# Is it an undisputed fact that machines can never be 100% efficient?

## Recommended Posts

In my high school physics class last year, I was dissapointed to hear that machines can never be 100% efficient. They will always lose energy through heat or some other medium because, to me, "that is just how it works." Is there any debate open to there being a 100% efficient machine, or is it literally impossible? I have had dreams of a magnetically powered self-sustaining engine, but it can't happen.

##### Share on other sites

Yes, it's impossible. Thermodynamics sees to that for you.

##### Share on other sites

What about the simplest possible machines, like a single molecule performing some action or something? It's not so much a law at the quantum level, is it? Rather it's more of a statistical thing, where the probability of performing a large number of actions without loss is statistically impossible, and the effort required to make a large system more efficient makes it more complex and even less likely to be 100% efficient. Either way, > 100% efficiency would not be possible, but does Newton's law hold in quantum mechanics?

##### Share on other sites

What about the simplest possible machines, like a single molecule performing some action or something? It's not so much a law at the quantum level, is it? Rather it's more of a statistical thing, where the probability of performing a large number of actions without loss is statistically impossible, and the effort required to make a large system more efficient makes it more complex and even less likely to be 100% efficient. Either way, > 100% efficiency would not be possible, but does Newton's law hold in quantum mechanics?

Wouldn't an atom or molecule transfer/lose photons, at the very least, in the act of performing some action?

##### Share on other sites

The only answer of thermodynamics is that energy is conserved, that is, efficiency defined that way would always be 100%.

So efficiency needs a smarter definition, something like "the useful energy divided by the costly energy".

If the useful energy includes the resulting heat, for instance if producing electricity from hydrogen and heating the building with the energy not converted to electricity, you can get 100% - this depends on what kind of heat is useable.

A heat pump that takes heat outside and brings it to a warmer temperature in the house has an efficiency around 300%.

Lossless... Type 1 superconductors have no minimum loss, under the proper conditions.

"Bearings" made of single molecules were also observed to have no loss.

With some superconducting bearings and vacuum, losses get really tiny - but type 1 superconductors achieve little. The stronger type 2 do have losses.

If you can store 1MW*3h during 10h with less than 3% losses, you already challenge that

http://www.scienceforums.net/topic/59338-flywheels-store-electricity-cheap-enough/

for that nice use, zero losses are not necessary.

Edited by Enthalpy
##### Share on other sites

I would warn against being too loose with terminology.

Some mechanical machines are theoretically 100% efficient. There is no theoretical barrier to this.

For example a simple lever, in the absence of friction, has an efficiency defined by the ration of the mechanical advantage to the velocity ratio which may be unity.

An simple example would be a simple lever.

In the real world no mechanical machine is perfect and therefore no real world machine is 100% efficient, but this is due to imperfections, not theoretical barriers.

Thermodynamic machines, called heat engines, on the other hand suffer the Carnot limit to their efficiency, which depends upon temperature difference.

The difference is that thermodynamic machines convert heat energy into mechanical work and vice versa, whilst mechanical machines put mechanical work in a more convenient form.

Please also note that the correct term in current usage for heat pump performance is 'coefficient of performance' or COP, not efficiency.

##### Share on other sites

What about the simplest possible machines, like a single molecule performing some action or something? It's not so much a law at the quantum level, is it? Rather it's more of a statistical thing, where the probability of performing a large number of actions without loss is statistically impossible, and the effort required to make a large system more efficient makes it more complex and even less likely to be 100% efficient. Either way, > 100% efficiency would not be possible, but does Newton's law hold in quantum mechanics?

Only when an atom has no extra states available to it can this be true. Once you involve additional states, entropy increases. In statistical mechanics, entropy is given by k ln(n), where n is the number of available states. So if n increases, entropy increases. An atom in the ground state has "nowhere to go" in terms of shedding energy, so it's 100% efficient. Only an atom in an excited state can lose energy.

That typically goes away once you add in a second particle.

##### Share on other sites

In my high school physics class last year, I was dissapointed to hear that machines can never be 100% efficient. They will always lose energy through heat or some other medium because, to me, "that is just how it works." Is there any debate open to there being a 100% efficient machine, or is it literally impossible?

Temperature of particle is connected to its velocity.

Hot particle is fast moving, cold particle is slow moving.

Once hot=fast moving particle is colliding with slow=cold moving particle, it's giving part or all of its kinetic energy.

It's never returning it back to source.

Device such as engine has chamber to which there is injected fuel which is mixed with air (or pure oxygen), and burn together forming some kind of fast moving oxides.

They're colliding with chamber giving it part of its kinetic energy, and heating chamber (unwanted effect). And moving f.e. piston (wanted by us effect).

Such explosions happen dozen times per second.

Engines are usually made of heavy durable metal to be able to stand these explosions.

Metals are good at absorbing heat, temperature is nicely spread across whole engine.

Air molecules after colliding with running hot engine from outside, are accelerated, and engine is cooled down, and it's lost forever energy.

Remaining energy is emitted as photons at various frequency ranges depending on temperature.

If temperature of metal is too high, it's starting emitting visible light (it's used in light bulbs), and in the worst scenario is changing to fluid.

Engines could be surrounded by Peltier cells to produce additional electricity.

Edited by Sensei
##### Share on other sites

Temperature of a single particle in not defined. Temperature is a property of a collection of atoms.

##### Share on other sites

If your using some form of any energy you cant replace it; For example you may capture the unused heat from a lightbulb and convert it back into electricity but the photons are used up in the environment. Same applies to heat, engines, light, radiation, electricity etc. If your not using some of the energy of the system you may be able to traverse it 100% efficently i suppose?? if the technology exists then its scientifically possible, for 99.9% of practical use you will be using some energy from the system to perform a task.

