Bar Joists on CD's by EOR 2

[STR04]

I'm reviewing some shop drawings for a set of bar joists on a 6:12 slope and the mfg'r keeps asking for the axial drag force due to the roof slope. I looked at the 2005 SJI Specification, Section 6.1 and could not find anywhere that this information was required to be specified by the EOR. The joist mfg'r's catalog states this requirement but I'm not sure I agree that it is my responsibility to provide this axial load if the roof loads are only DL+LL, nothing special. Has anyone else come across this issue?

http://www.njb-united.com/ncj/ncj20061.pdf

Section VIII

TIA

 

[haynewp]

I can't say I have seen anyone design a diaphragm for the downslope component plus lateral loads either, but that is what should be done where the joists can't resist the load themselves.

I was thinking about valley beams yesterday with joists framing in on each side. Even with a diaphragm, this should be a case where the joists could resist the downslope component on their own. So I can see where there may be confusion on when to include the downslope component and when not to, and that is why this manufacturer is probably trying to say he always leaves it up to the EOR to decide. If you tell him that the downslope component should be added to all joists and he still does not do it without you specifically telling him for every case, then maybe he is just being lazy. But it is finally the EOR's responsibility to check whatever the joist mfr. has done in the end anyway.




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

 

[miecz]

STR04-

In any case, the EOR should provide the axial force. The path of the forces is the responsibility of the EOR. These joists may have axial force, depending on how the EOR carries the loads.

 

[RARWOOD]

It has been a very long time since I specified bar joists. However if I remember correctly you can get horizontal joist seats for a 6/12 slope.

If I was reviewing shop drawings, that showed the joist with horizontal bearing seats; I would expect the joist mnf to design the joists to resist all gravity load. I would not expect the mnf to transefer part of the load in to the metal deck.

With horizontal bearing seats, under gravity load the only force that would be transferred to the supports should be a vertical reaction. The stresses developed parallel to the slope would be internal to the joist.

Requiring the joists to resist all the forces on the slope probably would not be cost effective. So I am not surprised that a mnf would make that assumption. If mnf is assuming that the deck is resisting the down slope component the mnf should make that clear to the EOR. However from experience the mnf may already know that the down slope component will not be critical to the deck design.

 

[knelli]

I am designing my first sloped steel joist roof structure. After reading the thread can I assume the horizontal thrust at the bearing wall is dumped back into the diaphragm and distributed into the masonry shear walls? I've been thinking about the wood structures I've designed and have never considered a horizontal thrust on the walls, but the trusses all had a horizontal bottom chord to tie the bearing walls together and internalize the horizontal force. What about wood scissor trusses? How are those supports designed?

My major dilema is that we have a 4:12 pitched roof over an auditorium area with a potential large span (if the architect has his way) which will produce some enormous horizontal forces at the bearing walls (like 7.5 k/ft!!) unless they can be assumed to be distributed into the diaphragm and taken out into the shear walls. I've looked through several of my books and cannot find mention of how to deal with this horizontal force. I even have a book with an example of a sloped roof joist with a ridge beam and it does not address the horizontal reaction.

I do have "Designing with Steel Joists, Jiost Girders and Steel Deck" and have read the section on sloping joists. The author notes that if translation is restrained that horizontal thrust will be imparted to the support structure. Back to the previous discussion, does this thrust go into the wall or can it be dispersed into the diaphragm and the lateral system?

 

[haynewp]

Dispersed into the diaphragm as long the diaphragm is adequately connected to the steel joists and designed to take the downslope component over to the LFRS. And the LFRS has to be able to take it. The 7.5k/ft sounds really high, you might double check that.

 

[RARWOOD]

knelll

On what page in "Designin with Steel Joist....", is the discussion about sloped joists? I tried to find something in my addition with out luck. Hopefully I just overlooked it, as I would be interested in reading that section.

