Topic for open discussion: Radial
Step-Tow Multi-r
|
Send
AWE notes and topic replies to editor@upperwindpower.com
|
?next? |
Sept. 30, 2020, post by Dave Santos Multi-r Form-Finding Notes Dynamic Form- radical morphing and topological surgery evolving on a tensile skeleton of persistent form. We can design a skeleton that supports evolutionary engineering of its supported parts. Economy of Scope- variability of form and diversity of function. No single dead-end design down-select. Dispatchability- responsiveness to season, peak demand, emergency disruptions of grid supply, etc. Bulk Modes v. Local Modes- bulk laminar v. local turbulent wind harvesting regimes. Local modes harvest energy that stall bulk modes. An ideal design will do both bulk and local mode harvesting. Many Wing System Validations- grid fins, starling murmuration, feathers, leaves, cellular/multi kites, etc.; all prove multi-wing lattices do bulk modes effectively. Echinoderm Biomimetic Form Model- better than we alone could imagine. Note metamaterial edge-mode optimization: https://www.georgehart.com/sculpture/sand-dollars.html [Ed adds: https://tinyurl.com/MandalasAWEformfinding ] |
Aug. 17, 2020, post by Dave Santros Radial Step-Tow Multi-r Inspired by Jellyfish Locomotion Dynamics: ============================= Max Langbein notes: Even if that fractal concept still had a low power-per-area-ratio, would you still think it would be worth working out ? I see the movement modes already being worked on by joe. Even a >100MW installation over hambach mine would be impressive, and could be re-powered afterwards. Of
course using a grid with self-organized motion modes, and power
transferred by self-stimulated mechanical waves running through the
grid instead of controlled movements, it would be more generic and less
control-intense, by that maybe also more robust. ============ Yes, We are able to point to progress in conventional wind power, where the latest turbines have bigger disc area but lower specific power.
They win by having a far larger capacity factor, due a lower cut-in
velocity and lower usable velocity, in "most probable wind velocity".
Lower specific power is less intuitive, but has its sweet spot for
lowest LCOE. And while multi-r may not at first use airspace maximal
intensity, the large sparse sweep patterns of single-line kiteplanes
are far worse. "Just add string" is the cheapest way to span Hambach at
a minimal power scale.
Yes,
1-10-100kW 1-10-100MW will all be a stepping stones in testing and
early operation, with no historical shortcuts in aerospace scaling.
Each scale-up is a challenge, an almost complete start-over. It will
take a big social commitment to go big by 2030 at about 100MW, and
maybe1GW by 2040, unless extreme world crisis moves everything faster.
Yes
please include Tassos, as he is very excited to participate, and
already thinking hard about the FEA CFD validation. His Thesis is
fairly relevant, and he starts from a many-connected metamaterial AWES
paradigm. Tassos represents academic cross-validation for TUK's multi-r
system identification. He brings the best of the hungry Greek
engineering science establishment.
==== ===== Max notes: Hello, Thassos, Thank you for sharing your thesis, containing deep thoughts about how to distribute forces, and algorithms to fill a structure in a way that maintains stability. Our specific problem here in Airborne wind energy, or our idea for a large airborne wind energy structure, would be a large dome-like structure,which is kept tensioned and upright by many little softkites. In the starting mode, the dome is moved by winches in a step tow fashion (each of the six pulley doing a sinus ,each pase-shifted by 60 degrees).afterwords, the kites move the winches still in the same fashion however with different force distributions. The
dome would be two-layered to make it possible to steer the kites
angles slightly (needed to switch between power gen modes) The challenge here is that the dome does not keep it's shape, but is layed flat initially, then slightly arched, etc., but still all strings should be kept tensioned. One
solution would be a strongly hierarchical structure, however that has
the disadvantages that it has only a low density, and is not very
flat, by that preventing "layering" of the domes which would be a nice
re-powering possibility of a site (just add more kite layers) The honeycomb structure in the images inserted here is just for illustration, also the kites on the dome are only illustration. Maybe you could come up with a structure for that. In the kites we still discuss if simple sails kept in shape and position by the structure, or self-adjusting kites with maintain aoa to be lifted, would be better. I
Tassos responds: Dear Max, Thank you for your email. It's a pleasure to meet you.
...
