Adam Dickson

Strengthen wing skin 3-4 connection - part 1 - proposal

Some time ago I noticed that all the wing skin-skin connections involve a double row of 3.2mm rivets - with the exception of the top skin 3-4 connection, which consists of just a single row of rivets.

Furthermore even the latest revision skin 4 (Rev 1) has a rather small hole-centre to edge distance, reducing to as little as 5.5mm near the front of the skin edge.

Approximate calculations show that the tensile strength of these skin-skin connections are rivet limited, as one gesipa 3.2mm rivet has a shear load capability of 720N, while the load capability of a 32mm width band (the amount of space between each successive rivet of the skin 4 edge connection - 0.64mm thick) of 6061-T6 is 240MPa * 0.032 * 0.00064 = ~ 4900N.

For all the other skin-skin connections, there is a double row of rivets, which a slightly reduced spacing, more doubling the strength of the connection. Conversely, the skin 3-4 connection has less than half the strength of all the other connections. This is a bit strange, given that this connection is the most inboard, where one could expect stresses to be greater.

I note the TSi has a double row of rivets at this connection. This and the following posts will document my effort at improving the skin-skin connection, more than doubling the strength and bringing to a level similar to all the other connections.

I have contacted Sling and they have to-date seen no problems with the existing connections, even with high-time training aircraft. However I have examined a local Sling 2 which has been used almost exclusively for circuit training for beginners, and has had a very hard life - no doubt had some very hard landings. It has easily completed 3000-5000 touch-and-gos. I know as a fact its companion aircraft has had a nose-wheel collapse and prop-strike. On the former aircraft I have noticed that the paint around the rivet heads comprising this connection has cracked. Is this evidence of some fretting and and some deformation of the rivets or the holes? Or is this the result of many boarding operations where people are placing their weight near these rivets and flexing the skin and underlying rib12? I will post photos at my next opportunity.

I was originally supplied with a Rev 2 rib 12 which had 4mm holes in the top flanges, modifying the original 3.2mm holes. I would have had to upsize the skin holes. I speculate that this Rev 2 was an effort to increase the strength of this join by upsizing the rivets, which have a greater shear yield strength of 1200N. The problem with this change is the very small hole centre-edge distance of skin 4. Upsizing these holes to 4mm would have left only about 3mm material left, well under the 2D or even 1.5D guideline. I purchased replacement Rev 3 ribs 12 which have returned to the original 3.2mm holes - possibly for the foregoing reasons. This is all speculation on my part.

With these thoughts mind I have proposed a modification to this connection by adding a 0.8mm doubler between rib12 and the skin 3-4 lap joint. This will bring the strength up to a level equivalent to all the other joins, and equivalent to the TSi configuration. Additional rows of rivets will be added to both skins 3 and 4 in staggered fashion resulting in two cascaded zig-zag lap joints. There will be three rows of rivets, placed to avoid both the anti-slip on the skin 3 side and the overhanging skin 3 edge on the skin 4 side. An internal 0.64mm shim will be placed between the overlaying skin 3 and the doubler, butted against the edge of skin 4. In order to maintain the wing surface profile, room will be created for the 0.8mm thick shim by dropping rib 12 by this distance. This can be achieved with appropriate upsizing and recentering of the 4mm mounting holes in the front and rear spar to 4.8mm, then using 4.8mm rivets for attachment. The doubler extends ~12mm beyond the new outer rows of rivets, well in excess of the 2D guideline.

The asymmetry of the rivet pattern is thought acceptable since the design does not attempt to balance the failure of the sheet and the rivets, in fact the rivets are still the limiting factor for yield by a large margin. The symmetrical zig-zag lap joints in AC43.13 are based on traditional rivet technique (boilers and bridges!) where the rivet and sheet failure levels were similar (for optimal use of materials).

Attached is a sketch of the proposed plan, and a photo showing the proposed locations of the added rows of rivets. Sling have ok'ed this scheme, although this change is "at own risk" like everything else about an E-AB build. Note the 21mm figure has been changed to 20mm.

I am moving very slowly before altering any actual parts - as of writing I have manufactured all the doublers and shims, simulated dropping ribs 12 by 0.8mm, and done a full trial clecoing. The concept is validated as of writing. I have not yet altered any skins or parts from the actual Sling 2 kit. The following posts will document the implementation.

This post is from Adam Dickson