Bar Joists on CD's by EOR 2

Could be just me, but that sure seems like something they could calculate on their own. They've got the loads and the geometry; the resulting drag force should be part of their joist design. I've never had that come up.

I just spoke with the technical director from SJI and he agreed with the joist designer...go figure. They basically want every load spelled out and diagrammed. I just wish their specification would spell this out to the EOR.

Are they asking for the down slope component of the gravity loads?

I don't design steel joists, but I do design glulam trusses. In a case where the trusses is only required to carry gravity loads, I can't imagen asking the EOR for the down slope component.

STR04, is there a metal deck diaphragm attached to the joists? I have a project going now where I just asked my local joist mfr the same question about whether I should show this component or if he would come up with it and design for it. He stated they don't consider this downward drag if there is a welded diaphragm to the joist top chord, that they consider this component goes into the diaphragm then into the LFRS and bypasses the top chord. I don't know if I completely buy that though.


We make a living by what we get, we make a life by what we give.
Sir Winston Churchill

STR04

The reason the joist mfr doesn't want to consider the downward component is that it will probably increase the cost of each joist.

I wonder how many people include the down slope gravity load component in their diaphragm design?

Do you think the diaphragm is stiff enough to be relatively close to the axial stiffness of the joist? It would have to be in order for the diaphragm to take load. I would guess that for typical flexible diaphragm this is not the case unless you have small spans.

If the joist is only connected to its end walls, or some beam, there theoretically is zero (or very little) joist stiffness in its axial direction - i.e. the joist has nothing pushing back uphill at its ends. Therefore, the diaphragm must take it all.

I agree.


We make a living by what we get, we make a life by what we give.
Sir Winston Churchill

JAE - Thanks. Got it. Don't know what I was thinking.

I guess one twist might be if it is a gable building with joists welded on either side to a beam at the ridge, the joist top chord could be acting as a tension drag bringing force into the top joist connection. The force at the ridge may then cancel, assuming the building is symmetrical and loaded the same on each side of the ridge.

But that would be more of a joist connection concern (if you assumed the diaphragm did not take it all in this case since you have a stiff joist reaction point due to opposite forces at the ridge) rather than buckling of the top chord.


We make a living by what we get, we make a life by what we give.
Sir Winston Churchill

Nevermind, the end reaction at the ridge connection would still be vertical in that case. The joist end reactions would only be not-vertical in the single slope case where the diaphragm is assumed to take all the downward component and you are left with the perpendicular to joist component reaction at each end of the joists.


We make a living by what we get, we make a life by what we give.
Sir Winston Churchill

I like to make some observations regarding the free body diagram in the joist catalogue:

1. The FBD shows the upper joist support/seat to be a roller. Theoretically this may be correct. The way to achieve roller end conditions the joist seat must be allowed to move horizontally. This can be achieved via various means such as bolting, providing slotted holes and lubricating the bearing surface. However, majority of joist seats, at both ends, that I see are welded to supporting beams (or bearing plates) with minimum of 3/16 by 2 inch long fillet welds.

2. If my observations above are correct, then both ends of the joist should be pinned; not pinned-roller due to the vertical and horizontal restraint that the welds will impose.

3. If you model any truss with pinned-roller end conditions you will get differing joist reactions and member stresses from those obtained from pin-pin boundary conditions.

Comments anyone?


Regards,
Lutfi

Lutfi,

You would have a larger stiffness requirement for a pin-roller than for a pin-pin, correct? That seems like a good reason to model a joist pin-roller. Plus you wouldn't have the horizontal thrust to take care of at the support. It makes it easier to use a joist in a larger variety of details. That's my guess.

Just trying to be the devil's advocate here-

If you have a 6 in 12 sloped roof, and somehow the snow sticks to it, where does the thrust loading go?

Pulling numbers out of the air:

(30 ft joist)x(8 ft trib)x(20psf Snow)x(SIN(31.3) [6 rise in 12 run]) = ~2.5kips per joist

I don't even know if that psf is reasonable, I live in Texas and design high-rises, snow usually doesn't control in my designs.. And I might have totally messed up the geometry- but I did stay at a Holiday Inn last night..

But a simple calc could give the manufacturer an idea of how much additional force to design the connection for, would reduce their potential liability, and should make you as the EOR feel more comfortable with the final product and your design. This also may be a subtle way for the manufacturer to tell you they see that something is off in your design.

My punchline is that it doesn't make sense to me to 'neglect' an effect, especially if you are the EOR, the manufacturer is asking for it, and in the real world it exists; plus the manufacturer may indirectly be trying to tell you something.

There is a very good book, "Designing With Steel Joist Joist Girders Steel Deck". The book was published by the Nucor Corporation and authored by James Fisher, Michael West, and Julius Van De Pas.

In chapter 8, Responsibilities, the book states "The Codes of Standard Practice of SDI and SJI govern in the absence of other contract requirements". The key element here is what are the contract requirements.

When I work for a structural engineering firm I designed a number of structures using press plate wood trusses. The specifications required that the trusses were designed by or under the direct supervision of a PE. In addition the specifications required the truss supplier to desig and supply all the truss to truss connections, truss hold downs and supplementaly bracing required for truss stability.

I often got the response from the truss supplier that they were supplying a component and that they were not responsible for the addition requirements. I would explain to them that if they wished to supply the trusses then they must meet the specification. And that if they were unwilling to meet the specifications, there were other suppliers who would.

RARSWC,

I agree with your statement and would say that I am very disappointed with this particular joist mfg'r, but after reviewing the SJI specifications, which has no teeth, I could not say where the responsibilities of the EOR stops and the mfg'r picks up. The technical director from SJI even agreed about the ambiguity in the code. This industry lives in a vacuum and wants every load spelled out or they will assume no special design work is required and pick out their canned joist design per SJI load tables.

STR04

STR04

I tend to agree with the joist manufacturer, that the EOR should provide this force. The axial force depends on factors unrelated to the joist design, i.e. is the ridge beam supported by columns? Is there a tension tie between the walls?

jmiec,

The ridge is supported by columns, therefore no tension tie is required. Keep in mind steel frame with steel bar joists. Not wood rafters & tie collars.

STR04

Since I work for a fabricator of glulam trusses, I partially understand why the joist mnf may have request load information.

In many cases the design load used is based on the engineer's judgement. If the fabricator determines the design load and completes the shop drawings, the fab may be forced into additional work later when the EOR disagrees with the fab assumption.

In the above case the fab will incurr additional cost for redesign and shop drawing revisions. So in many cases it is easier for the fab to ask the EOR for the load.

This is a very interesting thread with a lot of good comments that have been posted.

One aspect I am still interested in, is where does the load go?

Based on my experience, assuming the down slope component is resisted by the deck, probably is the best approach over all. However if the joist mnf makes that assumption, I think the mnf should tell the EOR.

I don't know any one who has designed a steel deck diaphragm for wind load and the down slope component of the gravity load that is not being carried by the joists.