Pokhran II not fully successful: Scientist - Part-2

[shiv]

Praful Bidwai's most enduring nickname on this forum is Purefool Bidwai. The reasons for such a nickname are not difficult to discover because he (like many other journalists) uses journalistic rhetoric to make a point that often does not stand up to scrutiny. Journalistic rhetoric is the literary equivalent of embellishing a lump of dung with cream and strawberries to make it appear edible so at least a few readers will swallow it.

From the article linked above:
http://www.hinduonnet.com/fline/stories ... 908000.htm

For instance, the nuclear-propelled submarine, recently released for trials, was to cost under Rs.1,000 crore when the Advanced Technology Vessel project was started in 1975. Thirty four years on, the meter has clocked Rs.30,000 crore.

May I indulge in dung decoration myself?

Thirty four years ago an intern who had just passed his MBBS used to get paid Rs 150 per month. Today he earns a hefty 7500 a month - a 3000 percent increase. This only reveals the avarice and insatiable desire to make money among young doctors - in a profession that was unable to handle Swine flu in 1974 and is still unable to do that 34 years later

[enqyoob]

shiv, it's the Cost Overrun, showing how inefficient doctors are becoming in the Capitalist Imperialist dung-heaps. What Louis Pasteur did for a croissant, a glass of wine and a kiss, now costs Rs. 3,000 and takes six people and 3 laboratories and 2 overpaid capitalist Pharmaceutical companies to deal with.

And the most shocking thing is that they still waste time focusing on the human who gets bitten by the mad dog, and kill the dog, instead of rewarding the mad dog as the Marxists would do and shooting the human for being in the way of its teeth.

[shiv]

Otherwise, I would have used, say, 5 permanent sensors at Khetolai and 5 at the logistics base, from which I record signals every time there is an explosive test conducted. As long as the test stays below the threshold, I don't need to worry about individual houses, just the signals at one sample location, or, for overkill, at all ten locations.

This brings me to a thought I have had about calibration of the geology of Pokhran..

Guess where eLlCeeyAy goes to test its bomb delivering cabability. And where is the Vaysena bombing range. I just wonder if the little phatakas that are delivered by Vayusena send out minute seismic signals that can be used for calibration?

[samuel]

Now you are talking down at me unfortunately..OK. First about hurricanes.

http://www.aoml.noaa.gov/hrd/ike/Files/ ... _final.pdf
Please read it.

The laws of physics dictate that the wind damage for
a given structure must be proportional to the rate
of work done by the wind (and not the force exerted
by the wind) on the structure. This depends on the
deflection of the structure under wind loading,
but it is certainly not proportional to just the force.
Because the deflection itself is proportional to the
force exerted, its dependence on wind velocity must
be higher than just its square. It could be equated to
the dissipation rate of the wind kinetic energy, and
hence taken as being proportional to the cube of the
wind velocity, as suggested by ...

You can come to us and give us a talk about what you know.
I am sure our department will be interested.

Back to system ID. Happy to engage and I'll be happy to read the wiki page, but please consider coming to one our classes!

[enqyoob]

Yes, that's right, I pointed that out 2 posts above - the weapons testing is done at the Pokhran range. They have decades worth of routine explosion test data.

Then again, as you know, none of the Indian planes has ever hit any target because the missiles are all fizziles and the planes themselves are called "flying coffins" etc. Everyone knows that all the Kargil peaks are still in Pakistani hands. So the "weapons tests" are done following the Sergeant York Air Defense Gun Testing Method where the VIPs sit around for the demo, the (unmanned) helicopter hovers above the AA gun for 5 minutes while they pretend to aim the gun, and then simultaneous charges are set off in the helicopter and the gun barrel. So similarly, all the "target practice" of the IAF must consist, per Bidwai & Co, of the planes making low passes, while the targets are blown up by buried explosives.

[enqyoob]

************** First let us dispose of these *************

Now you are talking down at me unfortunately..OK. First about hurricanes.

Sorry to hurt your sentiments, but can we debate this without getting down to such levels? If not, please accept my congratulations on your impeccable state of knowledge, and let us end this discussion. I only posted the fact that I have used System ID for complex things, to avoid hurting your sentiments and keep you from pushing an unsustainable line of argument and then getting all H&D-damaged. Not to keep you from understanding the concept, which is actually quite easy once you get the hang of it. The off-topic flamebaits about giving talks at your fine institution etc. are unwarranted, and will lead in short order to the damage to H&D that I was trying to avoid, so please delete those NOW.

Thanks

*********************************************************

Samuel: while reading, also pls do not hesitate to think:

Because the deflection itself is proportional to the force exerted, its dependence on wind velocity must
be higher than just its square. It could be equated to the dissipation rate of the wind kinetic energy, and
hence taken as being proportional to the cube of the wind velocity, as suggested by ...

Kudos to whoever "suggested" that, but that does not make it true. It just shows what sorts of things get published.


Solid objects typically have linear Stress-Strain relationships. "Strain" is usually defined as Deflection Divided by Initial Dimensions. Stress is Force per Unit Area.
Deflections ARE dependent on FORCE, not work, until the structure comes off its foundation.
After that, the distance it moves indicates work done on the structure by the wind, which still depends on the forces exerted by the wind, which still depend on dynamic pressure, which is 0.5*density*speed-squared.
May not be proportional unless the structure's elastic coefficient is constant over the range of force magnitudes. But for small forces, the deflection IS proportional to force.

Wind force is indeed proportional to square of velocity, not to cube. Hello????

