Upsize and recentre rear fuselage holes to resolve misalignments

This work was actually done Feb 2023, not posted until now

As documented previously there is a large fore-aft mismatch between the longeron 301 holes and the rear bottom skin between rib 601 and rib 5, and a large up-down mismatch between the rear top skin and the side skin between rib 5 and rib 4. The hole match is good in the former case between the rear top skin and the rear bottom skin, and is the latter case between longeron 301 and the side skin.

Another builder (Pascal) replaced the rear top skin with a custom skin with holes omitted to resolve the mismatch between rib 4 and 5, but still had the mismatch between the skins and the longeron between rib 5 and 601. He resolved the longeron mismatch with a doubler arrangement of some sort.

If I were to do this again I would probably follow the same path.

In the event I found that all the mismatches could be resolved by upsizing and recentering the 3.2mm rivets to 4.8mm rivets, a 50% increase. I found that no hold-edge to skin edge distance was reduced owing to a fortuitous aspect of the recentering. The only place where some residual elongation resulted was in the 3-4 rivet location forward of rib 5 where a ~0.5mm notch was left in the side skin on the bottom edge of the upsized holes. As the stress on this system of connections is predominantly fore-aft due to the twisting forces of the tail plane, and owing to the fact that few rivets were affected this was judged of little account.

As mentioned previously, I noted that some club aircraft show noticeable fretting of the original 3.2mm rivets. Is this due to some elongation of skin holes under the rivet head, or actual stress failure of the skins or rivets at the point of contact? The main force on the this connection will be fore-aft shearing induced by twisting forces on the tailplane.

Another question is that the ratio between the upsized rivet diameter and the centre-edge distance is reduced, dropping as low as 1.45 for the rear top skin between ribs 4 and 5 and 1.66 between ribs 5 and 601. Does this matter? The actual amount of metal between the rivet edge and the skin edge is not reduced, owing to the recentering. Also the shear forces transferred by the rivets is fore-aft. My intuition suggested that the actual stress induced in the skins should not be greatly affected, and in fact be slightly reduced due to a given force being transferred to a 50% greater rivet diameter resulting in less chance of the skin material deforming in the immediate vicinity of the rivet. Furthermore, this stress being lower at the rivet surface should lead to less chance of the rivet material itself deforming. Both of these factors should greatly reduce the chance of fretting of these rivet connections.

In Nov 2023 this work was extended:

To quantify these ideas Pascal Latten kindly ran some FEM modelling of such laterally stressed rivet connections fastening two skins together. The lower peak stress for a given shear load was confirmed, and the maintained edge distance resulted in a no discernable change in the stress field near the edge of each skin. The smaller ratio was of no account. In fact the stress near the edge of each skin fell well below the stresses in the immediate vicinity of the rivet. My conclusion from all of this is that the upsized connection actually represented an actual improvement of the ruggedness of this connection with greater resistance to fretting from repeated shear loading.

The deformation of the rivet material from forces transferred by overly thin skins (which may not themselves deform due to being harder) - and consequent fretting - is an issue found elsewhere with the Sling 2, being the basis for SB14 where stainless steel rivets are substituted. The upsizing to 4.8mm rivets - and a consequent dilution of given forces over a larger area - is a measure consistent with the aims of a SB14. I have done this elsewhere to solve an observed fretting problem, namely the connection between RF-BKT-004 and RF-RIB-601 via the RF-PLT-007 in the rib 6 assembly. Here I not only upsized but also increased RF-PLT-007 from 0.5mm to 1mm thickness to reduce the stresses on the rivet material.