Structural design: unsupported length

[Vinassman]

Hi All,

Just to be sure, I would like to ask some details on the design flow for structural members.

Let's take an example: I would like to design a main deck of a pontoon, with the following characteristics.

- deck load: 5MT / m2
- frame style: longitudinally-framed type
- stiffeners spacing: 600 mm
- (transverse) deck beams spacing: 2100 mm
- (longitudinal) deck girders spacing: 3000 mm
- maximum permissible stress: 130 MPa

Ok then, here is how I would do it:

1. plating:

minimum thickness: t = 0.77 x k x s x (p / sigma)^0.5
where:
k is depending of the aspect ratio of the panel
s is the stiffeners spacing
p is the deck load
sigma is the permissible stress

in our case: t = 9.06 mm

2. stiffener:

Here the minimum section modulus depends of the boundary conditions. For our case, since they are continuous through transverse members to effectively participate to the hull girder section modulus, we will consider them as fixed at both ends.

Thus:
SM = p x s x l^2 / (12 x sigma)

where:
SM is the minimum section modulus
p is the deck load
s is the stiffeners spacing
l is the deck beams spacing
sigma is the permissible stress
minimum section modulus

in our case: SM = 85 cm3

3. deck beam

Regarding the boundary condition, the transverse members are not continuous through the deck girders and therefore, we have to consider them as simply supported at both ends.

Thus:
SM = p x s x l^2 / (8 x sigma)

where:
SM is the minimum section modulus
p is the deck load
s is the deck beams spacing
l is the deck girders spacing
sigma is the permissible stress
minimum section modulus

in our case: SM = 909 cm3

4. deck girder

Because the length of span is too big, due to a very few number of transverse W.T. bulkhead fitted, additional pillars are fitted. They are fitted every 6300 mm.

For the same reason than for the stiffeners, the deck girders are consider as fixed at both ends.

Thus:
SM = p x s x l^2 / (12 x sigma)

where:
SM is the minimum section modulus
p is the deck load
s is the deck girders spacing
l is the pillars spacing
sigma is the permissible stress
minimum section modulus

in our case: SM = 3816 cm3

5. pillar:

Taking into consideration the buckling criterion, the pillars are designed. I don't develop this part.

6. Global effort.

Then, after the local design is done, the global moment has to be taken into consideration to check whether some structural members will face buckling or not.

Please tell me if I made some mistakes here, or if the workflow is correct.

I am afraid that there is nobody around me who can tell me whether I am completely off or not. And I must confess I am a bit lost in the classification rules when it comes about structures. Their empirical formulas are just losing me . And I much prefer understand the theory behind to be honest.

Thanks a lot anyway!

Cheese,

VM

[MikeJohns]

What class are you designing to?
You can also do this from first principles. Simple beam theory and apply the FOS the rules identify .
You'll find it's very close to the formula given by the rules. It's also a good check of the output from the rules based approach.

Just calculate each successive support/framing system .

[Vinassman]

Usually I am using all the major IACS members rules. But my purpose here is to design the structure based on first principle. Indeed.

Could you please tell me whether my calculations above are correct? Talking about the boundary conditions, the unsupported length, the spacing and so on?

Thanks a lot.

VM

[jehardiman]

Actually, you have done it reversed...you did the plating first, girder last. What you should do is work up a general arrangement with the primary gross structural girder (use the peanut butter method for long's and stringers, work in block sections between bulkheads/major sectional supports), work up the primary bending stresses in the girder blocks based upon hydro and weights, then look at the arrangement of the secondary structural bending in beams/transverses/longitudinals/stringers, work up their buclking and shear, add your tripping brackets, then move to the tertiary loads in the plates last. Doing it Primary, Secondary, Tertiary you will end up with a more rigid lighter structure.

[Vinassman]

Hi jehardiman,

Sorry but I'm not sure to be familiar with the "peanuts butter method". Sounds good anyway.

Otherwise thanks a lot for the advice!

Cheers,
VM

[jehardiman]

Originally Posted by Vinassman

Hi jehardiman,

Sorry but I'm not sure to be familiar with the "peanuts butter method". Sounds good anyway.

Otherwise thanks a lot for the advice!

Cheers,
VM

In the "peanut butter method" during the initial turn through the design sprial you just "spread" the longitudinals over the shell plating "bread" to work up your required sectional modulus. Subsequent turns through the design have the longitudinals placed and individualy designed for thier loads. Generally, the shell plating is the majority of the weight if you use the typical longitudinal spacing (i.e. effective flange) of 60t. However you may find that you can get less deflection and weight if you vary the stiffener SM and spacing relative to the shell. Additionaly, the 60t is only for a Uniform Distributed Lateral Load (like the hydrostatic load on shell plating) and should not be used for decks especally if subject to point loads. See PNA Vol I Chpt 4, Sec 3.6 or SD&C Chpt 6, Sec 5.2.