FEMM simulation showing COP 3 and 7

[Omnibus]

My understanding is that when we're dealing with pmm as with any other engine we're dealing with just one contiguous system no matter how many parts or cycles it consists of. Take, for instance, an internal combustion engine. It has many different parts, mechanisms to convert translational into rotational motion, two or four different strokes with their respective actions and so on. Nevertheless, when we do the energy balance we consider the energy obtained versus the energy content of the gas we put in, treating the whole engine as a black box, as you put it.

In our case, in order for the pmm to be self-sustaining the energy obtained when the magnets are closing (minus the energy to pull them apart) should be greater than the energy to pull apart the cores (minus the energy obtained when the cores go the other way). This is even in the ideal situation (without friction). Note, the machine itself has to both produce and spend that energy through negative feedback. No energy from the outside. Therefore, withing one cycle it is only the sum of the energies (with their minuses and plusses) that are at machine's disposal to function and therefore the balance which is done has to account for all these energies within that one cycle and determine if the produced overall energy within that cycle is more than the energy spent during that same cycle.

In our case it is more. However it is indeed only slightly more because the exces energy only comes from the magnets but the work for moving the cores in and out is substantially more and it determines the overall balance. It is not as much more as in your numerical example but it is still substantially more.

As for the scale, the scale for all the energies presented is exactly the same, as is seen from the figures you show and from the numbers for the joules you've calculated. You've given all your numbers in joules so there's no way that the scales would be different.

Thus, the conclusion is, yes, it's an OU machine but a very inefficient OU machine even theoretically, as many even simpler OU machines are. It is the inefficiency of these machines (the low level of OU, insufficient to overcome the friction in a real device) which is the culprit for the massive lack of success we're experiencing in making a working perpetuum mobile. Not that making of a working OU machine is impossible but so far we haven't found a way to make these machines more efficient or to reduce the friction down to levels which cannot obstruct the intrinsic OU property of these machines.

[broli]

In our case it is more. However it is indeed only slightly more because the exces energy only comes from the magnets but the work for moving the cores in and out is substantially more and it determines the overall balance. It is not as much more as in your numerical example but it is still substantially more.

As for the scale, the scale for all the energies presented is exactly the same, as is seen from the figures you show and from the numbers for the joules you've calculated. You've given all your numbers in joules so there's no way that the scales would be different.

Thus, the conclusion is, yes, it's an OU machine but a very inefficient OU machine even theoretically, as many even simpler OU machines are. It is the inefficiency of these machines (the low level of OU, insufficient to overcome the friction in a real device) which is the culprit for the massive lack of success we're experiencing in making a working perpetuum mobile. Not that making of a working OU machine is impossible but so far we haven't found a way to make these machines more efficient or to reduce the friction down to levels which cannot obstruct the intrinsic OU property of these machines.

Omnibus I can run a completely new simulation that would jack down the forces due to opening and closing the cores by just using balancing magnets, the fact that this can be done shows the way you calculate COP is not completely correct.

Here's an other example of such balancing magnets:

http://www.youtube.com/watch?v=EBvEXeRmxJY

[Omnibus]

That would be very interesting because increasing the efficiency is indeed the goal. That energy for pulling apart the cores has to be reduced substantially. Hope your idea will work because I know from experience that adding more and more magnets for the purposes of helping to overcome the sticky spot in fact gets the project further into a dead-end. Also I would discourage using the term 'coefficient of performance' as having no scientific meaning. The term used in thermodynamics when carrying out such energy balances is efficiency or efficiency coefficient.

[broli]

That would be very interesting because increasing the efficiency is indeed the goal. That energy for pulling apart the cores has to be reduced substantially. Hope your idea will work because I know from experience that adding more and more magnets for the purposes of helping to overcome the sticky spot in fact gets the project further into a dead-end. Also I would discourage using the term 'coefficient of performance' as having no scientific meaning. The term used in thermodynamics when carrying out such energy balances is efficiency or efficiency coefficient.

phoneboy gave me some good insights and inspiration. The balancing or scaling of the core forces is no longer needed in a complete system. This energy is stored in a flywheel and got back out of it when the cores close. If we are going to use cams then we only need to make sure the friction is low. But at 25J excess energy per cycle you need to have some serious friction for it to not operate due to friction. Either the concept and simulation is flawed and it doesn't work or they aren't and it does.

The opening and closing is provided by a cam that's on the same axle as the flywheel and the pistons connected to the magnets on a sliding rail.

@phoneboy: can you perhaps animate or explain your design a bit better.

[Omnibus]

Well, let's see it. That's quite interesting. So far, the excess energy you're referring to sinks into an enormous energy expense for the pulling apart of the cores, as in your slightly exaggerated numerical example.