Flaw in the starting strength model of moment arms (aka EPIC PHYSICS POST)

[spacediver]

Hi,
I appreciate, nay, *applaud* the effort by spacediver here, but just one bit I'd like to point out:

There are limits to how much useful (i.e. relevant to the technique of the lift) information you can glean from a single moment 2D model. What is worse, even a full 3D-model built from these ingredients would be inaccurate due to the mechanics of the shoulder not really conforming to these models in terms of force production. I agree fully that it is somewhat important, highly relevant and most of all very interesting to analyze the mechanics of every model in detail. The detailed analysis, however, is on only useful in one of two ways: Firstly, insofar as the model accurately describes the actual situation, and, in particular, if two actual situations are compared, the model accurately describes the relevant parts that make them different; in such a case the model and analysis are useful because they tell you something that can improve technique. Secondly, the detailed analysis of a model, when it does *not* describe an actual situation, may be useful if it reveals that we should not use the model in the said situation.

To achieve one of the two, we must be able to compare the results of the analysis to something that actually happens. Mind you, I am not taking the stand here that one should refrain from making analysis such as spacediver is doing, quite the opposite. But I am pointing out that perhaps such enthusiasm would serve a more useful purpose if directed towards finding criteria for validating the findings, or, alternatively, enrichment of the model to account for the 3D-nature of the press.

I sadly do not have time to analyze this model in detail myself. But from what I can say, it is impressive.

Don't give up and never stop.
-H.

I completely agree with everything you said, except that I think detailed analyses can be useful in a third way:

The purpose of my model wasn't to show that in the press, the forces and torques behave as I've described. The purpose was an existence proof of a situation that disproves the dogma of "moment arms between barbell and joint x always increase force requirements of muscles around joint x". I could have modeled any simple system, biomechanical or otherwise, to prove the same point.

Rip says that we should reduce the moment arm between hand and shoulder in the press. When asked why, I imagine he would say that this reduces the force requirements on the shoulder musculature. If you were to ask him how he knows this, he'd probably say "because physics".

This post is attempt to show how anyone who says "because physics" in response to that question is wrong.

[spacediver]

Seeing as how I also went to Stanford, but mostly on the dummy admission tract, I'd like to go through this and offer comments/criticisms.

I agree with Tiedemies regarding the comment about 2D versus 3D analysis, but I think that's smallish potatoes.

What I will say is that in just a skimming of this material I see no free body diagrams and no summation of force/moment equations. If you're going to try and prove that the application of a force at varying locations requires no appreciable change in the applied force in a mechanical system, then you might want to actually calculate the forces and see if they balance at various points. That beings said, it's probably in here, but I just don't see it.

Also, angular momentum is irrelevant in this case for two reasons; one, these are effectively statics problems because of low velocities and two, you're looking at linear translation of the mass, i.e., the weight moved by the system.

Again, I'm probably wrong.

You're correct that angular momentum is irrelevant. I perhaps could have done without the first of the two posts. It served a somewhat different purpose, and was mainly motivated by a comment by comrade Tom in the other thread, where he had trouble understanding how the humerus could abduct without the shoulder rotators doing any work.

As for the free body diagrams, they're unnecessary, when all we want to know are the force requirements of the two fibers. If we wanted to know internal joint forces (say we wanted to build an artificial contraption like this and know how much stress the joints can handle, or how much compression and tension the components can handle), then we would need a free body diagram. My goal originally was to learn how to construct a free body diagram for this problem (I have a text that Savs recommended on its way to me right now, and will hopefully help me learn how to do such a thing), but it became clear when working on this particular problem, that free body diagrams were unnecessary.

[spacediver]

I'm not sure if this has been mentioned before in this or the old thread, but if you are talking about a barbell movement (= not a dumbell movement), you would have what my mechanical engineer friend calls a "symmetry constraint", which could be modelled as the load moving in a vertical "rail" with all the reaction forces this implies.

This also means infinite solution using only two muscles. Either muscle could take all or none of the load. ...According to the model that is.

I don't think this applies here. Suppose you try to press a barbell with severed triceps. Assuming the elbow joint is at an acute enough angle of flexion, then the moment you exert any upwards force on the bar with your hands, the elbows would collapse (i.e. the reaction force of the bar would cause the elbows joint to undergo flexion).