The rate of dissipation of kinetic energy has units of kg*(m^2/s^3). Please show me how this "equates" to the dimensions or units of "deflection" which as I understand, would be "m". The paper you cite must give the constant of proportionality, and I am curious to know what that is, and why it would be constant.

As for the System ID aspect, the issue here is not whether I want to come to your university to give talks, but whether System ID is fundamentally impossible as you have been trying to prove by your examples, or whether it is done routinely enough that many books have been published by practitioners, and machines sold since the 1970s to do it efficiently.

Another example where a completely non-isotropic structure is characterized by System ID is composite aircraft components such as wings and tails, which have numerous elements of different directions and properties. The whole concept is to say:

Look, I only want to know the relationship between an excitation applied here and the response felt there. Something that includes all the effects of all the stuff in between, but does not require me to know its intricate details

So it needs only one sensor at the excitation and one sensor where you are interested in knowing response.

If you use more sensors, you learn more about the medium, that's all and can make more informed predictions.

[ramana]

Note Praful Bidwai writes in Frontline:

Fizzle?

And follow his main argument being made here also.

from the main headline in that article:

The claim that the May 1998 thermonuclear test failed should not be used to demand further testing. India does not need hydrogen bombs for security.

Saracsm is lost when one is on overdrive. I was referring to the dont test India doesn't need H bomb lobby.

[samuel]

The paper cited is this and doesn't appear to be so much junk, but who knows, Nature might be slipping:

Emanuel, K., 2005: Increasing destructiveness of
tropical cyclones over the past 30 years. Nature,
436, 686–688.

This is the paper you need to read. It appeared in nature...Would you like to talk to him?
http://www.mindfully.org/Air/2005/Cyclo ... 4aug05.htm
Read equation 1.
And this article:
ftp://texmex.mit.edu/pub/emanuel/PAPERS/hurrpower.pdf

I'll type up the derivation from my notes tomorrow.

As regards system id, you can do all you like to quantify local (in space or time) behavior, it doesn't solve the problem. People do it routinely. We will need to set up a real example, meaning with right functions etc. to see that. But let me type up a critique of the papers you cited, which are about local analysis, again.

S

[enqyoob]

Samuel, at the risk of continuing a discussion that is clearly off-topic, (but only because you dragged it there), may I request that you stop trying to make arguments by posting stuff that you have not thought through? Otherwise there is a risk of suffering what some other postors suffered, and I have no wish to make more postors suffer.

OK, I read your equation 1, and nowhere does it, or the associated discussion, say that the damage to any given house scales as v^3. The Total Power of a storm may be proportional to the cube of peak wind-speed, no problem there, and bigger storms may cover larger areas, no problem there either. So thanks for your offer to put me in touch with the author (another attempt at flamebaiting, no doubt), but no thanks.

The excerpt that you quoted before was self-contradictory, because it said that "deflections are proportional to (wind) force" which is true, and then claimed that (wind) force does not depend on square of wind speed, which would be nonsense.

If that was published in "NATURE", and if the quote was not taken out of context, well, I suggest that cancelling your subscription would be a good idea to avoid wasting more money. If you picked the quote just to win an indefensible line of argument and did not understand what you were posting, well... no comment..
but if you KNOW that wind force is proportional to square of wind speed and STILL posted that as your defense of the indefensible, well... no thanks, I don't want to be anywhere near your university or whatever, if it condones that.

The only other aspect that the paper points out near Equation 1 is that drag coefficients change with wind speed somewhat. This is a Reynolds Number effect, and as the paper points out, the increase occurs only over a small range of wind speed. This does not make force proportional to cube of wind speed.

The papers that I cited were just intros to different things in System ID. If you are already expert on the subject, then you should have no difficulty in seeing how to characterize Response at Points B, C and D, to Excitation at Point A, when the medium between these remains constant, so please don't waste bandwidth typing up critiques and treatises on different papers. Avoid what I call "techno-babble" and focus on what is important to the problem.

The interest in the papers readily Googled is about predicting damage to given buildings from earthquakes of given intensity. True. But they also show how "earthquakes" are simulated by pounding on the ground, as an example. I did not find any titled "System ID of the Ground Between Nearby Village And Nuclear Test Shaft" because those are probably not found easily on Google. What you CAN see from the papers is that using a (very) limited number of sensors, you can measure the response at given points, to a seismic-type excitation, and understand enough to make predictions of damage.

So I suggest that we get back to studying the issue of whether the response of the soil in Khetolai to explosions occurring near S1-S1 shafts, could have been well-characterized with reasonable, and reasonably obvious, methods and effort in the years between 1974 and 1998. I say, "sure!"

Look at it this way: If you hit the soil at some given depth at S1 with a (big) hammer exactly in the same place and same way, 5 times, days apart, why should the response felt at Khetolai, 1 foot below the surface at any given point, be any different between the 5? The relation should depend only on the characteristics of the soil between the two points. It doesn't matter if the signal went to Islamabad and back, it would do the same each time. And so you can consistently predict what will happen the next time you hit at S1, i.e., what will be felt at Khetolai. It's that simple.

Thanks!

