Join 3,415 readers in helping fund MetaFilter (Hide)

How much force to tear off a limb?
September 25, 2009 11:56 AM   Subscribe

How much force would it take to tear off a limb from the human body? Don't worry, I'm not planning on trying this at home. Yes, I realize that bodies are variable. If you are familiar with any research in this realm, please point me to it!
posted by Wordwoman to Science & Nature (21 answers total) 6 users marked this as a favorite
You'll need to define the nature of the force. A sword is force. Me grabbing your wrist is force. One method is easier than the other in separating your limb from your body.

Even if you're just, say, pulling your arm from your socket, how is your arm being held to do this? A rope around your wrist will provide a different fulcrum than, say, getting caught in a farm thresher.
posted by Cool Papa Bell at 12:04 PM on September 25, 2009

Oh, and which limb, at which joint? Despite both being ball-and-socket joints, it's significantly easier to dislocate your shoulder than to dislocate your femur at the pelvis.
posted by Cool Papa Bell at 12:06 PM on September 25, 2009

The question asks about "tearing" off a limb. Using a sword to sever a limb from the body wouldn't be tearing.
posted by dfriedman at 12:09 PM on September 25, 2009

Well, we know that it can be done with as little as four horsepower.
posted by ZaneJ. at 12:16 PM on September 25, 2009 [7 favorites]

This thread on the Straight Dope forum suggests that it's harder than one might guess. In particular, from multiple eyewitness descriptions of an execution by quartering as well as this Straight Dope column on the subject it appears that four horses, each attached to a limb, actually have a fair bit of trouble pulling the limbs off. In the case of Robert-Fran├žois Damiens it eventually required cutting the joints to make the job manageable for the horses.
posted by jedicus at 12:19 PM on September 25, 2009 [2 favorites]

How much force would it take to tear off a limb from the human body?

Well, there was one tug-of-war accident that resulted in two men having their left arms torn off. According to the article, the rope they were using could only withstand 26,000 kg of force, and snapped, and the injuries were from the rebound force.
posted by burnmp3s at 12:20 PM on September 25, 2009

Well, we know that it can be done with as little as four horsepower.

"Housepower" actually refers to an estimate of how much work a horse do over a normal workday, not it's maximum output.
posted by delmoi at 12:44 PM on September 25, 2009

delmoi - not sure where you heard that, but it's totally false.
posted by lohmannn at 1:01 PM on September 25, 2009 [1 favorite]

Clothes dryers can pull an arm off but I do not know how much force is involved.
posted by Iron Rat at 1:14 PM on September 25, 2009

I remember seeing Discovery documentary on medieval torture machines.
You might be interested in "The Rack".
They did some experiments with pig's leg, but I don't remember how much force was involved.
Machines of Malice - Going Medieval
posted by leigh1 at 1:16 PM on September 25, 2009

Depends. Twisting motions are particularly nasty. Don't get caught on a big lathe or a driveshaft on a truck.
posted by paanta at 2:07 PM on September 25, 2009

There are just so many variables here. For one, it's going to depend on whether the pulling force happens suddenly or gradually. Not just because a sudden force has a magnifying effect because of the impulse, but also because if presented with a sudden force (or if the person was unconscious or unaware) then the muscles are not going to have a chance to tense and absorb the impact, resulting in most of the force being borne directly by the socket. This I imagine would result in a significantly lower amount of force that the joint is able to bear since the socket joints have all evolved to work best with muscles stabilizing them. And therefore from that you can also conclude that the amount of force that can be sustained will depend on the level of musculature, leading to very different numbers across the whole spectrum of human physiology. As a thought experiment, if you took a very fit top athlete and had them hang from their hands in rings and slowly attached increasing amounts of weights dangling from their legs I would imagine that you could build up to quite a large amount of force being sustained through their shoulder joints -- an amount which would not be possible if the force was not built up to slowly or if it was a spindly 12 year old girl for example.
posted by Rhomboid at 2:26 PM on September 25, 2009

In one of the Mythbusters prison escape episodes they try to see if it is possible to get over a wall by putting the ball of your ball and chain into a cannon and shooting yourself over. They don't mention the exact force that it takes to pull a leg off at the hip joint, however they used a rope that was supposed accurately represent the tensile strength of a human leg at the hip joint. You might be able to get a specific number off of a Mythbusters forum or contact them directly.
posted by Procloeon at 2:35 PM on September 25, 2009

WAG-type numbers: based on the wikipedia links provided ZaneJ/jedicus and lohmannn we can assume the following:

i) two horses opposed can tear off at least a limb (four attached to each limb is a system composed of two sets of forces, normal to each other and thus independent to a first approximation)

ii) horses have a maximum impulse of just under 15 hp for a short while.


