steel shear connection modification 3

[a2mfk]

Sorry its a bit long, trying to answer all the questions I would be asking... I am not the EOR, just designing the steel shear connections.

-Standard AISC steel shear connection, single shear plate bolted to the web of a supported beam and welded to an embed plate support in a concrete tilt-panel.

-Existing building, finishes up and this will have to be a weekend job probably.

-Third floor beam, temp shoring and jacking pretty much out of the question.

-Dead loads and some live loads (obviously more or less depending on the time of day) in place, building occupied.

-It was determined the existing shear connection was about one bolt short or about 6 kip.

1. Does AISC allow you to add the strength of an additional weld from the shear plate to the web of the now-bolted beam?

2. To maintain a connection that allows some rotation to maintain a pin-pin assumption, the weld could be placed only near the bottom. I'm winging this email so pardon my mechanics of materials memory, can't remember where exactly the least amount of movement at the end of the beam would occur..

3. The bolts should be in full bearing condition and local
deformations for the most part should have taken place inside the connection, correct? I know it depends on % of dead and live load, size of the holes, etc.

4. Alternatively, what about adding a shear plate to the opposite side of the web and turn this into a double shear connection? They could take the nuts off of the existing connection, slide another shear plate on, weld it to the embed plate, then tighten up the nuts....

The problem I see with that is making sure the bolts are bearing on the new shear plate once it is welded up there , so it is not just there for peace of mind. I am worried the connection would have to yield slightly to begin bearing on the new plate, but maybe that is acceptable? Would I have to use a slip critical connection to ensure load transfer?

5. Other options include possibly a angle seat below the beam, or drilling the holes larger and replacing the bolts with 7/8" dia.

Any references to AISC or other advice would be greatly appreciated.


Thanks in advance,

Andrew Kester, PE

Florida

 

[Lion06]

1) See J1.8 in AISC 360-05, if you provide slip-critical bolts you can let the bolts take the existing load and size the weld to take the additional. If you don't have clip-critical bolts you can use 1/2 of the bolt capacity based on bearing type bolts.

2) I think the connection to the supported beam is the place where you DO want moment transfer - this will allow the connection to deliver a pure shear to the embed plate and not an eccentric shear (or shear and moment).


4) This would require temporary shoring, no?

 

[Lion06]

what about leaving the bolts in place and providing a fillet weld around three sides of the shear tab to the web of the supported beam?

 

[a2mfk]

StructEIT
Welding WAS my first choice, my concerns are #1 and #2... That was the reference I was looking for. The in-place connection is bearing, non-slip critical. According to Chapter J, I'd have to size the weld to take the entire weld, in which case it would be too rigid of a connection IMO. But I need to do more research, I understand what you are saying about pure shear, but as the beam deflects and rotates, it will transfer some moment into the embed plate which we do not want...

I think that makes these the two most logical:
a) Add another shear plate and make this a double shear connection
b) Drill larger holes and replace 3/4" bolts with 7/8"

I do not know if these would require shoring, that would be a huge problem.

 

[paddingtongreen]

You could replace the bolts, one by one, with bolts for a slip critical connection, torquing as you go (I would use TC bolts), and then do the make up welding.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Lion06]

A2mfk-

I think rigidity in the connection to the supported beam is a GOOD thing. It's taking demand off of the embed plate. If the connection to the supported beam is a pin, then the shear tab is a cantilever off of the embed plate and the plate is taking shear and moment. If the connection to the supported beam is rigid, then the load to the embed plate is a pure shear (not shear and moment). This is beneficial to the embed plated.

 

[a2mfk]

Michael,
I could ignore the existing bolts and just weld the existing plate to satisfy the required strength of the connection.

However, I still have concern with this now becoming a stiff connection and moment being transferred into the embed plate. But there was moment before as a simple shear bolted connection because of the eccentricity, so the EOR would just have to OK our fix.

 

[nutte]

Welding the shear tab to the web of the supported beam is not a good idea. There has been no research of this connection (a shear tab welded to both members, supported beam and supporting member), so we don't know how it behaves, as far as maintaining your pinned support assumption.

Adding a second shear tab and getting the bolts in double shear is a good idea.

In the original connection design, were the bolts designed as N (threads included in the shear plane) or X (threads excluded from the shear plane)? If they were N, you could check them or change them to X, gaining more strength. You could also go to a higher bolt grade, or to larger bolts, although this might pose edge distance problems.

Again, welding the shear tab to the supported beam is not recommended by AISC.

