Very Non-Standard Connection, Would like an Opinion.

[IngDod]

Greetings, I am currently designing a moment connection between an HSS column and an HSS beam, I am forced into this configurations and although I would very much prefer to use W-shapes it is out of the question. Initially I tried to simply weld the beam to the column, but this lead to a myriad of problems such as plastification of the column wall, punching of the column wall and such. Currently I am trying to use stiffening plates to keep the wall of the column from deforming. However I know of no hand calculation to design or even estimate the design of this connection, this led me to rely on a FEA program (SAP2000), which I very much dread since I have to rely completely on the computer. What I did is model the connection without the siffeners and check that its behavior is as expected (miserable failure of the the connection to behave as fixed) and then I proceeded with my stiffener plates scheme. The general behavior of the connection is as expected, the HSS walls are no longer deforming as the plates give them considerable resistance. I am seeing very high stresses near the corners of the plates.. But this is a linear FEM Analysis so no stress redistribution after yielding is being accounted for. I will attach Images and printouts of the connection and would very much like to hear your input, If anyone can suggest a method for hand calculation I would greatly appreciate it.

 

[dhengr]

IngDod:
Regarding the top stiff. pl.... The ears on my stiff. pl. (along the col. sides) would be 5-6cm wide, as your’s are, but they would be 25-26cm long (not your 22cm) before they started tapering into the 9cm beam tube top flg. I would increase your 18cm length to 24-30cm, further out onto the beam tube. At the column face, place the center of a compass at the face of the col. and at the center line of the beam and draw a half circle cut-out, the circle dia. is 20cm, the col. width. Except, I would like the cut-out to be elliptical in shape (minor dia. 20cm, major dia. 30cm), going out onto the beam a little further than 10cm. My top stiff. pl. is “Y” shaped. The large corner blue areas will go away, I’ll bet. I agree that tapered or square won’t make much difference on the stress flow in the top stiff. pl. but it is probably being cut out of a larger pl. by automatic cutting equip., so why not clean it up? Eliminate the middle stiff. pl. it just complicates the assembly. The bot. stiff. pl. should be tipped up about 30̊, made longer, and becomes the bot. flg. on the hunch, but it is cut square to fit the col. shape. Alternatively, the hunch could be cut from a piece of the beam tubing; then the stiff. pl. can be shorter, but I’d still tip it up to match the hunch slope.

Don’t weld around corners, stop welds on the flats of the tubes. The longitudinal welds btwn. the stiff. pls. and the corners of the beam tubes are really difficult welds to do well. Yes, you should include the load inputs from both the weld across the top of the beam tube and the two longitudinal welds btwn. the stiff. pl. and the corners of the beam tube, but do not try to connect them by welding around the corner of the pl. at the tube corners.

 

[wannabeSE]

IngDog - You may have already considered the following:

Make sure the architects won't have heart attacks when they find out the connection projects 50mm (2") beyond the face of the column.

Are the 3/4" fillet weld on both sides? Consider partial penetration of full penetration welds.

With such big welds, it may be worth checking for base metal rupture of the column (In the US, 5/8" is the maximum wall thickness for HSS8x8).

Check the beam for the concentrated load from the haunch. Stiffeners may be a good idea even if the beam can handle load. It could be informative to extend the beam past the haunch in the FEA model.

 

[IngDod]

@dik: Thanks, I already have this book and most of the CIDECT guides. They are a very good resource.. but they are remarkably light on hss to hss moment connections.. they do say that directly welding them is a big no-no for developing fixity unless the sections are of similar width.

@dhengr: I understand now, well honestly I rather leave them as squares cause most likely this things will be cut using a hand-held torch (I live in the third world).. so any fancy tapering could come back and bite me in the rear. I will most definitely keep your suggestions about welding in mind.

@wannabeSE: Im pretty sure he will, but he is the one who wanted to have a 40feet clear span while supporting a steel deck. (this beam is supported on another beam at 20 feet.. just don't want to scare anyone thinking this thing spans 40feet on its own). I will have a pretty tough time designing the welds. I will keep everything you told me in consideration when designing the welds. I had the beam going past the haunch before.. I did see some stresses going over the limit but only in the vicinity of the haunch (no more than 2cm i would say) the rest of the web was fine, I will have to remodel and check again.. the problem is that I am applying the moment to the connection by using forces on the joint of the top web, directly above the haunch, so this will cause an artificial high stress in the haunch.. I will have to do a separate model for this.

I finally modeled the whole connection, it has a very large beam framing on one face and two more beams framing perpendicular to this one. Since I had to greatly expand the plates to take the other beams Im seeing much better stress distributions and my plate is only 1cm thick. However I am having a hard time finding out the proper way to get the design forces for the welds.. I mean I can probably get it with the FEA, but I would like to be able to do some ballpark estimating by hand. Would summing the tension forces (as vectors) coming from all three beams top flanges and using the resultant be an appropriate way? I am attaching my latest results, please comment.. this easily the weirdest connection I have ever designed..... Now I have to do two more.

 

[IngDod]

@dhengr: I will definitely try to angle the bottom plate with the haunch, I will look much better and be more effective since its aligned with the load.

I have a question.. If I size the welds to be enough to develop the rupture stress of the base metals, would this not preclude a failure of the welds? I mean, the hss or the plates would have to tear before the weld fails.

