Engineers? Opinions on this please...

[NevisDog]

Originally Posted by Don C L

BTW, I am now not sure what the point load will be but I thought, or estimated, up to 10K# ..... Additionally, unlike a simple I-beam, the chevrons will have to go out, pushing the bulkheads apart, in order to collapse, no? So now I am wondering, by cutting the original arch out, have I weakened the cabin top to the point that it might actually push the bulkheads apart a bit? ...

Go back to Eigen's calcs. With the 1,000 lb load, the movement was .003" (case 2, case 3 similar), which for my metric brain is less than 0.1 mm. Multiply that by 10 and we still have only 0.75mm: sorry Don, but this ain't going to push bulkheads apart, this structure fails before any movement is visible, or even measurable. Chevrons actually weaken the structure (refer Eigen - but easily corrected if necessary, just by forming it into a simple I-beam - I-beams don't have to be 'I'-shaped).

For those still having difficulty visualising the 'humongous' loads we already discussed (which some here don't yet believe we've even considered), the loads imposed on a yacht's mast and rig are by far the simplest to calculate of any part of a yacht's structure. No need to make it sound complicated - it isn't. Picture this:

Static: no one wants their rig to fail because they grab the main halliard and haul down to a dock, say to scrub one side of the bottom. Anyway, that's the sort of load your mast will routinely experience from the normal wind loadings on a blustery day. In my case, 2 tonnes of lead ballast at let's say 1.5m below the centre of buoyancy, which happens to be the same distance from mast step to chainplates, hence 2 tonnes is the downward force applied by the rigging on the mast step (a little less due to shroud angles, a little more due to rig weight, but we'll keep it simple). 2 tonnes also happens to be the SWL of my shroud rigging (10mm upper plus 8mm single lower), so the designer got it dead right. So, picture 2 tonnes sitting on top of your mast pressing down, and that's just the static load the mast step must support. I guess Don has worked out his boat is quite a bit lighter than mine, and that's fine.

Dynamic: now take the biggest sledgehammer you can heft and smack the side of your mast, right at the top - don't worry about denting it. Inertia prevents the yacht from leaning over, while the rigging prevents the mast from leaning over and instantly transfers the load down the mast to the mast step; the load can be quite a bit greater than the static loads. This is what happens in a knock-down. You don't want this to happen more than once in the 10 year life of your rigging, but equally you don't want your deck collapsing when it does happen - better that the rigging fails. So the max load that can be transferred to the mast step is the MBL of your rigging wire side stays (the shrouds - we've all seen rigging fail in this scenario, right?), which for some crazy reason is six times the SWL, for stainless wire. So the wire won't fail until it reaches six times the designed load of 2 tonnes (approx - 13 tonnes in my case). Now the crazy guy who built my particular boat went one size larger on the mast section, so no chance of that failing, so I'm stuck with 13 tonnes loading on my mast step if ever this yacht is knocked down. Having experienced a knock-down once in my career, the last thing I want to worry about is my deck collapsing, so I'm mighty particular about this structure, but fortunately, as I've said, it is perhaps the simplest part of a yacht to predict maximum loads.

If you're heading deep offshore like Barny, then allow for the dynamic loads (as the original small boat design may not); if you're coastal sailing then forget the knock-down scenario and add a small safety margin for luck.

[Don C L]

Hello all: Eigenvector gave me the ok to share the email exchange we have had lately for those interested. I'm learning a lot!

me: I was just wondering how much I can expect the aluminum alone, with wood only as spacers, to hold since I wasn't figuring the wood would be as strong as the aluminum. But really on the whole I am wondering if it is up to 10K lbs as is.

EV, Don, it's really close to a 10,000 lb load per my analysis if you didn't bond and only bolted through the bulkhead. Normally there are margins built into the stress values to account for variations during fabrication and fatigue loading. For a non-bonded to the bulkhead you are running right on he edge with no margin.

Earlier you mentioned that you put the bolt hole right at the high stress area. This is not necissarily true. In general every time you put a hole in a stress field, you cause a stress riser of 2x.
A good indicator of the maximum vertical load would be based on the breaking strength of either the port or sbd shroud cable.

