Lectures on Chapter 22: Magnetic Forces and Magnetic Fields

Magnetism

Everyone has played with permanent magnets. You probably know there is a north pole and a south pole on a bar magnet. We can draw magnetic field lines just as we drew electric field lines. The rules: (1) the direction of the tangent to a magnetic field line at any point gives the direction of at that point, and (2) the spacing of the lines is a measure of the magnitude. Lines emerge from N and enter at S.

=1.0 true in dipoleBfield.eps

If you have 2 bar magnets, opposite ends attract, i.e., N attracts S. Likes repel, e.g. N repels N. The earth's north pole is a geomagnetic S because field lines enter and point into the earth at the earth's north pole. Antarctica is a magnetic north pole because field lines point up and out of the earth's surface. One way to test to see which way the lines point is to use a compass.

Force Law

What happens if we put a charge +q in a magnetic field ? If it's sitting still, the answer is nothing. But if it's moving with velocity , it feels a force perpendicular to and to .

q can be positive or negative. changes direction if . So

=4.0 true in LorentzForce.eps

There are several things to notice about :

1.

If doesn't change, only direction of changes in a uniform field. (Acceleration comes from change of direction .) So no change in the kinetic energy .

2.

If , . Particle moving parallel to feels no force.

= angle between and

. This applies if

=1.5 true in angleBv.eps

3.

Maximum force of deflection occurs for

4.

, Bigger Bigger F_B
Bigger Bigger F_B

5.

Magnetic forces do no work because the force is perpendicular to the direction that the particle moves:

W_B = =

Units

Since
Another common unit of B is the gauss (G):

The earth's magnetic field is T.

Going in Circles

Suppose lies in a plane . Here we are considering uniform . A charge of +q moves with velocity . implies that the charge goes in a circle. (Think of whirling a rock around with a string. The force of the string points radially inward.)

=1.5 true in cyclotron1.eps

=1.5 true in cyclotron2.eps

Since ,

(1)

Set this equal to the centripetal force:

 

(2)

r is the radius of orbit. This equation is called the cyclotron formula. Larger smaller orbit. Larger bigger orbit.

Period =T= circumference/speed =
Notice that the period T has no dependence on the velocity v.

Frequency

Angular Frequency

Hall Effect

Consider a metal bar with current flowing in it carried by electrons with an average drift velocity

=2.5 true in hallEffect1.eps

Now suppose we apply a magnetic field in the direction. This initially causes a downward deflection of the moving electrons.

=2.5 true in hallEffect2.eps

Negative charge builds up at the bottom; positive charge at the top.

=2.5 true in hallEffect3.eps

The transverse electric field counters the magnetic force so that the electrons again flow in the direction. Notice that if the charge carriers had been positively charged, would point in the opposite direction ( in same direction as before). Thus if we measure the voltage difference between top and bottom, the sign tells us the sign of the charge of the carriers.

It is easy to determine the magnitude of by balancing the electric force with the magnetic force:

(3)

We know J = nqv_d implies n= J/qv_d =JB/qE_t

=2.5 true in hallEffect4.eps

I = JA

Thus we can determine the carrier density n from quantities we can measure.

Force on a Current Carrying Wire

Consider a wire carrying current I. If we put the wire in an field, no force acts on it because the wire is neutral. On the other hand, if we put it in a field, there will be a sideways force on the current and hence on the wire.

=4.0 true in ForceonWire.eps

Remember that I points in the direction of positive charge flow.

Let's be more quantitative about this. Suppose we think of the current as consisting of a charge per unit length moving at velocity v down the wire. (Assume that there is a charge density that isn't moving so that the wire as a whole is electrically neutral.)

=1.5 true in wire1.eps

In the time , a length of charge flows past a point. Thus the amount of charge that flows past the point is . The current I is:

Consider a straight section of wire of length carrying current I. The amount of moving charge in this section is . Thus,

F = =

In general, we have

where is a length vector which points along the wire. length of straight wire segment. If the wire isn't straight, we can divide it up into little straight segments each of length ds. The force on each tiny segment is then

There is no such thing as an isolated current carrying wire segment. Current coming in one end must go out the other end.

