watertight bulkhead

Good advice Angelique: working did it for me! I understand your point Ad Hoc.

 

Hi Pammie

How ya getting on??

 

The primary structural arrangement is as shown attached. The crossbeams are supported by bulkheads/web frames, inboard/outboard girders (or primary longitudinals), and center girder (note that center girder can be eliminated in small boat provided an alternative structural arrangement is made). Following LR rule, the primaries must be supported by a bulkhead/webframes spaced no more than 6 meter apart. These are primary structures and is sized such that its effective length is the distance between the bulkheads. Where it ends, discontinued, or terminated, it must be bracketed (soft toe or hard).

The side longitudinals (secondaries) are to be continous and supported by bulkhead/webframes/ringframes. It is sized by the distances of the supporting transverse, not more than 2 meters. It must be tabbed to the supporting transverses and bracketed ONLY when it is terminated. LR rule, others may vary.

The sacrificial nose can treated in a variety of way provided it can withstand the slamming effect;
1. Using a solid foam cored laminate.
2. Using a honeycomb core so that it disintegrate on a normal (horizontal) impact.
3. Arranging the laminate in a single skin layup so that it spreads outward on impact.
4, A combination of any.

 

Rules vary. Stiffeners are secondary elements that is meant to stiffen or provide support to a plate. There are rules that apply and defines it as "intercostal" or broken. The primary use is when it is used to stiffen the top of a hatch cover where it stops before the bend of the plate. If used as a side longitudinal, LR defines it as continous and connected to the transverse(s). Perhaps there are other uses of the term "intercostal" that I am not aware of.

 

It just means it is not "connected" to the primary supporting structure.

As such, structurally, it changes the bending moment and thus end conditions.

So it goes from a built-in at both ends, M = q.L/12 to a simple-support M=q.L/8...so the BMt is greater and as such the longitudinal must be stiffer.
Then you just need to ensure the dfelections are not greater, mid-span, and also at the ends, the terminates are not stress concentrations.

 

Yes you are right. The auto edit think both is wrong. I was not able to use the correct term. Will edit my post.

 

Thanks AH. I follow the shear, bending, and deflection coefficient in Table 3.15. The software also defaults to the correct coefficient whenever I define the member as "stiffener" or not fully fixed. Not really sure on this one as far as the software is designed but I assume a plate is bounded by stiffeners on both sides and terminated in the ends by a primary. Am I correct? The only "not fixed" element I can think of are hatch covers/glazing, where it is simply supported.

But in the rule, the mention of intercostal appears only in Wet Deck Transverse Stiffeners and Side Outboard Transverse Stiffeners. There is also a mention of a Centreline Girder, a primary element, where it maybe formed as continous or intercostal. Here, I always follow the rule "whenever I break it, bracket".

 

I'm very curious about her build. She seems to be doing a lot of testing of methods, so is she building a designed boat or is she stepping away from the designer or is she just testing her results to the designer recommendation? Does anyone know what her build is? Just curious. It looks like habitable cat hulls from her facebook page, but I didn't see her mention the design.

 

She is building a boat meant to be raced. Testing material is required if the boat is to be built to a certain regulations. In the absence of a coupon test, predicted material strength is computed to 80% of calculated strength which is a lot to lose. Already, her choice of materials shows she is exceeding the predicted strength, both by LR and DNV.

 

Every boat is the same, no matter what material. See below for 'typical' structural arrangement.



The plate, whether steel, ally or composite, has transverse frames and long.t stiffeners. The frames support the stiffeners, the frame and stiffeners support the plate.

The stiffeners can be continuous, as shown here:


passing through the frames that support them, like in the typ arrangement above, via a cutout, or, the stiffenders can be intercostal, they stop short of the farmers, as shown below:



And, as noted above, so long as you ensure the stiffener is stiff enough to pass the checks, because the ends conditions are now different, all is ok

 

Haven't been on boat design for a few weeks. Thanks Ad Hoc bringing the discussion to live again. And Thanks RX for making me attentive to it.
Ofcourse I have tought about it in the mean time, but had no time to work it out.

I'll start with the conditions: worstcase is a collision with a heavy object on one of the hulls in such a way that the boat is not able to shed collision energy in turning. Eg the other hulls crashes into the same object. In this case it is necessary to limit the load to keep the essential boat structure intact. The question what load this should be is not answered yet. I worked the other way around: because of the length of my workshop I have to build the hulls in two parts one being the nose of 1,5 meter. Because of this structure it is reasonably easy to sacrifice the nose. 1,5 meter if necessary, 0,75 meter is another option. The deceleration from 20 knots to 0 in 1,5 meter leads to a force of about 72 kN.