The petrol in an internal combustion engine is used to move your car forward, you may be able to capture the remaining heat and convert it back into usable energy but the process of moving the car (the explosion in the piston which rotates the camshaft) has already used some of the energy and that is gone forever. This should help you understand efficiency.

I'm not sure the technology is there to convert energy and forces with 100% efficiency but i believe its theoretically possible, infact the science accounts for *almost* every possible way the energy will convert. (finding practical materials and methods is harder)

Regards.

EDIT:

This is the same reason perpectual motion devices dont work, the electrons that are somehow propelled are equal to the force used to propel them.

I'm not sure if you can convert energy and force with 100% efficiency but you'd gain nothing from such a system except the energy lost in translation. (which in some cases counts for alot)

Edited by DevilSolution
##### Share on other sites

I'm not sure the technology is there to convert energy and forces with 100% efficiency

What do you mean convert forces?

So long as there is no actual movement my lever example provides a 100% 'efficient' force 'convertor'.

##### Share on other sites

• 2 weeks later...

Law of thermodynamics says? there is definitely soething tansformed to heat energy. You apply force to a simple machine? you already burn energy and some tranformed to heat so there are still loses. But if we compute only during which the force is applied there are still loses. Eg: law of gravitation that each object aith mass pulls one another adds to something you need to overcome and will need you additional force. The scale might be mocroscopic but still loses ssuming .0001%

Edited by gabrelov
##### Share on other sites

What if you said "This is my machine it is designed to make a lot of sound, radiate heat, decompose over time" then your machine would be 100% efficient.

I think non 100% efficiency means the forces that any machine uses are non conservative, or at least some of them are, that is mechanical energy (U+K) is not conserved.

Edited by Subliminal
##### Share on other sites

What if you said "This is my machine it is designed to make a lot of sound, radiate heat, decompose over time" then your machine would be 100% efficient.

I think non 100% efficiency means the forces that any machine uses are non conservative, or at least some of them are, that is mechanical energy (U+K) is not conserved.

Then you're stretching the meanings of the words beyond the point of meaninglessness. Just because inefficiency is a *desirable* quality of that machine, doesn't mean that "inefficient" now means "efficient".

We'd have to agree on what "machine" and "efficiency" means, but I think the expected meaning is "something that performs a (mechanical) action" and 100% efficiency means the energy used is conserved fully in the intended (mechanical) action, and none is converted to other forms.

(Mechanical) might also be replaced by electrical, etc.

##### Share on other sites

What if you said "This is my machine it is designed to make a lot of sound, radiate heat, decompose over time" then your machine would be 100% efficient

Cheeky!

It was ever the prerogative of the young and inexperienced to push the boundaries, but rest assured, the professors who developed the rules over two centuries were cleverer than you an I put together.

Efficiency is defined as output over input.

Consider the following machine:

A completely sealed box with fully adiabtic rigid walls contains a quantitiy of substance A.

Energy is added to the machine, perhaps in the form of heat, perhaps in the form of electricity, whatever.

Since there is zero output the efficiency is zero.

So the positions and salaries of white haired aged professors are secured

phew!

Edited by studiot
##### Share on other sites

What if you said "This is my machine it is designed to make a lot of sound, radiate heat, decompose over time" then your machine would be 100% efficient.

I think non 100% efficiency means the forces that any machine uses are non conservative, or at least some of them are, that is mechanical energy (U+K) is not conserved.

My thoughts as well. No one seems to have seen the elephant; define machine.

##### Share on other sites

Why do even well oiled machines wear down?

in the words of r~claus-

def: entropy

in any natural process there exists an inherent tendency towards the dissipation of useful energy

def: in other words

what starts with a bang; ends with entropy

just sayin

ron

##### Share on other sites

The word "efficiency" usually relates to conversion of energy from one form to another - "machine" is a little more vague. If we consider engines, such as an internal combustion engine, these are heat engines which convert heat (from burning fuel) into mechanical energy by a cyclical process. The Carnot cycle is an idealised version of a heat engine. It has higher efficiency than a real heat engine, but its efficiency can never be as high as 100%.

Machines such as electric motors do not fall into the cyclical, heat engine variety so the Carnot cycle says little about them. An electric motor will clearly not be 100% efficient because of loss of energy by ohmic heating and frictional losses at bearings. However, my guess is that the original question concerns a general argument which can be used to rule out 100% efficiency in such cases.

Edited by JonG
##### Share on other sites

The only answer of thermodynamics is that energy is conserved, that is, efficiency defined that way would always be 100%.

So efficiency needs a smarter definition, something like "the useful energy divided by the costly energy".

If the useful energy includes the resulting heat, for instance if producing electricity from hydrogen and heating the building with the energy not converted to electricity, you can get 100% - this depends on what kind of heat is useable.

A heat pump that takes heat outside and brings it to a warmer temperature in the house has an efficiency around 300%.

Lossless... Type 1 superconductors have no minimum loss, under the proper conditions.

"Bearings" made of single molecules were also observed to have no loss.

With some superconducting bearings and vacuum, losses get really tiny - but type 1 superconductors achieve little. The stronger type 2 do have losses.

If you can store 1MW*3h during 10h with less than 3% losses, you already challenge that

http://www.scienceforums.net/topic/59338-flywheels-store-electricity-cheap-enough/

for that nice use, zero losses are not necessary.

300%? That would suggest that a net gain in energy wouldn't it?

##### Share on other sites

Is a photon a perfect device due to the endless reciprocation between the magnetic and electric components in a vacuum?

Edited by StringJunky

## Create an account

Register a new account