Current Nuclear Weapons Development

[Sferrin]

I understand your point, my point is that everyone seems reasonably satisfied about upholding the testing moratorium, at least insofar as supercritical yield producing tests go.

And they're currently building NEW nuclear weapons. We aren't and haven't for almost half a century. Maybe they have more confidence because they still have an industrial base. Also, why should we care if they're confident or not? That means exactly squat when it comes to our forces.

 

[whatismoo]

And they're currently building NEW nuclear weapons. We aren't and haven't for almost half a century. Maybe they have more confidence because they still have an industrial base. Also, why should we care if they're confident or not? That means exactly squat when it comes to our forces.

I really don't see how this is a relevant point. An untested design is an untested design, whether you've been serially producing other designs or not. You'd have a point if we were talking about assessing the reliability of aging warheads, you can avoid that by making new warheads, but that's not the point in question.

We should care if they're confident because they have a significantly less advanced non-detonation testing infrastructure and yet are confident fielding novel designs. We have significantly more advanced infrastructure to model, simulate and test below the threshold of all up nuclear detonations (via the >dozen high end supercomputers DOE has, NIF, Z Machine, LANSCE, DARHT, JASPER, Rochester LLE, BEEF, etc). If other nuclear weapons states are willing to rely on less thoroughly understood and tested novel designs, it is a point in favor of testing not being essential.

Of course, some people won't be satisfied until they can watch the double flash from the rooftop bars on the strip but I think that's a bit silly. It seems to me that the advantage the US has in understanding of NW physics stemming from a non-test environment is large enough that it would only benefit Russia and China if the US were to break the testing moratorium. If the Russians and/or Chinese break it, sure then start digging holes in Frenchman Flat, but why spoil a game rigged in our favor?

 

[Sferrin]

I really don't see how this is a relevant point. An untested design is an untested design, whether you've been serially producing other designs or not. You'd have a point if we were talking about assessing the reliability of aging warheads, you can avoid that by making new warheads, but that's not the point in question.

We should care if they're confident because they have a significantly less advanced non-detonation testing infrastructure and yet are confident fielding novel designs. We have significantly more advanced infrastructure to model, simulate and test below the threshold of all up nuclear detonations (via the >dozen high end supercomputers DOE has, NIF, Z Machine, LANSCE, DARHT, JASPER, Rochester LLE, BEEF, etc). If other nuclear weapons states are willing to rely on less thoroughly understood and tested novel designs, it is a point in favor of testing not being essential.

Of course, some people won't be satisfied until they can watch the double flash from the rooftop bars on the strip but I think that's a bit silly. It seems to me that the advantage the US has in understanding of NW physics stemming from a non-test environment is large enough that it would only benefit Russia and China if the US were to break the testing moratorium. If the Russians and/or Chinese break it, sure then start digging holes in Frenchman Flat, but why spoil a game rigged in our favor?

Agree to disagree.

 

[NMaude]

The reason why the US would need to conduct new underground nuke tests is if the were designing and testing a new small, high-yield design since by pushing the envelope that hard means that even small discrepancies in the design can cause a dramatic reduction in yield or even a fizzle. I remembering years ago that the W-76's radiation-case (Made of U-238) was as thin as the aluminium wall of a coke-can.
Now on the other hand say the USAF issue a requirement for a new TN warhead of, say, one-megaton, 18-20" in diameter, 48" long and a weight of 1,500-2,000lb* (In other words a conservative design typical of the late 1950s-early 1960s) it could be designed and built without any need for a live nuclear-test.

*Basically a 21st century W-28 or W-49 other old warhead designs that could be built would the likes of the W-47, W-56 and W-59 in a 21st century update.​

 

[Scott Kenny]

I am very uncomfortable deploying a new warhead design without an all-up test. Or at least a full-yield test, we don't have to live launch I don't think. For example, we haven't live-launched an SLBM since Polaris A-1. We have test launched hundreds of them short of live warheads.

But I'd be perfectly fine with someone ordering production of an old B53 or B41 high-yield design for planetary defense without testing it.

 

[Forest Green]

The reason why the US would need to conduct new underground nuke tests is if the were designing and testing a new small, high-yield design since by pushing the envelope that hard means that even small discrepancies in the design can cause a dramatic reduction in yield or even a fizzle. I remembering years ago that the W-76's radiation-case (Made of U-238) was as thin as the aluminium wall of a coke-can.
Now on the other hand say the USAF issue a requirement for a new TN warhead of, say, one-megaton, 18-20" in diameter, 48" long and a weight of 1,500-2,000lb* (In other words a conservative design typical of the late 1950s-early 1960s) it could be designed and built without any need for a live nuclear-test.

*Basically a 21st century W-28 or W-49 other old warhead designs that could be built would the likes of the W-47, W-56 and W-59 in a 21st century update.​

And the W93/Mk7 is really pushing the envelope, 250-350kT at W76 mass (~95kg) by most estimates.

 

[Josh_TN]

And the W93/Mk7 is really pushing the envelope, 250-350kT at W76 mass (~95kg) by most estimates.

Had not heard that. I am surprised by that; last time I had heard anything about new U.S. warhead design a coupe decades ago I had thought the emphasis was on reliability and low maintenance, not further miniaturization.

 

[Forest Green]

Had not heard that. I am surprised by that; last time I had heard anything about new U.S. warhead design a coupe decades ago I had thought the emphasis was on reliability and low maintenance, not further miniaturization.

Must emphasise, all estimates at this stage, they're supposed to have tested a HGV on a D5LE possibly too.

 

[zen]

Yes the Reliable Replacement Weapon.

It was actually suspected that the UK was doing something different (Krakaroa etc) and possibly at US military urging. To keep it out of Congressional oversight

 

[Forest Green]

Yes the Reliable Replacement Weapon.

It was actually suspected that the UK was doing something different (Krakaroa etc) and possibly at US military urging. To keep it out of Congressional oversight

The UK warhead will be called the Astraea A21 and share the Mk7 RB.

Astraea (nuclear warhead) - Wikipedia

en.wikipedia.org

Not sure if posted before. UK air-launched nuclear weapon/missile.

UK considers reintroducing air-launched tactical nuclear weapons

An air-launched tactical nuclear weapon capability would not address short-term security concerns of the United Kingdom and Europe.

thebulletin.org

 

[NMaude]

And the W93/Mk7 is really pushing the envelope, 250-350kT at W76 mass (~95kg) by most estimates.

