Fatigue Capacity of Bolted Tension Members

[TEDstruc]

I'm looking for opinions on whether or not bolted tension members that see cyclic load can use snug tight bolts (bearing type connection). In this case, lets consider an angle tension hanger with a bolted connection (bolts in shear) that supports a piece of equipment that sees some cyclic load. The company I work for (an equipment manufacturer) installs thousands of hangers like this using a specified bolt installation torque. I've estimated that for non-PT bolts we are getting the bolt to about 50% of a fully pretensioned bolt (on a 3/4" bolt say 16 kips of pretension vs 28 kips for a fully pretensioned A325 bolt).

Most of the hangers installed are seeing relatively low cyclic load that would not exceed the fatigue threshold stress (Fth) that is specified in Table A-3.1, Section 2.3 (for rivets or snug-tightened bolts). However, in the commentary under section 3.4, it states that non-pretensioned fasteners are not permitted for joints subject to cyclic shear forces... If non-pretensioned bolts are not allowed in this application, why is section 2.3 included? Is it just to cover existing connections that aren't fully pretensioned?

 

[WARose]

I'm looking for opinions on whether or not bolted tension members that see cyclic load can use snug tight bolts (bearing type connection).......However, in the commentary under section 3.4, it states that non-pretensioned fasteners are not permitted for joints subject to cyclic shear forces... If non-pretensioned bolts are not allowed in this application, why is section 2.3 included?

I've always been a little confused by that too. I've always suspected (based on the Journal articles I have read) that slippage (followed by slamming) is probably playing a role. (Along with concentrations and natural material weakness in shear.)

But to answer your question, in that vein, I typically don't use snug type bolts in tension applications either unless I am convinced the application will be free of any slamming type issue. (I.e. the tension cannot be overcome.)

 

[ProgrammingPE]

AISC 360-16, Section J3.1b(1) refers you to the RCSC to determine when bolts need to be pretensioned. RCSC 2014, Section 4.2 says that pretensioning is required for "joints that are subject to significant load reversal" and also "joints that are subject to fatigue load with no reversal of the loading direction." I imagine that that is the provision that the commentary for Section 3.4 is referring to.

If non-pretensioned bolts are not allowed in this application, why is section 2.3 included? Is it just to cover existing connections that aren't fully pretensioned?

Section 2.1 and 2.2 are for slip-critical connections, so it looks like pretensioned bolts should fall under Section 2.3. Regardless, Section 2.3 would still be provided to check riveted connections.

(For some reason, AISC 360-16 changed the description for Section 2.3 to be for just "riveted joints", but they didn't change the graphic to remove "snug-tightened bolts".)

Structural Engineering Software:

http://www.structuralcentral.com

Structural Engineering Videos:

http://www.youtube.com/channel/UCOj2mXXf3ZbZtgxTc6BtkGg

 

[TEDstruc]

Further to this topic, do you include shear lag when calculating fatigue capacity of a bolted tensile member? Per section 2.2 and 2.3 in Table A-3.1, it just refers to net section, not effective net section. Obviously in the case of section 2.1, you are simply looking at gross cross section.

 

[ProgrammingPE]

I would still use the effective net section which includes the shear lag factor since that accurately represents the stress distribution. The cracking originates at the side of the hole and the outstanding leg is not going to reduce the stress there if it is not attached to anything.

Structural Engineering Software:

http://www.structuralcentral.com

Structural Engineering Videos:

http://www.youtube.com/channel/UCOj2mXXf3ZbZtgxTc6BtkGg