Torque on a Current Loop

Consider a rectangular loop of wire carrying a current in a magnetic field . There will be forces on the loop, but the net force will cancel out .

=3.25 true in torqueLoop1.eps

= (4) =

Now suppose we tilt the loop by rotating it about the x-axis.

=2.0 true in torqueLoop2.eps

and want to rotate the loop so that points along . ( is a unit vector normal (or perpendicular) to the plane of the loop.) and produce a torque.

and are measured from the midpoint.

=1.5 true in torqueLoop3.eps

If the loop were replaced by a coil with N turns the torque would be

Getting a current loop to turn in a magnetic field is the principle behind electric motors. The loop keeps turning because the direction of the current is switched every half cycle. An electric motor is what starts your car.

Magnetic Dipole

We can associate a magnetic dipole moment with the current loop. The direction of is normal to the loop ( given by the right hand rule with the fingers in the direction of the current. The magnitude of for a flat loop with N turns is

Then our previous expression for becomes

The general vector relation is

(5)

The magnetic potential energy is

(6)

The coil aligns itself in the field so as to minimize its potential energy.

=4.0 true in uB.eps

Currents produce magnetic fields

How are magnetic fields produced? One source is permanent magnets (refrigerator magnets). Another turns out to be electric currents. The connection between electricity and magnetism was made in 1819 by Hans Christian Oersted. One day, while teaching class, he brought a compass near a wire with a current flowing through it. Since the needle was perpendicular to the wire, nothing happened.

=2.0 true in BfromWire1.eps

Later, after class, Oersted decided to try it again, but this time with the wire oriented parallel to the compass needle.

=4.0 true in BfromWire2.eps

The needle turned perpendicular to the wire. When he reversed the current, the compass needle reversed its direction. Oersted had shown that electrical currents produce magnetic fields.

=2.0 true in BfromWire3.eps

The magnetic field near a wire is much larger than the earth's magnetic field. So we can use a compass to map out the field of a current carrying wire. Denote the magnetic field by . circles around the wire. Right hand rule: if thumb points in the direction of current flow, your fingers curl in the direction of the field.

=2.0 true in BfromWire4.eps

Biot-Savart Law

Recall that to find the produced by a continuous charge distribution q, we divide the charge into little pieces dq. Each piece produces of magnitude

(7)

Similarly, to find the field produced by a current carrying wire, we divide the wire into tiny segments . length of segment. points in the direction of current flow. Note that is a vector while dqis a scalar.

Suppose a point P is a distance r from . Then the magnitude of the magnetic field set up at point P by the current element I ds is

=2.0 true in BiotSavart.eps

is the permeability constant (or permeability of free space):

is the magnetic analog of . is the angle between and . The direction of is given by . points from to P. Putting this together gives the Biot-Savart Law:

Let be a unit vector in the direction. Then

Recipe for Using the Biot-Savart Law

1.

Divide current into tiny pieces where points in the direction of current flow.

2.

Find direction of the field due to at a distance away:

3.

Calculate the magnitude

4.

Integrate over the entire current flow:

You may have to do this by components:

B_x =

B_y =

B_z =

Notice that falls off as just like . Notice also that and .

Due to Long Straight Wire

The simplest application of Biot-Savart is to find the field produced by a long straight wire. Notice that our earlier claim of curling around the wire agrees with what we got from .

=4.0 true in BiotSavWire.eps

To find the magnitude of , we use

(8)

Notice that has the same direction (out of the paper) for every current element into which the wire can be divided. Thus we can add the magnitudes

Note that and

dB = =

B = = =

Aside: We had to evaluate the integral . This is how we did this:

Let                          because

I =

I = = = =

=1.5 true in triangle.eps

So

So a long straight wire produces a field

where is the perpendicular distance from the wire. Notice that B gets weaker as you go farther from the wire. Also B increases linearly with the current strength I.

on a Current Carrying Wire

If we put a long straight current carrying wire in an external magnetic field the total magnetic field is given by

where is due to the current in the wire. However, the force on the wire is due only to because the wire does not exert force on itself. Thus

To see why

note that on one side of the wire is opposite to that on the other side (e.g. ). So the forces cancel out.