[samuel]

Hold on to your horses! Power dissipation is used as an index of the damage due to hurricanes and it goes as the cube. The papers are to show:
a) Power dissipation goes as the cubed of wind speed.
b) Damages go as PDI. The losses have been measured and "calibrated" against wind speed. They *are* industry practice.
Are we okay so far? This is NOT about the force exerted but about losses incurred!!!
I gave you the paper citation to read. Here it is, if you could not download it: ftp://texmex.mit.edu/pub/emanuel/PAPERS/NATURE03906.pdf

[enqyoob]

I did try to end this... DO read the paper, Samuel, and DO read Equation 1, as you advised me to do. It says that FINANCIAL LOSSES due to a hurricane scale as its size, and its size increases with its power, and the power is indicated by the cube of its peak velocity (er... the velocity in a hurricane is not uniform throughout, it tends to be higher at the edge of its core..but I am sure you know that?)

IOW, a more powerful hurricane is also BIGGER in its area of influence, and that's why the financial losses become higher. If you look carefully at Equation 1, you will see that one of the integrals is over the AREA that it covers.

(Note from Hurricane Andrew which slammed Miami circa 1991: most of the really expensive damage occurs when roofs of homes blow off. This occurs at some threshold wind speed, where the lift is sufficient to lift the roof, or tear panels off the fixtures. Directly dependent on the dynamic pressure of the wind, meaning v^2. So what happened in Miami is that this threshold was exceeded for a large number of houses. The insurance companies tried arguing that speed only reached 125mph, and no roofs should have blown off if they were built well, but the weather reports said that it reached 165 mph, which is about twice the force.)

"PDI" in your paper is just a measure of "how much power was in this hurricane total, on a scale of (something)? You are going to suffer financial losses corresponding to this Index." This was for hurricanes that passed straight over a given area (a more powerful storm that missed all habitation and farms, would hardly cause any financial damage). The analogy would be that Khetolai needed a "PDI" for blasts located at S1, that said: "Damage will be proportional to the Power of explosions in S1". IOW, a Response Function between Response At Khetolai (equivalent of storm passing over a given region) and Incident Power at S1.

This is a far cry from claiming that the amount of damage that a structure suffers scales with the cube of wind speed, not the square of wind speed, so I am glad we settled that. Now, it turns out, it is the size of the affected area that scales with cube of wind speed. No problem with that - powerful earthquakes and nuclear blasts devastate larger regions than less powerful ones. This is why they kept the S1 power down - to avoid devastating a region so large that it included Khetolai and the Logistics Base.

So why is this even relevant to the Khetolai case? If I recall, you went down this line to claim that damage from a hurricane could not be predicted because (as I understood you to think) the fine-scale variations and luck etc, are too complex to predict.

Completely different problem. But it WAS more complex than calculating one "PDI" because the blast signature of a thermonuclear blast (meaning frequency content of the signal) is substantially different from that of, say, a conventional explosion or a 1KT Paki falling on the ground.

But for Khetolai, all they had to do was understand the signature of response at Khetolai (more complex than a single "PDI", understand what frequencies were transmitting too well vs. those that were getting damped out, (house collapses are related to resonant frequencies that are excited) and accordingly predict what would happen when a given type and magnitude of impact occurred. Would there be a huge shaking at certain frequencies, above the threshold at which damage occurred? If so, they had to back off in the explosive power.

In the actual event, they ended up causing permanent deformations in the soil, which is what caused the brick separation cracks.

[samuel]

Yes, damages are related to PDI -- and this is what is practice! The EXACT point is or rather now was -- just taking the hurricane intensity (oh its a 100Knot hurricane) and applying pdi does NOT tell you what the spatial distribution of damages is and certainly just taking some sparse measurements of damages and using the cube law does not give you a reasonable spatial distribution of the hurricane winds!

sources -- winds at lots of points. medium -- the distribution of structures. measurements -- losses. Unless you assume that the medium is UNIFORM (that is there are a million copies of the average house) or there is one speed, this is not an easy problem.

Let me relate that back to khetolai. There are two fundamental sources. The intensity of the source and the spatial distribution of the density (or whatever function) of the medium. The contention is they have lots of measurements and simultaneously a claim that you don't need lots of measurements and what goes along with that is, well we need to do just local analysis around khetolai or that the medium is so smooth that few measurements (and time is not really the issue here it is time invariant and all that) suffice.

But what was actually done? How many measurements could they feasibly take and did they really have a cluster of instruments around khetolai as is required for just reconstructing the density there -- but what use is that in estimating the intensity of the source if that is not in the same area? You can dismiss these questions if you like, but can you say more on what basis? Couldn't the broken bricks simply be a matter of spatial distribution of density, otherwise, should not the response be "higher" everywhere?



S

[samuel]

Let me add a note to that. Even if we go the frequency route, which is natural when dealing with waves, the medium is anisotropic it absorbs waves in certain directions in certain ways and in other directions in other ways. Will a regular bomb produce the same kind of waves as their detonation? Is that a reasonable claim to make? If not, ok, what did they need to do to quantify the medium's response at the desired intensity? That's the system ID problem in my mind. You don't want to estimate model parameters in one regime and then copy them over to the other if it doesn't let you estimate the system's response correctly in the desired regime, right? Then, if certain waves resonated with structures at khetolai, why does that mean the intensity was higher? It could just be resonance because that frequency passed through there?

[enqyoob]

Thanks, now we are on the same wavelength. I am saying that they had 24+ years to make impulse-response measurements at as many points as they wanted, all over the test range. They also had hundreds, maybe thousands or tens of thousands, of munitions tests at the Pokhran range, plus one nuclear explosion (1974) to use as impulse sources.

This has two implications.

1) To protect Khetolai, and thus to establish an upper bound on the power that they could generate at S1, they needed only VERY SPARSE measurements - at most a few sensors at/around Khetolai.

