Scenarios: Unrest: Home invasion Pt13: Physical protection: Doors

Doors:

Tips

General guides

General advice

  • Ideally for a fairly secure normal door you want:
    • single-leaf (for rigidity, failing which use rebated meeting stile or ideally fit astragals),
    • outward-opening (against ramming),
    • fabricated steel or least solid timber (avoid steel extrusions, aluminium, upvc and glass),
    • hardened attack face,
    • surface-mounted multipoint locks,
    • cylinder guard,
    • autodeadlocking but unlocking by key only,
    • electric locks are better than electric strikes but worse than mechanical ones as they can be magnetised, shocked or wiring hacked and solenoids can be forced back, mag locks are vulnerable to power cuts or batteries running out but as an emergency temporary lock could be acceptable if 24V on an outward opener,
    • long throw bolts,
    • tiny gap between frame and leaf,
    • keeps that are part of the jamb – not a strip that pops into it,
    • deep and wide jambs,
    • four sided frame with sill,
    • stiff internal frame to resist crowbars,
    • dog bolts (not hinge bolts),
    • security butt hinges,
    • extra long hinge screws,
    • no door viewer,
    • no glazing if you want to delay many minutes (failing which, high security windows with equivalent protection such as laminated glass or polycarbonate), and
    • as a minimum for any security, wood infills need to be 15mm and steel skins need to be 1.5mm.
  • Having no glazing in an external door may mean a dark hall, but you could replace an internal door with a glazed version to make up for it or fit a sun tunnel through the loft into the landing in a house.
Sun tunnel to avoid glass in front door
  • External doors should average 1.8W/M2.k insulation, which is why many security doors are filled with foam. Ways around this are explained in the Windows chapter. Its not a crime as long as you have steel skins and stiffeners, just stay away from plastic skins.
Foam is OK behind steel – not so good behind plastic
  • Have pre-cut boards on standby to barricade ground floor openings.

Materials

  • Choice of materials can improve delay for core as well as skins.

Pecking order of materials

  • In order of increasing strength of composite door materials, go for: polycarbonate, plywood, concrete, steel. Obviously for single material doors you use wood or steel as poly is transparent and meltable around the lock and concrete is too brittle.
  • If you were designing your own security door then in increasing order of strength you could use a steel-polycarbonate-steel sandwich, steel-plywood-steel sandwich, concrete filled hollow steel or either exmesh reinforced concrete filled hollow steel or a steel-foam-wood sandwich.
  • Home Office tests found that mortice lock security doors can be breached by two men with manual tools in six minutes if wood, and twelve minutes if hollow steel. Also that steel doors are more vulnerable to battering rams, but less so to hydraulics if multipoint as they tend to bend leaving other bolts engaged. Wood doors are sometimes better against spreaders as the frame splits leaving the hydraulics with no fulcrum. Disc cutters are the most devastating against all doors, in fact they are the only cutting tool that works well against steel hollow doors. Thermic lances work but make horrendous smoke when they reach insulation.

Plastic

  • Extrusions, upvc and glazing are best avoided; UPVC corner welds are a weak spot and the frame is too flexible and the profiles too brittle for normal locks to survive even small crowbars or screwdrivers.
  • You do not have eggbox doors anywhere at home, right?

Proteus

  • One promising material is Dr Stefan’s Szyniszewski’s ‘Proteus’ biomimetic sonic metamorphic architected MFCC (metallic foam ceramic composite) material, a highly deformable material weighing 1.14t/m3, made of 13mm ceramic ball grid being 14% of the volume, suspended in 2mm cell aluminium foam reinforced with 1mm diameter 100mm long nickel chromium fibres, which can be made as panels sandwiched between steel skins. It is resonated by the cutting tool to turn ceramic and the aluminium into abrasive powder which cut the blade and enjoys a strain rate effect like sandbags against bullets, which cannot cut the powder as it is bouncing around in foam. It is 15% the weight of steel. It vaporised grinder discs, stopped drills and water jet cutters cut through the foam but bounced off the ceramic balls which also widened the jet and reduced its speed 50 times before stopping it at 15mm depth. It also allows slightly lighter weight ballistic and blast panels than armoured steel. It is available for prototypes in 2022, expecting to go to market in 2023. However, where ballistic protection is not needed fabricators are likely to go for cheaper home made resin ceramic pours.
Anti-cut ceramic aluminium foam

Armoured steel

  • Ballistic steel armour weight grows exponentially as you move up the ballistic grades, eg upgrading from FB4 to FB6 more than doubles its weight, and upgrading from FB6 to FB7 more than doubles it again.

Glazing

  • Ballistic polycarbonate is lighter than the equivalent rating of steel, whilst polycarbonate glass is heavier, and laminated glass is even heavier and probably too thick for the door if above FB4. Normally door manufacturers cap laminated glass thickness at 45mm before switching to polycarbonate glass.
  • Blast glazing easily fits in a door even if it is laminated glass.
Blast door with vision panel

Steel sheets

  • As a stopgap insurers advise businesses to cover doors with 1.5mm steel sheet at 150mm centres with 6mm dome head coach bolts filed down or welded tight inside. This could work for a SHTF scenario where looks no longer matter. JSP440 recommends an 18 gauge steel facing screwed into the edges every 4″. Avoid the locksmith scams where they sell a 1.5mm sheet for £400 that is screwed instead of coach bolted and doesn’t even reach the edges where the attack will come.
£400 for about £40 worth of galvanised steel. Once the burglar chisels the leading edge at the locks it is all over. Nor will the screws add anything to the strength of the fire door against a crowbar. At best it gives a small delay for any thumbturns if there were thin panels underneath.

UPVC

  • UPVC panels can be upgraded to slightly more robust ‘security’ versions, but in reality they are sometimes just 4mm MDF hidden between sheets of polystyrene, whereas police advice on womens refuges is to to reinforce door panels with 12mm plywood and building regs require at least 9mm timber walls next to doors.
  • Sadly uPVC is too flexible to be secure.
A good boot and it is all over for cheap and nasty upvc doors.
How a UPVC door infill is only held by clip-in beads
How to remove the whole UPVC door and walk through it while it is still closed. P.S. gravity pulls the panel down into the bottom rail so there is nothing holding the top of the panel except a slither on clip-in bead.
See how flexible uPVC doors are – even with a girl wielding a glorified spatula

Wood

Materials
  • Solid wood doors can withstand a short attack on the wood; the problem comes with attachments like locks, hinges and glazing. There is simply not enough wood to hold the lock, hinges or glass once it is inserted. Wood resists stretching along the grain and being compressed, but it is not much use against splitting along the grain in tension. A 44mm door with a morticed 18mm lockcase has at best 13mm left on one side to hold it with all the energy concentrated in one spot.
  • Hollow or particleboard wooden doors are useless.
Hollow particleboard. Don’t do it.
ASL Titan Doors callouts
  • No softwood can be used anywhere in a wooden security door.
  • Applied wooden stops are vulnerable so rebates should be part of the frame.
  • Avoid eggbox doors!.
Eggbox door fail.
Thin wood is useless.
Capabilities
Performance
  • Wooden doors particularly benefit from being outward opening, especially against knife attacks against electric strike plates, except knife attacks on wood slabs where inward opening 44mm lasts 5 mins against basic knife attack compared to 2 mins outward opening, or electric shear locks which can last fives times longer against a knife when inward opening.
  • Wood doors at best only stop a handgun forced entry attack for 30s.
  • CPNI found the strongest non security door was 54mm solid wood inward opening with flush bolts top and bottom, which lasted 10 mins against a lone attacker with a knife.
  • Plywood can act as a good energy absorption layer and is used as wall panels with exmesh so the same can be done as a door insert. The exmesh slows torch cutting and the wood makes it even slower, hopefully forcing attackers to have to change to a grinder after wasting time with the torch.
Properties
  • If using wood structurally, its strength depends on density, ring width and grain direction. You need clear wood without knots, splits or cross grain checks. Wood is orthotropic so much better in one dimension than others, eg stronger along the grain (longitudinal) than perpendicular to it (tangential). The direction of growth (radial) is another dimension. Nine elasticity moduli determine wood’s elasticity so it cannot be treated as a homogenous material like metal. Overall, the strength of wood depends what direction you ask it to work in. Wood is not tested by stretching but rather by compressing or bending, so it is more a material to give rigidity and impact resistance as the core of a steel case than one that cannot be torn apart.Black cherry is stiffer than oak.
  • Hickory can be as tough as oak.
  • Oak’s tensile strength is about between 78-112MPa along the grain where it is strongest, and its compression strength is 18MPa perpendicular to that. Its rupture strength depends on moisture content.

Aluminium

  • Aluminium is too weak against manual attack without steel reinforcement so you usually might as well go with steel in the first place. It is also colder, much easier to fracture and much softer, although almost 40% as stiff as mild steel.
  • However, it can be handy as an anti cut layer as it tends to melt and clog up drill bits and conduct heat away from cutting torches.
Aluminium door’s keeps flew off without even breaking glass.

