Short ED, again

[koopas]

Hi all,

I am still a relatively new and inexperienced airline structural engineer. I have a question about fastener short edge distance relative to a part's edge, again (I posted a similar message a few months ago, apologies if sections are redundant)

I'd like to know if the following methodology is correct when analyzing short ED conditions.

If the short ED is parallel to the load path (part's edge is perpendicular to load path), I'd use the empirical formula for shearout:

Pall = Fsu_material * t * (2ED + 0.766D) or
Pall = Fsu_material * 2t * (ED + 0.383D)

Then, I ensure that the Pall you just calculated (using the actual, shorter ED) is greater than the Pall from the SRM/MILHDBK 5 (based on 2D). If it is, the short ED condition is OK.

For short ED's that are normal to the load path (part's edge is parallel to the load path), I would do a net tension check using Pall = Ftu*A_net. Then, I would ensure that Pall is higher than the MIL-HDBK joint allowable Pall.

Please correct me if I am wrong. Do you agree so far?



The last time I posted this question, Philcondit responded with:

&quot;One thing to remember about the shear-out formula: It may be unconservative for e/D < 1.5. This can be appreciated if you envision the &quot;ligament&quot; as a beam, under a distributed load from the fastener. For e/D > 1.5, this beam is a shear critical beam, while a smaller e/D becomes bending critical (a picture here would be helpful, I know). So, how do you analyze a shy e/D less than 1.5 you ask? Treat it as an idealized lug and analyze according to Bruhn.

For e/D perpendicular to the load path, look at Petersons handbook for stress concentrations. They have a case for a hole near an edge. If you take the Kt's for various e/D's and normalize to a 2D configuration, you have a relative stress factor. To take it one step further, if you knew the fatigue life of the original part (maybe DT inspection, Service Bulletins, or fleet history), you could take the inverse of the relative stress to the fourth power (typical for aluminum)and get a relative life ratio to multiply your unmodified life by. Fun Stuff!&quot;

You lost me when you said &quot;ligament&quot; and &quot;Treat it as an idealized lug and analyze according to Bruhn.&quot; Could you digress a little? Oh, I don't own Bruhn.

Likewise ,for short ED normal to the load path, I don't know how to do the analysis with Peterson's Kt's. I have never dealt with Kt's. Is that related to crack formation, growth, and propagation? I can see that for this type of short ED, the part is susceptible of developing a crack from the hole all the way to the edge due to the tensile load and reduced net area. For this type of loading, you actually recommend performing some sort of fatigue analysis instead of a simple net area check?

(should I get Bruhn and Peterson's books?)

Thanks for your wisdom!
Alex

 

[Sparweb]

I would strongly recommend both books. What books did you use in college?

The point philcondit is trying to get across is that the closer the fastener gets to the edge, the thinner that edge material gets. This thin material will break out under progressively smaller loads as the fastener gets closer to the edge. How much load? That depends on the geometry, and which way the load is oriented in relation to the edge of the sheet. Philcondit refers to Peterson and Bruhn's books because the data in them are essential to a credible analysis.



STF

 

[Sparweb]

Grain direction is also an important consideration. Any load on the fastener to or from the edge could be in the short-transverse direction of the material, if it's an extrusion, for example. This direction is the weakest of all and the F_tu might be half the value in the Longitudinal direction.


STF

 

[KSshopper]

&quot;Ligament&quot; is simply the material left between the hole and the edge.

On the parallel edge margin problem: I've heard of engineers doing a similar calculation, but I think it's overly conservative. I've found I can redistribute the shear in the shy ED fastener to the fasteners on either side. To be sure I've stiffened these fasteners by oversizing them to draw load away from the problem. Coldworking is also a viable method to restore hole quality if you're worried about fatigue. If all hope is gone, (i.e. e/d < 1) the hole can be trimmed out, and fasteners added/oversized to pick up the lost load.

On lug analysis, I recommend buying Bruhn and familarize yourself with his lug analysis. It is an valuable tool for evaluating joints.

 

[koopas]

Sparweb,

I majored in aerospace engineering and never used Bruhn or Peterson. The classes I took were the standard statics, continuum mechanics, and two additional structural analysis courses (one of which was an FEA class). In most cases, we used the prof's notes.