2) They had the chance to make far more measurements than this, and in fact characterize impulse-response relationships between the shafts, and essentially many different points all over the range - and do the equivalent of a tomographic brain-scan of the whole place - and to do this over the whole range of frequencies of interest. Once they had this, then they had a very good grasp about the yield of the blast. Very little uncertainty.

Actually tomographic scan is a good example for the spatial complexity. AFAIK, a typical tomography system only measures at 20 degree intervals (allooooo Dr. Shiv???) and then builds up the full volume reconstruction from that, although there may be features in most brains that are a lot smaller than the space between two adjacent 20-degree stations. ( I remember being spun around in steps, and there were not that many steps before they concluded that only 1 point is needed to characterize a total vacuum). I remember them wishing that they had the TWO-SENSOR system so that they could process the dummy through there twice as fast, and also reminescing about the time that they left some poor soul strapped into that coffin-like thing and locked up and went home for the day, because it took so long that they forgot she was there.

Since they cannot keep one sensor steady and allow the POK range to spin around, they probably had to locate quite a few sensors all round, and at different depths, to characterize the whole medium of the test range.

But if they did put in, say, 18 sensors at each depth, and used ten depths, that's still only 180 sensors. If you used redundancy on each, that's 360 sensors. Put in one per day, it takes only 1 year and they had 23 left. Compared to the cost of nuclear device tests (Rs. 1,700,000,000,000,000,000 per Comrade Bidwai), what's 360 accelerometers and cables and power supplies?

The third point is that if you combine (1) and (2) you see that the uncertainty about the relationship between Khetolai shaking level and S1 yield, was very very small. IF you had access to the Classified data about the range instrumentation results.

[samuel]

Some basic info

http://www.ehealthmd.com/library/ctscan/CTS_work.html

As the scanner rotates, the detector takes numerous snapshots called "profiles." Typically, about 1,000 profiles are taken in one rotation. Each profile is analyzed by computer, and the full set of profiles from each rotation is compiled into to form the slice-a two-dimensional image.

A 1000 profiles per rotation?

The above information may not be current but I will reiterate. Let us take a simple spatially varying medium in 1D a source and a sparse measurement and quantify a) the reconstruction fidelity and b) the intensity of the source.
I will bet that
a) a sparse problem subject to normal regularization will produce bogus estimates of the reconstruction and intensity.

b) for nonlinear media, if we quantify the system in one band, it does not tell us much about others (necessarily).

c) The fundamental ambiguity between whether there were local or propagation effect or intensity remains.

S

[shiv]

CT scan machines are OT and out of my own area of work (though not the scans themselves). Need to ask - but resolution has certainly improved over the years. Still - it gets difficult to spot 5 to 10 mm lesions in some parts of the body where 5 to 10 mm is huge.

OT OT

[Arun_S]

I agree on the part that the yield verification by seismic data should be done by an unrelated organization.

If we do a 150kt phataka, the following would happen:

1. NPA will claim it is only 30KT, hence fizzle.
2. Hawks will claim it is FBF, too heavy to put on the tip of the missile etc, hence no deterrent
3. SuperHawks will demanding 1MT bum onlee.

And the political leadership will still be patting their own musharraff, seeing if the backbone is being grown yet.

Point 1 is not correct.

Given what we have heard from "Don't test, India doesn't need H bomb, lobby" on BARC competence in characterizing the Pokhran range with the stimulus response characterization prior to 1998 to determine exactly how much TN fizzle is need to protect the Khetolai village from any damages more than just simple cracks; BARC will next time do the 150 kT at the optimal burial depth (given they are super competent in this SHOCK 3D etc etc, and the POK-II experience under its belt), the crater from 150kT will be of a size that no 30kT test out of thousands done by N5, match. It will only match those test that produced 150kT or more. There is no two way about in competently showing a test that yields 150 kT. It cant be masked as 30 kT period !! (unless of course SHOCK 3D decides to give its progenitor a shock and they over-bury the 150 kT into a 900 m deep shaft, or they continue to walk beaten tradition of Pok-II TN fizzle )

[enqyoob]

Even if we go the frequency route, which is natural when dealing with waves, the medium is anisotropic it absorbs waves in certain directions in certain ways and in other directions in other ways. Will a regular bomb produce the same kind of waves as their detonation? Is that a reasonable claim to make? If not, ok, what did they need to do to quantify the medium's response at the desired intensity? That's the system ID problem in my mind. You don't want to estimate model parameters in one regime and then copy them over to the other. Then, if certain waves resonated with structures at khetolai, why does that mean the intensity was higher? It could just be resonance because that frequency passed through there?

Several good points:

1. Anisotropy of the medium is not a problem for the straight "S1-Khetolai response". It's like saying, the train from Dilli to Kolkotta travels at 180 kmph in UP, 60kmph in Bihar, and 30kmph in W.Bengal, but as long as the track conditions are the same at any given point every day, the total time taken is the same every trip. And if the thieves are the same, they are going to steal the same weight of mail etc every day. So anisotropy poses no error in estimating what yield would be safe for Khetolai.

2. Anisotropy of the material in the overall range means that simply putting a sensor at a given distance is NOT going to give a spy a good estimate of yield by "guessing" the medium properties. So only the guys who have access to the data from the 180 sensors or whatever and done the reconstruction can really say how the waves would have propagated. The others are bissing in the hawa, so to speak, with their "estimates". They are HOPING that there is no anisotropy.

