Wikipedia:Reference desk/Archives/Science/2015 February 17

= February 17 =

Modern battering rams
Let's say you take a modern battering ram to a locked metal door and break it in: how does it work? Moments after I took this picture, the firefighters did exactly that; I assumed that it was a simple wooden door, but the neighbors standing next to me said that it was metal. My uninitiated impression is that it would just bounce off, or simply cause the door to buckle and not fit in the doorway anymore (if you cover something with a piece of aluminium foil and then poke it in the middle, it no longer covers everything), but in this case, the door appeared to swing open as if they'd used a key and turned the knob. Nyttend (talk) 04:28, 17 February 2015 (UTC)
 * We have the articles Battering ram and Door breaching which cover this a bit although not perhaps in details of your question. I'm not sure about firefighters, but AFAIK whenever I have seen videos or pictures of real life or training (i.e. not fictional) situations of battering ram door breaching by SWAT like teams or special forces or whatever they generally use the battering ram near the lock which I presume increases the likelihood of breaking it. See for example the picture in our article or or  . Or even these failures  . Or perhaps even this ]. Note, whether special forces of fire fighters, they probably don't care if the door itself breaks, the point is the lock is I imagine generally a weak point compared to the door itself, or even the hinges. (Our article does mention either the lock or hinges are generally the target locations for ballistic breaching i.e. using a firearm of some sort with shotguns and the lock generally being the preferred options.) If the lock itself doesn't break and they have to break the door to some extent, then it would seem usually it wouldn't be the case that targetting another location would have helped. In fact their attempts to breach the lock may break enough that they can try to use a crow bar or some other tool to help them breach as the failures somewhat show. (I imagine particularly for fire fighters who unlike SWAT & special forces aren't often faced with doors intentionally designed to be hard to breach, the percentage of times when the lock breaks or frame holding the lock breaks or the door otherwise opens at the lock. Even for SWAT & special forces teams, I would guess they would normally try to prepare if possible and figure out whether their battering ram is going to work or they should use some other method.) Nil Einne (talk) 06:57, 17 February 2015 (UTC)
 * Thanks for the detailed response; I'd seen the battering ram article (I read it immediately before coming here), but not the door breaching article. I hadn't really considered the lock itself (I'd really only considered ancient/mediaeval ramming, which splinters wood and cracks masonry), thus the confusion.  Nyttend (talk) 07:23, 17 February 2015 (UTC)


 * It's all about defeating the weakest point. The door itself is unlikely to be that - generally, the place where the bolt passes through the frame is easily the weakest point.  So applying a lot of force for a brief amount of time (which is what a battering ram does) is likely to fracture the material on the far side of the bolt.   This is why modern dead-bolts come with 6" long screws that go through the striker plate and into the largest, heaviest pieces of timber in the door frame.  Those screws are what have to be defeated here - the screws either have to bend or break, allowing the wood holding them in place to splinter and the bolt to pop out.   Probably the best defense against that kind of effort to open the door would be one that opens outwards because then the entire frame can support the door and prevent the force of the ram from being concentrated onto the door bolt and striker plate.   The battering ram works because it weighs over 100kg - and since Force equals Mass times Acceleration, the more solidly the lock resists motion, the higher the deceleration of the ram at the moment of impact, and the larger the force applied to the bolt and hence through the striker plate to the frame.


 * Destroying the lock itself may not actually have much effect because the bolt is still likely to pass through the striker plate in the frame and to continue to prevent the door from opening. SteveBaker (talk) 14:39, 17 February 2015 (UTC)


