Jim Caldwell
Member
No rod or tube lengthwise support?
new stuff in the a-class
- Thread starter Lars Schrøder d 13
- Start date
juniordave nz
Member
Much more efficient to put the structure in the shape, rather than use something up the middle. They have a skin thickness of about 1.5mm. They're solid as.
I'm going with pretty much the same layup for the rudders, but I will add a bit of internal structure to them as the span is a lot greater. I haven't bothered with any internal structure in the winglets.
I'm going with pretty much the same layup for the rudders, but I will add a bit of internal structure to them as the span is a lot greater. I haven't bothered with any internal structure in the winglets.
F18 Sailor
Super Anarchist
Unbalanced layers is your number one reason, as 99% of carbon comes in a balanced fabric (unlike fiberglass, which you have to watch for!). Also, I'd put the 45-45 layers on the outside when you make the next set. Honestly I'd just do 1 layer of 45-45 on the outside, 3mm rohacell, then another layer of 45-45 on the inside and then a ~1" wide 10oz strip of uni-carbon on the very inside of each halve. Join with a 100 psi foam shear web wrapped with a 45-45 5.7oz carbon. Or use a straw and a carbon sock for the spar, or a carbon rod, bonded with epoxy splooge and foaming agent.
I'm afraid with no core at all you're looking at failure, but I assume you have some sort of shear web? Personally on parts this small and loads this high, solid core with 100 psi XPS or equivalent Rohacell (IGF71) would be my first choice. Hollow molded can be risky...
Finally, it's easier to polish the molds once then every part you pull out of them...you can shoot clear coat in the mold and have very little finish work when done. Just my opinion, since I suspect you'll build more than one set for testing!
Overall it's neat to see the homebuilt progress with no CNC, and the parts are pretty good looking!
I'm afraid with no core at all you're looking at failure, but I assume you have some sort of shear web? Personally on parts this small and loads this high, solid core with 100 psi XPS or equivalent Rohacell (IGF71) would be my first choice. Hollow molded can be risky...
Finally, it's easier to polish the molds once then every part you pull out of them...you can shoot clear coat in the mold and have very little finish work when done. Just my opinion, since I suspect you'll build more than one set for testing!
Overall it's neat to see the homebuilt progress with no CNC, and the parts are pretty good looking!
I may be able to contribute to this home build exercise since over a 6 year peirod I home built a heap of moth foils and eventually gave up and bought a Mach2.
I agree that making a good mold makes better products more easily, but unless you have the design right to start with the effort might be wasted if you decide to throw the mold away. Since I have no idea what ideal ACat rudders look like I will not comment on the design.
Making thick skins for structure is OK except at high sheer load areas, like the T joint and just below the hull or rudder box. Here you need solid core for sheer strength and if you discontinue it it needs to be tapered. I have used solid oak cores, balsa cores, and strips of pultruded carbon (available from CST Composites), which works best. You may think that your shells are thick at 1.5mm but that is no where thick enough for the high load areas. Moth foils need to be about 3mm, mostly unis but with a couple of diagonal layers of square. 1.5mm might be OK for the horizontals if you make the T joint incorporate a full span spar.
Making the horizontal in two parts is asking for trouble. It would be much safer to make the bottom skin continuous and thick to contribute heaps to the bending strength.
My most successful T joints were molded separately in special molds. Each branch was tapered from 60x8mm (Shaped to fit inside the skins) to 20x4mm over about 100mm in all three directions. The tapered thickness is so that it can fit into tapered slots in the spars of each component. These Ts were glued into the horizontal when the top and bottom shells were mated. But the vertical was assembled with the slot pre prepared, including a lot of hoop fibre around the socket. I tried a few with the horizontal and T removable from the vertical but never got it strong enough so I recommend it all be glued together.
In all the years I spent maybe enough to have bought 5 sets of factory made foils, I had fun, learnt a lot, but wasted heaps. I am happy now with my Mach2 foils and when I build a new moth I will not be trying to make foils again. So have fun, good luck but do not expect to be successful first go.