There isn't much information available about the W93 but given that it's supposed to replace the W88 it will no doubt have the same or very similar dimensions, shape (A lopsided peanut design) and similar weights.

In order to get these small but high-yield TN warheads their secondaries clearly use a LOT of Oralloy* (And Tuballoy** for the W87) meaning when detonated the resulting explosive yield would have a high fission-fraction, in other words very dirty warheads in terms of nuclear-fallout.

*Weapons-grade (93.5% U-235) uranium.
**Depleted uranium.

 

[bobbymike]

OK :-(

Not my opinion but an example

“I believe Trump calling for new testing of nuclear weapons is short sight and unnecessarily provocative given the current state of the world.

Besides that I feel given the number of tests the US performed during the Cold War, that testing is not required even to build new nuclear warheads.”
——————
The above is valid and will not illicit any concern from our members guaranteed.

I am far from perfect and over my 16+ years on this site have made questionable comments.

I think everyone here is trying to be as professional as possible insuring the continued comity [mostly] on the greatest aviation technology site on the internet.

 

[Forest Green]

There isn't much information available about the W93 but given that it's supposed to replace the W88 it will no doubt have the same or very similar dimensions, shape (A lopsided peanut design) and similar weights.

In order to get these small but high-yield TN warheads their secondaries clearly use a LOT of Oralloy* (And Tuballoy** for the W87) meaning when detonated the resulting explosive yield would have a high fission-fraction, in other words very dirty warheads in terms of nuclear-fallout.

*Weapons-grade (93.5% U-235) uranium.
**Depleted uranium.

It uses a W76 secondary with a new primary and is estimated to have 250-350kT.

 

[Scott Kenny]

It uses a W76 secondary with a new primary and is estimated to have 250-350kT.

That doesn't make sense.

It's the secondary (and tertiary if present) that provide the boom. A new primary should not provide any significant boost in yield, and using a 100kt secondary with a new primary should still give you about a 100kt yield.

Unless of course the Mk5 RB is acting as the tertiary because it's made up of significant amounts of DU. That I don't know off the top of my head and probably wouldn't be able to say if I did remember. If that is the case, it might give you something in the vicinity of 350kt, but I'd expect it to be closer to 475.

 

[zen]

UK Astrea may have tight constraints to fit US supplied shell and there will be weight/balance restrictions too.
But UK warheads be made in the UK with UK equipment to UK standards.
So there is inevitably some divergence and differences.
That said at the time the UK was funding sub-critical tests there were all sorts of doubts and political restrictions raised on the US military side. RRW looked at risk.

Since some of these tests were done in the US, there would have been a lot of sharing of results under the US-UK nuclear agreements. Convenient for the US military, since a lot of this is protected from political shenanigans.

 

[Forest Green]

That doesn't make sense.

It's the secondary (and tertiary if present) that provide the boom. A new primary should not provide any significant boost in yield, and using a 100kt secondary with a new primary should still give you about a 100kt yield.

Unless of course the Mk5 RB is acting as the tertiary because it's made up of significant amounts of DU. That I don't know off the top of my head and probably wouldn't be able to say if I did remember. If that is the case, it might give you something in the vicinity of 350kt, but I'd expect it to be closer to 475.

From memory designs like the W87 for instance actually had two rings of U235 and because they were short of U235 due to simultaneous W88 production they used non-enriched Uranium in the secondary, which led to 300kT instead of 475kT.

 

[Sferrin]

From memory designs like the W87 for instance actually had two rings of U235 and because they were short of U235 due to simultaneous W88 production they used non-enriched Uranium in the secondary, which led to 300kT instead of 475kT.

"Nominal W-87 yield is 300 kilotons; yield can be increased to 475 kilotons by adding a sleeve of enriched uranium-235 around the secondary: the oralloy adds considerably to total yield when it fissions after the secondary is compressed and ignited by its "sparkplug". The 300 kiloton W-87 uses much less fissionable and fusionable materials than the W-78 with its 335 kiloton yield."

 

[whatismoo]

It uses a W76 secondary with a new primary and is estimated to have 250-350kT.

Wouldn't this be like six different kinds of illegal to divulge if it had any bearing in reality?

Also isn't W93 still in phase 2A, which is requirements definition?

 

[TwoTribes]

At the latest meeting urging all countries to bring the Comprehensive Nuclear-Test-Ban Treaty into effect, the United States cast the sole opposing vote. The CTBTO is not a important international organization... but... it is still a signal?

 

[Forest Green]

Wouldn't this be like six different kinds of illegal to divulge if it had any bearing in reality?

Also isn't W93 still in phase 2A, which is requirements definition?

Sure but I think something very general has probably been released saying a yield between the current W76 and W88, the rest is likely educated guesswork.

 

[bobbymike]

Sure but I think something very general has probably been released saying a yield between the current W76 and W88, the rest is likely educated guesswork.

I thought I read somewhere that the W93 is based on a proven [I don’t recall the word “tested”] design developed in the early 2000s that pointed to work completed on the RRW during that period?

 

[NMaude]

I thought I read somewhere that the W93 is based on a proven [I don’t recall the word “tested”] design developed in the early 2000s that pointed to work completed on the RRW during that period?

The RRW was basically a modified W89 which although never put into production had been tested several times in the last few years of the Cold War prior to the end of underground nuclear-testing in Nevada in 1992. The W89's primary used recycled pits from the primaries of dismantled W68s.​

 

[Desertfox]

Agree to disagree.

Disagree all you want, as someone who has seen behind the curtain, you are dead wrong.

 

[Scott Kenny]

Wouldn't this be like six different kinds of illegal to divulge if it had any bearing in reality?

Not if the description has been published somewhere.

 

[JTR]

how many warheads per missile in the future D5 ?
16 of these new variat D5 missile may be the same than 24 legacy D5, no?

The D5, D5LE, and D5LE2 all have identical capabilities – up to a maximum loadout of 8x Mk5/W88, or 12x Mk4/W76.

As I understand it from public discussion, 1-8x W76s, 4x W88s.

There's going to be a couple of birds per sub with the tiny yield W76Mod2 (5-7kt, sub-Hiroshima) warheads for escalate-to-descalate, and those are almost certainly 1 warhead per missile with every single penaid the US has developed installed for balance if nothing else. If you can stuff a W76Mod2 physics package into some fancy maneuverable RV, it'll probably have that, too, less for accuracy and more for making interception harder. The mission calls for only launching one of those, the second missile is a spare carried in case there's a fault with either the missile or the tube.

