Engineers? Opinions on this please...

[J Clark H356]

Originally Posted by NevisDog

I'm sorry but your analysis is so wrong on so many counts that I cannot begin to help you. If you are indeed a yacht rigger than please try to stay within your limited field of expertise.

Nevis;
I happen to be a 70 year old Architect who has been calculating beams all my adult career. It is you who doesn't use the correct terminology, and may be spreading incorrect information that will get people in trouble. I certainly do not need your help in making this analysis of this beam.

This is a simple beam with a center point load (the mast). My analysis is dead on with respect to the method of calculating the working load of this beam section and if you knew anything about moments and beams, you would understand this yourself.

I have pointed out that I do not know what the contribution of the mast loads are as I don't know and don't want the OP to find himself with a cracked beam and failed mast. You should follow your own advice as I am sure you are leading him down the wrong path as much of what you have said is false.

This is my last post on the subject. To the OP, please don't depend on forums for analysis that could endanger your life. Go hire a professional if you have doubts about this beam. You are getting a lot of nonsense on this forum.

[captlloyd]

Originally Posted by Jim Cate

All this talk of humongous loads in the rig are interesting, but do they represent the real working loads that this poor beam will encounter? Seems that the dynamic loads (excluding inertial loading from wave crashes, etc) are limited by the righting moment of the yacht. The slack bilge, narrow WL beam of these older designs aren't very stiff overall, and thus the immense loads being bandied about are pretty unlikely, especially in a cruising situation, with no rail meat upping the ante! To assume loads similar to the breaking strength of the wire, or even that of its MWL do not (IMO) represent reality. The historical evidence of non-failure even with a rotted beam suggests strongly that even this non-optimal replacement will be more than adequate.

Give poor Don a break and stop the hyperbole and extreme examples!!

Jim

x 2. I digress back to my "Lord Jim " story.

[Snowpetrel]

Just to wade in where I am not qualified to... My calcs were done using the most conservative approach, point load, simply suported beam, and no canterlever support. And the using the weaker wood species strengths. Even then the beam looked capaple of supporting 8 tonnes or so before rupture. If we add in the mast steps contribution (as a distributed load rather than point load) and the fact that the ends are supported by being rigidly fixed to the bulkhead the real loads it would be capable of taking should be considerably more than this.

If we take the rigid end connection case another failure mode to watch out for is cracks in the plywood at the top of the beam near the bulkhead support. This is possibly the highest loaded part in this case due to the solid wood core not continuing into the bulkhead, and shear stresses. If you tie the beam rigidly to the deck this will significantly decrease loads in the whole beam, and especially in this part. If not then a bit of uniglass at this point would help if you are worried.

We still havent analysed or included the contribution the aluminium makes to the beam. Ultimately it has to help, but the calcs for that are way above my pay grade, due to it being a composite structure with widely different modulas of elasticity.

[Eigenvector]

So I’ve been watching this thread since its inception with interest. You can tell the engineers’ who respond by their nomenclature, and conversely those that stayed at a Holiday Inn.

I am willing to put this to bed.

I’m willing to run an FEA analysis of this problem and post the visual stress results (for the non engineers) in the form of pictures and animations of the stresses and deflections under load.

All I need is the geometry of the beam as fabricated with all the features. I’ll assume a linear elastic analysis given a 1000 lbs load. Which is linear scalable for higher loads.

I am an Engineer, PE, Mechanical, Structural, and do this for a living, currently.

[NevisDog]

Originally Posted by sailorchic34

While I agree that one need only analyse one half of the beam ... calling it a cantilever was incorrect as was Flaming the other poster, which was not called for.

I did not call the beam a cantilever, merely stating that the cantilever formula should be used in calculating the stength of a rigidly attached (as opposed to simply supported) beam. The other poster has unfortunately made so many basic errors in his assumptions that it is impossible to list them all. I know of no other way of reassuring the OP, when such unnecessarily pessimistic assumptions could cause him to rip out a perfectly satisfactory structure and start over. Perhaps if you, with your knowledge, can put him right in a more friendly, helpful, diplomatic way then I would be most grateful to you. I don't blame him for trying but he must understand he is wide of the mark - not letting him know is disingenuous.

