Ferry Dock Collapse 13

[dik]

"At least seven people were killed and several others injured Saturday after part of a ferry dock collapsed on Georgia’s Sapelo Island, according to the Georgia Department of Natural Resources.

It happened as crowds gathered on the island for a celebration of its tiny Gullah-Geechee community of Black slave descendants.

At least 20 people were plunged into the water when a gangway collapsed on the visitor ferry dock shortly before 4 p.m., Georgia DNR Capt. Chris Hodge said at a Saturday night news conference. A McIntosh County commissioner previously said a boat hit the dock but a DNR spokesperson later told The Associated Press there was no collision and it is unclear why the dock collapsed."

https://www.cnn.com/2024/10/19/us/sapelo-island-georgia-boat-dock-collapse/index.html

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Sym P. le]

A quick look yields the Sapelo Purchase Order.

I outlined the timeline of project progression earlier (Page 5, Wednesday at 5:21 PM). This is a quote and P.O. for the Meridian landing (Phase III), not the failed Marsh landing gangway. The question is whether the documents and overarching agreements pertain specifically to the failed structure. The previous invoice I posted indicates the contractor acquired the Marsh gangway to fulfil terms of their contract. All of these documents are valuable to understand the environment in which the contract was hatched. The quote you posted includes a price for stamped drawings and the P.O. includes same.

 

[Garbage1]

SymPle sorry if I reposted anything you already posted, but I must admit this Web Site makes it very difficult to keep things straight. You only get a few posts on anyone page, and each page loads slowly, and you have constantly jump back and forth between way too many pages to find previous posts then back to last page where you are trying to post a response! The design waste vast amounts of screen space, and repeats the OP at top of every page which waste more space...... Once I read thru all the posts, I have not bothered to go back and try to ensure I have not duplicated some previous post.......
Then a lot of space at bottom of page is wasted with so called "similar threads".
Also when logged in there is a post page at bottom of pages on each page, which is confusing at times.
Lots of Continuous improvement effort required, if that is possible with base software.....
Does not appear a lot of seasoned users/customers had any input to the design of this (whatever you want to call it)......

 

[Sym P. le]

No worries, this new forum does not work as efficiently as the old, at least from my user perspective. It is much more difficult to track long threads and even referencing previous posts takes forever, at least if I want to reference the poster and time of post. Thus when I state I posted "earlier", it's because I'm being lazy, not intending to be terse. That reference was to Page 5, Wednesday at 5:21 PM.

You didn't double post. I was trying to re-emphasize the progression of the Sapelo Island project as a whole through various phases as quoted and then executed.

Regarding "SECTION 05555 - ALUMINIUM MARINE STRUCTURES", that was an addendum (Addendum No. 2) to the Specs for the Phase II bid package. It created some confusion with the "SECTION 355113 - CONCRETE FLOATING DOCK SYSTEMS AND GANGWAYS". The distinction is that Crescent also has aluminum docks available and I believe they are used for smaller installations. The Bellingham concrete docks are preferred for larger heavy duty installations. Both are envisioned as systems c/w gangways etc.

Phase II Specs and documents are found in the folder "Bid Documents" (They refer to Marsh and Meridian as they were priced together before Meridian was deducted). The Phase III Specs and documents are found in the folder "Meridian/Bid Docs" (When the Meridian project was rebid).

 

[LittleInch]

Each post has a number in the right hand side so you can just use that?

So it seems to be 20 posts per page. I'm asking for the number or page to be able to be varied so you vac choose to get 20 or 50 or 100. Hasn't happened yet but might.

On a phone it loads much faster than before this way so that is better.

On the disaster, I think it will be found that the gangway was longer than ones they normally produced to limit the angle at low tide and no one actually designed it, they just bought a stretched 60 ft one, and never realised the trusses were different.

 

[Sym P. le]

It's not so much that the truss type was different (the Howe truss is actually more suitable than Pratt), it's that an "L" or inverted "T" has limited viability in any circumstance. That two different types were employed in essentially the same project just makes the head scratching more ...

 

[Garbage1]

It's not so much that the truss type was different (the Howe truss is actually more suitable than Pratt), it's that an "L" or inverted "T" has limited viability in any circumstance. That two different types were employed in essentially the same project just makes the head scratching more ...

Howe more suitable for aluminum than Pratt??? My take is Howe is better suited for wood with larger diagonal members in compression, and Pratt is more suitable for aluminum/steel with longer smaller diagonals in tension??? EDIT: In terms of efficiency and lower costs a Pratt looks the winner for aluminum. My concern is Pratt sized members in Howe configuration will not support equal loads. Perhaps Howe distributes loading better under certain conditions. Either one can work if members sized according to configuration for specific material used.

