Dents 4

[DavidWootton]

Hi Everyone,

I am looking to find some information on the allowable sizes and depths of dents in various sizes of panels and their stress concentration factors. Such things as stress concentration factors against depth and width of dent to diameter will prove useful. I hope some of you have looked into this scenario and can give me your input on this matter.

Thanking you in advance

David

 

[KSshopper]

Boeing and Airbus each have their proprietary methods to deal with dents that come through MRB. I've also seen some finite element work done on the effects of dents as well. Both approaches treat dents as a fatigue issue rather than a static strength problem. Quite frankly I don't think either company's method has much value added.

I think the best approach is to try and rework smooth dents (no sharp creases) back into place, then check for cracks using dye penetrant or surface eddy current. If the dent has sharp creases, or generally looks bad, it's a safe bet to cut out the area and repair.

 

[EasyAim]

Assuming we are talking solely about metallic structures:

Rarely are dents are concern. The first thing I ask is "is it a thru dent?" (i.e. is a bulge visible on the opposite side of the part?). If so you may have some buckling/stability issues.
Another serious concern is if a dent occurs at a hole, stiffener, or other geometric discontinuity. Consider a doubler repair to reduce loads in the area in that case.
Similarly, for thin skins, determine if it "oil-cans" (pops back and forth between concave and convex which light force). If so, you must add a doubler or similarly stiffen it.

ALWAYS perform a dye penetrant inspection for cracks.... in about 10% of the cases a crack developed - and sometimes not right at the dent. Be sure to inspect any stiffener radii and other geometric shapes in the surrounding area.
Dents are rarely broad or deep enough to change the gross area stress. They are rarely a fatigue problem either: A shallow dent (i.e. dish depth less than 30% of the thickness with no sharp edges) creates a Kt of about 1.1 so it is easy to buy it off since it is less severe than an access hole or fastener hole which is usually nearby. (I've never seen a report of a crack that initiated due to a dent.)
If at all possible, I suggest to NOT push or pull a dent - althought you sometimes have to do so due to aerodynamic concerns.

 

[ABXLiaison]

Easy Aim,

I would be interested in some of your substantiation for shallow dents if possible. We run into this occasionally with relatively minor dents that fall out of SRM limits. If I could show a Kt relationship to adjacent structure it would be a huge benefit.

 

[JohnDAR]

Dents may be a problem by themselves but repairing them can introduce more problems. The issue with metallic structure - especially aluminum - is the effect of cold working the dent to 'straighten' it. Cold working can increase the temper and dramatically reduce the toughness. Even very old repair manuals warn about this problem.

 

[riktoo]

Last summer A friend purchased a damaged set of wings for his Murphy Moose .
The leading edges had a total of 3 awfull dents ,the longest being 38 inches right on the curvature of the leading edge. There were serious creases and the leading edge was effectively flat to the airstream . The depth of the deformation was about 3 1/2 "in. with two right angles at the bottom and top of the damaged area.

The alloy is 5150 .We knew we would have to cut these areas out and rebuild them with a patch.Keeping the damaged area to attach the metal patch ,would be stronger . Assuming that we had nothing to loose as the damaged areas is usually cut out or patched over, I tried to hammer out some the damage.
I used a small body hammer and extremely low force. Dents and creases are areas of concentrated stresses and if you know how to relieve that stress gradually ,it is possible to eleminate most of it.
Finally after six hours of light hammering and more than 6000 hammer blows the leading edge looked almost perfect. As the aluminum was work hardened during the process we proceeded with the usual patch repair.
The other dents were repaired with the same procedure

 

[formentera27]

I just can tell you that philcondit´s approach is not what neither the DC-9 not the A320 SRM states in case of dents.
As you probably know, there are some dents which are acceptable as is. Others require repetitive inspections and other require a final repair.
Anyhow, you can never reform the dent´s shape to back to place before making an assesment.

 

[KSshopper]

I'm sure the DC-9 and A320 SRM's don't address reworking dents, but if the SRM's were perfect, maintenance wouldn't need liaison engineers!

The fact on dents is repair philosophy is based on "tribal knowledge" The reason for this is dents have not been seriously addressed by industry or academia. Most of the early A/d criteria was based more on aerodynamic concerns rather than structural.

I have seen both Airbus and Boeing tests and analytical techniques on dents. They only address "durability" concerns and leave lingering questions to static strength and damage tolerance. Until industry pays more attention to dents, we'll be forced to treat a dent according to the knowledge of their tribe.

 

[johnfalky]

I remember having a DC-9 in a heavy check with inspection writeups on "dents" in the elevator skins. When I requested tolerances from "Boeing Long Beach" they sent me the factory smoothness standards for these parts. They were so "tight" that more than likely, they could NOT be used even in manufacturing, since NO part could have passed the inspection requirements. The big problem here was a poorly written inspection jobcard that was triggering numerous writeups where there wasn't a problem, i.e. the "dent" writeups were really ripples caused by temp. changes on the ramp and the stresses causes by installation of the fasteners through the skins themselves. During maintenance, dent writeups, which are difficult to clear without "approved data" often generate unneeded rework/trimouts of fuselage skins and flight control surfaces. As noted in some of the responses, "leave it alone" is often a good solution

 

[EasyAim]

There is an excellent paper by Frank Simmons, Jose Veciana, and John Wallace titled "Effects of Dent Removal on the Design Properties of Fuselage Skin Material", 2000-APRIL, AIAA paper 2000-1467.