But turning that around, the guys who did the 180-sensor scan WOULD know how to take the anisotropy into account. Like the little spot that shows up in the brain scan with the 20-degree-step tomography, the rocks etc would have showed up in the reconstruction. So then they would have built a modeling/ simulation taking all these into account. I cannot say if this was indeed done, but consider that they went to the trouble of simulating the Bene*** disaster, modeling the strata and faults in detail, to achieve a successful simulation. Why would they not put the same effort into modeling their home range? The whole point of that 2006 BARC paper was to say, "hey, guys, we have the tools, and obviously we used them successfully on our stuff long ago".

Note that the 2006 BARC paper was submitted in 2002 - a good measure of how desi research is obstructed and delayed by having to seek "acceptance" in the gora journals. Or it was a really poorly-run journal, like "Acta Astronautica" in the Space business. Any journal that takes longer to review and publish than it takes the average baby to be made, is a lousy journal. Taking the delays in Official Approval, getting drawings made, proof-reading, etc, etc, this means the work was finished in 2001 at the latest. It takes a while to do stuff like this, so they must have started circa 1999 or 1998. After they had finished their analysis of their own tests, and got permission to reveal enough about the technology used to those who needed to be deterred.

3. If not, ok, what did they need to do to quantify the medium's response at the desired intensity? That's the system ID problem in my mind.

If the medium's response was "linear" then the coefficients in the response function are independent of intensity. Was S1 in the linear range of response? I don't know enough about the Richter scale to decide this, but per Sanku here, magnitude 5.8/5.9 is still fairly small/ moderate shaking, so it may be fair to treat magnitude 5.2 as also small and hence in the linear regime.

Then, if certain waves resonated with structures at khetolai, why does that mean the intensity was higher? It could just be resonance because that frequency passed through there?

That's not quite the way to look at it. All the frequencies pass through everywhere, but the intensity at the different frequencies may be different - this is the Transfer Function magnitude, a function of frequency. So at certain places in Khetolai, perhaps the intensity at certain frequencies was higher than, say, at the Logistics Base. Or lower.

As the 'quake intensity increases in the "moderate" range, this function's form would not change much, but the amount of energy at each frequency would increase. So the resonant frequencies of certain structures may have been given enough energy to really shake down.

If the intensity went beyond the "linear" regime, then there would have been distortions of the whole wave form, meaning a redistribution of the energy between frequencies. Thus in this regime, it is possible for certain frequencies to have much more intensity, than what was predicted from the linear-regime tests. Hence the concern about the school falling down, forcing the Army to ask the school headmaster to see that the kids were outside during the relevant time.

Now we come to the real problem: There was really no way to simulate a "quake" as strong as what a TN device would generate using conventional explosives - this would go outside the "linear" regime, and so predictions of what would happen, would become a bit "iffy". The simulation nerds could provide some answers, because they could take soil samples and subject them in the lab to the required impact level, then put those values instead of the existing values from the "linear" model in their simulation and see what happens.

The Bene*** data set was again a way of going into this regime and predicting development of cracks, venting etc., all highly "nonlinear" events. So they did have some capabilities even in this regime. This all may have got into some uncertainty, hence the damage at Khetolai. But my point again, is that this could only have happened at the UPPER bound of their predicted yield, not at a low level.

Given that all the desi netas cited the concern about damage to the village at the time, it is pretty clear that they were real worried about that aspect - must have been a certain amount of uncertainty.

Thanks for the discussion! I learned a lot!

[Austin]

What percentage of burn is needed from 3rd stage ( LiD ) to jump from fizzle to sizzle ?
PKI mentioned the current S1 burnt ~ 10 % , hence a fizzle.

[Arun_S]

What percentage of burn is needed from 3rd stage ( LiD ) to jump from fizzle to sizzle ?
PKI mentioned the current S1 burnt ~ 10 % , hence a fizzle.

Question not clear. LiD is second stage not third stage. In TN lingo Natural Uranium or fissile tamper is called as the 3'rd stage .

If it was a typo and you mean second stage ( and back in 1999 PKI used the words "perhaps upto 10%" Fusion LiD burn), the answer is at least 50-60 %. Although TN weapons are generally designed for 80% or higher burn).

EDITED.

[enqyoob]

OK, these 1000-profiles per rotation CATs are clearly much better resolved. However, even 4 cameras can be used to do tomographic reconstruction using scattered light from the volume of interest, in certain problems where the field has very high spatial complexity, but is also very unsteady so there is no way to hold it steady and spin it around etc. Check "tomographic velocimetry" for instance. Anyway, this is OT.

The item here is not the right way to look at it:

b) for nonlinear media, if we quantify the system in one band, it does not tell us much about others (necessarily).

Decomposing a signal into the frequency domain means that you are simultaneously capturing the response at all frequencies. So the CAT equivalent would be that you are getting the transmission at all frequencies, and hence have the equivalent of 1024 (or 2048 or 4096) images, one at each frequency, for each image.

The spectrum of an impulse has an infinite band of frequencies.

Actually, a "typical earthquake spectrum" is shown at this site. along with some info on shakers to test equipment.

[shiv]

It cant be masked as 30 kT period !!

Yes but Seismologists may be tempted to admit that there are error margins is all seismological estimates. If they say 50 to 150 kt we will have to test 1 megaton so that they can say 200 kt to 1 megaton.

This is the problem about testing to send unmistakable, unmaskable signals to others for an answer to "Woh log kya kahenge?". Even the Pokhran fizzle suffers from exactly this drawback.

