Canting keel forces

dtoc

New member
39
2
I have been thinking a lot about the transient forces, both resonance and shock loads. The biggest issue for me is not having good analytical tools, because doing the analysis by hand is slow, very error prone and sometimes basically impossible. The one tool that I have used is Working Model 2D, which is nice but quite limited. A definite improvement from Excel, which is where I did the initial steady state calcs.

Getting the basic torque balance correct was relatively easy either way. The harder aspect was how the loads are split between the 2 rams, as the system is statically indeterminate. The location of the ram pivot point (trunnion) and offset angle for keel arm above the keel pivot point are 2 unknowns that affect the fraction of the ram force that applies a torque to the keel. If the rams were at 90 degrees to the arm, then all of the force is applied, if 0 degrees then none of the force is applied as a torque. Another way to look at this is the minimum tangential distance between the ram (and its extensions) and the keel pivot point is the effective arm on which the ram force is applied.

What I found is that the there is a direct relationship between the height of the ram trunnion and the optimum keel offset angle. With a lower trunnion point you can have both arms connect to the same point, with a higher trunnion point you will offset the connection point +/- n degrees. While lower is better for the weight, there is the issue of fitting everything in.

How the load is spread between the rams is effected by this, as well as other factors. The simplest is to balance the pressure in each cylinder. This will give a force balance of 1.22:1 for the specified ram based on the cylinder to rod diameters. Then located the rams to maximize the possible force at max cant. This however does not consider the how the rams will load in the transient case.

In a transient event, such as landing off of a wave, the hydraulic system will act basically as a closed system. Pressure relief/control valves will most likely not be able to react fast enough. The hydraulic system will then act as a spring with a known k constant. The k will be dependent on the Bulk Modulus of the fluid, the initial volume of fluid and any flexibility in the system. System flex may be an accumulator, which is designed to add volume under increasing pressure, or hoses that expand under pressure.

When you adjust the pressure at steady state you are effectively changing the initial condition of the hydraulic spring with a very very small change in the spring constant. So when the transient occurs the additional loads are applied to each ram based on the spring constant and the displacement, not the load split under steady state.

This is a critical point because what may appear to be an optimum ram location will overly load the pulling ram in a transient. The max transient load case, 40 degrees cant, must keep the load ratio at 1.22:1 during the transient to utilize all of the strength of the system.

Updated calcs: (sorry but I don't know how to create a table in the post so the number can be hard to read)

cant angle 40

ram stroke (mm) 386

trunnion to keel max (mm) 836.4333

trunnio to keel min (mm) 450.4333

internal keel arm (mm) 300.2547

trunnion offset y (mm) -15

arm offset (degrees) 5

Trunnion (x) 620.9346

Trunnion (x) -620.935

Trunnion (y) 285.2547

push ram angle 5.002929

pull ram angle 5.005439

Push angle 50.00293

Pull angle 130.0054

heel 0 5 10 20

cant 40 40 40 40

bulb (kg) 6500 6500 6500 6500

arm (m) 4.1472 4.1472 4.1472 4.1472

fin (kg) 1054 1054 1054 1054

arm (m) 1.96 1.96 1.96 1.96

Keel torque (kg m) 18655 20522 22233 25134

arm (mm) 300.25 300.25 300.25 300.25

perpendicular force @ arm (kg) 62132 68349 74046 83710

ram force ratio (selected) 1.22 1.22 1.22 1.22

push ram force (kg) 44571 49031 53117 60050

pull ram force (kg) 36538 40194 43544 49227

FoS at working 1.44 1.31 1.21 1.07

FoS at Max 2.87 2.61 2.41 2.13

FoS at Ultimate 4.31 3.92 3.62 3.20

 

star-sailor

Member
151
0
Europe
google-translation of an article on the movistar-homepage:

The ground equipment of movistar completes the changes that the Spanish boat will undergo before returning to the water the next week. As they in the heat of announced in his daily reports stage between City of the Cabo and Melbourne, most important it will be to replace the hydraulic engineers who move the pivotante keel of the 70 VO Spanish, that has been the headache of the crew from the twelfth day of competition by waters of the South Ocean. The equipment of movistar is going to reject the titanium hydraulic engineers and to recover those of stainless steel that used when coming out throughout the previous training from Vigo, which they did not give any problem in more than 20,000 miles of navigation. Thus they will be able to go to top without worrying about its reliability and obtaining more speed. The hydraulic engineers with whom movistar has made the second stage of the Volvo Ocean Race are made in titanium, a very light material that allowed the equipment to save more than 100 kilos of weight. Steel titanium the hydraulic engineers have in front of not failed to be made in titanium, are a very valid material, have commented Greg Waters, expert in hydraulic systems for pivotantes keels. The problem arose because they were too light to even be of titanio.El titanium it has peculiarities that there are to consider at the time of working with him, and is better to make with an extra of weight that with the theoretically ideal weight. Greg Waters works in the specialized company Central Coast, and been has involved in the project to movistar from the beginning. It has developed hydraulic engineers for systems of pivotantes keels of emblematic boats like the Morning Glory, the Pyewacket, the Wild Oats X or the Wild Oats XI winner of the Rolex Sydney Hobart 2005. Those boats are examples of which the pivotante keel in himself is not problematic if it is designed and it made well. The specialist of Coast Power station is the father of the hydraulic engineers who mounted movistar in his initial passage from Australia to Spain. Initially, the idea was to mount carbon hydraulic engineers. While the equipment hoped to that they were ready, they made those of stainless steel, that are with that the boat sailed without incidences from Newcastle (Australia) to Vigo. Those of carbon mounted in Sanxenxo and as soon as 20 minutes lasted, reason why it became to those from stainless steel. Later it was requested to them that they matched second of steel in case they were necessary, and hardly weeks before the exit it was chosen by those of titanium, that they did not make. The hydraulic engineers throughout work the time that the boat is sailing, more than 470 hours in the second stage of the Volvo Ocean Race, even when the keel remains fixed. Those of titanium with which movistar to Melbourne arrived did not support the exigencies of the regata from City of the Cabo. Its safety margin was not the suitable one. This variable measures number of times that can to resist pressure theoretical to that 70 tons work altogether, that in conditions of normal navigation is of 35 tons for each hydraulic engineer. But sometimes that number is surpassed, for example when falling of a wave, and it is there where they receive greater stress. Then: those of titanium had a margin of between 1.5 and 2, whereas those that we are going to mount now have between 5 and 6, reason why can be said that with the hydraulic stainless steel arms we will triple the safety margin. Greg Waters speaks from the interior of movistar, where it works in the preparation of the system to mount the steel hydraulic engineers. They are draining all the oil of the deposit of the emergency device and replacing it by new oil. When breaking the arm took place a loss of oil, the boys added olive oil, the only thing whom they had by hand to preserve the system. The steel hydraulic engineers will mount throughout the weekend. The only thing which they must vary is the pressure of the system, because since the new hydraulic arms have greater dimensions, works to smaller pressure than the old ones. The rest will remain equal.

 

STYacht.com

Super Anarchist
1,691
0
Amsterdam
Vibroman, dtoc,

A point about the flutter theory for fatigue. What makes you think you are hearing the keel? I would be just as likely to suspect the boards, even if not fully extended. They are assymetrical, given a clear higher and lower pressure side virtually always. Also the rudders for that matter, though they use angle of attack and not assymetry to develop lift.

The cant keel should not have a significant angle of attack due to leeway because it is, well, canted out to windward. Unless the designers are trying to generate lift, which would induce drag. I suspect all the boats are counting on planing for dynamic lift, not on the canted keel foil. Also, would you not expect the vibration to change dramatically as the pressure field around the hull and near the free surface in waves changes, if it was coming from the keel fin?

D

 
Vibroman, dtoc,
A point about the flutter theory for fatigue. What makes you think you are hearing the keel? I would be just as likely to suspect the boards, even if not fully extended. They are assymetrical, given a clear higher and lower pressure side virtually always. Also the rudders for that matter, though they use angle of attack and not assymetry to develop lift.

The cant keel should not have a significant angle of attack due to leeway because it is, well, canted out to windward. Unless the designers are trying to generate lift, which would induce drag. I suspect all the boats are counting on planing for dynamic lift, not on the canted keel foil. Also, would you not expect the vibration to change dramatically as the pressure field around the hull and near the free surface in waves changes, if it was coming from the keel fin?

D
My guess too. A stiff carbon board with a little play in it's casing can easily cause that noise.

 

dtoc

New member
39
2
Vibroman, dtoc,

A point about the flutter theory for fatigue. What makes you think you are hearing the keel? I would be just as likely to suspect the boards, even if not fully extended. They are assymetrical, given a clear higher and lower pressure side virtually always. Also the rudders for that matter, though they use angle of attack and not assymetry to develop lift.

The cant keel should not have a significant angle of attack due to leeway because it is, well, canted out to windward. Unless the designers are trying to generate lift, which would induce drag. I suspect all the boats are counting on planing for dynamic lift, not on the canted keel foil. Also, would you not expect the vibration to change dramatically as the pressure field around the hull and near the free surface in waves changes, if it was coming from the keel fin?

D
My guess too. A stiff carbon board with a little play in it's casing can easily cause that noise.
Good point about the carbon daggerboards and designing the strut to have minimal drag.

I don't think the analysis will show any significant forces possible from flow induced vibration on the keel strut. However the analysis is interesting and we'll see what it says.