Side note: I just looked at the Pershing II, and that MARV is crazy heavy compared to the W88/Mk5 MIRV. The whole Pershing II MARV package is nearly 1500lbs! It is also about 14ft long, which I am pretty sure is nearly twice the length of the Mk5 RBA.​

There were only 400 total W88s (~475kt) made. So that gives you 32x W88s per Columbia-class (Okay, technically 33 1/3 W88s per sub, but I rounded. This would leave you with 16x leftover W88s to spread wherever or save for rotation). If you load them as 4 per bird that gives you 8 birds per boat as your heavy hitters, half your load.

That leaves 6 birds with some number of W76Mod1s (90-100kt) each, likely 8 per bird but fewer is possible.

Interestingly, the W88 in Mk5 RBA weighs twice as much as W76 in Mk4 RBA, so the missiles would have the same range if loaded that way. IIRC the W76Mod2 was intended for depressed trajectories, so losing a lot of range despite the low payload weight.

Operational loadings will only continue to be so heavily downloaded if arms control treaties continue to hold up, which is an open question. The missile was designed with the original intent of keeping all missiles loaded with 8–12 warheads. Technically the original intent was that all missiles would be loaded with 8x Mk5/W88 warheads, but that obviously changed after Rocky Flats was shut down. Once the W93 begins to enter the mix, it is likely that force structures will start to change even more dramatically.

It's not quite clear what the loadout of the missiles with the Mk4/W76-2 RBs looks like. I would personally expect them to have a minimum of two RBs per missile simply for ballast considerations, as the Trident annulus does not allow for single warheads loadings to be balanced. If only one RB is loaded, the missile would have to input active corrections for the asymmetric unbalanced payload mass throughout the entire flight path. I'm not sure if that's practically achievable on this missile. Alternatively, it is possible that one Mk4/W76-2 is loaded, and one inert ballasted Mk4 RB (similar to those used on test flights) is loaded opposite of it.

There are no penaids deployed on Trident missiles. All US SLBMs have lacked the capability to deploy traditional chaff penaids (the only form of penaids that have ever been fielded by US missile platforms), and while some other forms of penaids have undergone limited development programs, none have ever been deployed on any US missile platform, land or sea based alike.

You could probably build a Mk500 Evader RB with a Mk4 physics package inside it, but the development program for this MARV was canceled long ago, and there is no clear and pressing need to develop MARVs for any US missile platform, let alone the sea-based deterrent arm. The Mk500 evader is purely useful for penetrating ABM defenses. It (like most other penetration-enhancing MARVs) actually degrades RB accuracy.

The Mk5/W88 does not weigh twice as much as the Mk4/W76. It's about 75–85% heavier.

There is no evidence to support the assertion that the W76-2 was intended for depressed trajectories. It uses the same aeroshell, nose tip, and missile as all other Mk4/W76 variants. You would need a completely different missile if you wanted a warhead to be optimized for depressed trajectories. Some limited evidence suggests that the D5 with the Mk4 or Mk5 may possibly be capable of certain limited semi-depressed trajectories at very short ranges, but the evidence in support of this is mixed and of very poor quality, and I've seen actual submariners strongly disagree with the assertion that the D5 is capable of depressed trajectory shots. Personally, I am doubtful that there is any operationally useful capability to use the D5 on DT shots.

The Pershing II is considerably more than twice the length of the Mk5 RB. Estimates vary for the dimensions of the Mk5, but the Mk5 RB is likely no more than 5 ft in length. The exact limits of the Trident II D5 annulus/shroud are not clearly disclosed, but it appears the maximum length you can fit under it is likely less than 6 ft at the very most (assuming close to zero height for connection hardware and such – in reality, max height would be limited even further than this).

As I understand it, both the W76 and W88 are to be replaced by the W93 (or Astraea A21 for the UK)using the Mk7 RB, which is estimated to be 250-350kT and probably closer to W76 weight-wise, using the W76 secondary and a new IHE primary. Given recent tests there's a possibility each RB will be a HGV.

In other news:


View attachment 789845

The W93 is not a replacement for the W76 or the W88. It is intended to supplement both of those models. The end intention is for the SLBM missile pool to consist of a mixture of all three models.

I have a paper that explains this, but it is technically not declassified (although the original lacks classification markers) and was not intended to be released online (a redacted version was officially released, but the unredacted version got leaked later on), so I am not sure if it's okay to post a link to it on this forum given the possible legal difficulties with posting potentially improperly declassified materials. You can find the paper I am referring to by searching "W93/Mk7 Navy Warhead - Developing Modern Capabilities to Address Current and Future Threats". A link to the unredacted version is available in the comments of the reddit thread that shows up as the first search result.

There is no evidence to suggest that the Mk7 RB will be a HGV (or a MARV). A traditional HGV would not be capable of physically fitting inside the Trident II D5 annulus/shroud. All evidence suggests that the Mk7 RB will be highly similar to the Mk5 RB in design, and will likely also be similar in size to the Mk5 RB (although it may be a bit smaller or larger).

There is no evidence to support the assertion that the Mk7/W93 will be closer to W76 weight wise, and plenty to suggest otherwise. Given the available information about this warhead (higher design margins in particular), it is safe to assume it will be much closer to the Mk5/W88 in weight.

There is no reason to believe that the W76 secondary would be used on the W93. It is more likely that a W62, W78, or W89 secondary is used instead (there are a number of other options as well beyond these). It would be pointless to equip it with a W76 secondary. That would not achieve the stated goal of diversifying stockpile risk, and would not allow for an increase in yield beyond 100 kt.

It is unclear if an IHE primary will be used on the W93. While this is certainly desirable if it turns out to be feasible, the Navy has very different thinking about risk vs the Air Force. The D5 propellant is detonable, and all existing naval SLBM warheads use CHE. Still, it seems likely that they will aim to incorporate a IHE primary if it is possible to do so without compromising too much on performance goals (which remains highly uncertain).

Interesting to compare the lower french yield on M51 with the US and GB ones.
M51is supposed to carry up to 10 warheads, with a classical layout of 6.

I never understood why the ASMPA, made for a 'ultime message' is a 300kt yied and the warheads of subs only 100kt.... Why not the contrary? Why not a small warhead on ASMPA (15kt) ?

It's because the French are solely concerned with countervalue targeting from their survivable at-sea deterrent.

