Bearhawk Aircraft Bearhawk Tailwheels LLC Eric Newton's Builder Manuals Bearhawk Plans Bearhawk Store


No announcement yet.

The Military way to Buck a Rivet

  • Filter
  • Time
  • Show
Clear All
new posts

  • The Military way to Buck a Rivet

    Sometimes I get into minutia. But, this minutia in the above publication is a link to the Military specifications to aircraft riveting. I found kind of interesting.

    AC 43.13 calls for a rivet's driven shop head thickness of .5D and head diameter of 1.5D. But the Mil Specs give us some tolerances that I had not seen before. In fact a 3/32 (.093 D) shop head thickness tolerance is .038 - .050 whereas I was looking for .5D (.046) as a minimum only. (why is there a Maximum Thickness Tolerance?) The Mils Spec head diameter minimum head diameter is .122" rather than (1.5D) .187".

    When de-burring a rivet hole, I had never seen tolerances before, just "Not Too Much." Mil Spec says 10% material thickness. Rivet Edge Margins are also more specific and dependent upon the type of rivet being used...the mil spec 3/32 Universal Head rivet edge margin .156" (Thats equal to about 1.7D) whereas the AC43.13 standard 2D would be .188"

    I look at having some tolerances as liberating. But its not AC 43.13. Give me some feed back.

    Brooks Cone
    Chelsea, MI
    Scratch Building Patrol #303
    Brooks Cone
    Southeast Michigan
    Patrol #303, Kit build

  • #2
    I thought the maximum thickness was related to making sure you weren't using too long of a rivet. This could mean too much metal forming during the set, which could lead to cracking after the material is work hardened and then further worked. But ask an RV builder, those guys are the rivet experts.


    • #3
      It is virtually impossible to get the rivet exactly on the AC43.13 spec otherwise a person would have to trim every rivet & caliper it before driving it. It could drive a person to madness.

      The military spec is a plus minus tolerance that translates into an upper & lower range for an acceptable shop head. This is practical and reasonable to be able to complete work. It is common in machine tolerances to have + values for the tradesman making the parts which is the range for the acceptable fit & finish. The process of beating a rivet between an air impact & a bucking bar has no controls for exact precision so a range allows the work to be completed successfully. These are numbers that work for a riveter and inspector. Max thickness is the upper range and would be to ensure the rivet is driven tight as thicker could be a under driven condition. Likely in the manufacture of 100's of aircraft it may be practical to supply exact rivet lengths for the application to get the rivet within tolerances. We did our level best to get close to the AC43.13 as possible. We checked rivets occasionally to see if they were good but did not check them all. After a while it is possible to look at a driven rivet & see whether it is enough or over driven.

      We deburred by spinning a larger bit between the fingers with light pressure. This works quick and it took the burr off to leave a clean hole. The "not too much" likely means no countersink which is possible if someone is deburring with a power tool. A countersink with a regular rivet head reduces its clamping area & may put bad stresses into the corner between the rivet head & the shank.
      Last edited by Glenn Patterson; 07-02-2016, 10:32 PM.


      • #4
        Back when I was building my RV8, I saw a test of riveting that an RV builder did. He made several strips of .032 2024T3. He riveted some perfectly - totally to spec. Then intentionally under drove some rivets and also over drove others. Both AN470 and AN426 were part of the test. He pulled the strips to failure. Generally, an over driven rivet is stronger than an under driven rivet. But the worst under driven rivet was still within 15% of the strength of a perfectly driven rivet. An over driven rivet was within 8%-10% of the strength of a perfectly driven rivet.

        And a flush head AN426 on dimpled skin (like we have on all of Bob's designs) was considerably stronger than an AN470 round head rivet in all cases. All holes were properly deburred.

        Now I am not suggesting poor craftsmanship by any means. But as they say "Perfection is the enemy of completion". And of course some rivets are more important than others. We all want to do good jobs building safe airplanes for ourselves and our families. Mark


        • #5
          Agree with Mark on all his points. "Perfection is the enemy of completion".

          Bob Barrow has built 4 scratch aircraft that are flying successfully. Dollars to donuts he did not measure all the rivets or pursue absolute perfection. Build it strong and safe with simple processes in a good workmanlike manner. Some builders get lost in making all the tooling, fixtures and software programs to pursue perfection that just chews up life and does not advance the build quickly. A person is given one lifetime so a decision has to be made to either enjoy the build process and relish the build as long as possible with no completion goals. Or if it is to fly the airplane then make the build as short a line to that goal as possible while young enough to enjoy flying your proud work. Getting it done means sticking to it & staying focused. Set it aside for a while and the longer it sits idle the harder it is to go back at it. There will be down days but tough them out because suddenly all the parts built on those down days and good days come together to make good visual progress that lifts the soul and re-energizes for the next phase.

          I know from managing code piping and pressure vessel weld program that the number of inspections is proportional to the service. Usually it is a nominal inspection of 5% of the welds. The inspections increase in increments up to a 100% where there is high risk. Check some rivets as you go and perhaps more frequently in key structural areas. Your eye will soon see go or no go as good as the gauge.
          Last edited by Glenn Patterson; 07-03-2016, 06:59 PM.