Rotating Wing Mast – theoretical discussion

sigurd - do you think the mast will be lighter using a foam construction? Is this why you want to use foam? Peter

 

Yes, no aeroelastic modelling. I've been running a few different sail shapes. The leading edge of the sail seems to matter to get the high lifts. There are lots of different gap geometries that seem to work. I have found gap settings that have both high lift with a drafty sail and a deep drag bucket when the sail is flat.
What would be lovely, some experimental data on multi elements that I could reproduce in the software.

Yes I think bat cars are best. I figure some sort of spacer attached to each bat might keep the gap pretty constant.

 

I just have problems envisioning a 3:1 c/t top section, having two rather pointy ends, keeping its shape. Maybe it's no problem?

 

Sorry could you explain 3:1 C/t top a bit better? Peter you mean chord to thickness?

 

yep, that's it

 

Nope rob doesn't use foam in the mast, only in the booms and hull connective beams... yhe mast is a solid CF laminate...

 

Allright.

 

petereng, how does a client go about testing their mast laminate?

 

Hi Idkfa - I get the client to make a test panel 600x600mm and I send it to a local university for test. They cut it up into coupons and test for:
1) tensile strength and stiffness
2) flexural strength and stiffness
3) fibre ratio
4) compressive strength and stiffness
6) in plane shear strength and stiffness
7) interlaminar strength

I may not do all of these as this would cost about $2000AD. But half of them would cost say $600AUD which would get enough info for me to estimate the rest. Once the mechanicals are established then I can design a mast accurately. This may be uneconomical for a small mast but for a mast or a boat which will cost alot of money its well spent. Cheers Peter s

For a client doing a small mast who does not want to spend the money I get them to make a panel say 200mm by 1200mm. I then cure it properly and get them to clamp it onto a stiff bench. I then measure its deflection under gravity and hang a small weight on it and measure again. Then I calculate/estimate its bending stiffness. If its up to what I want I then say Ok for the client to make the mast and we then check the mast stiffness as well. Often amatuer builders do not acheive the target stiffness for the mast as they don;t have the vacuum bagging skills, the attention to detail about using as less resin as possible etc. Its easy to make a small panel but to make a big structure to spec takes experience. Infusion is levelling this problem with most infusion laminates having consistent high properties. Cheers Peter S

 

Just arrived. Interesting thread. I have used a number of engineers for masts and have found a lot of variation in weights, laminates and layouts. Not sure what this indicates but none of them have so far failed. The biggest differences are 1) the accuracy of the static (and presumably the dynamic ones as well) bend numbers. Currently, we are within a couple of mm and the masts are lighter than previous ones so pretty happy. 2) The ease with which you can communicate with the engineer, particularly when you have a different idea to what he is advocating. Find one who will listen, explain and if it makes sens, change and you have a very rare bird.

The mast in this boat http://www.youtube.com/watch?v=8chR6DAFjGA weighs 120 kgs, is designed for 10% tip lay off in 30 knots of breeze. More than that and the sail shape will be pretty awful.

I use foam in wing masts, not in tubes. Also sheer webs in some wings. Usually a bulkhead at high load areas, but we have also used extra laminate (including 90's) instead. It is partiucularly useful on masts with a straight taper (easy to build) where the extra laminate reverses the taper to make it easy to install.

The telescoping wing mast for this boat http://www.youtube.com/watch?v=DfhdRfBTt8o has a sheer web in the bottom section, none in the top. It does have an external flange at the overlap. The lengths were 8m and 9.5m, rm 4 ton metres, mast weight 66 kgs chord 600, thickness 200. I have a half size version of the lower mast in my 7.5m test boat. It was tied up to a jetty in 50 knots (probably saw less than this, but very swirly) and the boat would have capsized if both hulls were not tied down. No way could the mast have reacted wuickly enough to feather into the wind, particularly as it was from different directions at the top and bottom. Partly due to this, and partly due to the cost, we have decided to put a telescoping tube mast on the 15m in the video. Unlike the wing, this will be a conventional top mast smaller diameter than bottom mast arrangement. Not only safer and cheaper, but if it doesn't work, the two pieces can be glued together to make a conventional one piece mast.
Each section is 12m long, so it should be an interesting rig on such a light boat. Sorting the sail hoisting was a challenge, but if it works, it should enable multiple sections to telescope, which opens up all sorts of possibilities.


rob

 

