How would you analyse this beam? 1

[FraserB]

Hi guys,

Would you analyse the beam shown on the attached sketch as simply supported or with it's ends fixed in place?

For clarification its an I-beam with a simple welded end plate, then bolted into a concrete wall either side.

Also with the connection being hard up against the wall, would you assume there to be negligible pull out force/tension in the bolts, only shear?

Thanks ,

Frase.

 

[csd72]

you have not specified if they are cast in bolts or post installed anchors (chemical or mechanical). If they are cast in bolts then how do you fit the beam in?

The beam is never going to bear right on the concrete at both ends so you will need to have some gap that will be filled with grout.

I would definately analyse them as pinned and design the anchors for shear only with a little extra allowance for the gap to the plates.

 

[dtoli]

Regardless of your first comment csd72 that you're right (I would add some tolerances in length and geometry and so on - i.e. flat and parallel end walls)

I would like to comment for the design that:
Designing the beam as pin ended it is a conservative way for the beam.
On the other hand you under design the anchors. The reality is that you have a partial fix situation and there are moments to consider. It is not certain that at ULS forces will distribute because of the reduced shear capacity (due to tension) or fail due to concrete pull out.

So I would suggest to design anchors as fix ended and (if economy or space is critical - again I don't suggest but I have to say) use anchors with resistance as close to design forces.

Anchors are elements with uncertainties in placement. I may recall some of the problems:
- Placement of an expansion bolt close to a rebar, thus getting wrong indication that the end has been expanded.
- Inappropriately cleaning of drilled hole (specially for resin anchors).
- Local reduced strength of base material (concrete).
- etc.


URL:

http://www.diolkos-eng.gr

 

[dik]

I would design it as simply supported and provide shims to install the beam.

Dik

 

[JStephen]

It would perhaps be preferable to come up with chairs or clips that attach to the walls with the beam in a more conventional configuration. It might also be easier to install that way.

 

[hokie66]

Simple beam, but I never detail it that way. It is much better to connect something like a tee to the wall, then bolt the beam to the tee. Think of the erector when you draw something like this.

 

[rb1957]

ditto ... tolerances ... you'll never fab the beam to exactly fill the span. you can bolt one side directly (if this attachemtn is your preference). the other side you can bolt an attachment piece (a Tee to pick-up the web, as suggested above). or two Tee interface pieces ...

 

[ToadJones]

use Hokie's advice.
The more rotation you allow for in the connection, the closer you'll get to a simple beam....not to mention trying to erect the beam the way you show it will suck.

 

[Teguci]

As drawn, I would analyze the beam as simply supported.

As inferred above, there are better ways to detail this:
1. Single (or double) angle shear connection bolted to steel beam.
2. End plates only welded to web of beam (minimize moment transfer to anchors - may consider coping flanges). Further, extend plate to provide four anchors in single line.
3. Seat angle ledge with top flange keeper.

Not sure what the concrete walls are for, but, if there is a possibility for lateral motion (ie they are retaining walls) the beam and connection should be designed to either allow for this movement or resist it.

 

[paddingtongreen]

It's up to you!

First, I would use concrete inserts so that the bolts can be placed with the plates in place.

A better detail is have knife plates or clip angles on the wall plates and then design as pinned connections, slotted holes can make up any slack.

If you need them to be fixed, you simply follow through designing each of the elements for the force(s) it must resist to justify your assumption of fixity.

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

 

[SAIL3]

As others have indicated...
bm...design as simply supported..
conn...design as fixed end..
use built-up tee-section..more cost effective than using
an actual tee..unless you have a 100 of them..
use slotted holes in web of tee for fit-up..

 

[connectegr]

If your intent is a simple beam, do not weld the beam flanges to the end plate. Weld the beam web only to the plate. Tolerances for the beam fabrication are not the main problem. Tolerances for the concrete walls, are far from "Swiss watch". I would suggest knife angles with short or long slots in the beam web legs. They can be field anchored to the columns or field welded to embed plates. The beam ( with coped bottom flange) can then slide between the angle legs and the slots provide adjustment. The angle legs provide a flexible connection, and ideally a shear only connection to the concrete wall. This simplifies the anchor or embed design.

If you prefer long slots the slots can be in the beam web with standard holes in the angles. This provides the same adjustment, without requiring plate washers. If the eccentricity exceeds the 3" you should consider eccentricity in the bolts. If the outstanding legs are welded, weld the vertical lengths with only a return at the top.

http://www.FerrellEngineering.com

 

[connectegr]

If the flanges are welded to the end plate, even with fillet welds, the connection is not pinned. Bolts/anchors are not required above/below the flanges to create moment capacity. A "moment connection", even with reduced moment capacity, will induce moment forces.

IMHO, the question is whether your intent is to provide a simple beam or a moment connection. You determine the appropriate connection based on the intended design, not the other way around.

http://www.FerrellEngineering.com

 

[connectegr]

And your last question. Local bending/moment forces are not offset by "hard up" framing conditions. Unless the member has a continuous axial compression force.

http://www.FerrellEngineering.com