If you look at the last figure in my report, the aluminum is not displayed with the associated stresses. In this image, the aluminum is trying to cut through the plywood.

As NevisDog said early on, the analysis shows what the stresses are at given loads. Unless we know what the loads are, I don't really know what to recommend.

Feel free to ask more questions.

me: So as it stands, it should be ok for a load of 10K#. But if there is a load, a dynamic load I presume, of over 10K# then I may expect to see failure at the stress points noted in the report? I am not even sure if my shrouds can handle 10K#. I'll ask my rigger friend.
Now if a hole is located in a stress field, does the bolt that is in there provide some support for the metal or wood, as opposed to being an empty hole?

So I was looking at Flitch beams since someone mentioned them. If, or perhaps when, I rebuild this, I am guessing that doing it over as again as a Flitch beam is not the best idea, is that right?

EV: Should be OK to 10K. If you have a rigger friend, as for the estimated load by the vertical mast load. It's easily calculated from the shroud tensions but you have to know all the angles and dimensions to do it.

Think of it this simple way when it comes to dynamic loads. Lets look at a boat from the stern with two shrouds, port and stbd. They both have tension of say 1000 lbs. As the sail generates lift, it transfers this into the mast which is stayed by the two shrouds.

As the load increases, the leeward shroud load decreases tension while the windward shroud stays the same. This happens all the way until the leeward shroud goes slack. Only then does the windward shroud load increase. Thus the vertical load is constant even though the wind changes. This is a simplistic example but generally correct.

As far as the bolts and holes it complicated to visualize, but no, the bolt never replaces the material drilled out. The is always going to be a stress riser caused by any geometric discontinuity within the stress field. This is because you cant transfer load/stresses in all 3 directions across a contact interface. They would have to be welded to achieve a 3 dimensional bond.

The Flitch beam is similar as described previously in one of my emails. The wood is there for lateral stability of the steel plates. The steel or metal is allowed to flex and move independent of each other since there is no bonding between them. In your case you bonded them together which forces the aluminum to try and flex as much as the less stiff wood. You would have been better off not bonding the aluminum to the wood and not insetting (confining) the aluminum into the wood.

This being said, with a beam such at a Flitch, the beam capacity is determine solely on the steel portion, and the lateral stability (buckling constraint) is provided by the wood.

The next time you are in a warehouse such as lowes or home depot, look up at the exposed beams. Every 10-20 ft you will see some supports that run perpendicular to to the beam for lateral support. These are to keep the beam from buckling by twisting out to their weak direction under load. I our case, the wood provided this support.
This is easily shown by pushing down on a yard stick supported by the ends only. Every time you push down in the strong direction (wide side up and down) it will twist out to the week direction. If you lay it flat, it can't twist out to the stronger direction.

Hope all of this helps.

me: Yes, that definitely helps a lot.
FYI my rigger friend just sent me the figures, so it looks like I might hold off on rebuilding, as inelegant as the arch is...I now see..

My shrouds hold 6000 lbs. each, but he figures the highest dynamic load I'll find on my boat is 3000 lbs. I upgraded to 7/32 from 3/16 to reduce stretch.

[NevisDog]

Originally Posted by Don C L

Hello all: Eigenvector gave me the ok to share...
My shrouds hold 6000 lbs. each, but he (rigger friend) figures the highest dynamic load I'll find on my boat is 3000 lbs. I upgraded to 7/32 from 3/16 to reduce stretch.

Thanks for sharing Eigen's invaluable insights - very educational!

[Eigenvector]

OK. I'm back for a while. One comment to some of the previous comments when it comes to FRP (Fiber Reinforced Plastics). These are engineered materials with nasty directional properties and very difficult to analyze. That's why most of the time you see mat utilized or a balanced weave to try and get at least the planar directions with equal properties.

I've worked in the areas of engineering compelling technology (Price is Secondary to the Performance). This is reserved for the AC boats etc and Military applications.

