Re: Re: [PSUBS-MAILIST] Ribs

The hull optimizer is in limbo right now - I'm at a point where theoretically it works, but approximately 10^13 iterations are required to produce a solution, which is not reasonable for a home computing environment. To correct this problem, I am trying to implement logic to disregard particular nonsensical combinations of geometries (i.e. stiffener geometries which are not weldable in practice, stiffener web depths which approach the hull radius, etc.), as well as editing my algorithms to be more efficient to avoid unnecessary calculation steps, in order to reduce the order of magnitude of iterations required. I don't know what the rest of you consider reasonable, but I figure if the app can plug away in the background for twenty minutes or so while I have a cup of coffee, then I'm in the ballpark. Stay tuned.

As for your other questions, I'll defer to Cliff for the ones pertaining to his spreadsheet or the article he posted. As for your general questions, it is important to understand that structural reinforcement of any type attracts load due to the increased stiffness. This is why stiffeners can work at all, since if this was not the case the shell between stiffeners would have a particular strength that could not be improved without increasing the shell thickness. In reality, the stiffened section at the rib locations acts to strengthen the unstiffened shell in between. Thus, the literal meaning of inter-stiffener strength refers to the failure strength at the weakest point (the centre of the bay), but the stiffeners contribute to determining that value. The way the ABS Rules are presented, the length of the bay is from the center of one stiffener to the center of the adjacent stiffener, but you can't physically replicate the system by using half a stiffener on each end, so trying to define the physical extents is a bit disingeneous.

An inter-stiffener stress failure (interframe shell yielding) is the failure mode whereby the shell collapses inward between stiffeners, such that the resulting deformation gives the hull a characteristic accordion shape. This is due to bending stress along the longitudinal (axial) direction. Local buckling (axisymmetric local plastic collapse, hoop stress failure, interframe shell buckling) is the failure mode whereby the shell, in between stiffeners, collapses inward along the transverse (radial) direction in a plurality of lobes (this is your number of local instability nodes). Contrast this with overall buckling failure (axisymmetric general plastic collapse, overall collapse) in which the deformation mode is similar, but occurs over the entire length of the hull, stiffeners and all, in a number of lobes corresponding to your number of general instability nodes.

Stiffener stresses, as implemented in the ABS Rules, are considered in conjunction with a particular adjacent length of the cylindrical shell. Thus, your T stiffeners are actually considered with this shell section (effectively as I-beams) for the purpose of the calculations. Longitudinal stress as referred to by ABS is a bit misleading, as this is the longitudinal stress in the stiffener, not the hull, and is actually at right angles to the hull axis. This is the limiting hoop or radial stress in the stiffener combined section. Bending of the stiffener is considered separately in concert with the cylindrical shell in the overall buckling calculation.

You are correct that the lowest allowable working pressure for any failure mode will dictate how the hull will fail, and according to the numbers you posted, general instability is not a concern. As I mentioned in my previous message, the most efficient hull is one where the maximum allowable working pressures for each failure mode are pretty close to one another, but there is nothing wrong with having higher limiting pressures for some modes. In that case, they cease to be a worry.

The difference between heavy and non-heavy stiffeners is not necessarily related to web thickness. Stiffeners which meet the "heavy stiffener" requirements as per the ABS Rules reduce the value of Lc (length between heavy stiffeners or length between ends of the vessel) for use in the calculation of the limit pressure for overall buckling. Thus, if you have a low limiting pressure in overall buckling, beefing up a few of your stiffeners to meet the heavy stiffener requirements would push that value up. Obviously, stronger stiffeners will also increase the other limiting strengths to some degree, but since that occurs only at the particular stiffener under examination, the limiting pressures for the purpose of determining the maximum allowable working pressure still correspond to those at the weakest (i.e. non-heavy). Non-heavy stiffeners are simply any other stiffeners which do not reduce the length of Lc for the purpose ofincreasing the overall buckling limit pressure.

Stiffener tripping occurs due to local buckling of the stiffener web under the imposed bending and axial stresses. A stiffener web which is very deep in comparison to its thickness is likely to buckle out of plane under load, so stiffeners have to meet certain shape limitations. As implemented in the ABS Rules, this is just a condition that must be met before proceeding with strength calculations.

Similar to the tripping check, the inertia requirements check examines the sectional moment of inertia of the combined section (stiffener plus part of the adjacent cylindrical shell) to see that it meets a particular minimum value, and this is different for heavy stiffeners than for non-heavy stiffeners.

As for plastic deformation of the hull during failure - it is quite likely that this will not result in a hull breach; however, this can not be relied upon for safety, since the exact course of deformation is not deterministic, depending on the granular structure of the steel, the as-built weld geometry, and a number of other factors. It is true that some military submarines incorporate flooding bulkheads which are designed to deform plastically in service as a single-use emergency measure, but it is not valid to assume that you can dive a hull to failure and expect to be able to recover under all circumstances. Exceeding yield deforms the steel plasticly in response to the load, which ordinarily alleviates the stress before reaching the ultimate tensile stress, but stress concentrations which are created as a result of the deformation could possibly push you over UTS, resulting in a catastrophic failure. More pressing a consideration is the fact that large scale plastic deformation of a pressure hull may adversely affect your buoyancy, so there are other concerns which may make hull integrity moot.