Hollow steel doors

Why use steel
  • You cannot just laminate plywood to steel and mortice locks into the edges, as they will just burst under attack, instead you need a steel frame to hold a core such as plywood which can then be skinned with steel, unless you rim mount the locks and dog bolts, but even then the edges are vulnerable to being chewed up by crowbars. To see why, just watch videos of residential security doors with multiple bolts being prised out of the keep.
Mild steel
  • Mild steel is hard, strong, tough and stiff, yet not so hard that it is brittle, and is ductile enough to soak up some of the energy rather than tear or rip out of the wall though excessive stiffness. Unfortunately once steel exceeds about 700MPa strength it tends to be brittle with low toughness and can fracture under plastic deformation or even long before that amount of energy just from an impact. Roughly speaking, by moving from mild to stainless steel you double strength and halve toughness, and by moving from stainless steel to tool steel you double strength again and halve toughness again. Tool steel is about 50% harder than mild or stainless steel so does bring an anti-drill benefit, whilst stainless steel costs a fortune but despite its strength does not mean you can just halve thickness as all three steels have about the same stiffness.
Machinability
  • One rough indicator of metal strength is its machinability rating. 1% carbon steel’s is 100. D2 tool steel’s is 45. The lower the better assuming the fabricator can make it. Another guide is carbon content. Mild steel has only about 0.15% carbon. High carbon steel goes up to 2% before it becomes cast iron. The less carbon the easier to bend, cut and weld, the more carbon the stronger. Stainless steel is obviously more corrosion resistant and is an iron alloy with at least 10.5% chromium; it is easier to bend and harder to cut than mild steel.
Alloys
  • You can, however, consider whether other materials with extreme properties can add value to a door, whether as a plate or core. Material properties to look for include hardness against cutting, stiffness against levering and impact, and high ultimate strength to delay fracture and high yield strength to delay plastic deformation before you get to the ductile stage to protect against hammering. You might also want to move (copper, beryllium or aluminium) or stop (concrete, insulation) heat in a strongroom door (eg for a bunker) to delay thermal cutting.
  • Most ‘miracle’ materials are not as great as they might first appear on paper. For example:
    • Molybdenum is stiffer than steel, but almost as heavy as lead and still shrinks sideways when stretched longways.
    • Titanium is light, but ruinously expensive and softer, stretchier and weaker than steel so no use to man or beast in security.
    • Tungsten is stronger, stiffer and harder than steel, but too expensive to justify the thinness gained.
    • Tungsten rhenium is outrageously expensive, and cold and heavy, albeit stiffer than tool steel.
    • Carbides are incredibly hard but, like concrete, too brittle to be used for anything other than a sacrificial energy absorption layer protected by a tough strong attack face.
    • Graphene is lighter, warmer, stiffer and 274 times stronger than mild steel, but we do not yet know how to make it in thick tough layers that will not fracture. It is being experimented with as a reinforcing fibre to sprinkle in other materials.
  • Natural materials, polymers and ceramics lack toughness and strength found in composites and metals.
  • Diamond is sometimes used as a coating on safe drill plates but is a bit dear for door sizes.
  • Price rules out many of the hardest or strongest metals, leaving steel as the go-to skin.
Ceramics
  • If it was only down to hardness and stiffness, technical ceramics would outperform metals and composites, although carbides could still help as a restrained energy absorbing core especially for ballistics.
Foam
  • Polyurethane foam is a cheap, warm, light, rigid material purely to insulate the core with. Although weak in tension, it can muster 20MPa elastic modulus in restrained compression, such as when sealed in a metal door. Insulating cores of fibreglass or polyurethane foam are economical.
Concrete
  • Concrete is 100 times more insulating than steel, although 12 times worse than fibreglass, but if you were building a strongroom door filled with 24″ thick concrete it will still be warm as the 2″ of fibreglass in a normal hollow steel door.
  • As concrete is brittle it needs holding together so delays are far higher from steel fibre reinforced concrete. It is normally 2% steel by weight.
Ballistic cores
  • As for security cores, Kevlar (poly para phenylene terephthalamide) is dear but is only needed in thin sheets and adds protection against bullets and cutting as well as effectively adding a dash of fibreglass insulation; it cannot normally strengthen the door mechanically, as if it is anchored in tension it will perform worse ballistically – it works best free to blow backwards using the 2″ or so width of the door. It costs about £300-1,500 for a door panel depending on rating and can alternatively be bought as a rigid glued panel that could work on its own or be glued or inserted in the core. FB4 can be achieved with 4 or 5mm.
  • Beyond FB4 you will probably have to go for Dyneema, Spectra or fibreglass resin panels (aka Bristol armour / green armour). Fibreglass panels will be from about 7mm-32mm for FB2-FB6, 12mm for FB4. Ballistic hardened steel will be 3-13mm thick depending what rating from FB4 to FB7.
Polyethylene
  • Another candidate for security cores is polyethylene slabs; they are cheap, relatively light and warm, and in door thickness slabs, whilst with their softness and stretchiness they can obviously be torn and cut, they can provide extra strength and bulk for saws and drills to have to chew though, especially when intruders forget to bring a wide kerf negative rake ripping blade as the friction from normal blades chop-welds it back together, and even if they use a non-melt blade it will not be much use on steel, so can make a useful addition to low security doors below SR4 where power tools are not expected. A standard door size slab of about 40mm x 750mm x 2000mm weighs about 57kg, although they are typically used in thinner smaller panels for vulnerable zones. You could also alternatively add a layer of its denser cousin, UHMWPE, which combines toughness with lower, but still useful, melting tendency to smash and trap bullets. Piles of UHMWPE sheets are used in flexible ballistic vests.
UHMWPE
MaterialPrice
USD/kg
Weight
t/m3
Thermal
W(m.K)
Hardness
HB
Toughness
MPa.m 1/2
Strength
MPa
Stiffness
GPa
Kevlar501.440.046.633,60084
Tungsten carbide30014.511025,0002.5370600
PH stainless steel97.7518400801,000200
D2 tool steel27.6720630421,860210
Tungsten rhenium alloy3,00019.7705002,100400
Boron carbide1602.523038GPa3.5500460
Tungsten11019.251702,5707980750
Inconel408.26.5330801,200200
Mo 25 re alloy3,00011.2703501,100360
TZM alloy10010.22126220800320
Beryllium copper138.251151501,110131
Graphene1002.274130,0001,000
Titanium6184.522150450116
Molybdenum4010.281381.9GPa30565329
Mild steel0.57.8550420200475200
Polycarbonate2.21.20.2802.2652.15
HDPE0.90.950.5402.6301
Fibreglass3200.04
Aluminium2.72351.33Gpa215575
Copper8.95401250MPa170117
Polyurethane foam2.40.0960.0240.6
Concrete0.072.40.56 Mohs0.4260
Oak0.70.7040.1741009510
Ferosafe950HRC1100.44
Steel security door materials properties
Designing your steel door
How strong to go
  • A good start in designing a security door is to work backwards from the design threat.
    • Even if you only fear a crowbar, lump hammer and hand axe, that is enough to threaten the door and its fixings, leaving the intruder to either climb through a hole in the door or remove it by disengaging bolts or tearing screws out. This suggests your biggest priorities are skin and internal frame thickness and strength and number of fixings such as lock bolts, hinge bolts, hinges, screw threads, welds and reinforcements. Yet time and time again certification testers find thick skinned doors fail due to quality of build, or lack of drill plates allowing a cutting attack succeeding on the lock without having to cut a man sized hole, or lack of adequate cylinder guard allowing the lock to be neutralised.
    • Thick skins are not much use until you have multipoint locks.
    • Hardened steel skins may be too brittle unless wrapping a plywood core. However they may be necessary if the main threat is ballistic.
    • For post SHTF scenarios the threats are less ballistic, blast, thermal and power tools and maybe not burning or cutting, and more hand tools, especially for banging and prying, so skin thickness and hardness and anti-cut or anti-torch core materials matter less, and a lot less than multipoint locking. Just about every test fail you see, even on higher ratings than a high risk residential one like SR3, involves crowbars or hammer and wedges eventually rotating and disengaging bolts and hinges by crushing edges and cutting multipoint rods: thick internal frame channels and anti drill plates are your friend – you want to force intruders to crush all four edges or cut a hole as a big as a man or as big as most locks or bolts or wall anchors.
    • A hollow steel door, for example, is not automatically ballistic, nor will it stop cutting or such easy tearing, but is light, warm, looks relatively normal especially if wood-skinned over the steel plate skins, can be bought cheaply from many suppliers depending how domestic it looks, can easily be blast resistant and you can hide extra layers inside for further protection or functionality, eg fibre glass against cutting or bullets, or concrete against all attacks, or wood or foam against thermal attack. The main reasons they are used, regardless of protection level, is that they look more normal, fabricators are used to them for non security products, and they can be filled with insulation to comply with building regulations.
Typical 1.5mm mild steel security door skins can be cut in a few minutes by hammering an axe at an angle with a lump hammer. In this test the Teckentrup unrated door beat a suspected fake SR2 next to it which was opened in seconds with screwdriver.
‘Security door’
Testing security door with bottom bolt not thrown. The doors, bolt and single point lock are stiff enough that it only needed the top bolt thrown, plus the single point without which disaster would have ensued.
How steel doors work
  • Plates & beams:
    • In engineering terms a door is either mainly a plate or mainly a beam, and attackers want to either treat it as a solid by rotate bolts or as a fluid by putting the door in sheer.
      • Few steel doors are pure plates, as most need stiffeners that make it more of a beam. Even safes, gun room doors and strongroom doors are partly beams as their thick typically 10mm plates are reinforced with steel angles. However skins are largely plates that experience bending until they deflect by their thickness and membrane tensile stress beyond the extent to which they are turned into an array of smaller plates each supported by reinforcements. Incidentally, metal merchants only call sheets ‘plate’ when they are over 6mm thick.
      • Most steel doors are beams. Hollow steel door frames are beams in their own right and the doors usually also have stiffeners, which are also beams, connecting the frame, making the door something akin to a honeycomb beam (with intracell buckling and face wrinkling problems). Others are sandwich beam (cored laminate composites whose stretching stiffness is just the sum of its layers but whose bending stiffness needs about 30 calculations). Looking down into a hollow door from the top, vertical stiffeners are transverse beams and horizontal stiffeners are longitudinal beams. However, steel doors as beams act like a plate in that they bend in two directions, and, except in blasts, energy is not applied anything like uniformly across the door. So the out of plane beam behaviour is not the main deformation consideration in security, rather what matters is resistance to edge deflection and tearing or cutting of skins. However, stiffeners can help prevent the door being squashed away from the frame in sheer or being pushed in to rotate the bolts out of their keeps. So the door is also a beam in plane and that dimension is thin and deep so should be capable of being engineered with massive resistance to having its edges crushed at the hinges and locks. The reason they do not work that well is that the edge reinforcements are channels for cost, weight and sometimes hardware fitting reasons. If they were solid (and so was the frame), then, assuming the wall has enough bulk to it, intruders are realistically going to have to cut or smash through the door, its fixings or the wall. However, even safes only use thicker steel round the edges but not solid frames backed up by stiffeners that could stop frames being bent. Ultimately a door is not really a beam overall as it is oriented 90% in the wrong direction and even then it would still not have most mass near the ends. Door aspect ratios in plan are always dreadful, although strongroom doors get closer to 1:1.
      • To stop both the rotation and sheer attacks needs the same solution – to make the door a clamped beam rather than a simply supported rectangular slab, by somehow grabbing bolts on as many edges as possible to stop them sliding out as edges are crushed or the whole door is punched in. The closest thing to that the author is aware of is the Surelock McGill Moorgate, invented as a tube station state scissor gate hook lock, it not only boasts two opposite direction hook bolts but can drive extra bolts sideways up and down the locking edge and up and down into the head and sill, plus it is has achieved SR3 for a scissor gate, meaning a jack, wedges and crowbar failed to wrench it out in plane from the keep with nothing else to hold it, and that was just the two hook bolts. The door would then be highly pinned if not technically fixed or clamped, assuming you combine something like a piano hinge or incredibly heavy duty hinges and hook hinge bolts to lock down the hinge stile enough to make up up for the lack of hook bolts along the hinge side of the head and sill.
      • A thick plate is strong against blast, bullets and manual attack, especially in protecting fittings on the inside such as locks from cutting attacks such as drilling, but is cold, heavy, does not look domestic, is a giveaway that you are protecting something, may still need pockets for locks and a frame for glazing, and due to lack of supply it will have to be custom made. Unsurprisingly they are only really found on bunkers where the disadvantages do not matter.
Hollon gun safe showing door frame. A plate is never just a plate.
Surelock McGill Moorgate hook lock to SR3
Thickness
  • The most efficient thicknesses are hollow steel doors with three point locking, or plywood core laminated with hardened steel. This plywood/steel laminate is particularly good against explosives.
  • Laminations:
    • Laminated door slab thickness increase gives more delay than hollow door skin thickness increase.
    • Increasing number of plys in a laminated slab increases delay significantly until there are five plys when diminishing returns kick in dramatically.
    • Most of the more efficient designs laminate a soft core with hard skin. However at the highest threat levels, concrete core can roughly increase delay 50% as an alternative to relatively soft foam and/or oak, and against battery power tools it can double delay, or quadruple delay if reinforced with exmesh. Doubling the polycarbonate in a laminate can double delay.
Concrete core vault door manufacture showing concrete pour. Unlikely you would use this design thought given the mistakes.
  • Skins:
    • A skin can be too hard in that it tends to come with brittleness, for example armoured steel is best used as anti drill plates over locks rather than as a skin unless you are happy to line it with mild steel or mainly fear ballistic attack. Normally the mild steel goes outside.
    • Fitters often do not like working with metal they cannot drill on site, and not many fabricators are geared up for unworkable metals, for example, how will it be bent and welded?
    • Steel doors should be galvanised and at least 14 gauge.
    • To avoid welds eventually showing through paint, specify a wood skin.
    • Avoid cheap Chinese fake security doors made of bacofoil-skinned eggboxes.
Kid wrecks ‘security door’ with tin opener
Yes folks, you can buy a ‘security door’ for $40, complete with 0.4mm steel facings filled with paper honeycomb
Why you do not want a hardened steel structure.
  • Battery tools:
    • Against battery power tools, or even some heavy hand tools, door thickness has to increase to allow a gap (which is sometimes filled with insulation) to stop both skins being damaged at once.
    • Unrated security doors typically last 1 to 4 minutes against heavy hand tools or battery tools, depending whether made of 1.5mm or 2mm steel plate and whether hinges and gaps are protected.
    • With power tools a crawl size hole can be cut in a steel skinned door in 3 minutes, and three hinges can be cut in 1 minute with hand tools, or with power tools if hinge pins are welded.
    • To slow down power tools you may have to use a vault door or a series of security doors.
    • Skin thickness typically needs to nearly double from some 1.5mm to 2.5mm if battery power tools are a threat as otherwise delays massively shrinks by 80%.
    • Doors may have to be upgraded to resist modern cheap 36V battery tools, with materials such as manganese steel and ceramic (for hardness) and fibres (such as Kevlar, to clog up blades); this is the focus of a CPNI project, FARM, which initially pinned hopes on resin-suspended ceramic baking beans for grinders (as ceramic wears discs where fibres do not; a bit like Proteus except with resin instead of aluminium foam), and loose Kevlar fibres for drills.
  • Thermal tools:
    • Against thermal tools, a core of fireproofing or insulation is also needed.
    • Against the worst tools, hardened steel-skinned wood core is needed, possibly also with insulation, and no external keyway is allowed.
    • However, by the time you face threats which warrant specialist doors with fireproofing, insulation and wood cores, the alternatives are to either a) use multiple mass market hollow steel doors with deep gaps between skins to slow down cutting or b) buy a vault door, but again with no external keyways, so someone has to stay indoors to let you in unless the door is for a panic room that is usually unlocked.
Thermal resistant plate stops thermic lance
  • Hydraulics:
    • Something to beware of, even for steel outward opening security doors, is that an organised gang might use commercial hydraulics to come up with their own version of the likes of Libervit Blackline gear, such as their top of the range battery powered 7t-rated VE70 & Option T (sold as DoorRaider to the tactical community), which can rip out an RC3 security door with aluminium skin or 0.75 steel skin with morticed locking along the locking stile, but maybe not an LPS1175 door with rim bolts and which will usually have at least a 1mm steel skin.
    • For exceptional risks you should ensure a hard smooth floor surface outside the door that will not grip metal studs.
    • See the post on MoE kit.
  • Weight:
    • As a rough guide to what is in LPS1175 doors we know that a normal door size weighs about 120kg per centimetre thickness of steel, whether it is all plate or partly beam in the form of stiffeners and internal frame.
    • For example, Sunray’s Excludor 6 is the joint strongest LPS1175 door available, with R&D’s Eco VI; it is as heavy as 11mm steel plate, and they only admit to 1.5mm skins, so there is something as heavy as 8mm steel in the 42mm hollow to cope with 18V power tools, 3′ crowbar and steel wedges and sledgehammer. A tool steel plate backed by concrete would weigh the missing weight and one assumes would stop someone climbing through within 10 minutes with only 18V batteries in their saws.
  • Plates: Thickness is everything with door skins as far as loads are concerned.
    • The caveat with any formula estimating the strength of a steel door is that they assume infinite values at connections which are actually slightly wobbly or stretchy (hinges are not clamps), doors are hybrids beams and plates (with plates within plates and beams within beams), and load from an intruder will be repeated, impact, concentrated and eccentric as opposed to one-off, without shock, uniformly spread or central.
    • Concentrated force in centre example: One experiment by a security door manufacturer found that, assuming no stiffeners, doubling galvanised steel from 1 to 2mm reduced centre deflection from a centre load ten times, and thickening steel from 1.2mm to 2mm reduces deflection nearly five times and roughly doubles stiffness, whilst thickening from 1.5mm to 2mm more than halves deflection and more than doubles stiffness. A rule of thumb is steel strength against bending stress increases with thickness by square law and stiffness against bending increases by cube law. See Metador experiments.
    • Simply supported load:
      • Uniform load
        • Using this formula assumes bolts have clearance to absorb deflection before they hit their keeps and become pins or clamps. In reality hinges mess up the equation for one edge which is a pinned edge until they fail, and in reality hinges eventually rip off, often first displacing and deforming with their attachment elastically until they deform plastically.
        • Applying a uniform load over a simply supported plate door deflects it in the middle depending mainly on width, aspect ratio and load to the power of 4, and thickness and Youngs modulus cubed. On thicker plates the deflection problem is mainly width to the power of 4, but is also affected by Youngs modulus, thickness and aspect ratio cubed. As you cannot very well change door size, your variables force you to seek the best combination of Youngs modulus and thickness. Thickness is weight and cost but can provide other advantages, so you probably want to go for the thinnest steel that covers other requirements such as drill time and make up for it with Youngs modulus.
        • A door twice as tall as it is wide is stressed slightly more than a square would be. On thicker plates the stress problem is more aspect ratio than width and thickness, but, as you cannot really change door dimensions, all that is left to influence is plate thickness and you need a fair bit of thickness to make up for aspect ratio.
        • A door twice as tall as it is wide acts as if it is under four times the load of a square. This means a door is pretty much always a deflection problem needing stiffeners.
        • A door twice as tall as it is wide experiences force at the middle of the long edge 20% higher than a square, but this effect falls away exponentially with aspect ratio so is not a reason to worry about tall narrow doors. However, you probably want to create squares with stiffeners to chase that law of physics within a hollow door as a honeycomb beam to the extent it is possible given that they are unfortunately still having to be attached to long edge channels and the rectangular door plate and while the door is still ‘platey’ in nature, plus add a vertical stiffener to address the door’s aspect ratio.
        • Stiffness is mainly affected by Youngs modulus and thickness cubed but also Poisson ratio squared. Poisson ratio is how much a material narrows when stretched. Unfortunately steel’s ratio is only about 0.3 but anything better tends to be too soft or brittle like cork, iron, boron carbide, concrete or glass. So this is another factor that is hard to change.
        • Anyway, as you are stuck with a door size and whatever load an attacker chooses, all you can control is thickness and Youngs modulus, so you basically want a sheet of sufficiently thick and stiff steel to stop bolts disengaging. If deflection was your only problem you would just go for the cheapest combination of thickness or stiffness. In reality you have to worry about hardness too, but manufacturers tend to hide the hardness plate behind the skin to protect it as the cheapest hard metals are brittle. The other way you can cheat is to choose stiffeners almost entirely for their stiffness to effectively increase the thickness and stiffness of the plate. That is exactly what hollow steel security door manufacturers do: morph a plate into a beam, leaving it with weak ‘beaminess’ but strong ‘plateness’.
        • The uniform load deflection formula is a very rough way to gauge strength for a mob attack with people crashing into the door, and tells us something about how much bolts would be rotated against their keep, but does not tell us how far they will be drawn back out of the keep, which stiffeners are not so good at preventing as they can be caved in in the centre of the door more easily than at the edge which is forced to come with the centre to the extent the door is not a completely elastic membrane.
      • Circular load
        • If the force is applied in the middle, as with perhaps a ram, then stress becomes more of a thickness problem, as the main factor is thickness squared, although Poisson ratio, width and aspect ratio still matter. A door that is twice as tall as it is wide is more than twice as bad for stress as a square. As you are probably stuck using steel and a normal door size, again you can only really vary thickness.
        • Deflection in the centre depends on Youngs modulus and thickness cubed and width and aspect ratio squared. Aspect ratio is adjusted by a factor which says that a door that is twice as tall as it is wide is 40% worse for deflection than a square. Again, you are probably stuck with the Youngs modulus of steel and the door dimensions, so all that is left to do is choose thickness, but at least the material choice can have a big effect if you want.
        • This tells us that thickness is our main tool in the battle against deflection from point loads like rams, with Youngs modulus the runner up.
      • Overall, the priorities in descending order for simply supported loads are thickness, Youngs modulus and Poisson ratio.
    • Pinned edges
      • Uniform load
        • Clamping edges is not great for middle edge stress, as now the door is not just holding the weight of the attack but is not allowed to bend and instead is asked to try and deform instead. It depends on width and thickness squared, so at least you can influence thickness, but unfortunately the biggest influence by far is aspect ratio, to the power of 6. This means you need a lot of thickness to make up for a door being a big clamped rectangle. You can imagine a trampoline made of steel: with no bounce to speak of the stress goes pretty much straight into the springs and wants to tear out of the steel ‘fabric’ and frame, especially at the middle of any long edge – just where you do not want stress on connections like locks, bolts and keeps and their fixings.
        • As for deflection in a clamped plate under uniform load, it depends on Youngs modulus and thickness cubed (which you can choose but it will not make the biggest difference in the result), but more so on width to the power of 4 (which you cannot do much about), and mostly on aspect ratio to the power of 5 (which you cannot do much about).
        • Unfortunately this means that in a clamped plate more and more thickness and more and more Youngs modulus has to be piled in to make up for door size and shape. It is mainly thickness that will help.
      • Centre spot load
        • Suddenly things are different for a pinned door if a load is applied on a spot in the centre, as the aspect ratio reduces stress there. A door that is twice as tall as it is wide suffers nearly 14 times less stress in the centre when loaded there than a square. It then benefits (or suffers) from the square law with thickness.
        • Deflection, on the other hand, is worse with aspect ratio, which is what matters unless you are coming up to the stress capacity of your plate. A pinned door suffering a point load in the centre that is twice as tall as it is wide is 30% worse for deflection than a square, but you are unlikely to be able to change the shape. And width squared is a bigger factor, although you cannot change size either. What you can change is Youngs modulus and thickness cubed, which are the main factors.
        • So pinned edge centre spot load is mainly addressed through thickness and to a lesser extent Youngs modulus.
      • Overall, the priorities in descending order for pinned loads are thickness and Youngs modulus.
  • Hinges
    • Uniform load
      • If you treat the hinge side as a fixed long edge, then the factors determining stress are width multiplied by aspect ratio adjustment squared, divided by thickness squared. So thickness makes a big difference and aspect ratio adjustment means that for a door that is twice as tall as it is wide stress is 46% worse than a square, but fortunately already only half as important in the first place due to the fixed edge, so the hinges make up for the aspect ratio.
      • In this scenario Youngs modulus and thickness cubed are factors in deflection, but the bigger influences are an aspect ratio adjustment and width to the power of 4. The aspect ratio adjustment says it is 80% worse for deflection to have a door twice as tall as it is wide, but fortunately the adjustment has a reduction built in which still reduces effective width by about 95% where there is a fixed long edge. The hinged side would have improved a square by about a third but the 2:1 door shape is improved by about half. So it ends up that size and shape is not so important after all. There is a downside, which is that the deflection has to try and pull the bolts out on the lock side more instead.
  • Poisson ratio, the tendency to narrow as a material stretches, is a relatively minor factor, especially as security doors tend to be neither entirely pinned nor entirely simply supported as they have pinned hinges and slightly loose bolts, although it is strill relevant if you fear hinges coming off and bolts working loose in an attack. Your main concern however is thickness, then Youngs modulus, which is the force required to stretch. If using steel then Youngs modulus is about 200GPa and your only choice is thickness. Carbides are one of the few candidate materials that are much stiffer against stretching.
  • Steel will stretch out of plane when pulled or pushed and will stress when it happens, and will also narrow unless pinned with hinges and something like hook bolts, but should not break. The challenge is to make it thick enough not to narrow or stretch all the way through the opening, pulling the bolts and maybe even keeps with it.
  • Beams:
    • The stress in a beam depends on length, but mostly on depth and width squared.
    • The deflection in a beam depends on Youngs modulus, moment of inertia and length. The other problem is that for uniform loads (like a theoretical well aimed air burst almost is), beams (eg stiffeners) stiffeners suffer from their strength depending on 5 x length cubed (so exponentially worsens with length) as well as 384 x Youngs modulus multiplied by moment of inertia. A centre spot load deflects 60% more than a uniform load. With the door being a fixed size all you can do is seek stiff materials with massive end-heavy cross sections like I-beams or hat channels.
    • The moment of inertia of a hollow rectangle tube like some stiffeners, is outside width multiplied by outside depth cubed minus inside width multiplied by inside depth cubed divided by 12. The main result of a door being a horribly shallow beam is it is 90 degrees at the wrong angle for a beam; as a hollow rectangle with the long side instead of the short side facing the load its depth is factored in to the power of 1 instead of cubed. This is why you are better off with more narrow stiffeners than fewer wider square profiles ones.
Fittings
  • On a hollow steel door for residential risks, delay mainly comes from astragal dog bolts and drill plates, and to a lesser extent frame thickness, rather than skin thickness, as attackers do not go the long way of disengaging bolts by pressing the whole door.
  • Frames are usually hollow steel but are better if filled with hardwood or concrete.
  • Remember that most security door skins are just under the 1.7mm thickness expected in a heavy traffic non-security door, so they are not strongroom doors by any stretch of the imagination.
  • Leaf frames are made of steel channels thicker than the skins to provide stiffness and a modicum of protection to locks and hinges, although inserted sleeves (matching the channel profile), or strips at least 2mm thick and 32mm wide, are often used to reinforce fixings like hinges.
  • Extra thread depth can be provided by extruding a hole to double its depth or through bolts with spacers or sex bolts can be used instead of reinforcing.
  • Pilot holes should be two thirds the width of the screw ready for tapping.
  • A 1.5mm steel skin hollow door with single point mortice lock can buy 3 mins against a basic knife & fire extinguisher attack, 30s against moderate EFES (explosive) attack or low BFES attack (softened up with handgun). Upgrading to multipoint rim mounted locks has the advantage of 3 mins against low BFES or 2 mins moderate shotgun BFES or 1 min moderate rifle BFES.
  • Single point mortice locks gain little from thicker steel door skin. Three point locking can roughly quadruple delay.
  • Locks should benefit from custom cutouts and pockets for the chosen lock.
  • Even SR4 doors are in trouble once about 4t is pressing on the lock, so multipoints become necessary.
  • Hinges would ideally be crane hinges to eliminate the gap between door and frame built into a hinge with a pin.
Armour
  • The inside of low security hollow steel doors is typically consumed by steel spacers and fibreglass or Celotex insulation. As security ratings increase, armour is inserted as anti drill plates, infills for vulnerable zones or whole door size panels.
  • SR3:
    • If building your own door then if aiming at rough equivalence to SR3, as explained below, you ideally want to grout the frame against jacks and wedges, use high hardness steel door skins or inserts against wedges, hammers, shears, drills and saws, and use multipoint rim locks against crowbars.
    • Most fabricators however get away with only using high hardness steel for anti drill plates and mild steel or at best stainless steel, for the skin.
    • But you could address a wider range of threats using LDPE (low density polyethylene, basically a chopping board) to glue up saws, or a fabricator may go as far as extra steel behind the skin.
  • SR4+:
    • To reach SR4 for a door in reinforced masonry you would need thicker skins and harder inserts and maybe an internal panel to chew up drill bits and saw blades like Avertic Armour, 6mm of manganese steel or D2 tool steel, or a new graphene reinforced tungsten carbide from Diamorph, Tenmat Ferosafe, available in 4mm thick 20cm square tiles, or just use 10mm steel plate. Hiplok D1000 wraps it round aluminium and takes 2.5 discs to get through with an angle grinder and holds Sold Secure Motorcycle Diamond.
    • You can glue on Swedish firm Engtex’s Avertic Armour inside the skin (anti cut forestry protection resold by Burton Safes).
    • Extra steel and maybe also Avertic sheet is needed against drills for solenoid locks to protect cables, and for panic bars as even two tiny holes allows a fishing attack. Avertic weighs only 1.9-3.1kg/m2, is 5-10mm thick and is available at levels Shield (drills), level 1 against drills & chainsaws, Defense (drill, grinder, chainsaw), Armour (drill, grinder, chainsaw & cutoff saw), level 2 against grinders, level 3 against power saws and level 4 / Homeland / Secured against recips.
    • You can also fill metal frames with concrete. This are much stronger against impact when specified as aggregate Portland cement reinforced with 0.3-1mm thick 1-8cm long steel or polypropylene fibres and rebar mats of up to 16mm bars, and against torch cutting when dotted with fused and sintered aluminium oxide in 25-50mm sizes.
    • Water-filled ceramic has been used as insulation against torch cutting.
    • Any space not consumer by armour or insulation can be filled with redwood to slow down cutting. Plywood can also slow torch cutting as an insulator.
    • 9 gauge 3/4″ exmesh is proven to slow torch cutting dramatically as much as 3″ steel plate as the heat gets wasted through the holes. It can also delay grinders and boltcutters and provides linear delay increase as you thicken it by layering it. With 12 layers it estimated it could be equivalent to LPS1175 H20 (the toughest pedestrian door is rated E10 & G5 as at 2022). You would still need to stop edges being crushed, bolts cut or locks drilled with such a design, so in reality you would probably use less exmesh and add a drill plate and use thicker mortared multiple rebated frames instead.
    • For ATM safes body strength is assumed if steel plate is 12mm and 690MPA tensile strength or 24mm and 345MPa.
    • In a paper, Development of New Composite Materials for Secure Storage Units to Increase Burglary Resistance, Roebbecke reported:
      • Grinders:
        • Steel and copper were cut quickly by grinders without wearing the blade as much, boron carbide resin was not much better and tool steel was outperformed by several other steels including, in increasing order of cut resistance, hardened tool steel (HRC65), low carbon steel (370MPa HRB70), 1045 carbon steel (530MPa HRB90), white cast iron (HRC55), A514 alloy steel (690MPa HRC20), Hardalloy 1450/Si3N4 composite and cemented tungsten carbide.
        • The best material for wearing down blades were, in decreasing order of effectiveness, cemented tungsten carbide 20% cobalt, Si3N4 AS10 and SiAION AS500, outperforming metal like stainless steel over 50 times in blade wear / material removed ratio. However, Si3N4 AS10 and SiAION AS500 are too brittle due to microcracks under thermal stress caused by low conductivity, but still did better than, say, boron carbide, Cermet insert PV3010 or cemented tungsten K10 if used as a composite with Hardalloy 140. Boron carbide resin suffers high material loss from cuts but does destroy blades in the process and is light.
        • Vapour deposited carbon comes out on top comes out on top if the preference is a balance of hardness-stiffness ratio and resistance to thermal stress, with silicon carbide fibre matrix composite the runner up at USD5,500/kg. Boron carbide, aluminas, silicon carbides, silicon nitrides (like SiAION), cemented tungsten carbide and tool steel do not like thermal stress, but this can be solved by using fine particles.
      • Thermic lance
        • 110 copper 6.3mm prevented cutting for at least 34s (especially where an amateur maxes out oxygen pressure) as it has density of about 9t/m3 conducted heat too fast with conductivity of 390W, has specific heat capacity of 385J/(kg*K) and has low heat of combustion at only 2.45MJ/kg, and at 3mm it is still six times slower to cut than steel; it is some ten times dearer than steel though. Graphite, low expansion aluminium oxide, boron carbide phenol formaldehyde resin and cemented tungsten carbide were almost as good, with high melting temperature but not the highest thermal effusivity (conduction, density & specific heat capacity), which at 7mm thickness would take about 20 minutes to cut a 10″ square hole to try and climb through; in comparison stainless steel 12mm thick would last 3 minutes and A514 alloy steel would need to 23mm thick to match that. Graphite and low expansion aluminium oxide get too hot to stay strong so may be too weak for a strongroom door, but boron carbide resin and cemented tungsten carbide maintain strength.
        • Overall, cemented tungsten carbide looks like the best candidate to single-handedly resist abrasive and thermal cutting, with an additional layer of copper being your best bet against thermic lances.
      • Hammer
        • To soak up blows from a hammer requires high yield strength and strain. Concrete is hopeless, plastic and elastomers are too weak, fibres, cemented tungsten carbide, reinforced phenolic resin and technical ceramics do not like strain, aluminium alloys and copper are not too bad, nickel alloys are slightly better but a mixed bag with some suffering poor strain, titanium alloys and tool steel does well on strength but is not the best on strain, and low alloy steel, carbon steel and stainless steel are the best all rounders, although with stainless steel you have to specify the right type if you want to beat its nickel, carbon and titanium cousins. Another factor, though, is toughness, and titanium alloy perform consistently well here, while nickel alloy can be tougher but can also be weaker in yield strength and low alloy steel can be stronger but not as tough and carbon steel does slightly worse than low alloy steel, whereas stainless steel is a fairer compromise between strength and toughness although not as consistent as titanium alloy, copper is as tough as steel but lacks strength, tool steel is stronger than stainless steel but less tough, cemented tungsten carbide beats even tool steel for strength but lags behind cast iron, phenolic resin and aluminium alloys on toughness. Concrete, elastomers and plastics lack strength and toughness, the best technical ceramics can outdo cast iron on strength but are worse than tool steel on toughness, and fibres are the strongest option but are only as tough as plastic. It was found that 6mm construction steel could be enough to see off the sort of hammering involved in testing safes that use high tensile steel, although 12mm is advised and you also want to stop edges being crushed.
      • The best combination for a strongroom door looks to be at least 6mm stainless steel attack face against impact, protecting an anti-cut layer of cemented tungsten carbide, with at least 6mm copper behind that if thermic lances are feared, and a stainless steel protected faced tied to the attack face would be needed to hold it all together under attack.
      • Composites:
        • The report author suggests some composites:
          • Although copper cannot be asked to double up roles, the steel attack face could double up as anti-cut by inserting technical ceramic discs such as 10mm diameter 3mm thick silicon nitride with 20mm spacing into low alloy steel 5160, in which case the two layers can be brazed together with 0.12mm silver. The risks with that configuration are cutting around the discs, and that using multiple offset layers leaves thinner layers for drilling and a weakened steel plate.
          • Boron carbide particles 4mm and fibres (eg Toyobo Zylon-AS polyethylene as fibres and fabric) could be set in ground phenolic resin of formaldehyde novolac with hexamethylenetetramine, with a mix of 17% aqueous resin (for wetting the boron and fabric), 48% phenolic resin, 2% fabric, 35% boron powder. It weighs 1.6t/m3, with a Youngs modulus of about 6.5GPa, flexural strength of about 50MPa, costs three times as much as steel for a given thermal resistance, and 5mm of this composite is as hard to cut as 25mm 345MPa HRB95 tool steel. PVB can be added for ductility.
          • Cemented tungsten carbide of 1.4-2um powder sintered to cobalt matrix (for extra toughness, softness and stiffness) 6.5mm thick sandwiched by 1.5mm brazed high strength steel is another composite. The cemented tungsten carbide layer is 10-100 times better against cutting than steel but costs about £1M/m3.
          • TRIP or TRIPLEX steel is another possibility. Transformation induced plasticity steel is hard yet tough with about 700MPa tensile strength and so could handle abrasive cutting and impact, whilst thermal cutting presumably could be dealt with by a copper backing.
  • Other anti cut materials:
    • Material to counter drills, grinders and recips include stainless steel (for low threat eg SR3), D2 tool steel or the slightly cheaper and about 24% softer manganese steel (for higher threat eg SR4).
    • D2 costs £500 for a complete 4.5mm anti drill door liner compared to £35 for a 2mm galvanised door skin or £200 for a stainless steel door skin. Anti-drill steel such as D2 to steel is hard and thus brittle so needs to be protected by a softer door skin, eg S275 mild steel or pregalvanised mild steel which is even softer. Even 4mm K700 manganese steel can have a 10mm hole put through it by a 12V drill in 95s whilst a 10.8V L-ion drill takes 138s, in fact it is not much better than stainless steel.
    • Other options are glass fibre, Kevlar or coiled steel fibres to entangle drills, ceramic to blunt grinders (eg baking beans) and hard/soft laminates such as rubber & MDF to clog recips.
    • New technology may involve materials that melt onto blades or move with them (like rollers in BS3621 lock bolts).
  • Eventually a pneumatic drill and thermal lance will come through even a vault door.
  • Strengthening steel
    • Manganese steel gets harder when hot so is harder to drill the deeper you go, especially if an amateur puts their drill on high speed.
    • You could slip in a sheet of rubber to choke an intruder on smoke as they drill.
    • Another trick is a brass plate behind hard steel to melt onto a heated up bit and clog it and snap it.
    • Tungsten-alloyed steel or cast iron is hard to cut with a torch.
    • Extra boron or cobalt in steel adds hardness.
    • Extra carbon, nickel or manganese in steel adds strength.
    • Extra nickel or vanadium in steel increases toughness.
    • If the door is to match a strongroom wall such as in a bunker or panic room, to something like at least SR7-SR8, then metals may have to be treated such as:
      • Coating containing carbide chips to smash or dull drill tips.
      • Diamond dust encrusted plates.
      • Ball bearing plates, but even ball bearings can be dealt with by patience, something like the old Strongarm Ballbuster carbide bit and cellulose tamping, getting lucky between them, or making a hole bigger than the ball bearing.
      • Another possibility is to make a plate similar to Lips’ Ellox (of 1980s Chubb safe fame) by suspending alumina nuggets in elastomer.
      • Have your own equivalent of Maxoloy made from carbide chips suspended in copper alloy, or Aloxite (aluminium oxide chips).
      • Embed angled mild steel plates in concrete to clog up diamond core drills.
      • Have your own Adamantium made with iron shot rotating in concrete that left room around the iron when it cured.
Ceramic baking beans poured in a hollow door can cause havoc for blades
Liberty Safe ball bearing plate
Stiffeners
  • Stiffeners should be at least 1mm thick and either spot welded every 3″ or continuously welded, and on 4″ centres.
  • Occasionally you find ribbons (long narrow plates) used, eg the old Skydas design used by Fort and Munitus, but normally ‘hat’ channels (top hat shape beams) are used, interlaced between the two skins or as two layers meeting in the middle; less commonly you see C channels or L angles, or increasingly strips of steel plate are used as webs which using the skins as flanges turning the door into a row of I-beams. Alternatively square or round beams (tubes) are used, in which case the strength comes mainly from profile width rather than wall thickness.
  • A door needs most reinforcement vertically, as that is the longest span for a rectangle of uniform thickness, and it is weakest in the middle, so needs thicker stiffeners in the middle and the others can thin out as they near the edges.
  • They cannot be made deeper – which you would normally do in engineering – as you already want them as deep as the door’s thickness, since strength increases as the square of depth, but only linearly with thickness.
  • The problem, though, is that vertical stiffeners let the door twist, and most or all bolts are on the vertical stiles, so the biggest risk is from the door bending vertically. So you need horizontal stiffeners as well.
  • Consider that blast doors sometimes use honeycomb core – steel stiffeners are plenty strong enough to take the energy of someone running at the door, rather the risks come from energy density in the form of a sledgehammer to the skin or crowbar to the edge.

