Crack tip interactions for the cracks emanating from fuselage skin panel cut-outs

[banag]

When we do Damage Tolerance Analysis for fuselage skin panel cut-outs, We usually assume a primary flaw (0.05") on one side of cut-out and continuing damage on other side of cut-out (e.g. 0.005"). When we grow the crack, at some point of time, cracks on both sides of the cut-out are assumed to start interacting with each other (as 2-tip crack). As crack grows to a considerable amount, the cut-out itself will be acting as a crack. (After some time, total crack length is assumed as: [Crack growing from 0.05" flaw + Crack growing from 0.005" flaw + Cutout Dia]. We are aware that the primary objective of our analysis is to cover the worst possible scenarios. So, we have to go case by case and check for the possible worst case. But, I am curious to know when the cracks on both side of the cut-out start interact with each other. There are multiple parameters involved in that like the size of the cut-out, the loading on both side of the cut-out and symmetrical stiffener members.

We follow linear elastic fracture mechanics (LEFM) model. The stress intensity can be calculated by assuming that both the crack tips are independent and also by assuming that both tips interact with each other. The later part will give more stress intensity using hand book values. When we do FEA, we might able to find the expected stress intensity after doing multiple studies which may not be applicable for all cases.
Ultimate question: What would happen in reality? Is there any parameter within LEFM or any other model that has the parameter to get what will happen in reality.


P.S: I am new to this forum. Please elaborate your thoughts in this regard. Thanks in advance.

 

[rb1957]

asuming there isn't a separate frame around the cut-out (an alternate loadpath) ...

ok, you have a 0.05" primary flaw at one location (possibly assess several scenarios to see which is critical) and 0.005" secondary flaws everywhere else. it is important to note that the 0.005" flaws don't necessarily start at time = 0. It is a common relief from the cert authorities to limit secondary flaw growth to 1 lifetime of the aircraft, in order not to penalise good fatigue design. note too, good DTA design would have 0.005" (part-thru) flaws less than threshold.

but to answer your question, given that these flaws have grown in to the edge of the cut-out and are now growing away for the cut-out they may be more severe than a tip-to-tip crack if there are local stress concentrations. the test is to calculate crack opening size and compare with a tip-to-tip crack ... if they open like a tip-to-tip crack, then they are one !

Quando Omni Flunkus Moritati

 

[banag]

Thanks rb1957 for clarifying the exact location of the problem (no frames are near to cut-out).

Yes, the local stress concentration effect is significant closer to cut-out.

[highlight #729FCF]...the test is to calculate crack opening size and compare with a tip-to-tip crack ...[/highlight]
I am not aware about calculating this crack opening size analytically? If you have any reference kindly share it.

The same question from analysis point of view: Consider a cut-out (~2-4" dia) with fastener holes around cut-out. Two cases can be considered. 1) Growing the crack from the cut-out edge as 1-tip crack (after the crack from fastener hole edge reaches the cut-out edge) until the continuing damage on the other side of the cut-out grows for a considerable amount - either the continuing damage links with the fastener hole located near the cut-out with the cutout or it becomes through crack and 2) Growing the crack as 2-tip crack (tip-to-tip) after the crack from fastener hole edge reaches cut-out edge without worrying about the continuing damage crack length. Most of the times the second case gives lower life. To take the benefit of the independent crack tips, case (1) can be considered.
Again the same question, is it ok to take the advantage by considereing that the crack tips will act independently? Note that the crack is growing along longitudinal direction with bulging effect. Any reference to fatigue test results is very much appreciated.

 

[rb1957]

CTOD is a common fracture analysis calc. i have done it myself, but google "CTOD".

when growing as a one tip crack, are you considering it as an edge crack, ie using the 1.12 edge crack geometry factor ?

i guess a 0.005" corner flaw should take more than a lifetime to grow from the fastener hole to the edge ... i'd've thouyght it should take many lifetimes; not to suggests inadequate fatigue design or overly conservative analysis.

sure, you can model the crack growth as a 2tip crack across opposite fastener holes (spanning the cut-out), and that sounds somewaht conservative. if this doesn't work for you (a question is what is detectable damage ? sounds like it is a fastener hole cracked to the cut-out edge). i think it's somewhat reasonable to say that the two tips are not interacting when they are small, so you can reasonably grow the 0.005" flaw without considering load magnification due to the other crack. you can study this with a little FEA, modelling the crack as duplicated nodes. determining geometry factors for the continuing damage flaw (at the primary flaw location) is difficult enough ... a part-thru flaw at a hole cracked through to a near edge. your critical crack should be way longer than these lengths ... a serious 2 tip crack, spanning the cut-out.

it sounds like you're stuck with a design that has a short life by analysis. so either you ...
1) live with the short results,
2) keep analyzing, more intensive models (3D FEA ? crack tip elements, ...)
3) fatigue test.

Quando Omni Flunkus Moritati

 

[der8110]

Calculate the stress intensity (beta) factors individually, and combine by superposition. One of the downsides of software reliance is it takes away from having a "feel" for what's occurring. Eventually the [Crack growing from 0.05" flaw + Crack growing from 0.005" flaw + Cutout Dia] becomes a single crack, such as a through crack in an infinitely wide plate. If nothing else, this can serve as a boundary condition check.