Cable Rail - Guardrail system design - Analysis and thoughts

[amain002]

All, I have a question, more related to statics I believe, but confusing the heck out of us in the office. Different people have different opinion on this so wanted to have a broader opinion of great scholars in this forum.

I have a condition where cables rail system is used as a guardrails system. Per manufacturer, the tension on each cable is about 225 to 250 lbs (therefore 1000 lbs/ft for cables at 3” OC). The owner/arch wants to have a very slim look, thus the post sizes cannot be huge as we typically do for cable rails posts. Thus we are adding a rigid member at top and bottom of the post to make it like a frame. Typical member sizes are 3x2x1/8 top and bottom and post are T shape made with 3”x 1/2” plate.

My concern is:
(At end posts - where the cables are tied off and only these post sees the tension load from cables)

1) Does the 1000 Plf tension force get applied to only one post (condition 1, see attached) or can we say that the tension force gets distributed evenly - half (Condition 2) to each post?
2) If the second case is true, does that mean only 500 plf tension force is acting at each post or 1000 lbs at each post?
3) If so, the forces are acting in opposite direction which drastically reduces moment at base and also the member sizes.
4) Want to get your opinion of how can we analyses this case.

Another concern is, if I have a case where I have end posts with 3 to 4 intermediate post, which does not see tension loads from cables, however its connected at top and bottom with a rigid member to the end posts, thus can transfer loads between posts. Do you all feel comfortable designing the system as multi frame, so that we can reduce the load at the base connection? Or just ignore the frame analysis and design it as a cantilevered post condition. If so, the member sizes will need to be increased for the post.

Please share your thought on how would you analyze this condition and thank you for taking time to read this lengthy explanation.

 

[oldestguy]

You might explain a little. Is this along a walkway, a driveway, a street, a highway? Might be legal requirements. Cable systems make a good way to "grab a vehicle" rather than deflect it. Not good.

 

[amain002]

@ oldestguy: This railings are for a restaurant so loading is per IBC requirements only.

 

[MrHershey]

1. Tension will be to both posts (like you're showing on second page). First page is incorrect. With your rigid compression elements the tension is switched to an internal force, you won't get deflection like a moment frame.
2. Full 1000 plf tension at each post.
3. Yes, but incumbent on you having those rigid compression members that are drawn in blue. If those members aren't there then you're back high moments and member sizes as the posts act individually and cantilever from base.
4. Green posts have 1000 plf on them from cables (plus whatever railing load in perpendicular direction). For cable tension they span from top rigid member to bottom rigid member. For railing load in perpendicular direction they cantilever from base.

 

[amain002]

@ MrHershey: Thank you for a detailed response.

My analogy (which I know I am wrong but cannot justify myself why).: Why does the two end post see 1000 lbs each. why isn't the load being distributed into half (i.e 500 lbs at each end post).

On the contrary, going back to concept of statics, if I were to cut a section near the post and do a free body diagram (wow idk when was the last time I did so), the tension load from cable = 1000 lbs, so the resisting load equals 1000 lbs too. which compares/agrees with your response. But I can't still justify my question in the earlier paragraph why not 50-50 load on each post. Any insight on this?

 

[MrHershey]

Trust the free body.

Anchor a piece of rope to the ground. Tie a loop on the other end and attach it to a hanging scale and you hold the hanging scale and pull. If you want that scale to say 100#, do you need to pull with 100# of force or 50# of force? Clearly 100# (and if you put a scale between rope and ground, it would also say 100#), you don't get to cut your load in half because the ground is also pulling.

May be easier to think about if you flip to compression. If you take a free-standing column and dump 100# on the top, that means your compression in column is 100#. It doesn't become 200# just because the ground is also pushing back with 100# just like the post on the other end of the tension cable is pulling with 100#.

 

[BridgeSmith]

Trust the statics. Cut it anywhere you want and the answer is the same - there's 250 lb pulling in each direction on each cable.

That also means there's 250lb x half the number of cables in compression on the top and bottom rails. Be sure to check the buckling capacity for those in combination with the required bending loads from the horizontal and vertical railing live loads.

