Wave Tank Question

Not specifically boat design related, but still related...

If one were to develop a small-scale wave tank that was about 6X longer than in width (4' x 24'), and one wanted to be able to generate waves along the tank length as well as waves diagonally across the tank, would it be better to have a system that generates fewer transverse intervals with (graduated) depth; or gridded intervals that may have finer width-wise gradation, but less depth-wise gradation?

Following images attempt to clarify the question. The 'LH' is an adaptation Edinburgh Designs' shallow-water 'piston' generator. Regarding 'RH', I'm looking at a transition from round piston to square duct commensurate with a tunnel test section to impeller, assuming flow would be reasonably smooth (big assumption...).

Any advice appreciated.

 

What is your intended purpose? Wave generator design is generally selected to meet a specific requirement. And, BTW, you have to give a lot of consideration to the "beach".

 

This is the very beginning of an idea for a small tank that might be used in a science museum (I design and build interactive exhibits). I'm interested in more than the 2D generators I’ve seen in museums and was thinking something along the lines of 4'w x 2' deep x 32' long would be manageable to prototype yet wide enough to develop angled wave fronts with multiple generators.

I could also see a towing rig for models -something on the scale of an Edmond Bruce tow tank. Accuracy of results would be less important than the ability to demonstrate general principals and phenomena, and to have fun with experimenting!

The blue and purple sketch is based on the Edinburgh Designs wet-back piston generator. I'm guessing their generator is based on this S.H.Salter idea form a 1981 paper 'Absorbing Wave-Makers and Wide Tanks.'
If the inner cam was independently actuated, you might be able to emulate both shallow and deep-water wave forms:


The wet-back generator makes tank design simpler, but there is a lot of hardware and moving stuff up above the tank. That got me to thinking about using a grid of small, servo-driven pistons with transition ducts to go from round to square. No one wants to seal a square piston, right?


I understand the need to for a beach. I was thinking of closed-loop feedback for the generators. And while I’m at it, mirror the system for an active beach! This gets away from the analogy to a ‘real’ beach, but could give you more control over the waves in a much shorter tank, no? Guess a small 2D test section is in order first .

 

First, I would suggest you get a copy of Oceanographical Engineering by Weigel and read the first few chapters to better understand the difference between periodic and solitary/impulse waves. While this may make no difference for a simple demo "wave" tank, it has serious implications on what you are trying to engineer; i.e. the difference between paddle and pneumatic wave generators. Second, I would tell you to google "nswccd wave generators" for a description of the MASK (Maneuvering And SeaKeeping) tank and what it is capable of because it fits your needs to a T, just on a larger scale. Finally, if you want to combine flow and waves, I would suggest you google the old (circa 1969) Chicago Bridge and Iron (CBI) Marine Research Facility tank for the elegant solution of how to combine everything (i.e. flow, wave height, and water depth).

 

Thank you for the references! I am not familiar with the Weigel book, but will look into it. Along with numerous research papers I've gathered online, I have print copies of Dean & Dalrymple's "Water Wave Mechanics for Engineers and Scientists," Kinsman's "Wind Waves: Their Generation and Propogation on the Ocean Surface," and Bascom's Eaves and Beaches. The math in the first two is a bit of a slog, but I'm finding my way through it. Bascom is much more accessible, which is the intent.

I am familiar with the MASK tank and the sort of array with pleated gaskets between bottom-hinged paddles. I'm most interested in periodic waves or the ability to mix waves, and not so much with impulse waves and the pneumatic generators used in water parks. The two methods I put out are both periodic generators using 'paddles' in one form or another, aimed toward modular construction of smaller units that would avoid the complex seals and pleats common to dry-back paddle wave makers.

I'm intrigued by the idea of including flow with wave generation, but probably won't pursue it due to the huge power requirements. I've built a wind tunnel with a 10" sq. test section, powered with a 2HP axial fan and VFD. Pumping water of any quantity through a much larger tank section isn't in the budget, and I don't know where the power would come from in my shop! Towing small models through the wave tank is quite feasible.

 

Actually, the models you presented generate impulse waves, no matter how periodic you think the output is because there is water flow that is not compatible with the requirements of a periodic wave, hence the need for "complex seals and pleats". As someone who has used, and measured, the capability of the MASK before and after the renovation, I will say that the new system is not better or worse than the previous pneumatic system for most spectra, but it is capable of generating larger, and steeper, waves due to the vertical limits of the older system.

 

Calvin and Hobbes by Bill Watterson for November 16, 2021 | GoComics.com https://www.gocomics.com/calvinandhobbes/2021/11/16

 

Touché, John!

