Eccentricity on embed plate due to shear load

[NewbieStruct]

I have a difference of opinion with my senior engineer regarding eccentricity. We have a situation where a double angle connection is used to attach a steel beam to an embed at a concrete tilt panel wall. He says there will be no tension in the embed studs since the eccentricity of the shear load can be neglected as it is similar to attachment of beam to a column. The load is due to a composite beam attached to the face of tilt panel wall. The steel deck of the concrete slab is just above the tilt panel wall.The very first row of studs has an edge distance of 2" from the top of the wall. If I run the analysis using ACI 318 appendix D without the tension loads the connection is fine. However, with the tension loads the connection fails. If I ignore the first row of studs the connection is fine (edge distance is now 8") with the shear and tension loads. Is this ok to consider? Should the connection be designed without eccentricity ?

 

[Lion06]

If you meet the welding requirements for the outstanding legs (short returns at the top, not fully welded), then you can ignore the eccentricity as stated for the double angle connection per AISC. I would personally feel uncomfortable ignoring the eccentricity completely for the embed plate checks, especially with such a small edge distance at the top. Can you lengthen the embed plate and increase the edge distance at the top? 2" is too little edge distance, IMO.

 

[hokie66]

Connections to concrete panels are much more unforgiving than connections to steel columns. If the first studs are only 2" below the top of the wall, I would move the stud group down, or omit those at the top if the connection is adequate without them.

 

[NewbieStruct]

The connection has total 8 studs. The load is 28 kips. The studs are 6"o.c. If I ignore the top two studs, the connection does work in shear, tension and interaction.

 

[hokie66]

Then why not just leave off the top one? You don't want the top of the panel to crack.

 

[connectegr]

If the outstanding legs are welding vertically and with only a return at the top, then, the connection angles are considered flexible. Simple beam end rotation is takin in the deformation of the connection angles. In this case it is typically comsidered as shear only in the embed plate and studs.

http://www.FerrellEngineering.com

Providing fabrication and erection efficient structural design of connections. Consulting services for structural welding and bolting.

 

[hokie66]

connectegr,
I would agree for the steel part of the connection, and for this type connection to a steel column. But those studs which are only 2" from the top edge of the concrete panel will still tend to distress the concrete. Seen it happen.

 

[ztengguy]

I cant fathom how if you have an eccentricity, regardless of stiffness of the angle, you would not get a tension in the top row of studs.

 

[msquared48]

If the connection is not built yet, you could design a bearing plate at the top of the plate that could extend into the panel, or bear on top of it, removing the eccentricity to the anchors, and the problem.

Mike McCann
MMC Engineering

 

[TehMightyEngineer]

I'm with connectegr on this, a simple shear connection in a steel beam to a steel column will have the steel column much more rigid than the angles. Thus, the angles deform and the connection rotates. Same thing will happen with your connection, concrete embedded plate will be much stiffer than the angles and thus the angle will deform and release the flexural force. I'd say this would only not apply if your concrete embedded plate has such a low capacity that stud pullout failure would occur before rotation of the angles could occur.

Maine EIT, Civil/Structural.

 

[ztengguy]

Thus, the studs will have tension load in them along with shear.

 

[Once20036]

I vote no tension.

If you had a perfectly rigid connection: A load would be applied to the beam creating shear on the connection. As the beam deflected at midspan the ends would try to rotate. The rigid connection would try to resist this rotation an tension would be created in the bolts.

In an actual connection (as described by several folks above), as the beam ends try to rotate, the flexible connection allows this rotation. Because the connection doesn't resist this rotation, no tension is created and the embed can be designed for shear only.

 

[ztengguy]

Lets say the connection is as described, but the shear plate used to connect was 12" long, and the holes are at 9", flexible connection allows this rotation, still no tension?

In my mind, regardless of flexibility, you still have a load eccentric to the plate, so the bearing load would create tension. Perhaps someone could produce a free body diagram to convince me otherwise?

 

[hokie66]

ztengguy,
You can't convince these steel guys that a flexible connection isn't flexible enough not to cause distress in concrete. As long as they stick to steel, all will be fine.

 

[KootK]

I'll raise my hand in support of both Ztengguy and Newbie. I think that two issues are getting confused here:

1) Flexibility in the angle connection WILL reduce the moment transfer into the embed that would occur from the beam attempting to behave as though it were fixed ended.

2) Flexibility in the angle connection WILL NOT reduce the moment generated by the eccentricity between the embed plate and the location where the beam shear is delivered. I'm thinking of a column of bolts here (e~2.5"). If the bolts are slip critical, or the angles are welded to the beam such that moment can be transferred, then one could argue that this eccentricity approaches zero since the angles are merely a flexural extension of the beam. I suspect that the most likely scenario is long slotted bolts in which case I would not consider the angles moment connected to the beam.

It's a certainty that there is some moment in being delivered to the embed and thus some tension in the studs. By ignoring the moment / tension, you place yourself on the wrong side of the conservatism line. Beyond that, it's just a matter of degree.

Is there no direct tension on this connection from wind / seismic?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[ztengguy]

KootK, thats probably the best explanation that I have seen, but, as with a moment resisted by a plate with anchors, do you not end up with a tension load and compression load due to the couple?

I have done alot of embed plates, and ignoring the interaction equations, can get you in trouble. Lets hope not in this case.

 

[PMR06]

I agree with ztengguy & KootK, there is a moment on the embed plate. The same moment that is "deforming" the "flexible" angles outward IS THE MOMENT TO BE RESOLVED. The angles won't deform outward with shear only.

 

[KootK]

@ztengguy: I agree completely, all those moments that I was speaking of translate themselves into a T/C couple on the embed.

If I was newbie's boss and really wanted to win this argument, I'd claim that under moderate shear loads, even LSL holes will deform into a plowing condition and result in a fairly fixed connection between the beam and the angles. Thus the shear could be considered to be delivered near the face of the embed plate. I wouldn't actually do this in practice myself because quantifying it would be a beast and my level of confidence in the strategy would be low.

It seems to me, that even if one could justify delivering shear at the face of the embed plate, there's still a moment to be dealt with in the embed. In this scenario there is still an eccentricity between where the shear would be delivered (face of embed) and where the shear is ultimately resisted by compression in the concrete(presumably some distance along the embedded studs). This is basically what the shear breakout mode of failure addresses, no?



The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[connectegr]

Is the beam composite?

According to AISC, the welded connection angle is flexible and there is no eccentricity in the connection design. Theoretically the concern is simple beam end rotation and avoiding any fixed end moment. As said before these embed connections are typically designed as shear only. Typically simple beam end rotation results in tiny radians of actual rotation. If the beam is composite and therefore has studs along the flange this end rotation is even less.

I agree with hokie that 2" edge on the embed studs is minimal regardless of the direction of load.

http://www.FerrellEngineering.com

Providing fabrication and erection efficient structural design of connections. Consulting services for structural welding and bolting.

 

[jhenry]

Did you build the job yet?

If not, then design the connection for interaction of shear and stress. This moment fixity discussion has value, but you have to design for anticipated loading.
So, design the connection where the top to bolts are not effective due to improper edge distance. You have 6 studs left to handle spaced at 6" o/c to handle 28 kips.
So do your shear and tension analysis on the remaining 6 bolts. I will assume 3/4" diameter Nelson studs.
I think that the connection will only work if there some rotation is allowed, and to consider no interaction reserve for tension.