FS for Cable X-Bracing? 3

[jheidt2543]

I am working on the retrofitting of an existing pre-engineered metal building that is 40 years old. I plan to use high strength, 7 strand cable for the X-bracing and would like to confirm what the "normal" factor of saftey that is used for this type of application.

I have contacted the metal building supplier and some cable suppliers; they have been very helpful, but I would like to verify the factor of safety with a code citation or a reference to a design standard. Anyone have some insight on this?

 

[dik]

You might want to look at a safety factor of 2 or 3 for the cable and a safety factor of 5 for the fasteners... just a place to start, and be comfortable with what you decide.

 

[haynewp]

According to Structural Steel Designer's Handbook, a common design factor (breaking strength/expected load) suggested in the AISI publication "Wire Rope Users Manual" is 5.

Here is a link that may be of some assistance, I can't find bracing specifically though.

http://www.hanford.gov/docs/rl9236/rl9236tc.htm

 

[jheidt2543]

haynewp,

Thanks for the reference. What got me wondering about the "correct" FS is regular building components (beams, columns etc) the we use 1.7 for live loads; for lifting with cables we use 3 or 5. For the cable braces used in a building for wind bracing, should we use 1.7 (I don't think so) or 5 like in lifting (I don't think so)?

I guess I'm leaning toward using 3, but I was hoping there might be some design standard related to bracing in buildings.

 

[JAE]

I would suggest that you separate seismic safety from wind safety.

For wind, I would simply use something like 1.7 to 2 as appropriate. Be sure to adequately study the entire system to know where the weak link in the chain is. Follow the load path thoroughly.

For seismic, you would use the special load combinations used for connection design. This involves the Omega factor in the IBC or other similar factors. In addition, I would ensure that the cable would totally yield prior to the more brittle connection at each end would fail. For standard clevises or socketed connections this may be difficult but the key is to let the cable yield prior to the end connection failing. You want as ductile a connection as possible.

 

[dik]

The kicker with cabling is the connection... hence the higher safety factor suggested... You have to be careful in not reading too much into safety factors for cables used for handling materials... they are subject to a lot of abuse, and this is reflected in an increased factor of safety...

 

[flamby]

haynewp has rightly pointed out the general factor of safety of 5 (minimum) for steel wire ropes. However the comparison of this figure of '5' to '1.7' of general fabrication steel is meaningless as there is no 'yield point' for wire ropes. This FS of 5 is applied to breaking strength of rope, the only measuable property for ropes, whereas FS of 1.7 is generally applied to Yield Stres of steel for structural steel.

The point is, we can draw a elasticity curve for steel and mark out a yield point. In ropes, such exercise will not be of much use because the ropes un-twine and elongate even with slight load. An FS of 5 on breaking capacity of rope does not mean that it is far safer than structural steel having an FS of only 1.7 on its yield point.

 

[jheidt2543]

While doing some additional research on this, I skimmed through the book "Metal Building Systems, Design and Specifications" by Alexander Newman. On page 38 he shows a detail of a hill-side washer with an eye bolt for a connection with cable X-bracing - a standard detail I've seen many times. On page 40 he shows a photo of a fractured hill-side washer as a result of the Northridge Earthquake and the eye-bolt nearly pulled through the column web.

On page 39 he states "...we recommend that a steel reinforcing plate be placed under the washer. The plate should be fitted between the column flanges and welded to them (Fig. 3.17). The plate's thickness can be determined by calculations."

I'm not in a seismic zone, so I've never seen this before. However, it does bring up the question of checking the column or beam web for bending in connections like this, even from wind loading. Anyone actually doing this?

 

[Lutfi]

I would use a SF of 2.0 for cable. This is consistent with TIA/EIA 222 Standard which requires FS of 2.0 for guy cables. I see the bracing cable in metal building to be not any different from that of a guy cable (carries wind load in tension only).

The SF of 5.0 for material handling is used due to the criticality of the function and potential misuse, as stated by dik. OSHA requires these material handling cables (wire ropes) to be replaced when certain number of kinks, broken strands are evident.

Good luck.

 

[Polecat]

I would have to agree with Lufti on this one. And definitely consider using EHS 7 strand guy wire for this, not wire rope. One of the reasons that wire rope has a minimum SF of 5 is because of its flexibility and the continuous wear it is subjected to in moving over pulleys and sheaves. This is not the case with guy wires for utility poles or communication towers.

