I am slow motioning the Netanyahu presentation on Iran's prior bomb program (pre-2003), taking screen captures if technical items of note. Up to implosion system so far (which on first impression aligns with prior leaks and technical analysis thereof). Work in progress.
1/ Ok. Having finished with capturing screenshots, brief policy points:
A) None of Netanyahu's framing or fact claims exceed the conservative assumptions I and other NP people used to assess value of JCPOA.
B) Tech not worse either. Just better details, sources.
A) None of Netanyahu's framing or fact claims exceed the conservative assumptions I and other NP people used to assess value of JCPOA.
B) Tech not worse either. Just better details, sources.
2/
C) We all publicly assumed Iran was lying about prior military weapons programs in pre-agreement assessments. Not news.
D) I still believe JCPOA was the right first solution to prevent Iranian weaponization. Nothing here changes that.
C) We all publicly assumed Iran was lying about prior military weapons programs in pre-agreement assessments. Not news.
D) I still believe JCPOA was the right first solution to prevent Iranian weaponization. Nothing here changes that.
3/ See other comments by @ArmsControlWonk @aaronstein1 @mhanham et al (MIIS CNS staff, other experts) for more policy and technical stuff (and @DaveSchmerler undoubtedly putting his Big Geolocation Hat on right now...)
4/ Now on to technical stuff. This is going to take a little while please be patient.
5/ (transferring files to my Twitter platform) ..... 3% complete .... 🙄🤦🏻♂️🤷🏻♂️😹
6/ (50% downloaded! Cooking with WiFi!.... 🤷🏻♂️🤦🏻♂️)
7/ Ok got them and trimmed excess screen space. Here's an original capture with the YouTube URL, rest are more trimmed: 
8/ Ok trimmed. Serious technical discussions starting now.
Cyan is HEU fissile levitated pit.
Purple is reflector / tamper layer lining explosive, which is Orange and unlensed. Green is branching explosive channels Multipoint Initiation layer. Grey outside case.
Cyan is HEU fissile levitated pit.
Purple is reflector / tamper layer lining explosive, which is Orange and unlensed. Green is branching explosive channels Multipoint Initiation layer. Grey outside case.
9/ This is a small test unit for the branching explosive channel Multipoint Initiation layer. Top, single detonation Initiation point. Around it pins are electronic sensors to detect when detonation front reaches each point. Briefly: detonating explosive gas conducts electricity.
10/ Start of implosion simulation from Multipoint Initiation.
11/ The detonation waves moving inwards and merging. Unevenly; the explosive layer is not lensed and the waves stay uneven as they move inwards. The unevenness is why MPI unlensed designs are relatively rare in nuclear weapons.
12/ The detonation front reaches the tamper/reflector layer that lines it. It's moving inwards and deforming from the uneven impact.
13/ The tamper/reflector hits the fissile pit or core...
14/ ...and the core compresses. Core and reflector are hopelessly uneven at this point but it doesn't matter. Will measure the explosive compression asserted by simulation later but it's decent. Neutrons are multiplying fast now and the very bright lights come next.
15/ That's it for that (physical implosion, not actually neutron/yield) simulation. Next an integrated warhead design. Need to measure but with a 1.2 meter missile undit I think that's smaller than R265, the previously leaked design. Won't know until I measure.
16/ A little out of order, but a hemispherical half-weapon assembly mockup. Two of these with real material, proper pit mating interface and a simultaneous firing mechanism and you have a nuclear bomb.
17/ Back in sequence, blueprint of the underground areas at Fordow. Have to overlay on mountain pictures and maps but the entrances match my memory of the two portals.
18/ ...and as he approaches wrapping up, his overview of the program components. Nothing technical here. End of my screen capture run (I have more but aren't relevant to technical aspects, will let others focus on policy and politics while I catch up with replies and questions.
19/ I'm making lunch and will start measuring dimensions on relevant screen captures once on a laptop again. Details so far match prior information and my study variations ranges well with one exception. ...