 

[dik]

Since the Arch is being a little difficult, can you CJP a solid block of steel at the connection point and weld to your hearts content?

Dik

 

[IngDod]

Well i was just trying to be funny... This is a steel deck structure so most of the connections will be hidden behind a drywall ceiling... so the aesthetic of the connection are not very important. I will prefer CJP.. but I am concerned that the hss walls are very thin.. only 0.55cm... Also a concern of mine is the skill of the welder, here welders are not certified as in the US. Some are very good, some wont know the difference between a fillet and a CJP.. I was speaking to one the other day and I asked him.. how you make the weld a certain thickness? he told me... "I Just make many passes".

 

[wannabeSE]

The architect may be concerned if the connection interferes with the framing for exterior cladding. Or, if it conflicts with studs for walls used to conceal the columns.

 

[SAIL3]

These are really large loads for this conn....I get 60kips(approx) for T&C....60/.707=84.8 kips @ the haunch...how are you going to transfer this to the col?....the majority of the T & C loads will be reacted by the sidewalls of the col so the center portion of these pl's will not see any appreciable load...the load basically stays in the outer region of these pl's...I attached another concept that may be viable but you may run into problems with getting at some of the welds...another option is to play around with differnt width bms inorder to accomodate some of these welds....also this is imparting significant shear loads into the col sidewalls and when combined with any other existing loads in the col may need to be checked.

 

[IngDod]

@SAIL3: I greatly appreciate you taking the time to help me with this.

The connection in your sketch would be ideal, I believe the intention with this is to have the gusset transfer the load directly to the plate?. In my design the haunch loads the wall and then the wall loads the plate I believe (the wall tries to punch in and the plate restrains it). Yes, my intention with this stiffening plates is to allow the T and C force to travel from the front wall (the one to which the beam is framing) to the sidewalls without punching or deforming the front wall. The way I first envisioned the load path is: The beam flanges push and pull on the HSS walls, the walls then pull and push at the plates, which keep the walls in place and distributes the load to the sidewalls. Although I believe that by welding the flanges to the whole connected length of the beam I am transferring the load to the plates and then the plates transfer to the side-walls; however this is not the case with the bottom plate as I have it now, as the haunch bears on the wall without being connected to the bottom plate. I will definitely try to see if your connection detail is viable, I will have to talk with the welder.

This columns are loaded mostly by the moment coming from the beams.. their DC ratio is around 0.2 Compression and 0.8 Bending. I did not manually check interaction of axial and moment stresses with the shear coming from the T and C forces.. this is something I will have to do.

 

[charliealphabravo]

Due to the thin walls of the sections, moment connections in HSS will begin to look more like space frame connections as the moment gets higher. The stiffened connection will begin to look like a socket that the tube plugs in to.

Here is just an example I googled of a wide flange beam in an HSS column (see figure 3). You can see that the entire HSS section is wrapped to ensure the integrity of its shape...

http://www.atlasconnection.com/prof...tions-under-seismic-loading-for-moment-frames

I think the design you show with stiffeners on the top and bottom is a holdover from wide flange theory where most of the moment goes through the flanges. This will not be the case I think with HSS and there will always be tearing and distortion.

The most important thing to reaching the full moment capacity of the section is to ensure the integrity of the entire HSS section shape. I think you will find that using gusset plates and stiffeners to create what is effectively a welded socket is the best way to do this even though it appears redundant and impractical in wide flange world. The benefit is that the connections should be very straight forward to design.

I have sketched what I think the connection might look like if the beam and column are the same width. If you want the column wider than the beam then the gussets and stiffeners will need to transfer the load not the HSS.

 

[SAIL3]

another item to check.....assuming the t of the col wall is 0.55cm=.22 in......T=C=60kips...that is 60/2=30 kips per sidewall of col.....shear area=(8-3x.22).22=1.62in^2.....shear stress=30/1.62=18.6ksi....also I would check for shear panel buckling in the col sidewall between the top and bottom rings...I would either increase the wall thickeness or if that is not possible add a flat pl to the sidewalls in that area....
charlie brought up some good points and referenced an interesting site...

 

[IngDod]

@Charlie: Thanks for the link and for the sketch, I had read that website during my research. The connection you propose is exactly the same as the one proposed by CIDECT to develop a rigid moment connections in Vierendeel trusses, I tried that but there's no way I can get a beam wide enough to match the column and the architect would undoubtedly object to the haunches protruding from the steel deck. I had a conversation with the manufacturer of the HSS in my country, they tell me that in this cases the ring devices I am trying to develop have been used in the past, they told me: "The plates act as a diaphragm by distributing the stress to the walls of the tube", they also told me I should keep the thickness of the plate at no more than twice the thickness of the hss for stress distribution reasons.

@SAIL3: I have not managed to get to this part as I was just requested to modify the structure (architectural reasons), this presents me with the opportunity of increasing the beam size and getting rid of the haunches, which actually lowers the moment on the connection by about 15%, And would simplify the connection. Im afraid that's as thick as the walls of these tubes get, the steel is Fy=50ksi and Fu=62ksi so it might hold, welding a diagonal plate from top ring to bottom ring could help in this (similar to what is done with W-Shapes) and I will also check the panel buckling.

Again thanks to everyone who has contributed to this thread.