Is bulkhead tabbing now redundant?

Groper

With respect the issue is your own understanding and expectations. I can explain very well, but from the outset you have been doggedly set in your opines and ideas of what YOU expect and what YOU want to know or hear as an answer to your questions. Thus the conversation has gone down a very linear and myopic route, owing to an incalcitrance to listen and understand what you’re asking when it differs from your expectations.

And their in lies the issues. This is new to you, so rather than listen to those that have done these coupon testing before, RX, myself and probably Tunnels too (amongst others), you have consistently avoided such admissions to prove a point of clouding the issues further from progression down this simple request.

And here in lies the issue again. Published data is for general design and first hit calculations. BUT, and it is a very big BUT (as I have noted many times on this thread), if you deviate by a very small amount from such published data, which is inevitable if you wish to use a different mixture and/materials and layup/process etc from that which is published, you shall have to arrive at your own set of E, just like Peter did (noted previously). This includes the whole QA side too. The whole point is, the burned of proof is on you not Class to arrive at the values of E you wish to use. It is that simple.

If you wish to put your prejudice to one side, and listen to several very experience people posting on this thread, you will.

 

Ok now who is not listening here? I told you i already fabricated a 6m length of triangle section composite beam, loaded it, and measured the deflection! The agreement was very good with the E i used in the equation to predict deflection, which was 43 Gpa for my e-glass/epoxy laminate. To get the predicted deflection to match the actual measured defelction, i had to increase the E to 45 Gpa.

Are you saying i need to do another test to determine E? If so, why? The material we are working with has not changed. We are not interested in triangle beams tho, do i need to make a T section test peice, which would simulate a peice of hull with a stiffener behind it? Or does the amount of epoxy adhesive in the fillets change the E?

 

Hmmm…a never ending circle of repetition.

If you do not want to use an independent test facility to establish you values, that is your prerogative. Thus, do not expect anything that you have done to be accepted by a surveyor.

 

I dont care about surveying or anything else for that matter, this is an exercise on an internet forum, to determine if it is at least in the ball park of being doable. If it is, then we progress further and get the independent testing done, along with finalizing all the other design engineering for a future boat design - which doesnt exist yet!

For now, just need to know if we are in the ballpark, if not, we can save a fist full of cash, paying for testing which we know is a WOFTAM!

And you think im stupid... ever heard of the pot calling the kettle black?

 

Groper,
I was leading you to that but Ad Hoc beat me to it. AH is a degreed engineer and sometimes a university professor (sometimes also my mentor as his post leads me to take up the challenge) so he thinks like that. EXACT. Nothing more, nothing less. His vast experience is such that he knows the answer before we even start calculating.

I was going to suggest to you to make a full size test panel as it is practical since you are working with a small boat. Do it the way you would laminate it. Since you have already made the flexure test, make another one with the joint I described at ¼ span. Make another 2 test piece of the keel part with the T section or center girder and the floor attached (one with join, another without). About 2 to 4” wide test piece.

Invert the partial (keel) structure and load it by slowly adding sandbags until you see a failure. Note the failure if it is a joint, delamination, cracking, ect. Jot down the amount of weight and calculate the pressure. Load over area of the test piece. This is your “yield load” or “allowable load”. Any mode of failure is a failure and has no meaning at this point. Divide this by 3 (Factor of Safety) to arrive at a workable load. Still unusable at this point.

Next, you must learn to compute the “Design Load” for each zone in the boat. Keel pressure, bottom panel pressure, side panels, coachroof, ect. Marine Composites (downloadable free) shows the formula to calculate pressures. Class societies also published the formula for calculating local loads (DNV, LR, GL are available in this forum) with all the adjustment factors for the type of boat. At this point, you should be able to calculate the displacement of your boat, vertical acceleration, know your desired speed, area of operation, ect.

This will be your calculated “Design Load” and you can now compare if your workable load in the test piece is equal to or greater than the design load. This puts you on a comfort zone. If the derive test load is lower, you have reduced the factor of safety and your structure is likely to break.

If you are making production boats, hire an NA, or if your boat was designed by an NA but you are not comfortable with the strength or the quality level of the yard works, then it is time to make coupon test. Test pieces are numerous. 7 pieces for one test method, throw the highest and lowest values and average the remaining 5 pieces. Make one set for tensile test (from which you get the modulus, Poisson’s ratio, ultimate strength, yield strength), one test for compression (ultimate compression, modulus, yield strength), one set test for flexular strength.

Armed with these statistically derived material properties, you will have to learn how to calculate panel strength/deflection, stiffener strength/deflection, panel with stiffener, ect. It takes time but the Class Rule book will guide you. Give up? Hand over the test coupon results to a composite engineer. He will ask for it even before he even starts holding a pencil.

