Helieng asks how a bifilar works.
Here are some concepts on vibration absorbers in general, then on the bifilar:
If you put a weight on top of a coil spring, the spring will bounce the weight at a specific frequency, based on the amount of mass and the stiffness of the spring. Think of the hula doll sitting on the car dashboard, with the head on a spring and eyes that light up when the brakes are pressed.
A stiffer spring makes a higher frequency, as does a smaller mass. By varying either or both, we can make the spring-mass system "like" (resonate at) any given frequency.
If we place this spring-mass system in a helicopter, and tune it to "like" the exact n per rev frequency of the helicopter, it becomes a vibration absorber. That means that it picks up the vibration energy from its mountings and easily transmits that energy to the mass. The mount and deck become quiet, and the mass bounces like crazy. We can orient the mass for vertical or horizontal vibrations, although it tends to absorb both to some degree.
The bifilar is a spring-mass vibration absorber, but the "spring" is the centripital acceleration field that the bifilar arm swings in. This is easy to picture, since the mass pivots on a roller so that it must swing closer to the hub when it swings off center. It behaves like a pendulum in the centripital field.
The great strength of a bifilar is that it automatically tunes to the frequency of the rotor, since its "spring" is based on the rotational speed. This is better than the spring-mass system we first described, where we must change spring or mass to retune. Since the rotor rpm changes slightly as we maneuver, a bifilar stays in tune and stays efficient across a wide band of rpms, where a spring mass absorber falls off its peak and can be quite ineffective. A bifilar equipped helo is smoother in autorotation and power changes than a fixed absorber equipped helicopter.
Bifilars were first used on engine cranks to smooth out the vibration of the engine.
Regarding the orientation of the absorber, the rotor head bifilar is quite effective even though it is aligned in the lateral axis. This is because the rotor root shears I mentioned in previous posts act both in-plane and vertically, so there is enough vibration to go around. In fact, even a vertical vibration at the head is often transmitted as a roll mode, since the airframe resists vertical motions by its whole mass, while its rolling motion has much less inertia. <img src="cool.gif" border="0">