I agree that making a good mold makes better products more easily, but unless you have the design right to start with the effort might be wasted if you decide to throw the mold away. Since I have no idea what ideal ACat rudders look like I will not comment on the design.
Making thick skins for structure is OK except at high sheer load areas, like the T joint and just below the hull or rudder box. Here you need solid core for sheer strength and if you discontinue it it needs to be tapered. I have used solid oak cores, balsa cores, and strips of pultruded carbon (available from CST Composites), which works best. You may think that your shells are thick at 1.5mm but that is no where thick enough for the high load areas. Moth foils need to be about 3mm, mostly unis but with a couple of diagonal layers of square. 1.5mm might be OK for the horizontals if you make the T joint incorporate a full span spar.
Making the horizontal in two parts is asking for trouble. It would be much safer to make the bottom skin continuous and thick to contribute heaps to the bending strength.
My most successful T joints were molded separately in special molds. Each branch was tapered from 60x8mm (Shaped to fit inside the skins) to 20x4mm over about 100mm in all three directions. The tapered thickness is so that it can fit into tapered slots in the spars of each component. These Ts were glued into the horizontal when the top and bottom shells were mated. But the vertical was assembled with the slot pre prepared, including a lot of hoop fibre around the socket. I tried a few with the horizontal and T removable from the vertical but never got it strong enough so I recommend it all be glued together.
In all the years I spent maybe enough to have bought 5 sets of factory made foils, I had fun, learnt a lot, but wasted heaps. I am happy now with my Mach2 foils and when I build a new moth I will not be trying to make foils again. So have fun, good luck but do not expect to be successful first go.
juniordave nz
Member
Thanks for your feedback guys. I'm not planning on them being hollow foils. I will fill them up with foam after glueing, the rudder I may do before glueing.
The winglet is only an L foil not a T, that's why they are separate. I will build an angle bracket out of uni carbon and CSM core that will go inside the winglet and the rudder and will glue in. Hopefully that will be strong enough. I would say my foils are over built, but we will see.
This is the first time I've made parts out of female molds. I've done a lot of laying up over parts including a couple of surfboards and other boat parts and I used to work for a boat builder building high performance boats. I've also studied composites, so it's quite interesting actually putting that into practice.
I agree with your comments on the quality of molds. I would like to make them better, but am trying to get these out as soon as pos to test them. Because if they work I will use them for the worlds. But I am still waiting on my boat to get here right at the end of the year.
The winglets are now glued together. They are solid. I cooked them for a few hours at about 70degrees after glueing.
The winglet is only an L foil not a T, that's why they are separate. I will build an angle bracket out of uni carbon and CSM core that will go inside the winglet and the rudder and will glue in. Hopefully that will be strong enough. I would say my foils are over built, but we will see.
This is the first time I've made parts out of female molds. I've done a lot of laying up over parts including a couple of surfboards and other boat parts and I used to work for a boat builder building high performance boats. I've also studied composites, so it's quite interesting actually putting that into practice.
I agree with your comments on the quality of molds. I would like to make them better, but am trying to get these out as soon as pos to test them. Because if they work I will use them for the worlds. But I am still waiting on my boat to get here right at the end of the year.
The winglets are now glued together. They are solid. I cooked them for a few hours at about 70degrees after glueing.
juniordave nz
Member
Paradox isn't experiencing any weird helm feedback due to their L rudders. And I don't think you would, it might change the balance ever so slightly when one lifts out. But no real difference to a rudder lifting out under normal conditions. You may get some leeway force from the winglet, but this can be cancelled out by the rudder, so you shouldn't feel it.
Also I have set the length of the rudders so that both winglets should always be in the water, unless at high heel angles. The first reason I want these winglets is to be able to drive hard downwind in the steep chop we get in our harbour and I'm putting them on an ASG3 which has very fine bows. I am also hoping they will act like a flatter stern and allow me to trap downwind. In these cases I don't want the rudder winglets coming out of the water, remember what happened to ETNZ when the winglets let go....
The second reason I want winglets is to develop a set of J's or C's or L's for the dagger boards and start to get the boat foiling or semi-foiling or whatever seems to come out the fastest. For that bit of work I have a CNC on hand for and a full suite of ANSYS to test in.