I will attach an email string between Crescent Equipment and GaDNR. There were prior complaints that caused concern from GaDNR and required inspections. It is interesting to note that the loud popping and broken welds in attached email traffic was centered around 15-20' from the floating dock, or in the area that failed........
Camber process after tack welding decking in place.....see quote More interesting information contained in attached email traffic.......

Ty Barrow/Crescent Equipment Quote from attached email.
"What I would like to ask is that we let that one sit for a bit longer The decking is spot welded to the
stringers before the cambering process, and it is now out there 24/7 testing that camber. Its possible
that one section (for whatever reason) needed relief. I would like to go out and re-inspect it later this
year to make a conclusion."

 

[Sym P. le]

This is why more eyes are better. I hadn't accessed fully absorbed that file yet. I wonder what issues can arise in the cambering process.

As to the truss type, it was my impression that the Howe distributes the top chord compression across the length of the structure compared to the Pratt which concentrates the force at the center. Since buckling is likely the failure mode, that is the factor of greatest concern in this instance.

 

[Garbage1]

Camber would have pre-stressed the structure. Wonder how they accomplish this task with an 80' long dock? Hot or Cold forming?

I don't disagree with your Howe top chord compression handling, and I would argue that member was not sized properly either for the Howe configuration. My guess is actual designer did not understand difference between round and square tube load capacities.

The lack of engineering and understanding the affects of geometry changes are the issue along with quality control; rather than whether Howe or Pratt design works best in this situation.

It is also interesting that GaDNR had no engineering evaluations conducted by Ga's engineering division or their contract engineer specialist. Rather they tasked the 'biased' original manufacture to inspect his pre-fab design and built marketplace product.......When you buy a Pre-FAB product on just a Purchase Order, you get what you pay for it seems.......or in this case, more than you bargaining for.....

This image, posted earlier, of a Howe design has very different member sizing than the Sapelo 80' Howe.

 

[Sym P. le]

The supposed engineering drawings which were invoiced and paid will be an issue. I wonder if they will ever see the light of day.

 

[Garbage1]

Based upon the fact that St. Mary's investigation, conclusions and action taken has not made the light of day, I don't expect Sapelo's to make it either. IMO, the documents they are withholding will likely not shine favorable on anyone involved (Georgia Government Agencies, Crescent Equipment, nor probably the supposed engineering stamped drawings for Crescent's 'Modular/Expandable All-In-One' gangway design.

 

[FacEngrPE]

Setting camber in walkway like this can be done by blocking the lower cord to the desired shape before welding in the diagonal members. For a structure like this camber is mostly cosmetic ie it does not change the member stresses.

All of the questions regarding correct engineering calculations which could have been a deliverable, can be short cut by performing an acceptance proof test. Perhaps 2X design load.

Maritime Safety Services; INSPECTION AND LOAD TESTING OF ACCOMMODATION LADDERS AND GANGWAYS

Periodic proof load testing of a maritime vessel’s “means of embarkation and disembarkation” (accommodation ladders/gangways & davits) is required in accordance with IMO Regulation II-1/3-9 of the 1974 SOLAS Convention, and as per the Guidelines for Construction, Installation,
Maintenance and Inspection/Survey of Means of Embarkation and Disembarkation, effective January 1, 2010.

The SOLAS regulations require that the ladders and gangways should be static proof load tested and function tested following initial installation and then at 5 year intervals per SOLAS I/7 and I/8.

OSHA Code

29CFR1915.74(a)
Access to vessels afloat. The employer shall not permit employees to board or leave any vessel, except a barge or river towboat, until the following requirements have been met:

29CFR1915.74(a)(1)
Whenever practicable, a gangway of not less than 20 inches walking surface of adequate strength, maintained in safe repair and safely secured shall be used. If a gangway is not practicable, a substantial straight ladder, extending at least 36 inches above the upper landing surface and adequately secured against shifting or slipping shall be provided. When conditions are such that neither a gangway nor a straight ladder can be used, a Jacob's ladder meeting the requirements of paragraphs (d) (1) and (2) of this section may be used.

If employees had been on the gangway, OSHA would likely have become involved.

 

[Denial]

That reference was to Page 5, Wednesday at 5:21 PM

This intra-thread referencing is now quite a pain in the RRRs, and I have already raised it in another forum.
» Once a post is more than a day or so old, its date is shown but its time no longer appears. (Even more wasted white space.)
» I think, but have not yet confirmed definitively, that the date&time displayed when a person views a post has been converted to the viewer's time zone. So the date&time shown to YOU will usually NOT be the date&time shown to ME.

