Writing up some notes here, because I realize I've gone through this exercise several times and have gotten somewhat different answers.
Problem statement: For a pilot sitting in in a seat with a 5°-from-horizontal seat angle, 13°-from-vertical back angle, and relatively low seat height (so knees at maybe a 135° extension angle), what is the relative position of the median pilot's CG relative to the seat reference point (intersection of seat and back)? What is the standard deviation of this position in each axis, and what is the position for a 90th percentile weight (R=1.28, 260 lb) male?
Source 1: "Whole Body Center of Gravity and Moments of Inertia Study." The measured seated position is a vertical seat (0° seat angle, 0° back angle, 90° knee extension), arms at sides. Population is reasonable.
Source 2: "Biomechanical analysis of the dimensions of pilot seats in civil aircraft." Had this in my notes as relevant, but while it has some vaguely interesting info about pilot seat geometry in transport-category aircraft, nothing about pilot CG.
Source 3: "Determination of Centers of Gravity of Man." Fewer subjects (all male!), more positions. No subjects at our (current) 90th percentile. From table II, position (b) is the closest to the position used in Source 1, so let's look at that first to see how well the sources agree. Some risk of transcription error due to poor scan quality; ± is "across subjects", not std dev.
This is close enough, especially in x, to be comparable. Closer to the "both" numbers above, but not far off.
From this source, we see that both arms forward moves the CG forward about 2.2 cm, and up about the same; so this gives us a sense of how much CG varies for a single individual with normal movements.
Table IV goes into a slightly more interesting position for us -- a 110° knee extension angle. Standard hand-on-stick gives:
Plenty more positions after this, but only so much information is consumable.
Exercise for the reader: Are there any other sources worth looking at here?
Coordinate transform: It's not quite a trivial rotation to consider the more typical seat geometry, because the back and legs tilt by a different amount. However, it's reasonable to assume that the numbers will fall somewhere between those of a 5° rotation (seat angle) and 13° rotation (back angle). For a starting point, let's modify the Source 1 "both"numbers by +0.5 cm in each direction, corresponding roughly to change from Source 3 with knee extension from 90° to 110°. (We could do a bit more, since our angle is a bit more than this... but gotta start somewhere.
Finally, we must figure out how to choose a single number to go forward with, although the uncertainty will be significant. I can make an argument for the 13° being more heavily weighted (most of a person's weight is above the legs), or the 5° being more heavily weighted (78% of seated weight is typically on bottom cushion) -- but at the cost of adding a couple centimeters more of uncertainty, and in the absence of better info, I'll use a 50/50 split.
Exercise for the reader: Can you find more information to inform this transform?
So we get x ~= 19.02, z = 30.65. On top of the (1.4, 1.8) cm initial std deviation -- which we multiple by 1.28 to get our 90th percentile, or (1.8, 2.3) cm, we've added (2.0, 3.4) cm of process uncertainty, and about (2.2, 2.2) cm from posture range. So we reasonably expect most typical pilots to have a CG position of (19.0 ± 6.0, 30.6 ± 7.4) cm relative to the SRP.
Review, critique, and more info requested!
Problem statement: For a pilot sitting in in a seat with a 5°-from-horizontal seat angle, 13°-from-vertical back angle, and relatively low seat height (so knees at maybe a 135° extension angle), what is the relative position of the median pilot's CG relative to the seat reference point (intersection of seat and back)? What is the standard deviation of this position in each axis, and what is the position for a 90th percentile weight (R=1.28, 260 lb) male?
Source 1: "Whole Body Center of Gravity and Moments of Inertia Study." The measured seated position is a vertical seat (0° seat angle, 0° back angle, 90° knee extension), arms at sides. Population is reasonable.
| X cm (in the seat direction) | Z cm (in the back direction) | |
| Male | 23.48 ±1.4 | 26.22 ±1.4 |
| Female | 22.15 ±1.3 | 23.43 ±1.0 |
| Both | 22.82 ±1.4 | 24.85 ±1.8 |
| 90% male | 25.27 | 28.01 |
Source 2: "Biomechanical analysis of the dimensions of pilot seats in civil aircraft." Had this in my notes as relevant, but while it has some vaguely interesting info about pilot seat geometry in transport-category aircraft, nothing about pilot CG.
Source 3: "Determination of Centers of Gravity of Man." Fewer subjects (all male!), more positions. No subjects at our (current) 90th percentile. From table II, position (b) is the closest to the position used in Source 1, so let's look at that first to see how well the sources agree. Some risk of transcription error due to poor scan quality; ± is "across subjects", not std dev.
| X cm (in the seat direction) | Z cm (in the back direction) | |
| Male | 22.54 ± 2.2 | 24.45 ± 2.2 |
From this source, we see that both arms forward moves the CG forward about 2.2 cm, and up about the same; so this gives us a sense of how much CG varies for a single individual with normal movements.
Table IV goes into a slightly more interesting position for us -- a 110° knee extension angle. Standard hand-on-stick gives:
| X cm (in the seat direction) | Z cm (in the back direction) | |
| Male | 23.02 ± 2.2 | 25.08 ± 2.2 |
Exercise for the reader: Are there any other sources worth looking at here?
Coordinate transform: It's not quite a trivial rotation to consider the more typical seat geometry, because the back and legs tilt by a different amount. However, it's reasonable to assume that the numbers will fall somewhere between those of a 5° rotation (seat angle) and 13° rotation (back angle). For a starting point, let's modify the Source 1 "both"numbers by +0.5 cm in each direction, corresponding roughly to change from Source 3 with knee extension from 90° to 110°. (We could do a bit more, since our angle is a bit more than this... but gotta start somewhere.
| X (axis-aligned) | Z (axis aligned) | |
| Unchanged seat angle | 23.32 | 25.35 |
| 5° rotation | 21.02 | 27.29 |
| 13° rotation | 17.02 | 34.01 |
Finally, we must figure out how to choose a single number to go forward with, although the uncertainty will be significant. I can make an argument for the 13° being more heavily weighted (most of a person's weight is above the legs), or the 5° being more heavily weighted (78% of seated weight is typically on bottom cushion) -- but at the cost of adding a couple centimeters more of uncertainty, and in the absence of better info, I'll use a 50/50 split.
Exercise for the reader: Can you find more information to inform this transform?
So we get x ~= 19.02, z = 30.65. On top of the (1.4, 1.8) cm initial std deviation -- which we multiple by 1.28 to get our 90th percentile, or (1.8, 2.3) cm, we've added (2.0, 3.4) cm of process uncertainty, and about (2.2, 2.2) cm from posture range. So we reasonably expect most typical pilots to have a CG position of (19.0 ± 6.0, 30.6 ± 7.4) cm relative to the SRP.
Review, critique, and more info requested!