What I'm more interested in is looking at using a spring backed accululator to spread out the load from a landing into one that will increase the righting moment above the steady state value for long enough that has a positive effect. Cause the boat to roll to weather on landing rather than shaking and vibrating the energy away. Yes, it will decrease the righting moment just as much following that, but still a big net plus I think.

dtoc

 

born2sail

Super Anarchist
Also, with regard to the harmonics, there's a rudder, strut and board "singing" at the same time. Is it possible that the different "songs" could be causing a problem?

 

francescobussi

New member
16
0
Mah.....I don't think that boat engineers don't consider the evenience of the flutter...this is a weel known phenomenon...and having a symmetrical or asymmetrical profile is the same.....

The flutter phenomenon is induced by lift on a wing and by low torsional-flessional strenght.....

(both profiles symmetrical or asymmetrical produce lift....the difference is mainly that an asymmetrical profile produces lift even at negative angle of attack) For every structure you can note there is a point called center of cut (I don't know if the english spelling is right) ,this is the centr of torsion.

The flutter problem can be avoid by putting a D-box (looks at radiocontrolled light airpanes)..this D-box is located on the leading edge of the wing so the center of cut is translated near the leading edge.....now if you have that lift is located rear of center of cut you avoid flutter.....

And even I don't think that planing (like this boat do) induce on surfaces a frequency that can damage the keel...

If you want a more precise description in flutter let me know...i can try to explain...

Ciao

Francesco

 
Can it be it is not the deceleration of the bulb weight that is causing the load, but the boat landing on the keelfoil? Imagine slamming a plank flat on the water at 15 knots, that will be hard to force down.

 

vibroman

Super Anarchist
Vibroman, dtoc,

A point about the flutter theory for fatigue. What makes you think you are hearing the keel? I would be just as likely to suspect the boards, even if not fully extended. They are assymetrical, given a clear higher and lower pressure side virtually always. Also the rudders for that matter, though they use angle of attack and not assymetry to develop lift.

The cant keel should not have a significant angle of attack due to leeway because it is, well, canted out to windward. Unless the designers are trying to generate lift, which would induce drag. I suspect all the boats are counting on planing for dynamic lift, not on the canted keel foil. Also, would you not expect the vibration to change dramatically as the pressure field around the hull and near the free surface in waves changes, if it was coming from the keel fin?

D
My guess too. A stiff carbon board with a little play in it's casing can easily cause that noise.

Good points all.

The natural frequency would not change significantly with wave action. However the exciting forces (flutter vortex shedding or what ever you want to call it will) This could cause the structure to vibrate intermittently. Visualize running finger round top of wine glass. It only "sings" when the exciting forces are just right.

On one of the vids you do here 2 different frequencies...perhaps board and rudder.

Are there any reports from the crews of the other boats of "keel/board/rudder hum"? and if so any info as to to the location of the sound? eg aft keel area board etc?

Also saw some video on sail TV yesterday of another boat haulin A$% and I had a similar hum to it.

 
Last edited by a moderator:

vibroman

Super Anarchist
Also, with regard to the harmonics, there's a rudder, strut and board "singing" at the same time. Is it possible that the different "songs" could be causing a problem?
The only way that could happen is through a "beat" frequency. This occurs as the two signals alternately add to each other and then subtract. Based on the data from the video the beat frequency would be 63 Hz I did not see or hear that frequency. but that may be due to the frquency response of the microphone on the vid camera. In this case the beat frequency. Interestingly that frequency is in the general area of the 1 xRPM of a typical hydraulic pump. I have no idea what the speed of the pumps on the boats are.

 
I've now seen two suggestions that the keel strut will increase the loading on the rams: it won't, it will reduce. Slamming down results in upward lift countering the bulb. Same goes for aircraft: full wing fuel tanks is good for in flight fatique (hopeless for taxiing though, and that's a major culprit, other story).

As for fatique loading; anything more than 1 cycle is fatique lading. You can't say that 100000 cyclees isn't a lot for stainless, all depends on the stress levels. Get to know the sigma, than find the sigma / cycle curves to make a judgement about fatique level.

If safety factors were 3-4 for static cases and they are seeing close 2+g loading you'll very quickly end up with fatique failure. Add to this misalignment, either static (installation tolerances) and dynamic (boat flexing), and you're way up the creek.

Flutter would be less of an issue on a solid steel structure with such a heavy bulb; it will have very strong damping characteristics. The open 60 struts are carbon with lighter bulbs making it all a bit more critical.

I wouldn't recommend putting a D-nose on any structure; relatively inefficient due to reduced moment of inertia of the leading edge, and risky upon impact. Most wings and keels are box structures with a leading and trailing edge fairing (also makes for handy fuel storage, or moonshine if you're in that business). Good point about the central point of lift, approx 25% chord on most profiles. At this point you can separate out Cl (lift), Cd (drag) and Cm (torsional moment) and graph them against Angle of Attack.

