Canting keel forces

dtoc

New member
39
2
Much discussion has been put up in boards on why the canting keel systems are failing, including support structure, ram pistons and hydraulics. I thought it might be interesting to look at the forces involved and the equipment used to exert them.

The canting keel works in a lever mechanism, where the bulb sits on one end of the lever and the hydraulic rams push/pull on the other. In order for the keel to stay in place relative to the structure, the torques created by the rams and the bulb must be in balance.

In order to do these calculations the masses and geometries must be known. For this I used information posted on the web, specifically the ram info from Cariboni for their special Volvo 70 Titanium ram and the general rule dimensions for a Volvo 70 along with some estimates based on the CAD drawings shown.

The draft is assumed to be 14.5 ft, or 4.42m. I used a bulb mass of 6000 kg, but also calculated based on 4500 kg. While the draft is 14.5 ft, the distance from the pivot point to the cg of the keel will be less than that. This is due to the pivot point being below water level, even though it is inside the hull, and that the cg is not at the max distance from the pivot point. Initially I used 14 ft (4.27 m), but I will show where this is in the calcs in case you want to split the fin and bulb weights and calculate a detailed cg and effective mass. However, the exact mass of the bulb is a highly guarded secret for the teams.

The other end of the lever is easier to calculate. The max stroke of the custom ram is 386 mm. For it to swing the keel through 80 degrees (+/- 40) and have this range of motion the geometry can be solved for with the dimensions given. The 1 unknown is the trunnion pivot point of the ram. Its distance from the end of the cylinder, height above the keel pivot point and the lateral distance from the pivot point. Assuming a pivot point of 1/3 of the distance up the cylinder, an estimate from the cad drawings, the other 2 numbers can be solved for. There is a relationship between the pivot point and the length of the lever. I solved this to maximize the torque applied. The other assumption is that the 2 rams connect to different arms, at an angle away from 0 when the cant is 0 degrees. This is to maximize the pull force when at max cant. From the drawings I assumed +/- 20 degrees. The optimum value can be solved for, but I haven’t. These calculations gave me a lever length of 340 mm with the pivot point 20 mm higher. The combined force perpendicular to the respective lever connection points is 1.675 times the push force.

Steady state with 0 degrees heel and 40 degrees cant

6000*4.27*SIN40=Fpush*1.675*0.340

Fpush=(6000*4.27SIN40)/(1.675*0.340)

Fpush=28930 kg

Steady State with 20 degrees heel and 40 degrees cant

Fpush = 38976 kg

The ratio of these forces to the rated loads for the ram can be viewed as a factor of safety or as the number of g’s before that limit is reached.

Case 1: 2.2 to working load 6.6 to ultimate

Case 2: 1.6 to working load 4.9 to ultimate

My question to the naval architects is what is range of g forces a boat like this would encounter when landing? Or to anyone who was keeping track of the onboard data from the VOR website? I remember seeing 4 g’s.

This does not seem like much of a factor of safety, which may be why these titanium rams have been failing.

The next subject to discuss is the hydraulic pressure transients created by the shock loads and what can be done to control / utilize them. Which is the reason I did all of these calcs. As has been previously mentioned pressure relief valves are not effective.

 

snorky

Anarchist
760
0
Semaphore
YO... this is excellent work.... perhaps you can get some dosh from FARR.. or why not email your calcs to VOR sure they would love to have them!

 

nobody

Member
464
0
Sydney
Much discussion... [snip]

Excellent work. If the figures are correct it makes you wonder. The designers seem to have not much allowance for fatigue in there. The ram load would be based on a brand new ram. Repead the load over and over and ....

Nobody

 
Last edited by a moderator:

winchfodder

Super Anarchist
1,800
333
Carolina, USA
and i presume somewhere there has to be a calculation on the leverage on the ram structure from the fin area which attaches the 6 ton bulb to the keel when the 16 ton boat travelling at say 25 knots drops vertically 2 metres

 

snorky

Anarchist
760
0
Semaphore
YO... this is excellent work.... perhaps you can get some dosh from FARR.. or why not email your calcs to VOR sure they would love to have them!
Being a bit more serious.. having this massive bulb waving around, subject to both the vertical drops and perhaps more importantly horizontal and torsional forces.. need some form of metrics from the boats.... whilst they rams act in one plane, they aren't doing it in the others which is probably causing the failures...