 

[knelli]

Hi RARSWC,

I have the 2nd edition of the book from 2002. The discussion is on pages 102 and 103.
I still don't understand why the EOR is responsible for specifying the internal axial load on a sloped joist..... They cover that load calc on pg 103 in the last paragraph.

knelli

PS. I think I found a solution to my problem, I was getting the major kickout force because my joist was partially a scissor joist. I found a way to add a ridge joist girder without interupting the auditorium space.

 

[haynewp]

I was thinking about this today again.

Say you have a single slope building with steel joists spanning from a beam at the low end to a beam at the high end. There is nothing special about the beams (they are not intended to take thrust in the horizontal direction like a tension ring). Also, assume there is NO diaphragm capable of carrying the downslope component.

If there is only gravity load on the roof, the vertical reaction at the low end will have a component parallel to the top chord of the joist. So I think there still can be compression in the top chord of the joist even with only the vertical reaction at the low end acting on the horizontal seat. So there is no stiffness provided in the axial direction of the joist but there is the component of the vertical reaction at the low end which allows compression in the top chord to be generated. The joist is not stable if you look at it this way (no horizontal reaction is available) but there is only a vertically applied load anyway.

Or have I lost my freakin mind today?

 

[youngstructural]

Haynewp:

I think your mechanics for the situation are sound, but that it's a question of degree: There is almost no way you're going to generate enough load to have the joists fail out. It would be an extreeme case indeed, since the diaphragm is always going to have some capacity against this loading, and will allow the joists to help each othe out.

Or is it my turn to question my own sanity?

Regards,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[knelli]

I modeled a sloped member today with vertical loads with pinned supports and I did not get a horizontal reaction at the support. then I changed one support to a roller and it did not move horizontally. I think as long as the joist is sloped and not a scissor type then there is no horizontal reaction at the support? Like you said haynewp, there is only vertical loads applied and the horizontal forces would be internal to the joist....
SO that means that SJI is requiring the EOR to give an internal drag force??

Quote from "Designing with Steel Joists....." pg 103 in the sloped joist example:

"In addition, the manufacturer will need to design this joist for the affects of the load parallel to the joist. This load would be: {LL*cos(theta) + DL)sin(theta). This load will be applied as an additional top chord axial force in the joist by the manufacturer."

Back to Statics 101. Vertical load a sloped member. Give pinned connection one end and roller the other. Sum moments about the pinned connection and voila! No horizontal rxn!?

 

[haynewp]

I am just thinking that before the diaphragm has been connected to the LFRS, that the joists are building forces parallel to their top chord by virtue of their own vertical reactions at each end. Once the diaphragm is welded down and connected to the LFRS, the loads applied ato the roof after this point may be assumed to go into the diaphragm.

 

[RARWOOD]

Knell

Thank you for the information. I have an older edition of the book. 10 years ago when I specified a lot of bar joists I found the book very usefull.

 

[raman22]

My take is a sloped member with a support(wall or beam) at each end has only a vertical reaction at each end. Gravity load acts solely in the vertical direction. If you break this load into two vectors: one perpindicular to the beam and parallel to the beam, it still has a vertical resultant along the beam and at the supports. This will even apply to a beam with a sloped seat bearing connection. The loads are resolved within the connection. If you model this in a computer modeling program, you can verify there are no horizontal reactions taking place at the supports....A sloped beam does not have axial load from gravity loading, a sloped column(or beam-col) does since it does not have a support at one end, although this axial load would be resolved internally within the truss....

 

[jechols]

I have an old, 2 page (some of it hand written) article from Vulcraft illustrating how to calculate the axial component on the joist but it is only due to wind loads not gravity.

 

[JLNJ]

A sloped beam does have axial force from gravity load. It's hard to visualize, but it's there.

 

[michfan]

For interesting reading, go to:

http://www.sbcindustry.com/technotes.php

The Michigan Tech Notes was just updated Sept 13, 2006 so this is very current.

I've been on both sides of the issue - I've been the one sending plans out to component manufacturers and I've been the one designing the component. Both sides think they don't have the responsibility for calculating loads on a truss or joist. Here's the answer...