The US is the odd man out with high-yield high-accuracy warheads on our survivable at-sea deterrent.

I do not know if it’s firmly established what current D5s can carry. Eight is number most often given but there have been some allegations/rumors of a dozen (presumably limited to smaller and lighter W76 if true).

The D5 is limited to a maximum of 8x Mk5/W88 warheads, or 12x Mk4/W76 warheads. The warhead adapter ring must be swapped out to accommodate a change between the two configurations.

The reason why the US would need to conduct new underground nuke tests is if the were designing and testing a new small, high-yield design since by pushing the envelope that hard means that even small discrepancies in the design can cause a dramatic reduction in yield or even a fizzle. I remembering years ago that the W-76's radiation-case (Made of U-238) was as thin as the aluminium wall of a coke-can.
Now on the other hand say the USAF issue a requirement for a new TN warhead of, say, one-megaton, 18-20" in diameter, 48" long and a weight of 1,500-2,000lb* (In other words a conservative design typical of the late 1950s-early 1960s) it could be designed and built without any need for a live nuclear-test.

*Basically a 21st century W-28 or W-49 other old warhead designs that could be built would the likes of the W-47, W-56 and W-59 in a 21st century update.​

The W76 radiation case is not made of U238. It is thought to most likely be made of a thin layer of aluminum with an extremely thin layer of a high-Z element vapor-deposited onto the inner surface of the radiation case. The most likely high-Z element used is probably either gold or lead, although several other possible candidate elements exist.

The W93 is intended to be based heavily off of existing designs, with extremely conservative design margins, far higher than any other US design.

Some people believe that they may go as far as to directly reuse an existing primary, secondary, or both from an existing weapon.

The most popular theory is that it'll reuse W78 secondaries, although I personally have my doubts over this theory. I believe even if it does end up using the W78 secondary design, it will likely still use newly manufactured secondaries rather than using reclaimed secondaries from decommissioned W78s (which is the popular theory). I am also not fully convinced that the W78 secondary is the most likely candidate for use.

And the W93/Mk7 is really pushing the envelope, 250-350kT at W76 mass (~95kg) by most estimates.

We have no firm data on yield for the W93 beyond that it will likely be somewhere in between that of the W76 and that of the W88. That implies a yield in the 100–455 kt range. 250–350 kt is not necessarily likely. It may be much lower. 150–200 kt for example is not out of the question at all.

There is no reason to believe that the Mk7/W93 will have W76 mass, and plenty of reason to believe otherwise, given that we know it will be developed with zero nuclear testing, be based on existing tested designs, and will have extremely conservative designs margins, far higher than any other US design. This points to a device with a weight similar to that of the Mk5/W88 (likely a bit lighter), but at a much lower yield than the Mk5/W88.

I don't think it's possible to develop a warhead with the W76's mass and a higher yield without any nuclear testing. Plenty of other experts agree with me on this.

The fate of the LLNL 100 kt/lb SLBM warhead also strongly agrees with this theory. That one was developed back in the era of unrestricted nuclear testing, and yet they couldn't deliver even half of the promised yield to weight.

Why would you expect to be able to double, triple, or even more the yield to weight of an existing super-optimized lightweight compact low-volume design? Even if unrestricted nuclear testing was still available, I wouldn't bet a penny on that being achievable. Its an absurd leap in performance for a design that was already on the bleeding edge of what was achievable back when there were no limits on nuclear testing.

I thought I read somewhere that the W93 is based on a proven [I don’t recall the word “tested”] design developed in the early 2000s that pointed to work completed on the RRW during that period?

Yes and no. It was said that all of the W93's key nuclear components will be based on currently deployed and previous tested designs, and that no nuclear testing will be necessary to deploy it.

I'd post the full quote, but I don't want to run afoul of forum rules (or get this place shut down), given that the source is from a leaked unredacted version of a document that was only officially released in heavily redacted form.

I've posted directions on how to locate a copy of the original document further up in this comment (see the section above on the "W93/Mk7 Navy Warhead - Developing Modern Capabilities to Address Current and Future Threats" paper).

 

[Titus K]

The UK warhead will be called the Astraea A21 and share the Mk7 RB.

Astraea (nuclear warhead) - Wikipedia

en.wikipedia.org

Thanks !

The Mark 7 aeroshell and some other non-nuclear components will be shared with the US under existing treaty arrangements. There will be collaboration with the US to ensure that the warhead remains compatible with the Trident missile system.[1][16] The Mark 7 aeroshell will require specific dimensions and weight distribution of the warhead.[17]

What about the neutron generator and gas reservoirs ?
If i remeber correctly, these are bough from US suppliers on the current Holbrook warhead.

 

[Scott Kenny]

Operational loadings will only continue to be so heavily downloaded if arms control treaties continue to hold up, which is an open question. The missile was designed with the original intent of keeping all missiles loaded with 8–12 warheads. Technically the original intent was that all missiles would be loaded with 8x Mk5/W88 warheads, but that obviously changed after Rocky Flats was shut down. Once the W93 begins to enter the mix, it is likely that force structures will start to change even more dramatically.

Uploading the W88s would only reduce the number of birds per boat, and constrict patrol areas due to reduced ranges.

Once W93s come online that will be less of an issue.

It's not quite clear what the loadout of the missiles with the Mk4/W76-2 RBs looks like. I would personally expect them to have a minimum of two RBs per missile simply for ballast considerations, as the Trident annulus does not allow for single warheads loadings to be balanced. If only one RB is loaded, the missile would have to input active corrections for the asymmetric unbalanced payload mass throughout the entire flight path. I'm not sure if that's practically achievable on this missile. Alternatively, it is possible that one Mk4/W76-2 is loaded, and one inert ballasted Mk4 RB (similar to those used on test flights) is loaded opposite of it.

The reason I think it'd only be one live warhead on those missiles is the mission profile. Popping off a single small/low-fallout nuke in order to de-escalate a situation before the full weight of STRATCOM turns the receiving country into a glass self-lighting parking lot.

So yes, it may end up as an inert/ballasted Mk4 for balance instead of purely various penaids.

There are no penaids deployed on Trident missiles. All US SLBMs have lacked the capability to deploy traditional chaff penaids (the only form of penaids that have ever been fielded by US missile platforms), and while some other forms of penaids have undergone limited development programs, none have ever been deployed on any US missile platform, land or sea based alike.

That can be fixed. And I'm sure MDA would like some "advanced targets" to test against, which would free up some funding outside the usual USN/SP channels.