Perhaps we can 'crack it' where the others can't...

:blink:

 

lucas

Super Anarchist
1,550
0
Well, here goes your bulb weight clac... :p

111704_802_1138250726898_movistar_27s_bulb_165x165.jpg

The result after a meeting between a chainsaw an the bulb from movistar...

( http://team.abnamro.com/web/show/id=102954/contentid=2409 )

 

winchfodder

Super Anarchist
1,800
333
Carolina, USA
Now if they had used the cariboni system as used on an Open 60 as in the photo in the study then i don't think they would be having the problems we are seeing now !!!

cantingkeel1.jpg

CK88_2mini.jpg

 
Last edited by a moderator:

canstead

Anarchist
904
47
Looking at this paper, the carboni pictures and making some wild conclusions from what Don Jones, and the original Movistar failures...

Am I right in saying the Carboni mechanism is in AA1 and AA2. Now looking at this system it appears that it is a self contained unit, ie the ram trunions and keel bearing are structurally linked within the unit, and the majority of forces are transfered to the hull through the mechanism frame, and not through the rams? I would think that therefore any twisting of the keel through acceleration/decelration would not affect ram alignment?

Therefore those issues Don Jones was talking about don't apply to the Carboni method?

Also, when Movistar broke in the first leg, they talked about 'Ram Shelfs' ie the bit which connected the RAM to the BOAT. This is in-line with Don Jones comments, but also doesn't sound like the Carboni type set-up.

Wild Speculation?

(posted at same time as winchfodder above!)

 
Last edited by a moderator:

moody frog

Super Anarchist
4,292
108
Brittany
Excellent find Arthur, now I hv one more thing on my week-end agenda !

Now, Am I wrong ? :

The study is based on Cariboni.

Semcon (where the study comes from) is a sponsor and technical partner of Ericsson.

Ericsson uses Marine & Hydraulics systems

??

;) Incidentally the M&H webpage on "Custom Engineered Marine Hydraulic Systems for Maximum Yacht Performance" is " temporarily unavailable" B)

 

dtoc

New member
39
2
Three items on my previous calcs:

1. Add an optimization for the offset angle for the internal lever arm. Simple method, but it gave me 12 degrees rather than the 20 I assumed.

2. In doing so I found an error in my calcs which I corrected. Unfortunately this may the results worse, i.e. a smaller FoS.

3. I adjusted the bulb weight and arm length and added a fin weight and arm based on the Chamlers paper. My only disagreement with their number is that they assume the pivot point is at the bottom of the hull. We know that in the Volvo 70’s the pivot point is inside the hull, so I added 30 cm to the arm length. I have not read the entire Chamlers paper, but it does look quite interesting.

Here are my updated results:

heel 0 20

cant 40 40

bulb (kg) 6500 6500

arm (m) 3.92 3.92

fin (kg) 1054 1054

arm (m) 1.85 1.85

arm (mm) 342.53 342.53

ram force multiplier 1.41 1.41

ram force (kg) 36460 49123

FoS at working 1.76 1.30

FoS at Max 3.51 2.61

FoS at Ultimate 5.27 3.91

 
Last edited by a moderator:

dtoc

New member
39
2
Looking at this paper, the carboni pictures and making some wild conclusions from what Don Jones, and the original Movistar failures...
Am I right in saying the Carboni mechanism is in AA1 and AA2. Now looking at this system it appears that it is a self contained unit, ie the ram trunions and keel bearing are structurally linked within the unit, and the majority of forces are transfered to the hull through the mechanism frame, and not through the rams? I would think that therefore any twisting of the keel through acceleration/decelration would not affect ram alignment?

Therefore those issues Don Jones was talking about don't apply to the Carboni method?

Also, when Movistar broke in the first leg, they talked about 'Ram Shelfs' ie the bit which connected the RAM to the BOAT. This is in-line with Don Jones comments, but also doesn't sound like the Carboni type set-up.

Wild Speculation?

