Rotating Wing Mast – theoretical discussion

Are the fibres in a braided sock woven? If so, then doesnt this mean the fibres are not straight and therefore a substantial loss in mechanical properties?

 

Hi Groper - Yes the fibres are "crimped" due to the weave but the angle is very low and there will be a slight loss of properties due to this crimp. But there is a gain of toughness due the weave so depends on what you are looking for. Plus they are convenient if the cost is good. Cheers Peter S

 

Thank you very much Peter,

It is always about my recreational project, regarding the spars-system for A-Cat wing.

Weither to address, the torsion in some case, or the "bounding" of the UD or both, it was interesting to see what kind of stuff was available on the market for these 45° and/or 90° layers.

Rear upper tube can be so light that it seems that there is no alternative but socks used for glider or airplane model's wing.

Even for the main mast which bear all the compression load*, a standard sock (around 85mm/90/mm diameter) and 200g/square meter CF, one inside+one outside will represent ((60g+resin)x2)x 6 meters lenght, so already one kilo.

The targetted weight should be under 5 kg (target weight is not calculated, but guessed by analogy with a similar tube).

So it fits quite well the 30/70% ratio above-mentionned in this thread.

So remain around 4kg for UD.

For sure it's a bit far from FEA, but for a rookie in these matters, it is very important to have some benchmarks.

Thank you Peter, and everrybody; You make this thread a great education instrument.

Best Regards


*(but according to Hydros C-Cat engineers, for a wing the "dimensioning" parameter is not the compression or buckling, but flexion)

 

Hi Erwan - You could look at it like this. If you keep to the 70/30 UD/DB the E=80GPa as the CF lam fails at about 1% this means it fails in bending/comp/flex at about S=E.e 80000x0.01=800Mpa. Which is about right. So if you use conventional calcs this will work. The torsion is another matter. I'd have to figure that out a bit better. Peter

 

Woven socks ar3 inferior to a dbias stitched fabric. For ultimate performance or stength to weight ratios, woven fabrics have no place. A spar is an extreme example of this. How you fabricate determines what materials you will use, not only engineering theory. ..

 

Thank you very much Peter,

I ll try to dig in your calcul, it is like a corrected exercice for me and it will be a great help.

Groper, I guess with stiched dbias, you have to roll around the tube, and have some "recovery".

Also, f your tube is tappered according to some elliptical distribution of inertia, in case this kind of shape would be optimum, sock should be easier to use. Also depend if it's with female mold or mandrel

And (regardless of fabrication issue)with a double tappered solution, you can save area in the tappered ends and "re-invest" in UD or extra sock at some critical place. But of course all that is still a bit intuitive.

Thanks again and best regards everybody.

Erwan

 

Erwan - Perhaps if you make a homemade "winder" eg you make the mandrel and it is in a fixture which turns with a handle or a BBQ mortor. Then you guide the UD tape onto it by hand. You can get 1" and 2" UD tape. So use tape for 0deg and wind on for 45s. This will tight pack the layup. Peter

 

Low end filament winders uses a continous motorcycle chain attached to a driven and fixed sprocket. A fixed point is inserted in between a point in any of the chain sprocket space. This is the linear guide which moves back and forth by reversing the motor drive at the end of cycle. The length of the drive (or the stopping and reversing of the motor) is controlled by repositioning a microswitch on both ends.

Wind angle is controlled by the speed of rotation of the mandrel in relation to the linear speed of the filament feed. Either of the two can be controlled. The mandrel speed can be fixed and the linear "motor" varied or vice versa.

For smaller machines, a cross helix madrel is used with a guide roller in the groove. Similar to the mechanics of a bobbin winder. The disadvantage is the width of the linear travel is fixed.

 

Thanks Peter and RXComposites

In fact I already start to look for an old aluminium Serfiac windsurf mast for kids
it was a 3.4 m long, in one part, tapered all lenght and made of Zycral, so quite stiff and light enough, I guess after polishing, it should be perfect for a first experience with DIY mandrel/winder.