[Don C L]

Originally Posted by Eigenvector

So I’ve been watching this thread since its inception with interest. You can tell the engineers’ who respond by their nomenclature, and conversely those that stayed at a Holiday Inn.

I am willing to put this to bed.

I’m willing to run an FEA analysis of this problem and post the visual stress results (for the non engineers) in the form of pictures and animations of the stresses and deflections under load.

All I need is the geometry of the beam as fabricated with all the features. I’ll assume a linear elastic analysis given a 1000 lbs load. Which is linear scalable for higher loads.

I am an Engineer, PE, Mechanical, Structural, and do this for a living, currently.

That's very generous of you. Do you need more than I have already posted?

[Eigenvector]

Note to forum. The PM's have been running in the background. I am currently traveling, but should be able to post some results next week.

[NevisDog]

Originally Posted by Eigenvector

Note to forum. The PM's have been running in the background. I am currently traveling, but should be able to post some results next week.

It's a fine thing you are doing, Eigen.

I can see how extremely arrogant this next bit will sound, no doubt I'll be mauled again by certain others, but placing truth and accuracy way above caution and diplomacy (it's been my downfall throughout my working life, here goes:

Not sure how familiar with boat structures Eigenvector is so just want to list the factors that are sometimes overlooked in FEA. (I realise analysing a 'simple beam' is child's play, but the attachment points of that 'simple beam' can cause over/under-estimation of the loads it can and must carry, just as it can with simple formulae.)

1. Attachment of beam ends to bulkhead/pillars: is this a rigid connection (epoxy, +sheathing) or flexible bolted joint? (Rigid attachment will approximately double the load-bearing, all else being equal.)
2. Attachment of simple beam to deck: is this a rigid continuous connection, creating a T-section beam, or is one free to move relative to the other? (Again this may approximately double the load-bearing, all else being equal, but the analyst will need full details of deck structure in order to add an appropriate width into his analysis.)
3. What is the width of the mast step, and the width to the bolting down points of the mast step to deck? If the mast step is very rigid (most are) in comparison to the deck structure (and provided the beam is attached to the underside deck) then again this will greatly reduce the bending loads seen by the 'simple beam' and increase its load-bearing capacity by a substantial margin, compared to an assumed point loading (which is a very, very pessimistic assumption). This detail will totally alter the structure to be analysed.
4. Are you prepared to add unidirectional roving to the underside? The FEA can tell you the effectiveness of this (also I can assure you it will readily solve any possible weakness found by the FEA results).
5. Do you intend to seal the beam from moisture penetration? (Otherwise the analyst must allow for potential future degradation/delamination of ply.)


And finally, to assess the potential loads on this 'far from simple' beam:
6. Will rigging be properly tensioned by an experienced rigger? (Otherwise the beam will see unbalanced loadings that will be difficult to estimate.)
7. Do you think (as many others have suggested here) that your rigging is over-sized for the loads it may have to carry, or oversized in comparison to a rather lightweight mast section? Numerous posters have suggested rigging will never fail, the loads it can carry are 'humongous', etc. This is hogwash. Rigging is normally sized correctly by the designer, to meet the ultimate loads it must sustain. If it is clearly oversized in comparison to other parts of the boat then by all means perform the second part of the analysis based on one size smaller diameter wire (but note this means you should happily be prepared to exchange your current rigging for more lightweight stuff at next replacement - are you really that confident the designer got it wrong?).
8. Whatever size wire you decide on, this gives the maximum load your beam must carry. Anything less and your deck will fail before your rigging does - not a pleasant prospect in a storm. The loads may seem high, but equally your beam should easily support such 'humongous' loads (provided all the above factors are taken into the FEA calculations).