=1.8 true in SelforceonWire.eps

However if the wire was bent, one part of the wire could exert force on the other part.

=1.0 true in BentWire.eps

Force Between 2 Wires

Rather than considering a bent wire (with complications like bends and corners), let's consider 2 parallel wires with currents in them. Then the field produced by wire will exert a force on wire :

(9)

where segment of length pointing in direction of I_b. Similarly wire b exerts a force on wire a:

(10)

Let's see which way these forces point:

=6.0 true in Forcebetwires.eps

Let's calculate in case 1:

=2.0 true in Forcebetwires2.eps

Notice that the force depends on both I_a and I_b. Similarly,

(11)

The force per unit length is

(12)

Current Loop

We can use the Biot-Savart law to find the magnetic field a distance z above the center of a current loop of radius R.

=2.0 true in currentloop.eps

We divide the ring into segments . . So

dB = =

We want the z-component of dB. (By symmetry,

dB_z = = = = =

If the loop is replaced by a circular coil with N turns of wire, then

Far above the loop (z >> R)

Let's relate this to the magnetic dipole moment of the loop. Recall , where A = area of the loop. The direction of is given by the right hand rule: when your fingers curl in the direction of the current, your thumb points in the direction of . Notice that points in the same direction as .

where . This is the magnetic field produced by a dipole moment along the z-axis. The field lines are like those of a bar magnet with Nand S poles.

=1.0 true in dipolefield.eps

Note that our expression for is similar to the electric field produced by an electric dipole :

Both and fall off as 1/z³

Ampere's Law

Recall that in electrostatics, Gauss' law made it easier to find if we were given a charge distribution with symmetry.

(13)

Ampere's law is the magnetic analog of Gauss' law.

(14)

This is Ampere's Law. This tells us that if we encircle a current carrying wire with a closed loop C, the line integral around C will equal . ( does not go along the wire.) The direction of C (clockwise vs. counterclockwise) is given by the same right hand rule that gives the direction of : thumb along I, fingers curl in direction of C. Ccan be any shape as long as it doesn't cross itself and it is closed. It is easiest to do the line integral if or if . Thus Ampere's law allows us to find if the current configurations have symmetry. The following current configurations are commonly evaluated using Ampere's law:

1.

Infinitely straight lines

2.

Infinite Planes

3.

Infinite Solenoids

4.

Toroids (Donuts)

Recipe for Using Ampere's Law

1.

Determine the direction of produced by the current by symmetry. If the current distribution does not have symmetry, don't use Ampere's law.

2.

Draw Amperian loop surrounding the current. Try to make sides of loop such that or on each side.

3.

Evaluate line integral .

4.

Evaluate how much current is enclosed by the Amperian loop: .

5.

Solve for the magnitude of . (Direction of was determined in step 1.)

Notice that if , then . But this doesn't mean that B=0. In the figure on sides 2 and 4, .

=2.0 true in amperianSq.eps

The line integral for 1 and 3 cancel:

Example: Long Straight Wire Find B a distance r from a long straight wire carrying current I. Draw a loop C of radius r around the wire in the direction of . circles around the wire. B=const on the loop C because the radius of C is fixed at r.

=2.0 true in amperianCircle.eps

B =

This is the same result that we got using the Biot-Savart law. Notice that this was much easier than our previous calculation.

Suppose the long straight wire has radius R and carries total current I. Find both outside (r > R) and inside (r < R) the wire. Assume the current is spread uniformly over the cross section of the wire.

=2.0 true in amperianWire.eps

For r > R, the calculation is the same as above and

For r < R (inside), we draw a loop C of radius r. Then

= (current density)(area inside loop C) . So which implies

(15)

B = =

Notice that B increases linearly with r inside the wire.

=2.0 true in BvsrWire.eps

points azimuthally both inside and outside the wire.