 * The model used by London's Metropolitan Police is called an Enforcer; it weighs 16 kg (35 lb) and exerts an impact pressure of 3 tonnes. There's a whole playlist on YouTube calle Breeching Methods; all the ones that I looked at seemed to be targeting the lock area of the door. Alansplodge (talk) 18:12, 17 February 2015 (UTC)
 * Hmmm...the kind I've seen in use in the US military is carried by four guys - evidently intended for chunkier doors. Clearly there are multiple types in use. SteveBaker (talk) 04:58, 18 February 2015 (UTC)
 * I should clarify when I said lock above, I was including everything that made up the lock, including the dead bolt, and the entry of the dead bolt in to the door frame. Nil Einne (talk) 05:24, 18 February 2015 (UTC)
 * Outward opening doors is not something I really considered in my above answer. From what I can tell from sources like    (Yahoo Answers but has people which sound like they have experience)  (doesn't really discuss that well)   is other methods will normally be used, such as prying the door open or breaking the hinges. A battering ram could still be used, either to try and damage the door or to force in the Halligan bar (although a sledge hammer or axe or something else may be used edit: instead of the battering ram I mean). A notable point with outward opening doors is that the hinges may be exposed so could be a weak point (although I think more likely to be something law enforcement by which I'm including special forces etc and criminals). One thing I was thinking about but didn't mention is doors with a metal bar or similar across them. There is some discussion of how to deal with these in some of the sources. Also I was a little wrong above in suggesting there won't be any concern about breaking the door. I was primarily thinking of a real emergency. In less important cases like where there is an alarm in an apparently locked and empty office, an attempt may be made to minimise damage . (And another point I was thinking but didn't mention very well is in some cases law enforcement may not necessarily care so much about time taken, as opposed to a successful breach without detection until the breach. Of course, sometimes it's not clear what was being attempted ....  (from the above playlist).) With both firefighters and others, the sources generally remind to check and consider all possible points of entry (including checking that you even need to force entry), and have a quick look before forcing to try and determine the best method of doing so. Edit: Should clarify that even in the case of outward opening doors, if you aren't trying to compromise the hinges, you're normally still targetting the lock area. It's just that you still want to open it outwards if possible rather than trying to force it inwards. Although it may be slightly above or below the lock, presuming you can't break the bolt as suggested below. Edit2:  may also be of interest. It doesn't go much in to breaching methods, at least that I noticed but does mention the tools firefighters may use, at least in whatever area the person preparing it is from I presume. Nil Einne (talk) 06:44, 18 February 2015 (UTC)


 * Nobody has yet mentioned a Halligan bar, which is a very effective axe-like tool that can be used to shear a bolt. Here is a video clip of USAF tactical door breaching using this tool.  Nimur (talk) 07:45, 18 February 2015 (UTC)
 * Actually I did above although I didn't go in to detail about how it may be used and in particular didn't mention this sort of usage. Most of the examples I saw were more using it to pry rather than shear the bolt, I presume because it was decided this wouldn't work and/or a lack of training and/or would probably take longer than the alternatives. Nil Einne (talk) 12:50, 18 February 2015 (UTC)
 * My apologies - Nil Einne definitely gets credit for mentioning this first. Sorry I hadn't read carefully enough!  Nimur (talk) 16:31, 18 February 2015 (UTC)

About biological computer simulations...
Could a virtual human model be constructed by making a molecular dynamics simulation of the DNA using the appropriate conditions like making a virtual womb or a virtual fertilized egg?

similar to this: http://www.cell.com/abstract/S0092-8674(12)00776-3

41.235.27.248 (talk) 08:39, 17 February 2015 (UTC)


 * That model does not use Molecular dynamics. It just updates a large number of cell variables using a combination of "28 Submodels of Diverse Cellular Processes", "For example, metabolism was modeled using flux-balance analysis (Suthers et al., 2009), whereas RNA and protein degradation were modeled as Poisson processes.". Obviously (and this cannot be avoided) the model is a gross oversimplification of the actual molecular processes taking place in a cell. It would theoretically be possible to make a similar model that takes into account cell multiplication and growth of an organism, but the processes that give rise to human development are highly complicated and some aspects are not well understood.
 * Simulating a human being at a truly molecular level however is far beyond the capabilities of our computers and will probably never be feasible. - Lindert (talk) 10:34, 17 February 2015 (UTC)
 * In fact it breaks down at a very basic level. DNA contains the recipes for building the body's proteins.  The function of a protein depends on the way it folds up into a 3-dimensional shape.  Currently we don't have any computationally tractable way of computing the folding pattern of an arbitrarily chosen protein.  See protein structure prediction for more information on the "protein folding problem". Looie496 (talk) 14:00, 17 February 2015 (UTC)


 * It's certainly possible in principle. Whether it's currently (or ever) likely to be practical is a matter of degree.  If you tried to simulate every chemical reaction path at the atomic level, then the computational complexity would be insane.   Just figuring out how one single protein will fold up given it's chemical structure (See: De novo protein structure prediction) is a task that is so complex that it's currently unsolved.   Such partial solutions that we have require hefty amounts of super-computer time and are not yet 100% reliable (See Protein structure prediction).   So it's clear that doing it reliably at the level of atoms is not likely to happen for a very long time.