(posted at same time as winchfodder above!)
I don't think that the use of a self contained unit rules out the issues brought up by Don Jones. The structure used to hold the rams and the keel has still been optimized for weight and is still subject to torsion when the boat rapidly decelerates. At the pulling ram, with 40 deg cant, any out of plane movement of the keel will cause a high shear force with no room for error. What I mean is that if a torque parallel to the axis of the keel pivot is applied, due to rapid hull deceleration and bulb momentum, and the alignment of the connection point is now off by a few mm, there is no room for a small angular deflection of the ram in the cylinder or bending (elastic) of the ram. This could cause the failures shown. These calcs are impossible by hand because the structure connecting the ram trunnions and keel pivot is very complex.

 

wraith

Super Anarchist
2,195
0
Solo said:
My question to the naval architects is what is range of g forces a boat like this would encounter when landing? Or to anyone who was keeping track of the onboard data from the VOR website? I remember seeing 4 g’s.
g's of the hull have little to do with g's on the end of the flexible 15 arm.

s.
Unless the arm is sufficiently long as to move the bulb sigificantly away from the surface of the Earth a g is a g (leaving aside issues to do with a non spherical earth).

The torque induced as a result of the deceleration is absolutely to do with the lever arm.

 
Last edited by a moderator:

GybeSet

Super Anarchist
Excellent find Arthur, now I hv one more thing on my week-end agenda !
Now, Am I wrong ? :

The study is based on Cariboni.

Semcon (where the study comes from) is a sponsor and technical partner of Ericsson.

Ericsson uses Marine & Hydraulics systems

??

;) Incidentally the M&H webpage on "Custom Engineered Marine Hydraulic Systems for Maximum Yacht Performance" is " temporarily unavailable" B)
whattabout on their webpage 'Volvo BREAKING News' on? off?

go C.C.H systems

 
Last edited by a moderator:

born2sail

Super Anarchist
First off, thanks dtoc, this is one of the best threads ever. I didn't think starting the day with so much math (or whatever you call it) could involve so little pain.

Not being an engineer, maybe this question has been answered in some other lingo: If the fairing on the strut is out of whack, can sufficient twist be put on the keel system to cause the mechanism to go kaput before anything else?

 
Last edited by a moderator:

canstead

Anarchist
904
47
Looking at this paper, the carboni pictures and making some wild conclusions from what Don Jones, and the original Movistar failures...

Am I right in saying the Carboni mechanism is in AA1 and AA2. Now looking at this system it appears that it is a self contained unit, ie the ram trunions and keel bearing are structurally linked within the unit, and the majority of forces are transfered to the hull through the mechanism frame, and not through the rams? I would think that therefore any twisting of the keel through acceleration/decelration would not affect ram alignment?

Therefore those issues Don Jones was talking about don't apply to the Carboni method?

Also, when Movistar broke in the first leg, they talked about 'Ram Shelfs' ie the bit which connected the RAM to the BOAT. This is in-line with Don Jones comments, but also doesn't sound like the Carboni type set-up.

Wild Speculation?

(posted at same time as winchfodder above!)
I don't think that the use of a self contained unit rules out the issues brought up by Don Jones. The structure used to hold the rams and the keel has still been optimized for weight and is still subject to torsion when the boat rapidly decelerates. At the pulling ram, with 40 deg cant, any out of plane movement of the keel will cause a high shear force with no room for error. What I mean is that if a torque parallel to the axis of the keel pivot is applied, due to rapid hull deceleration and bulb momentum, and the alignment of the connection point is now off by a few mm, there is no room for a small angular deflection of the ram in the cylinder or bending (elastic) of the ram. This could cause the failures shown. These calcs are impossible by hand because the structure connecting the ram trunnions and keel pivot is very complex.
The point is on the Carboni the torsion is transfered directly into the structure holding the rams, then into the hull. The canting structure as a whole should flex, including the ram, maintaining the ram/keel alignment. If there isn't this structure the keel mount only flex's, while the ram mounting stay static as the moevement is absorbed by the boat structure.

Also when the boat is falling off waves and the hull is flexing, in the case with no self contained structure, that is causing misalignment as the hull around the ram mounts is moving differently to the hull around the keel mount. In the self contained structure, there would be no relative movement between the ram and the keel.

 
Top