Also Samc99us, I like that idea of putting the 45's on the outside instead of the 90's. makes sense. I'm going to stick with my current layup for this set, as I want them bomb proof (hope I'm not wrong), but I will look into other layups in the future.
Regarding the balanced fabric's. I'm really not keen on using weaves where you need high performance. There is a lot of waste with them as you end up with too much off axis carbon and you can't really taylor the layup very well. Stitched double bias is nice though. Makes the 45's a lot easier.
Also once I glue everything together, it becomes a balance laminate, just with a void/foam/whatever in the middle.
Also I have set the length of the rudders so that both winglets should always be in the water, unless at high heel angles. The first reason I want these winglets is to be able to drive hard downwind in the steep chop we get in our harbour and I'm putting them on an ASG3 which has very fine bows. I am also hoping they will act like a flatter stern and allow me to trap downwind. In these cases I don't want the rudder winglets coming out of the water, remember what happened to ETNZ when the winglets let go....
The second reason I want winglets is to develop a set of J's or C's or L's for the dagger boards and start to get the boat foiling or semi-foiling or whatever seems to come out the fastest. For that bit of work I have a CNC on hand for and a full suite of ANSYS to test in.
Also Samc99us, I like that idea of putting the 45's on the outside instead of the 90's. makes sense. I'm going to stick with my current layup for this set, as I want them bomb proof (hope I'm not wrong), but I will look into other layups in the future.
Regarding the balanced fabric's. I'm really not keen on using weaves where you need high performance. There is a lot of waste with them as you end up with too much off axis carbon and you can't really taylor the layup very well. Stitched double bias is nice though. Makes the 45's a lot easier.
Also once I glue everything together, it becomes a balance laminate, just with a void/foam/whatever in the middle.
from my limited winglet experience your hope is well founded. i used 1-2 degrees up on the leading edge of the winglet and got enough lift to support my weight.
but not on an Acat; it should work the same on all non-foilers i think.
but not on an Acat; it should work the same on all non-foilers i think.
juniordave nz
Member
Last photos of a winglet.
Going to go finish off the rudder mold's now. Hopefully lay them up over the next day or so. Then hopefully collect my boat on the 31st.
Nothing else to do round here at mo, just keeps raining.
Going to go finish off the rudder mold's now. Hopefully lay them up over the next day or so. Then hopefully collect my boat on the 31st.
Nothing else to do round here at mo, just keeps raining.
juniordave nz
Member
I would think that would be judged illegal, as you are not inserting the foil (in it's entirety) from above. Also a C-L foil would likely go past the 0.75m rule, which dictates how close the tip of the foil can be to the centreline of the platform
Sailingkid
Super Anarchist
There isn't anything weird going on with Ls, they just feel like regular (non foiled) rudders IMHO.
juniordave nz
Member
^^ Thanks for the feedback. Hope the Nats are going well for you.
Half a rudder done, another half to go. I had to order some more carbon to build the other rudder, so that will have to wait a week or so.
Carbon pre cut and ready for laying up, sitting on a coffee table that's taking me ages to finish.
Mold all waxed up and ready to go.
Under vacuum. I let the resin flow for about an hour before putting a couple of electric blankets on top with insulation to cook them. Gets them pretty hot.
I got bored yesterday so started to build a CFD model of the rudder to try and work out the optimum angle for the winglet to rudder. I hate building the computer models, but the results should be interesting. Will post them here when I've finished.
Half a rudder done, another half to go. I had to order some more carbon to build the other rudder, so that will have to wait a week or so.
Carbon pre cut and ready for laying up, sitting on a coffee table that's taking me ages to finish.
Mold all waxed up and ready to go.
Under vacuum. I let the resin flow for about an hour before putting a couple of electric blankets on top with insulation to cook them. Gets them pretty hot.
I got bored yesterday so started to build a CFD model of the rudder to try and work out the optimum angle for the winglet to rudder. I hate building the computer models, but the results should be interesting. Will post them here when I've finished.
Unless there is a known angle of rudder rake for your boat, i would think it would be a good idea to sail the boat with the new rudders sans winglets in order to test the helm first.
