This will be my last reply on the subject. On Fri, 23 Feb 1996, jim bell wrote:
At 08:05 PM 2/22/96 -0500, Black Unicorn wrote:
Your intermediary chamber, if surrounding the blasting cap, is likely to detonate to one side first, at a right angle to the axis of the chamber to the explosive assembly.
What I anticipated, to tell you the truth, was a long intermediary thin tube (again, 1 mm diameter for concreteness, just as an example) BETWEEN a chamber surrounding the cap, and the secondary chamber. (the secondary chamber would be carefully designed to spread the shock front evenly) I fully intended to avoid all of the possible consequences of weird explosive modes in common blasting caps.
Then you still have the problem of the connection of the exceedingly thin tube to your starfish/wagon wheel where the individual tubes branch off to the compressive explosive assembly. This puts you right back to square one, the need to mill these connections with exceedingly close tolerances.
What really mystifies me is that you would think somebody who was intelligent enough to be capable of building a bomb could possibly be unaware of the strange behavior of common blasting caps? Do you think we're all stupid out here?!?
Considering that you never discussed the specifics of blasting caps or the manner in which they might influence your design, I think the safe assumption is that you never thought of it. Your paragraph above is poor spin control in that regard.
Last time I talked to Dr. Edgerton in his lab (You _do_ know about Dr. Edgerton, don't you?!? EG+G?), in about 1978 or so, he showed me some interesting pictures he had taken of blasting caps exploding, and the weird patterns they made. Believe me, from that moment onwards I had no illusions about the predictability of the common blasting cap.
This is name dropping. The reader will notice the need to bolster credibility by association with a "big" name. The conclusions are obvious.
BTW, the reason Edgerton paid a bit of attention to ME, as opposed to every other lowly undergrad at MIT, was the fact that I did something he had tried many times and failed to do: For my strobe laboratory project, I decided that I was going to photograph a popcorn kernel opening up at 10,000 frames per second. He called it "impossible": I called it a challenge. That is why I did the project. I showed him 11 frames taken a few weeks later.
Dr. Edgerton was suitably impressed.
The reader will notice now how the author has gone from associating with the "big" name to being superior to the "big" name. The intent is to bolster holed credibility even further. This tactic will be recognized by the astute political observer as "link and exceed." (Most often used with our favorate "big" name, JFK.
1> Interference from the milling shape and accuracy o f the openings to the tubes containing the liquid explosive.
Quantify, quantify. How much of a problem?
Clever question given that I am without any information as to the exact shape of your tubes, if they are bowled down towards the explosive assembly, or what their exact width (excepting your vague 1mm figure) might be. You make some guesses as to material, but these two are fairly flimsy even by your own admission.
I don't expect you to be able to "use ESP" and anticipate all the exact mistakes somebody could make. Rather, you should be willing to accept the principle, and explain how much inaccuracy is "too much," and try to give an example of an error that would produce an inaccuracy of this magnitude. So far you've done none of this. I have to conclude you were simply trolling, or intentionally spreading FUD (Fear, Uncertainty, and Doubt) without genuinely trying to get involved in an interesting hypthetical idea.
Without more on your shape and size of the core, its essentially impossible to say what timing difference will be debilitating. I did infact give an example related to fractions of the branch tube lengths. Without more from you, the supposed originator of the idea, no one can tell you how much timing error is enough. The answer in general form is: Enough to cause the center shape to displace enough of it's mass away from the direct brunt of force from the opposed charge. Most excessively, some of the branch tubes might fail to initiate all together. The quantity of all this depends so specifically on the type of explosive used for the compression, the exact density of your core, the number of facets you use in your compressive assembly and the location of initiation on the facets so as to be entirely useless without precise details. What I can tell you is that it's enough of a problem to make high speed precise switches necessary. Be real, if the timing problem was enough to make even normal resistance matching a problem, how the hell are you going to solve it with a much more volitle and violent initiation method?
BTW, while I do indeed consider this as purely hypothetical, on the offchance you're a FUDmeister from the government, you should be aware that _I'm_ fully aware that while the main form of radioactive emanation from Pu-239 is alphas which can be stopped by a piece of paper or a few inches of air, I am also fully aware that the decay produces a substantial quantity of gamma radiation (whose exact wavelengths and energies I can easily look up in my trusty CRC Handbook). So don't bother flying an airplane equipped with a gamma ray detector over my house;
[additional paranoid rantings intended to bolster the authors image by substitute knowledge deleted]
All you need to realize to appreciate the problem is that if you do not have a precisely milled end, with a precise depth into the compressing high explosive outer face, you have differences in how and when the various faces of the explosive assembly are going to initiate. If you make your tubes narrow, it becomes very hard to mill the ends of your tubes, and if you widen the tubes, it exagerates the distortive effect of irregularities in the tube ends.
