The superstructure was another structural challenge, but for different reasons. It was clearly going to be a heavy section, and was planned to rest on four discrete locations without sagging.
The challenge was to have a stiff structure which avoided the volume of the upper gearbox, stayed within the diminishing profile of the superstructure, and bridged the wide opening of the centre section. It also had to allow the mounting of 10 smaller turrets, 4 gun director radars and their associated drives.
Ideally, a deep section would run centrally along its whole length, but the presence of the gearbox precluded that. The solution was to have a deep section for part of it, which transitioned to an open box section of equally deep sides.
To ensure this structure would sit in the right place on the centre section, when lowering down into position it is guided by two vertical pins (with tapered noses) which engage with structure. Two of the resting points (fishplates) can be seen either side at a distance from the taller pin.
When the superstructure is not on the model, it must be able to support itself above a flat surface leaving a gap below it, as some parts protrude below the support points. There are three pairs of ‘feet’ on which the superstructure will happily sit on any two pairs without sagging.
If it sits on the middle pair plus an end pair, a counterweight is needed to keep it from tipping of course. This allowed the underside to be worked on when the build was too far advanced to upturn (flat on my back like when I was younger working under my car!).
The positioning of the feet was dictated by where the centre section could accommodate them.
There were two more important considerations for the structural build. Firstly, there had to be two lifting points – hand holds for lifting it on and off the model – which needed to be firmly connected to the structure. The ship has two large radar booms set above the superstructure – one forward and one aft.
The design was for these to be the fixed lifting handles, but the load path back to the structure was still a challenge, especially the forward lifting point.
The second important consideration was prevention of damage if lifted incorrectly. There is a very inviting handhold just below the bridge that a well meaning soul might use to help lift the superstructure, without first being told of where to lift from.
Lifting it from here would be disastrous as it would buckle the bridge upwards, so I mitigated this possibility during the build of the bridge. This lower part of the bridge wraps around the tall conning tower, making a good load path difficult.
The solution was to make the ‘wrap-around’ multi-strip thick to resist bending, then anchor it firmly at the rear end to the top of the ‘box’ structure. This provided a strong cantilevered arrangement which has been tested successfully, though I wouldn’t want to rely on it too much!
With the basic structure in place, the systems needed to be installed for driving the smaller turrets and gun director radars. Both of these features needed to oscillate side-to-side, but with their drives coming up from the centre section. I also wanted the turrets to move apparently randomly.
The ship has a cluster of 40mm anti-aircraft guns (Bofors guns) between the funnels. Their arrangement provided a void beneath them which was convenient for housing the oscillation mechanism for the turrets.
It was about this time I noticed that some of the photos I had been referring to for the build were in disagreement. It turns out that some of them were of The USS Missouri's sister ships, and they were not all identical. For instance, the picture of the Bofors guns is of the USS New Jersey, so care was needed to stay true to the Missouri.
The drive for the turrets enters from below via a contrate gear. This contrate was originally a small one, but the necessary torque to drive all 10 turrets proved too high for it, as the teeth would ride over the other contrate as per the ‘jam protection’ function. This was resolved by using a pair of mating large contrates, which because of their greater radius of action could transmit more torque without slipping.
The input from this drives a shaft (sprung axially) and pinion, which in turn drives more shafts.
One of these extends down and has a small sprocket at the bottom. A chain – in the horizontal plane – loops around three more sprockets, each of which drives additional gears and cranks. The cranks are all short couplings, therefore each crank has a ½ inch throw.
Because there is very little clearance between the centre section and the underside of the superstructure, the lower cranks drive ‘strip links’ which move within this clearance and feed out to each of the four lower turrets. Note in the picture the smaller contrate at this time.
Because of the various gearing ratios and relative crank positions the turrets move at two different rates, and appear to be random in their motions. Actually, the crank positions are set so that the turrets’ peak friction positions (at the point where they reverse direction) do not coincide, thus reducing the overall peak torque required.
Chain systems need a tensioner, but this one has a requirement to keep the tension as low as practicably possible to minimise shaft friction. A traditional spring tensioner will not provide a constant load if there is fluctuation in the chain free play – which there always is due to the variability of the sprockets and shaft straightness.
The model employs a ‘gravity tensioner’, which comprises a ¾ inch sprocket wheel on the vertical end of a long–legged bell crank made from rods, and a stack of wheel discs mounted on the end of the bell crank’s horizontal arm. This keeps the chain tension and associated friction constant regardless of how much slack is in the chain. This mechanism is sandwiched between the crank drives in the very centre of the structure.
The shaft running above the bell crank pivot (pictured with the cover off) is for the three forward gun director radars.
The 10 small turrets are all identical. The base is a bush wheel, and the barrels are secured to the central rod using an arrangement of couplings which allow the angle of the barrels to differ between turrets.
A row of collars are used as captive nuts, and smaller headed set screws were needed in some of these to minimise the width of the barrel slots. The ends of the barrels were drilled to mimic the barrel bore.
With nearly all the systems in place and the beginnings of the superstructure, the model was getting to be a bit busy inside!