20/ The exception is:
The levitation gap between the U235 core ("pit" in nuclear weapons terminology) and the inside of the tamper/reflector layer in that simulation is wider than I expected. It's wider, more empty space than I expected but within analytic range I studied.
The levitation gap between the U235 core ("pit" in nuclear weapons terminology) and the inside of the tamper/reflector layer in that simulation is wider than I expected. It's wider, more empty space than I expected but within analytic range I studied.
21/ Neither my studies nor the shown Iranian simulation shows much velocity gained or unevenness relaxation in the extra space, but will think about it.
22/ The implosion package in the warhead design shown also seems smaller than the R265 design leaked before, but I will measure that picture in detail and veryify the missile dimensions before concluding that. Not much of a surprise if R265 wasn't the last design.
23/ Brief meta thread on truth of knowledge. Relevance: Iran claimed the prior leaked details on the concept and R265 design were elaborate forgeries.
24/ It's not possible for me or anyone else to prove or disprove the alleged origin of the documents. We have to assess in context. Contextually, current disclosure is aligned perfectly at technical level with public information from IAEA & Iran re other aspects of their program.
25/ The current disclosures match the R265 details very well as far as I can see so far. There's an incomplete but credible picture here.
26/ Based on R265 and Danilenko publications, I went in analyzing possible designs that matched the descriptions. Many of those were viable nuclear bombs as far as my analysis went. This info needs detail review but feels credible given my prior work.
27/ So the credibility of the picture is strengthened for it being legitimate stolen Iran historical documents and data.
But...
But...
28/ ... but, it's still possible that this is a detailed forgery. I find it hard to believe someone would release workable nuclear bomb designs at this details level to frame another country but it's possible. I can't rule that out.
29/ I can't assert absolute certainty or truth regarding the documents and design(s), but am reasonably sure they're credible as real workable bombs, matching other Iranian program information IAEA believes with high certainty.
30/ I can understand Netanyahu filing off / blurring the internal dimensions and a few key technical details, but as wide as possible uncensored access to the files will help other experts in fields from program management to Farsi language to Iranian political history to assess.
31/ I encourage widespread release and review by all relevant experts to help review credibility of the information claims.
32/ Mini-mini-meta-thread: This design is totally different than the two North Korean bomb designs we saw mockups of and believe we saw tested. No technical connection in the implosion systems at all. (End mini-mini-meta-thread)
33/ Aaron brings up Eugene Danilenko in linked RT; let's add context there.
34/ I think I misremembered Danilenko's first name. Will certainly be corrected soon.
Anyways, Danilenko is a former Soviet nuclear weapons engineer/scientist who worked on implosion systems. The elements that push inwards on bomb cores.
Anyways, Danilenko is a former Soviet nuclear weapons engineer/scientist who worked on implosion systems. The elements that push inwards on bomb cores.
35/ Danilenko worked for a while after the end of the Cold War doing "implosion nanodiamonds" which are formed in much smaller simpler versions of nuclear explosive implosion systems. They're used industrially, for metal cutting surfaces mostly.
36/ Danilenko's work on implosion nanodiamonds is openly published in technical journals, company reports. IAEA and various open sources have detailed overviews.
THEN he headed to Iran in 90s, to work on "implosion nanodiamonds" or so he said.
THEN he headed to Iran in 90s, to work on "implosion nanodiamonds" or so he said.
37/ Again, his presence there is public and he's admitted it. Long established, etc.
His techniques for implosion use the branching explosive channel Multipoint Initiation approach. He said he only gave Iran industrial nanodiamond technology, but R265 and these bombs used it.
His techniques for implosion use the branching explosive channel Multipoint Initiation approach. He said he only gave Iran industrial nanodiamond technology, but R265 and these bombs used it.
38/ The branching channel Multipoint Initiation is conceptually simple. Small explosions in a narrow channel (say, 1mm x 1mm) milled into a 1cm Aluminum plate won't directly set off insensitive explosive under the plate.
