Custom Steel Truss Connection

[designsimple]

I am working on connection detailing of very long custom steel trusses. They are about 168 ft long, 28 ft deep, 8 bays. I use W14-shapes as top and bottom chords, and steel tubes as vertical and diagonal webs. What is the best way to connect them? I am thinking use gussets to connect them, weld gussets to chords, notch ends of steel tubes, use long bolts to fasten them together. The sizes of square tubes about 12". However, I don't think it is good anyway. Anyone has better suggestion on the connection or even selection of truss members?

Thanks a lot fo any suggestion or idea.

 

[designsimple]

Any suggestion on how to break the top and bottom chords is greatly appreciated too. Thanks,

 

[Prestressed Guy]

For the gusset to HSS web connection long bolts are not a good idea. The bolts would be in bending instead of shear. Have you considered back-to-back channels for your webs? How heavy are your chord sections? If they are very heavy, bolted gussets may be a better choice to avoid the problems with welding thick sections.

Most of the long span steel trusses I have experience with use bolted splice plates to the flanges and webs for chord connections.

 

[kslee1000]

For truss size like this, I suggest to use open web tube for all main members, and double channels for secondaries. Bolted/riveted connections are preferred. The opening on tube serves as hand hole for bolt installation, air vent, and for drainage and maintenance purposes.

Chord members should be made continuous by cover plates, it could be made away from pannel point to avoid congestion, especially true for 3D truss.

 

[designsimple]

Thanks for replies. kslee100, what do you mean an open web tube? My loads are a little big larger. I am afraid of open web tubes are probably not strong enough. I think double chanels for webs is a good way too. My trusses have 9 vertical webs including two ends, and 8 diagonal webs which are symmetrical about the mid span. The maximum axial force of webs is about 700 kips. I am thinking to design it in this way: W14x120s as top and bottom chords, double channels as webs, cutted T-shapes (from a large W-shape)as connectors like gussets, bolt webs, chords and cutted T-shape connectors together, no weld. Any comment and new suggestion about this is truly appreciated.

 

[slickdeals]

If you are using a W14 member for your chords, they typically have an inside dimension of ~12.5". You may want to consider using W12 members for your diagonals and verticals and then welding gusset plates to the tip of the W14's flanges. This will give you the option of bolting the W12 to the gussets.

The gusset to W14 flanges could be shop welded and the W12's could be field attached prior to erection.

 

[slickdeals]

I am surprised as to why the 168' deep truss is 28' deep. You could easily get it to work with 14-18 foot depth. Is there an architectural reason for it to be 28' deep.

You will also be able to lower your web sizes if you reduce their lengths. A pratt configuration is your best alternative if you don't have uplift to deal with.

 

[designsimple]

slickdeals:

Actually the whole length of my trusses is 336 ft long but there are supports at the mid-span of the whole length of 336 ft, either by columns or by other deep trusses. Yes, 28 ft deep is for mechanical facilities and duct placements.

I don't like to weld gussets to the flanges of chords. I just feel it is not good especially considering seismic loading.

By the way, the W-shapes of top and bottom chords are placed vertically, not horizontally. If they are set horizontally, then a huge heavy W-shape will be required. This is a not option in my mind anyway.

Thanks for your inputs.

 

[slickdeals]

Are you carrying significant loads that are inducing bending in your chords, that require you to use them vertically as opposed to horizontally?

In addition, you might use intermediate bracing in your web members to reduce their sizes based on a large unbraced length.

 

[designsimple]

slickdeals, yes, the chords have got to carry 60-90 psf loads for 28 ft distribtion width. I can't place any intermediate bracing at those webs because mechanical ducts run through those spaces.

 

[designsimple]

I don't have my steel manual at hand now. I'm afraid that standard largest double channels might not be strong enough for the compression of 700 kips for 35 ft long webs.

 

[kslee1000]

It's really an open web box beam (tube with circular holes), there are many applications on long span truss bridges for its strength (especially for torsion) and ease of making connections (flush all sides). You can build your own using double angles (with web standing) laced together by benton plates and/or bar straps.

Wide flange can be used successfully for your case, just watch out its orientation to avoid difficulty in making good simple connections. You are correct in using all bolt connections, except for minnor areas, weld should not be used.

 

[designsimple]

I can't use open web box beams since there are lots of purlins running at 4-6' O.C. and connecting to the webs of the top and bottom chords. Also, the top chords at the roof level are slightly sloped more than 1 deg.

I just checked my STAAD analysis. It looks like I am not luck enough. The largest channle MC18x58 is just a little bit short for strength, about 5-8% less to meet the required design strength. Does anyone know there are some manufacturers make larger non-standard steel channels than MC15x58 in US?

Thanks,

 

[kslee1000]

You may consider using angles and plate to form a build-up channel.

 

[designsimple]

Built-up channels is an option but the fab/cost is a concern. Nobody want to pay extra dollars these days.

 

[slickdeals]

For long span trusses such as these, using Wide flanges is the way to go.

It seems like you cannot get any additional bracing for the weak axis. If you could, I would brace the weak axis at midpoint and design it that way. I am sure the mechanical ducts don't need 20' openings.

 

[kslee1000]

I agree that these days no one like the idea of build-up shapes because of high labor cost and quality issues. However, your case is one of the kind, may forced to do so no matter what shape you are using (wish it not going to be this way, but highly likely due to size and loads). The only reason I prefer to get away with wide flange is because of concern over connections. For strength and economical reasons, it is logical to set beam web standing and use single gusset plate to connect column and diagonals, you can see the geometry may give you a lot of head aches. As opposed to this option, double channels and box would simplify and strengthen the joint by allowing double gusset plates set to the side faces of members.

Is this a double trusses with duct running in between accross the 336' length? Without adequate bracing, you might need to stiffen your joints to assist in lateral stability (not a easy task).

Anyway, one more suggestion is to avoid mixing different shapes in main members (chord, column, diagonal), which would complicate the design

 

[designsimple]

Anyone may give better suggestion on the connection if steel tubes are used as vertical webs and diagonal webs and W-shapes as chords set vertically? I am thinking to notch tubes at ends, weld two gussets to the notch, and weld one gusset/T-shape to flanges of chords, and then bolt them together.

Thanks,

 

[LobstaEata]

designsimple

How about using a gusset connection shop welded to the WF and into a split tube stub. The remainder of the diagonal or vertical tube member may be continued from the stub using a sleeve splice and connected on both sides of the splice with holobolt fasteners. See

http://www.lindapter.com/page_product.cfm?pgroupid=3

 

[civilperson]

Weld a plate in a slot at the ends of the HSS members and bolt to a welded plate on the top and bottom chords.