Had I been a mechanical engineer, these books probably would have rung a bell.

So far, I've found the FlightSafetyBoeing (now Alteon Training) Structural Repair for Engineers courses (parts I and II), along with Flabel's Practical Stress Analysis the most helpful in my work, even though I don't do much design here at the airline, mostly repairs. Frankly, what you learn in school and what you have to apply at work are two separate things, I've come to harshly realize. I conjecture that most books mentioned on this forum are rather &quot;practical&quot; references with empirical formulas derived from tests. The books I own are simply &quot;typical&quot; college textbooks with not much relevence to my current position. Too many lengthy unsolvable PDE's and rather large tensors. The nice thing about Flabel's book is that the formulas are simple and readily applicable to the &quot;real world&quot;. I don't think I've seen an integral symbol yet.

I'll definitely purchase Bruhn and Peterson...quite pricy though! If I have any questions about them, I won't hesitate to ask here.

Thanks for all your help!
Alex

 

[Sparweb]

I experienced a similar mismatch myself when I started working, but with nothing but course notes at your disposal, ouch! I had Bruhn and Niu's books to start with, and I used them a lot.

Bruhn, indeed, is very empirical and hands-on; it's also obviously out of date, and the companion books by Bill McCoombs are recommended, particularly if you don't have a professor walking you through, page by page. Yes, more books to buy. Try used books on

http://www.alibris.com

(I only heard about them myself a few weeks ago, so I don't have them yet, sorry).

Another relevant question is how have similar repairs been done in the past. Presumably you work at an organization that does repair work regularly. Try to get your hands on drawings of other repairs. Be very cautious doing so, however, as there is no guarantee that someone else knows more than you.

I've heard stories of a particular engineer who repaired light twin aircraft (thin aluminum skins) using a thick stainless steel patch with three rows of #5 monel rivets all around. By his credentials, you'd think he knows what he's doing.


STF

 

[koopas]

Sparweb,

Pardon my ignorance, but is the thick patch of steel with the three rows of #5 monels bad?

Alex

 

[Sparweb]

First lesson I learned about repairs was to LOOK AROUND. The materials, the joints, and overall structures in the vicinity define the loads in the more local area you must repair. Your repair should &quot;look like&quot; the nearby structure. There might be locations in some aircraft where the aforementioned repair is appropriate, perhaps even in the SRM. I've never done a skin repair for a Part 25 aircraft (all that's done either by in-house engineers or the OEM). I only see Part 23 aircraft and 27/29 helicopters. Perhaps this engineer was trying to use a SRM repair from a heavy aircraft in the light twin.
To everyone else, it looked like a zebra walking on main street.

More specifically, any repair shouldn't be so stiff as to be a stress concentration of its own. By not permitting normal deflections of the skins and between the frames, the heavy doubler would increase the stress in the location, not reduce it.



STF

 

[rhodie]

I should have gone into aerospace engineering. *sigh*

Great thread, everybody. Good fun to read.

 

[Sparweb]

rhodie,

It's not too late to start!

STF

 

[JSF]

Koopas,
When you calculate your SED material shear allowable the formula Pso = Fsu_mat'l*2*t*(Edge Distance-0.383*D_hole)
Edge Distance should be measure from the center of the hole to the edge of the part. This allowable is for your material shear out. If you have ED is less than 2D, you need to check your bearing allowable of the part. of the critical ED. Pbru = Fbru(1.5D) or Fbru(2D) *D_hole*t. 1st thing you need to know is whether your fastener shear value if you do not know the load (only can be obtained from manufacturer or if you know how). If your fastener shear value is lower than your Pso, and Pbru your material is ok for SED. When Fbru less than 1.5D the material does not behave linearly (you can see from MIL-HB5). No one knows what is the number. You also need to check for the net tension failure. Usually you trim it out and or install a doubler or upgrade fasteners to make up the load.

Hope it clear your mind.

Rich.

 

[koopas]

Sparweb:

I am having trouble locating a cheap used book source for Bruhn. Also, could you confirm the companion book's specs.? Is the author Bill McCoombs?

