,...) m,) — J (har sin(ar +6) ifn = 2k s y= {ennai ifn =2k+1 Example 2.1.6 Calculate J’, f", and f" for f predict f from the pattern? VP FI. Can you Solution Since f(x) = (2? +1)? we have Ho) = $e? + Hn) = (a + 1)°8, Ia) = (8 + AYE al FN + 18x) = (+), : —32(2? +1)" f"(2) 2? 4 1)-3(22) ‘The pattern is not obvious to predict f without calculating it. In fact, fe) = 3(42? = 1)(a? +1)-7 So the pattern doesn’t become clearer even at this stage 2.2 Implicit Differentiation Curves are generally the graphs of equations in two variables. $ can be written in the form equations: F(x,y) =0 20where F(r,y) denotes an expression involving the two variables x and y. For example, a circle of radius 5 and centre at the origin has equation Flay)=2?t+y?-5. ‘To find the derivative of F', we use a technique called implicit differentia- tion. In this technique, we differentiate F with respect to x while treating yas a function of x having derivative # or y’. Example 2.2.1 Find # ify? = Solution We differentiate both sides o rule: he equation with respect to x using the chain dy Te) = Ge) 2 gh Note that y? = x implies y) = yZ and y. dy dx which agrees with the above derivative. Example 2.2.2 Find the slope of the circle 2? + 4) Solution ‘This equation of the circle combines two graphs namely y: = VZ5—z? and 2 = ~V%—2. Observe that the point (3, —4) lies on the graph of yx. So we can find the slope by calculating explicitly dy 26 6 dz 3% — a8 225-9 But we can also differentiate implicitly with respect to x: dy ay d d d dy “(2") + 2) = £023) 2 vo qe) + EW) = F(28) > 2x 4 BVT = 0 ‘Therefore, the slope at (3, —4) is So why implicit differentiation if we can explicitly differentiate with respect to x? Well, it is not all functions that we can easily express y as a function of x as the following example shows. Example 2.2.3 Find # if ysing = 23 + cosy, 21Solution Here, we can only use implicit differentiation. We have d d Flvsin) = d 5 : Fle!) + F(oosy) de a a > sine + ycosx = 32? — sinyS i hs dy => (sine + siny) = 32? — ycosr dy — 32? —ycosx dc sing +siny Example 2.2.4 Find an equation of the tangent to x2 + xy + 2y* = 4 at (-2,1). Solution Note that (2,1) lies on the curve. ‘To find the slope of the tangent, we differentiate implicitly with respect to x to get Qe +y+ cy! + Gy*y' =0. Substituting 7 = —2, y= 1 and solving for y’ we get —4+1—2y' + 6y! Osy ‘Thus, the slope of the tangent at (2,1) is §, and its equation is 3 4 y Git +2) +1 or 3x — Ay = -10 Example 2.2.5 Find £4 if xy + y? = 2x. Solution We implicitly differentiate twice both sides of the equation. Thus yt ay! + Quy! = 2 (2.2) and solving (2.2) for y’ we get y-2¥ (23) 22Implicitly differentiating (2.2) again with respect to x we got vty tay" +20) + 2uy" = 0 (x + Quy)y" = —2y' — 2(y/)? so that wy? nw 2y = 2 = ww! 2.4) y ety (24) Substituting (2.3) into (2.4) we get 22g) ~ HY? xy 2-y (2-y)? 2(5 1 ar) 3(E 7 ar) =2(2 = v(x + 2y +2—y) (ery = 222 = we ty +2) ~ + 2ype -2(2n — cy + 2y — y? +4 2y) ~ e+ ay (e+ 2y)* = me (since 2x — cy — G+ 2uP (We have used the given equation (in the question) to simplify the numera- tor.) 23