Solenoid

A solenoid is a cylindrical coil of wire. We are interested in long straight solenoids whose length L is much longer than the radius (L >> R). An infinite solenoid with tightly packed turns of wire has a uniform field inside the coil along the axis and B=0 outside the coil. You can see this by looking at a diagram.

=2.0 true in solenoid1.eps

The field between neighboring wires tends to cancel. Outside the solenoid, the field from the top tends to cancel that from the bottom . Inside the top and bottom add

=2.0 true in solenoid2.eps

We can systematically deduce the direction of . Let's use cylindrical coordinates .

: Is there a radial component B_r? No. Suppose there were an outward pointing radial component (B_r > 0). Then if we reverse the current, B_r would point inward. But reversing the current is equivalent to turning the solenoid upside down which shouldn't change B_r. So B_r = 0. Another way to see this is to note that B_r from adjacent turns of wire cancel.

=1.0 true in solenoid3.eps

Or consider z=0, say, and note that the current from produce contributions to B_r(z=0) which tend to cancel.

=1.0 true in solenoid4.eps

: What about ? Take an Amperian loop #1 around the cylinder. By symmetry, on this loop. The loop has a radius r.

=1.5 true in solenoid5.eps

because the no current pokes through the surface spanned by loop .

Only B_z is left. From the right hand rule, points along +zinside. (Same use of right hand rule that you used to get the dipole moment .) What is B_z outside? We can use Ampere's law to prove that B=0 outside. Draw an Amperian loop outside the solenoid. .

=2.0 true in solenoid6.eps

and because and . is constant along side 1; is constant along side 3. Let's assume and where B₁ > 0, B₃ > 0; i.e., and point in the direction. is antiparallel to . So since .

But infinitely far away from the solenoid, B=0 which implies that the constant = 0. So B₁ = B₃ = 0 outside the solenoid.

What is B_z inside the solenoid? Draw an Amperian loop as shown. If there are N turns per unit length, then .

=2.0 true in solenoid7.eps

Look at . on sides 2 and 4. along side 3. So

Magnetism in Matter

Electrons can generate magnetic fields in three ways.

1.

Currents in conductors.

2.

Orbital motion in an atom.

3.

Intrinsic ``spin" which produces an intrinsic magnetic moment .

Electrons orbiting around the nucleus of an atom are like a tiny current loop. Associated with this current loop is a magnetic dipole moment . The circular motion of the electrons is also associated with an orbital angular momentum . It turns out that .

=1.25 true in electronOrbit2.eps

Recall that the magnetic moment produced by a current loop is given by

(16)

=2.0 true in electronOrbit.eps

The current A and the area where Å. In most substances the orbital magnetic moment of one electron in an atom is cancelled by that of another electron orbiting in the opposite direction. So orbital motion produces very little, if any, magnetism in materials.

The other way electrons produce magnetism is via their intrinsic angular momentum or spin. In atoms or ions with many electrons, the electrons fill the orbitals in pairs with opposite alignments of their spins. As a result each pair has zero magnetic moment. But an atom with an odd number of electrons has at least one unpaired electron and a corresponding spin magnetic moment. A substance consisting of these magnetic atoms can exhibit different types of magnetic behavior.

=2.0 true in energy_levels.eps

Perhaps the most well known such behavior is ferromagnetic. Iron, cobalt, nickel, gadolinium, and dysprosium are ferromagnets. Ferromagnets are what you use to tack messages to your refrigerator door. In a ferromagnet the magnetic moments of the atoms tend to align parallel to one another.

=1.0 true in ferromagnet.eps

In ferromagnets there are microscopic regions called domains within which the moments are aligned parallel to one another. But a given domain can have its net magnetic moment pointing in a different direction from some other domain. All these misaligned domains yield a net magnetic moment of 0. This is why a piece of iron (like a nail) will usually not be magnetic.

=2.0 true in domains.eps

But if you put the iron in a strong magnetic field, the domains aligned along the field will grow at the expense of the domains which are not aligned along the field. As a result the iron will acquire a net magnetization and will become magnetized. Magnetization is the magnetic moment per unit volume.