Cheers.

 

Stumbled upon this, which is what I was looking for all along...

Shear stress in the glued plane of 2 perpendicular elements....

 

Ok so further to this, some basic calculations id like to run past you, to verify no mistakes.

First need to determine moment of inertia, I. Would it be fair to assume that if we are calculating for a hull to stringer joint, with 450mm stringer spacings, that we assume a T section of 450mm width plate, and we shall use stringers of say, 100mm depth in the center of the plate. Therefore, if we duplicate this section over and over, we essentially end up with a cross section of the entire hull with its stringers @ 450 crs?

So, for this T section i use a top plate of 450mm width and 16mm thickness (the hull plate) and a stringer of 16mm width and 100mm depth.

So for this T section i get an Area centroid of 97.45mm from the stringer top ie. 0.09745m. And thus I is then calculated as 5.890715*10^6 mm^4 or 5.890715*10^-6 m^4 to get the units into Meters. I wont go into the math for calculating this moment of inertia...

Then we take the equation from the above video, shear stress = VQ/It

Lets assume a load, V, = 1000N or 100kgs - we would first need to consult class rules to calculate the design hull pressures if we were doing the real thing, but for excercise lets assume 1000N.

So for the Q, we find A and y, so use the area of hull plate = 0.45m*0.016m = 0.0072m^2. And the y , distance from centroid, = 0.09745m-0.108m = 0.01055m
So Q, = A*y = 7.596 * 10^-5m

Thus the stress can be now calculated via VQ/It
= 1000N * 7.596*10^-5 / (5.890715*10^-6 * 0.016)
= 805 056 pascals or 0.805Mpa shear force acting at the 16mm stringer edge - where the bond line is.

Considering the rated shear strength of epoxy glue is 16Mpa for bond to epoxy e-glass substrate and verified by testing -> http://www.atlcomposites.com.au/files2/epoxy_products.adhesives/techniglue-hp_r15-eng.data.pdf

Thus the bondline should be able to handle this load with a safety factor of around 16.

If we added a radius fillet to either side of the bondline, equal to the stringer thickness per side, we would then have 3 times the area and thus the shear force per area would then be divided by 3 no? in this case we have more safety factor...

Is this correct so far?

Next we need to calculate the actual hull design pressures and do the above with those?

Is there anything else i need to consider - besides the testing of my own we already spoke of?

 

Groper

What those videos don’t explain is what/how an I-beam or Tee section actually works in practice.

When a load is applied to an I-beam, the flange carries the tensile load and the web carries the shear load. So if you have a vertical web with a horizontal section added on top to make a Tee, and then this is placed onto the hull plating to make an I-beam sectional shape, the shear is carried by the vertical web section only. The hull plate and the flange (horizontal part) of the Tee carry the tensile and compressive load when an applied load produces a bending moment on said section. Unless the section is so stiff there is no bending just pure shear.

The shear stress is simply the applied load divided by the area of the vertical web.

 

HPR25 adhesive

BTW, here is an adhesive recommendation I have put on my list. This is a toughen epoxy product. I have asked them to compare with a Plexas type product.

http://www.boatdesign.net/forums/fiberglass-composite-boat-building/delemma-hull-deck-joint-30869-2.html#post640391

 

Protruded Fiberglass Beams

I wonder if anyone has considered the use of protruded fiberglass beams for internal framing on glass or metal hull vessels?....I-beams, H-beams, T-beams

 

Yes, been done before, about 15years ago by Maunsell in the UK. I was on the panel overseeing the project.

 

Protruded Beams

What type of vessels was it principally aimed at? I saw one reference here on the forums from a fellow called EuroCanal I believe his call sign name was. Was it canal boats?

I believe I have seen some other reference to its being used in support of composite decks?

 

where do you get the data for slamming loads?
Is this the great unknown that causes some modern race boats to fail under stress?

 

Any. It wasn't type specific, other than any commercial vessel.

All commercial vessels must be Classed.

 

The class rules have methods for estimating slamming pressures, some of them are rather complex to calculate. Race boats are generally not built to class rules however, they would end up far heavier and no competitive if they were. So they are designed on the edge of failure, thus we do see failures when conditions become extreme such as the Volvo 70's surfing down 10m seas at 40kts and crashing through waves, impressive stuff....

If you want to see how the pressures are calculated, you must read the class rules, only online free stuff available that i know of, is the BV stuff located here http://www.veristar.com/wps/portal/!ut/p/_s.7_0_A/7_0_19FU?aboutArticle=aboutbvruleshome