M

mcencula

Well Known Member
OK, I'm planning on adding one large access hole to the top forward skin on my -7A (tipup). It will be a large access hole and consume as much of the space between the tipup support ribs as possible. Since it's going to be a single panel rather than two smaller panels (as others have done) it will span (and be screwed to) the center rib. I'm trying to calculate the number of fasteners (rivets and screws) to meet accepted practices and not compromise the strength of this skin.

According to AC 43.13-1B, Table 4-10, I need 6.9 3/32 rivets per inch (RPI) all the way around the access hole. Using three rows of rivets on 4D spacing (3D spacing row-to-row) results in 8.33 RPI. So far so good.

I plan to use #8 screws to attach the panel (AN509-8R8 or MS24694-S5).

To check the strength of the various interfaces, I have five sets of calculations:

1. Shear strength of rivets
Alloy 2117-T4
Shear strength 26,000 psi
Tensile strength 38,000 psi
Rivets per inch 6.9 (this is conservative since actual will be 8.33 RPI)
Shear strength of rivets per inch = 6.9 * (3/64)^2 * pi * 26,000 = 1238 lb/in

2. Bearing strength of skin on rivets
Alloy 2024-T3
Tensile strength 58,000 psi
Rivet dia = 3/32
Skin thickness .025
Rivets per inch 6.9 (this is conservative since actual will be 8.33 RPI)
Skin bearing strength per inch = 3/32 * 6.9 * .025 * 58,000 = 938 lb/in

3. Tensile strength of skin (this applies to both the top skin and the doubler)
Alloy 2024-T3
Skin thickness .025
Tensile strength 58,000 psi
4D rivet spacing results in 75% material remaining
Tensile strength of skin per inch = .025 * 1 * 75% * 58,000 = 1087 lb/in

4. Shear strength of screws
Screw spec AN509-8R8 (MS24694-S5)
Tensile strength 125,000 psi
Shear strength = 58% * Tensile strength = 72,500 psi (per Young's Theorem aka Distortion Energy Theorem aka Von Mises Theorem)
Diameter of screw = .164 inch
Shear strength of a single screw = (.164/2)^2 * pi * 125,000 * 58% = 2640 lb/screw

5. Bearing strength of skin on screws
Alloy 2024-T3
Tensile strength 58,000 psi
Diameter of screw .164 inch
Skin thickness .025
Skin bearing strength per screw = .164 * .025 * 58,000 = 238 lb/screw :eek:

So everything was looking good until I performed calculation #5. In order to make the screw interface as strong as the next weakest link (calc #2), I'll need 4 screws per inch. That "feels" like way too many screws, so I think using the grip diameter of the screw in calculation #5 may be way too conservative.

So here's the question: Since the screw locations will have nested dimples, is it reasonable to use a different diameter (larger than .164) to calculate the bearing strength of the skin on the screws? How do you aero engineers do this?

Thanks.

OK...I checked with a fellow engineer. He had some info that was helpful. He looked in one of his reference books, "Analysis & Design of Flight Vehicle Structures" by E.F. Bruhn and although he couldn't find the ultimate shear strength of a #8 dimpled hole, he found that a 5/36 flush dimpled hole was good for 415lb. That's a lot better and it get's me down to about 2.25 screws per inch.
 
Last edited:
Ultimate shear strength of alloy is around 40 ksi so for .025 sheet max ultimate shear flow is 1,000 lb/in. So you would need 4 screws per inch.

However I'd bet that the in the worst load case the skins see no where near this. It is probably designed around panel buckling or inter rivet buckling.

I'd check out what others have done. I suspect that 1 to 1.25" spacing works fine.

Jim Sharkey
RV-6 Down for Mag SB :(
 
Stress analysis

The number of fasteners should be based on the load being carried, not the properties of the material. We don't have load information of course, only Van's does. Use the handbook spacing for the rivets and screws (4D to 6D) and you will be fine.
 
