From a post to DAVIDWEBER.NET forums on 5/25/2011

Launcher design: PODs vs. tubes

    A few points about your analysis.

    (1) The actual "spinning up" interval on the Mark 16 is actually closer to 5 seconds from initiation to completion.

    (2) I keep talking about the problems when "capacitors" get hit onboard ship, but I don't think people are quite getting what's being hit. These are plasma capacitors, and each missile pod is equipped with sufficient capacitor capacity (ouch) to initiate fusion in all of its missiles, whether they be Mark 16s or Mark 23s. Hence a hit on one of them is going to have all sorts of, ah, negative consequences.

    (3) The loading/arming/firing sequence on a pod actually requires the pods' capacitors to be "charged" for considerably longer intervals, including the intervals in which they are being held "in the queue" before launch, than the interval a "live" Mark 16 spends in the tube. In both absolute terms, and certainly on a time-interval-per-missile-launched, the pod represents a considerably greater danger to the ship than the tube does. Missile tubes are also fitted with massive containment armor and a "blow out" design to vent plasma outboard rather than inboard. (I realize "venting" is a relative term when we're talking about plasma, but it is an aspect of the design philosophy, as is the additional armoring of the core hull inboard of the tube. This isn't really a practical design feature where a pod is concerned. Oh, and I should point out that without the additional, heavy armoring which is part of the tube installation, the mass and volume cost advantage of the tube over the pod would be even more pronounced.)

    (4) This should probably be part of (3), but that paragraph was getting too darned long so I decided to split it. [G] The actual exposure of the ship to destruction due to rupturing of fusion containment is much higher in a pod-layer, and each of those ruptures will be substantially more powerful/destructive than the loss of a single charged Mark 16 in its tube. This is an important aspect of the RMN's survivability thinking. I realize you were also talking about a far more rapid "cascade launch" in the part of your post I snipped, and that would at least have the advantage of getting the fully charged pods out of the bay in a hurry, but the energy handling requirements to charge all of those plasma capacitors will still be a problem. It takes time, and while the process is underway, the pod-layer is rather like a Japanese carrier at Midway with fully fueled and armed aircraft on the flight deck, more being worked on on the flight deck, and trolleys of bombs and torpedoes lying around "just until we can get them into the magazines." (see also my point below)

    (5) The number of tractors and the management systems to coordinate them for the simultaneous launch of a few hundred or a few thousand pods would be… burdensome to say the least. In addition, when the capacitors on a pod fail during the initiation stage — or before (and it does happen from time to time) — you will lose a fair number of pods if they are in too close a proximity.

    There are other factors involved in the survivability trade-off between "conventional" tubes and pods, and strong arguments can be made in favor of adopting either of them. Indeed, the RMN has adopted both of them, in a manner of speaking. They'll almost certainly continue to use both of them, although they are also rethinking the architecture of the podnought in terms of survivability features.

    BTW, one thing the Manties are having to sit on themselves pretty hard over is the difference in lethality between their hardware and anything the Invincible Solarian League Navy has so far brought to the table. The truth is that the SLN would require an advantage of… well, let's just say that it runs to three-numbers-to-one in missile launches just to break even on damage inflicted… and their designs are inherently less survivable when hit, in addition.

    Sorry. "Management systems" wasn't referring to "systems" as in control loops or links; it was a poor choice of words when a better one would have been "traffic management of constellations." I was referring to how you "manage" that many pods (and places for them to lock tractors) in a finite space, especially if you are trying to keep them inside the wedge for protection before you actually fire the things. It's not an electronic function or a software function; it's a space availability function.

    I'm not a-gonna tell you how far from the hull you can put a buckler. (Nyah-nyah! [G]) If you wish to assume this is because that very question may become, um relevant in the not too distant future, you are of course entitled to do so. All I've got to say is tum-te-tum-te-tum.

    On design aspects from the forums, the truth is that by and large I tend to stay away from them on purpose. I've already worked out pretty much where I'm going, and I have this really bad problem. Once I look at a forum, I start replying to it. And replying to it. And replying to it. Oh, you'd noticed already? [G] Anyway, it turns into a really dangerous time sink (in fact, I'm going to have to get back to writing — and staying away — Sometime Real Soon Now), and it also gets me into revealing stuff I don't want to reveal yet. There are a few threads I've poked my nose into because of an ongoing interest in them and because I've already been involved in public discussion of them, but by and large I try to avoid that. Sorry! I'd really love to spend the time to do it, but I'm not going to do it at all unless I have the time and availability to do it right.

    I will say that the next class of Manty SD(P)s (which will not be built to the design originally intended to replace the Invictus, since the Admiralty's been doing some more rethinking) has already been pretty much finalized. Details like sizes and masses and exact proportions of tonnage devoted to different functions remain to be worked out, but the main features are effectively fixed.