[Ndavies]

Looks like a really nice job! I am a civil engineer and have done a fair amount of structural work. My gut tells me this is way sufficient. What you have actually built is a "Flitch Beam" where you are relying upon the composite strength of the wood and aluminum. The key thing is to ensure they act in a composite way and this is achieved by adequate bolting - which it looks like you have done a good job of. The only improvement I could think of would have been to try to incorporate an angle section (L) or T so that you get the benefit of the strength of the flange and not just the web. However, since you are creating an arch and don't have a flat top either - this would be difficult to achieve. In the case of your beam, the stiffer member is likely carrying the majority of the load (the aluminum) and the wood is providing lateral stability (i.e., preventing the member from twisting or buckling).


If you want to check things as you load it up, I do like the suggestion of measuring the deflection under load. For serviceability, deflection of a simply supported beam is typically limited to L/360....however, at ultimate load, deflection would be greater than that. I suspect you will be well under the serviceability limit, but it may provide some re-assurance to measure deflection (probably using a dial gauge) as you tension the rig.


Good Luck!
Neil

[Eigenvector]

NevisDog,

I get what you're saying. My intent in volunteering to do this is mostly informative to the community and especially for those that have never been exposed to numerical stress analysis. You can't see or measure stress directly. Stress is a result of strain (deflection) x Modulas (stiffness). For the analysis of curved, discontinuous surfaces, holes ....etc.....geometry, this is where FEA (finite element analysis) is comes into play.

Since I'm on the road I can't respond to all your questions. But you are spot on, the fabricated beam is just a means to disperse the mast loads back into various structural areas of the boat, and hopefully in a pleasant manner. The boundary conditions and constraints are just as important as the loads.

When it comes to numerical analysis, there are now wrong answers. The computer solves the equations without hesitation. There are however, many ways to ASK THE WRONG QUESTION. This is where you comments and observations about boundary conditions come in.

The trick in engineering is to make the problem as simple as possible and still get meaningful results. We can add fibers in various directions and all that stuff that results in an optimized beam with anisotropic properties. We can characterize the epoxy bonds by using gap elements with breakout friction, but why if we don't have to.

My intent was to just do some basic analysis just to see where we stand. I have the tools available and I can make some time. My intent is the same as when I get call as an expert. My jobs is to help the Jury understand what the analysis shows, it's an education process.

Depending on how the initial round goes, we can play some what if's with the boundary conditions to see how it influences the result. We may be surprised or scared

(I did a failure analysis once that showed that the structure should have broken while they still building it. Published material strength values are minimums, people did eventually get hurt.)

I'm not going to do a $20K design and analysis job with a stamped report. Just something that burns a few hours.

As far as me being a boat designer.......sort of, but under the waves, but that was a previous life at a black R&D facility for 10 years. Done a lot of stuff rated for 20,000 fsw. Lots of experience with Ti, S-glass, Spectra, SS all the usual salt water materials. Nobody knows everything, but people have been paying me handsomely for the last 30 years for some reason. I guess I know enough.

[NevisDog]

Originally Posted by Eigenvector

...(I did a failure analysis once that showed that the structure should have broken while they still building it. Published material strength values are minimums, people did eventually get hurt.)

Yes, I've had similar experiences, poor FEA results that required urgent explanations as to why folks were almost killed - boundary conditions underestimated, calcs can err both ways - I can see you're onto it.

[Snowpetrel]

I am looking forward to seeing the FEA results. Especially how the aluminium core and the plywood behave together, the bolt induced stresses, and how the compression and shear stresses work out. Thanks Eigenvector.

[slowpoker]

Do not know if you have already actually installed the mast support arch you made and filled the gaps to the cabin top under deck with the epoxy putty you mentioned which will also help the strength of things. It does look stronger than the original set up.
Question: Would it be very difficult at this time, to add two more of those 1/4" thick 'chevrons', one at each side of your plywood, on the exterior sides of what you have already made and use longer bolts to sandwich the whole arch together? You did say it is now a little narrower than the original or that at least you have room to widen the arch....use this for more strength.
This is actually a simple solution to all the guessing and calculations and you get to use most of what you have already built.
Do insulate the bolt holes in the aluminum 'chevrons' with some silicone or 'liquid tape' goop against those dissimilar metals from the SS bolts if you are concerned about that, since you said you can take it apart now without major destruction. Even the epoxy glue will get between the bolts and the holes when squeezing the assembly together anyway.
If the plywood gets messed up with this taking apart process, then replace the plywood with some hardwood solid boards sandwiched together maintaining their grain in a horizontal plane and epoxy them together as well. You have a perfect pattern made already, may take you a day of work to make this up.
Those spaces you made to pass the electrical wires will weaken the structure and no one has addressed this void space which this extra set of 'chevrons' will also make stronger.
Then finally dress the arch with some nice wood veneers fore and aft to cover the bolts and take her sailing in full confidence and filled with pride!. Nice work you have done.

