Don't know what code(s) may be in effect for your location or application, but......
ASME B31.3 335.2.3 says that one thread short of full engagement is acceptable. From a practical point, I like to see at least one thread past the face of the nut. It's much easier for an inspector to see.
Keep it to a minimum if at all possible.
The more threads you leave past the face of the nut the more threads there are to rust/corrode and bind, possibly ruining the nut.
It depends on what code you are building to.
donf gave you the B31.3 requirement.
AWWA C207 required that "minimum bolt length shall be the sum of the mating flange maximum thicknesses, the gasket, and the depth of the nut plus 1/2" before torquing."
it depends whether you need a hydraulic tensioner to preload your bolts or not. Large sizes of bolt sometimes need a hydraulic tensioner to apply the loading so you need enough threads for the hydro-unit to engage on the threads. two threads showing will be nowhere near enough.
jtseng132, do you maybe have some reference for that? although it sounds quite straightforward to have sufficient threading prior to hydrau-tensioining, Id like to have some background reading on that for more details.
(basically if you take maximum axial (unbolted) thicknesses of mating flanges, gaskets and nuts etc., bolting should extend only 1/8" beyond nuts.
There are however other standards, and many references further explain that from a strength of materials or mechanical perspective far lesser engagement can normally develop full bolting strength. On the other hand, you may well need more protruberance for good reasons, e.g. when bolt and nut material vary or when tensioners are used etc.
If your piping design will be federally regulated under the DOT (transportation pipelines under ASME B31.4), you'll want a min of two threads showing past the nuts on all torqued studs. This makes it easier for DOT inspection to verify. If you order flange studs per the ASME codes, you'll end up with nuts flush with the stud end - which is not ideal. We always order studs slightly longer as a result.
I have a project many years ago that bolts to be tightened by bolt tensioner on many high pressure (>1000 psig, Div 2)vessels. The field contratcor told us the length to have one diameter minimum so the tensioner can grip on it. It is not a guessing. And it makes sense because normally the high of the nut is less than the diameter, so one diamter extra length is enough to engage the bolt tensioner. You can check with bolt tensioner supplier to confirm.
