hss tubes - blind connection or through bolts

huangyhg

hss tubes - blind connection or through bolts
the application i have is to connect a large, square steel tubular section (hss) to a connection "socket" made of plate steel.  the item is being used for a cantilever arm 38 feet in length, supporting a 3700 pound sign panel at it's end. the hss member is ts 16 x 16 x 1/2", and the socket "flanges" will be at least 3/4" thick.
the difficult part of this connection is the tubular member will require a blind connection -- the nut side of a bolt will not be accessible! the connection forces should be primarily shear, without applied tension.
does anyone have any experience with these types of  connections, or should i use through-bolts instead?
thanks in advance for your consideration of this problem!

i would be inclined to thru bolt, but there are a couple of issues that i can forsee:
1. you would need very long bolts, 20" or so of grip, so getting a325 is out of the question.  probably need to use high strength rod w/ threaded ends.
2. lining up the holes during erection may (or not) be problematic.
3. need to connect both vertical (gravity) and horizontal (wind/seismic) axes of tube, so connection area will be very congested with bolts going in two directions.  carefull planning of holes locations required.
all the above need to be considered, but can be easily addressed.  so i would still use thru bolting, since you know for sure what you will end up with.
if you want to blind-bolt, there is a product on the market specifically for blind bolting of tubes, called the lindapter hollobolt.  i don't know much about it, as i have never used nor researched them.  but here are a couple links:
manufacturer:
as i said before, i am not a great fan of huck bolts. used a zillion of them.  i would still through bolt or weld nuts with high strength bolts but if you wish to use the huck system (it does not sound plausable in this instance) duncan bolt company of santa fe springs, ca. supplied ours and their people were of great help when we needed it.
thanks for your advice, andrew & rod !
andrew is correct the connection will be very congested. currently, it looks like 15 through-bolts in each face, staggered 2" apart (4" gauge & pitch). while adding sleeves would be great, the interior of the hss is inaccessible during installation. sounds like internal stiffeners might help, though.
i am not sure the grip will be 21", since the plies connected only add up to about 2½". the overall length of the through-bolt will need to be about 21", with 1½" or so threaded at each end. it appears a449 rod should be used in leiu of a325 bolts.
i had reviewed the hollobolts previously. their reported shear capacities are closer to what an equilivent hs bolt provides. i'm not sure they approve of reversing the expansion cone, as discussed in the referenced report.
thanks for sharing your insights!   raf
i have never used the blind side bolts, but they are probably worth checking into.  on similar smaller tube type connections, i have used a307 bolts as thru bolts.  the only problem i see with thinner wall tubes is that pretensioning the bolt might start to bend the tube.  i know that a325's don't necessarily have to be pretensioned but a lot of building departments want to see you "torque" high strength bolts wether or not it is needed.  they see a325 on a bolt, and they automatically want to torque test the bolt.  you could of course explain to them why they don't need to torque test the connections in a subsequent letter.  some building departments will take it some won't buy it.  so i whink it is better to avoid the issue in the first place.
jcox --
you are correct about tube distortion under full torque. aisc's design spec for hss indicates snug-tight installation only for through-bolts. evelrod suggested using sleeves, but they would have to be fastened prior to fitup (see thread 725-68729). thanks for your advice!  raf
just another thought:  what about welding the socket flanges to the tube faces?
fit up might be a problem.  but you could cut slots in socekt flanges along axes of tube, to gain more weld length, if three sides weld not adequate.  
i figure you get a bending moment from dead load of about 141 kip-ft.  so, you would get a couple force of about 106 kip.  this would require at least 23 inch of 5/16 fillet weld (@ 4.64 kip/in allowable)  thus, if your socket flange was 12" wide and covered over tube about 8", you would have enough weld length to develop the required force.  
obviously, other conditions need to be checked, but this sounds like it is in the realm of being feasible.
just my 2.5 cents worth.
andrew
whatever system you use consider corrosion of the connection inside the section, there may be no indication of failure until it's too late.
greetings to andrew & "wahmit" :
the issue of corrosion is a concern for these huck bom's -- i don't think the mandrel-sleeve junction is watertight. while the structure will be galvanized, the connectors won't be.
as far as welding, the gravity loads create about 128kft moment. wind loading produces about 200kft. my numbers indicate a force of 400k at the tube corner, which will require about an 1-1/4" weld to resist!
my concern with field welding is that i'll get one great weld (top face), two crappy welds (vertical) and one worthless weld (overhead). that's a lot of field welding! hence, a bolt connection would be much more reliable -- if it can be done.
thanks for the ideas and issues!  regards,   raf
if it is acceptable an external pipe type flange, with stiffeners if required, would save you a lot of worry.
i like the pipe flange deal.  most of the cantilever signs in socal are of this design, but not 38 feet long that i have seen!  anyway, with a pipe flange connection all you really have is a 'demountable' welded connection with the top flange taking ~80% of the load!
rod