The WTCA has published Tech Notes for each state and a National version specifying the Design Responsibilities for Commercial Construction Projects. The WTCA represents the structural building components industry. This version specifically mentions trusses, but I bet you'll find a steel joist version out there somewhere too.

Here is what the Michigan version says:

"WTCA has developed this Technical Note to clearly outline a component manufacturer's role and responsibility for commercial construction in the context of the building code and professional engineering law applicable in the State of Michigan. This Technical Note is based on conversations with and questions from various Michigan local building officials and registered design professionals. The analysis is based on the current engineering laws of the state Michigan and the 2003 Michigan Building Code (MBC), which is based on the nationally recognized model building code the 2003 International Building Code (IBC)."

Skip to Page 4 -

"In preparing the Construction Documents, the Building Designer needs to provide the Truss Designer with the information necessary to properly design the Structural Building Components for the Building. According to ANSI/TPI 1-2002 (TPI 1) Chapter 2 (see Appendix C for complete text), which is adopted by reference in the MBC Sections 102.4 and 2303.4, and Chapter 35 "Reference Standards" provisions (see appendix B), the following information should be provided:

........skip to... 2.5.2.4 The location, direction, and magnitude of all dead and live loads applicable to each Structural Element and Truss...

2.5.2.5 All Structural Element and Truss anchorage designs required to resist uplift, gravity, and lateral loads;

2.5.2.6 Allowable vertical and horizontal deflection criteria and any specific criteria...

2.5.2.7 Proper transfer of design loads affecting the Structural Elements and Trusses;

2.5.2.8 Adequate connection between Trusses and between Structural Elements...but not Truss to Truss girder connections...

2.5.2.9 Permanent bracing design for the Building...and permanent bracing for all Structural Elements and Trusses...



I think this is pretty clear. It is the Building Designer, not the component designer who had better specify the loads to be applied. I read that as the EOR, if a structural engineer is involved, or the Architect if no SE is involved.


Right below that section reads:

"The Truss Manufacturer and Truss Designer must rely on the Building Designer to take the information provided by the Truss Design Drawing, the Building Component Safety Information (BCSI 1), and the Building Designer's analysis of the flow of loads through the Building to design a Permanent Building Stability Bracing system that takes the reisted truss member buckling loads (if any) and tie these loadsas off tto the Building's load path system... The Building Designer is the professional who is most intimately familiar with the flow of loads through the entire Building and is the one who can use this knowledge to ensure bracing load transfer and overall Building performance success..."

It is recommended that a copy of this be included with contract documents. I think that is a great idea.

And to any structural engineers out there who don't think they should have to take responsibility for figuring the loads on a truss (me included...), this should help explain why truss designers make $15/hour and engineers make $30+, don't you think?

 

[michfan]

Here is similar info from the SJI regarding joists:

steeljoist.org/graphics/assets/media/Whats_New/NASCC_2006_Final_Paper.pdf

 

[tdsull]

I like the statics 101 example.
I too want to be simple.

You have a mono sloped member with supports at each end.

The only way a horizontal thrust is generated, from vertical loads, is if the support at the high end moves vertically down and the member rotates about the high end.

 

[AggieYank]

Look at the following perfect example, taken from a good old college textbook:

http://tinypic.com/view/?pic=42i37s7

Conclusions:
1) There is no horizontal thrust at either reaction.
2) There is an axial force in the member. This force balances out in the joist, but will range from compression at one end of the joist to tension in the other end of the joist.

Model a typical sloped beam in an analysis program. There will be no horizontal reactions. The member axial force will definitely be non=zero, and vary depending on the load.

 

[knelli]

I think the point of argument here is that the EOR is responsible for specifying the APPLIED loads and that the axial force due to the geometry is an INTERNAL load. The engineer doesn't tell the joist/truss supplier what the internal web and chord forces are, so why is this internal load any dirrerent?