But you should be able to estimate the magnitude of the errors. Given a certain detonation velocity, for example, and assuming some sort of localized slowdown/speedup to to this velocity, you should be able to estimate (even if only accurate to a factor of 2-3) the amount of error present at that particular junction.
I did that with the example related to tube length. (Which you quite cleverly deleted) I'm hardly going to sit down and do models of shaped explosive dynamics and hydraluic shock analysis simply to tell you you're way off. If your so interested, examine the problem yourself. Again, the fact that you have never
done even this rudimentary analysis is quite telling. You've revealed nothing that I wasn't aware of, and that was apparently quite intentional.
What you were and are aware of is a matter for significant speculation, complicated by your characterization of shape charge analysis (especially with liquid explosive) as "rudimentary analysis."
I'm not in the business of designing nuclear initiators. I expose poorly thought out explosive engineering as a hobby. Your best solution is to mill each tube exactly alike, right down the the degree of bend and slope of arc as well as shape of either end. But you could have figured that out without me spelling it our for you.
More likely, I would have velocity-tested sample configurations down to 10 nsec accuracy, which would have revealed any unexpected error sources from temperature and/or pressure variation, as well as mechanical considerations such as bent tubing, etc...
But since you haven't tested it, and don't plan to, we can't know can we? Moreover, since your design was only marginally if ever interested in these problems, one has to assume you're way off the ball. Your 10ns accuracy figure is pulled right out of the air and in no way represents a figure you know you can obtain. I don't believe such accuracy can be obtained with this kind of physical inititation. I'm not telling you to stop trying, you can do what you like. I will, however, expose your oversights in order to facilitate reputation capital distribution by the list. (To which I now apologize for the massive spam).
[stuff deleted]
But they don't. The timing problem is quite significant. Why do you think high speed and superaccurate switches are so well guarded? There isn't an easy grassroots substitute, if there were, the switches would be fairly useless.
Maybe that's the secret. I already anticipated this.
[You keep saying that, (I already anticipated this.) I'm not quite sure you know how much you sound like an apologist] If it's the
government's motivation to keep "terrorists" from trying to build a bomb, then their first line of defense might be to make it appear more difficult than it really is. They also know that secrets which are actually turned into running, installed hardware eventually leak to the public, meaning that it might actually be better to keep THEIR bombs complicated, and to not use simplifying hardware.
Oh, whatever. Get a grip. These paranoid rantings intended to deflect and distract from the main failings in your design are more than obvious.
Remember, kryonic switiches are necessary even when dealing with the speeds of electric conductivity. The velocities of even hydrazine based explosives are signigicantly lower. The margin for error is similarly lower.
How low? Be specific.
Again, I don't know what your dimentions are. Hydrazine explosives tend to detonate around 8500-10000 m/s. The speed of transmission of electric impulses through a given conductive medium is certainly much higher.
Why do you keep mentioning "hydrazine explosives" when I didn't? Are you some sort of "one-trick pony"?
No, they are just the simpliest liquid explosives to obtain and make, and happen to have the highest detonation velocities. I assumed you would be using them. Perhaps I overestimated the depth of your thought on this matter.
You are correct this time. My fault. Uranium should have been in there. Typo on my part.
Finally! He's able to admit a MISTAKE!
When I make them, I admit them.
Hey, be my guest. If you had a critical mass worth of plutonium you're playing around with the wrong list, and, I might add, wasting your time with anything but the black market for the material.
If I had some, or for that matter if I even wanted some, would I be advertising the fact on an "NSA-required-reading" list BEFORE I'd done all this work?
As I said before, the degree of your expertise and wisdom is the primary issue in debate. --- My prefered and soon to be permanent e-mail address: unicorn@schloss.li "In fact, had Bancroft not existed, potestas scientiae in usu est Franklin might have had to invent him." in nihilum nil posse reverti 00B9289C28DC0E55 E16D5378B81E1C96 - Finger for Current Key Information