Site AtomHistory.ru published a very interesting document from the early days of the Soviet nuclear weapons program. It's the first page of the Decision of the Council of Ministers regarding the program of nuclear tests in 1953. Click on the image to go to the full-size image at the atomhistory.ru site. (Thanks to the Centre for Analysis of Strategies and Technologies for posting the document on their blog.)
Among other things, the document provides some details about the amounts of fissile materials used in the first Soviet nuclear weapons. Below is my translation of the document.
It is clearly marked as "Declassified" on top of its original "Top Secret/Special Folder" classification. As it was the case with most documents at that level of secrecy, some words were written in by hand - the information was too sensitive to entrust it to typists. The hand-written words are in italics. These precautions notwithstanding, the document does not refer to fissile materials by their actual names - plutonium is "tellurium-120" and uranium is "tin-115" (this is a bit strange - in earlier documents plutonium was "ametil" and uranium - "kremnil"). "Test site No. 2" is the Semipalatinsk Test Site. I translated "izdeliye" as "device".
The Council of Ministers of the USSR
of "__" _____ 195 3 No. ___
On the goals and the program of tests at the test site No. 2 in 1953.
The Council of Ministers of the USSR HAS DECIDED:
1. To accept the proposal of the First main directorate of the Council of Ministers of the USSR (Zavenyagin, Slavskiy, Kurchatov, Pavlov) to carry out the following tests of the RDS devices at the test site No.2 in July-September 1953:
a) A prototype of the "RDS-6s" device with the purpose of carrying a thermonuclear reaction, measurement of the speed and duration of the termonuclear reaction, validation of the calculations that formed the basis for the "RDS-6s" design and obtaining the physical data required for validating the design of a combat "RDS-6s" device.
It is expected that the full TNT equivalent of the device will be no less than 250 thousand tonnes.
b) An RDS-4 device with the main charge from tellurium-120 with the weight of 4.2 kg and tin-115 with 90% concentration with the weight of 6.8 kg with the purpose of determining the full TNT equivalent of the RDS-4 device with full weight of 1.2 tonnes.
It is expected that the full TNT equivalent will be 25 thousand tonnes.
v) An RDS-5 device with the main charge from tellurium-120 with the weight of 2 kg with a normal neutron background with the purpose of validating the possibility of obtaining an explosion of the main charge of a hollow design containing 2 kg of tellurium-120 and determining the full TNT equivalent.
It is expected that the full TNT equivalent will be 9.2 thousand tonnes.
g) An RDS-5 device with the main charge from tellurium-120 with the weight of 0.8 kg with a normal neutron background with the purpose of validating the possibility of obtaining an explosion of the main charge of a hollow design containing 0.8 kg of tellurium-120 and determining the full TNT equivalent.
Here is where the first page of the document ends. Apparently, the document continues and describes at least one other test that was conducted in 1953.
This appears to be a working document - the paragraphs that describe yield estimates are marked by a blue pencil, as if someone decided that the decision should not commit to specific outcomes of the tests. It is also possible that the the estimated were revised later. The actual yields were indeed a bit different from the estimated in the document.
RDS-6s, the first Soviet thermonuclear weapon (this is the "Layer Cake" design), tested on 12 August 1953, yielded 400 kt - 60 percent higher than the first estimate. Apparently there was a reason RDS-6s was considered a prototype rather than a weapon ready for combat deployment.
RDS-4 was tested on 23 August 1953 in an airdrop. The yield of the explosion, 28 kt, was fairly close to the original estimate. This weapon would later become known as "Tatyana" - the first Soviet tactical weapon. (By the way, the online version of the list of Soviet nuclear tests gives the yield of 23 kt. In the hard copy of that book as well as in other sources, the yield is given as 28 kt.)
The final three tests of the 1953 series were conducted in quick succession - on September 3, September 8, and September 10. We know that two of them were different versions of the RDS-5 design. The 2-kg core weapon is most likely the one that was tested on September 3. Its actual yield was 5.8 kt - quite a bit lower than the estimate. The 0.8-kg weapon is probably the September 8 test - it yielded 1.6 kt (we don't know what the yield estimate was in this case - it's on the second page of the document). The third device, the one tested on September 10, was probably a version of the RDS-5 design as well - as far as I can tell, there was no other design in the development at that time.
The "hollow design" in the description
is apparentlymight be what is known as a "levitated core" design with a gap between the core at the center of the device and the imploding shell - at least this is the kind of design that was discussed in 1949. Or it could be the "hollow pit" design, without the solid core in the middle. [UPDATED: It doesn't look like it was the design described in the 1949 documents]. I'm not quite sure what to make of the reference to the "normal neutron background", though. They must have used some kind of a neutron initiator.
Overall, this document gives an interesting data point - it is not very often we see documents with amounts of materials used in actual nuclear explosive devices, even though these amounts are not very hard to estimate.