Is bulkhead tabbing now redundant?

Thanks adhoc, last post was very helpful.

All of the sandwich panel failure modes such as buckling, skin wrinkling, crimping, and more, have equations for checking located in here -> http://www.hexcel.com/Resources/Dat...eets/Honeycomb_Sandwich_Design_Technology.pdf

Checking these with the current scantlings we know already work, all seems to be ok, as you would expect.

The yachts im referring to have already been built, and have been for many years, and have NA stamp of approval. The changes im looking to make, is simply the method in which the stiffeners and frames are installed into the boat - mainly to reduce labour and time. Cost is not such an issue - if it were, they would not be built using expensive cores and epoxy resins etc. I suspect everything is glass taped, simply because nobody has really looked at the problem in detail, theyre just doing what they know works -as tunnels would do, and like ive said before, theres nothing wrong with it... well nothing except for the time it takes to build it with those methods... i have a different approach to building, using different methods than most people use, being able to bond rather than tape will make a huge difference to me, but perhaps no help at all to someone else who is building with different methods.

So the main focus is not what the laminates, spacings, sizes of stiffeners are as we already know these...same goes for the laminate E, we already know it. but rather constrained strictly to the bonding / joint areas of these transverse and longitudinal stiffeners, and whether or not we can eliminate the over laminate fibreglass tapes in the bond fillet areas. We may have specialized adhesives with enough mechanical properties to get the job done without it. Considering the loads are relatively conservative, were only dealing with maximum displacement vessels of around 6000kgs and speeds not exceeding 25kts. Forward hull bottoms have very small area, with steep deadrise angles, and yachts cannot sail directly to windward, so slamming pressures are kept low organically. Not sure what class would say about the above reasoning tho, i havnt read through the slamming pressure calculations yet.

Brian, i have no desire to try Nomex honeycomb core, which seemed to be the issue Nigel Irens was referring to... i dont beleive its the right core material for building boat hulls, its also not compatible with my infusion process. It would have a place in other areas of the boat however, many use it for furniture panels, and other structural stuff. I think ill stick with foam, in varying densities, depending on required application - Im using 300kg/m3 in my current transoms as well as other highly loaded areas where you would normally find a plywood core for resistance to crushing and through bolting etc. I find foam is very user friendly to build with and the methods i use.

 

This is still a 2 fold process.

1) Mathematical analysis
2) Experimentation, to validate the mathematical model.

 

Ok...

We have determined (with the help of another forum member) that the hull pressure as defined by Lloyds Register rules, that the topsides hull panels have a design pressure of 19Kn/m^2. The bottom pressures are 25kn/m^3 - so these are our design loads for our yacht based on G2 classification and a speed of 16kts in 1m wave height, vessel length 12m and max displacement = 6 tonnes.

Using Plate theory, as defined in the hexcel reference i posted earlier, for a simply supported plate on all 4 sides. Our hull panel was sized as 1.5m length 0.5m width - length between frames and longitundinals.

So using a thickness of 17mm panels - 15mm core with 1mm skins - An E of 40Gpa - derived from panel testing data, I am getting deflection of 4,26mm on the topsides panels - how much is allowable deflection?

Based on analysis thus far, the internal stiffeners are not required to be tabbed as they are not considered structural - they dont even need to be there as the hull is stiff enough in its own right...

We to go from here?

 

25 kN/m2.

Limiting span/deflection ratio is length/midpoint deflection and varies from 100 to 250 depending on the location. Tables are in Part 8, chapter 7.2.1- Failure Mode Control.

4.26 mm deflection seems just about right but where is it located? for hull shell sandwich construction, it is 100 but for main/strength deck, it is 150.

 

I cannot get access to LR rules? The link in the class rules section doesnt work for me... i went to LR website, and it requires login etc etc


deflection is in the center of panel... so if the width of panel is 500mm, then we are just under 1:100 so ok, but if deflection is ratio of length, then 4.26 : 1500 so not ok...???

Ok, so how do we then determine the shear in the stiffeners we have, spaced every 1500mm x 500mm, to see whether fibreglass tape is required on the joints? Does each stiffener see the same pressure, or is it double the plate pressure each side. ie sum of pressure on both sides of the stiffener, which is 0.75m^2 plates each side.

So that would be area of plates*2*defined hull pressure for our load, then use this value to calculate shear in the stiffeners?

 

Groper, bear with me as this is "rusty" from memory not direct reference, the deflection is measured at the center of panel & expressed against the shortest span. In the Aus standard & again from memory, solid glass max deflection allowable was 50:1 & for sandwich panels 75:1, these were max allowable & looks from the postings of RX, that LRs apply a higher standard.
also, this is not my field, when it comes to design & engineering I leave it to the NAs, makes life easy, just stick to the drawings & schedule.
All the best in your enquiries from Jeff.

 

how can you expect to get a answer of you dont persent all the facts ??

what happened to the thick core and thin glass you were going on about making ??
15mm thick core (what type of foam and what density ?) with 1mm skins each side ( what glass you using and what resin ??)- I am getting deflection of 4,26mm on the topsides panels (width between frames of 500mm spacing ??)- how much is allowable deflection?
you are leaving out foam type the actual glass used plus what resin so you cant get a sensible answer without all those things??

So you have carrot panels !!
what happens to a carrot when you try to bend it ??
or
what happens to carbon skinned panels if you try to bend them ,is possibly more relevant to your thinking !!

internal stiffeners are not required to be tabbed as they are not considered structural !!
what ???? stiffeners are not considered structural ?? oh my God !!

 

Scratch my last comment, it doesnt make sense to use the panel areas twice, as each stiffener supports only half a panel, the next stiffener does the next half.

