Login required to started new threads

Login required to post replies

Re: cadence [tigermilk]
In Reply To:
In Reply To:
In Reply To:
In Reply To:
The simple pendulum is not a machine. All the parts are in equilibrium. The only energy conversion that takes place is between kinetic and potential energy of the individual molecules. There is no energy transfer between the parts. Swinging back and forth using an idealized model does not cause an energy variation in the system which would violate the 1st law since all the kinetic energy changes that occur can be accounted for by potential energy changes. I accept that analysis as valid. Unfortunately for your position in this discussion though, it cannot be applied to the pedaling model.
I must say this entire thread has been quite amusing. I've been left scratching my head at numerous times and muttering "huh?" I particularly love the quote "The simple pendulum is not a machine." Sure it is. Look up machine in the dictionary and tell me what you find. I'll give you a hint - lots of references to mechanisms. The pendulum is a mechanism. A one-bar simple mechanism.

The thigh, lower leg, foot, crank system is nothing more than a 4-bar linkage. The point at the hip is constrained against translation and rotates in a plane parallel to the plane of the bike. Likewise the bottom bracket. In a frictionless system, give either the "hip" or the "bottom bracket" point an initial rotation via some external torque and you set the wheels in motion (pun intended). That system will continue to move per it's kinematic constraints ad infinitum unless acted on by some other external force/moment.

Now you can argue all you want about the kinematic constraints, but if the problem is properly constrained (i.e., it's not an unstable mechanism), you can fully define the kinematic motion of every point in the system, whether it's rigid, flexible, or a combination of flex and rigid. And you can "transfer energy" from one rigid member to another.

A whole industry has been established based on the above. You can purchase quite complex codes for performing multibody dynamics, be it flex, rigid, or flex/rigid. Look up a code like ADAMS as an example. But I guess companies like MSC are completely clueless and should have consulted with you first. Their codes must be all wrong....
ma·chine (m-shn)n.1. a. A device consisting of fixed and moving parts that modifies mechanical energy and transmits it in a more useful form.b. A simple device, such as a lever, a pulley, or an inclined plane, that alters the magnitude or direction, or both, of an applied force; a simple machine.2. A system or device for doing work, as an automobile or a jackhammer, together with its power source and auxiliary equipment.

Exactly which one of these definitions of "machine" does the simple pendulum satisfy? I simply don't see the simple pendulum as a machine. I guess you could claim that a spinning disk is a machine also. A spinning disk is something that should continue forever unless acted upon by an outside force. Both the simple pendulum and the spinning disk is in equilibrium with itself. The MMF system converts an up and down motion to a rotational motion satisfying both definitions 1a and 1b. If loaded, as when riding a bicycle, it also satisfies definition 2.

Yes, it should be possible to work this entire problem out mathematically. The only issue would be the assumptions used for calculating the losses in the materials used. What would one use for the soft tissue of the leg for instance?) If there is an adequate system for analyzing this why doesn't some one do it and prove me wrong. All we have is hyperbole coming from the believers in the preposterous notion that this machine has no losses.
If you were an engineer or a pragmatist you'd recognize it as a machine.

Regarding the "MMF", you do realize that if someone were to model this, the FIRST step would be to assume a rigid body system don't you? That's engineering 101 at work there. Same thing with a mass-spring system. You start with the "unrealistic" model of no losses which actually tells you a WHOLE lot. You can bump up the fidelity and add damping and such as you get smarter or more adventurous, neither of which I'd suggest in your case for obvious reasons :) But as you add those "losses" you better have some good data. For this particular problem you could merely parametrically investigate the effects of various system damping, friction losses, and such.
Yes, I do realize that. However, the drawback of that system is it is not a very good model if one is trying to model internal losses. How could it be? Why can't anyone but me here see that? The rigid body model assumes internal losses are negligible. But, experimentally we know that when pedaling internal losses are not trivial, at least when the model is unloaded, according to Papadapalous. How is it posible that a model that assumes losses are trivial is still being touted as being appropriate model to answer the question as to whether they really are?
In Reply To:

Any "machine" is a construct of your own devise. As demonstrated above, you can set the analytical problem up to have no losses. If you throw this mechanism in a multibody dynamics program and specify no losses, then you will get a perpetual motion machine. You absolutely have to as you've specified no losses due to damping, friction, heat, hysteresis, etc. If you're simulation comes back otherwise, the program isn't implementing the equations of motion properly.
I understand that. A rigid model is useful for looking at forces and stresses on materials but not much else. As I said above, how is such a model useful for modeling real world energy losses if that is what you are interested in studying?
In Reply To:

Your obvious mistake is trying to interject what you see in nature (the lack of perpetual motion machines) and pigeonhole it into idealized models that we engineers utilize every day in design and analysis. You see that as a flaw in the methodology. We see that as an idealization that gets us 99% of the info we want.
No, my "mistake", at least from your view, is not falling into line and accepting the prevaling view that an idealized model that ignores the possibility of internal losses is a good model to analyze internal losses in real life. I don't buy it. I am surprised that you are taking this side. What is your evidence that this is a good model for this purpose?

--------------
Frank,
An original Ironman and the Inventor of PowerCranks
Last edited by: Frank Day: Oct 27, 09 6:53

Edit Log:

  • Post edited by Frank Day (Dawson Saddle) on Oct 27, 09 6:53: clarity as to meaning