[Don C L]

Originally Posted by slowpoker

Do not know if you have already actually installed the mast support arch you made and filled the gaps to the cabin top under deck with the epoxy putty you mentioned which will also help the strength of things. It does look stronger than the original set up.
Question: Would it be very difficult at this time, to add two more of those 1/4" thick 'chevrons', one at each side of your plywood, on the exterior sides of what you have already made and use longer bolts to sandwich the whole arch together? You did say it is now a little narrower than the original or that at least you have room to widen the arch....use this for more strength.
This is actually a simple solution to all the guessing and calculations and you get to use most of what you have already built.
Do insulate the bolt holes in the aluminum 'chevrons' with some silicone or 'liquid tape' goop against those dissimilar metals from the SS bolts if you are concerned about that, since you said you can take it apart now without major destruction. Even the epoxy glue will get between the bolts and the holes when squeezing the assembly together anyway.
If the plywood gets messed up with this taking apart process, then replace the plywood with some hardwood solid boards sandwiched together maintaining their grain in a horizontal plane and epoxy them together as well. You have a perfect pattern made already, may take you a day of work to make this up.
Those spaces you made to pass the electrical wires will weaken the structure and no one has addressed this void space which this extra set of 'chevrons' will also make stronger.
Then finally dress the arch with some nice wood veneers fore and aft to cover the bolts and take her sailing in full confidence and filled with pride!. Nice work you have done.

I have installed it, and I have sailed with it and I did push it pretty hard too, and it gave me no complaints. But then after thinking I may have underestimated the forces considerably, doubt has crept in... horrible thing, doubt!
Yes, I could add wood, aluminum or SS to the outside pretty easily, perhaps that is a possibility. That may be a very good idea. Now that you mention it, I could have SS chevrons or arches made to sandwich the whole arch together. Realistically though, even then, it may turn out to be best to just take the mast down and take out the arch and start over. It is very important to be sure all the bolt holes are lined up perfectly with no slop. Originally I clamped it all together and used a drill press to drill through the wood and aluminum together. It would be tough to be so sure of that now if I added more to the outside. I'd really want to use a drill press again. In my mind originally I wanted the support rigid directly underneath the mast which is where the bulkheads are. That was the reason for the aluminum bolted directly to the bulkheads and not just out on the sides, thus the composite sandwich.
Oh and for the electrical I routed a canal through the wood spacer between the bulkheads, I did not count that spacer as contributing to the strength of the arch except that it contributes to widening the beam.
Thanks for your input!

[NevisDog]

Originally Posted by slowpoker

... add two more of those 1/4" thick 'chevrons', one at each side of your plywood, on the exterior sides of what you have already made and use longer bolts ... Then finally dress the arch with some nice wood veneers fore and aft ... Nice work you have done.

In truth this is a very poor suggestion, from an engineering standpoint.

Think about it once more - if we add meat (of whatever variety) to the sides of a beam then the increase in strength is mainly in the web, which contributes nothing at all to the ultimate load capacity of this particular beam, the web being already vastly over strength; but if we add just a tiny amount of meat to the flanges then the strength increases exponentially.

Already the deck forms the upper flange (if epoxy puttied to the underside, as mentioned somewhere) - a very strong flange; the 'simple beam' becomes merely the web of the I-beam (or T-section) - already excessive; the lower flange requires only a single layer of fibres (glass/carbon/whatever) to increase the strength in an exponential manner.

Forget the wood veneers - let's engineer this properly! I'll shut up now, everyone hates a know-all