Tension rod sag 2

[DEL2000]

We have some big cantilevered architectural canopy, where we have 1-1/2" diameter tension rods splaying out from a vertical column (similar to a cable stay bridge). Well, the rod that goes to the end of the canopy is about 31' long at a 45 degree angled slope, doing a beam deflection calculation (for beam on a slope) gives a deflection of 12.8" just under its own weight.

This sag seems too excessive . . . is there any problem with using a tension rod with this much sag in it? The roof of the canopy is going to be poured concrete. I am just having trouble envisioning this rod with that much of a deflection. It would seem like it would start straightening out when they start pouring concrete on the roof and that it would be impossible to get the roof set level and at correct elevation.

We will take a long time to look at this, so I am just posing this question to see if someone has come across this situation before, and if my fears are founded in how the sag would affect the construction of the roof.

On a side note: I've been asking a lot of questions lately, so thank you to everybody who has taken the time to respond to me. I've been getting into a lot of "one of a kind" and "new to me" type projects, and also find out that a lot of them are new to my bosses as well. Just running things by you fine folks on the more difficult issues.

 

[UcfSE]

You can't use beam theory to calculate the deflection of a rod like this. Classical beam theory only works when deflections are small compared to the beam depth. Since 12" deflection is much much greater than the 1.5" rod, you will have to use something else, such as cable theory perhaps or large deflection theory. If you remember some assumptions made in classical beam theory, such as sin(theta) = theta for small angles, that will tell you why the equations for classical beam theory are not valid for large deflections (theta is no longer small). There are others assumptions also. Using the proper theory for your deflection calc will also allow you to account for how much the tension will help keep the rod straightened out.

How are your axial deflections working out for you? Will the rod stretch much and cause the roof to sag down? Are these rods exposed to wind? You might have some vibration problems or encounter some vortex shedding from the wind. Rain water on the cables may also exacerbate the problem by increasing drag. I believe AISC recommends limiting tension only members to a KL/r of 300 to help with these issues.

Can you add some sort of adjustable device such as a turnbuckle? I'd bet the architect wouldn't like that though.

Is the roof of a shape such that ponding stability could be a problem, or is it sloped or drained well?

These are some things that come to mind. I don't have any "real world" experience with a structure like this to be able to give you some real advice. It sounds very interesting, let us know how it goes

 

[Wopshistos]

UsfSE is correct In that the beam theory doesn't apply here.
You might however also want to look at things like Icing, creep and reversal due to wind as well as vibration and harmonics which can be a problem for tension structures. You should also seriously consider inspection and maintenance which I think are often overlooked by Architects on low budget stuff.

 

[marisse]

How thick is the roof by the way? Why do you need to use a concrete roof? Since it will be a poured concrete roof, why dont you try to add a cantilever beam (steel or concrete) to reduce the deflection?

You can not treat it as a simple beam, you have to model it using any design software.

 

[flamby]

UcfSE is also correct in stating that tension-only members sould not have kl/r of more than 300.

Ciao.

 

[DEL2000]

Thanks a lot, everyone. I can see now that I was using the wrong theory to get this deflection. What steered me that way was RISA 3D was giving me a really high deflection as well, 12.5 inches and a moment was being applied to the middle of the tension rod (which goes against what the cable theory assumes).

A couple of points:

1) The code says: "For members in which the design is based on tension, the slenderness ratio for l/r preferably should not exceed 300. The above limitation does not apply to rods in tension." We had a 1-1/2" diameter solid rod and were taking this exception to make this work. Is there any problem with using this exception?

2) Using RISA 3D, the tension rod, even as marked as a tension only member, showed this 12" deflection and a bending moment. I'll have to read up with RISA 3D to see how to model those members as actual cables. Any thoughts?

3) As for the canopy itself, we have steel framing supporting concrete deck spanning between the steel members. It is sloped and drained well; since the framing is all steel, creep won't be a problem; and we have provided enough dead load with the concrete roof, to counteract the straight wind uplift on the canopy. What I am not too sure about is if anybody has looked at the more dynamic aspects of the wind loading, such as vibration of those tension rods (I'm just started reviewing this canopy designed by others in the office, so I've only looked at it for a day). That is something I'll have to check into.

By the way, this canopy is on a building, about 650 sf, and the roof is about 13.5' tall. This is a pretty low budget (construction wise, anyways) canopy.

 

[TFL]

what about uplift from wind on the canopy?

 

[flamby]

DEL2000,

I may argue that your rod is not only in tension but with bending as well. So you are not complying the code to the letter. I will interpret the code to mean vertical rods for lifting purposes.

Ciao.