You could probably build a Mk500 Evader RB with a Mk4 physics package inside it, but the development program for this MARV was canceled long ago, and there is no clear and pressing need to develop MARVs for any US missile platform, let alone the sea-based deterrent arm. The Mk500 evader is purely useful for penetrating ABM defenses. It (like most other penetration-enhancing MARVs) actually degrades RB accuracy.

How many places that would need to be smacked with a W76-2 have reasonably competent ABM defenses?

All of them.

The Mk5/W88 does not weigh twice as much as the Mk4/W76. It's about 75–85% heavier.

The (unclassified) paper I saw gave 8x W76 the same range as 4x W88s

There is no evidence to support the assertion that the W76-2 was intended for depressed trajectories. It uses the same aeroshell, nose tip, and missile as all other Mk4/W76 variants. You would need a completely different missile if you wanted a warhead to be optimized for depressed trajectories. Some limited evidence suggests that the D5 with the Mk4 or Mk5 may possibly be capable of certain limited semi-depressed trajectories at very short ranges, but the evidence in support of this is mixed and of very poor quality, and I've seen actual submariners strongly disagree with the assertion that the D5 is capable of depressed trajectory shots. Personally, I am doubtful that there is any operationally useful capability to use the D5 on DT shots.

Same paper that said 8x W76 had the same range as 4x W88s had a line or two for depressed trajectory ranges for D5.

As to why I think it's a probability for DT shots for the W76-2, that goes back to the mission profile: Smack a relatively valuable military target to let someone know they are stepping on the US very last nerve. Most of The Usual Suspects(tm) for getting such a smack are close to an ocean, and generally close to what are openly understood to be sub patrol areas. So either the missile goes very very high and everyone and God sees it coming for 20-30 minutes, or it goes on a depressed trajectory and there's 7-8 minutes flight time.

 

[JTR]

Uploading the W88s would only reduce the number of birds per boat, and constrict patrol areas due to reduced ranges.

Once W93s come online that will be less of an issue.

Patrol areas are already tightly constrained by other factors, such as the limited area with high-resolution gravimetric and bathymetric survey data. Historically, it has not been considered feasible to significantly expand patrol areas because of the extortionate expense of surveying new patrol areas.

The reason I think it'd only be one live warhead on those missiles is the mission profile. Popping off a single small/low-fallout nuke in order to de-escalate a situation before the full weight of STRATCOM turns the receiving country into a glass self-lighting parking lot.

So yes, it may end up as an inert/ballasted Mk4 for balance instead of purely various penaids

You can also have two live warheads, but set them up to fuse either one or both warheads depending on the actual mission profile. That way a single missile can carry out two different roles.

It's not too difficult to set it up so that the second warhead either detonates at unboosted "fizzle" yield (typically around 0.3 kt), or simply fails to arm and detonate at all (in which case it'd explosively disassemble upon impact with the surface, with zero nuclear yield).

And if you have a mission that calls for two warheads, then you would change the profile on that missile to arm and fire both warheads at full yield instead of wasting one of the warheads.

Given the absurd cost of the missile and the tight limits on missile tubes available, it'd be sensible to ensure you maximize the available options for this low-yield warhead variant, which means giving it more than one possible mission configuration. If you're going to dedicate an entire D5 missile to this (frankly idiotic) low yield warhead, you might as well stick at least two of them on the missile in order to get the most bang for your buck in terms of targeting options.

That can be fixed. And I'm sure MDA would like some "advanced targets" to test against, which would free up some funding outside the usual USN/SP channels.

The NMD/MDA already procures their own targets based off of historically developed US designs (and mirror-imaged Soviet/Russian designs). They have no need for a production program, they're already fully capable of building and testing examples on their own, and have already tested some examples against GBIs.

How many places that would need to be smacked with a W76-2 have reasonably competent ABM defenses?

All of them.

You and I must have very different definitions of "reasonably competent ABM defenses", because I can literally only think of four countries that have demonstrated the capability to deploy reasonably competent ABM systems of any kind:

• United States
  • GBI ABM – midcourse interception
  • SM-3 Blk II ABM – midcourse interception (however with much more limited capabilities compared to GBI)
• Russia
  • Moscow ABM – midcourse/terminal interception
• Israel
  • Arrow 3 ABM – midcourse interception
• China
  • HQ-29 ABM – midcourse interception (however with limited capabilities most comparable to the SM-3 Blk II, and with little firm information available to date)

(I am excluding all ABM systems that have sub-ICBM capabilities, as we are discussing ICBM defense here, not SRBM/MRBM/IRBM defense.)

So of these four, we clearly aren't going to be targeting ourselves, and it's exceptionally unlikely that we'd ever end up getting into a nuclear conflict with Israel.

So there's really only two countries that we might hit with nukes that have any level of competency in ABM systems – Russia (which only has Moscow protected – the rest of the country is completely unprotected), and China (which is unlikely to have more than a couple critical areas at most protected by the HQ-29 ABM system).

And it's unlikely that we'll be nuking either of those countries anytime soon. Iran and North Korea are far more likely targets, targets which are totally lacking in ICBM-class ABM systems.

The (unclassified) paper I saw gave 8x W76 the same range as 4x W88s



Same paper that said 8x W76 had the same range as 4x W88s had a line or two for depressed trajectory ranges for D5.

As to why I think it's a probability for DT shots for the W76-2, that goes back to the mission profile: Smack a relatively valuable military target to let someone know they are stepping on the US very last nerve. Most of The Usual Suspects(tm) for getting such a smack are close to an ocean, and generally close to what are openly understood to be sub patrol areas. So either the missile goes very very high and everyone and God sees it coming for 20-30 minutes, or it goes on a depressed trajectory and there's 7-8 minutes flight time.

If it's the same paper I'm thinking of, that paper has some major errors in it.

For example, it proposes some comedically implausible propellant performances as alternative options for the Trident (literally higher specific impulses than any tested propellant has ever achieved, and with zero consideration of the issue that propellants with even a fraction of that level of performance are so horrifyingly unstable and hostile to human life that it'd be outright impossible to safely field (let alone launch) a missile using them), and then at one point suggests that the entire Trident bus could be offloaded from the missile (wtf? talk about non-credible!).

I don't think the entire paper is trash, but with major errors like those in it, I am very skeptical of the conclusions they drew (which they even acknowledge have major limitations. and would require levels of nosetip performance well beyond what the RB was likely designed for).