Location

  • Minimise the number of exterior doors. Ideally only have one, in fact ideally have no external keyways and always have someone indoors to let you in. Failing that, have only one door unlockable from outside, with any others having a flush face.
  • Sometimes building regs, fire safety, disability, occupiers liability, mean you can only achieve security by using a lobby of two slightly too weak doors.
  • Avoid homes with set back doors that can hide burglars in a porch, unless you are sure it is necessary to prevent levering or vehicle ramming.

Fixings

  • Fixings should be M10 bolts (ideally epoxyed eg Hilti) if you want openings to have any chance of being as secure as the wall. Avoid the usual nylon rawlplugs that contractors may try to use.
  • Bear in mind that certification tests bolt the doors to RSJs so they never fail by the wall failing.

Threats

Manual attack

Standards

Housebuilder standards

  • In addition to the standards covered below, there is BS8220-1:2000 for domestic security and BS8220-2:1995 for commercial. As a guide, BS8220-1 only asks for one BS3621 lock and any cylinders to BS EN1303:2005 grade 5 key security and grade 2 attack security.
  • Aside from security standards for doors and windows, UPVC products should be to BS7412 and BS12608. Aluminium profiles should be 1.2mm thick and to BS EN12020 and BS4873 using temper T5 or T6 and alloys 6060 or 6063. Steel profiles should be to BS6510.
  • To cut a long story short, the standards PAS, SBD, low grade STS and EN1627 RC1-4 are hopeless for WROL scenarios with mobs, weapons and no police, whereas RC5-6, DOS SD-STD-01.01 for (brute force mob attacks) or LPS 1175 ratings can potentially see off a noisy tooled up gang attack, including trying to bust through glazing (there is also the USA’s HPW TP-0500.03 ‘Prolonged’ for 3 hour glazing resistance). The gold standard though is to be found in the counter terror ratings from CPNI, although for example that can be met at Base level against forced entry by the right construction of materials such as 12mm polycarbonate glass laminate eg Tecdur. Mob attacks can also be mitigated by the USA’s ASTM F3038, which specifies protection for 15 or 60 minutes against multiple attackers, eg 1.75″ thick AssaAbloy Forced Entry doors (also rated to STD-01.01), and is basically the commercial non-ballistic equivalent of government SD-STD-01.01 standard.
  • Buildings regulations treat PAS24, STS 201, STS 202 level 2, LPS 1175 SR2 and LPS 2081 SRB as equivalent. These are all stealth attack standards with small quiet tools without breaking glass, except LPS 1175 SR2 – which gives three minutes against bigger noisy tools (like hammer, hand drill, screwdriver and hacksaw) and allows testers to break glass.