Just to back up some, It seams like most here have a good understanding of a beam under bending. The normal compression on top and tension on bottom sorta stuff. However, we are getting close to the D/W (Beam Depth / Width) ratio where the it become a "Deep Beam". Short Deep Beams are dominated by the Shear Stress and not the bending Stress. The implication of this is that having a weaker core (Where the Shear is) hurts you.

Additionally, running fiber top & bottom, skinned core is cool and all that BUT you have to be careful that the compression side is still adequate. These engineered materials also have different properties in compression than they do in tension. Fibers suck in compression, is only the matrix that keeps them from splaying out.

Then we go the other way say in the instance of a mast which is a long slender beam. It will buckle under compression long before yielding of the material. Thus intermediate supports, as in the analysis done referring to the lateral support provided by the plywood.

Oddly enough, buckling is a function of Modulas of Elasticity (Stiffness) and NOT the strength of the material. I've designed many things where the use of 316L Stainless resulted in a lighter design than did TI 6-4. Only because the first failure mode of the object was buckling and not stress.

The first failure mode in the analysis that I did was the plywood and not the aluminum chevrons.

[NevisDog]

Originally Posted by Eigenvector

... Short Deep Beams are dominated by the Shear Stress and not the bending Stress. The implication of this is that having a weaker core (Where the Shear is) hurts you.

Additionally, running fiber top & bottom, skinned core is cool and all that BUT you have to be careful that the compression side is still adequate. .....

The first failure mode in the analysis that I did was the plywood and not the aluminum chevrons.

Thanks Eigen, but most of that was over my head so, unless I'm unusually thick (and it's beginning to look that way), others may feel likewise. To be clear,
1. are you suggesting that solid ply is not strong enough in shear to form the core (web) of a very short, deep, composite beam formed by a GRP deck above and unidirectional fibres in tension below? A simple, solid ply (or timber) beam here will fail in shear? Seriously?
2. Are you saying a GRP deck cannot take compression? Surely this is its normal mode, with balsa below and an inner/lower skin providing tension?
3. We know ply is weaker than alu, but doesn't the alu concentrate the load, causing the ply to fail prematurely? If not then I've completely misread your FEA, sorry.
Darn, and I really thought I'd learnt something here.

[Eigenvector]

Originally Posted by NevisDog

Thanks Eigen, but most of that was over my head so, unless I'm unusually thick (and it's beginning to look that way), others may feel likewise. To be clear,

Sorry Guys. I read my own post afterwards and it barely made sense to me. My comments are more to remind us all that that you really have to be careful with materials that have non homogenous properties.

1. are you suggesting that solid ply is not strong enough in shear to form the core (web) of a very short, deep, composite beam formed by a GRP deck above and unidirectional fibres in tension below? A simple, solid ply (or timber) beam here will fail in shear? Seriously?

Not at all. What I was trying to convey is that just because we call it a "beam" the highest stresses within the beam are not "compression top" "tensile bottom" for "deep beams". These stubby beams act more like brick and are shear stress dominated and not bending stress.


2. Are you saying a GRP deck cannot take compression? Surely this is its normal mode, with balsa below and an inner/lower skin providing tension?

Again sorry for the confusion.........When I was talking about compression I was talking about compressive stress which is equal and opposite to the tensile stress in a beam top and bottom under simple bending.

Glass has a very high compressive stress. However when you start pushing it it wants to buckle. It is restrained from buckling by the epoxy or similar matrix.

An extreme (but dumb) example is a loose fiber sling used in rigging. 1000's of loose fiber wound in a loop with a cover. Very strong in tension / no compressive value what so ever. Soak in Epoxy, Still strong is tension and 1000x better in compression.

The major reason for the foam core is to separate the stress bearing members to increase the area moment of inertia which is used in the bending equation. Stress = Moment x Outerfiber / Area Moment.
Same reason I beams have thick flanges separated by a thin web.

(This is getting back to 1.) If we have only a vertical load and no bending (start going deep beam) the load is reacted only in the verical direction and not dumping out into the upper and lower flanges. It reacts straight down through all three members: Flange (FRP) + Web (balsa core) + Flange (web). As shown, the weakest link is the Web of the I beam or the Balsa Core. Hope this clarifies.