 * However, we can start at a higher level by simulating the known chemical pathways - that provides a considerable speedup, but at the cost of a loss of fidelity. Since an individual's genome will have small differences in their DNA compared to the average of the entire population, some of their proteins are guaranteed to have subtly different chemical structures, to fold slightly differently than the 'standard model' for these chemical pathways - and to get that right, you're back into the problems of protein folding.   So it's likely that until we can solve the problem at the atomic level, we won't be able to reproduce the exact processes that (for example) produce a particular facial appearance or predict brain structure or whatever.


 * The approach of attacking the problem at the level of chemical pathways would allow us to skirt the issues of exactly how the chemical structure of one organic compound interacts with another would solve that problem and greatly reduce the computational complexity to the point where I think we could say that it might one day be possible.  But doing that will inevitably produce a 'generic' human from the virtual womb because tiny differences in a real person's genome would change some of those pathways in subtle (or perhaps not-so-subtle) and un-researched ways.   But even with the more modest goal of simulating a generic human, would depend upon us knowing every single chemical pathway in sufficient detail to reproduce it...and clearly we don't yet have that knowledge because new pathways are discovered all the time.  But there is a good chance that studies of human biology will eventually uncover 100% of those pathways for a 'generic' human, leading to at least the possibility of coming up with a workable (albeit generic) simulation.


 * But if you want to predict (for example) how an unborn child will turn out given just the DNA, that level of abstraction is useless.


 * We could go yet more levels higher - perhaps understanding various cell types and how they replicate and interact - but that starts to look a lot like we're "rigging" the results to come out right...and would require yet more generic information and therefore more generic results.


 * So I think the answer is basically "No" right now...and probably "No" for the immediate future. But the processes involved are (at least statistically) amenable to computation - so we probably just need a MUCH bigger, faster computer to run it all on.   Whether we ever get a computer that large and fast is hard to know - but the limits on the size of transistors is known - and making a computer physically larger always makes it slower (because the speed of light is finite) - so there are limits to how powerful they can get...and getting powerful enough to fully solve de-novo protein folding by brute force in anything like fast enough time to produce a simulation that would run to completion within decades may never be possible - so we're left hoping for an algorithmic breakthrough that may or may not happen.  Protein folding is very likely to be an NP-hard problem...so we shouldn't expect solutions that are both fast and accurate.


 * Bottom line, best guess: "No, it'll never happen".


 * SteveBaker (talk) 14:24, 17 February 2015 (UTC)
 * One of the problems is that we don't have a full understanding of how MUCH of the stuff in the DNA works to create biological traits. As noted, DNA only does one thing, make proteins.  But there are lots of second-, third-, fourth-, and umpteenth-order stuff going on here.  Some bit of DNA may make some tiny protein, which directs the assembly of some other bit, which itself directs the assembly of some other bit, and so on down the line recursively.  We're really only good at the first step, that is telling what specific amino acid sequence DNA codes for, basically what we call primary structure.  We don't know entirely how such proteins can reliably fold to higher levels of structure.  Take that out ten or twenty levels of "we don't know what happens next" and you see how far we are from constructing life from first principles using ONLY the DNA code.  And we haven't even gotten into things like epigenetics, which is a fruitful area of research looking into heritable traits which are NOT even coded for in nucleic acids.  -- Jayron 32 17:30, 17 February 2015 (UTC)
 * Erm no, one thing DNA patently doesn't do is make protein. It's a template to make RNAs, many of which then get turned into protein, but many of which also never get to be protein, and are perfectly happy doing their work as RNA (ribosomal RNA, micro RNA, snoRNA etc etc etc etc)Fgf10 (talk) 21:44, 17 February 2015 (UTC)
 * Fair enough. I skipped a few steps for the sake of brevity.  It's difficult to teach an entire class on cell molecular biology in the space of a few lines on a website.  Forgive me.  -- Jayron 32 03:43, 18 February 2015 (UTC)
 * There's no forgiveness in Hell. μηδείς (talk) 22:58, 18 February 2015 (UTC)

MIT study about living on Mars
I didn't understand this limitation put forth by the MIT study:


 * "For example, if all food is obtained from locally grown crops, as Mars One envisions, the vegetation would produce unsafe levels of oxygen, which would set off a series of events that would eventually cause human inhabitants to suffocate. To avoid this scenario, a system to remove excess oxygen would have to be implemented — a technology that has not yet been developed for use in space."