Insurance standards

  • LPS1175 approved installers have to be certified to LPS1271 which requires them to register your door or shutter with LPCB. Of the manufacturers of LPS1175 security doors that might suit a home only Ascot and Sunray have LPS 1271.
  • Look for LPS 1175 or EN1627 rated doors to match your wall strength, usually SR2, SR3 RC5 or RC6, although RC4 upwards assumes dense 24cm thick masonry with compressive strength of 12MPa, and RC5 requires compressive strength of 20MPa and density of 1.8t/m3 (typical brick density); unfortunately housebrick compressive strength is more like 7MPa although can be up to 14MPa (3,000PSI concrete is 20MPa). Even more worrying, housebricks tends to have tensile strength of only about 0.38MPa (3,000PSI concrete is 2MPa) so fixings can rip out with ease. Occasionally these doors will also have PAS, STS or SBD, which they should far exceed, but are not normally aimed at the housebuilder market so it is common for manufacturers not to bother with those entry level certificates.
  • The accompanying standard to EN1627 for security glazing is EN356, rated from P1A to P8B. Glazing up to P5A weighs 16-22kg/m2 and is safety glass to stop steel balls coming through accidentally; it only needs glass a maximum of 10.3mm thick so is not even anti bandit glass. P6B to P8B needs laminated glass 15-26mm thick weighing 23-30kg/m2 or polycarbonate laminated glass 10-18mm thick. This glazing standard EN356 is even worse than the barrier standard EN1627. You probably want to go for LPS1270 security glazing which is rated to match LPS1175 SR levels and measures how many minutes it will stop a finger hole, hand hole and body hole.
  • You may as well know how big and ugly an SR3 door is, so here are the tests:
Teckentrup censored the couponing with hammer and chisel, which Technocover revealed in a scientific paper
SR2-3
SR3
SR3
SR3
  • If you worry your walls are too weak or your pocket too shallow for an SR3 door here is what an SR2 door can look like:
SR2 test on single point lock
LPS security door market
  • Security doors as strong as brick, rated SR3, will set you back £2,500-£8,000 and SR rated doors tend to look like prison doors (unless upgraded with decorative skins) and be outward opening with no glazing (unless you can accept a cell-style vision panel), and are usually made from 1.5mm steel. They are only as strong as the brickwork holding them, which can be gouged to release straps and screws, eg by percussion chisel or a concrete saw three inches around the frame to miss the fixing bolts – but those are higher threat tools than a SR3 rating which could cut a man sized hole in the door anyway.
  • The highest rated available door used to be SR4 (as government did not buy higher rated doors as extra security is provided by alarms and guards), but now, for example, AccentHansen and Sunray do an SR6 door (the latter Excludor 6 is basically a cell door, still only uses 1.5mm steel skin, the leaf is only 45mm thick and it weighs ‘only’ 135kg), although even if anyone started making an SR8 door it would only be certified for 20 minutes attack time and would only be as strong as the M20 bolts holding it to the reinforced concrete. The arms race to an SR7 came one step closer in 2022 when BastionDoor achieved G5 thanks to a new SR5-rated Surelock McGill GSS7 cylinder guard, model G5-LG05, but not quite G10 for all the door, so missed out on equivalence to SR7; until then the only way to get a G rating was two cylinders making already heavily armoured doors even more heavy and expensive, hence perhaps not much of a market.
  • Even an SR6 door can be gapped by steel wedges and a sledgehammer: the goal is to use locks strong enough to hold it to the frame nonetheless and skins that cannot be cut through fast enough to let someone in within a few minutes. During attack, the door and handles will shake, covers will fall off the lock, the bolt will start pulling out of the keep in both planes and the case will twist, hence it needs to be long throw. Incredibly the high security locks fitted to top end security doors usually only have a throw of 0.75-1.5″ (eg Stafford Bridge Wilton / Halton to SR4 uses multipoint 25mm throw), but tests on steel doors have shown top and bottom bolts can hold until the middle spreads 10″.
  • Only Fort Engineering makes LPS 1175 residential doors (retested EN356 doors), so you either have to buy from a monopoly supplier in that sense or buy to a different standard and beef it up with a grille, or accept an industrial look, or pay to upgrade to a decorative skin knowing the fabricator isn’t that experienced in it.
Police
Housebuilding / SBD
  • Police recommend, and building regs specify, a minimum of:
    • wooden door weight of 600kg/m3 (40kg for a normal door size slab),
    • 44mm thick,
    • solid or laminated,
    • frame rebating leaving 32mm of thickness,
    • panels 15mm thick and glued and mechanically fixed, and
    • panel smaller dimension to be under 23cm (eg can be high as you like but no wider than 9″).
  • SBD only requires doors to be A3 rated – protecting for 3 minutes against quiet light small pocket tools. SBD lasts 2 minutes longer against quiet tools than SR1 but assumes burglars won’t break glass or whack one tool against another for mechanical advantage. With no fewer than 342 SBD door suppliers at the last count, you probably need to look for additional and higher ratings to narrow your search. Police rarely require LPS1175 for newbuild estates, and rarely expect anyone to go for a door rated over SR2 and anyway advise that the addition of a grille could be OK instead of an SR3+ door, although that is a bit dodgy in the event of a fire and liable to not be used due to inconvenience.
  • SBD counter terror rating usually coincides with SR3 or SR4 doors but some go higher up to SR6 and occasionally a product is CT rated because it is also ballistic, eg Teckentrup 62 Secure B FB4 is only LPS2081 SRB for manual attack. Typical weights are around 200kg as frame and skin thickness is roughly doubled. SBD counter terror doors include:
    • Assa Abloy:
      • Powershield metal door range:
        • Clandeboye, SR2 & SDB, 44mm, outward opening, with 1.5mm skins and reinforced frame and allows vision panels;
        • Glen SR3, 46mm, 3mm skins and reinforced 1.5mm frame;
        • Strangford, SR4 is similar spec but with stronger reinforcements inside;
        • The Glen and Strangford were replaced by the 40% lighter Torr SR3 and Titan SR4, allowing vision panels.
      • Safeguard wooden security door range:
        • Trenchard, SR3, 56mm, multipoint, wooden with plywood core, outward opening, with anti cut reinforcement and allows vision panels;
        • Ultimum, SR4, 67mm, inward or outward, multipoint, with hardened steel reinforced solid hardwood frame and allows vision panels.
    • BastionDoor
      • BastionDoor, SR4
      • BastionWall 5 Doorset, CPNI, 136kg.
      • These are part of a bolt-together demountable saferoom, with wall thickness of about 50mm.
    • Bradbury – M2M range, SR3 to SR4, outward opening
    • Burton Safes – Kronos, SR4, inward or outward
    • Charter Global – Obexion, SR3 & SR4, 66mm, 1.5mm skin (SR4 armour option), 1.5mm frame (SR4 2mm) with 3mm reinforcement, multipoint and vision panel options, inward or outward.
    • Lathams – 70mm steel high security doorset, BR3, 2mm frame, 1.2mm skin, 2 vertical & 2 horizontal stiffeners, c. 100kg.
    • Robust – TuffDor4, SR4, 61mm, outward opening, timber core, vision panel options, PVC wood effect option
    • Rotec – Securedefend 4, SR4, 1.5mm frame & skin
    • Safetell Stronghold 4, SR4 & FB4
    • Stafford Bridge
      • Sandhurst, SR3, 54mm, FB1, C25 blast
      • Wilton, SR4, FB7, C15 blast, 55mm, hardwood plywood core, CPNI option (‘2016’, outward opening, no vision panel), inward or outward, wood panel option, multipoint, GridGlass vision panel option, FB7 option
      • Halton, SR4, FB7, C15 blast, 56mm
      • Whitehall, CPNI, FB7, C15 blast, 55mm
    • Sunray – Excludoor SR3 upwards
    • Technocover (owned by Hill & Smith alongside Barkers Fencing) – Sentinel / Ultrasecure SR3 upwards
    • Teckentrup Secure 2 upwards: 2 FB4 / 3 / 4 / B FB4
    • Warrior Doors – Warrior 4 upwards
SBD counter terror door at Scotland Yard, demonstrating how fabricators can skin a steel door with beautiful wood panel and clever dummy glazing
  • Police advice on refuges and gun storage gives hints on how to make your own security doors, below.
Refuges
  • Police advice on domestic violence sanctuary rooms is:
    • frames fixed to wall on 400mm centres with 5 gauge 80mm screws into wood stud or 8 gauge 120mm frame fixers into masonry,
    • door to be outward opening 44mm solid fire door with BS8621 hook multipoint with 3 star cylinder plus 2 star handles, anti thrust plate, three EN1935 grade 11 102x76mm hinges, two hinge bolts and a TS002 door viewer, plus fire and smoke seals and threshold and 25x33mm fire door stop.
  • Morticing leaves only about 13mm wood either side holding the door shut, so this refuge door spec does not stop each bolt being crowbarred out, but being outward opening it would be all but impossible to kick in before help arrived.
  • An alternative might be a steel security fire door for around £800 such as a PAS FD120 from Lathams, although discretion is part of security in a real DV refuge scenario.
Gun rooms
  • Police advice on gun storage doors is at least two Hela Tech 26/5 or BS3621 equivalent locks and:
    • gun enclosure doors (not in rooms) can be security rated 44mm solid timber with three hinges (no hinge bolts needed unless outward opening) and reinforced keeps;
    • gun room doors should, depending on risk, be:
      • 44mm solid timber or laminated security door,
      • timber faced with 16 gauge steel wrapped round the edges and screwed externally with security screws on 125mm centres, with frame screwed 150mm deep into wall at 400mm centres,
      • SR4 door
      • 10mm steel plate, :
        • with three steel hinges continuously welded to door and frame and 20x40mm section 20mm deep hinge bolts only if hinges external,
        • door overlapping frame 20mm if outward opening,
        • keeps continuously welded to frame if inward opening,
        • locks in 60x10mm steel carriers continuously welded to door,
        • frame of 10mm steel 200x75mm T section rag bolted on 400mm centres either side of T section with external bolts splayed or welded,
        • if armoury door then bolts should throw 20mm, three instead of two locks needed and if lock drives boltwork then one bolt needed along top, and
        • if steel armoury door then reinforced with 6mm steel 50x50mm angles welded close to edges and hinges welded to it plus 6mm steel angle bolt shrouds welded to the frame plus timber walls floor and ceiling of flooring strength with 75mm gap between steel and boards.
        • eg, Independent Safes charge around £2,400 for a gun room Coventry 10mm plate door (or 8mm with 6mm reinforcement) or £3,500 for a 25mm Birmingham plate door, these can all have internal locking including lockout feature.
  • There is the curiosity that police imply unprotected hinges will slow down gun thieves.
CPNI
  • CPNI set the government standard for counter terror and counter espionage barriers. These ratings used to be called SEAP before 2009.
  • Their main standard for manual forced entry is MFES. Ratings range from Base through Enhanced to High, typically reinforced ballistic & blast versions of at least SR3 doors. SR3 automatically qualifies for Base level on the assumption it provides some delay against non specialist hand tools by a determined opportunist. Enhanced level provides longer delay against non specialist tools and some delay against specialist powerful tools. High level provides substantial resistance to specialist power tools. CPNI uses the LPS1175 225mm x 400mm ellipsis for bodily entry test. MFES’s predecessor until 01/04/15, PBAS, was similar to the ‘resistance units’ in EN1143 strongroom standard in that instead of tool groups it applied tool coefficients of noise, weight, portability, availability and effectiveness. CPNI can give further ratings for penetration standards: ‘base’ (equivalent to SR3, holding off an amateur terrorist for 5 minutes), ‘enhanced’ or ‘high’ (what used to be called SEAP 1 / 2 / 3), and detection (of compromised asset inside) 1 to 4. CPNI is a more severe test, for example a product with SR4 might only last 1 minute against Base CPNI and 0.5 minutes against Enhanced CPNI.
  • As an add-on to MFES, their MTAS standard addresses marauding terror attacks by lone attackers with explosives, guns and makeshift tools followed up by a pair with tools, where the aim is to demoralise terrorists, make them use up bullets and give citizens time to hire or escape via short delays – by which time hopefully armed police have neutralised the jihadis. Ratings range from Basic (lone attacker with knife & fire extinguisher) through Low (add handgun or shotgun) & Moderate (add rifle, IED & hand tools), and High (add accomplice, battery tools & plastic explosive) to Expert (add specialist MoE tools). Ratings include separate ratings for any achievement from Low to Expert against BFES (ballistic) or EFES (explosives). Unusually for security standards, delay ratings go as low as 30s as that is enough to escape or make a terrorist seek out a new target in the last few minutes before armed police arrive.
  • CPNI counter terror advice is to use auto-deadlocking so you are always locked, or at least to have remote locking so you do not need to wait or risk going to the door to lock it, and thumbturns are allowed as it is more important to be able to lock instantly in a panic without fumbling with keys than to not be vulnerable to manipulation through tool holes later made.
  • They also recommend surface mounting of locks to give it the protection of the door thickness and space for thicker bolts, and any single point lock to be at middle height to minimise twisting, ideally use multi point locks, and ensure 17mm bolt engagement in keeps.
  • They rarely award MFES High rating to a door with a vision panel. Likewise they advise against door viewers – in fact any glass, plastic or aluminium, unframed doors, double doors, inward openers, latches, single point locking, thumbturns, missing thresholds plain meeting stiles on double doors. Where doors must have vision panels, eg for sunlight or entry control, they suggest considering video entry or a lobby with another door – ideally with no vison panel.
  • They usually expect security hinges and dogbolts, astragals, anti drill plates and sometimes extra armour. They do not allow lock cylinders without cylinder guards, even at MFES Base level.
MoD
  • MoD advice at JSP440 Chapter 7 Section 5 Annex A on doors for personal counter terrorism is simple:
    • look after keys,
    • don’t open the door if you’re not expecting someone, and
    • keep the porch brighter than the hall.
  • For military installations Chapter 5 Section IV advises:
    • 44mm solid hardwood doors, or retention of softwood or panelled doors if wrapped in 18 gauge steel plate facing screwed into the edges on 10cm centres,
    • dog bolts,
    • strong hinges,
    • masonry bolts or steel brackets,
    • outward opening,
    • any letterbox sealed,
    • consideration given to locking bars inside and replacing wooden sills with concrete, and
    • any glazing to barred, or protected by mesh, grille or shutter, or replaced with security glazing if necessary for presentational reasons.
  • Buildings
    • A reinforced steel door, or solid hardwood door faced in steel sheet, would be class 4 in a building. Confusingly, this would need to be set in a reinforced concrete wall, whereas a stronger steel strongroom door could apparently be set in unreinforced concrete. And equally confusingly it could be paired with windows as long as they are resistant, whereas a strongroom cannot have windows.
    • A class 3 door needed only offer some resistance and simply needed to be as strong as a class 3 wall, ie a cavity brick wall.
    • JSP440 para 05118 set class 2 at 4″ brick or even lightweight block and a class 2 door needed only be as strong as that.
    • Class 1 buildings would be a shed as it only offered weatherproofing, so it would either not lock or could be walked into with the door still locked.
  • Strongrooms
    • Strongrooms have concrete walls, steel doors with boltwork, class 3 or 4 lock, and no windows.
  • Secure rooms
    • Secure rooms are treated like class 3 containers resistant to limited hand tools, with laminated glass, solid wood door and lightweight walls. Confusingly there are also secure room types of much higher strength.
    • For ‘type A secure rooms’ – to replace a security container to prevent espionage and to lesser extent stop manual attack – it specified government doors called the Oxford or Cambridge fitted with a Medway lock, teamed up with the equivalent of 6″ reinforced concrete or 14″ unreinforced brickwork (which police treat as about SR4 for firearms storage) and is usually measured as more like SR3.
    • Type B secure rooms could make do with an Ashford government door teamed up with 4″ reinforced concrete or 9″ unreinforced brick, which is SR2 to SR3. This was also class 3.
  • Containers
    • Class 4 was to delay prising or fishing by power tools, eg safes.
    • A class 3 container was of stiff construction that would require an attacker with limited hand tools to open and had a class 3 lock that would need a locksmith to pick, eg cabinets
    • Class 2 was supposed to keep out opportunities but would not stop twisting by tools, eg document boxes,.
    • Class 1 was only for privacy, eg document pouches,
    • These classes used to be called categories and groups before that.
  • Locks
    • Class 4 locks were typically Manifoil combinations, and were supposed to require techniques not even available to locksmiths.
    • Class 3 were typically keypads or Mersey key locks (cousin of the Manifoil), and were supposed to require locksmiths to open.
    • Class 2 were certain Ingersoll or Assa 5 detainer or 10 lever locks, although nowadays rim locks are seen as class 1, and were supposed to require skill but few resources to open.
    • Class 1 were padlocks and Chubb mortice deadlocks – even the £150 3K70 5 detainer was dismissed as class 1; class 1 was supposed to moderately delay unauthorised opening.
LPS 1175
  • LPS1175 standard is for security barriers, but does not address thermal shock, chemical, blast, ballistic attack or static load. To resist static load, LCPB recommend multipoint locking.
  • LPS1175 Ratings SR1-2 are for domestic and low commercial risk, SR3-4 for commercial risk, SR5-6 for high risk and SR7-8 for extremely high risk. For comparison, SR1-2 (or PAS or RC2-4) is normally recommended for secure hospitals, SR1 inside and SR2 outside, with SR3 reserved for the most dangerous inmates in high security hospitals equivalent to cat B prisons, and the highest rating ever needed in that setting being C20 to cover noisy tools wielded for over 20 minutes, although ratings up to RC5 or SR5 may be specified for static load resistance and robustness to take years of abuse.
  • SR1-2/3 is for amateur burglars, SR3/4 upwards for professionals, and SR5/6-8 for high value internal storage areas, so the assumption is that not many buildings will have facades exceeding SR4/5, in fact, as SR4 requires reinforcement, few facades could even exceed SR3. Luxury jewellers go for SR3 or SR4 as they fear moped and ram raid gangs with sledgehammers who would threaten witnesses with tools so will attack for longer than anti bandit glazing was designed for.
  • LPS1175 assumes one attacker uses tools in group A-E and two attackers use tools in group F-H.
  • Tool groups A-E are normally for external facades (group B including a 3.6V drill, group C including a 12V drill, group D including 12V power tools, group E including 18V power tools), while groups F-H are normally for strongrooms (group F including 36V power tools, group G including 54V power tools), and H tools (like diamond cores, thermic lances, 5kW grinders, concrete chainsaws) being a precursor to professionals driving, shooting or blasting their way in. It would be unusual to spend enough to make the whole building resistant for many minutes to the biggest sledgehammers, pickaxes, crowbars, drills and saws, as by the time those tools work the response force should intervene, but in case they do not, perhaps because the a layer of detection failed or they have been neutralised, an inside room would be hardened against expected threats such as hydraulics, enforcers, cutting torches, the most powerful battery tools and petrol tools.
  • Consider that even group G 54V power tools are available from your local Toolstation: £470 for a reciprocating saw, £460 for a 5” grinder, £750 for a hammer drill, £870 for a cutoff saw. Group H tools are in the shops for a few hundred or thousands: a 5kW concrete chainsaw will set you back £2,000, a 5kW cutoff saw goes for some £1,200, 125mm diamond core bits are around £40, Arcair Slice thermic lance costs about £2,600, enforcer rams are on ebay for £280 and 5t toe jacks for £150, Stihl 20″ 2.4kW rescue chainsaws sell for £1,700 and 4.8kW 16″ petrol ring saws for £3,600. The rule of thumb is that group A tools fit in a pocket, group B tools fit under a coat, group C tools fit in a sports bag, group D-E tools fit in a backpack and group F-H tools need a vehicle.
LPS1175 tool groups
  • LPS 1175 also covers ‘security enclosures’ which are ‘security containers’ if self contained.
  • Fixings to substrates should be by M6 for SR1, M8 for SR2, M10 for SR3, M12 for SR4-5, M16 for SR6-7 and M20 for SR8.
  • Glazing, infills and external fixings can all be attacked in LPS1175 testing.
  • Testers often defeat products by breaking weak glazing, fishing through holes, cutting multipoint rods, hacking solonoid wiring or breaking escutcheons.
  • SR rated products are still sold, but LPS1175 version 8 in 2020 changed to separate tool group and time scores, so, for example, lasting 1 minute against tool group A gives an A1 rating equivalent to SR1, B3 is the old SR2, C5 is the old SR3, D10 is the old SR4, E10 is the old SR5, F10 is the old SR6, G10 is the old SR7 and H20 is the old SR8.
  • The test is passed if someone (a 16”x9” ellipsis block) could not squeeze through a hole made in the product, or if the product is smaller than the usual 16×9 hole then a hole cannot be made 5” in diameter or big enough to pass what is protected if smaller than that.
  • Melissa Ahmed wrote a paper on how to simulate LPS1175 tests to increase likelihood of passing first time. We can conclude from her observations that doors tend to fail LPS1175 because a hand hole is made (usually a 9cm triangle with hammer and chisel) allowing manipulation perhaps of a handle, or due to cutting, levering or impacting, meaning hammers, saws, drills, wedges and crowbars are the leading problems.
  • Most LPS1175 doors are industrial looking, but, if you want the choice from the whole market, below is a list of what you could buy and make do with or adapt for residential use, either buying the fabricator’s paint, PVC, wood or wood effect finish or adding your own skin, or leaving the slab as it is for use in a bunker or perhaps a garage interconnecting door.
  • LPS 1175 doors, excluding communal doors, include:
    • AccentHansen with their Secureshield using a honeycomb-filled 1.6mm steel skin ranging from SR2 to SR6,
    • Ascot with their honeycombed steel Centurion A36 range rated SR1 to SR5 (they imply they are CPNI but obviously the SR1-2 could not be). The SR3-5 are SBD rated, SR4-5 for terror.
    • Assa Abloy do doors from SR2 to 4. The (Adams Rite) Safeguard wooden range can be painted, veneered or panelled and can have vision panels so might be discreet enough to pass as a residential door. The SR3 rated Trenchard is steel reinforced 59mm laminated hardwood which can be glazed. Meanwhile their SR4 rated Ultimum is also steel reinforced wood at 67mm thick and can open inwards but at most in terms of glazing can only have a vision panel. These two appear to be based on the old hardwood/steel laminate AdamsRite Concept 2 & 3. Neither comes in a ballistic or blast version although they do sell timber-based ballistic and blast doors without manual attack security rating. The new CPNI versions in this timber range are the Cromwell (formerly Concept 2, 54mm, 4 x 16mm hinge bolts, single or multipoint rim lock, blast rated to EXV25, vision panel option), Fenwick (58mm, 4 x 16mm hinge bolts, rim multipoint, hardwood lamel & steel plate core, ballistic to BS5051 G1, blast rated to EXV20, said to be SR4, vision panel option) and Bowden (formerly Concept 3, 68mm, hardwood lamel & steel plate core, 4 x 16mm hinge bolts (can be reinforced into masonry) & multipoint, blast rated to EXV10, ballistic to FB4 & G2, vision panel option). There is still an old video on youtube of real British soldiers trying to sledgehammer, axe, pickaxe, enforcer and shoot their way past the old Safehaven door exit only version used for a panic room, which looks like it is based on a Concept 3 with solid hardwood, steel & resin core, morticed 3 point locking, upgraded to G2 / FB4 ballistic and EN1523 blast. Their doors from SR3 upwards and FB1 are SBD terror rated. The top of the range used to be customisable, the CPNI blast / ballistic / manual attack Wellington up to 1,500kPa-msec impulse with vision panel option and the Wiltshire ballistic door. Resold by Dennison Doors.
    • Bastion do doors up to SR4 for their BastionWall modular enclosures. In 2022 they launched a new lighter Bastion Door rated G5 using a new cylinder guard.
    • Bradbury do LPS1175 doors in their SBD M2M range, 45-66mm thick with 1.5 steel skin, to SR2 to SR4 which can come with vision panels, inward opening and wood effect finish. The SR4 is SBD terror rated. However, as with most high security doors, the lock is rim mounted and massive and the doors look industrial. These are resold by Theam (also resell Extendor grilles and Seceuroglide roller shutters). Lathams, Security Care (as Demon Doors) and Vaylia resell Bradbury doors.
    • Burton Safes do an SBD rated Kronos range from SR2 to 4 which can be inward opening but only industrial style. The SR4 is SBD terror rated. Incidentally, they do a domestic Janus range of RC3 & RC4 doors to EN1627 which could see off a battery drill, club hammer or axe for 10 minutes, but not a sledgehammer or power tools.
    • Charter Global do a 66mm thick 1.5mm steel skin Obexion door rated SR2 to SR4, which can have vision panels, although the SR2 is outwards only. The SR3-4 are SBD terror rated.
    • Crittal Fendor used to make Secureline SR2 doors (and SR2-4 windows) but did not reveal much online to reassure around domestic suitability. They made the windows for Broadmoor and Rampton, but withdrew their LPS1175 certificate on 25/11/20, as the author understands it because their only client asked them to turn the windows inside out to stop prisoners escaping.
    • Design & Supply do a rockwood core 1.5mm steel skinned 45mm thick Desray door rated SR1 to SR3 (Security Care resell the Desray 2 to SR2 and Bradbury doors rebranded as Demon).
    • Doortechnik do outward opening X15 and X20 doors 45-55mm thick 1.5mm steel skinned rated SR1 to SR3 with vision panels.
    • Fort Engineering Ltd do a retested range based on the EN356 Skydas door that Shield Security Doors used to sell, which combining conflicting brochures mixing RC grades with SR grades appear to be:
      • FORT001 to PAS24 & SBD (to get PAS should need to be a tweaked version of the Standart3 as PAS requires RC3)
      • FORT002 LP2S to SR2 (Fort may have muddled the FORT002 LP2S with FORT003 LP3S judging by installation guide and website).
      • FORT003 to PAS & RC3 (Skydas Standart3)
      • FORT003 LP2S to SR2 (Skydas Standart3 upgraded), inward opener, 75mm thick, 1 drill plate, 16 x wall screws 7.5mm x 72mm or 152mm with 24mm steel covers, 2 rolling dog bolts, Iseo Fiam 678G multipoint & Iseo R6 cylinder, 10mm (up to 20mm) gap around frame, £2,940 for four panel (two glazed).
      • FORT004 to RC4 (Skydas Premium)
      • FORT004 LP3S to SR3 (Skydas Premium upgraded), inward opener, 91mm thick, 2 manganese drill plates, dual grates, 17 x wall bolts 11mm x 70mm with 24mm steel covers, 8 hinge bolts, Iseo Fiam 678G multipoint with 8 multipoint bolts (6 locking stile + top & bottom) & Iseo R6 Plus cylinder).
      • FORT004 FB (Fort level 7) (Skydas Premium FB4) to RC4, SR3 & FB4.
      • This lineup includes the only SR3 residential security door on the market, other than arguably Teckentrup whose doors are not available in the domesticated milling, carving, veneers and films of the Forts. The prettification is done with painted or unpainted wood, or painted or 1.5mm Okoume plywood veneered 12mm Umidax-HL or EDM, or painted stainless steel.
      • They must be double locked for SR2 or SR3 rating.
      • Fort Security Midlands posted a video in 2015 claiming to be ‘LPS1175 certified’ but BRE had never heard of them by the looks of it. Another reseller, Arris Fort Security, fell into similar error.
      • Glazed versions are not rated.
    • Harlech do Crimegard SR1-4 doors.
    • Hampton Conservatories say they do an SR2 wooden outward opening door.
    • Hodgson Sayers do maintenance and access doors, the Epsilon (1-3), Apeiron (2-4) and Colossus (2), rated SR1 to SR4.
    • Kingsley Composites do GRP doors SR2-SR4 for environments where steel would rust, but are really only aimed at substations etc.
    • Lincoln Security doors are 66mm thick with internal Dufaylite honeycomb and come in SR2-SR4 including inwards opening and glazing. There is the Victor 1.5mm skin, Vigilant (2mm skin) and Vanquish, while the Vulcan (3mm skin) has reinforcement and armour with CPNI rating aimed at critical infrastructure.
    • Metador do Defender Xtreme doors rated SR1 to SR3 with vision panels available but only outward opening.
    • Parkside Group do a Comar 7 SR2 RC3 door.
    • Premier Security & Fire Consultants do SR2 rated doors including inward opening, the Industrial 75&81, Jansen 75 & 81, SBD2 & SBD2 81 and SBD D75 & SBD D81.
    • Prestige Fire Door Services do doors including inward opening rated SR1 to SR2.
    • Protect GRP claim to do an SR3 door but this is outwards only and appears to be for the tollbooth market.
    • PSF Wales do outward doors rated SR2 to SR4 in their Hercules range but aimed at the enclosure market.
    • R&D Sheetmetal do doors in their Security Evo range rated from SR3 to SR6, with vision panel options with grilles on the SR3-4. They say they have a new SR3 & 4 certification on an inwards opening model, and appear not to market the SR5-6 versions.
    • Rhino Systems do doors rated SR2 to SR4 in their HF (fire) and WT (flood) ranges including inwards openers and vision panels, 66mm thick with 1.5mm steel skins and blast protection. They also have RhinoHC doors with CPNI rating to deter blast, ballistic and ram attacks, 40-60mm thick with vision panel option.
    • Robust UK (owned by Novoferm owned by Sanwa) do an outward only TuffDor range with steel skinned timber, with vision panel and wood finish option, rated SR2-SR4, resold by Security Care. The Tuffdor 4 is SBD terror rated.
    • Shadbolt do a Shadsecure internal door (eg for flats) rated SR2 with a chipboard core and wooden finish.
    • Stafford Bridge Doors do 1.5mm steel skinned plywood ranges from SR2 to SR4 46-59mm thick with vision panel option, the Wilton, Halton, Sandhurst and Hendon, including steel or wood finish, inward openers. They also do CPNI outward only versions of the Sandhurst, Wilton, Halton and Whitehall with no vision panel. The Sandhurst SR3 and Wilton SR4 were seen in a Youtube video defeating a police MOE team in training with an enforcer, halligan and hydraulics. Their SR4 door was in a youtube video seeing off more powerful Holmatro rescue tools which tore the skin off but it stayed locked by a SurelockMcGill Stirling. The Sandhurst and Wilton can also be rated for blast and up to FB7 ballistic. The locks are surface mounted Surelock McGill Stirling single point.
    • Steelway Fensecure do outward doors rated SR2 to SR4 but the firm is mainly in the gates market.
    • Strongdor do a 1.5mm Zintec skinned rockwood core outward opening Securidor range rated RC1-3 (75-100m thick) or Securidor+ range SR1 to SR3 (45mm thick), and with LPS2081 B rating (PAS24 equivalent) on their Lumidor SRB inward openers with vision panels.
    • Sunray Doors do an outward only ExecDoor range from SR1 (A1) to SR2 (B3) and an SBD rated Excludor range rated SR3 to an astonishing SR6 (SR4-6 are SBD terror rated), with surprisingly all models only 45mm thick using only 1.5mm steel skin. All except the SR1 take vision panels. At SR6 it looks like a cell door with bars on the vision panel and a humungous cylinder guard, weighing in at 21 stone. The SR5 is CPNI rated. The SR2 can have a vision panel without bars. The SR3 & SR3 typically rely on a Pickersgill-Kaye HS100 rim multipoint. Pickersgill-Kaye was bought by Assa Abloy in 2017 and is now part of the High Security & Safety Group within Opening Solutions.
    • Technocover do the SBD rated outward opening Sentinel range from SR2 to SR5 and UltraSecure range from SR3 to SR4. These are SBD terror rated from SR3 upwards and can have vision panels and be wood clad. The SR2-3 doors come in 2mm steel skin and SR4-5 in 3mm.
    • Teckentrup do a blast and fire resistant SBD rated Teckentrup 62 range from LPS2081 SRA (outward opening Secure A) to LPS 1175 SR3 and up to RC4 42-62mm thick with 1.5mm steel skin, plus now LPS 1175 D10 to replace SR4. The SR3 can be inward opening and all can be glazed. They also do FB4 bullet resistant and cell doors also to RC4. The SR3-4 and FB4 (even if the FB4 is only SRB or SR2 manual attack rated) are SBD terror rated.
    • Torterolo do SR2 and RC3 rated doors including inward opening, the fire rated Protection and the Comfort UK. They also do RC3 rated Chrome, Dream and Wave (wood finish option) and RC4 rated Gold Plus (wood finish option) residential doors.
    • TS Designs do outward opening SBD rated B5 and C5 rated doors with vision panel option, the PS2 and PS3. Spec for both is 45kg/m2 45mm thick with foam core, 1.2 steel skin and Mico’s Abryll (SR6) or Elite (SR3) single point or INT multipoint (SR6 & 69kPA blast) locking.
    • Urban Front do an SR2 inward door, the E80S. Their entry level front doors start at £4,500.
    • Warrior Doors do SR2 and SR4 doors including inwards opening. Warrior3 and Warrior4 can come SBD terror rated. The infamous Warrior jewellers door that stopped a ram raid used a Pickersgill-Kaye HS270 mortice multipoint.
Vision panels create break in stiffeners as seen in Fort model
Fort directory entry claims EN1627 classes are LPS1175 ratings
Fort distributor pretend they have an SR4 door

This FB video shows the Fort Engineering FORT001 PAS24 SBD door taking a beating from a fireman’s axe. https://www.facebook.com/Fortengineering/videos/644982938938350/