 

[fepc]

Kl/r less than 300 does not apply to rods in tension. See AISC ASD Section B7 page 5-37.

 

[ml13]

A construction idea: have the GC shore the ends of the cantilever until the concrete has set up. Tension your rods, accounting for the deflection you will get from the initial load and any additional live load "camber" you need--and then gradually pull the shoring. I would also get on the phone to the best erector in town and ask them how they would do it.

Good luck.

 

[UcfSE]

kl/r<300 does not apply to rods in tension is correct. kl/r<300 does not strictly apply to anything since it is a recommendation for tension members and not a requirement.

A 1.5" diameter rod 31 feet long has a kl/r near 1000. I would expect this to be highly sensitive to dynamic effects from wind. It's going to require a much more involved analysis than just sizing rods for static loading. Limiting the slenderness ratio helps alleviate problems you can get with such slender members, though the limit is of course not required.

As you see you have to be a little careful how you model things in software. RISA uses beam elements for everything (except plates of course). Checking the T/C box does not change the element to a bar type but rather only reduces or eliminates its ability to carry load in a certain direction. This is a great example demonstrating that you have to be aware of what the software is doing. You can get any answer you want but that doesn't make it right I checked briefly in the RISA help file and didn't find anyway to model bars as bars, much less having members with large deflections.

 

[Grizzman]

I would calculate how much the rod will 'stretch' when the concrete is poured and pretension the rod to offset this. Since your rod is purely a tension member, the reaction should be pretty straight forward to calculate. Remember to add the pretension force to the final tensile load.
Next, I would calculate a wind load then apply a multiplier to it to account for vibration.(similar to multipliers for the number of expected cycles for a crane)Calculate the bending stress then add to the previously calculated tensile stresses for a maximum tesile stress in the rod. Make sure your rod is amply sized.
If possible, you might consider a tube shape to stiffen the element as a 1.1/2"Ø rod is quite a chunk of steel. But then the kl/r may become an issue.
For an easy non-intrusive adjustment mechanism with your rod set at a 45° angle, thread the rod and set it through the column and an equal angle leg.
|
|\/
|/\
/ /
/|/
/ |
(this type configuration)

In retrospect, it seems for a low cost job this is a mighty elaborate design. columns at the ends would be a much more cost effective solution.

 

[JoeTank]

flamby,
Your opinion to DELl2000 about tension rods only being used in a vertical orientation for lifting purposes seems like quite a limiting opionion on the use of tension rods. I think the code relaxation of kL/r limits for tension rods is also based on assumption that the tightness of the rod can be controlled somewhat by the use of a turnbuckle. This helps prevent problems like fluttering in the wind. If you look at the (possibly millions) of sway rods used for elevated water tank construction you will note that the code use of tension rods is not limited to rods in a vertcal orientation only.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[JAE]

RISA and Tension Rods:

We have met this problem of the RISA Beam theory (its not only RISA by the way but most all software) and excessive sagging rods in the output. What you can do is set up a fictional material property that matches the steel properties except that you set the density to a very low number (0.00001) and assign that to the rods. The sagging most likely came from the self-weight (-1) in your basic load tables for the Y direction. You can still use this but with the low density assigned to the rods they won't deflect much at all - basically you are "fooling" the computer to ignore the sagging as it will actually be taken up by the tension in the rod. Do add joint loads at each end of the rods to make up the lost rod weight so it is still included in your model.

Now the dynamic concerns above - you still need to check those.

 

[LPPE]

DEL2000 - I've done a few canopies like you describe, except not that large. I used a concrete deck to counteract the wind uplift, and tension rods with turnbuckles. The rods and turnbuckles are galvanized, and it looks pretty sharp when its all done. Never had a 31 foot rod, though.... Tell the architect that they're f*#*@#g nuts and to cut that canopy span down, and that you're an engineer, not a magician!

 

[DRC1]

First of all the deflection of the rod due to its own weight is negligable.Second it it is a two force tension member, the tension member will not sag. If this is a 2 force member with pin connectios, or a reasonable approximation there of, Finite element software is not needed, and as it appears here you can spend more time trying to model a simple problem than if it were just solved with paper and pencil.
I would consider either havig a turnbukle or theaded end to allow for adjustment of the roof when you are done. If you do use threaded ends, check to see if the ends should be upset.

 

[JoeTank]

DRC1,
Rods do sag... often several inches. Go look at your mother's or grandmother' clothes line. Turnbuckles will indeed lessen the sag, but not eliminate it.

LPPE.
31 foot rods are not really all that long. Elevated water tank tower rods are usually in the the 50 ft to 70 ft range. Good comment about the upset of the turnbuckle ends. That's sometimes overlooked.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com