moment frame investigation 1

[calculor1]

I'm doing an investigation of a large transit building. The building consists of primarily rigid frames which on the surface look to be very light. Upon reviewing the drawings it appears the frames were designed based on wind load only. I spoke with the original engineer to confirm the loading, upon which he became very elusive when questioned why he did not consider the various load combinations. Has anybody encounter rigid frames designed based soley on lateral loads ...i.e no consideration of gravity + lateral????

 

[JAE]

I've seen where older engineers (from the 50's and 60's) designed entire structures such as grocery stores without any consideration of lateral loads at all other than presuming the structural clay tile set between the perimeter wall columns was adequate.

 

[spats]

They could be Type 2 connections. See attached.

 

[calculor1]

I have checked some of the connections and it appears they were only designed for wind load. I have spoken with a couple of the older guys in the office and they say they typically designed for wind and considered the gravity load to be taken by the shear connection, the connection was simply designed for the moment due to the wind load. Seems strange to me because there obviously has to be some additional moment (from the gravity loading) due to p delta. Am I missing something?

 

[Lexatus]

This isn’t only strange, it is an error.
When you design a MRF building (Moment Resistant Frames) you must consider horizontal loads and vertical loads, because the axial force given by DL will reduce the moment resistance. You can easily see it if you consider a M-N domain, in which you can value the interaction of axial loads with the moment.

So, you can have 3 cases:
- A frame with only axial load => your axial resistance will be full (Nd);
- A frame with only flexure => your flexure capacity will be full (Md);
- A frame with axial load and flexure => you have interaction between M-N and you can’t get full resistance (Nd,reduced and Md,reduced).

How can I value this interaction? It’s very simple; you can use:
- a linear interaction in the easiest and most simplified case by taking Md and Rd on a diagram, then trace a line joining them two.
- a quadratic interaction, that represent better the real behavior of the section.

For the linear interaction you can use:
Ms / Md + Ns/Nd <= 1
In which:
Ms is the moment given by forces;
Md is the flexure resistance;
Ns is the axial load;
Nd is the axial resistance.


I said it in very simplified way (because you have to take in account shear, buckling, and so forth). SAP program operates in this way: with demand/capacity ratios.


So, when you design only with wind loads, you are brushing-off the axial load increment given by DL and LL.
Hope it helps.

 

[asixth]

When you superimpose gravity loads with lateral loads, the knee joint will be subject to a larger design moment than under gravity or lateral load alone.

Agree with Lexatus, reduction in bending capacity will result when a member is subject to a compression force.

 

[civeng80]

Cant post why not?

 

[DaveAtkins]

I think calculor1 is correct. It is perfectly acceptable to design the double clip angles for gravity loads only, and the angles at the top and bottom flanges for the wind loads only. The angle at the top flange, if welded properly to the column and the beam flange, will deform sufficiently under gravity loads to allow the beam connection to be designed as a pin for gravity loads. But these same angles will create a moment connection for wind loads. Blodgett covers it quite well in his textbook.

DaveAtkins

 

[csd72]

Dave,

all that is fine for the connections as long as the members can take the combined loads. By the sounds of the OP the members are light also which as noted previously would definitely be unsafe.

 

[apsix]

Keeping the lateral and gravity loads as separate design load cases was certainly seen as a legitimate design method in the past.
I doubt if that is still the case, but it would depend on your design code.

 

[Lion06]

lexatus-

This is not an error. We design buildings with "wind clip" connections every day. I didn't chime in earlier because you said the FRAMES looked light. In actuality a "wind clip" building will have heavier beams and lighter columns, not lighter members all around. The reason for this is that while the connections only take wind load, you assume a pin-pin design for the beam for gravity. That is obviously the worst case for gravity and will result in a heavier beam than assuming the beam is fixed at the columns. It also lightens up the columns because there is less moment getting thrown into the columns (wind load only, not wind plus gravity). It is covered quite well in Blodgett's book, but he really only touches on top plates with a bottom angle. We typically design them with top and bottom angles. Lou Geschwindner has several good papers on this topic (including one that I posted a long time ago involving the frame stability under these conditions.
While it is true that the moment in the connections won't always be due to the wind load effect, if designed properly it will never have more moment in it than the wind load moment (they're typically never right on the money, so you'll get some nominal amount over the wind moment).
Quick synopsis............ The connections are designed for wind load only. The beam is loaded with gravity and both connections yield at the wind load moment. The wind blows and the leeward connection tries to load further, but it can't because it's already yielded, so the total wind load moment gets thrown into the windward connection which can now absorb twice the wind load moment because it is trying to unload so it accepts "its own" wind moment to unload back to 0, then accepts "the leeward connection's" wind moment to get to the absolute value of the wind moment with the opposite sign.
That's the worst case. You can have the situation where the connections don't yield under gravity loads and some of the leeward connection does see some amount of wind moment before it yields and starts dumping additional moment into the windward connection. That is less common and not worth the time to look at (typically in our office, anyway).
As I said, though, that only makes the columns and connections lighter, but it actually makes the beams heavier.

 

[Lion06]

I suppose you can have a wind + gravity end moment greater than (wl^2)/8, so it could cause the beams to be lighter as well (though not always the case for the buildings I've been involved in). We don't typically use this lateral system for buildings over 4-5 stories. That being said, the reasoning and theory above still holds true.

 

[Lion06]

One last check is to make sure that the beam can take the negative wind moment with the appropriate unbraced length. In the 5-6 wind clip buildings I've done I've never had to bump up a beam size for wind moment.

 

[JLNJ]

It seems to me that even in type 2 connections the columns need to be designed for the wind + gravity condition, not wind only or gravity only. I suppose if it is a large building with a myriad of bays in both directions then the columns won't get sized up much from the gravity only condition. If it's a long narrow building the columns could get quite heavy as part of the wind frame.

Type 2 connections used to be the "norm" long before AISC was called them type 2. I believe the Empire State Building is what we would now call a type 2 connection building.

 

[Lion06]

The empire state building is "type 2 with wind" (that's what I believe they are actually called). The only gravity load the columns sees is axial and whatever moment gets dumped in from eccentricity. It does not see fixed end gravity moment (well, it does, but only up to the wind moment).

 

[haynewp]

I assume you guys mean Blodgett's "Design of Welded Structures"? If so, can you give the page number?

 

[Lion06]

Section 5.5. My edition is copyrighted 1966. I know some in my office have a newer version than I do - so page numbers may be off slightly.

 

[haynewp]

ok thanks