[spacediver]

spacediver, good stuff! I have minor quibbles here and there (you still aren't drawing proper free-body diagrams which leads to JC Superstar's comments and Satch's complaint, I'm not entirely sure of your undetermined claim, and a few other minor things), but in my opinion your main points still stand! That's why I said I don't think the (SS) statics analysis (torque about knee and hip in the squat, for example) is thorough enough. Also, now you get to think about the "uses the most muscle mass" statement. Hahaha. Aw hell, I really don't want to get dragged back into all of this. I have very strong opinions about the lack of scientific rigor coupled with the arrogance of many of the claims made in these forums. Some of the things are very difficult to simply call falsehoods.

I'm swamped this week, but for whatever it's worth maybe I'll take a stab at a dynamics model this weekend, and work out what approximations are necessary to show when a (proper) statics treatment is good enough. It's something I've been meaning and wanting to do. Maybe I'll find time for it. Maybe somebody else will beat me to it!

appreciate the input, and also the encouragement from everyone.

[spacediver]

Are we assuming a straight-line bar path?

yep

If not, can a curvilinear path be introduced?

Yes, but then you'd need free body diagram, since you'd need to calculate forces required for a resultant force at C that isn't parallel to gravity.

Is it worth considering the facet design of the joints themselves in the analysis?

yea, this is something that might turn out to have some interesting effects. For example, in "real life", the point of rotation shifts slightly. Also, in real life, the limbs have mass, and this does change things slightly.

[spacediver]

... and while I agree that there is room for improvement in the model (like there is with all theoretical models), I still feel that the book does a more than adequate job of presenting a sound rationale for the performance of the lifts as they are described.

I agree with this, and it's part of what motivated me to want to improve it. I'm actually more concerned with the credibility of the SS model than how much my analysis would actually practically change things.

[Murelli]

yea, this is something that might turn out to have some interesting effects. For example, in "real life", the point of rotation shifts slightly. Also, in real life, the limbs have mass, and this does change things slightly.

First I want to congratulate you on the job done. Great stuff there.

About limbs having mass: ignore it. Think how much your forearms and arms weigh compared to a 60-120 kg loaded barbell. Mostly irrelevant and unnecessarily complex, IMO. If you want to increase complexity, I believe a numerical simulation with more muscle elements or the 3D analysis would be more useful. Also, torso rotation would greatly improve our analysis of the limbo press, and how keeping the barbell over the midfoot and changing the moment arms a bit can help you during the different phases of pressing (think press 2.0 vs. strict).

After I changed my mind about 5 times in the latest belt thread, I started to see SS as Petrizzo has put - a book for novices, in a novice language.

[Lasse Høi]

I don't think this applies here. Suppose you try to press a barbell with severed triceps. Assuming the elbow joint is at an acute enough angle of flexion, then the moment you exert any upwards force on the bar with your hands, the elbows would collapse (i.e. the reaction force of the bar would cause the elbows joint to undergo flexion).

With the symmetry constraint you've only got one degree of freedom, thus you only need a single muscle. The point being of course that with more muscles you get infinite solutions.

Regardless, we still need to define what the optimal solution is. You have proved that the deltoid force minimum is not where the load is directly above A, for a particular scenario. But the force minimum force is not what is limiting the weight lifted. A better proof of your claim that ss is wrong, would be to show the maximum of the deltoid force vs. bar position, for different grip widths.

[John Petrizzo]

I agree with this, and it's part of what motivated me to want to improve it. I'm actually more concerned with the credibility of the SS model than how much my analysis would actually practically change things.

Very cool. Thanks for your efforts.

[spacediver]

With the symmetry constraint you've only got one degree of freedom, thus you only need a single muscle. The point being of course that with more muscles you get infinite solutions.

This constraint issue is interesting. I think I can see how a barbell introduces constraints that are not present, for example, with a dumbell overhead press.

Regardless, we still need to define what the optimal solution is. You have proved that the deltoid force minimum is not where the load is directly above A, for a particular scenario. But the force minimum force is not what is limiting the weight lifted. A better proof of your claim that ss is wrong, would be to show the maximum of the deltoid force vs. bar position, for different grip widths.

Completely agree that reducing deltoid force requirements does not necessarily mean optimal (I raised this point at the end of the second post). It would be interesting to show the grip width that maximized deltoid force requirements, although I'd need to explore a wider range of joint angles, and this is not straightforward when modeling muscles with straight lines, as you run into boundary cases, such as when the triceps becomes parallel to line BD (see for example top left of Fig 5 where this boundary case is approached). But again, the purpose of my approach wasn't to demonstrate which grip width is best, but rather to demonstrate how a particular style of physics thinking is misguided.