The need to test a 150 kiloton plus bomb is based on the need to show someone that we have a proven capability to make Deuterium fuse and that we have an almost limitless capability of making destructive bombs. If there is any other signal that this sends out I would like to know about it.

The next step after demonstrating Indian capability to cause fusion on demand is to demonstrate capability of delivering it.

After that it requires the demonstration of willingness to use that power.

And finally one needs to demonstrate that people who act funny with us by infiltrating men into our territory or sending in terrorists will be punished by our demonstrable fusion power, delivery capability and steely resolve leading to a permanent (long lasting) end to transgressions on India.

That in my view would be a holistic picture of India's nuclear forces.

There are two problems with this picture

1) India does not seem to be heading in the direction of doing all of the above
2) I see very little indication that India's problems will usefully addressed by taking the above route.

I would love to hear (from anyone) an exposition on the role of 150 kiloton plus nuclear weapons in enhancing India's overall security scenario - where "security" is Military security, economic security, food security and energy security.

I am serious. I am not saying that I would not be proud of seeing a demonstration of 150 or even 1500 kiloton weapon capability. But India has people who think in different ways and every action is based on varying world views.

[shiv]

What percentage of burn is needed from 3rd stage ( LiD ) to jump from fizzle to sizzle ?
PKI mentioned the current S1 burnt ~ 10 % , hence a fizzle.

If PKI said 10% and said 400 grams of LiD - then he implied that 4 kg of LiD must have been used.

In R Chidambaram's talk he said:

Chidambaram said that if 1 Kg of Pu is fully used up in a nuke - the yield will be 16KT (less if less is used up, of course). If 1 Kg of Deuterium is used up fully - you get a yield of 80kT, and if 1 kg of Lithium Deuteride is fully used up in a fusion bomb the yield will be 50 kT.

That would imply that PKI expected the bomb test to be at least a 200 kiloton test expecting 100 percent efficiency in LiD burn. Not counting fission trigger and spark plug. And perhaps no tamper at all? Because IIRC tamper yields a huge percentage of the total yield. Or tamper of PUF?

That needs to be correlated with a burial depth of "Over 200 meters" (Chengappa) and "300 meters" (BRF quoting Jaswant's interlocutor or someone)

But we have no proof that the bomb was not tested at a depth of 900 meters or some similar deep depth.

[samuel]

OK, let's instead of writing Y = A X, write using the transfer function X(w) H(w) = Y(w). They are really not different. Let's take a few statements:

1. All waves pass through -- no zeros -- well it's invertible then? If that's true then why is it not trivially reconstructible for any source destination pair? just poor snr?

2. All modes are produced by all bombs and X(w) is white. This makes no sense but I don't know what X(w) looks like. And if X(w) is actually white then the estimation problem is for 1 scalar value and that's trivial. May be that is why others are confident of their measurements, too? If X(w) is not white and the modes of the nukes are different than conventional "practice", what do they do?

3. All frequencies are detectible -- this is not true.

4. There are no multipath effects due to density and local effects and these effects are stationary irrespective of input intensity (by wavenumber). This makes no sense either.

5. The transfer function is linear (even if anisotropic) -- decidedly not true.

So why is the problem any easier?

I'll pull some idealized soil response functions from literature to see.

[Austin]

That would imply that PKI expected the bomb test to be at least a 200 kiloton test expecting 100 percent efficiency in LiD burn.

PKI mentioned even the most efficient design burns LiD with 50 % efficiency , that would probably imply the latest light weight , high yeald warhead from US,Russian stable.

But he thinks it burnt only ~ 10 % , so my question is what is the % of burn needed to call a TN bum a success , certainly one can increase the quantity of LiD and can burn at 10 % and give a desired yeald , but that will not mean TN is successful.

My little understanding on this topic is , I can get a yeald of just ~ 40 kT lets say its lower value for a given weapon design than can scale up to say 300 kT and if my LiD burns say with an effeciency of atleast 25 % then my TN design is an unqualified success.

But for a given design I can get a yeald of 150 kT but the TN will be a fizzle if my LiD burns at at efficiency of ~ 10 % and it just turns to be FBF and not a true TN.

[John Snow]

n guru ripped apart the ADA and Kaveri for various annamolies, I wonder if BARC folks are made incharge of Kaveri and LCA we would be having 100s of squadrons of Tejas blazing away.....

The only thing that come close to BARC is the manufacture of T-35 tanks by SU rolling them out staight to frontline... at one point the assembly crew themselves drove to the front which was few yards away if not few miles.

May be next time Kargil like thing happens in Arunachal , BARC wallahs will directly take the PU balls U balls D balls and finally deliver the Great Balls of fire, into the mouth of dragon no?

[ramana]

T-34 tanks.

That talk cancellation is not a good sign.


Real dhurgatna.

[shiv]

That would imply that PKI expected the bomb test to be at least a 200 kiloton test expecting 100 percent efficiency in LiD burn.

PKI mentioned even the most efficient design burns LiD with 50 % efficiency , that would probably imply the latest light weight , high yeald warhead from US,Russian stable.

But he thinks it burnt only ~ 10 % , so my question is what is the % of burn needed to call a TN bum a success , certainly one can increase the quantity of LiD and can burn at 10 % and give a desired yeald , but that will not mean TN is successful.

My little understanding on this topic is , I can get a yeald of just ~ 40 kT lets say its lower value for a given weapon design than can scale up to say 300 kT and if my LiD burns say with an effeciency of atleast 25 % then my TN design is an unqualified success.