The
approach to AWE you describe certainly sounds very interesting. With
regards to the double-layered dome you are referring to, it reminded me
of some experiments I had done during my MSc. In fact, it was a project
that preceded my thesis (that you mentioned), but it was based on
the same principles, i.e. particle-spring simulation algorithms.
However, in that case, I investigated certain fixed topologies (or
uniform connectivity patterns) of nodes arranged on a 2D grid (whereas
in the thesis the topology generation was the point of focus). In the
few cases of more dense connectivity patterns I had tested, the nodes
would fall into a 3D 2-layered (or 3-layered) space frame structure on
their own in order to achieve equilibrium. I had actually made a short
video of this project at that time that may perhaps
help illustrate what I am trying to describe. I just managed to
find it and uploaded it in the following link:
I
suppose that the overall isomorphic hexagonal shape of the structure
you are investigating (shown in the images you sent) is to accommodate
the 6 points of support and associated winches along with the phase
difference in motion. I think that other, perhaps even more complex
forms could also be feasible, although simplifications may be required
to accommodate the dynamic motion (and possibly other complex)
aspects, at least as a starting point.
If
I understand correctly, you want the structure to maintain a certain
(controlled?) degree of flexibility and allow deformations, while
maintaining tension in all connections. Provisionally, I would think
that this could be achievable through the configuration of the
connectivity pattern (topology) and the possible combination of rigid
and tensile elements, but granted the complex nature of the structure's
resulting motion due to wind and mechanisms (winches), it may not be a
straightforward task.
In any case, I think that particle-spring algorithms could be a suitable tool for form-finding and simulating such a structure.
Let me know what you think.
Best regards,
Tassos
Max notes: ... On 04.09.20 09:40, Tassos Kanellos wrote:
Nice user-friendly application. Impressive.
The
inner structure is not fixed being hexagonal, the only constraint is
that the cell centers of the upper layer being near nodes of the lower
layer. (the kites being thought of residing in the cells of the upper layer and steered by the nodes of the lower) Initially
I looked at a tri-winch setup (being the simplest setup allowing
movements in all directions) , however then the area that could be kept
tensioned was a bit small.
Maybe a simplified sim of the kite aredynamics has to be combined with the network sim and tested to find a good structure.
Maybe
the links/springs would have to be more elaborated, using some simple
analytical model of the actually used material of the tethers,
including drag. After all, tether drag is a non-negligible property of an AWE system. Dave Santos responds:Tassos,
That's
a fantastic interactive spring-mass lattice sim video! It even supports
travelling waves, to drive Multi-r legs for generation. This is why we
were looking for you. All our higher polygons are triangular sets, and
larger lattices are properly >3 multi-r. You were already even doing
layers in 2007.
Not
clear just how gravity was supported, but in first approximations we
can mostly ignore gravity, since kite-lift negates lattice weight. We
can also ignore tether-drag to start, as a minor damping factor, since
our sails operate at relative low velocity (low Re). Because
most-probable wind velocity is constant, dimensionless relative
velocity and drag ranges lower at larger scales. Tether-drag to
tether-cross-section-strength also falls in our favor at bigger scale.
Gravity
setting might be used as analog wind acceleration, setting the
spring-mass model sideways. If there is no field gravity, but just the
moving ball, then the ball could be pushed upwards at an angle to
deform the model kite lattice somewhat as wind would.
Spring-mass
lattices spontaneously resonate harmonically when excited. The
engineering challenge is to tune the parameters to match ideal
load-motion output across all wind-velocities and load-demands. Even
your old tool could sort of do this by tedious manual interface
manipulations played-back at presentation frequency (faster than
ponderous kite lattice would actually wave)
There
is a wonderful opportunity to do quick-and-dirty visualization, as
updated tools are slowly set up for rigorous simulation. A starting
visualization could at least begin to match starting GW-scale
calculations three magnitudes beyond most current AWES scaling
limitations of single-line topology "energy drones".
Very Excited!
daveS
Note-
This is our 2D primitive spring-mass cell, from 1975. A recent insight
is that it embodies 1/2 of a tensegrity cell, where one spar is the
Earth surface, and the other half-spar is the air pressure-field. A
mirror balance of forces resides in the Earth.
Payne Fig. 5
|