Designing a MCT.

[Andert]

First, I am new at this forum and if this is in the wrong place I am Sorry.

I am work with building a proof of concept for a marine current powerpland. And right now I have come to the design of the turbine.

An overview of the device so far:

http://www.andert.se/eng/over.jpg

http://www.andert.se/eng/exp1.jpg

Size:

http://www.andert.se/eng/matt.jpg

Inserts

http://www.andert.se/eng/exparm.jpg

http://www.andert.se/eng/expblade.jpg

I am a mechanical enigineer but the composite area is new to me. I have read a couple of books.

I am working in SW2010 and use the shell composite simulation
Forces and result

http://www.andert.se/eng/sim1.jpg

http://www.andert.se/eng/sim2.jpg

I am using data from Fiber Composite Analysis and Design:Composite Materials and Laminates Vol 1. Unidircectional 0.6 vf.
E-glas
Ex:45GPa
Ey,Ez:12GPa
G:4.4GPa
Vlt:0.25

T-300
Ex:132GPa
Ey,Ez:10GPa
G:6.5GPa
Vlt:0.25

Glass for the arms and Carbon for the wings.

Carbonfiber and stainless steel, can it be galvanically isolated with a layer of glassfiber?
How will water ingress in the structure and can it be avoided?

Can i combine glass and Carbon fiber in the same laminate?
What is best to use vowe or unidircetional?,

what stacking and angles sequence is best? How thick should i choose the different layes in the simulation?
How many layer are feasible to have?

I am also alooking for generell feed back, is It possible to make moldins this way?, is there a easier and better way to connect the blades?

 

[RPstress]

From over.jpg you appear to be designing a vertical axis turbine. I guess that can be used on a yacht to provide wind power in a marine context, or is the 'marine current' a river or tidal flow or similar?

Your quoted material properties are about right for undirectional glass and carbon fibre polymer matrix composite (poisson's ratios look a touch low).

As to your queries, some are a bit general. However:

- Carbon can be isolated from material that it might galvanically corrode with a layer or two of glass. However, good stainless steel (such as A-286 or 300-series) will probably not need such protection. (Watch out for other aspects of salt-water corrosion with stainless steel, such as crevice corrosion or stress corrosion cracking.)

- Protecting the composite from moisture ingress is not really practical in a marine environment. Just using allowable strengths that allow for the presence of moisture should allow for ingress.

- Carbon and glass can be combined to make a so-called 'hybrid laminate.' However, usually this is done becasue a purely glass or purely carbon laminate lacks some combination of properties wanted for the design.

- The layup (the combination of fibre directions) is a matter of the desired laminate stiffnesses and strengths. For initial designs and trials a 'quasi-isotropic' layup might be advisable. Optimising it can be done in due course. You should choose an approximate thickness based on the usual methods of sizing structures to resist the forces upon them. The number of plies is then the number needed to make that thickness. From four or five plies to several hundred is quite possible.

- Making the general shapes shown would be quite practical. Almost all composites are made by molding of some sort. The manufacturing method is up to the designer (you!).

- The way the blades are connected is not really depicted in any of the available jpegs. NB: 'expblade' and below give page not found.

For more information on marine composites try the 'usual' recommendation of

http://www.marinecomposites.com/

.

Finally, if this is not a wind turbine, i.e., if it is submerged, metal rather than composite may be a better bet to get adequate strength and stiffness with low volume (and drag in water).

 

[Andert]

Thank you for the reply!

It is a river flow unit, vertical axis. It will be positioned in a river,(sweet water).

The reason i might want to combine carbon and glass in the blades is that a collision with debrie might crack the carbon.

The reason I do not wish to go with metal is the weight and production cost. The weight will lead to difficult handeling and extremely low eigenfrequencys.

A "quasi-isotropic" sounds like a plan and as you pointed out I am not looking to optimize materialus just to get it strong enough.

I fixed the pictures as well. Please have a look!

I am greatfull for all comments and ideas!
I will check out Marine composites.

 

[RPstress]

If you're worried about cost (and who isn't?) but not weight then dare I say carbon is a questionable way to go? Maybe the vibrational characteristics make stiffness/weight a design driver, in which case carbon fibre makes sense.

Also, if large impacts are a worry then laminated composite has some 'interesting' downsides. However, GE's carbon turbofan fanblades seem to work well.

It will all depend on your design objectives and how they can be realised with different material choices.

