~ ·~ u v u onto ~ v: Surfaces Given z=f(x,y), the partial derivative of 1 pr~ D x ex = ex tan(sec (x)) v~u = ( ||~ v ||2 )~ v Ellipsoid z with respect to x is: p @z @f (x,y) Dx sin(x) = cos(x) =( p x2 1 if x 1) x2 y2 z2 fx (x, y) = zx = @x = @x Cross Product a2 + b2 + c2 =1 Dx cos(x) = sin(x) =( x2 1 if x p 1) likewise for partial with respect to y: Dx tan(x) = sec2 (x) u⇥~ ~ v @z @f (x,y) sinh 1 (x) = ln x + px2 + 1 fy (x, y) = zy = @y = @y Dx cot(x) = csc2 (x) Produces a Vector sinh 1 (x) = ln x + x2 1, x 1 (Geometrically, the cross product is the Notation Dx sec(x) = sec(x) tan(x) For fxyy , work ”inside to outside” fx tanh 1 (x) = 12 ln x + 11+x x, 1 < x < 1 area of a paralellogram with sides ||~ u|| Dx csc(x) = csc(x) cot(x) p then fxy , then fxyy Dx sin 1 = p 1 2 , x 2 [ 1, 1] 1 1+ 1 x2 and ||~ v ||) sech (x) = ln[ ], 0 < x 1 Hyperboloid of One Sheet @3 f 1 x x ~ u =< u1 , u2 , u3 > fxyy = @x@ 2 y , 1 1 ex e x x2 2 2 Dx cos = p , x 2 [ 1, 1] sinh(x) = 2 ~ v =< v1 , v2 , v3 > a2 + yb2 z c2 =1 3 @ f 1 x2 For @x@ 2 y , work right to left in the ex +e x (Major Axis: z because it follows - ) Dx tan 1 = 1 , 2⇡ x ⇡ cosh(x) = 1+x2 2 2 î ĵ k̂ denominator 1 p1 , |x| > 1 u⇥~ ~ v = u1 u2 u3 Dx sec = |x| x2 1 Trig Identities v1 v2 v3 Gradients Dx sinh(x) = cosh(x) 2 2 sin (x) + cos (x) = 1 The Gradient of a function in 2 variables Dx cosh(x) = sinh(x) 1 + tan2 (x) = sec2 (x) ~ v =~ u⇥~ 0 means the vectors are paralell is rf =< fx , fy > Dx tanh(x) = sech2 (x) 1 + cot2 (x) = csc2 (x) The Gradient of a function in 3 variables Dx coth(x) = csch2 (x) sin(x ± y) = sin(x) cos(y) ± cos(x) sin(y) Lines and Planes is rf =< fx , fy , fz > Dx sech(x) = sech(x) tanh(x) Hyperboloid of Two Sheets cos(x ± y) = cos(x) cos(y) ± sin(x) sin(y) Equation of a Plane 2 2 2 Dx csch(x) = csch(x) coth(x) tan(x)±tan(y) tan(x ± y) = 1⌥tan(x) tan(y) (x0 , y0 , z0 ) is a point on the plane and z c2 x a2 y b2 =1 Chain Rule(s) Dx sinh 1 = p 12 < A, B, C > is a normal vector (Major Axis: Z because it is the one not Take the Partial derivative with respect x +1 sin(2x) = 2 sin(x) cos(x) subtracted) to the first-order variables of the Dx cosh 1 = p 1 ,x > 1 cos(2x) = cos (x) sin2 (x) 2 x2 1 A(x x0 ) + B(y y0 ) + C(z z0 ) = 0 function times the partial (or normal) 1 1 cosh(n2 x) sinh2 x = 1 derivative of the first-order variable to Dx tanh = 1 x2 1<x<1 1 + tan2 (x) = sec2 (x) < A, B, C > · < x x0 , y y0 , z z0 >= 0 Ax + By + Cz = D where the ultimate variable you are looking for Dx sech 1 = p1 ,0 < x < 1 1 + cot2 (x) = csc2 (x) summed with the same process for other x 1 x2 1 cos(2x) D = Ax0 + By0 + Cz0 1 sin2 (x) = 2 first-order variables this makes sense for. Dx ln(x) = x 1+cos(2x) cos2 (x) = 2 Equation of a line Example: let x = x(s,t), y = y(t) and z = z(x,y). Integrals 1 cos(2x) tan2 (x) = 1+cos(2x) A line requires a Direction Vector Elliptic Paraboloid 2 2 z then has first partial derivative: R 1 sin( x) = sin(x) ~ u =< u1 , u2 , u3 > and a point z= xa2 + yb2 @z @z R x dx = ln |x| + c (x1 , y1 , z1 ) @x and @y x x cos( x) = cos(x) (Major Axis: z because it is the variable x has the partial derivatives: R e xdx = e 1 + cx tan( x) = tan(x) then, NOT squared) @x @x a dx = ln a a + c a parameterization of a line could be: @s and @t R ax 1 ax and y has the derivative: x = u1 t + x 1 R e dx = a e +c Calculus 3 Concepts y = u2 t + y1 dy p 1 dx = sin 1 (x) + c dt R 1 x2 Cartesian coords in 3D