Ok back to my original equation for the shear in the bonding area, we originally had 0.8Mpa shear for the 1000N force. Now we have derived the design load from the LR rules, it is in fact 19kN/m^2.

So, if we have frame spacing of 1.5m, each meter of frame edge supports 1.5*19kN = 28.5kN, So this is our load for our shear calculation, ignoring longitudinals. If we then split the span of the hull panel in half with a longitudinal to get our 1.5m*0.5m hull panels, now these longitudinals take half of the shear aswell, correct? So now each longitudinal and transverse frame, now sees 28.5kN/2 = 14.25kN. So this is what we need to use for our frame and longitudinal shear equation load, correct?

If so, we continue and assume each frame and longitudinal is 150mm deep, 17mm width - this is our 15mm core now with 1mm laminates per side.

Using the same formula as described earlier, that is Shear stress = VQ/It for calculation of our stress in the shear plane occuring between the hull panel and the frame/longitudinal.

The centroid of this new section was at 150.91mm as our T section was now 1500mm width, 150mm depth. *edit* i think this should be perhaps 1000mmx150mm?

So our Q was worked out accordingly, = 1.935.10e-4
The V now = 14250N
the I = 2.1558e-7

And thus the new stress in shear plane of hull to frame and longitudinal joints = 7.5Mpa

Epoxy adhesive mechanical properties = 16Mpa, so we have a safety factor of 2.13.

Does this still sound reasonable?

If so, this is telling me its doable, albeit with a reduced safety factor of 2.13. I looked over the engineering design of a freestanding carbon mast the other day, it was engineered to a safety factor of 2.25, so i dont feel so bad about it.

So now i need to build a test panel and verify, correct?

 

Sorry tunnels, to clarify, stiffeners are not required in addition to the frames and girders/longitudinals as hull panels are stiff enough in their own right (thanks to the cored construction btw).

Additional stiffeners in the form of furniture or other hat stiffeners etc are not required at all and anything added would not need to be considered structural in this instance.

 

sorry but every single piece of anything attached to the skin of the hull needs to be taken as structural and glassed and or stuck in as it should be !! even shelves inside a cupboard or a kitchen bench becomes structural and stiffens a panel it is attached to !!

not going off topic but what is the glass that is used for the skin ?? and how was it used !!!! one millimetre is surf board glass almost !!stringers and stiffeners what shape ?? just flat cored panels 1 layer of glass each side placed in on edge ??what goes on the top edge ?? anything or just a raw cut end ??

 

I might just be a country boy out of my depth with all this boffin buffoonery , but it seems to me if you want to compare some new-fangled idea with what has already proven itself, make up test samples of both varieties and subject them to the same stresses to see how the new idea stacks up against the old, seems highly fanciful to me to be drawing final conclusions from what a calculator says. If not, the test will come on the water, and I wouldn't want to find out at that stage that it didn't work, for want of some simple, inexpensive testing at an early stage.

 

Right. If the panels designed are stiff enough to pass the rule within the 1:2 ratio, there is no need to add plate stiffener. But nothing in the rule says you can't. Adding it will make the panel stiffer and a little over designed.

If you happen to design the panel within the optimum ratio of 1 width, 2 length and find that the panel is still flexible, you can insert a stiffener in between the width. That stiffener can also be a part of a table top or cabinet. The caveat is the panel size is no longer "efficient" as the w/l ratio is now 1:4. Plan carefully. In boats it is ok. In ships, you have to be very efficient in design.

 

The point was, to use the calculator first before spending money on test panels. Now it seems at least plausible, one can now justify spending money on further investigation. If that's successful, then one can justify building the new boat with the new process...

 

We are getting lost here somewhere. Panel flexing is just that, a panel (single skin or sandwich). Panel with stiffener is a different ball game. Calculation is difficult especially with sandwich laminate. I don't have my book with me but I think this is where the 1:50 ratio comes in. Not sure. I also posted the tabulation somewhere in this forum sometime ago.

The resolution of the vertical shear and the horizontal shear (glue line) I think has been resolved and it seems you have a good handle on that.

What has not been resolved is the panel stiffness and the fillet/tabbing flexing. As had been originally posted by Samsam and further explanation by AH. Take a look again at the picture.

To visualize, take a piece of 2x2 and lay it flat on a table. Take a piece of heavy paper (representing a flat panel) and lay (centered) on top of the 2x2. Apply pressure to the sides of the paper and it will bend, It will be supported at the edges of the 2x2 but at the center, it will curve upward, separating from the center of the wood. This indicates that in flexible panel the tensile strength of the glue matters. Bonding becomes a problem if the stiffener core is a foam.

Now take a corrugated cardboard and do the same test. Align the flute of the cardboard to the 2x2. Apply pressure until it bends. You will get a nice clean crease where the cardboard meets the edge of the 2x2. This implies that you need a flexible flange (tabbings) on the 2x2.

Take a second cardboard and align the flute perpendicular to the 2x2. Bend. The cardboard is relatively stiff, it will not buckle, the gap in the center of the 2x2 and the cardboard will not be much. Less tensile strength needed for the glue. Good balance all around.

Knowing now the relative design pressures, you can make small test samples to destruction.

As AH says, develop a mathematical model, then test. Maybe some members can help develop a mathematical model or formula that we can discuss/prove.

Here is the tabulation link. Post 22. http://www.boatdesign.net/forums/cl...t-scantling-iso-12215-5-6-2011-a-47239-2.html

 

How about spending US400 for LR SSC rules? You get a software and about a foot thick of documents. When you become proficient with composites DNV, ISO, BV, GL rules will become a breeze.