Thanks!  DRosenbach  ( Talk 21:45, 17 February 2015 (UTC)


 * Is it something to do with oxygen toxicity? The plants produce too much oxygen in an enclosed space and create an atmosphere that's too rich to support human life? --Kurt Shaped Box (talk) 21:55, 17 February 2015 (UTC)


 * Your link was to a press release from MIT, but here is the full study, thirty five pages (presented as a conference paper at the 65th International Astronautical Congress, Toronto, Canada). An Independent Assessment Of The Technical Feasibility Of The Mars One Mission Plan.
 * It appears the primary concern is that there is no sustainable way to control the molar fraction of oxygen or its partial pressure. State of the art control capabilities depend on ready access to large amounts of nitrogen gas - and when that runs out, the atmosphere becomes uncontrolled.  Around that time, the mole fraction of oxygen will be above that level considered safe from a standpoint of fire hazard, and the partial pressure of oxygen will be below that level considered safe from a standpoint of hypoxia.  The study authors know how exactly much gas such atmospheric control units need, because NASA has already built state-of-the-art spaceships (like Space Shuttle and International Space Station).  Gas leakage from such environments has been well-parameterized.
 * Cited source #30 is A Cabin Air Separator for EVA Oxygen; it was published in 2011 to investigate an oxygen partial-pressure management system suitable for International Space Station; and it explains details of the current state of the art technology. Significant risks include whether the outlet air is safe to breathe; whether the machine is a fire hazard; and the hazardous noise level that the machine may produce.
 * Nimur (talk) 22:09, 17 February 2015 (UTC)
 * It's quite apparent that a lot was left out of lecture during my periodontics training. :) Is this seen as a fundamental flaw in the idea, or is this just with, what we might call, current technology, such that all of this may be dealt with by whatever is developed over the next 10, 20 or 50 years?  DRosenbach  ( Talk 02:52, 18 February 2015 (UTC)
 * The report primarily concludes two things: that the cost will actually be much much higher (perhaps many orders of magnitude higher) than the proposed cost put forward by the Mars One team; and secondly, that the Technology Readiness Level is "relatively low." TRL is a methodology that organizations like NASA use to estimate technology evolution on a decades-long timescale.  Very low technology readiness levels imply that the concept has not even been demonstrated in principle - and that means that there is not presently a path to spaceflight capability for it.  In other words, we don't yet even know what we need to work on - so it's fruitless to try to estimate a schedule.  Will it be ready in the next 10, 20, or 50 years?  Reputable scientists decline to even speculate!
 * The way I read this report, the MIT authors are not actually saying that a Mars mission is fundamentally flawed. Rather, the MIT authors are taking a somewhat conservative position that any such mission - even using future technologies, so long as we're confident we could build them - will be much more expensive and much more difficult than the Mars One team believes - and then presenting 35 pages of details why.  If NASA does decide to commit to such a mission in the near future, you can bet that there will be a lot more study put into these and other details!
 * Nimur (talk) 03:08, 18 February 2015 (UTC)


 * Mars One is a put-on, by the producers of Big Brother. Saw this tonight on inside edition (which I was forced to watch, with my eyes clamped open as in A Clockwork Orange). μηδείς (talk) 01:24, 19 February 2015 (UTC)


 * check out the novel The Martian, as I believe the author directly addresses your question in the book using current technology. Viriditas (talk) 19:49, 19 February 2015 (UTC)

Tacking an untethered lighter than air vehicle?
We all know a hot air balloon goes where the wind takes it, controllable only by gaining/lowering altitude or dragging a rope on the ground. Whereas a kytoon, tethered to the ground, is capable of much more control. Still, I wonder... the inertia of a balloon, or in particular its heavy gondola, means that a kite tethered to the center of mass ought to exert a pull on it whenever the velocity vector of the wind changes. Has there ever been a successful use of this inhomogeneity to offer meaningful steering to a lighter than air vehicle without relying on propulsion or outside force? This peculiar offering is the result of one of my playing-VR-in-a-parallel-universe dreams; the design I dreamed involved a streamlined row of methane balloons supporting a base with fermentation tanks from which a telescoping tail of concentric rings of fabric louvers could be extended, drooping downward and in the direction of the wind; the operator, aboard the tail, used the controls so that it pulled at an angle to the wind. When the wind was completely unfavorable it could be retracted by using the weight of the tanks, with variable mechanical advantage, to retract the hawser from which the tail hung; the methane also provided some weak propulsion by propellers and allowed regeneration of lift gas permitting frequent changes in altitude. The dream was made more visually impressive in that, I thought, it was necessary to stay very near the ground to have the most variation in wind velocity... Wnt (talk) 23:37, 17 February 2015 (UTC)