z = u3 t + z 1 In this case (with z containing x and y 1 2 dx = tan 1 (x) + c as well as x and y both containing s and R 1+x1 given two points: t), the chain rule for @z @z @z @x p dx = sec 1 (x) + c Distance from a Point to a Plane @s is @s = @x @s (x1 , y1 , z1 ) and (x2 , y2 , z2 ), R x x 1 2 Distance between them: The distance from a point (x0 , y0 , z0 ) to Hyperbolic Paraboloid The chain rule for @z@t is p R sinh(x)dx = cosh(x) + c (x1 x2 )2 + (y1 y2 )2 + (z1 z2 )2 a plane Ax+By+Cz=D can be expressed (Major Axis: Z axis because it is not @z @t = @z @x @x @t + @z dy @y dt R cosh(x)dx = sinh(x) + c Midpoint: by the formula: squared) Note: the use of ”d” instead of ”@” with |Ax0 +By0 +Cz0 D| R tanh(x)dx = ln | cosh(x)| + c x +x y +y ( 12 2, 12 2, 12 2) z +z d= p z= y2 x2 the function of only one independent tanh(x)sech(x)dx = sech(x) + c A2 +B 2 +C 2 b2 a2 R 2 Sphere with center (h,k,l) and radius r: variable R sech (x)dx = tanh(x) + c (x h)2 + (y k)2 + (z l)2 = r 2 R csch(x) coth(x)dx = csch(x) + c Coord Sys Conv Limits and Continuity R tan(x)dx = ln | cos(x)| + c Vectors Cylindrical to Rectangular Limits in 2 or more variables R cot(x)dx = ln | sin(x)| + c Vector: ~ u x = r cos(✓) Limits taken over a vectorized limit just R cos(x)dx = sin(x) + c Unit Vector: û y = r sin(✓) Elliptic Cone evaluate separately for each component sin(x)dx = cos(x) + c q (Major Axis: Z axis because it’s the only R z=z of the limit. p 1 dx = sin 1 ( u u|| = u21 + u22 + u23 Magnitude: ||~ one being subtracted) a 2 u2 a) + c Rectangular p to Cylindrical Strategies to show limit exists R Unit Vector: û = u ~ x2 y2 z2 1. Plug in Numbers, Everything is Fine 1 dx = a 1 tan 1 u ||~ u|| r = x2 + y 2 + =0 a +c a2 b2 c2 R a2 +u2 tan(✓) = xy 2. Algebraic Manipulation ln(x)dx = (xln(x)) x + c Dot Product z=z -factoring/dividing out u·~ ~ v Spherical to Rectangular -use trig identites U-Substitution Produces a Scalar x = ⇢ sin( ) cos(✓) 3. Change to polar coords Let u = f (x) (can be more than one (Geometrically, the dot product is a y = ⇢ sin( ) sin(✓) if (x, y) ! (0, 0) , r ! 0 variable). vector projection) z = ⇢ cos( ) Strategies to show limit DNE f (x) ~ u =< u1 , u2 , u3 > 1. Show limit is di↵erent if approached Determine: du = dx dx and solve for Rectangular p to Spherical Cylinder dx. ~ v =< v1 , v2 , v3 > ⇢ = x2 + y 2 + z 2 from di↵erent paths 1 of the variables is missing Then, if a definite integral, substitute u·~ ~ v =~ 0 means the two vectors are tan(✓) = xy (x=y, x = y 2 , etc.) OR the bounds for u = f (x) at each bounds Perpendicular ✓ is the angle between cos( ) = p z 2. Switch to Polar coords and show the (x a)2 + (y b2 ) = c Solve the integral using u. them. x2 +y 2 +z 2 limit DNE. Spherical to Cylindrical (Major Axis is missing variable) u·~ ~ v = ||~ u|| ||~ v || cos(✓) Continunity u·~ ~ v = u1 v1 + u 2 v2 + u3 v3 r = ⇢ sin( ) A fn, z = f (x, y), is continuous at (a,b) Integration by R R Parts Partial Derivatives NOTE: ✓=✓ if udv = uv vdu û · v̂ = cos(✓) z = ⇢ cos( ) Partial Derivatives are simply holding all f (a, b) = lim(x,y)!(a,b) f (x, y) u||2 = ~ ||~ u·~u Cylindrical to Spherical other variables constant (and act like Which means: Fns and Identities p u·~ ~ v = 0 when ? p ⇢ = r2 + z 2 constants for the derivative) and only 1. The limit exists 1 sin(cos (x)) = p1 x2 Angle Between ~ u and ~ v: ✓=✓ taking the derivative with respect to a 2. The fn value is defined z cos(sin 1 (x)) = 1 x2 1 ✓ = cos ( ||~ ~ ·~ u v ) cos( ) = p given variable. 3. They are the same value u|| ||~v || r 2 +z 2