EIA does specify a SF of 2.0, and I believe the main reason for this is the assurance that the cable never gets close to its yield point (and, yes, guy wires DO have a yield point that can be measured by the tangent offset method). The main point is that a guy wire and a pole, tower or frame in a metal building all will act a a single structure and will take load according to their initial tension and relative stiffness. Hence, the resulting deflections and member stresses can be predicted as long as the wire does not yield. That is the real issue here. Once the cable yields, you don't know what the member stresses are in the structure.

I also totally agree with those who empahsize the proper treatment of the connections. All else being equal, if you have a problem with a guyed structure, it will likely be in the connections.

http://www.spiraleng.com

 

[PXC]

I suspect that for lifting gear, the factor of safety of 5 is also intended to cover fatigue.

 

[VoyageofDiscovery]

With a lower FS, one should consider some sway in the system as cables will stretch to some degree.

Regards

VOD

 

[Polecat]

When dealing with frames, especially metal buildings, you will always have to contend with sidesway. With the codes as demanding as they are today, one should not attempt this without a non-linear finite element analysis program. With an FEA, you will be able to see the resulting deflections and the extent to which they affect the cables and, mor importantly, the column moments.

If, however, you computer just crashed (as mine did recently) you may have to dig out your old Hardy Cross reference book and resort to moment distribution for this ----- good luck!

http://www.spiraleng.com

 

[j19]

For the cable itself, ASCE 19-96 (Structural Applications of Steel Cables for Buildings) indicates the use of a 2.0 safety factor is acceptable for wind or earthquake loads. The end connectors should be capable of developing 110% of the nominal cable strength.

 

[PXC]

VOD mentioned cables stretching. They certainly do "stretch" more than solid steel, but it is not a problem for design. The Modulus of Elasticity varies with the ropes construction. I have some published values that vary from 86 to 138 GPa, with most in the range 103-115GPa, and a representative value of 110 GPa. (For imperial units readers, solid steel is 200 GPa). When analysing/modelling the rope, this reduced value of E is to be used with the "metallic" cross-section area of the rope.

 

[dik]

PXC:
Can you fax me or eMail me the published values? If faxed, I'll scan them and convert to *.pdf's for general distribution. Fax 705 324-5852, eMail dikcoates@alpha.to

 

[Lutfi]

dik,

if you scan and pdf, I would love to get a copy. please e-mail to LM_1959@hotmail.com.

Many thanks,
Lutfi

 

[jheidt2543]

dik,

I too would like a copy of the cable values when they are available @ jheidt2543@aol.com, thanks in advance.

This has turned out to be quite a good discussion. Thanks to all for contributing.

 

[PXC]

Details of wire ropes:

A wire rope subject to a tensile force initially extends as the overall diameter reduces until the adjacent wires contact each other. This is a permanent extension, and can not be accurately predicted. As a guide, assume 0.25 to 1% extension.

After this extension has occurred, the rope behaves approximately elastic up to yield. Values are given for this below:

E apparent modulus of elasticity of rope GPa
A Metallic area of rope
d Rope diameter

Construction A E
6x7(6x1)/F 0.405d^2 110
6x19(9/9/1)/F, 6x25(12/6F+6/1)/F 0.405d^2 100
6x19(9/9/1)/IWRC 0.475d^2 110
8x19(9/9/1)/F 0.355d^2 86
6x36(14/7F+7/7/1)/F 0.410d^2 96.5
6x14 Triangular strand rope 0.465d^2 103
6x26 to 6x29 Triangular strand rope 0.450d^2 103
6x30 to 6x33 Triangular strand rope 0.457d^2 110
Non spin winding ropes 0.500d^2 110
34 LR UHP 0.493d^2 115
18x7 Non-spin UHP 0.500d^2 115
Half locked coil 0.640d^2 138

Reference - Haggie steel wire ropes for mine hoisting 1987

Note that these are typically 20 to 70 mm diameter ropes used for mine hoisting. They are designed, among other things, to minimise rotation and stretch. If you had a general purpose rope with a simple construction and you did not have details of E, I would guess that an E value of 85 GPa or even less might be appropriate.