 * When in doubt, consult the Balloon Flying Handbook. The FAA classifies a "balloon" officially as a non-steerable aircraft, while using other terminology (for example, "thermal airship") to refer to aircraft that are lighter than air, kept aloft by hot air, and are steerable by some means (most typically, powered propulsion).  There are also weight-shift-control aircraft, powered parachutes, and so on.  For the purposes of regulatory classification and categorization, these aircrafts are not balloons.
 * Refer to 14 CFR §1.1 Definitions for more information. It is important to know exactly how steerable an aircraft is, because a steerable airship has different right-of-way rules than a balloon or a glider or a weight-shift-controlled aircraft (14 CFR §91.113).
 * Nimur (talk) 00:20, 18 February 2015 (UTC)
 * An hot air balloon moves with its surrounding air so there is no air current/-movement difference for the kytoon to work. It might work if the kytoon could reach a different air layer but the goal of changing direction is much easier reached simply by driving the hot air balloon up or down there. --Kharon (talk) 04:02, 18 February 2015 (UTC)
 * maybe I should have started with a simpler question: when you're on the ground, it's quite common to see the weathervane twist this way and that, for gusts of wind to pick up and die out. This is wind shear, AFAIK.  How much wind shear affects a lighter than air vehicle that is close to the ground? (contour flying, as Nimur's FAA manual describes it)  The question then is whether this provides enough force, when properly tapped, to significantly alter its course from that of the overall wind. Wnt (talk) 13:56, 18 February 2015 (UTC)
 * Wind shear is hazardous to all aircraft, and it is particularly hazardous to a balloon. Wind shear can cause a balloon to fly in unusual attitudes; it can induce lift that tugs the balloon downwards (in the opposite direction to that which pilots normally want to be lifted!); low level wind shear can make take-off and landing dangerous.  Pilots of light aircraft - and ligher-than-air aircraft - typically try to avoid flight into known conditions of wind shear.  Wind that abruptly changes its direction and magnitude is dangerous, unpredictable, and most importantly, invisible.
 * In fact, if you read the chapter on contour flying, you'll see how very very strongly the handbook emphasizes that it must be conducted safely and legally. "All aircraft should be operated so as to be safe, even in worst-case conditions. Every good pilot is always thinking “what if...,” and should operate accordingly."  What if flying into known wind-shear causes the balloon envelope to collapse?  What if the wind shear or turbulence inverts the aircraft and dumps the pilot or passenger out of the gondola?  If you can't guarantee control of the situation, you aren't making great aeronautical decisions.  It would be unwise to design or operate an aircraft whose entire principle of operation runs counter to well-established common-sense guidelines.  Nimur (talk) 21:46, 18 February 2015 (UTC)
 * I greatly appreciate your realistic pilot's common sense! All this is true.  I was thinking of this more theoretically (specifically, in terms of a simulated "first propelled aircraft" from a dreamy parallel universe)  After all, I was dreaming of a historical computer game, my parallel self riding in the swinging tail of the gondola pulling on the rigging to twist the louvers this way and that to put the tail's pull at an angle to the wind, the other member of the team sitting in the gondola controlling fins on the main body and running the tanks.  And not a very serious gamer either ... the dream started with him inserting in free-fall heading right at the spot to control the tail where I had set up, and me dodging and getting knocked out against one of the smaller square fabric louvers barely hanging on.  And later on he fouled it by somehow tilting too far out of horizontal and getting air backed up into the fermentation tanks, which did something to foul up the lift gas regeneration... it was a good dream :)  More seriously, in the days of computer control of things - including perhaps unmanned lighter than air vehicles - and in any case perhaps allowing better real-time control than a human can do, with better sensing of oncoming wind changes - many of these safety considerations might be substantially reduced. Wnt (talk) 00:53, 19 February 2015 (UTC)
 * I don't know much about balloons, but I've logged lots of hours with 1,2, and four-line kites. The single line fighter kites in particular are a fascinating example of control through dynamic instability. I see no reason why a balloon with a heavy gondola shouldn't be able to use two or more quad line kites to achieve some amount of tacking or otherwise upstream movement movement that is not the exact same as the average wind velocity for the local region. Think of the SkySails system, but with more kites and variable length lines. You'd hypothetically be able to tap into different airstreams, and also fly each kite near the edge of its Flight_envelope (see here for kite-specific discussion) to achieve various torques and forces, that could then interact with whatever foils are on the gondola and balloon. A sort of intermediate step between a fixed anchor and a free-floating kite/balloon would be kite surfing, and those guys can definitely do some weird things in the air. There's also some mildly related stuff at Kite_applications. SemanticMantis (talk) 15:18, 18 February 2015 (UTC)