Directional Derivatives Double Integrals Work Surface Integrals Other
p p Information Let F~ = M î + ĵ + k̂ (force) Let a p = a Let z=f(x,y) be a fuction, (a,b) ap point With Respect to the xy-axis, if taking an b b in the domain (a valid input point) and integral, M = M (x, y, z), N = N (x, y, z), P = ·R be closed, bounded region in xy-plane Where R R P (x, y, z) p a Cone is defined as û a unit vector (2D). R R dydx is cutting in vertical rectangles, ·f be a fn with first order partial z = a(x2 + y 2 ), The Directional Derivative is then the dxdy is cutting in horizontal (Literally)d~r = dxî + dy ĵ + dz k̂ derivatives on R In SphericalqCoordinates, R derivative at the point (a,b) in the rectangles Work w = c F ~ · d~ r ·G be a surface over R given by = cos 1 ( 1+a a ) direction of û or: (Work done by moving a particle over z = f (x, y) Polar Coordinates ·g(x, y, z) = g(x, y, f (x, y)) is cont. on R Right Circular Cylinder: Du ~ f (a, b) = û · rf (a, b) curve C with force F ~) This will return a scalar. 4-D version: When using polar coordinates, Then, R R V = ⇡r 2 h, SA = ⇡r 2 + 2⇡rh pn dA = rdrd✓ g(x, y, z)dS = limn!inf (1 + m n ) = emp Du ~ f (a, b, c) = û · rf (a, b, c) R RG Independence of Path R g(x, y, f (x, y))dS Law of Cosines: Surface Area of a Curve q a2 = b2 + c2 2bc(cos(✓)) Tangent Planes Fund Thm of Line Integrals where dS = fx2 + fy2 + 1dydx let z = f(x,y) be continuous over S (a let F(x,y,z) = k be a surface and P = C is curve given by ~ r (t), t 2 [a, b]; Flux ~ Stokes Theorem (x0 , y0 , z0 ) be a point on that surface. closed Region in 2D domain) r 0(t) exists. If f (~ ~ r ) is continuously R R of F across G Then the surface area of z = f(x,y) over ~ · ndS = F Let: Equation of a Tangent Plane: di↵erentiable on an open set containing R RG ·S be a 3D surface S is: R [ M fx N fy + P ]dxdy rF (x0 , y0 , z0 )· < x x0 , y y0 , z z0 > R R q C, then c rf (~ r = f (~b) f (~ r ) · d~ a) R ~ (x, y, z) = ·F SA = S fx2 + fy2 + 1dA where: Equivalent Conditions ~ (x, y, z) = ·F M (x, y, z)î + N (x, y, z)ĵ + P (x, y, z)l̂ Approximations Triple Integrals F~ (~ r ) continuous on open connected set ·M,N,P have continuous 1st order partial R R R M (x, y, z)î + N (x, y, z)ĵ + P (x, y, z)k̂ let z = f (x, y) be a di↵erentiable D. Then, ·G is surface f(x,y)=z derivatives f (x, y, z)dv = ~ = rf for some fn f. (if F ~ is function total di↵erential of f = dz R a Rs R 2 (x,y) (a)F ·~ n is upward unit normal on G. ·C is piece-wise smooth, simple, closed, 2 2 (x) f (x, y, z)dzdydx dz = rf · < dx, dy > a1 conservative) ·f(x,y) has continuous 1st order partial curve, positively oriented 1 (x) 1 (x,y) R This is the approximate change in z Note: dv can be exchanged for dxdydz in , (b) c F ~ (~ r ) · d~r isindep.of pathinD derivatives ·T̂ is unit tangent vector to C. The actual change in z is the di↵erence any order, but you must then choose R Then, , (c) c F ~ (~ r ) · d~r = 0 for all closed paths H R R in z values: your limits of integration according to ~c · T̂ dS = F ~ ) · n̂dS = (r ⇥ F z = z z1 in D. R R s that order (r ⇥ F ~) · ~ndxdy Conservation Theorem R F~ = M î + N ĵ + P k̂ continuously Remember: H R Maxima and Minima Jacobian Method R R di↵erentiable on open, simply connected Unit Circle ~ ·T F ~ ds = (M dx + N dy + P dz) c