Quote from: fletcher on September 03, 2012, 12:37:24 AM
Yes, I can understand your approach of physical verification against model behaviour.

I was suggesting that cross sectional analysis of PE + KE v's Work Done is highly indicative that your formula's are indeed correct or near enough to not matter - if something is way out of kilter then it would be cause for concern & looking closer at the formula.

One thing you should be prepared for Dennis - that is, that criticism will rain down whether the sim [or sonsofsim] shows OU or does not - on one hand you will get comments about not capturing the true effect rigorously therefore not a reliable predictor of behaviour - And, on the other hand, that should it/they show OU that the formula's need work or that a sim that follows physics laws can never show OU.

That's why I admire your determination to accurately model a simple system first which matches real world behaviour & extrapolate the evolution thereafter with more sims from a solid first principles foundation.

fletcher,  I may grumble a bit when it looks like you want me to do more work, but I appreciate you inputs on my sim.  I take them all with due consideration. 

Yesterday, I spent most of the day trying to make a plot for the energy in vs potential energy out, or captured inside.  I broke down every individual change in pressure or head inside the ZED and all actual work done.  I then tried to show every combination that could be taken together.  I could not get anything that looked even close no matter how I played with the combinations.  That is when it dawned on me.  I was never going to be able to make it make sense using that approach. 

My whole simulation is based on balancing the input force with an internal state that creates an equal and opposite force on the input.  I was avoiding using the input forces and direct "push back" forces in all my formulas, because I knew that was going to show 100% by definition.  Yet that was the only calculation that could be used.  Either that or I am way off base in my relationships between the internal geometries.  Otherwise, I would have to show that I am creating and destroying virtual water on demand.  The real test will have to come from matching up the sim and actual build results

Lacking any further inputs from builders, I will push forward with completing my revised PDF with new animations. 

@ ALL BUILDERS,

I am quite interested in having measurements from tests that I can duplicate in the sim by adjusting the sim dimensions to match your models.  Please PM me with any details that you would not want to share publicly yet.  I don't care if it shows O/U or not.

I have been thinking about how to take measurements of the heads in a clear ZED.  The way I am leaning towards is to mount a LASER pointer on a right angle stick that can slide up and down on a fixed vertical post, wall, etc.  A poor man's height gage!  I happen to have a LASER pointer that puts out a line rather than a dot, which might make it easier to see.  The other thing was to get some colored plastic floating beads to put into each head to make the level easier to see, or a cutout sheet plastic ring might be a better way.

The needed data other than the ZED dimensions is the water heads at the beginning (down), the heads when the pod starts to lift, the heads at about half stroke, and the heads at the top of the stroke (just at the top stop if you have one).  The data would only be used to verify the accuracy of the sim, not to be shared by me.

My sim is currently limited to a single layer, but I will add more layers in time.  So, even if the model has more layers in it today, The data would be useful to me to evaluate the next iteration of the sim. 

TIA,
Dennis

Hi Dennis
If nobody else is that that point yet, I may be able to give you a hand by the end of the week.
I should have a single layer demo available to work with by then.
It will be 8" OD x 12" tall with clear .125 walls and .125 gaps.
Any assistance I can provide.

I'd already planned to work from the outside in, taking as many measurements as I can along the way.
It's set up to be 5 risers and a 2.5" pod fully assembled.
Clear, single top plate so it will be pretty simple to monitor what's going on inside - for better or worse

Dale

.

Red-Sunset said


The Pivot lies in the ratio between the outer layer and the pod area. A pod area as big as the outer layer surface would be unity, if we ignore losses. As you make the pod smaller than the outer layer, the outer to pod area ratio will become your advantage over unity.
 
1. The pod is a cylinder. By pod area , do we mean the area of one end of that cylinder?


2. As it would be physically impossible to have a pod with an area as big as the surface area of the top of the outer riser, and still fit the pod inside he riser , can we assume that this is a theoretical question?


3. So making the pod area smaller than the area of the top of the outer riser, is advantageous in terms of gaining OU. Is there, therefore an "ideal" pod to outer ratio,  Or can we enlarge this ratio to an absurd extent, and still gain? Can Red_Sunset, or anyone please answer these questions?