You’re thinking of the other bomb, the U-235 one, which they didn’t test at Trinity and which was dropped on Hiroshima. That is two separate pieces of Uranium that are slammed together to create a critical mass. The Pu-239 core was a single sphere of metal. It was subcritical until you compress it down with a spherical implosion from explosive charges all around it (from the size of a grapefruit to the size of a lime), at which point it reaches a high enough density to go critical.
The gun-type bomb (where a subcritical mass is shot into another subcritical mass) is very simple to build once you have the materials to do it. They didn't think it needed a test since it was pretty obvious that it would work.
The implosion design is tricky. You need to arrange and detonate the explosives precisely to compress the core evenly from all sides, otherwise it shoots out the side or otherwise doesn't go bang the way you want it to. Hence the test.
That trickiness can be a good thing. Almost all modern weapons use the implosion design, partly because it's much safer. With a gun-type design, an accident could easily cause the two pieces to contact each other, resulting in an unwanted detonation. With an implosion design, accidentally setting off the explosives is very unlikely to set them off with the correct timing, so you'll probably just lose the core.
The implosion design is also a lot more efficient. Little Boy used 64kg of uranium. Fat Man used just 6.2kg of plutonium and even got a bigger bang out of it.
It is all true, but one needs to take into account that because of the different properties of the materials, the critical mass for uranium-235 is intrinsically much greater than that for plutonium-239.
For a bare sphere, it is about 10 kg for plutonium and 50 kg for uranium.
>(from the size of a grapefruit to the size of a lime)
Whoa. Its hard to imagine you could have enough conventional explosives to compress a dense metal by ~10x (?). You'd need some serious containment to direct that energy inward rather than outward. I suppose I have some reading to do.
Plutonium was compressed about two-fold by volume.
There is a story about it. When they first brainstormed the ways to make the bomb, even before Los Alamos, in 1942, one of the several ideas was to use explosives to throw smaller pieces of material together, to make the super-critical mass. This was dismissed as too imprecise, but it was still listed in the April 1943 as one of the possibilities in the Los Alamos Primer, which was the orientation booklet for the scientists joining the project.
One of the scientists, Seth Neddermeyer, fell in love the the idea and talked the bosses into letting him try it. He consulted with the explosives experts in Pittsburgh and started some crude preliminary experiments.
When von Neumann was told about these experiments in October 1943, he immediately pointed out what when the pieces of metal slam together at a high velocity in the center, this creates extremely high pressures. Teller then remembered that at such pressures, iron in the Earth's core becomes slightly compressed. They instantly realized that compression makes the exponent in the chain reaction greater, and that this is a new way to make the bomb. They explained the idea to Oppenheimer, and he pivoted the project to the new method.
This did not work. The material did not assemble into a neat ball, but was just making a mess. But Robert Christy, the guy who was making the calculations for this, realized in September 1944 that the slamming of the pieces together at high velocity was not strictly essential, and that a solid ball of metal could also be compressed by an inward going shock, although not as efficiently. Because this was guaranteed to work, this was chosen as the design for the "Gadget".
Ironically, Seth Neddermeyer, who was instrumental for this to happen, has never accepted that the metal could compress.
They struggled with many things, often time the minutiae of accomplishing something conceptually rather simple. For example, making an explosive with a significantly slower detonation velocity turned out to be very tricky. The concept was simple -- just add some barium nitrate to the TNT. But if you just did that, the mixture stopped flowing nicely, and it still was either not slow enough, or refused to explode at all. Extreme technological nuances were required just to prepare a mixture of two simple ingredients before satisfactory results were obtained. This one thing was its own research project.
Accurately casting explosive in odd shapes, without different ingredients separating, and without producing voids when the melt solidified, required developing a whole new technology with careful gradients of temperature in the molds.
They tried lots of different commercial and handmade detonators to find which ones would work most consistently. That took an awful lot of time.
The electronics itself was probably least difficult -- a microsecond was already a very long time for the electronic circuits even in 1945. One could use an off the shelf oscilloscope to see if the detonators worked simultaneously or not. Incidentally, 2/3 of the cables in the famous picture of the "Gadget" are not the detonators, but the simultaneity sensors -- reporting the difference between the earliest and the latest detonation fronts.
Everything was tested extremely extensively. Tremendous resources were spent on testing and test equipment. All in all somewhere between 20000 and 40000 explosive tests were performed at Los Alamos during the project.
It is not often emphasized how much of the work was done in the explosives laboratory in Pittsburgh before passing it on to Los Alamos. They have developed the slow explosive. They also reproduced from the earlier British work and further developed and tested the concept of the lenses, together with many other more advanced things which did not find an immediate application in the bomb. The director of the laboratory, George Kistyakowsky, took over the explosives work at Los Alamos, once the implosion became the main focus of the project.