If for example you had a touch of lee or weather helm that you wanted to fine tune; changing the rudder rake would change the angle of the winglet.
Conversely, if rudder rake doesn't effect the helm much then you can play with the angle of the winglet by changing the rudder rake. Real sailing time beats a computer simulation for that.
Your angle it will be either 0 degrees or "up a little" in the front. your weight placement will do the rest. the definition of "a little" is best determined by trial and error.
If for example you had a touch of lee or weather helm that you wanted to fine tune; changing the rudder rake would change the angle of the winglet.
Conversely, if rudder rake doesn't effect the helm much then you can play with the angle of the winglet by changing the rudder rake. Real sailing time beats a computer simulation for that.
Your angle it will be either 0 degrees or "up a little" in the front. your weight placement will do the rest. the definition of "a little" is best determined by trial and error.
juniordave nz
Member
Good points. I will sail with the rudders without winglets if I have time.
Modelling is for the other angle relative to vertical, tip down or tip up. I want to look at the difference in drag and the difference it has with regard to lift.
Modelling is for the other angle relative to vertical, tip down or tip up. I want to look at the difference in drag and the difference it has with regard to lift.
juniordave nz
Member
Boats sitting in a container awaiting clearance, so still building foils.
Testing flow directions before adjusting angle of foil. This is effectively bow up mode and therefore winglet is lifting, in this case about 10kg of support.
Testing flow directions before adjusting angle of foil. This is effectively bow up mode and therefore winglet is lifting, in this case about 10kg of support.
juniordave nz
Member
Lots of modelling going on. Getting some interesting results.
Here are some images of a 5 degree horizontal flow (effectively leeway, although a bit to much as the rudder is producing 600N of side force). The images are of the leeward rudder. The upper boundary loosely models the surface interface by allowing water to move through it. This shows how the effectiveness of the rudder diminishes the closer it is to the surface without a endplate.
The other interesting thing is that the L and similar would happen with a T, is producing lift due to the join between the wing and rudder, this is pronounced in this flow but happens even when the flow is perpendicular to both rudder and wing.
Front view
Looking from windward
Looking from leeward
Here are some images of a 5 degree horizontal flow (effectively leeway, although a bit to much as the rudder is producing 600N of side force). The images are of the leeward rudder. The upper boundary loosely models the surface interface by allowing water to move through it. This shows how the effectiveness of the rudder diminishes the closer it is to the surface without a endplate.
The other interesting thing is that the L and similar would happen with a T, is producing lift due to the join between the wing and rudder, this is pronounced in this flow but happens even when the flow is perpendicular to both rudder and wing.
Front view
Looking from windward
Looking from leeward
Last edited by a moderator:
juniordave nz
Member
Looking down at wing
Looking up at wing
juniordave nz
Member
I need to finish sorting out the numbers from the CFD runs, but it looks like 90 degrees is the most efficient. The models gave me 4.5GB of data and I only want something like 20 bytes. There are ways I could have set it up to make getting the numbers out easier, but I worked that out after I ran them. Learning for next time.
In the mean time. First rudder almost finished. Next one about to be built.
And I now have a boat to sling them off.
a bit of eps holding the winglet on, that'll do, following the flexy foils trend...
In the mean time. First rudder almost finished. Next one about to be built.
And I now have a boat to sling them off.
a bit of eps holding the winglet on, that'll do, following the flexy foils trend...
F18 Sailor
Super Anarchist
juniordave,
Looks good! There are two types of balance in laminates; actually balancing the number of layers/axis, and balancing the cloth itself. Worry about the latter when using lighter weight fiberglass.
The ANSYS runs are interesting, but how much that will help vs. sailing->testing->sailing I'm not so sure. I for one cannot setup the boundary conditions when waves are involved.
Looks good! There are two types of balance in laminates; actually balancing the number of layers/axis, and balancing the cloth itself. Worry about the latter when using lighter weight fiberglass.
The ANSYS runs are interesting, but how much that will help vs. sailing->testing->sailing I'm not so sure. I for one cannot setup the boundary conditions when waves are involved.