Grrrrr.

 

[Sym P. le]

... and the time stamp disappears so now I don't even know what I talking about. !?!?

Ok. Found it. Page 5, #97, starting with "Some clarification."

If you press and hold the posting number you get a link which you can apply liberaly or conservatively, which ever way you lean. Or you can just be a pirate, RRR!

https://www.eng-tips.com/threads/ferry-dock-collapse.523612/post-9032544

Edit: Fixed applying the wrong link. RRRRR!!!

 

[GregLocock]

"All of the questions regarding correct engineering calculations which could have been a deliverable, can be short cut by performing an acceptance proof test. Perhaps 2X design load."

For man rated stuff in Oz I think it is 4x by annual test and 10x by calculation. For things that don't get used much this annual test is the major contributor to their fatigue life. Not my field, I may be wrong.

 

[Garbage1]

Ty Barrow/Crescent Equipment's text with Georgia officials, post # 146, also revealed perhaps one of the primary reasons for cambering deck/beams assembly after deck was 'tack' welded to beams. The other reason likely to prevent natural droop in flexible assembly under dead loads and actual added live loads. Perhaps 'dampen' some of the 'Bounce' Feature...........

"Those welds in the picture are spot welds (literally, just a tack) to keep the deck board from ìslappingî
potentially against the stringer . We could have not put any of those tack welds and not affected the
rigidity. Combined , all those probably equal about a foot of weld.
(but it would be squeaky, and you could get metallic ping, if you heavy stepped on just the right places.)"

"Actually thatís a practice more required on docks (flat surface) The camber on the gangways, holds the
decking down fairly tight."

 

[Sym P. le]

Cambering the deck pretensions the top chord and likely has minimal effect on the deck weld connections at the bottom chord. It also precompresses the vertical elements. As FacEngrPE (#151) pointed out, the camber is within/close to the elastic limits of the material. Once diagonals are applied and the camber stops removed, the diagonals acquire an induced compression from the structure's self weight. The key takeaway being that the structure is preloaded opposite to anticipated loading.

An interesting stress redistribution occurs when an applied load is skewed to one end of the truss. (In the example below, the load arrows do not reflect the magnitude of applied load.) I applied increasing point loads to the right, as if a crowd of patrons bottleneck at the right end of the gangway. Note vertical member 30 becomes compressed while diagonals 8 and 9 become tensile. The peak compression on the top chord also shifts to the right. This is consistent with the failure witnessed in the collapsed gangway.



I'm curious why the top chords IRL all tore (including center rail) to the immediate left of the equivalent node 23/24 connection.

 

[human909]

"All of the questions regarding correct engineering calculations which could have been a deliverable, can be short cut by performing an acceptance proof test. Perhaps 2X design load."

For man rated stuff in Oz I think it is 4x by annual test and 10x by calculation. For things that don't get used much this annual test is the major contributor to their fatigue life. Not my field, I may be wrong.

That is well off base and not correct for Oz structural engineering or most structural engineer. But like you said, not your field so the misunderstanding is certainly understandable. But for an explanation:

We used limit state engineering for structural with a 1.5x for live loads (eg people) and for aluminium a material reduction factor of 0.8. Furthermore you don't often "proof test" structures. There are multiple reasons for this. It is often completely impractical and potentially dangerous if failure is ever possible (and if it wasn't then what is the point of a proof test). And also a full proof test for ultimate design loads could damage a structure.

(Some "structures" do get proof tested. Fixed cranes, lifting beams and other lifting devices. Likewise personal safety anchor points.)

 

[GregLocock]

I'm talking about gantries, cranes and hoists, not fixed structures, but again, not my field.

When using rigging equipment identifying the working load limit is essential in determining your lifting limitations. Working Load Limit is the maximum load that can be applied to a sling or component. Rigging equipment in accordance with Australian Standards will have a high safety factor of 6:1. This is an important safety feature. The safety factor helps to counter possible problems from shock, vibration, fatigue, wear, damage and corrosion. It is important that the safety factor is maintained.

RIGGING EQUIPMENT SAFETY GUIDELINES

It is important that every time rigging equipment is used, it is visually inspected prior to use. Find out the full process and what you need to consider to ensure ongoing site safety and compliance.

www.alllifting.com.au

 

[human909]

None of which is relevant to this current situation. Whether it is in the USA or Australia.

 

[Garbage1]

SymPle, I think it would be educational to tweak your diagonal layout on your Howe model to a Pratt layout, to see affects of identical asymetric loading. Perhaps your investigative mind has already done this?