This is also the beauty of Engineered materials and especially FRP. You can change the material properties on the fly as a function of what loads/stresses exist at that particular region. Even to the extent of Green Curing the the resin (not a total cure) in certain areas of our objects to increase the toughness while sacraficing some strength in compression.


3. We know ply is weaker than alu, but doesn't the alu concentrate the load, causing the ply to fail prematurely? If not then I've completely misread your FEA, sorry.

You are correct. The big issue in beam presented to me for analysis is that the aluminum was not only bonded to the plywood but also inset. I assumed it was bonded. Because it was inset, the aluminum was trying to chisel it's way out of the cavity because of it's superior stiffness.

I'll pull up the analysis and animate the stresses which will will yield a lot of info to all. I'll post a video back here.


Darn, and I really thought I'd learnt something here.

Hopefully not true, I just didn't do a good job explaining. Hope this clarifies.

[Eigenvector]

Guy's, I just noticed that the Quote Message thing didn't work like I thought. Most of my dialog is contained in Nevis's quote above.

[Don C L]

Originally Posted by Eigenvector

Guy's, I just noticed that the Quote Message thing didn't work like I thought. Most of my dialog is contained in Nevis's quote above.

I put your words in blue there for clarity.

[Eigenvector]

Guy’s

Here two animations of the same model while the load is increasing from 0-1000lbs. Both of these are of the same analysis with all the components, I have just selectively turned off the results for some of the components so you can see the interior stresses and see the difference in who’s taking all the stresses.

I tried to pick a good view to see the holes and stresses with a single orientation. As a result, these are not the same stressed areas as identified in the report as the critical or governing stresses..

In the video below, the aluminum chevron is visible and embedded into the plywood cavity..

https://expirebox.com/download/07c3f...ad37fefed.html


In the following video, the aluminum chevron is hidden and you can see the cavity in the plywood.

https://expirebox.com/download/6f353...07f56e53e.html

[Don C L]

Originally Posted by Eigenvector

Guy’s

Here two animations of the same model while the load is increasing from 0-1000lbs. Both of these are of the same analysis with all the components, I have just selectively turned off the results for some of the components so you can see the interior stresses and see the difference in who’s taking all the stresses.

I tried to pick a good view to see the holes and stresses with a single orientation. As a result, these are not the same stressed areas as identified in the report as the critical or governing stresses..

In the video below, the aluminum chevron is visible and embedded into the plywood cavity..

https://expirebox.com/download/07c3f...ad37fefed.html


In the following video, the aluminum chevron is hidden and you can see the cavity in the plywood.

https://expirebox.com/download/6f353...07f56e53e.html

I haven't been able to open it on my two old Macs or my iPhone yet. Can you give me the Cliffnotes version? Did it hold 10K?

[Eigenvector]

Same analysis just a different way of showing it. I think you're OK.

[NevisDog]

With this technology restricted more to aircraft wings, nuclear submarines, or perhaps America's cup boats, in my day - nice to see it used on Don's mast support.

Don't know what stresses the colours represent (darned imperial measurements - who uses them nowadays?) but if this is only from zero to 1,000 lbs load, then ... ???

Nice work anyway.

[Eigenvector]

The highest stresses as represented in Red are as follows per 1000 lbs load:

13.79 MPa - For most of the world

2000 psi - For the non adopters

42 1/3 fulrongs/fortnight - For those areas with legalized weed

[NevisDog]

Now I'm very, very comfortable with Don's design - finally I can see why he did what he did. (Different, but not stupid)Thanks Eigen - that was far from an easy problem to solve.

[Don C L]

Thanks Nevis! I was really not sure there for a while! I mean about that stupid part. Like I said I can see how could have, and can, pretty easily re-do it and improve it, but for now, today, I'm going sailing! Btw I added some epoxy and two new supporting blocks and I cranked down on the rigging and so far no measurable changes. We have a decent breeze today, I'll report back if I see anything.