An old Assa Abloy Safeguard door shot and sledgehammerd by soldiers
Fort SR3 test
Salsbury SR4
Sandhurst SR3 (bolts top & bottom + 1 on side, Government level 2 so probably equivalent to new CPNI Enhanced version) & Wilton SR4 (bolts top & bottom + 3 on side), both with SurelockMcGill Slimline Multipoint
  • The only residential LPS1175 doors for houses are Fort’s retested versions of Skydas RC3 and RC4 doors to SR2 and SR3. The only other options with potential residential suitability via wood effect or panel options are Technocover’s Sentinel / Ultrasecure, Assa Abloy’s Trenchard, Bradbury’s M2M or Stafford Bridge’s Sandhurst. The Sandhurst is the thinnest, most domestic looking and may well be the strongest judging by the youtube video suggesting it can exceed its rating if teamed up with a multipoint, and costs about £8,000.
  • Prices start from around £1,100 for SR1, £1,600 for SR2, £3,000 for SR3 and £4,200 for SR4 for basic industrial ‘steel slabs’, and several thousands more for a wood finish and glazing, easily north of £4,000. Samson Doors, for example, say their Hormann ThermPro is rated SR1 and LPS 2081, costing at least £1,342 for the SBD version (which thus is equivalent to LPS1175 A3). The reality is you have to pay over £2,000 if you want a decent looking door to stop someone wielding tools for a few minutes.
EN1627
  • If you cannot afford an LPS 1175 door or don’t like their industrial look then you could consider EN1627. This will probably look more domestic but you have to be careful what EN1627 actually tests. It might be the realistic option if you are not sure about the door reveal in terms of how the brickwork would take the impact through the jamb or fixing bolts or how long it would take to smash out the bricks. Consider that there is a youtube video of police accidentally taking a one-brick wide pillar with the door using an enforcer – can your wall take an SR3 battering?
  • Also consider that the UK voted against EN1627 as we would be forced to adopt it as a British Standard.
  • LPCB published a comparison of LPS1175 and EN1627 and the Flint paper.
  • SBD have issued an interpretive guide to EN1627, warning that EN1627 does not test lockpicking.
  • EN1624 only requires cylinders to EN1303 with ratings ending 41 for RC1 to RC3 and ending 62 for RC4 to RC6, and locks to EN12209 to class 3 for RC 1 to RC2 and class 7 for RC4 to RC6.
  • Nor does it cover some common attacks, or barging out glass for RC1-RC4 – yes that is right, testers are allowed to hammer the rest of the door but not the glass.
  • Whilst up to RC3 stealth attack is assumed, RC4 is supposed to handle banging and crashing, but if you buy RC4 be careful what separate rating any glazing has as breaking glass is not tested up to RC4.
  • EN1627 assumes an intruder is larger when crawling through a hole than LPS1175, LPS2081 and PAS, while the LPS standards test for hand holes as well.
  • It gets worse: RC1 does not require laminated glass, RC2 allows P4A safety glass, RC3 makes do with P5A safety glass – which only really needs to stop a crowbar for 5 minutes but is only tested with a metal ball and which LPCB say is not even SR2, RC4 fobs you off with P6B which would need to stop a hammer for 10 mins but is only tested against 30 axe blows and would not even stop a screwdriver, RC5 assumes P7B will stop an angle grinder but is only tested against 51 axe blows, and RC6 says P8B will see off a percussion drill but is only tested against 71 axe blows.
  • Check ESG’s ‘just how tough is security glass’ video of these European standards like 30mm P8B going down with almost the first punch, against a sledgehammer anyway. Even if you specify locks and glazing that you know are more secure than what EN1627 requires, the SC rating does not test it all as one door.
  • LPCB say that for glazed products RC1 is equivalent to at most SRA, RC2 varies from SRA through SRB to SR1 but probably just SRA, RC3 is a slight improvement varying from SRA to SR1, RC4 varies from SRA to SR2 but usually SR1, then there is a big step up to RC5 which varies from SR1 to SR5 but usually SR2 to SR4, RC6 varies from SR2 to SR6 but usually SR3 to SR5. For unglazed products they say an RC2 has an even chance of being in SR1 territory, RC3 ranges from SRB to SR2 but usually SR1, RC4 ranges from SR1 to SR3 but usually SR2, RC5 ranges from SR2 to SR5 and usually SR3 or SR4, RC6 ranges from SR3 to SR6 but usually SR4 or SR5.
  • To be fair, the few products that have dual rating tend to equate RC3 with SR2 and if you look at videos of RC4 doors like the Skydas Standard 3 RC4 they can take a hell of a pounding from a sledgehammer (more than the club hammer used in SR3), however, BRE claims that 90% of RC4 doors fail SR3 and only last about 4 minutes against a hammer and screwdriver, meaning you probably do need need a RC5 to equate to an SR3, although the Skydas/Fort Premium RC4 did pass SR3 (and their Standard RC3 passed SR2).
  • The twist is that EN1627 RC4 requires a 8.5″ brick wall so implies it is equal to it and if you only have a cavity wall or lightweight blocks then there may be little point having a door stronger than RC3, which is awkward as that pegs a brick wall at about SR1. You can see from the tool groups that RC4 is similar to SR3, and the standard even attacks for twice the time, so is not to be sniffed at, but the methodology has flaws including that it is hard to check UK accreditation, hardware is not tested in the product and testers are limited in how they are allowed to use tools.
  • In short, go for RC4 with no glazing if you want an almost affordable residential looking entry level security door that might be as strong as your wall, but to replicate SR3 for high risk domestic risk or any kind of targeted attack you probably need RC5 unglazed or RC6 glazed – which are hard to come by if you want residential style, in fact the author has yet to find any sold in the UK. You can buy industrial RC5 doors from MetalQuartz (outward opening no or fully glazed, RC4 no half or fully glazed inward or outward opening), Fichet (who also do communal half glazed versions and fully glazed up to RC4) and Doruksafe (who also do wood faced versions), and industrial RC6 doors from Sommer and Salzer. The closest you will come to the strength of brick in a residential door is an RC4 from suppliers like Abbey, Hormann, Novoferm and Skydas’ successors Fort or Munitus, but unless your home is made of wood, breeze block or thin brick you may as well forget about EN1627 for now, although you could always hide an RC5 slab inside as an internal door or interconnecting garage door. As these high threat level doors are ruinously expensive, horrendously ugly and probably overspecified by tool group for your needs, you are probably better off buying an RC4 with a mob attack accreditation such as FE5 to give assurance against sledgehammers and crowbars, in fact even RC3 doors can also come with that rating even though they don’t have an EN1627 rating against even a lump hammer.
  • Suppliers include:
    • Gunnebo do a MagTek RC4 door.
    • Hormann do RC3-4 ThermoCarbon (from £4,876) and ThermoSafe Hybrid front doors achieving RC3 with glazing or RC4 without glazing (from 4,400 EUR) which is steel skinned foam. Their ThermPro is PAS & SBD.
    • Munitus is a manufacturer in Lithuania apparently launched by ex-Skydas staff, led by former production director Vaclovas Bieksa and commercial director Markas Vysniauskas, from the ruins of that bankrupt manufacturer when it stopped paying staff, using Skydas blueprints hence they have the same products up to RC4 & BR6. They won a misuse of information lawsuit when Skydas’ administrator withdrew the claim after being asked for security for costs (suggesting they cannot come after Fort either). They cited Shield Security Doors as their reseller the same as Skydas. Top Security Doors claim to be a reseller in the UK, but their director Adomas Saltenas is named on a draft Shield Front Doors website so it is not clear whether this is a diversion. They seem to have teamed up with German manufacturer Gerlock for some new window designs in return for Gerlock acting as reseller deal but also rebranded the Skydas doors as the Gerlock Classic RC3 (upgradeable to ballistic FB4), Max RC4 (upgradeable to FB6), FE5 (upgraded version of Gerlock to SD-STD-01.01 FE5, upgradeable to FB4) and FE15 (upgraded version of Max to FE15, upgradeable to FB6). Notably they almost completely redesigned the Skydas SD-STD.01.01 doors by replacing the horizontal and vertical ribbons with narrower spaced horizontal bars and vertical box sections forming a square grid.
    • Novoferm do RC2-4 internal and external doors available in wood effect. The external 64m 1/1.5mm steel skin Novoporta Premio ES1 and 40mm 1mm steel skin Novosecure ES6 Slimline (can be glazed) are RC2; the external 63mm 1.5mm steel skin ES63 and internal 51mm thick 1mm steel skin ES9 are RC3; the 51mm 1 mm steel skin internal & external ES11 is RC4.
    • Robust UK sell the Secur-Dor range: Secur-Dor 2 with STS202 BR2 & PAS (55mm thick, 1.5mm skins, 1.6mm frame, 11.5m laminated vision panels on PAS24, 2 x CEN7 deadlocks on BR2), the Secur-Dor3 with BR3 & RC3 (55mm thick, 2mm skins and frame, multipoint) and the Secur-Dor4 with EN13124-2 EXR2 blast resistance and tested but they simultaneously say passed and not passed RC4 (48mm thick, 3mm attack face (1.5mm steel + 1.5mm stainless steel) + 1.5mm protected face, 3mm frame, timber core, 3 x deadlocks). They claim the RC4 is stronger than the SR4 as it is tested for 30 mins, which is a bit misleading as EN1627 RC4 tool contact time is 10 mins.
    • Shield Security Doors used to resell 75mm thick Lithuanian residential doors from Skydas rated RC2 (Standard2), RC3 (Standard3 / eg £2,400 for four panel (two glazed)) & RC4 (Premium Plus / Standard 3 RC4), and built panic rooms. These, or remarkably similar doors anyway, are now only available through a new UK factory Fort Engineering which is run by the same production director who used to run Shield and Skydas. Because Skydas was the Baltics’ biggest armoured door manufacturer, and Fort now makes the only LPS1175 residential doors as at May 2022, the saga and products deserve special attention, with Fort now offering a leading product for domestic security doors if you can get past their unfortunate history with questionable business associates and allegation after counter allegation. The author hoped to finally be able to recommend a residential LPS1175 door, but the CEO was perhaps understandably sceptical of clarifying the issues below with an unknown blogger on an unmonetised site who therefore cannot visit to build relationships of trust, so the author sticks to what is public domain. Fort are not unique in rebuffing enquirers who are not distributors or introducers, or in being opaque on who makes and sells what.
      • Associates:
        • London celeb ex-gangster Dave Courtney endorsed Skydas doors from Shield from his ‘Camelot Castle’. Potentially worryingly, though, his Skydas Standart3 by the looks of it has a letter box and viewer. In theory a letterbox could be a dummy for aesthetic reasons, built into the wood skin for heritage replica reasons, but another reseller says it may have restricted opening (one customer has even put a catflap in). Incidentally, although Dave says his is bulletproof, if it was then it would have eight hinge bolts not four and be 91mm thick not 75mm if it is a Skydas, as the author understands it, and as it does not have six hinge bolts it is a higher model than the Standart2 to RC2, and the shootbolts top and bottom suggest it is at best the Embassy to SD-STD.01.01 FE5 & EN1627 RC3, Standart3 RC4 or Standart3 to RC3 (likely as he did not specify ballistic). It is also not an ‘inch thick’ as Dave suggests, the steel plate is more like 1mm or 1.5mm. Whatever, it looks better than his previous doors, an old flush white slab and a red panelled glass job.
        • The other questionable connection is that at one time one of the main UK resellers was a cannabis oil salesman and escort agent.
      • Specifications
        • Standart1/StandartLS?: RC2?, 2 rolling 16mm dog bolts, 8 wall bolts. Fort still sell this. Price used to start at EUR794.
        • Standart2: 75mm thick, RC2, 1 steel lock box?, 1 steel lock plate, 2 x 16mm rolling dog bolts, 4 x 12mm fixed dog bolts, class 3 lock, 16 wall screws with 24mm covers. Fort still sell this. Price used to start at EUR1,254.
        • 2000: 86mm thick, 1.5mm steel, RC3, 2 x 16mm dog bolts, Passivhaus
        • Standart3: 75mm thick, RC3, 1 steel lock box, 1 steel lock plate, 2 rolling dog bolts?, 4 x 16mm fixed dog bolts, class 5 lock, side lock bolts + 1 top & 1 bottom (in steel tubes), 16 wall screws with 24mm covers. Price used to start at EUR1,432, now typically around £2,000.
        • Embassy: Standart3 + FE5, 1 x Manganese steel lock plate. Price around USD9,000. The Embassy also has the American US-DoS standard SD-STD-01.01 5FE rating to last 5 minutes against a mob with big crowbars and sledgehammers.
        • Standart3 RC4: 75mm thick, 2 steel lock boxes, RC4, 4 x 16mm dog bolts, class 5/7? lock, 16 wall screws with 24mm covers. Fort still sell this despite suggesting they don’t as it is listed in their Fort Security Doors distributor page.
        • Premium: 91mm thick, RC4, 2 Manganese steel lock plates, 2 steel lock boxes, 8 x 16mm dog bolts, class 7 lock, 6 x lock bolts + 1 top & 1 bottom (in steel tubes), grating (trays & ribbons) both sides, 17 wall bolts with 24mm covers. Fort still sell this. Price used to start at EUR2,722.
        • Fortress: Premium + FE15. Price around USD12,000. The Fortress also has the American US-DoS standard SD-STD-01.01 15FE rating to last 15 minutes against a mob with big crowbars and sledgehammers. Either the Embassy or Fortress could be handy in a mob attack as the FE rating is against a 6-man team (they have to be ‘muscular’ and over 11 stone) with a 12lb sledgehammer, 120lb ram, 5’ crowbar, wood splitter maul, 3.5lb axe, 4’ boltcutters, propane torch, cold chisel, punches and saws, albeit allowing up to a 12” hole. They can be fireproofed for an extra $3,500.
        • The Standart3 can be upgraded to FB4 ballistic and the Premium can be upgraded to FB6 ballistic. Their ballistic glazing was advertised as 25mm so the FB6 alleged more recently must be a different spec. The only cutaway photo of their ballistic glass shows a rebate of barely a few millimetres so that may be something to check.
        • Their forced entry glazing is only 11mm as the doors were designed according to EN1627 which does not test glass, so they are only anti bandit; likewise they do not have an SLPS1175 glazed door.
        • Locks are usually Fiam 678 or Mottura 8571, class 7, powering four bolts sideways and one bolt each up and down, plus optional Assa 565 class 2 day lock.
        • Their doors are typically spot welded steel sheet reinforced with stiffeners, typically two vertical C profiles cut into, or two ribbons laid over, four horizontal C profiles, on one side for most models but both sides for the highest rated products. However the attack face sometimes has the stiffeners exposed with the steel sheet on the protected face, but this has not stopped them passing numerous certifications.
      • Fort – the new manufacturer
        • Fort’s owners now also run Fort Security Doors Ltd offering ‘half price wholesale’ and, tellingly for EN1627 proponents, charge £450 extra to upgrade from RC3 to SR2 and now sell PAS24, SR2 and SR3 doors; with extras and VAT they come to easily £2,000 for a SR2. Strangely they also claim to run Fort North Ltd or Shield NW Ltd which the Registrar of Companies says do not exist despite their claimed 10 year pedigree – perhaps they are really typos and actually divisions or branches of Fort Engineering Ltd, but their Ts&Cs point to a company number for Shield North West Ltd, launched in 2011 as Crime Protect Ltd, now run by Alan Hanson & Lee Cunningham, which appeared active from 2011-2019. Fort is owned by Remigijus Guibys and Indre Stankute with Nigel Stevens and Fred Drabble as additional directors, although J&F Holdings Ltd (Drabble, Stevens and O’Connor’s joint venture vehicle) used to own 20% and James O’Connor was a director during 2014-6.
      • Skydas – the original manufacturer
        • There has been plenty of litigation (Fort trademark) between Skydas (the original manufacturer in Lithuania launched in 2002), incarnations of Fort Engineering (the copycat manufacturer in UK launched after it seems production director 2002-2014 Remigijus Guobys fell out businesswise with his co-owner brother Rolandas Guobys at Skydas whom he had already fallen out with personally since age 12) and numerous incarnations of Shield (the Skydas reseller in UK) and even Munitus, the Lithuanian firm which replaced them once insolvency set in.
        • Skydas itself was accused in 2020 of fraud, a fake land deal that siphoned off about £500K, hiding debts, not paying staff and taking deposits knowing it is going bust. Launched in 1999 with five staff in a garage, in its heyday it turned over some £4M in 2015, represented in 20 countries, but headcount plunged from 97 to 20 between 2017 and 2020 while staff sued for wages and by 09/12/2020 it officially had no staff. Remigijus had also sued it for fraud when it was asset stripped but ended up being ordered to let the remaining shareholder buy him out. It appears to have gone insolvent on 12/11/2020 and liquidated on 17/02/2022, just as he predicted it would. It used to sell EUR4M a year.
      • Shield – the reseller
        • Shield Security Doors Ltd was Skydas’ UK exclusive reseller launched by Remigijus in 2008, which sold EUR6.5M worth of Skydas doors by 2015, but fell into administration in 2017 amid ‘questionable transactions’ including fears around deposits and payments to Shield Luxury Doors Ltd (which Remigijus says hid his profits from Skydas and Shield Security Doors and undermined exclusivity), with £10,000 recovered from Lithuanian director Nijole Mazeikaite (Valdas’ wife), and owned by Rolandas Guobys. This was triggered by director Remigijus Guobys walking out on the other director Valdas Macys in 2014 to launch Fort Engineering Ltd, cutting out Skydas from the loop and taking his technical knowhow with him. Shield pretended Fort was their idea and the client database and order book was leaked to competitors behind the administrator’s back. Bizarrely, Fort sales execs were simultaneously acting as consultants to Shield Front Doors Ltd. Remigijus also runs Fort Security Doors Ltd co-owned as CEO with the CMO Vytautus Kajackas launched in 2017. We can deduce from a company video that most of their revenue came from panic rooms rather than doors as they were only selling about 100 doors a year. We are told in one case that distributors were making 400% markup on doors, but in the UK this would now be way under the 40% that Skydas normally made on UK orders.
        • Another company in the maze was Shield Steel Doors Ltd owned by Remigijus Guobys and Rolandas Guobys, presumably another UK Skydas reseller, launched in 2009 long before their falling out at Skydas.
        • Yet another was Shield EU Security Doors & Windows Ltd, launched in 2014 and wound up in 2018 by the board, run by Nigel Stephens, Fred Drabble and James O’Connor, although originally the third shareholder added in 2015 was Gary Mumford, not Nigel Stevens the company secretary. They used to say that the Fort doors they resold were classified under EN1627 to match product numbers so that FORT001 was RC1 and FORT 007 was RC5, which was nonsense as the FORT007 was an FB4 RC4 and FORT001 was RC2 before it mysteriously became their PAS product, which would mean RC3. They also said their ballistic glass was tested to LPS1270, which has never been mentioned in other literature across the network of companies.
        • Incredibly there was also Shield Front Doors Ltd from 2011-2017, later becoming Magnificent Security Ltd, run by Indre Stankute (Remigijus’ new partner) and Ainis Radzevicius, and owned by her, Robin Carr, Ainis Radzevicius, Clara Natera and Laura Veizbonaite (Remigijus’ then wife), which has been selling Skydas doors rebranded as Magnificent Security, with Gold (Premium), Platinum (Premium FB), Silver (Standart3) and Bronze (Standart2). Indre Stankute also popped up as director of Fort, who were about to be struck off in Nov 2021, alongside Remigijus Guobys and formerly Nigel Stephens, Fred Drabble, James O’Connor.
        • Drabble, Stephens and O’Connor also previously ran Impregnable Security Ltd, which, having morphed from a decorating firm who tidied up after installation, claimed to be Skydas’ reseller for Home Counties and seems to have been active from 2013-4, before leaving for Fort. Stephens, O’Connor and Drabble left to launch J&F Holdings Ltd (dormant company nearly struck off in Jul 2021). Although soon Drabble left, Stephens and O’Conor also ran dormant company Imedia Management until they shut it in 2014 shortly before moving to Fort. Stephens and O’Connor also ran OOC Marketing Ltd. In Jan 2022 O’Connor’s Intimate Leisure (GB) Ltd closed.
        • Stephens and O’Connor still run JOC Marketing Ltd and Exclusive London Models London UK Ltd. O’Connor left to launch Cleveland Carter Ltd, selling security doors with Nigel Stephens. In 2015 Drabble’s brother and nephew set up Cadogan Doors as a Fort reseller, which closed in 2019.
        • Other resellers include Stronghold Security Doors who have been selling similar doors and also used the names Ortus Engineering Ltd (run by Lithuanians Karolis Zilius, Kestutis Vieikis & Justas Samalionis), and Security Doors Factory Ltd (run by Lithuanians Karolis Zilius & Darius Virokaitis).
        • Securehouse have also been selling Skydas doors, until into at least 2019-2021, claiming to be the UK exclusive distributor (which might come as news to Lock Services London Ltd), run by Lithuanians, Dainius Buzinskas and the MD Edmundas Zaunieravicius. Their design options were branded Munitus, with the range branding the Premium as Max, Standart3 as Classic, and sold the FB4, FE5 & FE15 versions, and said they use ESG for their anti bandit glass – which is odd because ESG are based in Essex whereas Securehouse claimed they were importing from Lithuania.
    • Strongdor do a 1.2mm Zintec skinned rockwood core outward opening Securidor range rated RC1-3 (45mm thick, similar to Robust’s SecurDor).
Celeb endorsement for Shield’s Skydas security door
Allegedly Dave Courtney’s door rolling off the production line
Inside a Fort Engineering door
Munitus RC3 door
Munitas double glazed RC3
Munitas RC3 P6B glazing
Munitas RC3 dynamic load
RC3 patio door
RC3 door
Shield/Fort/Skydas forced entry test revealing barrier is really down to one plate, as with most security doors
RC4 door formerly sold by Shield
Gunnebo MagTek RC4
RC5 door
Salzer S4 RC6 (RC5 with safe lock)
STS202
  • WarringtonFire’s Security Technical Schedule 202 gives ratings from BR1 to BR6, similar to LPS1175 in terms of toolsets, but with shorter attack times at the top end with BR1 lasting 1 minute, BR2 3 minutes, BR3 5 minutes and BR4 upwards lasting 10 minutes. Products pass if they stop a cylindroid passing through 500mm long and 380x225mm across.
  • It would be unusual for a security door not to have SBD as well as any insurance standard, so a good starting point for finding suppliers of doors to STS is the SBD door database. SR2/BR2 is the standard requirement for communal doors so unfortunately most BR2 doors are for blocks of flats rather than for insurance requirements in commercial or industrial sectors.
  • Suppliers include:
    • Eurobond Doors do a BR2 steel door.
    • Lathams sell Chinese steel security doors including the 1mm skin BR1 Heavy Doorset (c. £440), 1mm skin BR2 Ultra Multipoint (c. £620) and 70mm thick 2mm skin BR3 High Security Doorset (SBD terror rated, c. £1,100). Some models are not advertised but available to order.
    • Neos Protect do a range of BR2 doors for communal entrances some of which might at a pinch suit a modern house made of concrete or metal.
    • Robust UK sell the Secur-Dor range with SBD: Secur-Dor 2 with STS202 BR2 & PAS (1.5mm skins, 1.6mm frame, 11.5m laminated vision panels on PAS24, 2 x CEN7 deadlocks on BR2), and Secur-Dor3 with BR3 (2mm skins and frame, multipoint).
  • Unfortunately only Lathams and Robust offer BR3 doors but with no wood finish option, meaning STS can safely be ignored for private homes unless perhaps as a slab for an internal door or interconnecting garage door .
BR2 door
LPS2081
  • LPCB have a lower domestic standard, LPS2081, like PAS24, against opportunistic burglars who are not prepared to make a noise or spend over 3 minutes or carry big tools.
  • It requires locks to have 1,000 differs for SRA and 5,000 for SRB, and to use LPS1242 class 1 or BS3621 cylinders.
  • But once SHTF, burglars may not worry about being seen with tools or caught red handed if there are no police or if they are desperate.
  • LPCB say SRB is sometimes as strong as SR1 otherwise sits just below it, while SRA is always below SR1.
  • LPS2081 has two toolsets: A and B.
    • Toolset A has to be resisted for 1 minute and includes cable cutter, glass cutter, knife, spanner, 1’ crowbar, pliers, punch, rubber mallet, screwdriver, traction screws and wedges.
    • Toolset B has to be resisted for 3 minutes and further includes boltcutter, claw hammer, drill bit, hand drill, junior hacksaw, tin snips, pipe wrench.
  • The tools cannot hit each other as silence is assumed. Other than that, they are the tools from LPS 1175 groups A and B.
  • LPCB LPS standards come with a health warning which is they are a monopoly scheme operated by testing house BRE Global Ltd which you cannot inspect and reputedly include rest times. In contrast the theoretically weaker standards PAS and lower EN1627 ratings are set by BSI. In any case, you probably want to avoid LPS2081 for the same reason you avoid EN1627 RC1-3.
  • LPS2081 doorsets are available from Bradbury, Strongdor (glazable inward or outward opening Lumidor), Teckentrup, Design & Supply, Premier SSL (inward or outward opening Jensen 81, SBD 81 & Industrial 81) and Warrior (11.5mm glazing, bare steel communal entrance fully glazed only). Most seem to be communal entrance fully glazed doors or sheet clad, and only Teckentrup and Premier SSL mention LPS2081 on their website. Even police should not accept LPS2081 for low risk blocks of flats as normally the minimum is SR1/BR1.
PAS24
  • Police say PAS24 is as good as SR1 or SR2, but LPS175 allows noisy attacks and breaking glass so it is not clear why they think so, other than that disregarding noise and breaking glass it is similar to an LPS1175 A3, ie longer delay than an SR1 (if the burglar plays the rules of being silent).
  • LPCB say for glazed products PAS24 is equivalent to RC2 and SRA or sometimes SRB, and for unglazed products they say it is worth RC3 and between SRB and SR2 but usually SR1.
  • So-called security doors without LPS1175 or STS 202 or equivalent certification may well be strong, but at the end of the day are not rated to even stop a screwdriver for a minute.
  • PAS24 assumes the burglar won’t break glass or pick the lock or rip the frame out, and only tests quietly wiggling hand tools for 3 minutes; it is merely one way of achieving the bog standard requirement in building regulations Approved Document Q.
  • It even allows thumb turns so a burglar can break a door light and let himself in, although for door glazing it does require 6.8mm laminated glass near thumbturns.
  • PAS24 2016 requires accessible doors and windows without locks to have 6.8mm laminated glass and locking windows to have 6.4mm laminated glass as a minimum; it also requires British Standard locks.
  • Europe has their copycat standard called EN1627 but this does not include locks. PAS24 can be achieved with RC3 plus extra requirements (Annex A Security hardware and cylinder test and assessment, cylinders falling within the scope of EN 1303 shall meet key-related security (digit 7) grade 5 and resistance to drilling security grade 2). In other words RC3 is not even enough for a new build let alone a prepper. That is a reminder why EN1627 doors need RC5 to begin to call themselves a high security door. But nor is PAS24 anywhere near as strong as a brick wall either so it leaves you with a weak spot.
  • Examples of good PAS24 doors include Solidoor, or Rockdoor which whilst not a security door is solid core and can come with mesh reinforcement which forces even firemen to take two minutes cutting out locks with a circular saw. Another is Endurance with their 48mm Kerto LVL-Q plywood cores of 17x 3mm plys. Solidor use skins of Veka’s Coolskin foiled UPVC. None are any use with glazing as it is toughened glass that would safely let intruders help themselves to the thumbturn or climb through.
Foam core plastic doors vs plywood core
Pitiful: PAS24 testers tickle a door

MoD armoury

  • The MoD JS440 Chapter 6 Section III Annex A armoury standard specified products such as a Benweld with hookbolt. But it also contemplated makeshift versions using outward opening wooden doors, faced with steel sheet screwed at 10cm centres along edges and carriage bolted through the door in the centre and corners, frames rag-bolted to 50mm depth on 60cm centres, capped and recessed steel hinges, and two locks from the approved list and Yale WS7 rim mount hinge bolts.
  • Annex B implied that the armoury standard, adopted for explosive stores, used 1.6mm steel sheet and 44mm wooden doors.
  • The Annex C approved locks included old 1950s rim hook locks for legacy doors – 10 lever Ingersols: the SC71, SC73 or D20, or old Chubb 5 detainer models: the Union 3R35 autodeadlocking latch, hookbolt 3M50, sashlock 3K70 (sashlock) or deadlock 3G110, or ASSA cylinder Twin 6000 with deadlock 9788.

