The $1000 Seastead Design Contest

[Jackson]

Thanks Hebrides.

Unfortunately the more I look at this skp model the more questions I have.

Why aren't the supporting legs/columns at the corners? Not only would that create a much more stable structure, but the trusses would then form a square frame around the perimeter making it far easier to design an efficient layout for any added structures. After all, oil production platforms follow this format and they know more than anybody about living off-shore.

I've never seen guywires used at this scale in above-surface marine structures (other than ships' masts of course). Guywires tend to be used to brace very slender structures which have high compressive strength, but little lateral rigidity. The pillars and trusses in the skp model however are extremely deep/broad in comparison to their length/height so it would seem more logical to provide the bracing in the form of rigid connections at their intersections. I'm no marine engineer, but it doesn't seem at all logical that the trusses aren't structurally connected to the columns other than (presumably) within the flat "raft" base. Why not design it as a single connected structure? If marine engineers did design it they appear to have gone against generally accepted practice in a number of rather inexplicable ways.

[fimp]

Hi Jackson,

@jackson said:

The guywires in the skp model make it pretty much impossible to build anything on top of the trusses and lateral guywires (as in the image) make it extremely difficult to design anything other than a series of unconnected towers with fairly small footprints. Can you clarify please?

Can you provide a screenshot of the horizontal radial wires? It sounds like an error but I don't see them anywhere.
For an example of how a design can look without interfering too much with the lateral wires, take a look at the sample design: Images and SketchUp file (which is still being polished).

@jackson said:

Why aren't the supporting legs/columns at the corners? Not only would that create a much more stable structure, but the trusses would then form a square frame around the perimeter making it far easier to design an efficient layout for any added structures. After all, oil production platforms follow this format and they know more than anybody about living off-shore.

I've never seen guywires used at this scale in above-surface marine structures (other than ships' masts of course). Guywires tend to be used to brace very slender structures which have high compressive strength, but little lateral rigidity. The pillars and trusses in the skp model however are extremely deep/broad in comparison to their length/height so it would seem more logical to provide the bracing in the form of rigid connections at their intersections. I'm no marine engineer, but it doesn't seem at all logical that the trusses aren't structurally connected to the columns other than (presumably) within the flat "raft" base. Why not design it as a single connected structure? If marine engineers did design it they appear to have gone against generally accepted practice in a number of rather inexplicable ways.

I send your question to our staff and the marine engineering company, which made the design. Please see the following initial answer to your question from our staff:
*"Putting the legs at the corners results in a larger span, that is, more cantilevering, and thus requires a bigger truss which wastes materials. Right now, the largest span is 200', which means the furthest cantilevering is 100'. If spars were at the corners there would be a 400' span and thus 200' cantilevering.

As for wires vs. trusses, materials are generally more efficient in tension than compression. Hanging things saves on materials - see the work of Buckminster Fuller."*

And more details from the engineering company:*"Semi-submersible columns are in general connected at the top (by beams under the deck, and at the bottom, using large pontoons. We're not, so increasing the separation makes it difficult for the structure to wistand the "hogging" and "sagging" modes due to wave loading. Moving structure outside the column also helps in distributing the load directly above the column center , and minimizing the offset moment, again because we only have one connection point.

The cables are there to reduce weight and redistribute the weight again above the buoyancy.

Why not design as a single structure? Mostly because it would be too large for most shipyards in the world, and we want the flexibility to use smaller fab yards, and improve economics, in the assembly and integration scenario.

Lastly, this is not an oil and gas platform. The design basis requirements are different, and there is a coolness factor that needs to be included. We're trying to combine an more efficient structure, a larger footprint, an optimized payload, good motions, without increasing cost."*

I hope this answers your questions satisfactory.

[Jackson]

Thanks for your reply Lasse.

See images of radial cables in the skp model below.

The design you linked to shows the cables attached not to the columns, but to bowed vertical beams which are attached to the columns. Following your engineers' logic that cables are an inherently efficient means of bracing a structure (although I'm not aware of any suspended structure on this scale which is suited to a moving foundation, i.e. ship or oil platform), it would seem that the material savings allowed by the use of cables are then spent on the extra bowed vertical beams. If I had the choice I would take the material costs from the extra vertical beams and the cables and spend it on rigid connections between the trusses and columns, thereby freeing up almost all the space above the trusses for development.
Unfortunately it's extremely difficult to even begin designing (even a sketch design) when the underlying structure is according to your engineers both a) very well thought through and therefore a coherent structure in which every element is dependant on another, but also b) large elements of the structure can apparently be moved around to suit planning. I understand (in spite of the comprehensive engineer's report) that the whole idea is fairly conceptual, but a moving target is very hard to hit.

@unknownuser said:

"Putting the legs at the corners results in a larger span, that is, more cantilevering, and thus requires a bigger truss which wastes materials. Right now, the largest span is 200', which means the furthest cantilevering is 100'. If spars were at the corners there would be a 400' span and thus 200' cantilevering.

I follow the logic of reducing the span of the trusses, but surely this is at the expense of stability? As I said, I'm no marine engineer, but the same principle applies at sea as on land: the farther apart the legs/supports are the lower the centre of gravity and the more stable the structure. Besides, in the current structure the trusses aren't cantilevered- they're suspended from the columns via the cables. In section the platform is effectively a suspension bridge, but AFAIK suspended structures rely on an extremely stable foundation whereas cantilevered structures are better suited to less stable conditions.

Sorry for banging on about this, I'm just intrigued about the many ways in which this structure strays from standard engineering practice.

[hebrides]

I want to modify the model (move the cables) to accomodate my structure but need a few hints to hobble up something that is feasible.
I would like to know what percentage of the total weight of the truss system is supported by the wires. I would guess that it is less than half. This cable arrangement would have affected the stability testing to a great degree as it sends a fair portion of the weight to the upper portion of the columns.
Seems to me the centre of gravity is very high. Is there a balust and if so is it used to fine tune the stability?

[fimp]

Hi,

Jackson:

@unknownuser said:

The layout of the guywires (horizontally radial) in the downloaded skp model appear to differ from those of the image on your website (lateral).

@unknownuser said:

See images of radial cables in the skp model below.

I don't see how they differ from the images on our website. Except if you mean the image of the sample design which differs only because the wires are attached to the vertical beams (which in turn connect to the spars) instead of connecting directly to the spars. This is an acceptable minor alteration of the design and, while practical in some cases, doesn't change any properties of the platform.

Regarding your further engineering inquiry, unfortunately I cannot pass on all questions to our engineering team. I ask you to carry on the discussion in our forums where quite a few users have engineering knowledge.

hebrides:

@unknownuser said:

I want to modify the model (move the cables) to accomodate my structure but need a few hints to hobble up something that is feasible.
I would like to know what percentage of the total weight of the truss system is supported by the wires.

In general, we don't allow modifications of the base platform unless they are minor cosmetic alterations. How much is minor? Can't tell you exactly unless you tell me exactly what you want to change, but anything that requires calculations like you want to do is definitely too much.

• Lasse