Inertial Navigation Systems — Page 235, Lesson 275

Let's pick up with the accelerometer and why it can't just be bolted to the airframe. Normally, the accelerometer lives on the gyro-stabilized platform, and there's a good reason for that. If you hard-mounted it directly to the aircraft, it would suffer problems in the pitch and roll planes. Here's the mechanism. The accelerometer is essentially a pendulum. If the aircraft tilts, the device tilts with it, and gravity pulls that pendulum away from its null position — the centered, zero-output position. That tilt makes the accelerometer output an erroneous acceleration signal. And that false acceleration integrates into an erroneous velocity, which then integrates into an erroneous distance travelled. So a simple pitch or roll angle creates a false acceleration problem. The whole point of keeping the accelerometer earth-horizontal is to prevent that error from ever appearing. Now, the gyroscope used in an inertial navigation system is a special animal. It's an integrating gyroscope — a one degree of freedom gyro. The key contrast is in the restraint. A rate gyroscope uses mechanical restraint, a spring. The integrating gyro uses viscous restraint — fluid drag instead of a spring. That's the defining difference. Picture the construction. It's basically a can within a can. The outer can, the frame, is filled with a viscous fluid. Inside it, the inner gimbal is pivoted about its vertical axis. That fluid supports the weight of the inner gimbal, which reduces bearing torques — less friction fighting the motion. That's the rate-integrating gyro. Now, how do we keep the accelerometer level? We mount it on a gimbal assembly called the platform. The platform is a mechanical device that lets the aircraft go through any attitude change while the innermost element — where the accelerometers sit — stays earth-level. The gyroscopes that stabilize the platform are also mounted on that innermost element. They feed signals to amplifiers and motors, and those motors drive the gimbals to keep the accelerometers level. Let me walk you through the control loop, because it's a closed loop. The gyro and the accelerometer share a common gimbal. If that gimbal tips off level, the gyro's spin axis stays fixed in space — that's the gyroscopic property. So the gyro's case moves off level relative to the spin axis. The amount the case is tipped is detected by the signal pick-off inside the gyro. That signal gets amplified and sent to a gimbal drive motor, which restores the gimbal to level. Since the accelerometer is always kept level, it never senses a component of gravity — it only senses the horizontal accelerations of the aircraft as it travels across the earth's surface. In reality, the platform carries three integrating gyros, with their input axes mutually perpendicular — at right angles to each other. Three gimbal motors drive the platform gimbal rings about the pitch, roll, and vertical axes respectively. The gyros sense incipient displacement — the very beginning of a tilt — and activate the appropriate motors to move the gimbal rings, keeping the platform stable as the aircraft maneuvers around it. So the chain is: gyro senses tilt → signal pick-off → amplifier → gimbal motor → gimbal restored → accelerometer stays level → only true horizontal acceleration is measured. That's the whole architecture of the stabilized platform.