Internal Points f (g(u, v), h(u, v))|J(u, v)|dudv = set D. (cos, sin) R RG 1. Take the Partial Derivatives with f (x, y)dxdy F~ conservative , r ⇥ F ~ =~ 0 R respect to X and Y (fx and fy ) (Can use (in 2D r ⇥ F ~ =~ 0 i↵ My = Nx ) @x @x gradient) @u @v J(u, v) = @y @y 2. Set derivatives equal to 0 and use to solve system of equations for x and y @u @v Green’s Theorem 3. Plug back into original equation for z. Common Jacobians: Rect. to Cylindrical: r (method of changing line integral for Use Second Derivative Test for whether Rect. to Spherical: ⇢2 sin( ) double integral - Use for Flux and points are local max, min, or saddle Circulation across 2D curve and line Vector Fields integrals over a closed H R R boundary) Second Partial Derivative Test M dy N dx = (Mx + Ny )dxdy 1. Find all (x,y) points such that let f (x, y, z) be a scalar field and H R RR ~ (x, y, z) = M dx + N dy = (Nx My )dxdy rf (x, y) = ~0 F Let: R 2 2. Let D = fxx (x, y)fyy (x, y) fxy (x, y) M (x, y, z)î + N (x, y, z)ĵ + P (x, y, z)k̂ be ·R be a region in xy-plane IF (a) D > 0 AND fxx < 0, f(x,y) is a vector field, ·C is simple, closed curve enclosing R local max value Grandient of f = rf =< @f @f @f @x , @y , @z > (w/ paramerization ~ r (t)) (b) D > 0 AND fxx (x, y) > 0 f(x,y) is ~: ~ (x, y) = M (x, y)î + N (x, y)ĵ be ·F Divergence of F local min value ~ = @M + @N + @P continuously di↵erentiable over R[C. (c) D < 0, (x,y,f(x,y)) is a saddle point r·F @x @y @z ~: Form 1: Flux Across Boundary (d) D = 0, test is inconclusive Curl of F ~ n 3. Determine if any boundary point î ĵ k̂ H = unit normal R R vector to C ~ ·~ F n= r·F~ dA gives min or max. Typically, we have to ~ = r⇥F @ @ @ c H R R R parametrize boundary and then reduce @x @y @z , M dy N dx = R (Mx + Ny )dxdy M N P Form 2: Circulation Along to a Calc 1 type of min/max problem to Line Integrals Boundary solve. H R R ~ · d~ F r= r⇥F ~ · ûdA The following only apply only if a C given by x = x(t), y = y(t), t 2 [a, b] c H R R R R R , M dx + N dy = (Nx My )dxdy boundary is given f (x, y)ds = ab f (x(t), y(t))ds R 1. check the corner points c q AreaH of R dy 2 1 1 2. Check each line (0 x 5 would where ds = ( dx 2 dt ) + ( dt ) dt A = ( 2 ydx + 2 xdy) q give x=0 and x=5 ) dy 2 or 1 + ( dx ) dx On Bounded Equations, this is the q Gauss’ Divergence Thm global min and max...second derivative or 1 + ( dx 2 dy ) dy test is not needed. (3D Analog of Green’s Theorem - Use To evaluate a Line Integral, · get a paramaterized version of the line for Flux over a 3D surface) Let: Lagrange Multipliers (usually in terms of t, though in ~ (x, y, z) be vector field continuously ·F Given a function f(x,y) with a constraint exclusive terms of x or y is ok) di↵erentiable in solid S g(x,y), solve the following system of · evaluate for the derivatives needed ·S is a 3D solid ·@S boundary of S (A Originally Written By Daniel Kenner for equations to find the max and min (usually dy, dx, and/or dt) Surface) MATH 2210 at the University of Utah. points on the constraint (NOTE: may · plug in to original equation to get in ·n̂unit outer normal to @S Source code available at need to also find internal points.): terms of the independant variable Then, https://github.com/keytotime/Calc3 CheatSheet R R R R R rf = rg · solve integral F~ (x, y, z) · n̂dS = ~ dV r·F Thanks to Kelly Macarthur for Teaching and @S S g(x, y) = 0(orkif given) (dV = dxdydz) Providing Notes.