I'm not arguing that a DT capability wouldn't have some military uses. It clearly would.

But I'm arguing that there's little to no credible evidence that the D5 missile (as well as the Mk4/Mk5 aeroshells) were designed with any significant DT capability.

And there are reasons to avoid achieving DT capabilities, as having such a capability is highly destabilizing. The main use for a DT capability is to perform a first strike decapitation shot against a peer state. It's hard to argue that this capability is a necessary part of a peaceful nuclear deterrent, and it tends to undermine arms control treaty efforts to have this kind of capability.

I know arms control treaties are effectively no longer relevant in the modern world, but they were still highly relevant back when the C4, Mk4, D5, and Mk5 were being developed, and would have influenced their development.

 

[JTR]

Thanks !

What about the neutron generator and gas reservoirs ?
If i remeber correctly, these are bough from US suppliers on the current Holbrook warhead.

Virtually every single part in the UK's warhead was designed and manufactured in the US. The only real exceptions are that the physics package is manufactured in the UK (although the design used is 99% US content), and final assembly occurs in the UK.

The only UK-manufactured part is the warhead physics package (primary, secondary, and radiation case). Even that is basically just an off the shelf US warhead design with a slightly different formulation of CHE used. The UK didn't even do the testing on the switch to a different formulation of CHE, they had to get US nuclear weapons labs to handle the qualification testing to enable changing the formulation of CHE used.

I see no reason that the Mk7/W93 will be any different from the Mk4/W76. The US will settle on a final warhead design, then hand a copy of the blueprints to the UK. The UK will maybe decided to go with a slightly different variant of HE, have the US do the qualification testing to confirm it's okay to change out the HE, then the UK will manufacture the physics package domestically, and procure all other warhead/aeroshell parts (the AF&F, the aeroshell, the neutron generators, the gas transfer system, the release assembly, etc) from the US as they have always done. They'll assemble it domestically, call it a "angelicized W93", and refuse to answer questions about just how much UK domestic content is in the warhead (spoiler: next to nothing).

The UK is shockingly dependent on the US for virtually every aspect of its at-sea deterrent. Even the HEU fuel in their submarine reactors is sourced from the US.

It really makes it starkly obvious by contrast just how incredibly impressive the French program is, where every single component is truly domestically designed and produced.

 

[Scott Kenny]

Patrol areas are already tightly constrained by other factors, such as the limited area with high-resolution gravimetric and bathymetric survey data. Historically, it has not been considered feasible to significantly expand patrol areas because of the extortionate expense of surveying new patrol areas.

Except we're not expanding patrol areas.

We're losing them due to shorter missile range.

(frankly idiotic) low yield warhead

No argument there. The logic is really stretched tight, and I'm not sure I see a situation where using one would not immediately lead to a full nuclear exchange.

But it exists, and I think what I've said is the CONOPS they're running from.

The NMD/MDA already procures their own targets based off of historically developed US designs (and mirror-imaged Soviet/Russian designs). They have no need for a production program, they're already fully capable of building and testing examples on their own, and have already tested some examples against GBIs.

Which means those (top-end) designs need to get developed into versions usable on missiles.

You and I must have very different definitions of "reasonably competent ABM defenses",

[...]

And it's unlikely that we'll be nuking either of those countries anytime soon. Iran and North Korea are far more likely targets, targets which are totally lacking in ICBM-class ABM systems.

I'm not convinced NK is lacking in ABM defenses. Yes, I know that ICBMs come down a hell of a lot faster than TBMs.

Plus I'm still assuming that Xi is going to start something over Taiwan which will likely end up with someone trying that peewee nuke to get their attention.

I'm not arguing that a DT capability wouldn't have some military uses. It clearly would.

But I'm arguing that there's little to no credible evidence that the D5 missile (as well as the Mk4/Mk5 aeroshells) were designed with any significant DT capability.

And there are reasons to avoid achieving DT capabilities, as having such a capability is highly destabilizing. The main use for a DT capability is to perform a first strike decapitation shot against a peer state. It's hard to argue that this capability is a necessary part of a peaceful nuclear deterrent, and it tends to undermine arms control treaty efforts to have this kind of capability.

I think it would be very poor planning on anyone else's part to assume that the US cannot do Tridents on a DT.

I know arms control treaties are effectively no longer relevant in the modern world, but they were still highly relevant back when the C4, Mk4, D5, and Mk5 were being developed, and would have influenced their development.

100% agree here.

 

[NMaude]

The most likely high-Z element used is probably either gold or lead, although several other possible candidate elements exist.

Gold? Talk about literal gold-plating, now as for Lead there is one other alternative that I do believe has been used in past US TN warheads (The SUBROC's W55 for example) and that is Thorium. Thorium (Th) and Lead (Pb) have similar mechanical properties but Th has a Z of 90 as opposed to Pb's Z of 82, having a much higher atomic-number (Z) means that a thinner (Hence lighter) radiation case can be used.​

 

[Desertfox]

Gold would probably not even crack the top 5 in expensive materials used...

 

[NMaude]

Gold would probably not even crack the top 5 in expensive materials used...

Plutonium I gather is a great deal more expensive than Gold.

 

[zen]

Virtually every single part in the UK's warhead was designed and manufactured in the US. The only real exceptions are that the physics package is manufactured in the UK (although the design used is 99% US content), and final assembly occurs in the UK.

The only UK-manufactured part is the warhead physics package (primary, secondary, and radiation case). Even that is basically just an off the shelf US warhead design with a slightly different formulation of CHE used. The UK didn't even do the testing on the switch to a different formulation of CHE, they had to get US nuclear weapons labs to handle the qualification testing to enable changing the formulation of CHE used.

I see no reason that the Mk7/W93 will be any different from the Mk4/W76. The US will settle on a final warhead design, then hand a copy of the blueprints to the UK. The UK will maybe decided to go with a slightly different variant of HE, have the US do the qualification testing to confirm it's okay to change out the HE, then the UK will manufacture the physics package domestically, and procure all other warhead/aeroshell parts (the AF&F, the aeroshell, the neutron generators, the gas transfer system, the release assembly, etc) from the US as they have always done. They'll assemble it domestically, call it a "angelicized W93", and refuse to answer questions about just how much UK domestic content is in the warhead (spoiler: next to nothing).

The UK is shockingly dependent on the US for virtually every aspect of its at-sea deterrent. Even the HEU fuel in their submarine reactors is sourced from the US.

