1. Functions of 2 Variables: This animation merely graphs a function of 2 variables and then rotates it around the origin so that it can be seen from all sides. One comment: if the function is symmetric via a 180 degree rotation around the origin, k=0...2*Pi*(n-1)/n on line 6 can be changed to k=0..Pi*(n-1)/n, that is, the 2* can be removed, either speeding up the generation process or making the animation smoother.
    2. restart: with(plots):

rm := 0.01: rn := 2:                    # r domain (cyl. coords.)
p := -2: q := 2: n := 10:               # z range, frames
f := (x,y) -> (x^2 - y^2)/(x^2 + y^2):  # function

animate3d([r*cos(t), r*sin(t), f(r*cos(t+k),r*sin(t+k))],
         r=rm..rn, t=0..2*Pi, k=0..2*Pi*(n-1)/n, view=p..q,
         frames=n, grid=[12,36]);
  2. Traces: This animation demonstrates the idea of the trace of a function of two variables. The point is to observe the curves of intersection of the horizontal plane with the surface. Note that with a little fiddling you can change the horizontal traces to vertical traces by switching the x,y and k around in the "an2 :=" line. For example, [k,y,x] instead of [x,y,k] would graph vertical planes parallel to the yz-plane. However, because k has twice the range of x, you may want to use [k/2,y,2x] instead to get a better appearance.
    3. restart: with(plots):

a := -2: b := 2:          # x domain
c := -2: d := 2:          # y domain
p := -4: q := 4: n := 15: # z range
f := (x,y) -> x^2 - y^2:  # function

an1 := plot3d(f(x,y),x=a..b,y=c..d,view=p..q):
an2 := animate3d([x,y,k],x=a..b,y=c..d,k=p..q,
       frames=n,grid=[2,2],color=white):
display(an1,an2);
  3. Contour Lines: This animation takes the idea of the trace one step further and shows the creation of contour lines for a function.
    4. restart:with(plots):

a := -1:  b := 1:         # x domain
c := -1:  d := 1:         # y domain
p := -1:  q := 1: n := 8: # z range & frames
f := (x,y) -> x^2 - y^2 : # function

an1 := plot3d(f(x,y),x=a..b,y=c..d,grid=[10,10],view=p..q):
for k from 0 to n do
  an2[k] := plot3d(q-k/n*(q-p),x=a..b,y=c..d,grid=[2,2],
            color=white):
  an3[k] := implicitplot3d(f(x-b+a,y)=q-(q-p)*k/n,
            x=b..2*b-a,y=c..d,z=p..p+.001,numpoints=100):
end do:
an4[0] := an3[0]:
for k from 1 to n do
  an4[k] := display(an4[k-1],an3[k]):
end do:
an5 := plots[display]([seq(an2[i],i=0..n)],insequence=true):
an6 := plots[display]([seq(an4[i],i=0..n)],insequence=true):
display(an1,an5,an6,scaling=constrained);
  4. Partial Derivatives: This animation shows that a function of 2 variables can be considered as a function of one variable if the other is held constant. Then we can take a derivative just as we could for single variable functions. We note that the last line shows an example of a partial derivative with respect to x; it can be changed to read "display(an1,an3,an5);" in order to demonstrate a partial derivative with respect to y; of course, the "#" symbols must be eliminated also. 
    5. restart: with(plots):

a := -1: b := 1:          # x domain
c := -1: d := 1:          # y domain
p := -1.4: q := 1.4:      # z range
r := -.5: n := 16:        # y position, frames
f := (x,y)->-x^2+y^2-x*y: # function

fx := (x,y) -> diff(f(x,y),x):
fy := (x,y) -> diff(f(x,y),y):
 an1 := plot3d(f(x,y),x=a..b,y=c..d,view=p..q,grid=[13,13]):
 an2 := plot3d([t,r,s],t=a..b,s=p..q,color=white,grid=[2,2]):
#an3 := plot3d([r,t,s],t=c..d,s=p..q,color=white,grid=[2,2]):
 an4 := animate3d([t,r,fx(k,r)*(t-k)+f(k,r)],t=a..b,s=p..q,
        k=c..d,frames=n):
#an5 := animate3d([r,t,fy(r,k)*(t-k)+f(r,k)],t=c..d,s=p..q,
#       k=a..b,frames=n):
display(an1,an2,an4);
  5. Tangent Planes: This animation shows the tangent planes at several points on a surface. 
    6. restart: with(plots):

