NAS1097 vs MS20426 1

[2ndThermoLaw]

Hello,

can these be substituted for each other? ie. could MS20426 be installed in lieu of NAS1097 and vice versa?

 

[berkshire]

2ndThermoLaw (Mechanical)
The Nas1097 rivet has a smaller head than the MS20426 for the same shank diameter.
So you could not put an Nas1097 rivet in a hole where an MS20426 has been.
B.E.

 

[rb1957]

NAS1097 are typically "repair" rivets. like berkshire says the 1097 has a smaller head than the 426 (something you should be able to find out on your own). you could replace a 426-4 (BB4) rivet with a 1097-5 (LZ5) by drilling out an oversize hole (5/32" in lieu of 1/8"). youo'd use the same rivet material, (AD, DD).

 

[Sparweb]

The allowable joint loads aren't the same, unless the materials are so thick you don't need 1097's any more.

Steven Fahey, CET

 

[debodine]

we tend to use them interchangeably for non-structural applications, like mounting a disconnect bracket that holds an electrical connector, etc.

debodine

 

[planedr]

I never really considered them a repair rivet. Small and narrow body aircraft use NAS 1097 (BACR15CE or BACR15GF), as they can be countersunk in thin sheets without knife edging.

For example a -3 426 is knife edged in 0.036 sheet. A -5 1097 is knife edged in that same sheet and carries more load. More load and a better fatigue life.

All that varies of course with sheet thickness, assembly methods, hole size, driven condition, and prep.

 

[2ndThermoLaw]

I would be comparing them at design stage not during repairs and so on... Just wondering if there's any "mysterious" cons to a 1097 apart from the obvious factors (knife edging, load... etc).I'm thinking interms of fatigue performance...etc.

On the subject of knife edging... as a general rule how much (of the thickness) uncut material be left on sheet... by this i mean can you punch to, say, 80% material thickness to countersink the head???

 

[MNLiaison]

Optimum countersink depth for Boeing aircraft is 67% (or less)of the sheet thickness. They state this percentage helps to reduce possibility of fatigue cracking. During repair they allow up to 80% of the sheet thickness for countersink depth.

The NAS1097AD fastener is used through out the Douglas DC9 Fuselage to attach the skins to most underlying structure due to the thickness of the skin (0.050"). Using their countersink charts for the NAS1097 and their minimum sheet thickness requirements the % for the 5/32" dia. is 58%, for 3/16" dia it is 63%, and for 1/4" dia it is 74%. All of the previous percentages leave the fastener head a little high and then would be shaved flush.

So it looks as though 67% would be a typically value for countersink depth to sheet thickness.

Regards

 

[2ndThermoLaw]

Yeah I've heard that 67% floating around!!! Where would I be able to find some hard evidence??? Any generic guidelines out there?

 

[kontiki99]

You might try looking in FAA Advisory Circular 4313.1B, ACCEPTABLE METHODS, TECHNIQUES, AND PRACTICES AIRCRAFT INSPECTION AND REPAIR. It's available at the FAAs Advisory Circular website.

I'm not structures, and don't know how well the topic is covered, but it's the first place I look for generic guidelines.

 

[berkshire]

kontiki99 (Electrical)
theNAS1097AD is not even mentioned in AC 43 13 1B change 1 The only countersunk rivet mentioned is the MS20426.
Also I do not think there is any reference to percentages of metal thickness on countersinks in that repair data.you would have to look elswhere for that
B.E.

 

[rb1957]

for good fatigue performance CSKs should be less than 2/3rds of the thickness. i forget the original reference for that, and accept it now as good practice.

there is also a practical aspect, in that you need to some straight shank to allow some tolerance on installation.

btw, LZ4s in 0.04" skin, pretty typical constrcution are just outside this guide (0.029" CSK in 0.04" thick). my approach is to treat these as "transition" knife edge and add a fudge factor into the fatigue analysis (Kt = 1.5) to account for this.

 

[WhiteRabbit]

You can also use (Csk Dia - Shank Dia)/[2*(tan 50)] to calculate the csk depth.

Douglas uses 70% csk depth. It seems that it is more dictated by specific program philosophies. One half dozen...

 

[737eng]

I have the same data as MNLiaison for the BAC aircraft. I took a Boeing Structures class years ago where they state that the BAC standard for countersink max depth is 67% thickness for new designs and 80% thickness for repairs. This 80%t is reinforced by going to Chap 51-40 of the SRM and evaluating the min thickness tables for countersinks, you can check these tables and find that they all ensure min 80%t.