US military

  • Manual attack specifications as per the ‘walls’ post advise:
Level II low threat
  • Hollow metal 1.6mm steel 4” centres hat-stiffened 1.6mm steel faced doors with 1.9mm steel channel with recessed edge, in double rabbet 4” jamb fully welded 1.9mm steel frame. Add more doors rather than upgrade one door for delay beyond 1 minute. Hinge bolts and two locks, one being a mortice deadbolt and the other a rim drop bolt with anti wedge and anti drill plates, rated to 5.45t; or
    • Take the following delay times from constructions of:
      • 1 minute for garage door from 1.2mm galvanised steel roller door,
      • 1 minute (also level IV) from 3/8” steel plate on 20 gauge (0.81mm) steel plate with two padlocks,
      • 1 minute from 16 gauge (1.34mm) metal rollup door,
      • 2 minutes from 16 gauge (1.34mm) hollow steel door,
      • 2 minutes (also level III & V non-explosive) from ¼” steel plate with padlock,
      • 2 minutes (also 4 mins at level V (sic) non-explosive) from ¾” steel plate on 3” rigid insulation on 1/8” steel plate,
      • 4 minutes (also 2 mins at level V non-explosive) from US class 5 vault door,
      • 5 minutes (also 1 min at levels III-V) from 16 gauge (1.34mm) hollow steel door with hinge bolts, astragal and drill resistant deadbolt,
      • 6 minutes from two layers of 10 gauge (3.12mm) hot rolled steel on ¾” plywood,
      • 6 minutes from 3/8” steel on 3” redwood on 0.036 steel,
      • 11 minutes from ¼” stainless steel on 0.5” polycarbonate on 10 gauge (3.12mm) mild steel,
      • 12 minutes from three layers of 10 gauge (3.12mm) hot rolled steel on two layers of ¾” plywood,
      • 14 minutes from sandwich of two layers of 9 gauge (3.5mm) hardened steel and one layer of ¾” plywood,
      • 15 minutes from 12 gauge (2.36mm) hollow metal door filled with lightweight concrete with hinge bolts, astragal and drill resistant deadbolt,
      • 20 minutes from sandwich of two layers of 9 gauge (3.5mm) hardened steel and one layer of ¾” plywood lined with two layers of 0.9lb per square foot gravel finish roof paper,
      • 20 minutes from sandwich of 3 layers of 10 gauge (3.12mm) hardened steel and 2 layers of 0.5” acrylic,
      • 20 minutes from sandwich of 3 layers of 10 gauge (3.12mm) mild steel and 2 layers of 0.5” polycarbonate,
      • 27 minutes from sandwich of 9 layers of ¾” plywood and 8 layers of 10 gauge (3.12mm) steel plate,
      • 30 minutes from 12 gauge (2.36mm) hollow metal door with hinge bolts, astragal, drill resistant deadbolt and 3 point locking.
Level III medium threat
  • Prison door to HMMA 863-14 with 4 minute delay in concrete block wall; hollow metal 1.9mm steel 4” centres hat-stiffened 1.9mm steel faced doors with 2.7mm steel channel with recessed edge and 7 gauge (4.24mm) astragal, in 3,000PSI grouted double rabbet 8” jamb fully welded 2.7mm steel frame. Hinge bolts and two locks, one being mortice and one rim with anti wedge, and anti drill plates rated to 5.45t; or
  • Construct from:
    • 1 minute from 16 gauge (1.34mm) hollow metal door with hinge bolts, astragal and drill resistant deadbolt,
    • 4 minutes for garage door from 1.75” hollow metal door with 12 gauge (2.36mm) skins and stiffeners on 6” vertical spacing (ammunition dump door),
    • 5 minutes from 12 gauge (2.36mm) hollow metal door with hinge bolts, astragal and drill resistant deadbolt,
    • 10 minutes (also 4 mins at level V) from ¾” steel plate on ¼” steel plate,
    • 14 minutes (sic) (also 4 mins at level V) from 3/8” steel plate on ¼” steel plate,
    • 15 minutes from 10” thick 0.5” metal sheet skinned door, with internal locking,
    • 22 minutes from 3/8” drill resistant steel plate GSA class 6 vault door,
    • 26 minutes (also 17 mins at level V) from ¾” steel plate on 3” redwood on ¼” steel plate,
    • 30 minutes from 6” thick 0.5” metal plate door filled with lightweight concrete, with internal locking.
Level IV-V
  • Requires vault doors or use multiple level III doors; or
  • Construct from:
Level IV:
  • 5 minutes from 12 gauge (2.36mm) hollow metal door with hinge bolts, astragal, drill resistant deadbolt, filled with lightweight fireproofing,
  • 15 minutes from a 10” thick 0.5” plate skinned door, filled with lightweight concrete, with internal locking,
  • 30 minutes from a 10” thick 0.5” plate skinned door, filled with lightweight concrete reinforced with exmesh, with internal locking.
Level V:
  • 3 minutes from sandwich of 3/16” steel plate, 1” insulation and two layers of 22 gauge (0.66mm) sheet metal,
  • 5 minutes from 6” thick hollow metal door with 0.75” steel front plate and 0.25” back plate, filled with lightweight concrete reinforced with exmesh, with internal locking
  • 12 minutes from sandwich of 0.5” hardened steel, 1.25” oak, 0.5” plywood and two layers of 0.25” hardened steel,
  • 15 minutes from 10” thick hollow metal door with 0.5” skins, filled with lightweight concrete reinforced with exmesh, with internal locking,
  • 16 minutes from sandwich of two layers of ¼” hardened steel, two 3” layers of urethane foam, ¼” perforated steel plate and 1.5” oak,
  • 17 minutes from sandwich of two layers of 3/8” A36 (low carbon) steel, 2.125” polyurethane foam, 0.25” perforated steel and 2.5” oak,
  • 18 minutes from sandwich of two layers of A36 steel plate, 4” polyurethane foam and 3” oak,
  • 19 minutes from sandwich of two layers of 3/8” A36 steel, 3” oak and 4” silicate foam,
  • 23 minutes from sandwich of two layers of 3/8” A36 steel, 4” silicate foam and 3” oak,
  • 30 minutes from a 10” thick hollow metal door with 0.5” skins, filled with lightweight concrete reinforced with exmesh, with internal locking and welded C steel grating vestibule for standoff.

US prison

Brig
  • USA brig doors are expected to be hollow metal and withstand 6t overall or 3t on an unbolted corner.
Civilian
  • US hollow metal prison doors must pass the ANSI/NAAMM HMMA863-14 test for static load in the centre, a rack test for load on unlocked corners, an impact test against ramming for as long as a riot would last, and bullet resistance to UL752 level 3 handgun.
  • There are four grades: grades 1-2 need 2.3mm skins while grades 3-4 only need 1.7mm skins.
  • Door thickness to be 2”.
  • Vertical stiffeners to be spot welded at 3” centres and spaced at 4” centres and 1mm thick.
  • Airspace to be filled with fibreglass or rockwool slabs.
  • Edges to be reinforced with 3.1mm thick channels.
  • Reinforcing plates are required for hinges and strikes.
  • Hinges to be 5” from top and 10” from bottom plus equally spaced intermediate hinges.
  • Keyways to be 46” up.
  • Jamb anchors to be every 18” + 2 per length.
  • Frames to be filled with mortar grout with 4” slump.
  • Masonry straps to be 2”x10”.

Sandia research

  • Nuclear safety findings from the 1970s says:
    • GSA grade VI vault door (3/8″ steel plate protecting insulation and 1/10″ plate) lasts 9m30s against a 200lb rocket torch.
    • Insulated steel plate door of 0.75″ steel sheet behind and 0.25″ in front with hinge bolts and deadlocks lasts 1m30s against a few pounds of explosives.
    • 18 gauge hollow steel non-security doors are vulnerable to the handle being taken off with a pipe wrench in 24s, a hole being cut out by a fire axe in 3m48s, or, the fastest of all, being jemmied open with a 15lb crowbar in 12s. Thicker plates protect against the slowest risk, whereas what is most desperately needed is bolts on all edges, keylocking deadbolts on the leading edge and maybe stronger frames.
    • 16 gauge corrugated steel roller shutters last 48s against a crowbar and 2×4.

US security commercial hollow metal doors

  • US hollow metal security commercial doors must pass the ANSI/NAAMM HMMA863-14 test. There are six grades from 1 to 6, with 3-6 being potential anti terror level.
  • Testing is done to prison standards ASTM F1450 and ASTM F1592, embassy ballistic & manual attack standard SD-STD-01.01, ballistic standard UL752 levels 3/7/8 and British insurance manual attack standard LPS 1175.
  • Tests include static load at corners and at and between locks, soft impact (shoulder for classes 1-3 simulated by cushioning the ram with 2” thick 6” diameter 32kg/m3 rigid foam polystyrene) and hard impact (sledgehammer for classes 4-6 simulated by 36kg 4” square (2.5” diameter) ram delivering 271J), jamb-to-jamb stiffness, beading removal, edge crush, forced entry and ballistic.
  • Ballistic testing is up to AP rifle.
  • Static load testing includes trying to push out glazing by hitting the fixed stop side so pushing against removable beading.
  • Frames are tested against spreading with a hydraulic ram.
  • Edge crush is tested with a hydraulic ram with 1.5” diameter.
  • Skins for grades 1-3 are 1.7mm and for grades 4-6 are 2.3mm, continuously welded.
  • Minimum door thickness is 1.75”.
  • Reinforcement and anchorage to be as for prison doors.
  • Grades 3 upwards use a mob of 6, with 2 personnel for grade 1 and 4 for grade 2.
  • Each grade uses a higher tool group from A to F with increasing attack times from 5 through 10/15/20/40 to 60.
  • Static loads at edges only needs to be about 3kN deflecting 30mm for grades 1-2, about 6kN deflecting 20mm for grade 3, about 12kN for grade 4 and 20kN for grades 5-6 with deflection up to 10mm for grades 4-6, suggesting vulnerability to hydraulic rabbit rams.
  • However, locks have to resist around double that load at about 6kN at grades 1-2, about 12kN at grade 3, about 20kN at trade 4, and about 29kN at grades 5-6, and at all grades deflection is limited to 10mm.
  • Impacts have to be resisted to two blows of 80/120/160J for grades 1-3 soft body and 200J for grade 4 and 271J for grades 5-6 hard body, with 30 blows to glazed points and mulled systems, 30/75/100 blows to hinges and 100/200/300 blows to locks.
  • Edge crush of 0.25” has to be resisted to about 6kN for grades 1-2, about 9kN for grade 3, over 15kN for grade 4, over 25kN for grade 5 and over 35kN for grade 6.
  • Jamb to jamb stiffness has to resist 6kN with less than 0.375” deflection at grade 1 or over 8kN at grade2, while grades 3-6 are allowed up to 0.5” deflection from 16kN at grade 3, over 22kN at grade 4, over 26kN at grade 5 and over 33kN at grade 6.
  • This standard uses longer attacks so exceeds the old LPS1175 SR1-6 and instead is closer to the new LPS 1175 A5/B10/C15/D20/E40/D60.
  • There is also the civilian standard ASTM3038-14 which is similar to SD-STD-01.01.

SD-STD-01.01

  • The USA state department (‘DoS’) have a standard for embassy doors called SD-STD-01.01 which is available with FE (forced entry) or FEBR (plus ballistic) rating for 5, 15 or 60 minutes. They are designed to stop ‘social disorder’ with a mob attack by young men with guns or makeshift rams and levers. Perhaps surprisingly, an FE5 is only roughly equivalent to an SR3 (slightly weaker against LPS1175 tools and techniques but stronger against less sophisticated heavier tools). Even FE60s tend to use morticed locks, although this makes sense as more severe tools are not tested against locks but instead the door must withstand a battering so has to be strong enough to protect the lock unlike a typical LPS1175 door which uses rim locks behind thinner metal. However, Skydas secured SD-STD.01.01 FE15 with a door using a central plate and lock carrier exposed once the decorative skin is torn off.
  • SD-STD-01.01 doors must open when unlocked with under 12lbs of force, to ensure usability.
  • SD-STD-01.01 doors fail forced entry (‘FE’) rating if any mounting fails or a 12″ x 12″ x 8″ cuboid or 12″ x 12″ cylinder can be passed through. This would appear to unfairly mean it fails even if it is still locked solid by bolts both sides but a hinge has been chopped off. DoS can test components that failed or passed to see how long they last, and, for example, you sometimes see FE15 doors that actually lasted, say, 56 minutes on ‘test to failure’ so just missed out on a FE60 rating. The tool group includes a two man 4″ x 4″ 120lb ram, two 30″ 10/12lb sledgehammers, wedges and two 5′ crowbars, plus for ballistic versions a rifle and shotgun.
  • The issue with SD-STD-01.01 for householders is that it exceeds the strength of a brick wall without protecting against severe tools, so is mainly only suitable for reinforced masonry buildings where only brute force is expected or where you just need delay to run to a bunker. However, some FE5s & FE15s come with other accreditations, as did the Skydas Embassy FE5 (with RC3) and Skydas Fortress FE15 (which they said lasted 77 mins against RC4 tools, but that was spot testing under ERN1627), giving at least some protection against a wider toolset and thus more suited to maintain consistent protection across the facade of a reinforced masonry building. This might also come with options for wood finish or bullet resistant glazing where the product is doubling up for a more residential market where, for example, there is a EN1627 accreditation. The Skydas FE5 centre bolts were taken out with a giant crowbar after 210s but the top bolts held – perhaps due to testing an 8′ tall door. If testers could have used a saw against the top and bottom bolt rods and C channel stiffeners a crowbar would have taken care of it, but it was a great design for the standard which got the last laugh when a tester ran up to it with the test block thinking he could shove it through at any moment only to find the door literally put the tester on his arse when the five footer crowbar sprang off.
SD-STD-01.01 FEBR 5 & 15
Skydas Embassy SD-STD-01.01 5FE
Skydas SD-STD-01.01 5FE
Skydas Fortress SD-STD-01.01 15FE
Skydas FE5 & FE15
Harden Architectural Security Products (formerly Ross) FEBR60

EN1143

  • If a security door such as the Salzer S4HS to RC6 or the Burton Safes Orthrus to SR6 is not enough (for your bunker, strongroom or modular strongroom), then your next level up is a strongroom door.
  • Strongroom doors are certified to EN1143 grades 0-XIII, with CD (for grades VIII upwards) and EX upgrades (for grades II-XIII) available against diamond core drill and explosives. A T2 rating is additionally available for a new toolset, but which for strongroom door tool category S only adds a concrete chainsaw.
  • Grades use ‘resistance units’ which roughly increase 50% each grade so a grade XII is 100 times stronger than a grade 0.
  • Examples include:
    • Salzer can do a 300kg 117mm thick S4W1 up to grade I.
    • DorukSafe have the DorukSafe Level XI CDEX (rated against diamond core drill and explosives) for the Bond villain lair touch.
    • Independent Safes sell Robur and Securikey modular vault doors and 8mm / 10mm / 25mm plate gunroom doors.
    • Burton Safes make Cerberus vault doors from grade I (790kg) to grade XIII (4.3t) (which do not seem to lock from inside) and Orthus security doors (16cm thick) up to SR6 & G5 and Kronos security doors up to SR4, and resell Janus residential security doors up to RC4. Barrington Security also mention Burton Safes grade V strongroom doors made of 2mm steel plate each face with internal locking.
    • Securikey Modulprim vault doors range from 29-170mm thick and 380kg-2.3t for grades 0-X, but cannot lock from inside.
    • Kaso in Finland make vault doors grade II-XII weighing 320kg to 2.8t and 62-300mm thick.
  • Strongroom doors typically weigh from 300kg for a grade I to 2.8t for a grade XII, with thickness from 15cm for grade 1 to 50cm for grade XIII. As examples, Robur’s grade I weighs 420kg and grade II weighs 800kg.
  • The boltwork is only covered with a plate to stop spying on it when open.
  • There is also a similar UK safe standard LPS1183 which covers strongrooms but LPCB have never certified any.
  • Both EN1143 and LPS1183 require locks to EN1300. This requires one class A lock up to grade II, one class B for grade 3, two class B up to grade V, two class C up to grade X, and two class D up to grade XIII (electronic combinations) or three class C for grade XI-XII (so you can use key locks). Safe (and thus strongroom) key locks are typically from Germany such as Dormakaba (Mauer) or France, Fichet Bauche. The ECB maintain a directory of EN1300 safe locks. A main supplier for Mauer and LA Gard is Safelock Systems. The downside of avoiding combination locks is that a clever enough locksmith could pick all three key locks, hence the importance of being able to override locks from inside. Perhaps surprisingly, 7 or 8 lever safe locks only cost £20-40.
Panic room using EN1143-1 T2 strongroom door disguised with mirror

Ballistic

Standards

  • The UK ballistic standard for barriers such as doors is BS EN1522 and EN1523, ranging from FB1 for .22 rifles, through FB2 for 9mm Lugers, FB3 for .357 Magnums, FB4 for FB3 + .44 Remington Magnums, FB5 for 5.56 x 45 rifle, FB6 for FB5 + 7.62 x 51 rifle, to FSG for 12 bore x 70 shotguns firing 31g solid lead slugs and FB7 for 7.62 x 51 AP rifle. For glazing it is BS EN1063 with equivalent prefix BR ratings which are paired with the EN1522 door if glazed and FSG mist be matched with SG2 glazing. Except for SG1 for glazing, all ratings in EN1522 and EN1063 test three shots.
  • Another European standard, administered from Germany, is VPAM 2006. Its levels stretch from PM1 to PM14. They add level PM2 for low speed 9mm handguns between FB1 and FB2, PM4 covers FB3 & FB4, PM5 is for high speed .357 Magnums and sits above FB4, PM6 for Kalashnikov sits above FB4, PM7 covers FB5 & FB6, PM8 for Kalashinkov AP sits above FB6, PM9 matches FB7, PM10-PM14 covers high speed heavy AP rounds such as 7.62 calibres, .50×99 Browning and 1.45×114 anti tank machine gun. It does not have shotgun levels, although a PM7 ought to handle it.
  • The US equivalent is UL752 which goes up to level 10 for a .50 rifle, or there is NIJ 0108.01 going up to level IV for a 30-06 AP rifle. The FEBR combined manual attack standard SD-STD.01.01 also covers firearms.
  • Ballistic products often also come with fire, blast and manual attack ratings.
  • Threat energies of small arms roughly cluster in bands up to 250J, 500J and over 750J. This is relevant to glass, as bullets pass through already broken glass so you measure total energy dumped into the pane. For other materials you measure energy density, which for small arms tends to concentrate into bands of up to 5J/mm2, 10J/mm2 and 20J/mm2, and in respect of which Magnums are one of the biggest threats. The energy density of army rifles largely concerns 5.56 (1,700J muzzle energy) and 7.62 calibre (around 2,350J), both of which have similar energy densities of about 70J/mm2. Hunting rifles are a bigger problem for armour as not only does their muzzle energy typically range from 3,000-5,000J (like a 7.62×54 rifle) but their energy density can go up to about 125J.mm2 such as the 7mm Remington.
  • VPAM have a guide to choosing protection level.
  • Check what is promised about products: was it just tested, or actually certified, or was it subjected to a computer simulation? Whether you can rely on missing certification depends whether bullets are your main priority and how obvious it is that the manual attack strength will more than handle bullets.

Tips

  • Beware weaknesses at edges, joints, locks and hinges. This is why security doors mount hinges and locks on the surface so they are not gouging out strength from the leaf core. EN1522 tests the joints more than centres of leafs.
  • Handguns and some rifles do not blow unarmoured mortice locks out, but shotguns and some rifles do. Locking knobs may be easier to blow away though.
  • In ballistic attacks you usually want to hide occupants and consume ammunition, so any glazing should have removable screens (eg curtains) and barriers ideally strong enough to not just stop one bullet to pass a test, but to enthusiastically eat up the whole magazine.
  • You can buy electric glass linked to gunshot detection which triggers into privacy mode on hearing a gunshot.
  • For high threat levels the most practical door protection may be a medium threat door backed by a high threat lobby wall.
  • For FB5 or NIJ0108.01 upwards you might want to use high hardness steel so, for example, you stop a Nato rifle with only 6.5mm weighing only 52kg/m2. Otherwise normal steel weighs 23kg/m2 for FB4, 48kg/m2 for FB6 and 120kg/m2 for FB7.
  • To reduce weight you may have to pay double for Kevlar instead of armoured steel.
  • Normal steel skinned doors will not stop any bullet.
EN1523 target directions – testing for weak joints
Privacy glass has a controller which you can connect to gunshot detectors or IoT devices
Hardened foyer layout from UFC direct fire design guide

Products

Doors
Aluminium foam is a candidate material for ballistic panels
Fittings
  • You can buy ballistic & manual attack escutcheons such as from SurelockMcGill.
Architectural Armour ballistic door viewers
Citytsafes Modul-X SD-STD.01.01 FEBR60 vault door (reinforced concrete)
SurelockMcGill locks holds a FEBR60 for an hour

Materials

  • Ballistic glazing in doors is accredited to EN1063 up to BR7 for rifles or SG2 for shotguns. If using laminated polycarbonate glass the weight for BR4-BR7 is from 52-170kg/m2 and thickness is 24-74mm, and 52/83kg/m2 and 24/38mm for SG1/2. If using laminated glass the thickness for BR2-BR7 NS (no spall) is 31-81mm and weight 73-196kg/m2. If using laminated polycarbonate you need 18mm Lexguard MP750 for a 9mm handgun, 26mm BBPOL25 for a .357 to UL752 level 2, and 32mm MP1250 for a .44 or 9mm Uzi to UL752 level 3, weighing 23-38kg/m2.
  • Door leaf ballistic protection is usually provided by steel, technical ceramics (like silicon carbide, boron carbide, titanium diboride, aluminium nitride, silicon nitride, aluminium oxide, tungsten carbide) or fibre (like Kevlar, Spectra or fibreglass). Glass in itself is only used against shaped charge jets. Ballistic materials have also been discussed above under Materials in terms of ballistic cores and armour, as some double up as anti cut materials.
  • A candidate ballistic material is composite metal foam. Good results have been obtained from 2mm steel spheres in steel foam backed by 7075-T6 high tensile strength aluminium and faced by boron carbide with layers glued with aerospace grade Loctite EA 9309NA, ideally with the boron carbide protected by an attack face of rolled homogenous armour (1.8% nickel 0.65% chromium 0.2% molydenum steel alloy).
Ceramics
  • DSTL have been investigating silicon boron carbide as a solution to the brittleness of boron carbide (whose weak polytypes suffer phase collapse under high velocity impact over 20GPa, behaving like graphite and a liquid) by doping it with silicon carbide, and looking at nanocomposites with powdered silicon carbide and boron carbide to reinforce alumina armour, and aluminium filled ceramic foam as a backing for brittle boron carbide.
  • Other work has been done on using laminated ceramic because multiple thinner tiles produce smaller tensile reflections which damage the bullet more, as the ceramic survives a fraction of a second longer since dwell time (time bullet spends on surface before tile breaks) depends on thickness rather than density.
  • One way in which ceramic brittleness lets it fail on impact is that its spall strength, unlike metal, is way lower than its Hugoniot elastic limit.
  • Candidate ceramics include:
    • Silicon carbide, which is usually hot pressed, with about 4um grain size, about 3.2t/m3 density, about 420GPa Youngs modulus, about 0.17 Poisson ratio, about 175GPa sheer modulus, about 4GPa compressive strength, about 0.345GPa tensile strength, about 12GPa HEL and about 0.6GPa spall strength.
    • Boron carbide, which is usually hot pressed, with about 3um grain size, about 2.5t/m3 density, about 400GPa Youngs modulus, about 170GPa shear modulus, about 0.17 Poisson ratio, about 4GPa compressive strength, about 16GPa HEL and about 0.6GPa spall strength.
    • Titanium diboride, which is usually hot pressed, with about 30um, grain size, about 4.5t/m3 density, about 530GPa Youngs modulus, about 240GPa shear modulus, about 0.1 Poisson ratio, about 4.5GPa compressive strength, about 10GPa HEL, about 0.45GPa spall strength.
    • Aluminium nitride, which is usually hot pressed, with about 2um grain size, about 3.25t/3 density, about 320GPa Youngs modulus, about 130GPa shear modulus, about 0.238 Poisson ratio, about 3GPa compressive strength, about 9GPa HEL and about 0.5GPa spall strength.
    • Silicon nitride, which is variously sintered, hot pressed or reaction bonded, with about 0.5um grain size, about 3.5t/m3 density, about 269GPa Youngs modulus, about 106GPa shear modulus, about 0.267 Poisson ratio, about 2.7GPa compressive strength, about 0.81GPa tensile strength, 2-12GPa HEL, about 0.5GPa spall strength.
    • Aluminium oxide, which is typically hot pressed, with about 15um grain size, about 3.85t/m3 density, about 370GPa Youngs modulus, about 150GPa shear modulus, about 0.24 Poisson ratio, about 3GPa compressive strength, about 0.25GPa tensile strength, about 9GPa HEL, about 0.45 GPa spall strength.
    • Tungsten carbide, which is typically sintered, with about 1um grain size, about 15t/m3 density, about 645GPa Youngs modulus, about 270GPa shear modulus, about 0.21 Poisson ratio, about 5.1GPa compressive strength, about 0.25GPa tensile strength, about 4GPa HEL, about 3.1GPa spall strength.
    • Glass has about 2.5t/m3 density, about 75GPa Youngs modulus, about 30GPa shear modulus, about 0.23 Poisson ratio, about 1GPa compressive strength, about 0.15GPa tensile strength and about 6GPa HEL, so is relatively floppy, brittle and weak, albeit cheap and hard enough to help blunt bullets.

