Scribd
Upload
13 views

Differential Forms: X D Dy X y

Differential forms are mathematical objects that generalize the concept of directional derivatives and can be used to describe multivariable vector calculus concepts like divergence and curl. Some key points: - Differential 1-forms like dx, dy, dz assign vectors to points in a way that describes vector fields. - Higher order differential forms can be constructed by taking exterior products of 1-forms, generalizing the cross product. - The exterior derivative d maps forms to forms one degree higher and can be used to compute derivatives like via the chain rule without diagrams. - Important properties like d^2=0 allow differential forms to prove theorems in vector calculus.

Uploaded by

Zaratustra Nietzche
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
13 views

Differential Forms: X D Dy X y

Differential forms are mathematical objects that generalize the concept of directional derivatives and can be used to describe multivariable vector calculus concepts like divergence and curl. Some key points: - Differential 1-forms like dx, dy, dz assign vectors to points in a way that describes vector fields. - Higher order differential forms can be constructed by taking exterior products of 1-forms, generalizing the cross product. - The exterior derivative d maps forms to forms one degree higher and can be used to compute derivatives like via the chain rule without diagrams. - Important properties like d^2=0 allow differential forms to prove theorems in vector calculus.

Uploaded by

Zaratustra Nietzche
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Differential Forms

Note: The following contains some advanced topics, not necessarily part of Calc 3. However, Some
of the things we will later learn (Jacobian, Divergence, Curl, Chain Rule) will either become easier,
or you can gain a deeper understanding with forms. I will be referencing these occasionally in class,
so it is helpful to know this, but not required. When I mention them in class, I will only use small
bits and pieces, and I will give a quick refresher on what is needed. However, it wont make too
much sense without first glancing through this.
The objects dx, dy, dz, df, called differential forms, are not just notation; they do have important
meaning in math, but to really know what they are, takes a lot of sophistication. The easiest way
to think of these is through vector fields, which we will learn about later on in the course. In short,
a vector field assigns a vector to every point. So, for example, the vector field hx, yi would have
the zero vector at the origin, and radially, outward-pointing vectors at every other point, growing
longer as you move further away from the origin.
Every (2D) vector field can be written as a combination of the constant vector fields, h1, 0i , h0, 1i,
d

, dy
. That is. every vector field can be written as hf (x, y), g(x, y)i =
denoted more formally as x

f x + g y . The differential forms are defined by





+g
=f
dx f
x
y




dy f
+g
= g.
x
y

In higher definitions, we have the expected definition


!


n
n
X
X

dxk
fi
=
fi dxk
= fk .
x
x
i
i
i=1
i=1
So, in short, differential forms take in vector fields and spit out functions. So, in some sense, forms
are the opposite, or cancel out, vector fields.
Now, we can define the product of 1-forms dx and dy, called a 2-form, by
dx dy (u, v) = dx(u)dy(v) dx(v)dy(u).
From this, we get


dx dy f1
+ f2 , g1
+ g2
x
y
x
y



f1 f2


= f1 g2 f2 g1 =
g1 g2

The operation is a lot like the cross product. Namely, it is anticommutative, and the distributative law holds. Also, it is associative. In general, for higher order products, it is easiest to use
the fact that forms split over sums and differences, and just use the fact that



dx1 dx2 dxn f1


, f2
, . . . , fn
= f1 f2 fn .
x1
x2
xn
Again, as a reminder, since is anticommutative (dx dy = dy dx), we have that dx dx = 0.

Forms In 3D
In 3 dimensions, with variables x, y, z, every form can be expressed like the following:
f
f dx + g dy + h dz
f dx dy + g dy dz + h dz dx
f dx dy dz

(0-form)
(1-form)
(2-form)
(3-form)

There are no higher forms since we only have three dimensions, so a 4-form would have to repeat
a variable, making it 0.
Now, we define the derivative, d, which takes in forms, and spits out forms of 1 degree higher. Let
f be a function and be a form just involving dx, dy, dz, and (That is, like above, but where
f = g = h = 1 or 0). Then, we define
f
f
f
dx +
dy +
dz
x
y
z
d(f ) = df .
d(f ) = df =

Again, the definition for higher dimensions works exactly how you would expect.
Now, the explicit computation of the derivative of a 1-form:
d(f dx + g dy + h dz)
= (fx dx + fy dy + fz dz) dx + (gx dx + gy dy + gz dz) dy + (hx dx + hy dy + hz dz) dz
= 0 + fy dy dx + fz dz dx + 0 + gx dx dy + gz dz dy + 0 + hx dx dz + gy dy dz
= fy dx dy + fz dz dx + gx dx dy gz dy dz hx dz dx + hy dy dz
= (gx fy ) dx dy + (hy gz ) dy dz + (fz hx )dz dx.
Note: The above result looks like something you would get by taking the cross product. This fact
will be important later.
One can show that, in general, if you are working in any dimension and is any form, then
d2 = d(d) = 0.
This is a VERY important result, although showing it is a bit messy. It is mainly because of this
result that we will get all of our theorems in the last third of the course.

Applications
Hodge Star and the Cross Product
Define the Hodge star operator, ?, in 3-dimensions, which sends a k-form to a (3 k)-form, by
?(f ) = f dx dy dz
?(f dx + g dy + h dz) = h dx dy + f dy dz + g dz dx
?(f dx dy + g dy dz + h dz dx) = g dx + h dy + f dz
?(f dx dy dz) = f.
Note: ?2 () = ?(?()) = .
Now, if we identify a vector ha, b, ci with the form a dx + b dy + c dz, we can define
~u ~v = ?(~u ~v ).

This is in fact the cross product that we already know!


The Chain Rule
Suppose you have a function f and compute its differential, df. If one of the terms is g dx, and all of
f
f
the other differentials are independent, then we have that
= g, by where
I mean to be the
x
x
change of f with respect to the variable x. As a consequence of this, consider a function f (x, y, t)
with x(s, t) and y(t). Then, we would have
f
f
f
dx +
dy +
dt
x
y
t


x
f dy
f
f x
ds +
dt +
dt +
dt
=
x s
t
y dt
t


f x
f x f dy f
=
ds +
+
+
dt
x s
x t
y dt
t

df =

f x f dy f
+
+ . Using other
x t y dt t
F
notation, if we consider F (s, t) = f (x(s, t), y(t), t), then this is just
.
t
The main note here is that you can compute derivatives using form, that would normally require
the chain rule, but now do not need any diagrams to compute.
So, the overall change of f with respect to the variable t would be

You might also like