Great thread, loads of different engineers providing input in different areas, keep it going.

 

dtoc

New member
39
2
Can it be it is not the deceleration of the bulb weight that is causing the load, but the boat landing on the keelfoil? Imagine slamming a plank flat on the water at 15 knots, that will be hard to force down.
With the keel fully canted and the boat heeled 5 degrees, the bulb is still 3+ meters below the waterline. So the boat would really have to be launched. Even then when the strut re-enters the water it will be at a 45 degree angle and with significant forward speed. Angle of attack for the foil would be <5 degrees for most of it. Aoa increases from 0 as the boat accelerates down. (If the hull ever reached a vertical speed of 15 knots, a flat landing would probably break almost any boat)

The resistance of the water (lift), whether splashing or already submerged, will slow down the bulb and fin as the boat falls off the wave. However I don't believe it will be significant relative to the force when hull hits the water.

 

Steve Clark

Super Anarchist
Greg Waters on DailySail suggested that he titanium rams had a factor of safety of 1.5-2.

I would regard this as completely inadequate in a keel structure.

I wouldn't be happy with keel bolts this close to the edge.

When we redesigned Red Herring's keel we used a worse case scenario of the keel out of the water and bouncing up and down. Added a 5x FOS to that and felt it was OK because the vessel is primarily for inshore use.

I'm not going to point fingers, but I think I see the problem.

SHC

 

dtoc

New member
39
2
Greg Waters on DailySail suggested that he titanium rams had a factor of safety of 1.5-2.I would regard this as completely inadequate in a keel structure.

I wouldn't be happy with keel bolts this close to the edge.

When we redesigned Red Herring's keel we used a worse case scenario of the keel out of the water and bouncing up and down. Added a 5x FOS to that and felt it was OK because the vessel is primarily for inshore use.

I'm not going to point fingers, but I think I see the problem.

SHC
Steve,

I have just been reading that article and I agree with you on the FOS margin.

What is interesting is that I have done my calcs based on the Cariboni system, which as the article states is used by ABN1&2 and Brasil, which have not failed.

He also states that the load on each ram under normal sailing conditions is 35 tonnes, inline with my calcs based on a 5,500 kg bulb, which is what is reportedly being put on Brunel. Then with only a FOS of 1.5 - 2, ouch!

ABN & Brasil - FOS 3 - 4.5 (Max - Ultimate load @ 5500kg)

Movistar - FOS 1.5 - 2 now 5 - 6

dtoc

 

sucky bus

Super Anarchist
1,034
0
After all of our hard work.. just wondering why they don't use the hydrolic system to move the keel, then a physical lock / bolt (would be a big one though) to lock the keel in position.. surely this would mean that the hydrolics are used only during tacking and gybing.. not a huge amount of this in offshore... and use, without locks in the inshore..

Thoughts please as I'm sure this scenario must have been considered?

:ph34r:

 

dtoc

New member
39
2
After all of our hard work.. just wondering why they don't use the hydrolic system to move the keel, then a physical lock / bolt (would be a big one though) to lock the keel in position.. surely this would mean that the hydrolics are used only during tacking and gybing.. not a huge amount of this in offshore... and use, without locks in the inshore..
Thoughts please as I'm sure this scenario must have been considered?

:ph34r:
Heavy extra mechanism

Difficult to lock in place without slack, which causes impact loads

Hard/impossible to adjust trim quickly

No shock absorbation in system (not sure current systems are taking full advantage of this)

Any other thoughts?

 

Lima November

Super Anarchist
1,330
7
Nantes, France
No shock absorbation in system (not sure current systems are taking full advantage of this)
Maybe not full advantage, but they use it. They have some valve in the hydraulics that opens when the pressure in the cylinder gets too high, so that the keel drops a little on heavy impacts (I remember Bouwe Bekking mentions it somewhere). That limits the peak loads on the rams, by the way (for those of you who are computing those loads).

 
C

Codger

Guest
After all of our hard work.. just wondering why they don't use the hydrolic system to move the keel, then a physical lock / bolt (would be a big one though) to lock the keel in position.. surely this would mean that the hydrolics are used only during tacking and gybing.. not a huge amount of this in offshore... and use, without locks in the inshore..

Thoughts please as I'm sure this scenario must have been considered?

:ph34r:
Heavy extra mechanism

Difficult to lock in place without slack, which causes impact loads

Hard/impossible to adjust trim quickly

No shock absorbation in system (not sure current systems are taking full advantage of this)

Any other thoughts?
Since the shock load is predominant, consider using an accumulator. Here is a page with a drop down menu giving many options. http://www.parker.com/Products/EPD/EPDResu...+5%3A24%3A14+AM

Essentially, it works like the shock absorber on your car. You know what a bendy rig is. This would give you a "bendy keel".

 
Last edited:

Latest posts




Top