Blast

Standards

  • The blast door standard is arena test to EN13123-2 EXR1 to EXR5 for 3 to 20kg TNT at 3 to 5 metres, with glazing to EN13541 or ISO 16933 for arena tested glass, or EN13123-1 for shock tube test. ISO16933 has grades EXV45-EXV10 for pressures of 30-800KPa at 45-10m and grades SB1-SB7 for pressures of 70-2,800KPa (equivalent to 3-20kg TNT) at 3-9m, with SB3-SB7 equivalent to EXR1-EXR5. For glazing there is also the US blast safety standard GSA-TS01:2003 which simply rates damage for a given blast so is not a strength rating. US military standard UFC4-010-01 requires exterior doors to be outwards opening if no other standard applies and sets damage grades from very low to high, as does ASTM F2247-11 with response categories from I to IV, and ASTM 2912-11 for glazing from no break H1 to high hazard, and ASTM2927-12 for doors from category I-V.
  • If you want a blast door to still resist manual attack afterwards then you need to specify what in USA standards would be ‘no break’ for glazing and ‘category 1’ / ‘high protection’ for doors so it is not damaged by your blast design threat.
  • Blast ratings are for transported terrorist bombs at a distance. You will not buy a door off the shelf that can see off explosives attached to it, such as an Alford Strip, Gatecrasher or the old police favourite the Breachers Boot. You will struggle to buy a door off the shelf stronger than the highest blast standard, but you can commission them, eg against closer or bigger bombs like nukes, but of course they need bunker walls as by coincidence normal house walls collapse under bigger blasts than the highest blast standards, and for example, a blast door rated for 1-2MPa weighs about 1t/m2. Check wall strengths in the wall chapter.
UK military urban door breach grenade launcher
17.5PSI test

Materials

  • Glass:
    • Blast glazing as non spall laminated glass for ER1-ER4 is 18-33mm thick and weighs 40-83kg/m2. To stop fragmentation needs it upgrading to ballistic rating too.
    • Glazing is normally best avoided in doors as it prevents stiffeners spanning between edges.
  • Wood: Standard wooden doors can be reinforced with bolts and bars to reduce flying fragments but it still lets blast pressure in and makes little difference to safe scaled distance. Hardwood ply is, however, used as the core in many counter terror doors, especially internal where there are no insulation regulations.
  • Plastic: Chinese scientists found success with a blast door resisting 0.45MPa made of sheet molding compound (resin) reinforced with carbon fibre reinforced plastic (Basalt, resin, calcium carbonate etc).
  • Steel:
    • Steel security doors tend to avoid fragments down to scaled distance of 6m/kg0.33; closer than 4m/kg0.33 you need blast doors (by which point a tiled roof will have come off).
    • Doors that are strong against manual attack may be too stiff to take a blast, hence one design is the USA Air Force Research Laboratory ‘s Accordian-Flex design that can expand from 48″ to 114″ with damping chamber, and frames and doors benefit from being weak enough to act as a springs in series (see Keene).
    • Steel blast doors are usually made hollow and filled with stiffeners: either a honeycomb slab or steel channels. Structural rigid insulation core has been tested to work better than steel reinforcements, but would not be acceptable for manual attack or ballistic purposes – which blast doors often have to also serve.
    • Although concrete filled steel doors do better than hollow ones, curved hollow steel doors using arches or even multi-arches could do better.
    • Blast faces should normally be something like what in USA would be steel to ASTM A515 grade 50 with 200GPa Youngs modulus, 0.3 Poisson ratio, 265MPa static yield, 492MPa ultimate tensile strength.
    • Steel stiffeners:
      • Steel stiffeners are increasingly fins or strips, but typically hat profiles which are strongest for blast.
      • Not too many stiffeners should be used or the door will tear where plastic deformation is prevented, whereas it would have been better to have fewer and let the skin’s plastic deformation soak up the energy.
      • Stiffeners for blast work best when horizontal, but best of all when teamed up with structural rigid insulation. However, tests on square plates found that not only are stiffeners better than thicker plates, but cross stiffeners are best, ie forming squares – see Matsagar.
      • Hat stiffeners work best with thick loaded skin and deep stiffeners on it; the protected skin can be thinner and use shallower stiffeners.
      • Vertical stiffeners need extra strong welding to the hollow steel section acting as bottom rail as the door bottom is vulnerable to being blasted past the threshold as explosion tends to come from bombs on the ground or reflect off it.
      • Hollow steel sections are sometimes used instead of hat profiles, and absorb blasts by deformation along the centre of its length on the blast face as well as deflection in a slightly more V than U shape due to the middle fairing worst, but not tearing as they are mainly simple supported on bolts – so not clamped. They work best when deep, narrow (to minimise surface area) and long (long beams deflect less than short ones as they are heavier and soak up more energy in deformation). This suggests a hollow steel blast door using hollow steel sections needs to be thick to allow deep stiffeners to cover the whole span in one piece.
      • Unless stiffeners are welded strongly enough to the loaded skin it will detach under compression, causing wrinkling, pushing in on one side of the failed stiffener and pulling out on the other.
      • Welds should ideally be fillets rather than plug, and at 3″ rather than 6″, especially on the loaded skin (attack face under compression).
      • Stiffeners partly turn the plate into many plates, each of which deflects into the rectangle between stiffeners, whilst the stiffeners deflect with the door in the one overall ballooning inwards (see Salomoni, Mazzucco & Xota).
      • Stiffeners should normally be something like what in USA would be steel to ASTM A36 with 200GPa Youngs modulus, 0.29 Poisson ratio, 7.8t/m3, 248MPa static yield, 400MPa ultimate tensile strength.
Torterello door hat stiffeners being welded
  • For a blast door, according to the Steel Door Institute you need to specify:
    • Blast pressure: The maximum pressure (psi) expected to be exerted on the assembly by the projected blast event.
    • Blast duration: The length of time (milliseconds) required for the blast pressure to decay to zero.
    • Blast impulse: The blast energy as described by the area under the pressure vs. time curve (measured in pressure-time units such as psi-msec).
    • Blast direction: The direction of the blast load relative to the door assembly.
    • Rebound: The percentage of the initial peak blast pressure that is reflected back to the blast resistant unit.

Products

  • Assa Abloy do a Powershield steel door range that includes their strongest non-manual attack blast doors like the Goliath and Causeway. Their Fenwick and Wellington timber security doors can also be upgraded for blast. They can also commission doors beyond 12PSI for the public and beyond 100PSI for government.
  • Blast & Ballistics do blast doors.
  • Bastion do a CPNI blast door as part of their BastionWall dismountable strongroom, also available in SR4.
  • Safetell do the 80mm thick Stalwart to EXV25 & FB3/4 and EXV28 & FB4.
  • Sommer do the 142mm thick OST2 blast door to RC6 and FB7 ballistic with glazing option, and the OST2 RC5 with blast and FB7 ballistic (needs 24cm brickwork).
  • You will struggle to find a residential blast door but some manufacturers of residential manual attack and/or ballistic doors can add blast rating, such as Assa Abloy’s Safehaven and Robust’s Securdor.
  • A blast hinge that performed well in a blast test is the Surelock McGill S2HA7, as used on manual attack, ballistic and counter terror doors like Stafford Bridge’s Whitehall 2016 CPNI door.
  • Hinges need full length fillet welds as opposed to tacks.
Surelock McGill HA range heavy duty security hinges

Accessories

Locks

Standards

  • The baseline EU lock standard is EN1303 for cylinders and EN12209 for locks.
  • The baseline UK lock standard is BSx621, which upgrades EN12209 and EN1303 with the General Vulnerability Test for bumping etc.
  • The UK high security lock standard is LPS1242 but you will struggle to find any such certificates. The standard upgrades BSx621 with a manual attack test derived from LPS1175. In practice, LPS1175 fabricators schedule approved locks for their certificates which they say are as strong as the lock tested as part of the doorset. This list is well known and trusted, including by CPNI, although they tend to stick to the same few rim multipoints from Surelock McGill, for example, for MFES certificates.
  • SBD accreditation requires TS007 3 star, SS312 Diamond, BSx621 or PASx621.
  • The Guild of Architectural Ironmongery offer a specifier’s guide to security hardware.

BSx621

  • BSx621 dates back to the 1960s and is the UK insurance standard for domestic locks. It is basically EN1303 plus the BSx621 General Vulnerability Assessment, meaning there is not much point buying EN1303 as you still need BSI Kitemark or LPCB to run a LPS1242 test to check it is as good as BS3621. The 2017 version was updated to include changes to EN1303 from 2015 and EN12209 from 2016.
  • Locks should be to BS3621 if locking inside or BS8621 if using a thumbturn, and single locks should morticed a third of the way up plus a rim or mortice lock a third of the way down.
  • BS3621 requires cylinders to be to BS EN 1303 grade 5 key security and grade 2 attack security.
  • PASx621 standards is BSx621 for multipoints.
  • BS1021 is for external doublelocking whereby you can disable the inside keyway so a burglar cannot use a key to escape through the door if they broke in elsewhere, found your keys and want to stroll out the door with their swag.
  • BSx621 is tested in 2″ thick oak, so such locks are probably stronger than any non-security door.

EN12209

  • Locks can additionally be to BS EN12209. EN12209 has security ratings called box 7 (grades 1-7) and 11 (grades 0-H) for resistance to drills, wrenches and crowbars, and for bolt projection and key security. These only cater for locks up to about a high security deadlock you would see on a warehouse such as a Chubb 3K77 7 detainer.
    • For projection, grade 1 has 10mm, grade 2 = 12mm, grade 3 = 14mm and grade 4 upwards = 20mm.
    • For side load, grade 1 = 1kN, 2 = 3kN, 3 = 5kN, 4-5 = 7kN and 6-7 = 10kN.
    • For end load, grade 1 = 1kN, 2 = 2kN, 3 = 4kN, 4-5 = 5kN and 6-7 = 6kN.
    • Grade 5 upwards are drill resistant for 5 minutes before the bolt is forced.
    • For key security in terms of detainers, grade A = 3, B-C = 5, D-E = 6, F-G = 7 and H= 8.
    • For differs, grade A = 100, B =1,000, C = 10,000, D-E = 20,000, F = 6,000, G = 50,000 and H = 100,000.
    • BSx621 equates to EN12209 security rating 7B which under EN12209 is maximum manual attack security although EN12209 also caters for over eight levers compared to five needed for BS3621.
Adams Rite EN12209 grade 6 metal door deadlock

LPS1242

  • LPS1242 is the LPCB standard for security cylinders and ties in with BSx621, EN1303 (cylinders), LPS1175 and PAS24.
  • It is effectively EN1303 plus ratings for bumping, manual attack and key patent. The EBN1303 elements are really only key security grades 1-9 and attack resistance grades 0-5.
  • It sets minimum differs, detainers, steps at same level (two allowed at grade 1-5) by grade.
  • Keys can only be directly coded if grade 1-2.
  • Keys must be patented at grade 6-9.
  • Cylinders and keys must be registered at grades 5-9. Keys at grade G-H must be registered in an LPA1224 secure database register. Keys for cylinders at grades 7-9 must only be issued by the cylinder manufacturer to registered keyholders requesting in writing. Keys for cylinders at grades 5-6 must only be issued by the cylinder manufacturer or their authorised agent to registered keyholders requested in writing.
  • Master keys are not allowed at grades 8-9.
  • Cylinder and key must have the strength and durability in EN1303.
  • Cylinders to grade 1-6 must meet EN1303 para 4.8.6 for mechanism security plus grades EN1303 4-6 for LPS1242 grades 7-9.
  • Cylinders must have torque resistance to EN1303 para 5.8.6 and must resist a drill, chisel or wrench turning the cam 360 degrees, or retracting the bolt, at 5Nm and must not project by 3mm.
  • Attack resistance breaks down into:
    • Cylinders have to resist 1.5Nm torque by a key of the next differ.
    • Drill resistance is tested to EN1303 para 5.9.1 with LPS1242 table 1 bits up to 0.5″ and must not allow a torque of 5Nm to retract the bolt.
    • Chisel resistance is to EN1303 para 5.9.2 and LPS1242 table 6 with a 6kg hammer dropping 70cm.
    • Twist resistance is to EN1303 para 5.9.3 to number of twists in LPS1242 table 6.
    • Plug extraction resistance against traction screws is to EN1303 para 5.9.4 with the force and time in LPS1242 table 6.
  • Cylinders must resist manual attack and ratings are available from 1-8 reflecting LPS1175 tool groups A-G (with G including a pneumatic drill, oxy cutter and rescue saw), which usually involves testing cylinder guards.
  • A manipulation rating of 1 (binary yes) against bumping is available and is tested by a panel of three locksmiths.
  • As there are no such certificates, in practice if you want a lock to LPS1242 you have to choose from a LPS1175 certificate lock schedule in the hope that testers correctly spotted if the lock was the weak spot and tested it.

Recommendations

  • BSx621 locks are all you need for non security doors. For security doors choose locks from the equivalent LPS1175 certificate schedule.
  • Rim locks allow beefier bolts and are protected by the whole thickness of door and frame. Once you get to LPS1175 SR3 it becomes hard to pass testing with mortice locks. There are plenty of highly rated doors using prison-strength single point rim locks, even up to SR6 rating, although multipoints are handy against hydraulics.
  • Back doors need a mortice lock at mid height, backed up by mortice bolts top and bottom at least 4” from the top and bottom edges.
  • Mortice bolts in the top or bottom of a wooden door should be 50-100mm from the edge. If on the leading edge then they should be halfway between the lock and the top and bottom.
  • Slide bolts should be to the highest rating under BS EN12051. EN12051 has a security rating called digit 7 with grades 1 to 5 for saw resistance, side & end load and projection. For example, grade 1 can be sawed, does not lock but resists 500N side load and projects 12mm, whereas grade 5 resists sawing for 5 minutes, resists 5kN end load & 10kN side load and projects 17mm. Grades 1-3 are not saw resistant. Unfortunately you will struggle to find anyone selling a bolt claiming to be security graded to EN12051, apart from a few flush bolts such as UAP flush bolt (grade 1), Briton 4735 flush bolt (grade 1) & Zero Z003R (grade 2). A couple of rack bolts also claim to have EN12051 but are quiet about the security grade.
  • All external doors need hinge bolts. Wooden doors need them at a quarter of the door height from top and bottom but not within 150mm of hinges. External hinges should have non-removable pins and all hinges should be backed up with hinge bolts.
  • Locking bars if used should be hardened steel and 10mmx60mm.
  • Do not let cylinders protrude more than 3mm.
  • Patios should have anti-lifts and three locks.
  • French door slave leaves should be held with bolts projecting 20mm into the head and sill. French doors can have their handles locked together and there are locks specially to do this such as the SBD approved Patlock, although you still ought to protect the glazing and use the included anti tamper spindle.

Market

  • Do not be fooled by apparently competing brandnames. Assa Abloy includes Chubb, Pickersgill-Kaye, Yale, Union, Mul-T-Lock, Henderson, Fichet, Adams Rite & Nemef.
  • Big names to look for in the domestic door and window lock market with many SBD products include ERA, UAP and Yale. Chubb locks were rebranded Yale or Union in 2010.
  • Rim deadlatches cost from £15-290, with British Standard versions mostly starting at £60, from brands like:
    • Ingersoll (eg London Line SC100 autodeadlock with 12″ keep to BS3621 at £283 comes recommended),
  • Mortice deadlocks cost from £15-280, with 5 lever British Standard versions from £17, from brands like:
    • Willenhall for £30,
    • Imperial G5054,
    • Marcus York BS5DD.
    • High security 5 detainer locks which are harder to pick and come with 74,000 key differs start at around £100, such as the Union 3G155, and 7 levers from £190, such as the Banham M97 to BS3621.
    • A good EN12209 mortice deadlock is the Gerda ZW1000 rated grade 7 security; this takes a euro cylinder.
    • Microswitch deadlocks for alarm setting are available for around £115 or £125 for a detainer version.
    • Deadlocks which trap the key until locked are available for £70 to help stop you accidentally leaving the front door unlocked.
    • Common 5 lever mortice deadlocks can be picked in 4 minutes by an amateur and two minutes by a practised locksmith with a decoder like the GJ Locks Fortress Decoder which go for £225, or they can use a carbide drill bit and lever lifter.
  • Multipoints (and various door hardware) to SBD are available from firms including GU, ERA, DGS, Maco, Ingenious (see 1009 Duplex 5 hook), FUHR, Fullex, Kenricks, Winkhaus & Avantis (used in Solidors, see also their FMT bidirectional mushrooms for windows). Multipoints should use hookbolts in upvc doors as plastic frames have so much give. A few multipoints are certified to PASx621, such as Yale Doormaster PAS3621 also to PAS24 and SBD with three hook bolts, two rollers and cylinder guard handles for £142, although whether a uPVC door can survive the one tonne force it tests with is questionable. Another brand is Era Vectis Fortress PAS3621, which uses levers instead of a cylinder so has less sticking out to be attacked. Such multipoints are also available for wooden doors.
  • Patio locks are typically either:
  • French doors may be suited to the Cal ResiLok.
  • High security rim locks for security doors are typically Surelock McGill Stirling or Mico Abryll or Elite if single point, or Surelock McGill Slimline or Mico INT up to SR6 & SBD or Evo if multipoint. Mico single points are general also SBD. Surelock McGill also do the Loddon SR2 rim nightlatch and Mico do the similar SR2 Rimlock. Picksgill-Kaye are a favourite with Sunray and Warrior doors.
  • The Mico INT, what looks like the AD3 SATE model although oddly the two bolt version is not available with lever exit, featured on the double doors to the briefing room in Sky TV series COBRA, a closeup visible in Series 2 Episodes 3 & 6, with the distinctive logo, blind rose and without the third point (it only locks top and bottom); the doors are mockups with normal glass and no dog bolts.
  • CPNI counter terror locks include Pickersgill-Kaye HS100 Series 3 L8709/L8710, SurelockMcGill SL313, Mico 942-3L INT AD, Mico Abryll AD3, Mico Elite Combi P7 & Mico Evo 3 AD3. All are rim mounted and only one is single point.
  • You could look for suppliers starting with directories provided by MLA, GAI and BMF.
Yes you can have BS3621 locks for uPVC multipoint doors, it is called PAS3621.
Patio door blocker
Slimline lock allows a SD-STD-01.01 door to last an hour against a mob
Multipoint survives 15 min mob attack on Krieger door
Stirling lock survives SR2 manual attack
Slimline Multipoint on Stafford Bridge Halton SR3
SR3 scissor gate hook lock
  • Mico have a video of the Abryll lock surviving an SR6 test with wedges, sledgehammers and a 5’ crowbar, and another showing it holding an Ascot SR5 double door even where the bottom bolt was left unlocked, and another where it holds a Bradbury SR4 door, and the USCH5 cylinder guard surviving an SR5 sledgehammering, disc grinding, chiselling and drilling, and the USCH4 protecting a PSF Wales SR4 door, and the L2 cylinder guard surviving chiselling and hammering to SR2. Their videos are a way to see various common LPS security doors such as Surelock McGill Benweld (a CNI armoury door fabricator) under attack.
  • Below are photos of rim locks to SR2-SR4 &FB7 for single point and SR6 / RC5 / FB7 / SD-STD.01.01 FEBR60 for multipoint, plus a mortice single point to SR2.
Multipoint on Quinshield SR2
Single point on Bradbury SR4…
Mico Abryll holding Benweld SR4
Mico Abryll holding Bradbury SR4
Multipoint on R&D Sheetmetal SR5
CTS Secure SR6

Limiters

  • Also consider a Sold Secure Silver:
    • Securibar Innovation’s Securecoil 2 tonne and Sold Secure Silver rated door limiter to stop intruders bursting in when you open the door; it can only do this by fixing to a perpendicular wall. It is awaiting manufacturers to licence it so any investors should get in touch with Leslie at Securibar.
    • There is also a Sold Secure door limiter that locks to the door frame rather than the wall, the Sold Secure Bronze rated Links Tech Loxout for £50.
    • Howsarlock for £5 is SS Bronze rated and locks internal doors from inside. Buy several and use them for travel and bugging out in case you need to lockdown a premises. An alternative for clamping shut overhead door closers is the more versatile Handcop HC1265 for $13, which is a 960lb-rated cord restraint.