laminate buckle and stacking sequence

Hi All - I have just run the buckling model for everyones enjoyment. I built a flat small column just like a timber ruler. I applied 1kg on the end of it and calculated the buckling load. The section is 3.6mm thick by 30mm wide x300mm long. I built two laminates one (DB/UD/DB) 600gDB/2400gUD/600gDB) the other (1200gUD/1200gDB/1200gUD) The ply properties are as a normal std modulus (prepreg or infusion) CF 300g ply. E1=110000Mpa, E2=7000Mpa, G=4000Mpa nu=0.3 t=0.3mm. The laminate properties are therefore Ex=78777Mpa, Ey=14956Mpa and G=12120MPa t=3.6mm. Both laminates have the same in-plane stiffness data. I also made a panel that had the same stiffness isotropic with E=78777Mpa as a reference. Using Eulers equation we can predict the first column buckle using eulers equation for the ficticious material (a stiff aluminium say as Al E=70000Mpa)which is 102.7kgf. In the FEA the "metal" buckles at 102.8kgf so our benchmark is proved by calculation and by FEA. Now for the interesting part. As I have used software that uses classic laminate theory and can correctly couple all the plies, The DB/UD/BD buckles at 40.9kgf and the UD/DB/UD buckles at 74.0kgf. So which laminate do you want to use? I'll leave that for you to think about. This is easily replicated by someone who has the knowledge and software. So this tells me to put the UD on the outside and you will double your buckling capacity. Next we have to consider geometry. Now for you thinkers out there why did the CF buckle at a lower load then the isotropic material even though the longitudinal stiffness is the same?? Answer - the shear stiffness and flexure stiffness of the CF is less then the isotropic material hence the buckling load is less. When I design a laminate for whatever use I try to get the flexure stiffness and tensile/compression stiffness the same, this usually means putting UDs on outside of laminate. Cheers Peter S

I have convinced a couple of clients to use sheet aluminium for their mast vs CF. Most find it hard to get a CF laminate past 70GPa. The winders are at about 90GPa and can get a bit more. Infusion acheives thery so you can get from 70-110 easy. But to keep it short. If you are doing a wet bag layup its unlikely you will get to aluminiums stiffmness. Since al is 2700kg/m3 and Cf is about 1600kg/m3 if you dont get over 80GPa in CF you may as well make it in Al and it will be nearly as light but no mould needed. Just some basic sheet metal tools, aircraft rivets and your away. Theres heaps of info on how to make aluminium plane wings on the net and its the same process. Cheers again

 

i see...

This also seems intuitive with regard to making a sandwich panel with low modulus material in the center and high modulus material on the skins...

The reasoning ive always read, about encapsulating the UD fibres with the -+45deg fibres, was to keep the UD fibres in column and stop them from breaking out of the laminate when stressed - presumably by the same buckling stresses your talking about.

How can you justify whether the resin matrix is enough - say with a good epoxy for example - to resist these UD plys from breaking out of the laminate, when no -+45fibres are wrapped around the outside of them. Can you show us how to calculate that?

 

With a mast laminate I usually spec a glass or CF plain weave on the outside. Fibres don't usually "burst" unless the laminate is at flexural breaking load. Over the years I've had lots of laminates tested and the ones with UD on the outside regulary test much stronger then the other way round. With a composite mast if it buckles it usully springs back no damage. Infused glass and Cf laminates fail at over 1000Mpa stress in flexure. It takes quite a load to do this!! Using the FE program I can predict when the ply stress is at its ultimate stress hence at failure. By the way the 45s on the outside will burst well before the UDs on the outside as the resin does the work. Testing of DB glass for instance puts it at about 200MPa tops and UD at over 1000Mpa so the DB fails much lower the the UD in tension, compression and flexure. To be precise Groper I can do a ply by ply analysis and predict when each ply will fail. I think you are interpreting the fibres as ropes but this is incorrect once you hold it all together with resin. It stops being rope and becomes a laminate with different properties to a rope.

 

That doesnt sit quite right with me for a few reasons... as you know, i do alot of work with infused panels, i often fold them into rather tight curvature by removing the laminate from one side of the sandwich or just bending a solid glass laminate. When the fibres are at 45degrees to the bend axis, the laminate behaves nicely and there is little noticable stress when its bent. If i try to bend it at right angles to the fibres within the laminate, there is notable stress, and if too tight a radius is bent, i can see the fibres starting to turn white and what looks like stress cracks appearing in the laminate.

So, if we start to bend a mast with UD fibres in it, from what ive seen bending infused laminates, these UD fibres will fail long before the 45degrees fibres will. The 45deg fibres do not take the bending stress as much as the UD fibres as they are not working in the stress direction.

Another question, why do i always see a heavy UD laminate, interleaved with 45degree fibres every ~2mm or 2000gsm of UD laminate thickness - when used in a tension or compression flange of a beam? Again, i always thought it was to hold the laminate together and prevent it from "bursting" as you seem to have termed it?

 

Ok Groper - 1) When you bend a laminate you stress it. The stress is a function of the stiffness and the radius. If the laminate was a piece of rubber you could bend it more then if it was a piece of timber. Same as 45 deg laminates vs UD laminates. You are assuming that the stress is the same in each laminate. This is incorrect.

Re "interleaving" there are a couple of rules of thumb about having 45s every so often in a laminate. Lloyds are prescriptive about this. But from a theoretical point of view if a laminate is in bending there is a shear stress across the thickness. This shear needs to be resisted by the laminate by interlaminate shear strength. But the real reason is that applied stresses are rarely perfectly along the fibes. So the 45s allow the strains to be distributed into the Uds. Laminate design can get complicated but a good laminate will be as interleaved as possible and try not to have more than 45degs between fibres. EG the DB is not ideal to use inside the UD stack its better to have UD/45/UD/-45/UD. or say if I have a thick UD stack instead of stacking it all at 0deg I'd stack it at +/-5degs per layer to improve its toughness. Cheers Peter s

Hi Groper,
I looked up some tst results. Infused UD glass 740MPa flexure E=30GPa, CSM 338MPa E=15GPa DB would be less off axis ie 0 or 90deg.