It really makes it starkly obvious by contrast just how incredibly impressive the French program is, where every single component is truly domestically designed and produced.

Grand claims there, so perhaps some references to back it up.

 

[JTR]

Except we're not expanding patrol areas.

We're losing them due to shorter missile range.

My point is that patrol areas likely were never expanded significantly in the first place due to the prohibitive financial barriers on surveying new patrol areas, so we aren't losing much in terms of real-world capabilities.

Also, the Trident II D5 was explicitly designed as a 4000 nmi missile, not as a 6000 nmi missile. If increased range was so critical, it would have been designed differently. Any benefit from the downloading forced by arms control treaties is incidental at best.

Note also that the Trident is designed for 4000 nmi ranges. At longer ranges, accuracy degrades linearly. It is not as credible of a hard target counterforce weapon when it is heavily downloaded (unless of course extra ballast is loaded to replace the offloaded RBs, in which case you lose the benefits of increased range, but gain the original accuracy performance back).

Obviously this doesn't matter for the assured destruction (countervalue) mission, but it does greatly affect things for counterforce targeting missions.

Which means those (top-end) designs need to get developed into versions usable on missiles.

The US arrived at the conclusion that the best decoy is another warhead many decades ago, and has followed that policy ever since arriving at that conclusion.

While decoys and MARVs have been developed to hedge our bets and ensure our policy remains grounded in real-world facts, it is extremely telling that despite sinking an incredible amount of resources into multiple programs to develop numerous different types of decoys and a plethora of different MARVs, we have never deployed a single one of the many different types of developed decoys (with the exception of chaff), and the only MARV ever deployed was the terminal homing MARV on the Pershing II.

The only decoy ever deployed by the US has been chaff, and even that remains limited to only the MM III and its predecessors. The Peacekeeper/MX had no capability to deploy chaff, albeit likely in part because adding on said capability would make the missile too large to fit into a MM III silo.

No naval SLBM has ever had the ability to deploy chaff, due in part to tight space and weight constraints in launch tubes, and in part to the extremely limited specific impulse of the gas generators used on naval PBVs. Decoys have never been deployed on (US) naval SLBMs. Multiple MARVs were developed for naval SLBMs, but never deployed.

The strategy used on both land based ICBMs and sea based SLBMs has consistently focused on implementing the US policy of using more warheads in lieu of decoys, as the best possible decoy is another warhead, and historically we have been able to develop missile platforms capable of carrying multiple extremely miniaturized RVs/RBs.

At one point we were considering deploying a MM III variant with at least 7x MIRV RVs, in part as an anti-ABM measure. There is far more depth left in the existing strategy of overcoming ABM through warhead count alone than you would expect.

Will we need to consider deploying decoys in the future? Unlikely, as most decoys are highly fallible and can be easily discriminated by the combination of X band radar and IR seeker kill vehicles. Saturation of ABM defenses with salvos of heavily MIRVed missiles is the most credible approach available with the highest chance of penetrating said defenses. Decoys only help against shitty sub-par systems. Are you wanting to bet that China isn't going to shell out the cash for X band interceptor radars? I wouldn't.

Will we need to consider deploying MARVs in the future? Probably not for decades to come. And if arms control agreements are truly a thing of the past like they appear to be, then there is no real need to develop MARVs when building more warheads is a generally superior approach. Our current warhead counts are already more than adequate to penetrate any existing ABM system with plenty left over for other targets and additional strikes.

Also, MARVs have little to no utility against midcourse interceptors. They are really only useful against terminal interceptors, and there is only one terminal interceptor system capable of intercepting ICBM warheads in the entire world (the Russian Moscow ABM system). Even that system seems to lean heavily on midcourse defense nowadays, so it's unclear how useful MARVs would even be against it.

I'm not convinced NK is lacking in ABM defenses. Yes, I know that ICBMs come down a hell of a lot faster than TBMs.

Plus I'm still assuming that Xi is going to start something over Taiwan which will likely end up with someone trying that peewee nuke to get their attention.

I don't think you understand just how atrociously expensive and high tech ABM systems capable of intercepting ICBM RVs are. There's a very good reason why only four countries have managed to build these kinds of systems, and why even those four countries only have extremely limited systems in place.

It's possible that China might lend NK a HQ-29 battery at some point, although I can't see the benefit to China to doing so anytime soon given that domestic defense will be a far higher priority for the next few decades, HQ-29s are likely to remain in short supply for quite some time, and it would be foolish to gift NK with your one type of system capable of handling ICBM intercepts until most of your domestic military targets are protected.

Even if they do lend NK a HQ-29 battery, you can exhaust the interceptor missiles quickly enough by simply firing more warheads at the target. The presence of a HQ-29 battery would be easy enough to see on satellite images, so if one is present, we'd just adjust our firing plans to account for it, no different from how like we've long adjusted the firing plans against Moscow to account for the Moscow ABM system.

If we're fighting China, again, this is no different from the Moscow ABM system. Bypassing these ABM systems is a solved problem. We have plenty of warheads to spare. If we need to strike a target that's protected by a HQ-29 system, we'll simply allocate more warheads to the target in order to ensure the ABM system is overwhelmed and the target gets struck successfully.

Also, again, as this is a midcourse system, MARVs would not be helpful against it in the first place. And as it uses IR seekers, most forms of decoys are unlikely to be useful. It's unclear if it uses X band radar. If it does not, some forms of decoys may still be useful, but there's nothing stopping them from upgrading it to use X band radar in the future, rendering any effort put into developing anti-IR decoys completely useless.

I think it would be very poor planning on anyone else's part to assume that the US cannot do Tridents on a DT.

Argumentum ad ignorantiam, seriously?

We have no reason to believe any significant DT capability exists.

Sure, Russia and China will probably assume a DT capability exists for contingency planning purposes. But that is not the same thing as a DT capability actually existing.

Gold? Talk about literal gold-plating, now as for Lead there is one other alternative that I do believe has been used in past US TN warheads (The SUBROC's W55 for example) and that is Thorium. Thorium (Th) and Lead (Pb) have similar mechanical properties but Th has a Z of 90 as opposed to Pb's Z of 82, having a much higher atomic-number (Z) means that a thinner (Hence lighter) radiation case can be used.​

Funnily enough, most US warheads actually cost more than their weight in gold to manufacture.

Older warheads used thick rad cases, usually made of uranium or thorium. Even older designs used lead-lined steel cases. However these have been dead for quite some time, as the weight and volume penalty from using a thick rad case is quite severe.