39/ You drill small (5mm) holes from the channels down to the underlying explosive and fill them with explosives to start the underneath layer exploding.
40/ The channels are initiated (detonated) starting at one end and then will detonate to their other end. Pretty predictable, even speed especially if you put a flat cap plate over it to tamp it.
41/ The channels can go in multiple directions from the Initiation or first detonation point.
Channels can curve or zigzag if you need, with slight predictable detonation slowing.
Channels can also split or branch off, so 1 becomes 2 becomes 4 becomes 8 etc.
Channels can curve or zigzag if you need, with slight predictable detonation slowing.
Channels can also split or branch off, so 1 becomes 2 becomes 4 becomes 8 etc.
42/ One can with a little effort draw branching patterns that spread out with equal length paths to many of the explosive plugs going down to the underlying layers. You more or less simultaneously set the underlying explosive off at all those plug points. Multipoint Initiation.
43/ Doing a bomb implosion system requires a round sphere not a flat plate, so the spherical geometry is a little different than the flat plates. But you can do the same branching patterns with relatively even spherical coverage.
44/ The Iranian R265 reportedly used four quadrants on each hemisphere; one common Initiation point at the top or bottom, distribution channels in 4 directions presumably down roughly to each quadrants center, then back out to evenly cover the surface.
45/ The initial detonators top and bottom need reliably simultaneous detonation, but that's easy with any of Exploding Bridgewire, Exploding Foil, or Laser Detonators.
46/ The advantage of this is that with one simple flat or curved hemispherical plate you get from one detonation point to near simultaneous detonation at points all over the hemispherical surface. You can put a solid, uniform sphere or hollow sphere of explosive under.
47/ The disadvantage is that unless the explosive layer is thick compared to the detonation plug spacing, you get uneven shape of the inwards moving detonation front. Which their simulation shows clearly.
48/ But with suitably robust, high margins designs (a deformed tamper/pusher plate hitting a large highly enriched Uranium solid core counts...) the unevenness won't cause a failure. So you have a rough implosion that's robust and very very easy to design, test, &manufacture.
49/ Simplicity of science, design, engineering to product, testing program is so good that none of the tasks is beyond a single well motivated wide skill base individual. I know several...
(Lifetimes of effort involved, but it's just work not impractical to understand).
(Lifetimes of effort involved, but it's just work not impractical to understand).
50/ It is going to be using more Uranium and be bigger and heavier than alternative implosion technology. But not every nuclear weapons state needs cutting edge light or minimal fissiles use per bomb.
51/ Late tech note: Netanyahu said (or strongly implied) this was a MPI implosion test unit. I am now strongly suspecting it was misinterpreted (by them and then my using their term).
I now believe this is a Pin Ball Array.
I now believe this is a Pin Ball Array.
52/ The varying pin heights should have tipped me off earlier, but I went with what he said. My bad.
A Pin Ball Array sits inside the complete implosion assembly and electronically measures shape of imploded tamper/pusher over time, as the T/P hits the pins.
A Pin Ball Array sits inside the complete implosion assembly and electronically measures shape of imploded tamper/pusher over time, as the T/P hits the pins.
53/ Basically the varying height pins electrically short with the tamper/pusher when it hits them, and using precision oscilliscopes or timers you can tell what was where when during the implosion.
54/ There's a great picture of one floating around somewhere and I think I had an archive copy but still looking. Will post it when I find it.
55/ aha. str.llnl.gov/str/Grissom.ht…
That ones fiber optic and the term they're using is Pin Dome, but I've seen some variations and documented both electronic and fiber optic pins.
That ones fiber optic and the term they're using is Pin Dome, but I've seen some variations and documented both electronic and fiber optic pins.
56/ Found a DOE test method and results report on testing detonation in XTX-8004 (RDX/Sylgard 182 rubber) in thin channels using flat electronic switches on a printed circuit board. Can use multiple sensor methods; conductivity pins in bulk explosive also common in testing.