Cylinders:

  • Pin tumbler cylinders for standard security locks (ie on non-security doors) should be rated:
    • SS312 Diamond (eg Avocet, Brisant, ERA, Federal, MulTLock, M&C Condor, Schlosser Technik, UAP), SS312 Gold plus SS301 Bronze security handle, STS217, or (TS07) THF DS07 3 star, or
    • 1 star plus 2 star/BS EN1906 handle (such as UAP or Hoppe at £46) or cylinder guard (such as UAP HSE to SBD), or
    • the higher standard of LPS1242.
    • The minimum for high security residential should be LPS1242 (exceeding the building lock standard BS EN 1303).
    • SS312 Diamond is BSI kitemarked, seen as the highest protection of the domestic standards as it uses a wider toolset from BSx621 plus vice grips and cylinder snapper and attacks the door and the cylinder via a 45mm hole, and is based on EN1303 and the GVA from BSx621, although you can get slightly different but equally strong protection from pairing a 3 star cylinder with kitemarked handle or cylinder guard. Sold Secure is MLA’s test house. SS312 requires a UKAS accreditation against EN1303 plus MLA expert attack to GVA plus two minute manual attack.
    • TS007 is similar to BSx621 and is based on PAS24, EN1303 and the GVA from BS3621. Cylinders get 1 star for passing all tests other than snapping (bumping, picking & drilling) and 3 stars for also passing the snapping test. It includes snapping, bumping, plug extraction, picking and drilling.
    • STS217 is Warrington Certifications’ version of TS007.
    • Secure By Design is a scheme run by ACPO which tends to adopt other accreditations as police-approved. SBD accepts SS312 Gold as equivalent to a 1 star kitemark and SS312 Diamond as equivalent to a 3 star kitemark.
  • LPS1242
    • LPS1242 is based on EN1303 and so has key security grades 1-9 and attack security grades 0-5 for cylinders. However it adds the manual attack ratings from LPS 1175 so additionally uses the tool groups A-G.
    • Key grades 6-9 need patented keys and grades 5-9 need registered keys with issue only by manufacturer for grades 7-9.
    • Key security requirements include ranges from 100-1,000,000 differs for 2-9 detainers, and 2.5-40 Nm torque resistance.
    • The requirement for key differs is tool group A = 1,000 = BSx621, B = 5,000 = LPS1214, C/D = 30,000 = BS EN1303 class 4, E/F = 100,000 = BS EN1303 class 6 and BS EN1300 (safe lock standard) class B, and G/H = 1,000,000 = BS EN1300 class C. BS EN 1303 class 4 is thus the benchmark for high risk residential.
    • LPS1242 adds a pass or fail grade for manipulation, and manual attack security grades 0-8. A manipulation grade of 1 is available against bumping.
    • For attack grades the cylinder must resist 5Nm torque after drill, chisel (driven by 6kg hammer), pliers and plug extraction (by traction screw) attacks, although attack grade 0 is exempt from those attacks. Attack security includes ranges for drilling for power and times from 3-10 minutes, chisel blows 30-50, chisel twists 20-50, plug extraction times 3-10 minutes and power 15-17kN, and torque resistance 20-50Nm.
    • Manual attack grade tools are categorised similarly to LPS 1175. Category A tools are for stealth attack, including traction screws and pliers, category B adds tools of an opportunist including hammer, hand drill and pipe wrench, category C adds tools of a planned attack including chisel and 7.2V drill, category D adds tools of an experienced burglar including 12V drill and lockpuller, category D+ adds tools such as an 18V drill, category E adds the tools of a professional burglar targeting a saferoom such as 750W drill, category F adds tools such as a 2000W drill and category G adds tools as a precursor to vehicle, blast or ballistic attack such as concrete demolition gear.
  • Beware that escutcheons cannot stop drilling and EN1303 cylinders are not snap or drill proof, so ideally use a cylinder guard with a spinning disc – for what it is worth, CPNI do not allow anything less. Instead we have DHF TS 007 – which takes EN1303 but subjects it to the BS3621 vulnerability test, and an alternative from Warrington, STS 217, and Sold Secure SS312 Diamond. BSx621 attack grade 2 was pre-snapping so is useless without something like a security escutcheon. TS007 1 star requires the cylinder specification in BSx621 Annex A A5 & A6 and BS EN 1303 clause 7 with 16000C5D rating against drilling, torque and plus extraction. 2 star handles requires the PAS24 A3 parts 1&2 cylinder test and if subject to EN1906 must achieve a rating of 16 – – – 3 – -. 3 star cylinders have to meet those BSx621, BS EN 1303 and PAS24 cylinder tests for an attack time of 150 seconds.
  • Cylinders should be flush with the handle: if set back the handle they can be gouged, if protruding the cylinder can be snapped.
  • Beware cylinder guards that fit inside a upvc door like UAP’s Maximum Security Cylinder Guard A at £4, which potentially only stop a weak cylinder being snapped, leaving it open to picking, bumping and drilling – you might as well replace the cylinder with a 3 star instead of only curing one vulnerability and having to route a 24x60mm hole in your door for the privilege. You can buy a 1 star cylinder, which you pick up for £11, like the UAP Plus Zero Lift that is TS007 rated against bumping, picking and drilling.
  • Brands with SBD rated products include:
    • ABS ATK & Master & Secura & Ultimate (TS007 3 star),
    • Abus X60,
    • Adams Rite (TS007 1 star),
    • Titon Asterion TN62 (TS007 3 star)
    • Schlosser,
    • Mult-T-Lock MTL300 3XP (SS312 Diamond)
    • Sancta,
    • Federal,
    • EVVA Sicherheitstechnologie (TS007 1 star)),
    • Dormakaba (Gege pExtra, pExtra+, Quattro Plus, ExpertPlus with security grade 6 key level D & SBD & TS007 1 star, pExtra Guard & pExtra+ Guard are their TS007 3 star & BS3621 versions with grade 6 EN1303 key security),
    • Carlisle Brass (TS007 3 star),
    • Ingenious,
    • Brisant,
    • Kenrick Excalibur (TS007 3 star),
    • Asec,
    • Yale,
    • ERA,
    • Assa,
    • Ingersoll,
    • Union,
    • Cisa,
    • Avocet,
    • Iseo,
    • Burgwachter BS3 Sterling Doorminder (TS007 3 star)
    • DGS Elite (TS007 3 star & EN1303 grade 2 attack resistance) &
    • UAP.
    • Prices range from £10-290 with those over £80 tending to be for Banham specialist locks, although DormaKaba and Adams Rite cost more too.
  • Brands with SS312 Diamond cylinders include:
    • Assa Abloy Yale Platinum,
    • Avocet ABS Master & ATK,
    • Abus Ultimate & X60,
    • Apecs Consult AP,
    • Brisant Ultion XWM,
    • Deseo ALPS 65H,
    • Era SSD,
    • Federal Cherubim UCH & YCH & 11KDCH & Gabriel 6YUGTP & YCF & UCFTP & 11KFCFTP & 11KDCF & YCHTP,
    • Iseo F6 Extra S3 & R700 S3,
    • M&C Protect Condor,
    • Masterkey Adapta Prime HS,
    • Mila ProSecure,
    • Mul-T-Lock MTL300 & BS3XP & BreakSecure 3DS,
    • Security Hardware Schlosser Technik,
    • SKS Maxus Pro,
    • UAP Kinetica.
    • There are also SS Gold cylinders available from Brisant, Federal, Davenport Burgess & Dom UK.
    • There are also SS Silver cylinders to SS301 available from Brisant, Federal, Dom UK & Davenport Burgess.
    • There are also SS Bronze cylinders to SS301 available from Brisant – the Ultion 2 star.
  • Triple rated cylinders with SBD, TS007 and SS312 Diamond include the Era Fortress and Schlosser Technik.
  • LPS1175 doors need matching scheduled locks listed on the certificate, and the cylinders tend to only be rated subject to fitting their LPS1175 cylinder guards, such as the Surelock McGill GSS2 to SR2 & SBD. The certificate lists numbers for each door model which correspond to listed locks, and those locks then have letters that correspond to listed cylinders. For example, Assa Abloy’s Trenchard SR3 door can take locks (mostly their own) such as the Pickersgill-Kaye AA16, which takes cylinders (mostly their own) such as the Euro profile CY304PT, ANSI standard 11-disc Abloy Protec CY403TT for £100 (grade 6 key security, grade 2 attack resistance), Bramah C17 for £65 or Kaba K20 for £170.
  • Normally you want locks and cylinders for key and handle both sides; this dramatically quickens your search of compatible products as most are for panic bars, no key or no handle.
  • LPS1175 compatible cylinders with key both sides include:
    • Abloy Disklock Pro or Novel or Protec
    • Kaba Quattro Plus
  • LPS1175 compatible locks with key & handle both sides include:
    • Surelock McGill
      • 2 point SL211-00-(E1)P5W
      • 3 point SL311-00-(E1)K0P5W (to SR4)
      • 5 point SL311-00-(E1)K0P5UU1W (to SR4)
    • Mico USCH3 cylinder guard +
      • 3 point:
        • 942-3L EPC INT AD (to SR3)
        • 942-3L EPC INT MD (to SR3)
        • 942-3L INT AD (to SR3 & CPNI)
        • 942-3L INT MD (to SR3)
      • single point:
        • Elite AD3 (to SR3)
        • Elite MD3 / HS3 (to SR3)
        • Abryll MD3 (to SR3)
        • Abryll AD3 (to SR3 & CPNI)
    • Pickersgill-Kaye:
      • single point AA16 (to SR4)
      • three point AA50 (to SR4)
    • Some lock ratings depend on the cylinder guard they are teamed up with, for example, the Mico Abryll MD3 is SR3 with an SR3 cylinder guard, but with an SR6 cylinder guard is SR6 despite being single point, so is much stronger than most security doors it is sold with.
    • To match normal home brick strength at SR3, a Mico Elite MD3 single point rim manual deadlock is more than enough to hold an SR3 door, and a Mico 942-3L INT AD three point rim auto deadlock would not be overkill in brute strength but would resist a wider range of attacks; the question will be, after it dumps blows into the frame through the rawlbolts into the wall, will the bricks take the beating?
  • Often the same cylinder (and any armoured escutcheon that comes as part of it) is available in various key systems or ‘cylinder platforms’ or ‘mechanisms’ so barrels can vary, for example, you could swap Abloy’s Protec barrel for the cheaper, less secure and less flexible Novel barrel.
  • Cylinders also come in various profiles such as Euro profile made famous by UK importing continental plastic or composite doors, but there are also options like oval UK profile, Scandanavian oval, Finish round.
  • What manufacturers call the cylinder is often really a cylinder guard with whichever key platform barrel being protected inside.
  • Remember that some types of cylinder guard can be removed if a hole is made in the door, eg Avantis TwinFix can be taken out with a 4mm Allen key.
  • Fire brigade advice is in Guidance Note 11 which basically recommends internal keyless unlocking for doors and windows, and cautions against armoured security doors that can’t have a handhole cut through quickly to release the thumbturn. Most non-CPNI security doors would allow this.

Cylinder guards

  • Perhaps the strongest cylinder guard is Surelock McGill GSS7 which is rated LPS1175 G5 so can survive five minutes of petrol grinder. Their longest delay model is an SR5 at 10 minutes.
  • Below LPS1175 in the domestic standard security market you can buy 3 star cylinders that should not need cylinder guards at that low threat level, but you can add delay by fitting a 2 star cylinder guard anyway, such as SBD/PAS24 security handles from someone like Mila, in fact you could also fit a Winkhaus Armourshield but that gouges out material from the uPVC door so is of debatable value.
  • SBD / TS007 / PAS24 cylinder guards include:
  • An example of an unrated cylinder guard is Federal Lock Y6157 Euro cylinder escutcheon for about £20.
SR2 cylinder guard
SR4 cylinder guard
SR5 cylinder guard
GSS7 test

Fittings

Frames

  • Steel frames in masonry can be grouted, but only with 4″ slump mortar grout in braced frames and hand trowled during construction – you can’t pump it into the frame in an already built wall. They should be continuously welded rather than spot welded or knockdown. Ideally they are wraparound so cover the gap between frame and wall to resist access to fixings.
Welded wraparound frame

Hardware

  • Avoid panic hardware, and even handles and thumbturns, go for key operation, and even then you might want to protect the internal end of the cylinder in case a hole is made in the door to allow manipulation. The door must be tough enough to stop any hole whatsoever if you have internal handle operation. There are tools to turn handles through tiny holes. Bear in mind that Warrior‘s jewellers SR4 doors are specified to replace glass infills with steel if they use thumbturns and to make it manganese steel if they use panic handles. You will struggle to find an example of BRE passing glazing with panic bars – of any thickness.
BRE panic bar defeat in LPS1175 test
  • Use only bolt-through door furniture secured from inside.
  • Fit astragals to protect the locking edge and a Z-strip to protect the hinge edge, such as from Zeroplus. An alternative to the Z strip is to weld a steel angle to a steel angle welded to the steel door plate or to the frame that would bump up against a jamb or U-angle in the door edge if hinges were removed and the door pulled.
  • Avoid split spindle multipoints like the plague, as they encourage occupants to not deadlock the door.
Handles can be opened from outside through letterbox
Thumbturns can be turned through a letterbox
Thumbturns can be opened without a hole with vulnerable Euro cylinders
Panic bars can be opened through the edge of the door

Reinforcement

  • Wooden doors can be reinforced with Kickstop screw-on anti-thrust plates (£32) and stronger bolt-through guards (£45) for outward openers, Birmingham or London bars or shorter hinge guards (£36) and frame guards (£37-42), frame plates (£37), staple guards (£42) and deadlock guards (£37). A softwood door would shatter if not reinforced while the frame is reinforced.
Kickstop door reinforcers
If you reinforce the frame you have to reinforce the door
  • Companies like Birmingham Bar sell screw-on bars (£19 each) and Kickstop DG2 grilles of 5mm x 20mm bars (£130 on Amazon) to give at least a fig leaf of security for weak panels.
  • Inward opening doors can be reinforced against opening with bolts known by names such as ‘lockdown’, ‘barricade’ or ‘Nightlock’ or door stoppers on the bottom rail to bolt into the floor, popular in the USA. Solon Security DJ3 Door Jammer is SBD rated and Era Lockdown (a bigger barricade weighing in at 4kg at £330 aimed at the school market), Nightlock Original and Nightlock Lockdown 2 are recommended by CPNI, but they warn other brands advertise a room is occupied. However, some brands of telescopic jammer can snap. Abus do an inward opening door bar for about £150.
  • Other gadgets include the sheriff-invented Barracuda sliding wedge jamb to slot under an internal door to lock it from the inward opening side. There are versions to lock door closers and outward opening door bars that clamps the handle. The Secure-Ring was a chain for composite and upvc doors costing £28 which loops the handle to the wall, but nowadays only a cheap copycat product exists, the Asec lock ring which fixes to the door frame with a sticky fixer!
Abus door bar
  • Another option is the drop bar, such as from Strongdor. Usually limited to fire doors, they could make a great doomsday reinforcement for any solid door.
Strongdoor dropbar
  • The front door should have a limiter to DHF TS 003 and video intercom – voice identity is not enough. A cheaper alternative to video intercom is a door viewer to DHF TS 002 such as the UAP CVPL, although it is not entirely safe as there are locksmith tools to remove viewers to put a wire through to undo handles and TS002 only tests functionality, not attack resistance. The only SBD rated door viewers are from UAP. TS003 for chains and limiters requires 200 x 100N impacts for wear and tear then 3 attack impacts to PAS24 (30kg sandbag and 50kg ram swung from 80cm above impact site). Beware that limiters can be defeated with mica cards or Lock Jockey.
Yale 180 degree door viewer
  • Chains are useless.

Hinges

Glazing

  • Security glazings’s gold standard must be polycarbonate glass as polycarbonate is great against blast and impact whereas glass is better against cutting and bullets as long as it is held together with interlayers and spall liner.
  • Glazing units in aluminium or upvc can be refitted with security tape or clips to foil a de-beading attack.
  • Conservatories probably have to be written off and the door from them to the house treated as the external door, especially if the glazing is thin polycarbonate.
  • Vistamatic are a leading vision panel supplier. Their Chieftan is SR2 and Sherman is SR3 and are used in some Sunray and Metador doors. It is possible to buy vision panels for doors rated for 15 mins against a mob.
  • Manual attack glazing is accredited separately to LPS1270, or LPS1175 as part of the overall door, with the same ratings SR1-8 in terms of tool groups, but with the addition of number codes for minutes delay against particular size holes.
  • Avoid anti bandit glazing. For example, 11.5mm laminated glass lasts only 4s against a group C (SR3) tool, the lump hammer, before a hand hole is made.
  • Rated products include polycarbonate (‘Lexan’), laminated polycarbonate (Lexguard ‘detention’, 10-32mm of 2-4 ply, weighing 11-38kg/m2), laminated glass and laminated polycarbonate glass (sandwich of glass and polycarbonate).
  • See the Windows post for high security glazing advice.
SD-STD-01.01 15 min vision panel

Letterboxes

  • Letterboxes should be on the wall and to SBD and DHF TS 009, eg the DAD009, and ideally 10m from the building, but if you must keep a letterplate it should be to DHF TS008 (eg UAP Sotarian for £23 which only opens 37 degrees to stop fishing, or Yale Postmaster for about £40 also with restricted opening) and BS EN 13724, or be no more than 40mm wide and unremoveable from outside. If you use a thumbturn the letterplate should be at least 40cm from it and have a deflector. But having a thumbturn or letterplate in a security door is a contradiction in terms.
  • Any letterplate taken out of service should be protected with Safety Letterbox’s 1.5mm steel Flaplock LF05. There are blanking plates and locking flaps available, but they mostly only screw on or lock from one edge, so are rarely strong enough.
  • Visor guards or cowels disrupt hand access but are only a flimsy deterrent, unless rated such as ERA Fab&Fix Nu-Mail or Royd & Tucker LP08 cowls to SBD, PAS24 & TS008 for £50.
  • Arson through letterplates could be stopped by an extinguisher bag like the SBD rated Prestige Products’ Homeguard Plus for £77, or there is the SBD rated DeRaat’s Letterblox blanking kit (resold by Safelincs) at £108 – which at 3kg is probably quite sturdy.
  • SBD rated wall mounted letterboxes are available from Post Boxes UK. TS 009 letterboxes are tested for 1 minute attack by tools like screwdrivers and pliers to open it or fish out mail and must resist pulling off with 1.5kN, and keylocks should have key security grade 3 to EN1303.
  • TS008 letterplates are rated with digit 3 for resistance to thumbturn manipulation with tools like screwdrivers and pliers (1) and fishing of keys from hall table without pliers (2) for 3 minutes, digit 4 for lockability and digit 8 for arson protection against 15g firework, newspaper or 500ml petrol. Letterplates must resist fishing for 1 minute without tools. The 2022 version is TS008 has tightened up the key fishing test.
DAD UK’s DAD009 SBD letterbox
Yale Postmaster
ERA Fab & fix Nu-mail cowl
Royd & Tucker LP08 SBD letterplate
Mailguard Plus
Deraat Letterblox

Door entry systems

Door viewers

  • The NHBC standard 6.7.8 and Building Regulations 2010 Approved Document Q 2015 D10 requires an outside view, whether by window, CCTV or viewer.
  • Door viewers are advised against for doors with thumbturns or panic bars or anti-cut armour as they can be unscrewed and the hole used for a fishing tool or saw.
  • However, UK MoD standard JSP440 allows one model to be used, eg on key-only doors up to SR3, and that is the Chubb WS8, simply because it is wide angle, although at 15mm diameter it is way beyond the 6mm limit CPNI would ever tolerate if one was essential for safety, and most viewers need at least a 12mm hole. An example of a slim one is the Yale 9401 or UAP CILB120 at 12mm. The WS8 became the Yale WS8 before being discontinued but they still pop up for a tenner or less on ebay or amazon.
  • Door viewers should be to TS002 with box 7 for security to grade 3, meaning you get at least 160 degrees angle of vision, visual clarity of 8 point text at 6″, mist resistance (comparing reading smallest point size possible under visual clarity test after 3 days of humidity to BS3900) and mist clearance in 3 days at 30 degrees celsius. Most door viewers mentioning TS002 in their adverts only claim to be ‘tested to’ it – they actually failed or achieved a low grade in the security box.
  • Beware relying on SBD rated door viewers as, for example, the so-called ‘Wide Angle’ UAP 14mm door viewer only actually achieved 95 degrees viewing angle, just over half the pass mark for grade 3 security to TS002, and having misted up too fast only achieved grade 1 security (as was the whole product range).
  • For while you have batteries the safer alternative is a video door viewer, sometimes called a smart viewer, digital viewer or peephole camera. None are rated though, so there are no SBD or Sold Secure models for example. Typical specs are 2″ screen with 720 resolution and SD card. Some have features such as night vision, app, motion detection. Products include:

PIN, card & biometrics

  • PIN & card entry is usually a no-no for high security applications without guards and physical security in another layer in front of or behind it. Firstly, they usually rely on low security electronic locks and secondly, decent ones need encryption, anti-tamper, heartbeat and to be beefed up with biometrics, so are expensive. But in combination with guarded physical security it can be great as it adds something you know with something you have. That can be combined with something you are like your fingerprint, voice, signature, veins or eye, or allegedly at MI5, your weight (probably actually just anti-tailgating). Some security doors can optionally have a second lock so you can keep one for key only with the main one also on a key but with a relay for PIN, fob, fingerprint and/or remote release button.
  • If you use proximity fobs you ideally need PIN too.
  • Automated entry must have backup power, fail shut, PIN lockout and key backup option.
  • Biometrics are usually advised against for panic rooms due to high false rejection rate and slow activation speed.

Video entry

  • Video entry is safer than leaving a weak spot just so you can ID visitors through the door, as there are tools to unscrew peepholes and turn handles through the hole. Unfortunately the market is aimed at blocks of flats so suppliers are not big on explaining how their system could work for a house.
  • SBD rated domestic door entry systems typically cost about £600 just for the intercom before you even integrate it into a remote relay, £250 for a monitor, £150 for a power supply, £360 for a keypad with its own power supply, and you can pay £400 for a vandal resistant mullion proximity reader to read a fob. You are basically looking at £2,000 to supply and fit a rated door entry system.
  • SBD accredited door entry suppliers are:
  • SBD accredited domestic video intercom suppliers are:
    • Ring (Video Doorbell 2, Floodlight Cam (mains), Stickup Cam & Spotlight Cam (battery or solar) are SBD),
    • Videx (4000 Series suit a house but not SBD rated),
    • Comelit (VIP or Visto (£230) suit a house but not SBD rated).
  • The only SBD accredited video intercom is Ring. They charge £25pa for storing each camera’s video or £80pa for unlimited cameras.
  • Consider whether you need to link your door entry to any alarm and CCTV.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: