Flapper Locking video

By request, today we’re going to look at one of the less common locking systems used in firearms design: flapper locking. The idea was first patented by a Swede named Friberg in 1870, but a practical rifle was not built on the design until the 1907 Kjellman. (the Mauser 06/08 pistol was also developed at this time, and was produced in very small numbers, but more than the Kjellman). The most extensive use of the system was for the Degtyarev series of machine guns (DP, RPD, DShK), although the Germans also used it in the G41 (both Mauser and Walther) and G43 rifles. Finally, the Mauser company made a small number of early guns using the system – the 1906/08 pistol and 1916 rifle. The later roller-locked system is in many ways similar to flapper locking, but we won’t delve into that today.


  1. Excellent video!

    I remember reading about this system years ago. It was referred to as “the Kjellman Flap” and was mentioned only briefly. I’ve been wondering what it was and now I know.


  2. Arthur and I (you met us both, and our German friend Axel, at SHOT Show 2012) are not convinced that any selfloading/automatic weapons have primary extraction. It’s not that we think they couldn’t, but we have yet to confirm an example that does.

    The AK is often cited as having primary extraction, and it does have beveled surfaces that look like the primary extraction surfaces on a bolt action, but we are not convinced that’s actually what they do. Likewise, the AK has an angled cut in its trunnion that looks like a primary extraction cam, but as far as we can tell, only serves to knock the bolt into battery during locking.

    If there are any selfloading or automatic firearms that you know firsthand have primary extraction (even the AK! We’re agnostic about that rifle), perhaps do a video, so we can all see.

    • Well, if you watch inside the magwell of an AK while opening the bolt, it does definitely retract slightly while still rotating.

      That said, I must admit that I haven’t really looked into primary extraction closely – I just took it as a given. Clearly I need to learn more in that area before I open my mouth again. 🙂

    • I know for a fact that my M1 garand has primary extraction. If my understanding of it is correct. But during the unlocking phase the bolt travels back just a bit

  3. I disagree with your assertion that the safety factor is larger with the Russian flapper design (though it might be right). The two designs load the flappers in very different ways. The Russian method is in compression whereas the German design is in shear. Since allowable compressive stress in design is lower than shear stress (usually by a very large margin) the factor of safety might be very close in both designs. The Russian design has a risk of columnar collapse whereas the German is shear. Neither design has ant primary extraction but neither does most semi/automatic guns. It has to do more about what is the residual pressure in the chamber and the speed of extraction when the case is being pulled than if it has primary extraction. Being notorious for case heads being ripped off has more to do with that or extractor size than primary extraction.

    • Hmm – I didn’t think about the different failure modes that would govern. However, anecdoally, I’ve heard a lot about cracked G43 flaps, and nary a peep about problems with Degtyarev guns.

      • With the Degtyarev being effectively “rear locking” the same amount of strain is occurring over a greater distance, so for any given differences in clearence between the locking flaps and their seats, the “rear” locking action will take them up at lower stresses (even without allowing for the flaps bowing) than a “front” locking action will.

        There is the hint of a free lunch there too,

        The greater flex in the locking system results in more of the force of the gas pressure being taken by the brass case.

        For an illustration, imagine an infinitely rigid locking system, the infinitesimal movement of the case head at max pressure will transfer the entire pressure to the locking system.

        Now imagine no support at all behind the case head on firing(I think the ref for this is Hatcher’s Notebook – Hatcher fired a .30-30 Winchester model 94, assembled so that it had zero locking, the case head held, and the case stayed in the chamber) even assuming that the case doesn’t stick in the chamber, up until the instant that separation of the head, or sliding occurs, all of the pressure is being taken by the case head web.

        It is possible that the Degtyarev was operating as a glorified blowback, with minimal load being taken by the flaps’ abutment surfaces

  4. All Tilting Barrel Browning Lock pistols have
    “Primary Extraction” feature. Case in the chamber
    is forcefuly exerted back by the leverage of barrel
    rotated on the axis of barrel hood against wıth the
    stationary extractor on the breechbolt front face.
    However, early two link barreled pistols like Colt
    1905 had not this feature.

  5. That was wonderful! I’ll be passing this one on to some friends.
    …and I am just about positive that the AK has primary extraction. I too have witnessed the bolt coming back “a peep” as it rotates. Why did I sell that rifle?

  6. a slower “primary” phase of extraction is claimed for some tilting bolt designs, and certainly for the BAR.

    It can also be incorporated into BREN and VZ58 locking systems.

    I generally took guns like the AK, the Garand type actions and the AR10 to 18 actions to have zero or next to zero primary extraction, accellerating the case to full bolt speed almost instantaneously.

    Which raises the question;

    If primary extraction was designed into some self loaders, does it serve any real purpose?

    • One of the best systems for a slow and powerful primary extraction is a recoil operated toggle lock (as opposed to toggle delayed blowback, e.g. Schwarzlose and Pederson).

      The cam which begins to bend the toggle works as an accelerator, transferring momentum from the barrel and slide to the toggle and the bolt.

      The unlocking toggle is operating like a “compound linkage” reloading press at maximum leverage.

      Browning and Lahti both understood this well, hence their designs which were intended to directly replace toggle actions, incorporated accelerators.

      Lahti’s designs do this to a ridiculous degree, much as I love his L35 pistol and the Swedish M40 copy, he’d have been far better off using a Browning combined slide and bolt.

  7. A very interesting point about “primary extraction” has been raised. I don’t have an informed opinion on the subject, but I haven’t seen anyone making a convincing case for its existence.

    It’s one thing to say that the bolt or breech block “moves” during unlocking, it’s another thing altogether to prove that this movement provides any extraction. To prove primary extraction exists, you would need to measure the amount of movement during unlocking is greater than:

    A) the combined mechanical tolerances in the fit-up between the extractor and cartridge case,
    B) The mechanical tolerance in how the extractor attaches to the bolt,
    C) the elasticity of the cartridge case.

    In other words, although the bolt does “move” while unlocking, perhaps this has no effect on the cartridge case because the movement is smaller than the combined mechanical tolerances.

    It may to be possible to measure some of this (without the elasticity effect) as follows:

    1) Clamp a rifle in the vertical position.

    2) Chamber a dummy round (not a live one!) in the rifle.

    3) Slide a close fitting rod down the barrel (or use a smaller rod in some bushings) until the end of the rod rests on the tip of the dummy round. We don’t want the rod to have any play which might affect our measurements. We do want the weight of it though to help push the dummy round to the rear of the chamber.

    4) Set up a pair of dial indicators to measure the movement of the bolt and the rod while *slowly* unlocking the bolt.

    Someone might have a better idea than the above, but the idea is to make some actual measurements. The most difficult part in the project would likely be getting measuring access to the bolt.

    A pair of LVDTs or linear potentiometers connected to a two channel chart recorder would be better than dial indicators, as it would let you see the relative movements more clearly.The average machine shop though doesn’t have equipment like that.

    I don’t own the equipment to do this, but a study on it by someone would make a very interesting subject. I don’t think the debate can really be solved without someone actually making some measurements.

    • I’m replying to myself here, but I just thought of a flaw in the above proposal. The dummy round will fall (or be pushed by the rod) to the back of the chamber without the extractor pulling on it.

      The position of the rifle would need to be reversed (so its muzzle down) and the rod screwed into the tip of the dummy round. This would mean that the extractor must *pull* the round in order for it to move. Other than that though, I think the idea would work.

  8. Ian, I discovered your website about 5 years ago and I love it! I have always been interested in the mechanics behind different systems. Thank you for all the hard work you put into this site.

    Having said that, I an suprised that a very experended weaponeer as your self would make the common mistake of classifying two totaly different systems as the same.

    The Friberg so called “flap” is not a flap at all! The Friberg system, in reality uses “retracting locking lugs”. The Degtyarev system does in fact use flaps. One locks at the front, the other locks at the rear.Also,please note that the locking surfaces of the two systems rotate in opposite directions when locking and unlocking.

    Again, Thanks for a great web site with a welth of interesting and useful information.

    God Bless, Max

  9. Ian great job on this rare “grandfather” of a the Roller locked and Flapper locked family of weapons. I would like to point out that the great great great great grandfather of all of them was the 1907 Roth Steyr pistol. The Roth Steyr doesn’t have a roller. It does have a bar on the right side of it’s bolt that is cam ed into a recesses in the barrel extension by action of the firing pin upon firing. The firing pin actuated Locking flap was used in the 9 mm and 45 ACP Mauser 1906 and 12/14 pistols and MG 34 and a host of other machine guns from the Russian RPD to the Dishka 12.7 MG. This is the missing link in the roller locked firearms family. The 1907 Roth Steyr and earlier Roth’s also pioneered the rotating barrel delay last used on Colt 2000 pistol and others. Roth’s also held their firing pins in a half cocked position to make their double action trigger pull lighter. This system is now very modern and is used in every Glock pistol. I think the 1907 Roth Steyr is an extremely interesting pistol. It is fun to shoot but it is more fun just to take apart and marvel at the machining. How did they machine the very thin barrel shroud? and the inside camming surfaces in the end cap? It is a wonder considering they probably used machines powered by leather belts running up to shafts turned by water wheels. The men that made them were SKILLED. I hope to see an evaluation and shooting Video by you on the 1907 Roth Steyr in the future. Bob

  10. It should be stated that, most of the auto loading small arms “Primary Extraction”
    claims retain on the “Theoretical” side, rather than “Practical”. Tolerances of both
    related gun parts and the case construction will realise or fail on doing so. In fact, mechanical ability for “Primary Extraction” can be added on any kind of “Locked Breech” firearm as maintaining “Engaged Situation” of barrel and breechbolt via camming action just a little further after unlocking occurs. However, modern
    clean gun powder composition and aid of conical body case shape, togetherly precludes to use this facility on most of the modern auto loading small arms.

    As regarding the “Browning Tilting Barrel Lock”, as again, remaining mostly on the
    theoretical side, “Primary Extraction” feature is healtier on “Ejecting Port Locker”
    versions than standart “Inside the Slide” counterparts.

  11. FYI,

    The Chinese Type 77 and Type 85 12.7mm heavy machine gun and the QLZ88 35mm automatic grenade launcher are all using the Degtyarev flap lock. All three are currently in service and the QLZ88 is still in production. Btw, all three are direct gas impingement operated.

    The Chinese QJC88 12.7mm heavy machine gun features a smaller Kjillman flap lock, which the flaps opening in the front vs at the rear of the Degtyarev. It uses a long-stroke gas system and it’s still in production.

Leave a Reply

Your email address will not be published.