Ofcourse I do not want to break the nose in normal conditions. The design side load is 23,4 kN/m2 (ISO, I started reading Larsson! Should have done that a year ago :$). The most forward weak spot should be at 20 cm from the nosetip as this has the least width. When it breaks at this point the unsupported plates will break until the next bulkhead (or further). The load on this 20 cm tip is 4,7 kN. The next bulkhead is 75 cm from the nosetip. With a weak spot at 80 cm the load on this section is 10 kN. The bulkhead after this is at 150 cm and is a double one.

Breaking is dependant on the angle of the collision as combined bending/ kink (is this the right word?). I wanted to calculate the necessary moment of inertia for both situations but the numbers don't make sense yet... TBC in short term now!.

Another aspect is that I partly finished the nose. Inside laminate is 800 gr triax. My idea was to cut a groove in it and make the total sandwich thinner. Which is another method as the method proposed by RX based on laminate shear. Would that work? Ofcourse cutting sections of the nose is no problem.

 

Pammie,

I plotted the forces you will be experiencing on the hull(s).

The first thing to do is to design you hull to withstand the pressure loading. You may be using a different rule but just the same, the operating conditions are almost the same. See inset of the properties and operating condition. I am getting less pressure (20 kN/m2) than what you have derived. Note that this pressure rises as it moves forward. It not only rises, note that the pressure is for the bottom hull (side pressure is much less) but rises above the waterline as it reaches the stem (follow the orange dotted line). This suggest that you have to strengthen your forward hull. The blue vertical line are probable solution. Add more transverse frames.

The hull is designed to survive at 2.4 g vertical acceleration at LCG which you designed at 4 meter from transom. The g increases to 5.7 at the stem.

This rise of acceleration results to shearing of the attachment you plan at 1.5 meter from stem. Note the inset.

Once you have calculated the strength of the forward hull, you now can analyze the impact of collision/horizontal shear and design it if you want to shear at the connection or allow it to disintegrate. Most sacrificial nose is only 5% of the length of the hull or starting from a collision bulkhead. Note my previous post by LR.

 

OK, had to do that anyway. Larsson (eo) claim to use ISO 12215-5. As I understood part 7 on multihulls is no available yet. My idea was using ISO12215.part 5 and make an adaption for slender hulls as described in DNVGL P3.C3.S5.2.2.2. In my earlier calculations I used a design wedge of 100%.

You are using LR I think? Maybe we used different numbers in our equation? In your file I some differences (eg design speed 9 knots). I calculated the loads specifically for the nose (Kl; loads after 60% Lwl are less). I also used the load at half height (Kar) because I was interested in the total load over the complete height.
Larsson divides between planing power boats and sailing boats. For the first he uses added hydrodynamic factor ncg equivalent to your acceleration numbers. For the second he uses a factor (Ksls) to incorporate slamming. This is for slower monohulls.... So using the acceleration approach seems a good idea.

You talked earlier about a ISO12215 spreadsheet? Do you have that for me? I will first make an overview of the loads for the complete hullso we can better compare.

 

ISO spreadsheet deals only with plates and plate with stiffeners. LR is sophisticated. So many inputs and many control on the output. Whenever LR or ISO cannot handle specific use I design my own spreadsheet. For the vertical acceleration at any point away from the transom, I have this in the spreadsheet for quick analysis. Though that is reinventing the wheel. LR has this in the software. I believe GL has this method also as one of the input is distance away from transom (Daft and Xa). I just did not incorporate this in the spreadsheet.

Note that moments are calculated from the LCG as the hull rotates along that axis.
Pressure also differs depending on the height from the baseline and distance from aft. Bottom has the highest pressure. Sides not so much. Inboard hull, much less. Wet deck, depending on height, sometimes as much as bottom due to slamming.

 

ISO is ontensibly for those not doing a Class build and/or not wishing to plough through endless Class rules either.
It is aimed at those who are "small time" builds/designers in an attempt to get some minimum baseline level of acceptance.

You can pick and choose, but don't mix and match....the origins of their formulae and scantling derivation is different. Results are generally similar but the ruotes at which one arrives, very different!