The power to tear off a limb is less than the peak power provided by 2 horses (15 hp)

This is about 22 000 J/s, or, if they do that in 1 second, 22 000 J.

Applied over a short distance, say 10 cm, that gives up a force of (F=E/d) about 200 kN.

So 200 kN is more than enough to get the job done.

Now, a sanity check:
The yeild strength of muscle (the closes approximation I can find with a quick google) is about 1-4 MPa or MN/m2.

Assuming an average arm has a diameter of 10 cm, for a cross-sectional area of about 80 cm2.

The the yield force for an "arm" made of muscle (not stronger tendons and ligaments) would be something like 7-30 kN.

So, lets put the force at somewhere between 30 to 200 kN. A wide range, I know, but close enough for government work (as the Royal Executioner of France will attest).
posted by bonehead at 3:11 PM on September 25, 2009

From what I recall, a lot of the various medieval tortures didn't work especially well, or were even possible, rather they were the product of a "what's some sick shit we can do to someone" brainstorming session. When it finally came time to tear someone apart with red hot tongs, the executioner would get frustrated and end up resorting to more conventional means.
posted by electroboy at 3:49 PM on September 25, 2009

There is research done on this, actually quite a lot of it. It's done with cadavers and used to calibrate crash test dummies and the like. No point measuring how much your car messes you up during a crash if you don't know how much force is needed to break a person.

I'm not sure of the search terms to use to find exactly what you're looking for but spend some time with google scholar and you should be able to find the numbers you're looking for. I read about this research in Stiff by Mary Roach. Here are a couple of references she gave which may be relevant (can't find pdf links, sorry)

Patrick, LM, Kroell, CK, and Mertz, HJ, Forces on the Human Body in Simulated Crashes, Proceedings of the 9th Stapp Car Crash Conference, SAE, pp 237-260

AI King Occupant Kinematics and Impact Biomechanics, NATO ASI Series E Applied Sciences, 1997 - Boston [Mass.]: Martinus Nijhoff, 1983-c1999
posted by shelleycat at 4:06 PM on September 25, 2009

I have always heard that you can remove an ear with just 8lbs of pressure.
posted by pianomover at 10:28 PM on September 25, 2009

I found this article that discusses the load required for failure on repaired cadaver rotator cuffs. They found mean failure strengths in the range of 273-325 Newtons, and in all cases the failures came from either suture/tendon-interface tears or just suture tears. One would assume that an actual, undamaged shoulder could withstand more force. Some internet conversion tool tells me that this amounts to about 61-73 lbs of force, so I think you could take that as a definite lower bound.

F. Alan Barber, Morley A. Herbert, Michael H. Boothby, Ultimate Tensile Failure Loads of a Human Dermal Allograft Rotator Cuff Augmentation, Arthroscopy: The Journal of Arthroscopic & Related Surgery, Volume 24, Issue 1, January 2008
posted by vytae at 11:03 PM on September 25, 2009 [1 favorite]

load required for failure

That's a really good search term for this.
posted by shelleycat at 4:10 AM on September 26, 2009

"Housepower" actually refers to an estimate of how much work a horse do over a normal workday, not it's maximum output.

delmoi, a minor correction. Horsepower clearly refers to the amount of power a horse produces, not the integrated power over time (how much work over a normal workday).

Of course, it does not estimate the maximum power output possible from a horse, but the average power output expected from a "standard" draft horse.

Interestingly, estimates of the equivalent "manpower" range from ~75 to ~300 W. The latter is for workmen aged 20-29 (peak of physical prowess), and of course the word's gendering is not sexist, but a realistic qualifier.
posted by IAmBroom at 1:30 PM on September 28, 2009

hmmm, i think a living or dead body is held together quite closely, and there is some work involved in detatching a limb (ever take down a chicken from the farmyard to fried on the table?) Each case is different, depending on living/dead, animal type, effective forces, etc...
posted by bebrave! at 10:02 PM on September 29, 2009

« Older What is the best Electronic Ne...   |  Immigrant iPhone: When I move ... Newer »
This thread is closed to new comments.