Regards,

EK

 

For those interested in DIY Carbon sock

Cheers

EK

 

Erwan, try to get or borrow from a library, "Filament Winding Composites Structure Fabrication" by ST Peters, WD humprey, & RF Foral. It is a very good book about filament winding.

I have posted the Excel spreadsheet on this subject a long time ago but said spreadsheet deals with vacuum/pressure vessels. The good part is it computes the tow thickness and the stress. For it to be usable for your application I need to adjust the load model. This is a cantelever model with diminishing load and a varying section using an elliptical cross section.

The most difficult part (for me) in making the spreadsheet would be that this is a combined load of bending and torsion of a thin walled tube for unstayed mast.

 

Thank you very much RXComposite, I ll try to get the book asap.

Actually it's not a cantilever "mast", the biconical tube as dreamed above is hold with 2 front stays, 2 lateral stays and 2 trapezes with 85 kg meat on the hook for righting moment.

Thank you, but to adjust the load model, I must be able to provide the good inputs, the "dimensioning" parameter, (not sure it's the appropriate vocabulary).

For those interested with wingsail, there is a point which seems interesting looking at the pics of the C-Cat wings, this is the "Arthur" which is the lever arm which control the twist of the first element almost at the bottom of the wing.
This lever arm is so short, that its full range of rotation angles remains within the thickness of the wing section, and nothing goes outside, but the control line on each side.
It is an interesting insight for the torsion loads which I still need to appraise more acuratly, before to start digging further.

Thanks and best regards

EK

 

If it has stay, it becomes more complicated as you have to deal with compression and this compression varies because no matter what, the cables are elastic. So it is three combined stresses at the same time.

As the mast gets thinner and wider to become aerodynamic, buckling will be a problem. The solution I see is to borrow from aerospace technology, re: Composite Wing with a spar and cored skin. Computation can be simplified as stresses can be isolated. This was mentioned by Eric Sponberg in post no. 5.

The main spar caps (in unidirectional fibers) take care of bending loads, the web takes care of shear loads, the "wing" skin takes care of torsional loads. Because this is a C spar, the aft spar (or main spar closeout) helps in the shear load. There are many spar sizing program. I have one written in Excel and the materials are carbon fiber uni with E glass biax. But this is a cantelever wing.

Parts can be fabricated separetely by infusion and assembled together by gluing,taping or tabbing. The low density core can be hot wired.

Note that i have inserted the "engineering constant" graph for modulus (you can find more explanation in filament winding book). Note that CF uni is very sensitive to off axis load and you have to pay extra attention to E glass and CF as it has very different strain characteristics.

 

Thanks RX,
Now I have a lot of stuff to chew.
Yes there are 3 different loads at the same time: bending+compression+torsion.

It's not a teardrop mast or wingmast, just a tube with circular sections and the larger section in the middle becomes thinner going to both ends.

Thanks again

Regards

EK

 

Hi RX - I caution the use of cores and webs in masts. I have not yet designed a mast that needed a shear web. They are needed in designs that use the web as a construction aid, eg tortoured plywood. Cores present a couple of problems. 1) They reduce the global inertia of the section. 2) Strains like electricity take the shortest/stiffest path in a structure. So the risk is that once a mast is bent the load travels down one skin only. This is called assymetric loading of the sandwich. This then fails the skin in say buckling or the core fails (called rolling shear). Even FEA can miss this failure mode. To correctly model for these modes of failure the core needs to be modelled as a solid and skins need to be modelled as plates tessellated to the solids. If the mast is just modeled using plates these modes are neglected. In a conventional mast stress are very very low. If the mast is stiff enough to take the sailing loads its stress is very low. If you are aiming at a very light mast the trick is to build the biggest section possible (to maximise stifness/rigidity) but have adequate buckling reserve. In metal masts this is usually accomplished by keeping the radius/thickness ratio at less thea 50. In composites its hard to put a number on it due to the variable quality laminates out there and the often misunderstanding about assymetric laminates and poor laminate stacking as discussed earlier. Cheers Peter s