I'll admit to being naïve about the difference between periodic and impulse wave generators as you termed earlier. My understanding is this:

1. An impulse wave is formed by a mass moving into water. A tsunami is a prime example.
2. The above would make any wave generated by a piston, hinged flap, or plunger an impulse wave, regardless of the arrangement of the generators and whether there are water-tight seals between adjacent generators.
3. The distance between generators in an array and the volume of water that can flow between generators vs. being directly impinged upon by the generator surface will affect the wave form.
4. Any regular sequence of impulse waves would be periodic, regardless of the wave generator format.
5. Impulse wave generators are the typical method for generating periodic waves in wave tanks, with the the generator format prioritized to emulate wind-generated waves with shallow water or deep water characteristics. There are some tanks that incorporate wind and current generators, but they aren't the norm.
6. This one is a stretch on my part, but I'd assume any periodic wave induced by a wave generator will take some distance to 'settle' into the form it will hold down a tank of constant cross section, before being affected by the beach or tank end. Force feedback and active absorption applied correctly can decrease the distance required for this 'settling', allowing more wave intervals of uniform profile to be accommodated in a shorter tank.

Quite a bit of it it is summarized here: Wave Generators | Edinburgh Designs http://www4.edesign.co.uk/waves/some-wave-1/
Edinburgh Designs does a nice job of summarizing information for a lay person, while you know a ton of mathematical and engineering genius goes into the projects.

I'd appreciate you debunking my wrong assumptions!
Chris Krumm

 

Ok, so this may get a little persnickety, but the real difference between a periodic wave maker (NOTE: I have decided to use the term "maker" vice "generator" to help those that may follow from sliding down the wrong rabbit hole; i.e. a maker puts energy in, a generator takes power out) and an impulse wave maker is in the shape and distribution of the orbitals which controls the second order and higher drift terms. A true periodic wave has no mass drift, and in order to replicate this, there must be no net flow in the horizontal plane; typically pneumatic or continuous serpent-type wave makers are used to get this and even can generate oblique periodic wave forms. Impulse type wave makers (by this I mean individually operating flappers/pistons/etc.) get flow in both the horizontal and vertical plane between the operating units. This is not to say that the proper periodic orbital form cannot be generated with a sufficient number of small units, but it sort of gets self limiting due to the complexity.
So this gets us back to my first question..."What is your intended purpose?" If all you are trying to do is replicate a particular sea surface, then either type will do. If you are trying to look at mooring offshore structures or sub-surface effects you will need something that responds closer to a periodic wave maker.

 

Thanks for the clarification of terms. As to 'intended purpose,' control of surface state is the top priority. Towing models or testing small wave power generation devices would be natural visitor activities. Observing differences between orbital distributions at depth is valuable, so I am also interested in being the ability to vary between a full-depth piston and bottom-hinged flap wave maker. I would also like the ability to generate oblique periodic waves or to experiment with adjustments to a straight wave front that might be made with multiple wavemakers responding to wall effects. It seems from videos I've seen of 'smaller' research tanks using arrays of discrete wavemakers, this is achievable. By 'smaller' I'm talking about something at this scale...

https://www.livescience.com/4206-wave-scientists-write-water.html

...but more like this configuration:
http://www4.edesign.co.uk/portfolio/bergen-towing-tank

Definitely a few notches in size, control, and instrumentation above this:


Again, something with a tank on order of 4'w x 2'd x 24-32' l, to allow for ~16" static water depth. Being able to top-load and move the wave maker mechanism(s) makes sense. When I prototype an exhibit, I prefer to have flexibility in swapping mechanisms in and out. Having to start out with a dry-back tank dependent upon a watertight seal would be a PITA as an initial condition. If wavemaker(s) can load in from the top, it makes it simpler to effectively change tank length, add a model towing rig, or switch from a fixed beach to another absorbing wavemaker at the far end.

Re wavemaker options:

1) I'm sure I could make a pleated fabric seal to fit between an array of 5-6 bottom-hinged flap-type wavemakers, and get that installed into a top-loading, ballasted box that eliminates back-wave issues. There will still be gaps between paddles to accommodate movement of the pleats. This option probably forgoes any adjustability toward a piston/shallow depth wave configuration. The pleat design would become very complex.

2) I could also see developing something like the 2-cam Salter/Edinburgh mechanism into a 5-6 wide wavemaker array. This would be a top-loading, wet-back configuration. One would have to accept volumetric disturbances near the bottom of the cam system, slight vertical translation of the face of the inside cam, and gaps between elements to allow for movement.