But for a given design I can get a yeild of 150 kT but the TN will be a fizzle if my LiD burns at at efficiency of ~ 10 % and it just turns to be FBF and not a true TN.

PKI said fission trigger 20 kt and fusion 20 kt (He did not swear - he guessed - so I am no trying to lambast him) But if 20 kt is 10% efficiency and 400 gram burn, and suppose he was implying 25 % burn then 1 kg should have burned (out of 4 kg) - so about 70 kt test. (20 fission plus 50 fusion) This fits in well with Santhanams 60%.

But it also means that both accept 40 kt - making Sikka and RC telling the truth and NPAyatollahs and their CTBTO wrong.

This is one explanation which is consistent with what RC, Sikka, Santhanam and PKI are claiming about yield.

The controversy may be how much efficiency was expected of fusion burn - assuming that Santhanam and PKI have access to actual amount of LiD used and actual size of fission trigger and radiochem data.

[ramana]

We will find out when Sikka delivers his talk.

[shiv]

^^
If he delivers

[SagarP]

Irrelevant post deleted.
Stick to the point or quit posting.
Rahul.

[krishnan]

Totally off topic and unwanted post IMHO

[Mahendra]

^x2

Perfect bakra for the adminullahs iftar

[enqyoob]

n guru ripped apart the ADA and Kaveri for various annamolies, I wonder if BARC folks are made incharge of Kaveri and LCA we would be having 100s of squadrons of Tejas blazing away.....

Snowgaru: Thx 4 pointing that out. Only thing it shows is that I decide what to post based on what I c. Thinking of course would be optional - as some1 pointed out, it is impossible to draw conclusions if one introduces anisotropy in the medium of the thread...

If GTRE had said: "v r not yoojing supercruise yet because v are afraid ellceeya will accidentally break the sound barrier above Airport Road and overtake the traffic there" I would have been 400% believing them onlee.

BTW, I believe the SU did win the war with Marshals Zhukov's and Koniev's T-34-equipped tank divisions, hain? T-34 was "specifically dejigned phor Russian condishuns" like Kaveri is for Indian conditions. If the tread came off 5 minutes into the battle, there was no possibility of retreat.

I believe that even Pakistani nuclear weapons have been put into the military storage without any testing. All weapons, zero mileage. In fact I have been advising them to test each of their weapons fully by firing them at Sargodha, Peshawar etc, but will they listen?

[samuel]

Of course anisotropy is just one problem and in fact it is misinformation to suggest that anisotropy is what causes the confusion. the confusion arises when both source and medium are not known and anisotropy increases the difficulty of estimating the medium. There are other problems, like arrogance which seem to show up irrespective of thread though, quoting bogus literature and not reading what's provided. That may be more widespread than I thought.

[samuel]

To verify the claim that identifying the medium -- that is produce a seismic model of its response-- is an easy exercise, I will sample contemporary practice drawn from seismic reconstructions. There are many dimensions to this problem, I contend, that are to do with a) nonlinearity of the response function, b) dimensionality of the estimation problem, i.e. how many measurements and in what part of the spectrum and c) uncertainty of measurement, detection and knowledge of medium.

If this is indeed as easy as it is made to appear then that would be wonderful:
Johnson, P. A., P. Bodin, J. Gomberg, F. Pearce, Z. Lawrence, and F.-Y. Menq (2009), Inducing in situ, nonlinear soil response applying an active source, J. Geophys. Res., 114, B05304, doi:10.1029/2008JB005832.
Note: Heterodyning itself is an old idea, but this is a
very contemporary paper and can be used as a benchmark for what may be in 1998 or earlier.
Here is first paper:

It is well known that soil sites have a profound effect on ground motion during large earthquakes. The complex structure of soil deposits and the highly nonlinear constitutive behavior of soils largely control nonlinear site response at soil sites. Measurements of nonlinear soil response under natural conditions are critical to advancing our understanding of soil behavior during earthquakes. Many factors limit the use of earthquake observations to estimate nonlinear site response such that quantitative characterization of nonlinear behavior relies almost exclusively on laboratory experiments and modeling of wave propagation. Here we introduce a new method for in situ characterization of the nonlinear behavior of a natural soil formation using measurements obtained immediately adjacent to a large vibrator source. To our knowledge, we are the first group to propose and test such an approach. Employing a large, surface vibrator as a source, we measure the nonlinear behavior of the soil by incrementally increasing the source amplitude over a range of frequencies and monitoring changes in the output spectra. We apply a homodyne algorithm for measuring spectral amplitudes, which provides robust signal-to-noise ratios at the frequencies of interest. Spectral ratios are computed between the receivers and the source as well as receiver pairs located in an array adjacent to the source, providing the means to separate source and near-source nonlinearity from pervasive nonlinearity in the soil column. We find clear evidence of nonlinearity in significant decreases in the frequency of peak spectral ratios, corresponding to material softening with amplitude, observed across the array as the source amplitude is increased. The observed peak shifts are consistent with laboratory measurements of soil nonlinearity. Our results provide constraints for future numerical modeling studies of strong ground motion during earthquakes.

[enqyoob]

Thanks for finding that! Great stuff. This is why BeeAref is the One&Only for finding the most relevant and most buried knawlidj on any subject under the sun, and in this case, some inside the Sun.

