Rule of thumb for any soft metal (aluminum) is the thread depth for blind holes should be 2D.  For hard materials (steel) 1.5D should be sufficient.

--Scott

CNCCustom,
Swertel's rule of thumb is good advice. You also asked about "better" alloys and tempers, by which I guess you mean stronger alloys and tempers.  You didn't provide tempers for your 2024 and 6061 alloys, but there are stronger alloys available, 7075 being one of them.  

Suggest you use the applied load to determine the strength you need in the joint, then select a combination of alloy,  temper, and thread which will do the job. Be careful, also, that drilling and tapping, or other machining in the vicinity of the threaded holes, does not cause local annealing, which can result in a drastic loss of strength.  

The rule of thunb we use is:
Don't tap Aluminium alloys, use helicoil inserts.

We routinely tap 6061-T6 and 2024-T3 or T4, and 7075-T6.  They have reasonable strength and are hard enough to tap cleanly.  There are formulas for determining the tread depth so the tapped thread has strength equal to the machine screw used in it.  I try not to go less than 2 x diameter in depth and use helicoils when the parts will be frequenty disassembled.  As a reference point, my motocross bike makes extensive use of cast aluminum that is tapped.  It is pretty soft, so their thread depths are very deep, typically over 3x diameter.  With enough depth. you really have to abuse it to strip the threads.

Another factor in softer materials is not only the depth of thread, but also thread depth, and I'm not being redundant.  Most tap charts call out tap drills which will give you from 65 to 80% full thread depth.  It is a very common practice to reduce your tap drill sizes to achieve "fuller" threads.  You must really watch your process in aluminum, clean and lube the tap with EVERY hole, or you'll just be making scrap (or extra work for the edm dept.)  Another method to achieve stronger threads in wrought alloys is to use threadforming taps.  With this method you are literally cold working the metal, which gives more strength than a cut thread.  The grain pattern of the metal now follows the "vee" pattern instead of being chopped through on every thread.  Good Luck!

As a matter of interest has anyone measured the additional strength you get from a full thread form rather than an 80% one? If you draw out the deflected shape of a thread it is hard to see how the tip of the soft thread really contributes.

Cheers

Greg Locock

Greg,
Nice answer or post!
Do you mean calculate or do you really
mean measure?  I have seen calculations
in machineries handbook that calculate
strength based on the actual diameters.
It would nice to have a graph and post
the gain from 68 to 80 percent threads.

If you can tap thru there are some
special taps roll form type that require
larger size hole as Greg Locock mentions
and does a nice job.  The tap depth is
dependent greatly on the grade of bolts
that you are using.  No one has mentioned
class of threads.  I think this too would
be a factor not so much for strength but
for dynamic reasons.

Folks-
Tapping into soft metals such as aluminum is a poor design practice:
A.  Since most screws are steel, you end up with dissimilar metals (steel on aluminum) which will cause corrosion, pitting, galling, etc. at the threaded interface.
B.  Aluminums are usually softer than the steel screws and are apt to strip.

Best practice is to use clinch nuts or helicoils where the contacting metals are galvanically similar.

Well diamondjim, being a cynic I meant measure, but since it is Christmas I'll take calculate! I can see that there is some benefit in filling as much of the hard threadform with soft material as possible, to stabilise it as it fails, but would honestly have thought that the shearing behaviour at the root of the soft 'tooth' was not much affected by the presence or lack of material at the soft tip. Therefore I would expect only marginal gains in strength by increasing the proportion of thread depth in the soft material. I'm basing this on a hand-waving model of how the tooth deflects under load - I've never studied it.

The way we do it in practice is to tap the aluminium the first time and expect to helicoil if the bolt ever comes out.

Cheers

Greg Locock