3) Finally, we come to the multiple piston array. Its a lot of individual units, but I can see fabricating the modules from plastic tube and plate, with o-ring sealed plastic pistons being driven via lead screws and small servo motors from the back side. One big manifold with lots of repeat parts. The rub would be the transition from 'round to square' for wave propagation if I can't deal with sealing square pistons in 'square cylinders' (nice oxymoron?). Would transitional diffuser duct, perhaps with a piece of flow straightener (honeycomb) at each square end help? Would the covering and uncovering of discrete ports as the water surface rises and falls make this approach a non-starter? Just thinking out loud...

I hope the intended use, feature list, and scale of the tank is now clear. If so, do you have any recommendations to the best option for the wave generator(s)?

 

Ok, but the one thing I don't get is why do you think you need pistons in cylinders at all? It is the differential horizontal displacement of the vertical wave maker "wall" that drives the orbitals. I can think of 4 or 5 ways to implement that with flexible materials and the 48 actuators you proposed above. Given the size and depth of your tank, we are not talking about multiple inches of differential deflection between each actuator. With a depth of 16", a ~1 sec, ~3" amplitude wave is about all you are going to get before shoaling occurs, unless you have no beach and set up a driven seiche like Calvin did.

 

Thank you for the critique. I considered semi-flexible paddles, using something like a sheet of HDPE or polycarbonate with actuators , and some sort of flexible membrane or pleat/bellows separator between paddles. With the criteria of avoiding back waves, I assumed I'd need a wet-back wavemaker or a top-loading unit that would contain leaks around paddles to a 'sump box' that could be easily drained without impacting the main tank. Gasketing between conformable paddles seems formidable. Anyway, I imagined a sequence of of 5-6 units like this across a 4' wide tank:

Coordination of the actuators would not be trivial, but that's a software problem!

Going in the direction of a 3D, adaptive surface is feasible. It might be like this ADAPA mold, with lower resolution and simpler construction:


I was hoping discrete pistons as individual 'paddles' would be an option primarily from a fabrication standpoint. Consider a bunch of round discs with O-rings inside off-the shelf tubing vs. attaching the same number of actuators to a surface that must combine qualities of both an elastic membrane and rigid panel. I've built many cylinders like the first option that for moving water and air at other exhibits and know they work! The second option seems like it adds another layer of complexity. But it sounds like generating waves from an array of discrete 'walls' as an approximation to a conformable surface is a non-starter.

BTW, from the standpoint of scale, this Omey Labs 'Teaching Flume' is close to what I'm looking at, although it is not as wide as I would like:
https://www.omeylabs.com/teaching-flumes

From their specifications:
Section: 600mm wide, 600mm wall height
External: 750mm wide, 1.5m high, 8.1m long (or any length)
Delivery: 670kg, in two crates, each 2.2m x 1.1m x 0.9m
Scale: 1:300 scale nominal
Wave maker: dry-back, hinged paddle:
• Peak; 0.1m Hmax, 0.5sec Tmin
• Dimensions; 400mm wide, 350mm stroke, 300mm deep.

Based on their videos, wave size and ability to control sea-state look good to me. From an exhibits and experimentation standpoint, building a tank with additional 'dead' length to accommodate wavemakers and beaches may be the best place to begin. From an ADA and accessibility standpoint, the top of their tank is too high.

 

Nominal 4'w x 2' d x 8' l tank module, to make efficient use of sheet material and aluminum stock lengths:




Each 8' module would be field assembled from 2 leg frame units, 1 bottom panel unit, and 2 glazed side panel units. Top rails would be supplied at 128" or 144" to stagger joints from module seams. Probably polycarbonate for the glazing, VHB taped to welded aluminum box tube side frames. Aluminum sheet for the bottom panel, VHB taped to welded aluminum box tube side frames. Leg frames units are a single box tube weldment. Tanks seams closed with marine sealant backed by silicone rod, secured by connector bolts.

The 5 largest longitudinal members would be post-tensioned through the end panels with stainless steel rods (not shown). I've used this method when I want to avoid weld distortion or ugly bolting flanges. Here it is on my CNC plasma cutter:

 

This example in case it is close to your purpose, with it they can show regular waves, soliton and breaking waves (see video) :
LEGI - UMR 5519 - Le canal à houle https://www.legi.grenoble-inp.fr/web/spip.php?article744&lang=en

 

Worth a watch
If it is a museum, perhaps an interactive arm to let the visitors operate a linkage mechanism to push the top of the panel out though there would be someone who might try to overdue it