So you see that pushing on the ground in a sinusoidal fashion (instead of dropping Pakis which would be an impulse loading, albeit nothing like the sharpness of a TN shock) and measuring the response is indeed a valid, state-of -the-art technique that ppl have been using. From a 1-minute BRF-ishtyle in-depth examination (BRFIIDE) of the abstract you have kindly found and posted, it appears that the research is on identifying the nonlinearities. Therefore, the researchers carefully used one frequency at a time, and showed that the response (a) depends on the amplitude when the amplitude is beyond some threshold and (b) that the dependence on amplitude varies from one frequency to another.

If they were not focusing on this frequency-by-frequency separation of effects, the "dhamka" technique which produces multiple frequencies would have been the way to go.

Now in most of engineering, when one encounters nonlinear (but still continuous) phenomena, what one immediately does is linear-ize - that is, take the easy excuse. Specifically, one promises to only apply results to one small range of amplitudes (however large the mean of that might be) and coolly use the well-known linear methods. So in the POK case, this means that the linear "transfer function" for Dusshera firecrackers would be different from that for 1 dynamite stick, 2 dynamite sticks, a 200lb bum, a 500lb bum, a 1000lb bum, 2000 lb bum, Prithvi payload, and the POK 1974 results. Beyond that they had to be flying blind, which is why they went to the published US test effects at Bene**** to see if they could capture massively nonlinear effects, even including the presence of serious cracks/vents in the ground (discontinuities). They achieved "pretty good" success as demonstrated in the 2006 paper.

To do the nonlinear simulation, they would also have used laboratory soil tests as the authors here say, and then use the simulations to see what it would have been for various levels of POK-2 S1 "Fizzles", "Fyoozzles" and "Dhamkas".

Obviously, ONCE ONE GETS INTO the nonlinear regime, and BEYOND the POK-1 data which is the biggest they had b4 1998, the uncertainty mounts.

Which explains why they went too far and caused significant damage at Khetolai - the nonlinear simulation had some uncertainties, and the effects were MORE than they predicted.

Makes perfect sense, thx. Note, BTW, that the POK-2 planners had a very specific task, and were not developing "general techniques and theory" to gain fame. They were tasked precisely to find out what level of TN dhamka at S1 would leave Khetolai and the Logistics Base unscathed. So then one does not sit around wringing one's hands about the impossibility of general prediction of nonlinear systems, and the non-uniqueness of solutions. One puts the firecrackers in the hole, or in nearby holes, and proceeds. And one uses all the yield and response sensor data from all the Test Range munitions tests. Who knows, maybe they even dropped some megaton Pakis on the ground near Khetolai just to create nonlinear effects.

[enqyoob]

Just to add to that, samuel: note the phrase

material softening with amplitude, observed across the array as the source amplitude is increased.

This is one of the terrible realities of massive earthquakes. In the San Francisco quake (1989?) a whole area suffered what is known as "soil liquefaction" like in a "fluidized bed" experiment. That area suffered essentially 100% destruction because the building foundations were suddenly in what felt like a liquid medium, with large waves running through. Finis. Other regions close by were virtually unscathed in comparison, being built to tough CA quake standards.

But such things happen at enormous quake magnitudes. As Sanku suggested, 5.2-5.4 is probably not nonlinear enough to cause significant error in linear predictions, though they MIGHT hit resonant frequencies of some large/long halls like schools.

One other point about the sinusoidal excitation technique in that paper: They can reach large amplitudes because ALL their power is going into just one frequency, unlike the "dhamka" where the power is distributed over a broad spectrum. Huge difference! If u have the time and patience to wade through that paper, you might find that they are talking about 6.5+ quakes. Just a guess. I am too lazy 2 read, since ignorance is so "bissphool".

[John Snow]

We Indians are clever by half and one fourth chanikyan onlee. (this includes me too )

We should not have given numbers after the POK II series. aka maha scientists like RC AK APJ KS et al
Just say all tests went according to plan we will conduct further analysis which is contnious process.

But then we got carried away in the dust of Mushroom topping Pizza ( A cylinder supporting, a speherical nearly cloud)

like this http://www.wetasschronicles.com/NukeTest.jpg. Hence the use of cylindrical coordintes )

But we did declare the numbers, then we stated fitting the curve which always is difficult as curves change as they gain weight. no.

Now If people say we are nuclear power and are completely satisfied with results whats the big deal in declaring clearly what the yield was. It increases credibility and deterrent in one shot (test).

If small bums is wave of the future then why worry admit if H bum failed and increase our credibility to say we do not threaten big boys like Uncle aunty frenchi and Ruskie, only Paki and Chinee.

If the current deterret (assumed by opponents) works then to say H bum failed does not create any hole in the deterreny no?

If H bum failed and we acknowledge then the enemies also understand that we will improve designs and make it work and all the more reason to keep our options open.

But then Indians be it Kashmir or terror or Nukes we tie ourselves in pretzel knots even before uncle or aunty glares at us.

Jai bolo Hind ki.

[NRao]

We should not have given numbers after the POK II series. aka maha scientists like RC AK APJ KS et al
Just say all tests went according to plan we will conduct further analysis which is contnious process.

Precisely what I said.

Whatever the yield was .................................. was planned yield. We have the prefect instrument of deterrence. And, we can dial. Watch - test in two days - and, again, perfect yield. We dialed and got the perfect yield.

Let them compute, recompute, 10000 CPU hours, whatever. What do we care about computations, sigmaX values, confidence levels, etc?

All we care about is our enemies are 100% confidence or close to it that India HAS a deterrence.