f := (x,y) -> y^2*x-x^2: # the function
n := 16:                 # frames

fx := (x,y) -> subs(q=y,subs(p=x,diff(f(p,q),p))):
fy := (x,y) -> subs(q=y,subs(p=x,diff(f(p,q),q))):
an1 := plot3d([r*cos(t),r*sin(t),f(r*cos(t),r*sin(t))],
       t=0..2*Pi,r=0..2,grid=[36,12]):
an2 := animate3d([ t*sin(k)+s*cos(k),-t*cos(k)+s*sin(k),
       f(cos(k),sin(k))+.07
       +fx(cos(k),sin(k))*(t*sin(k)+(s-1)*cos(k))
       +fy(cos(k),sin(k))*(-t*cos(k)+(s-1)*sin(k))],
       t=-1..1,s=0..2,k=2*Pi/n..2*Pi,grid=[2,2],
       frames=n,color=white):
display(an1,an2);
  6. Directional Derivatives: Here we graph a function and half planes rotating around the vertical line x=1, y=0. In those half planes we show the directional derivative vectors of the function. If you imagine all of them together, you will get an ellipse lying in the tangent plane. 
    7. restart: with(plots):

f := (x,y) -> 4-x^2-y^2: # function
p := 0: q := 5:          # the z-range
n := 16:                 # frames

fx:= subs(b=0,subs(a=1,diff(f(a,b),a))):
fy:= subs(b=0,subs(a=1,diff(f(a,b),b))):
s := (sqrt(3*tan(k)^2+4)+(k-Pi/2)/abs(k-Pi/2))*
     cos(k)*(Pi/2-k)/abs(k-Pi/2):
an1 := plot3d([r*cos(t),r*sin(t),f(r*cos(t),r*sin(t))],
       r=0..2,t=0..2*Pi,grid=[13,25]):
an2 := animate3d([1+cos(k)*s*t,sin(k)*s*t,z],
       t=0..1,z=p..q,k=(4-n)*Pi/(2*n)..3*Pi/2,
       color=white,frames=n,grid=[2,2]):
an3 := animate3d([1+cos(k)*t,sin(k)*t,
       f(1,0)+fx*cos(k)*t+fy*sin(k)*t],
       t=0..1,z=0..1,k=(4-n)/(2*n)*Pi..3*Pi/2,frames=n):
display(an1,an2,an3);
  7. Vector Derivatives: This animation demonstrates the idea of the derivative of a 3-dimensional parametrized curve. 
    8. restart: with(plots):

p := 1: q := 3: n := 12:            # z range, frames
x := t -> (2 + cos(p*t))*cos(q*t):  #
y := t -> (2 + cos(p*t))*sin(q*t):  # vector function <x,y,z>
z := t -> sin(p*t):                 #

an1 := tubeplot([x(t),y(t),z(t)], t=0..2*Pi, radius=.1, numpoints=48):
dx := s -> D(x)(s):                 #
dy := s -> D(y)(s):                 # vector derivative
dz := s -> D(z)(s):                 #
r := s -> simplify(sqrt(dx(s)^2+dy(s)^2+dz(s)^2)):
vx := s -> dx(s)/r(s):              #
vy := s -> dy(s)/r(s):              # unit vector deriv.
vz := s -> dz(s)/r(s):              #
for i from 0 to n-1 do
  k := 2*Pi*i/n:
  an2[i] := tubeplot([x(k)+t*vx(k),y(k)+t*vy(k),z(k)+t*vz(k)],
            t=0..2, radius=.05, numpoints=2):
end do:
an3 := plots[display]([seq(an2[i],i=0..n-1)], insequence=true):
display(an1,an3);
  8. Spherical Coordinates, phi: This animation demonstrates the phi part of the spherical coordinates, which is the part that students seem to have the hardest time visualizing.
    9. restart: with(plots):

n:=15:                # frames

for k from 1 to n do
  an1[k] := sphereplot(1,theta=0..2*Pi,
            phi=0..Pi*k/n,grid=[16,k+1]):
  an2[k] := plot3d([s*sin(Pi*k/n)*cos(t),
            s*sin(Pi*k/n)*sin(t),s*cos(Pi*k/n)],
            t=0..2*Pi,s=0..1,grid=[16,2]):
end do:
an3 := plots[display]([seq(an1[k],k=1..n)],
       insequence=true):
an4 := plots[display]([seq(an2[k],k=1..n)],
       insequence=true):
display(an3,an4,scaling=constrained);
  9. Spherical Coordinates, theta: This animation demonstrates the theta part of the spherical coordinates.
    10. restart: with(plots):

n:=15:                # frames

for k from 1 to n do
  an1[k] := sphereplot(1,theta=0..2*Pi*k/n,
            phi=0..Pi,grid=[k+1,16]):
  an2[k] := plot3d([s*cos(2*Pi*k/n)*sin(p),
            s*sin(2*Pi*k/n)*sin(p),s*cos(p)],
            p=0..Pi,s=0..1,grid=[16,4]):
end do:
an3 := plots[display]([seq(an1[k],k=1..n)],
       insequence=true):
an4 := plots[display]([seq(an2[k],k=1..n)],
       insequence=true):
an5 := plot3d([s*sin(p),0,s*cos(p)],s=0..1,
       p=0..Pi,grid=[4,16]):
display(an5,an4,an3,scaling=constrained);