Can somebody explain “Gimbal Lock” in such a way that an idiot like me can understand it? Talk to me like i’m 5.
The Apollo missions, and the Mercury and the Gemini flights were interesting in that the hardware as well as the software was “custom” tailored to one, and only one mission. The rope-hardware core memory was woven right into the circuit boards.
The gyros were final aligned at the pad a few seconds before launch, a moment frozen in time on a specific day and launch window, when all the relative positions of the earth and stars had been determined. They knew where they were at that point. That was the “reference point” or stable platform? And they had done the calculations to determine where the moon will be 3 or 4 days from now.
I thought “Gimbal Lock” was when two of the spinning gyros reached the same plane, and couldn’t “get back out” or something like that? It nulls the reading in two axis or something??
Gimbal lock occurs when two gimbals (the frames that allow a gyroscope or IMU to rotate freely) are aligned, and a sensor can’t distinguish between rotation in roll and yaw directions. They aren’t physically restrained in any way but it becomes impossible to distinguish between the two motions and back out to a previous state, or to apply a unique rotation in one specific axis. Mathematically, it is a type of singularity which can be avoided by using quaternion algebra (basically complex numbers in four dimensions) rather than Euler angles, and physically by adding another gimbal or a floating spherical bearing. However, all modern systems just use digital sensors like a laser ring gyro where rotation about each axis is measured independently and there is no possibility of a ‘lock’ or confusion.
The Wikipedia page linked above has some good animations but I think the text could use some work.
What if the orientation or attitude of the spacecraft just keeps going past the null point and back out the other side? The gyroscopes are not restrained, but why can’t the math or navigation pick up from there? Or the null point is zero spatial data for purposes of navigation for however many seconds they are in a state if gimbal lock? It gets “lost” somehow? So two of the gimbals are parallel with each other, how do it know and why does it matter. I didn’t even think i understood it, now I really really don’t understand it much better.
You need three independent axes in your gyro to measure three possible dimensions of rotation. If one goes away because of gimbal lock, there is no way for the system to mechanically step its way out of that state, because which of the two (e.g.) y-axes is the “real” one?
To fix this situation, you might physically move the gyro axes to positions that are all 90 degrees from each other. But then, your gyro would not be calibrated, the axes would not line up with your desired frame of reference.
By adding a 4th gimbal, it prevents the other three from ever getting into this situation in the first place. IIRC, Apollo did not use a 4th gimbal because it made the instrument a lot larger and heavier.