A laminate stiff enough and strong enough is going to be a bit thicker and heavier in a hybrid of glass and carbon, and if it's thick enough to take the hydrodynamic loads and it's adequate for the impacts encountered, then well and good. I.e., you may be better off with a hybrid whereas pure carbon would need to be much more costly for the impacts. Quantifying that sort of impact thing is rather hard, though, even with testing.

NB: the proof-of-concept marine current turbine at Strangford Lough in Northern Ireland seems to be fairly conventional steel, judging by its appearance when I saw it on TV. (That's a two-rotor horizontal axis design.) I'm guessing that each of its rotors is about 30 or 40 ft in diameter. I'm sure its details are on the Internet somewhere.

Zooming in on expblade.jpg it looks like metallic fittings to spread the load from the attachment arms into the blade laminate. As a concept this seems fair and might be able to be made strong enough. On the other hand it all looks a bit undercooked and flimsy to my eyes. NB: can't really see any detail on the FEA pics, but that 20 kN load (two tons) is not to be sneezed at.

Roughly how big is this? That may affect your production choices.

 

[Andert]

It is quite big. The size in mm is found in

http://www.andert.se/eng/matt.jpg.

The diamter of the turbine will be about 6m and the height 3.5m.

Large impacts I am not worried about, but say small tree branches or so. Since the turbineblade will be moving about 3 times as fast as the water i figured i would put some glass in the front of it like this:

http://www.andert.se/eng/nacalaminat.jpg

Should i use a corematerial?, which?

I figure that this profile can be made either hand layed up devided in two parts and the glued or use pulltrusion to get it solid (and change the way of attaching the blades and arms). Is there other ways and which would be the best way to go?

And yes! It is a bit undercooked!, I am just starting to looking in to it as I have just finished the generator part of the device and orderd it from a mechanical workshop. I will start working on this full time the week after the next, after finishing the foundation.

20kN is alot!. I just found out from our hydrodynamics guy that the load on the arms is somewhere 3-4kN not 600N and 400N. So they will probably also have to be made out of carbon.

20kN on a solid profile Should give 70-80 MPa using a 10mm laminate should give about 80-100MPa, which should be manageble with a good laminate.

And and 20kN is momentary overload so it does nog have to withstand fatigue at this load.(~~2m/s flow velocity in the river and we will shut down the device then)

The Strangford turbins is encountering much higher flow velocity(4-5m/s mine 0.5-1,7m/s), therefor it can be made smaller and thicker more like a propeller, however they have had their blades broken off. In norway they have a tidal mill with glass/carbonfiber blades, Hammerfest strom. They have hade problems with material fatigue due to the currents induced by the waves in the ocean.

Much appriciate the answers!

 

[RPstress]

That 'Hammerfest strom' tidal mill sounds interesting. And I hadn't heard that the one at Strangford had had a bit of blade trouble. At a guess that will have involved fatigue and maybe some unforseen interactions between the blades and seabed or flow around the tower. Even if the Hammerfest installation has had problems it's reassuring that someone has made a composite tidal mill work.

Most hybrid laminates combining carbon and glass either use cloth with a carbon warp and a glass weft (or vice versa) or alternate layers of glass and carbon. Other hybrids such as tape-laid (or 'fibre-placed') material where adjacent tapes are different materials are possible, though I'm not aware of any work being done on such.

I think that the sort of hybrid in your 'nacalaminat' picture is not much used, though GE's fanblade has a titanium leading edge for a similar effect (composite helicopter rotor blades also often have a titanium leading edge, though much thinner than GE's fanblades have). I don't know if using glass in this way will work well or not. Perhaps someone else has relevant experience or an opinion.

NB: with a turbine the incidence of the blades to the flow may mean that protection as illustrated is not quite right.

As to cores, there are a number of marine cores available (such as those from Gurit and DIAB) which might work reasonably well. The sorts of stiff, brittle foams used in Aerospace are probably not appropriate to such a damage-prone environment. A honeycomb will have better properties than all but the heaviest foams, but is not really appropriate to a submerged structure. Note that any core will be weaker (but cheaper) than a solid composite. (GE fanblades use no core to maximise aerodynamic efficiency and bird-impact damage resistance.)

Your allowable stresses are probably appropriate to conservative initial sizing. I would expect final wet, damaged laminate ultimate allowables for carbon to be a bit higher. I don't really have experience in designing with glass to comment.

A carbon laminate should also have a fatigue allowable that is higher than your quoted stresses. Again, I cannot comment about glass.

 

[Andert]

The reason for wanting to have the impact area in the front of the blade is that a side collision would be more prone to bounche and have a less hard impact.

I will be back with more "cooked" simulations and so on.

Thank you for replies!