Note also that your use of the term "fatigue" is not correct. In an engineering context, fatigue refers specifically to the reduction in life of a component due to the effects of cyclic loading. Fatigue cracks occur due to weakening of the metal surface over time under the cyclic loading (this can be non-reversing, where the load simply varies in magnitude, or reversing, where the load alternates between compressive and tensile). A typical fatigue failure usually starts with a microscopic crack at the material surface. Due to the extreme stress concentration at the crack tip, the crack tends to propagate outward from the initiation site, advancing a tiny bit with each load cycle. As the crack advances, the remaining tensile area in the component correspondingly decreases, increasing the stress even further. Eventually, the crack progresses to a point whereby the remaining tensile area in the component is no longer sufficient to sustain the load, and the part sustains a catastrophic brittle failure. Examining the parts after such a failure, you see a very smooth "clam shell" surface centered on an initiation site at the part surface, extending some distance into the part, with the remainder of the fracture surface being rough and haphazard, having failed rapidly. The ABS Rules do not address fatigue failure other than to require that if the design life cycles of the hull exceed a certain value (determined by a calculation), then a complete fatigue analysis must be performed.

I'd like to see Cliff or Sean do a presentation on the calculator at the
next convention.

Based upon this discussion I went back and did some recalculating of my
own and notice that some of the terms used in the text book and this
discussion do not directly correlate to the hull calculator
spreadsheet. I wonder if Cliff or Sean could connect the dots for me.
For example, using the reference that Cliff gave us
http://books.google.com/books?id=rv0QXKI0HvMC&pg=PA288&lpg=PA288&dq=failure+modes+for+stiffened+cylindrical+shells+pressure+hulls&source=bl&ots=WYLCbtL-U4&sig=B9Z-NTwzBMYq-HiOetBdWuO8StE&hl=en&ei=i38BS-XEKdKonQfezd0X&sa=X&oi=book_result&ct=result&resnum=2&ved=0CA0Q6AEwATgK#v=onepage&q=failure%20modes%20for%20stiffened%20cylindrical%20shells%20pressure%20hulls&f=false

1) Where is local shell instability in the spreadsheet calculator? It
looks like it must be either lines 43-51, or 43-57? If local shell
instability stops at 51, what are lines 53-57 providing? If local
instability is 43-57, can you explain the difference between
Inter-stiffener strength and longitudinal stress?

2) Does Inter-stiffener strength literally mean between the stiffeners,
or does it include (physically) any two adjacent stiffeners?

3) What does "longitudinal stress at the frame" mean? Does this mean
hull and stiffeners, just hull, just stiffeners? Or does it mean the
stress at the flanges as measured between two adjacent ribs?

4) General Instability is lines 59-64, correct? I notice that for the
numbers I am using for calculations that General Instability is quite
large relative to inter-stiffener strength and longitudinal stress
calculated just above it. I'm getting 5138psi for General Instability,
808psi for inter-stiffener, and 883psi for longitudinal stress. I
assume this means I'm likely never going to get to a General Instability
failure because local instability will occur first. Am I right?

5) Where is "axisymmetric local plastic collapse" (from the book
reference) located in the xls hull calculator?

6) What does "non-heavy stiffeners" (from the hull calculator) mean?
Does this mean 1/4 inch rib webs are "non-heavy" whereas 1 inch rib webs
are "heavy"?

7) What does "stiffener tripping" mean? I notice mine says "Tripping".
This sounds bad.

8) What does "Inertia Requirement" mean?

9) Is "instability" and "buckling" the same thing and use interchangeably?

10) Just an observation given the photos in the book. The pictures give
me the impression that the failures shown are not necessarily
catastrophic (ie, hull ripping open and splitting). Instead, it appears
as if a huge hammer has dented the cylinders. Can we assume then that
failure of the hull does not necessarily mean a hull breach. In other
words, and I know I'll catch grief for this, but, acknowledging ahead of
time that there are always exceptions, is it expected that buckling
(instability?) typically does not translate to fatigue (in the sense
that the metal splits and allows in water)? Or is that assuming too much.

Sorry for all the questions, but I agree with Brian that this is a
fascinating topic once you get into it.

Jon

Brian Cox wrote:
> Cliff, Thank for the detailed information , I appreciate the time
> you put into explaining this. It's fascinating, and it's something I
> would like to get better educated on. These various modes of failure,
> I gather, all need to be balanced around the same ideal design
> pressure so you're not leaving yourself open to a mode of failure in
> one area. Does it make sence to have the first mode of failure be the
> buckling of the shell between the rings stiffeners? That way you get
> some warning before you reach a general instablility? I know the
> senerio for that would mean that you have already gone beyond your
> operating depth, so maybe that should not really play into it. I
> guess the goal is to evenly design around the three modes of failure.

************************************************************************
************************************************************************
************************************************************************
The personal submersibles mailing list complies with the US Federal
CAN-SPAM Act of 2003. Your email address appears in our database
because either you, or someone you know, requested you receive messages
from our organization.

If you want to be removed from this mailing list simply click on the
link below or send a blank email message to:
removeme-personal_submersibles@psubs.org

Removal of your email address from this mailing list occurs by an
automated process and should be complete within five minutes of
our server receiving your request.

PSUBS.ORG
PO Box 53
Weare, NH 03281
603-529-1100
************************************************************************
************************************************************************
************************************************************************