Mike, there are two ways to look at this problem: 1) Knowing the loads we design a structure that supports the loads 2) Treat it as a repair job.

I don't think anybody, including Van, knows the loads here. The primary fuselage bending loads are carried by the longerons, with shear dispersed into the side fuselage skins. The piece you're talking about modifying carries only secondary loads and I really don't have a clue as to what those might be. Supporting the canopy while open might be a major part of it, with secondary consideration to fore-aft support to the instrument panel and upper firewall. And a major job is just to hold shape for the upper fuselage forward of the inst panel.

So, lacking accurate loads data, I wouldn't have any idea how to handle it from that point of view. When looking at rivet pitch, and loads through relatively thin skin, the source is (or was) MIL HBK-5, chapter 8. I use a copy from Dec 1998, but its been revised and renamed since (with a subscription fee, of course). Using the MIL HBK, you can look up factors such as rivet shear strength, and bearing strength for the skin. The lowest of these will drive the structural design.

If I was doing this, I'd just follow the general guidance in AC4313 and make it look pretty. Keep it stiff enough for the panel, canopy, and firewall. With nutplates in there, the screws are transferring the stress from the nutplate into the rivets that attach the nutplate to the skin. That's two rivets per screw doing that job - or about 400 pounds shear strength per fastener. Way way more than enough to do this job, even with wide spacing.
 
Thanks Bill. Yes, I'm essentially trying to treat this as a repair, and have a good reference for single lap joint repairs with rivets (AC 43-13). Unfortunately I don't have such a reference for using screws. It sounds like it'd be good to get a copy of MIL HBK-5 or that Analysis and Design of Flight Vehicle Structures that they used at Northrop.
 
Got it!

Bill you were right on. I found a copy of MMPDS-01 (the FAA maintained version of MIL HDBK-5) That shows the joint strength of a 3/32 rivet to be 217 lb. and the joint strength of a 5/32 rivet to be 474 lb. There is no entry for a #8 screw, but using the data for a 5/32 rivet should be close enough.

Soooooo...I'll need 217 * 6.9 / 474 = 3.15 screws / inch.

It's amazing how much easier this was once I had the right reference material.

Thanks again!
 
mcencula said:
Bill you were right on. I found a copy of MMPDS-01 (the FAA maintained version of MIL HDBK-5) That shows the joint strength of a 3/32 rivet to be 217 lb. and the joint strength of a 5/32 rivet to be 474 lb. There is no entry for a #8 screw, but using the data for a 5/32 rivet should be close enough.

Soooooo...I'll need 217 * 6.9 / 474 = 3.15 screws / inch.

It's amazing how much easier this was once I had the right reference material.

Thanks again!
Click to expand...

2 cents - That's a heck of a lot of screws to remove to gain access - and if after a few years only one of them siezes or strips you'd be lucky. Not to mention the cost and weight. It's not that difficult to crawl into the foot well :D

Jim Sharkey
 
jsharkey said:
2 cents - That's a heck of a lot of screws to remove to gain access - and if after a few years only one of them siezes or strips you'd be lucky. Not to mention the cost and weight. It's not that difficult to crawl into the foot well :D

Jim Sharkey
Click to expand...

Hrm...good thought...gonna have to let this one roll around in the noggin a bit more.
 
Here's an example of how to do something like this:

http://www.meyette.us/FuselageSep05.htm

This is no big deal but is it REALLY worth the effort? I think you'd see a lot more do it if it was. Having said that, didn't Dan Checkoway do this on both sides of his forward fuse skin to get behind avionics panel access? I think if you just 'make it look right' by following the screw spacing Van's already does you can't go wrong. Look at your fuel tank and other similar areas that use screws for spacing suggestions... Bottom line, the loads can't be that much in that area relatively speaking and if you treat it like a repair job you should be fine if you still want to do that.
 
Last edited:
You must log in or register to reply here.