I would be careful of shaving NAS1097 heads, it's been awhile since I worked a Douglas aircraft (and I miss it!!), however, I thought there was a note in the SRM that restricted shaving NAS1097 heads.

 

[rb1957]

back to 2ndThermoLaw's post (on the 16th).

an obvious design aspect of using 1097s is their reduced tension capability. i know you're not going to be designing them to carry tension loads, but the issue may arise when you've used them in a thin gauge skin which then buckles (under shear and compression). off the top of my head i don't remember how NACA TN2661 calculates the tension loading on the rivets due to buckles forming, but i think you'll need tension allowables which are going to be hard to find (but they are pretty easy to generate).

regarding planedr's post (back on the 15th), i'd wouldn't put a 1097-5 in 0.036" skin, nor a 426-3, and i'd never describe either as "better fatigue" ... both are going to very limited in fatigue life and i think the 1097-5 with it's higher static allowable might beguile the designer into thinking things are ok. if i had to use a CSK rivet (in 0.036" sheeet, primary structure) i'd opt for a 1097-3, but i'd rather a 470-4.

 

[crackman]

The basic cutoff of 67% or 70% is based on much of the published data by the various OEMs (internally published of course). I have spent much time developing some of this data including performing batteries of test samples for various % Cntsks. Essentially the Kt effect around 67% to 70% is about 1.2, above this, it starts accelerating and gets up around 1.7 at 100% Cntsnk depending on the design of your joint (always develop your own test values if possible).

As for the NAS1097 vs MS20426 I can say this much. The 1097 does exhibit a big improvement in fatigue life because of the slight rise in the center of the head. For a 3/16 diameter the 1097 has a basic head height of 0.046 (vs 0.073 for 20425) and an additional 0.008 crown in the center. This crown results in better over bucking of the rivet and in swelling. Lockheed developed their own improved version of the 1097 and named it LS10052. The improvement is that the crown extends across the entire head rather than just in the center. If you look up AIAA Paper No. 72-776 entitled "Structural Development of the L-1011" by D.J. Mackey and H. Simons you will see a comparison of joint fatigue lives. The curves presents fatigue life comparisons of the MS20426, NAS1097 and LS10052. The data presented is for DD6 rivets but I have seen the remainer of the data and it is pretty uniform even for standard sizes and non-ice box rivets. Basically, the 1097 shows about a 5 to 1 improvement over the 20426 and the LS10052 shows an additional 1.36 over the 1097. This type of data has been developed over decades and has formed the basis of many OEMs methods alas no one is left alive to remember where it comes from.

As for shaving heads, dont do it without testing.
Hope this helps.

James Burd
Avenger Aircraft and Services

 

[planedr]

Rb 1957

You missed the point completely. That was 100% countersink!

In a given sheet assembly you can use a bigger rivet period. The prefered failure method in primary structure is bearing. For that you need a bigger rivet.

The deeper the countersink the more likely fatigue cracks are going to start at the inner surface along the straight shank of the fastener hole. The reduced head size of a 1097 provides greater inspection ability for the eliiptical crack formations in the future.

If you are in a position where your repair parts will buckle your repair is inadequate. If you must still use that thin a gauge you should be using 470's or Hi-Loks.

I would never go more then 67%. Boeing AR's will not accept 80% even with a gun to their head. Put in a countersink repair washer and then use a button head rivet. An already fatigued skin will crack out way to quickly. Better yet after inspecting the hole you need to o/s anyway just go button head.

 

[GregLocock]

Are you guys discussing alooominum plate? or does this apply to real metal as well?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[rb1957]

i'd say all metal plates would have the same knife-edge criteria. mind you, there are design solutions which would allow exceptions (non-primary structure, non-fatigue loaded, special rivets and hole treatments)

 

[crackman]

Just as an aside, the 67% is a good solid guideline to follow but note that it is not all encompassing. One has to consider the design and aircraft to which it is being applied. There are several aircraft that employ greater than 67% and even 100% successfully. Of course, they employ other mitigating design features to offset this (ie butt joints vs lap joints). Obviously delta pressure, fuselage radius, and service goals are other factors which directly influence this as well. All that said, without additional testing, certainly the 67% should be followed otherwise test and base your design on those results.

James Burd
Avenger Aircraft and Services