ISO 12215-5 problems

[killwater]

Hi
I have some big problems with scantling determination using ISO 12215-5.
First of all is measuring of the convex curvature. In different sections of the panel this value is diffrent. Should i use an arithmetic mean, the bigger value or the smaller one? There is no explanation in the standard that I could find.
The second problem is when we have panel that is longitudinally twisted. How can I figure out b and l value (unsupported panel beam and length) when the panel (without considering curvature) is three-dimensional? For example if we have panel limited from midship to bow and from gunwale to centerplane).
I would be gratefull for any help.

MK

[fcfc]

In the texte I have, it is written "there is a certain degree of interpretation".

l is limited to 330 * Lh. (9.1.2)
It is written id that l * b must match actual panel area. (9.1.3)

Now, there are limitations in ISO 12215. You cannot have panel thickness change without a stringer (say cored topside, single skin bottom). Curvature is badly handled in cored panels.

[killwater]

Thanks for reply.
Certain degree of interpretation is in regard to the angle of adjacent panels of natural stiffeners only.
I preprared some pictures to ilustrate what i mean:
So we have a twisted plate:

We need to measure the curvature but which one?

Simplified panel:

Panel with vertical projection:

Twisted panel side length difference:


Where have you found information that there cannot be thickness change without a stringer?

[fcfc]

10.1.4 Final design pressure and panel analysis.

NOTE At the time of publication,this part of ISO 12215 has no provision to consider specifically variable pressures or variable scantlings.

There is also a note 9.2.3 saying "non structural" stiffeners are not considered as a good practice. ie stiffening effect on panels, that are ignored during panel assement.

[dougfrolich]

Find the geometric centroid of the surface created by the 4 corners and measure normal to the surface. make a plane thru the mid point, and the offset from centroid, to the opposite midpoint. That intersetion is your curvature.

[killwater]

@fcfc
Yes, but you can red also that in such a case you should asses all scantlings and use the weakest one for chcecking compliance.

@dougfrolich
I think I do not understand your idea. This four points does not form a plane.

[fcfc]

I do not think twisted panels are an issue.
Just handle them as trapezoidal panels. (9.1.3).
l and b = average of opposite sides. , l the long, b the short.

For curvature, I would choose the flatter, most conservative one on the short side. (BTW whatabout when curvature is on the long side, kind of fantail stern).

NB problem arise for mainly unstiffened boats (natural stiffener mainly) for boats longer than 6m. For boats shorter, you have the drop test and no need for formal calculations.

[Alik]

Panel curvature: measurement of c should be made at same section where You measure b (width of panel). There are clear guidelines how to measure b in standard.

Twisted panels: measure straight distance between stiffeners, do not measure along panel. Also please note that length of panel does not matter if l/b is more than 2.