Old photos of the SHB show massive cables holding the arch as it was built and once they met and were secured, the cable were removed.
Although like many I have been around the SHB but have never seen any signs of where the anchor points were.
No doubt they were covered as part of the design once the Bridge was finished.
As the arch is basically one piece, the deck has expansion joints to allow it move and as the whole bridge sits on massive rollers, in theory with enough 'Huffing and Puffing' like 10,000 politicians promising at Election time, the whole bridge could be moved backwards and forwards.
I always thought the rivets used had a head and a solid shank not hollow or split (bificated).
As we have already heard, the cables, 128 on each side, were fed from underground anchor points, and supported the northern and southern half spans during erection. When both half spans were completed, the cables were gradually slackened off, and the halves met. The meeting was absolutely spot on for alignment, and the halves were joined. A few hours later, they had separated due to a temperature change, and a final meeting and joining took place at about 10 pm. The cable anchor points are now well buried under the approach spans on each side.
The arch is one piece, but is not supported on rollers. The main support pieces are the king posts; one on each side of the arch at the south end, and the same at the north end, i.e. four king posts. Each king post is close to vertical, and supported on a massive, cylinder-shaped bearing. The arch is self-supporting, and may expand and contract up to 18 centimetres due to variations in ambient temperature. Each bearing can support 20,000 tonnes.
Once the arch was complete, the "creeper cranes" which had lifted the steel components into place were at the top of the bridge. Rather than move them down the arch, the construction of the road deck started from the middle of the bridge and moved outwards towards both ends.
The rivets are indeed solid shank and not bifurcated.