AN Standard 37 degree taper fitting

[Edward Camic]

Does anyone know the reasoning behind the decision to use 37 degrees for the conical taper on an AN fitting? I understand this is a standard dimension for the fittings, but I am hoping to get a little background on how this came to be.

 

[jrmichler]

Early flare fittings had a 30 degree angle. Then the US government got involved and developed a family of 37 degree fittings for use in aircraft. Some good information from Parker: http://blog.parker.com/an-37-flare-vs-industrial-37-flare-fittings-whats-the-difference.

I don't know how they came up with 37 degrees, but we can look at the extremes of possible angles:
1) Very low angle - 5 to 10 degrees. It would be difficult to make the flare, difficult to assemble in a tight space, and difficult to get a good clamp fit while still able to disassemble.
2) Large angle - 90 degrees. Difficult to make the flare (tube cracking, need multiple steps), easy to assemble in a tight space, and easy to get a good clamp fit.

The extremes of angle have different problems, so there is an optimum in between that balances ease of flaring, ease of assembly and disassembly, and getting a reliable leak free connection. Finding the best angle is an optimization problem - the exact angle varies according to how each of the parameters in the optimization is weighted.

 

[Baluncore]

I am hoping to get a little background on how this came to be.

Machine tapers are specified by radius change per axial length. That is because the cross-feed on a cutting lathe is geared to the lead-screw by an integer ratio gearbox. The USA units are “inches per foot” so US tapers often show submultiples of 12” per foot.

Tan( 37° ) = 0.753554 which is close to 3/4.
Atan( 3/4 ) = 36.87° and I think that is where that choice of taper comes from.

To cut a 30° taper would require an irrational gear ratio of 0.57735 which is not possible with a simple gearbox.

 

[Edward Camic]

Super helpful and interesting, thanks everyone!