Modern warheads generally use rad cases that are only a few mm thick. The outer layer is usually an aluminum alloy, although magnesium or even plastic may be used instead. The inner layer is a high-Z element, and is likely applied using PVD (or a similar process).

It's unclear what material is most commonly used for the inner layer in modern thin wall rad cases. Uranium is obviously ideal when possible, but there are indications that lead and gold may have been used in certain weapons.

The biggest question is if it's practical to use PVD to deposit a thin layer of uranium on a thin walled plastic, magnesium, or aluminum radiation case. If it is, then PVD deposited uranium is probably the preferred option. If it is not, then it's likely that an alternate element is used instead, one which is more conducive to being applied using PVD. That may be thorium, or it may be lead, or a lead-bismuth alloy, or gold, or something else entirely. It doesn't necessarily need to be the highest-Z element available.

Obviously this area is still very highly classified, so it's difficult to obtain much in the way of firm information on the specifics.

Plutonium I gather is a great deal more expensive than Gold.

Weapons grade plutonium makes gold look cheap by comparison.

Grand claims there, so perhaps some references to back it up.

What specifically do you want references for?

These are far from grand claims. If you've done even the slightest bit of research on this topic area, you should already be familiar with many of these issues.

I have literally dozens of papers covering different aspects of this topic, please be specific about what you want me to provide citations for.

It takes a great deal of time to go through my archive, locate the relevant papers, dig them out of my archive, locate a live version online, format the reference, and write it up in a post. I am more than happy to do this if you can give me a specific scope that you want refs for. But I am not interested in wasting my time on an open-ended request. This is a hobby for me, not my job.

 

[BonPlan2]

Had not heard that. I am surprised by that; last time I had heard anything about new U.S. warhead design a coupe decades ago I had thought the emphasis was on reliability and low maintenance, not further miniaturization.

Theorical max yield per kilo is 6kT / kilogram.

Max real is in the 5.2 kT / Kg. The heavier the warhead, the most effective it is.

 

[TwoTribes]

 

[Josh_TN]

if the W76/88 are remaining in service, what need does the W93 address? You mentioned diversifying stockpile risk - since two types of warhead are already available, what are the concerns? If the design is much more conservative, is this something of a return to reliable warhead in design philosophy, even if not it actual design?

Previously were there not efforts to increase the accuracy of weapons using very small control surfaces/tabs to nudge the RV dozens of meters close to the aim point? Do you think there is any chance this be used with the smart fusing system to adjust downrange accuracy?

Regarding W76 mod2, I had always assumed two warheads were loaded and fired on the same target, with the second being disabled if the first fused correctly. I figured even with low yield, a double tap is in order, especially since such a weapon would likely be used at a pivotal moment to restore deterrence (I strongly disagree Iran or North Korea would be likely targets).

 

[Desertfox]

You can't change the fuzing on the fly, but you could launch two warheads close enough that the second one gets fratricide if the first one works.

To add to ABM systems, nuclear-tipped ABMs (like Russia's) have the issue that the first nuke going off will blind the radars. They can deal with a single incoming warhead, but a well timed coordinated attack can easily make it thru, since the system can only engage one RV with high PK at a time.

 

[JTR]

if the W76/88 are remaining in service, what need does the W93 address? You mentioned diversifying stockpile risk - since two types of warhead are already available, what are the concerns? If the design is much more conservative, is this something of a return to reliable warhead in design philosophy, even if not it actual design?

Previously were there not efforts to increase the accuracy of weapons using very small control surfaces/tabs to nudge the RV dozens of meters close to the aim point? Do you think there is any chance this be used with the smart fusing system to adjust downrange accuracy?

Regarding W76 mod2, I had always assumed two warheads were loaded and fired on the same target, with the second being disabled if the first fused correctly. I figured even with low yield, a double tap is in order, especially since such a weapon would likely be used at a pivotal moment to restore deterrence (I strongly disagree Iran or North Korea would be likely targets).

Currently there is a severe over-reliance on the W76 in the SLBM arsenal.

There's something like 400 W88 warheads vs 2000 W76 warheads (actually 3400, but only 2000 of these went through LEP).

Introducing the W93 will allow for diversification, reducing the risks associated with the current stockpile (which is virtually a W76 monoculture).

Both the W76 and the W88 are due to age-out and will require either subsequent LEP programs or replacement pretty much simultaneously in the mid term (2030s–2040s). Introducing the W93 will help mitigate risks associated with having to run new LEP programs for those warheads. If unexpected issues crop up during the LEP program for either warhead, the W93 can help fill the gap. It will also make it easier to schedule and process the LEPs for such a large number of warheads. It also offers the option of scaling down the LEP program for one (or possibly even both) of those warheads.

In the long run, the Navy wants to eventually try to phase out the W76 and the W88. The W93 is the first small step toward this goal, however it is far from the last step. IIRC the plan is to introduce at least 1–2 additional naval warhead designs after the W93.

A lot of concerns are raised regarding the transition to the much smaller Columbia fleet, and the need to ensure adequate capabilities despite this significant downsize in the number of available tubes. It's implied that the goal for the W93 is to be a lighter warhead than the W88, with a premium being placed on lighter weight instead of on higher yields, but also with the goal of having higher yields than the W76 in order to provide semi-adequate counterforce capabilities. The W76 yield is apparently unacceptably low for this use case.

There's also a serious issue with W88 attrition over time. The number of available W88s is slowly decreasing over time, which is much cause for alarm, and a large part of why the W93 is so urgently needed. If adequate numbers of W88s were available, much of the justification for the W93 wouldn't exist.

There are various other reasons and justifications, but I'm not going to list all of them. You can read the original document if you really want to know. I posted instructions on how to locate the leaked unredacted copy in one of my earlier comments. I highly recommend giving it a read through, it's quite educational.

Yes, it is likely quite similar to the RRW in terms of the design philosophy. It is quite possible that they may end up just using one of the old RRW design options for the W93, although as of now there is no way to tell for certain.

I doubt that we'll see any form of active accuracy MARV. It'd take up a lot of space and weight that they're clearly already planning on using for increased design margins, higher yields, and minimized weight. The main goal is a warhead of higher yield than the W76 but of lighter weight than the W88. If you wanted an accuracy MARV, you'd be looking at W76-like yield at best given the space constraints of the Trident II D5 annulus/shroud, and you wouldn't be able to meet the stated yield or weight goals for the W93.