Odd specification of FEA analysis method

Why would an FEA analysis guide for ship construction say:

"extra scantlings, such as owner-specified additional thickness, included in the vessel’s design specifications, should not be used in the finite element models"

Last sentence of Section 2:1 of ABSGUIDANCE NOTES ON SAFEHULL FINITE ELEMENT ANALYSIS OF HULL STRUCTURES .
2004

http://www.eagle.org/eagleExternalP...lFEAforHullStructures/Pub134_SHFEAGuideGlobal

 

Because you want to know whether the structure satisfies the Class minimum structure requirements as designed. Extras tend to give a false result.

If the Class rule was say 5mm, for the deck..but the client wanted say 7mm, as a +2mm corrosion allowance, then obviously an FEA will yield lower stresses with the increase in scantlings. Whilst this may sound fine...what is the stress in the structure like at the 5mm, rather than the 7mm thickness??...and that's the point!

 

Interesting topic, I'd like to comprehend it better, so please forgive the ignorance which will transpire from my question.

So, I understand that you are saying the structure has to be calculated without the corrosion allowance. Which means that thickness resulting from the Class rule scantlings do not contain that allowance, is that right? So the class scantlings should be intended as the bare minimum for structural resistance of a hypothetical ship which is not subject to corrosion, and then on top of it you add what you want to add...

Am I reading your words correctly?

Cheers

 

The concern that drove me to post this was worry over stress concentration effects of varying thickness. Sure the thicker section might be under less stress but an adjacent thin section connected to it might have dramatically increased stress and therefore be much more likely to fail from corrosion, buckling, or fatigue.

 

Classification Societies, Bureau Veritas at least, I know best, determine certain increases in thickness, of certain elements, due to corrosion that may be subjected, and which vary according to the existing fluid on either side of element.
As said Ad Hoc, the element must meet the minimum regulatory values​​, regardless the increase of thickness.

 

I used a corrosion allowance as an example, it could be anything. But yes, any “extra” that is added which is over and beyond that is required for minimum Class approval is not considered. Since if it passes a global FEA with the minimum scantlings required, as per the rules, then any addition, of whatever that may be…is only going to lower the stress/deflection. Notwithstanding localised stress raisers and fatigue of course, as that is a separate issue.

 

Rules are not always logical or good from the design point of view. They are what you need to follow though. I can see the problems than can be caused by increasing the plating at the deck without changing anything else. For example, the CG will be raised. Also, a more rigid deck structure may induce localized stresses at the sheer. ABS seems to sometimes take elements in isolation.

 

Ok, very clear. Thanks

 

in general you will not weaken an adjoining structure by stenghting some of the panels, you will not cause more load on the adjoinging structure by making some of them stronger.

The only excpetion to this is when designing for deflection, when you have a very flexible structure, and for what ever reason you over stiffen some of them, you could cause a failure to occure at the transision from the stiff portion to the flexible one since you tend to get a stress concentration where there is a discontinuity in the stiffness. It can still be done, you just have to make design provision for destributing the deflecton differences.

Most heavy strucutres however are not flexible enough to make a difference. something like a snow skis, an archery bow, or other devices that have extreme flex, it can be a killer.

 

The class society do not want extra complications due to "owners wishes"?
FEA is quite sensitive to quite small alterations in thicknesses, arrangements, etc.
Most probably the ABS want simply exclude all the possible variables out of the equation; than structure is approved to "pure" class requirements and all the "extras" are considered as corrosion addition.

For the owner, in the end effect this mean, that there is no need to replace plates/frames until all the "owner additions" wear out. Because strength is investigated and approved to class with minimum thickness.
The $ catch is here:
Lets say, your ship will need some plates exchanged after 10 or 15 years of use. With all the additions, it is possible that plate replacement will never happen. Or, at least, when selling the ship for subsequent owner, you will have a strong argument for higher price -"no major hull works will be ever necessary!".