Thanks,
Alex

 

[koopas]

Sparweb,

One more thing before I sign off tonight. Another post in this forum discussed the poor fatigue qualities of a thick repair doubler. Why is that? By thick, do you mean thick compared to the original skin or thick beyond a certain thickness? The SRM allows you to go same or one gage higher. By the way, have all SRM repairs undergone a damage tolerance assessment?

Say you have 0.090&quot; skin and slap on a 0.100&quot; external doubler, and calculated that you need three rows of fasteners around it.

I guess the goal here is to introduce the load gradually into the doubler so what are my options?

-Taper/chamfer the doubler edges and shoot an extra row of stabilizing fasteners with self-aligning washers/nuts? So I'd end up with three rows around the cutout and an extra row of field fasteners in the tapered area?

-Use two doublers. Say a 0.071&quot; and 0.032&quot;. You'd put the thicker one at the bottom (closest to the skin) and then the 0.032&quot;? Since your goal is to gradually introduce the load, you'd make the thicker doubler larger by say one row of fasteners? So you'd end up with three rows of fasteners through both doublers, and then another row only on the thicker doubler?

Do you agree with what I've presented? What are some other options?

Thanks,
Alex

 

[astroclone]

You want the thinnest sheet closet to the skin. That's why you would taper the edges of the doubler if that was your method. The first fastener row takes the most load so you want it going through a thinner sheet.
T

 

[Sparweb]

koopas,

All in all, you are getting the picture, though a tapered sheet can be tapered gradually enough that you can rivet it fine - just put the BUCKED head on the tapered side. Your most recent questions are answered fairly well in a FAA paper by T. Swift. I have a pdf copy. Supposedly it's on the web, but I've not found the link yet.

The thickness of doublers is a matter of debate because different aircraft should get different treatment. A pressurized transport category aircraft is designed to a &quot;damage tolerance&quot; basis, which is very different from the &quot;fail-safe&quot; designs of smaller lighter non-pressurized aircraft. A simple fail-safe aircraft tends to only need a doubler that's one guage up in thickness. If you're dealing with a damage tolerant aircraft then the thickness of the doubler must be considered carefully, it might be the same thickness, or you might need a &quot;layered&quot; repair, and if you're installing an object on the skin, like an antenna, then there are a raft more loads to consider.

I've got access to a bunch of stuff you might find handy before your copy of Bruhn arrives.

sparweb@hotmail.com

Also link to thread2-55653, my buddy Ed has the Bill McCombs info you need.


STF

 

[koopas]

Hi Sparweb,

I think I found a reasonably priced copy of Bruhn. What edition/revision is the latest?

Thanks,
Alex

 

[Sparweb]

Mine is 1973, and I got it new at the campus bookstore.


STF

 

[koopas]

Sparweb,

When you said you have access to a whole bunch of stuff before my copy of Bruhn arrives, do you have a personal website url? I am really trying to get up to speed on this stuff.

I just ordered the book, btw. 1,200 pages?!?!

Alex

 

[Sparweb]

1200 pages: they don't call it the &quot;bible&quot; for nothing.

Sorry, no home page you can skim, but drop me a line and I can give you some stuff via e-mail.

Since it's pretty much a guarantee that any e-mail address on Eng-Tips will eventually get picked up by spammers [flush](did you know that these postings come up in Google searches?), I recommend you get a &quot;freebie&quot; e-mail address from Hotmail or Yahoo, like I did. My home address has remained nearly junk-free for 5 years due to my caution.



STF

 

[Sparweb]

Though the mere mention of e-mail addresses often causes many threads to go dry, I'm going to try to revive this one, because something still doesn't sit right with me.

Two posts have given the shear-out equation to use while solving the short e/d problem as:

Pso = Fsu_material*2*t*(Edge Distance-0.383*D_hole)

Is that in Bruhn? If so, could somebody politely rub my nose in it? If not, is there some other reference?

I've taken a quick peek at the 0.383 coefficient and it appears to make a 50^o angle from the line of the rivet row.

ARCSIN(2*0.383)=50^o

Why 50^o? I thought 45^o is the typical angle in a shear-tear-out plane.

Using 50^o looks more conservative, but not by enough to be worthwhile.

Anyone care to elaborate?


STF