# How do you find the center of gravity of an Anthropomorphic Test Device

In regards to the aircraft seat testing process in which the occupant's center of gravity is required, I am quite interested to know how people calculate the C.G. of an Anthropomorphic Test Device (commonly known as crash test dummy) It would also be really helpful if I could get relevant resources in the answer.

• Finding the CG (barycenter) of anything (rigid) can be done by hanging the object at successive points. The CG is at the intersection of the vertical lines crossing the points. 2 points in 2D, 3 points in 3D.
– mins
Nov 21 at 13:51
• @mins I was about to Post exactly that but are you saying we can be sure the CG won't change if one arm is extended, or the legs tucked up? Nov 21 at 22:04
• @RobbieGoodwin: In a constant gravity field, the CG is the barycenter of mass, if any change occurs in the shape or in the density, the CG location changes, and a new determination is required of course. That's why the CG can't be provided as a single location by the manufacturer. This method is similar to the one used to balance wheels, except the wheel is rotated to use momentum/torque in place of weight to detect smaller CG offsets.
– mins
Nov 22 at 9:00

I would expect that the manufacturer would be able to provide you with the location of the CG among the specifications of the dummy.

Unfortunately, the CG of a flexible object depends on the shape the object is taking at the moment. The CG is obvious for a uniform bar of steel, but bend it into the horseshoe and the CG is no longer located within the bar.

If I had an oddly shaped rigid object that I wanted to find the CG of, I would hang the object from multiple points. Draw a vertical line down the object from each hanging point. Where the lines cross is the CG.

The center of gravity of a standing human body is roughly about an inch below navel: https://www.sciencedirect.com/topics/physics-and-astronomy/center-of-gravity

For a seated person the CoG is shifted forwards and up of navel, total distance and angle will vary according to the specifics of the posture: https://www.researchgate.net/figure/Center-of-gravity-of-a-person-in-sitting-posture-Body-segmental-Lengths-are-reported-in_fig2_256203106

The antrophomorphic test dummies mimic the proportions, weight distribution and kinematics of a human body as close as possible, for obvious reasons: to accurately replicate the response of a human body to forces, the ATD must resemble it accurately, not just as a whole, but down to its individual part also.

Example information of THOR ADT: Anthropomorphic Test Devices; THOR 50th Percentile Adult Male Test Dummy; Incorporation by Reference

If you wish to build a more accurate model, here is an example of a further break-down of proportional masses of average human body parts (Source):

Segment Male Female
Whole Trunk 55.1% 53.2%
Thorax 20.1% 17.0%
Abdomen 13.1% 12.2%
Pelvis 13.7% 16.0%
Total Arm 5.70% 4.97%
Upper Arm 3.25% 2.90%
Forearm 1.87% 1.57%
Hand 0.65% 0.50%
Forearm & Hand 2.52% 2.07%
Total Leg 16.7% 18.4%
Thigh 10.5% 11.8%
Leg 4.75% 5.35%
Foot 1.43% 1.33%
Leg & Foot 6.18% 6.68%
• For some aviation-related purposes (but not necessarily the OP's), it might also be useful to know the CG location of a sitting human body-- but not surprising that your sources didn't cover that-- plus obviously would be dependent on how upright the seat back was, etc-- Nov 21 at 13:57
• Thanks for the heads up, table fixed. As for the seated posture: exactly as you wrote, there are a lo of variables; angle of seat, position of hands and legs etc. I added one "version" of this case. Nov 21 at 14:16
• Awewsome answer! TY Nov 21 at 21:36
• The center of gravity of a standing human body is roughly about an inch below navel: is of course also true of a prone or supine human. That's testable with gym equipment such as a foam roller if your abdominal muscles are in decent condition. Nov 21 at 22:03
• Yep, and also for left and right alteral recumbent and any other straight posture regardless of orientation : ) That was just how it was described on the site. As a sidenote I was contemplating using "erect" instead of "standing", but... Nov 22 at 7:50

There is no more need to calculate the CG of an Anthropomorphic Test Device than there is to calculate the actual CG of a live body.

Aircraft weight and balance calculations generally don't require that level of precision. Roughly centered on the seat, "X" inches from the datum is close enough.

• More precision is needed for weight-shift trikes, hang gliders, and the like. (But I've no idea how that would connect with the question's tag accident-investigation.) Nov 21 at 15:08
• @CamilleGoudeseune, it isn't just the tag, it's the central theme of the question. Are you saying that crash test dummies are used to test hang gliders? Nov 21 at 16:25
• Tee hee! No, I'm just lacking context for the question. Nov 21 at 17:14
• Yes, I agree that understanding "why" helps to answer appropriately... Nov 21 at 17:24
• @SG-005, Interesting... I am curious why, because the human body comes in an infinite variety of physical forms compared to a factory dummy, and it would seem that an average sized ATD made of material with a uniform density approximately equal to meat and bone would be sufficient. What do you do with the CG data? P.S. You haven't accepted an answer or commented on others yet, are you getting any useful information? Nov 21 at 22:32