Carbon rudder stock: what about torsion?

[Stingrig]

Hi,
My first post on this very informative forum. The coming years I will enjoy building a 13 metre monohull sailboat. I have started with the small parts. Carbon boom and plywood folding dinghy are finished. At the moment I'm building the carbon fibre spade rudder.

The rudder stock has plenty of UD carbon fibre to deal with the expected bending moment. I can do those calculations, at least I think I can ....

My problem is dealing with the torque. The guys assisting me with the design also have a little difficulty translating torque in a laminate schedule. Their recommendation is based on experience. That's alright but I want to better understand what I'm doing.

I have already added a carbon biax 45/45 laminate. It's time to put the rudder stock in the blade and I am getting cold feet. Is it enough??? The biax is completely on the outside (unusual I know, but that's the way it worked out for me).

Some data:
-I want the rudder stock to be able to handle a torque of at least 2000 Nm;
-the dimensions of the rudder stock are approximately 11 x 9 cm around the lower bearing, it's BIG;
-the rudder stock is not tapered above the lower bearing, that should reduce stress (as a result from torque) on the laminate between lower bearing and tiller attachment;
-see pictures.

Any thoughts on determining the rudder stock laminate to handle the torque? I can add more biax if necessary.

Thanks,
Stingrig

 

[JamesG123]

You can get the spec sheet from whomever made the stock, but the only real way to know for that particular tube would be to take a test peice of sufficient length (30cm or so?) and create a test jig to see how much torque it can take. This could be as simple as one end clamped into a vice and then putting something square into the other end with a socket that can accept a torque wrench and then twisting on it until you either reach your desired torsion or it fails, which ever happens first.

 

[Mr Efficiency]

I don't understand why you'd need such high values for torque, in a 13 metre boat, where are you plucking 2000NM from ?

 

[Stingrig]

There's no spec sheet because I made the stock myself from scratch, around a wooden core. More than 10 mm UD carbon (port/starboard, less on front/back) to resist the bending moment and (for now?) a couple of mm 45/45 biax to resist the torque. Rectangular section 11 x 9 cm (where it counts).

The test piece crossed my mind. Thanks for reminding me! I could simply saw off a piece from the top and use that. I made it a little too long. Testing is ultimate reassurance. However, I'm still interested in something of a theoretical basis (for me that's half the fun of DIY).

 

[Stingrig]

I don't understand why you'd need such high values for torque, in a 13 metre boat, where are you plucking 2000NM from ?

Thanks. I am not really sure if 2000 Nm is too high. The calculation in Principles of Yacht Design (Larsson, older version of the book) for my rudder geometry gave me 1800 Nm.

Another approach:
Tiller length: 1.25 metre
Force on tiller: say 1000 Newton (when I get angry)
1000 N x 1.25 m = 1250 Nm

Both numbers could be way too high. In fact I hope they are but this is where I'm now ....

 

[JamesG123]

However, I'm still interested in something of a theoretical basis (for me that's half the fun of DIY).

Its not particularly DIY. The modelling is very complex involving the tension strength of the fibers, the sheer strength of the resin, all on a modulus of the outer layers first and working inward since that is where stresses accumulate. Crossed with that you made it square section as well. People specialize and spend entire careers doing this kind of analysis. And they spend a lot of time breaking test samples with torque arms.

 

[Stingrig]

Its not particularly DIY. The modelling is very complex involving the tension strength of the fibers, the sheer strength of the resin, all on a modulus of the outer layers first and working inward since that is where stresses accumulate. Crossed with that you made it square section as well. People specialize and spend entire careers doing this kind of analysis. And they spend a lot of time breaking test samples with torque arms.

Thank you, makes sense. I probably shouldn't go down this rabbit hole. My google searches confirm your remark about the rectangular section in relation to calculations for torsion. In relation to bending however I'm very glad it is rectangular (and big)! I will probably join the other mortals and destroy a test piece. Thanks again for your feedback.