SUMMARY:

Material: 2024-T3 clad aluminum
Skin gage: 0.040, 0.050, 0.063
Dent depth: 0.025 to 0.060 deep

* Sheets of 24 inch X 48 inch were dented.
* None of the materials cracked during impact to create the dent. (Rubber mat was behind the specimen.)
* Specimens then had the dent removed by "succession of controlled blows using a hammer at room temperature".
* Some specimens were subsequently rotary peened.
* Dogbone specimens 0.50 inch wide were cut from the dented sheets.
* Three groups of specimens were created: virgin baseline, dented and reversed hammered ("processed"), and processed with subsequent rotary peening.


CONCLUSIONS:

* Static strength properties of all of the specimens met MIL-HDBK-5 B-basis range.

* However, there was a reduction in fatigue life of 5% to 20% for processed specimens.

* Rotary peening for shallow dents had a large effect and returned the fatigue life back to original blueprint range.

* Rotary peening had a smaller effect to large dents. It was assumed that this was due to the deeper tensile residual stresses that would have been affected less by the shallow compressive residual stresses.

One of my aerospace clients have also performed similar tests with thick spars (reference Vought tests TR TR-97-44300-033 and AVO 2-51200/7M-105). These tests also confirmed that dents in metallic structures are a concern for fatigue and crack growth but are not a serious concern for static strength. (Assuming no oil-canning and load eccentricity concerns.)

 

[jetmaker]

EasyAim,

Do you have a link to that paper?

Thanks,

jetmaker

 

[HSThompson]

Well, This subject has probably been beaten to death, but here is another crack at it.

Take the case of a gouge in a skin panel (not too deep). To remove the gouge, you typically blend at a 30:1 ratio to restore most of the fatigue resistance properties. You might even possibly locally flap peen.

Now, I agree that this isn't exactly the same thing as a dent as the opposite side of the gouge skin surface is not deformed. However, isn't the resulting stress concentration of a 30:1 blended gouge similar to a dent condition?

I believe it is.

 

[jetmaker]

Dents left as dents are primarily a fatigue issue, but not due to the Kt effect. By leaving the dent you have residual stresses from the plastic deformation of the part that need to be accounted for in your fatigue analysis.

Also, large dents can start oil canning, which will result in a fatigue issue. Or in the case of shear resistant webs, the dent may result in the web buckling before ultimate load due to the eccentric load path.

Reworking a dent back to its original contour presents it's own problems in that you have just coldworked the material, likely lowering it's fracture capability. Therefore, it is best to reinforce these dents.

Regards.

 

[Sparweb]

Dents and scratches, even if the skin hasn't been penetrated, are just as bad as cracks and punctures. A repair must remove unacceptable damage, restore original strength, and be inspectable within a reasonable schedule.

Otherwise, we'll have another China Airlines wreck all over again.

I concur that a stress-concentration factor doesn't really apply to this situation, although if you don't remove the core of the damage, the Kt could be very high. Drilling or cutting out the center of the damage is the only way to take control of the residual stresses, unless you have a bulletproof way of demonstrating that you know the residual stress as-is.

On very small nicks, I expect that a single rivet solves the problem nicely. Drive a solid rivet through the skin. No load transfers through the rivet, so you don't have a fatigue problem, and use some sealant to close everything up neatly. The damage and contaminants are removed when the hole is drilled, and you could even use a csk rivet to keep things flush, meanwhile allowing you to use a smaller rivet while still taking away the same area at the surface.

If you have a larger dent to deal with, then the aluminum has been stretched, hence work-hardened, so the crack growth rate for the panel is now unknown. That AIAA paper sounds like just the thing for that situation. (I doubt you could find the paper for free)



Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

 

[Sparweb]

EasyAim,

Could you verify the reference to the AIAA paper, please? It was not published in the April 2000 AIAA Journal.

Thanks


Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

 

[737eng]

Sparweb,
AIAA 2000-1467 is the correct reference. It appears that it must have been presented at the 41st AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamic, and Materials Conference in ATL, APR 2000. There is another document number listed at the top - A00-24546.

 

[Sparweb]

Found it this time - didn't appreciate how many different periodicals they circulate. Thanks.

Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

 

[KeesG]

This thread is exactly what I needed to solve some problems. I am currently doing some research into why certain dent allowable limits are used. Apparently Airbus has a method listed in their SRM that allows you to evaluate the severity of the dents and how it will affect the remaining life of the structure/component. Does anybody have access to an Airbus manual with this dent evaluation method in it? Thanks!

 

[Oexen]

Concerning Airbus approach:

There is a stress increase due to a dent. Depending on the ratio of skin thickness / dent depth there may occur a stress increase of up to 80 % for deep dents.

You will find in the A320 SRM (it seems you have access to it) in chapter 53-X1-11 pb.101 (X for the specific a/c section 1 - 5)the diagram of allowable values for dents and their life limitation.

You will also find in the SRM in chapter 51-73-00 a process for dressing back dents. This is limited to thin skins (up to 1.6mm) as there is a risk of crack initiation due to dress back (cold working).

Another point for your studies. The stress increase is only determined for tensile stresses as compression is mainly transfered by the stringers. You just have to ensure there is no oilcanning effect and in areas which are mainly under compression (e.g. lower shell of the fuselage) the dent is likely to be acceptable.

I hope this may help you on your studies.

 

[KeesG]

Thanks for your reply Oexen. Unfortunately I don’t have access to Airbus manuals. I am very interested in what they have written down, do you know how I could possibly get access to these sections? I know the Airbus approach is focussed on fatigue. Another of our concerns is that static stability in both shear and compression could be affected as well depending on the stiffening ratio in the area where the dent(s) occur. We will be investigating this more. Thanks again for your help.