 

[paddingtongreen]

a2mfk & nutte, the beam is already loaded. The slope is already in the end of the beam. The only reason I can think of for not doing it is if you have seismic or fatigue problems.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[a2mfk]

Michael
Dead load is in place, portions of the live load will be in place. But if this is done in an empty building on the weekend, then additional live load will go into the newly welded connection, so additional beam rotation will occur. I just don't know if this is negligible or otherwise OK, as nutte indicated its not in my knowledge prequalified....

Trust me, I want to do this if I can justify it.

No seismic or fatigue issues.

 

[a2mfk]

StructEIT,
I understand your point exactly. Welding the shear plate will reduce the moment (already partially loaded) caused by the cantilevered bolted shear plate, and replace it with a "partially"-rigid moment connection. Thereafter, additional loading will produce moment into the PRM connection... Sorta robbing Peter to pay Paul in terms of the embed plate.

I am going to look into some on Nutte's suggestions about N vs X, and using higher strength bolts. We may be able to milk an additional 6 kips out of one of these techniques. OR, I like adding the shear plate and making this a double shear connection.

I'd rather not change the loading into the embed plate as this is by the EOR, who can reject any of my proposed repairs.

Thanks for everybody's suggestions!

 

[connectegr]

The original single plate connection lacks flexibility. This moment should be considered in the design of the embed. If this is considered, then the addition of weld to the beam web should be the simplest solution. The weld can be designed for all the load or in combination with slip critical bolts. Or an additional plate can be added on the far side, putting the bolts in double shear. The additional plate should be partial pen welded.

I would have recommended knife angles or single angle welded to the embed. This is a flexible connection and therefore the embed can be designed for shear only.

http://www.FerrellEngineering.com

 

[connectegr]

nutte is correct, welded/welded single plates are not recommended by AISC. This is primarily due to the lack of research, due to the impracticallity of the connection. The connection lacks flexibility, so the moment must be considered in the embed plate and studs. Knife plates are similarly inflexible.

These are considerations that should have been made in the original connection and other similar locations.

http://www.FerrellEngineering.com

 

[a2mfk]

connectegr,
Thanks for the input. This is a fab bust, I am trying to fix it for the fab, I am not the EOR. As it is, I am about 6 kips short. I can only assume the EOR originally considered the moment from the cantilevered shear plate.

As stated above, I am concerned that the moment I am adding due to fixing the connection would be more than that from the effects of the bolted shear plate that I am eliminating...

Pretty much have my answers though, I am keeping it simple and adding another plate to the other side of the web. That way I have not affected the original connection design intent and fixidity condition.

 

[connectegr]

Remember to use a partial pen weld for the new plate, not a single fillet weld.

http://www.FerrellEngineering.com

 

[a2mfk]

If I can get higher strength bolts or "x" type bolts to work, does anyone have experience with replacing bolts one at a time in a constructed building? It would seem the bolts would be in bearing and it would be nearly impossible to get them out, but I am not a steel worker with tricks up my sleeve.

Third floor beam- temp shoring and jacking are off the table...

 

[a2mfk]

connectegr- why the partial pen weld vs fillet? I may end up having to use that solution and I can't off the top of my head understand why a fillet isn't OK in this situation..

 

[connectegr]

"x" and "n" bolts are the same bolt. Unless your bolts are threaded full length. If your outer ply is greater than 5/16" thick and the bolt length is sized according to AISC, then you have threads excluded from the shear plane.

AWS prohibits the use of one sided fillets in tension. The eccentricity of this connection applies shear and moment on the weld. A partial pen provides weld more concentrically about the centerline of the plate.

http://www.FerrellEngineering.com

 

[dhengr]

A2mfk:
Could you weld a small seat angle below the beam to pick up the 6 kips? I would jack this up to be sure it was seated before welding, but you don’t have to unload the whole beam, and you won’t change the stiffness of the existing connection conditions appreciably.

RE: “AWS prohibits,” Connectegr’s last statement: it’s the prying action and multi-directional stress conditions at the root of the fillet weld that’s the killer. Loading parallel to the root works fine, but loading across the root works less well, it is the weak link in the weld when the stress field is perpendicular to the axis of the fillet. Although, the partial penetration weld root can be just as problematic under like loading and stress fields. So, as Connectegr says, it’s a matter of degree for you to puzzle over.

 

[a2mfk]

Interesting... But why is it allowed in a standard double angle connection that is welded to the support on only one side and a slight return, and bolted to the beam? I would think this is the same condition. Because its been tested for years and prequalified?

And since that is allowed by AISC, I would think that would be a solution if I have room on the embed PL.

I will obviously check to see if the threads are in the shear plane first. Thanks to both of you...