 

[KootK]

- For both two post and four post conditions, I see only the end posts absorbing the cable load.

- I believe that both end posts will see the full 900 plf tension.

- You end posts will predominantly see uniform-ish load between top and bottom rails. This seems different from your displacement sketches.

- I see no significant base reactions on your intermediate posts.

- I would expect your end post bases to see outwards shearing thrusts as they are effectively back-spans to the part of the posts spanning between rails. Slight oversize in the base plate holes should eliminate that as an anchorage consideration.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[BridgeSmith]

The bottom compression strut could be eliminated and it would change very little. There would be a slight increase in bending moment on the end posts as the span increases to the full height of the post from the distance between the struts. You'd also have shear to the base plate and anchor bolts, but an anchorage sufficient for the aforementioned out-of-plane live loads should be more than adequate for a few kips of shear.

 

[amain002]

All thank you for your detailed comments. Appreciate the help.

@ MrHershey: Thank you for the explanation. Was able to get a better understanding of the load path.

@HotRod10: "Be sure to check the buckling capacity for those in combination with the required bending loads from the horizontal and vertical railing live loads." - The post T shaped - 1/2"X3" Pl. It does meet the requirements considering the top and bottom rails adds rigidity.

@kootK: "For both two post and four post conditions, I see only the end posts absorbing the cable load." - Correct, only the end posts where the cable are tied off see the tension load from cables. However, since we are adding rigid top and bottom rails, the intermediate posts does see some force from the ends posts, which helps to relieve some forces at the base for the end posts.

"You end posts will predominantly see uniform-ish load between top and bottom rails. This seems different from your displacement sketches" - Yes correct. My bad. My intention for the sketch as to show which way the members behaved - one post vs two posts taking the load. Now we all agree that both ends see equal and opposite force.

"I would expect your end post bases to see outwards shearing thrusts as they are effectively back-spans to the part of the posts spanning between rails. Slight oversize in the base plate holes should eliminate that as an anchorage consideration." KootK - Can you please explain this one more time.

 

[BridgeSmith]

"The post T shaped - 1/2"X3" Pl. It does meet the requirements considering the top and bottom rails adds rigidity."

My concern was specifically the top rail. It will have axial compression in addition to out-of-plane bending moments, similar to a column that carries lateral loads as well as axial loading.

"...the intermediate posts does see some force from the ends posts, which helps to relieve some forces at the base for the end posts."

Most of the load will go directly to the rails, unless as KootK said, you lock up the base plates on the end posts before tensioning the cables. Then there will be shear pushing the base plates out, making the posts unequal 2 span beams, with a short span between the bottom rail and the post base that as the stiffer span will draw a large moment and induce large forces in the bottom rail and base plate. (draw an exaggerated deflected shape of the end posts under load and you'll see it). That's why he suggested slotting the base plates to make the that short span a free cantilever.

 

[EZBuilding]

This is a design guide for barrier cable system's for vehicular impact but it could be useful for your project.

http://www.post-tensioning.org/Uploads/Technote14.pdf

 

[amain002]

@HotRod10: Yes, the rails will see combined loading, sorry didn't catch that the first time. I have a HSS3x2x3/16" at 4 ft max, I ranc numbers and it seems to work for the combined loading.

"Most of the load will go directly to the rails, unless as KootK said, you lock up the base plates on the end posts before tensioning the cables. Then there will be shear pushing the base plates out, making the posts unequal 2 span beams, with a short span between the bottom rail and the post base that as the stiffer span will draw a large moment and induce large forces in the bottom rail and base plate. (draw an exaggerated deflected shape of the end posts under load and you'll see it). That's why he suggested slotting the base plates to make the that short span a free cantilever"

The connection will be face mounted base plate to the wall with 4 bolts. Post are connected using double angles. I don't think slots will work in this case.

I have revised the deflection sketch. I believe this should be correct now.

@EZBuilding: Thank you for the paper.

 

[BridgeSmith]

As long as the cables are tensioned before the nuts on the anchors are tightened and there is sufficient displacement capacity at the base plate, the position of the base plates are not locked in. They will be able to achieve the displacement. If the deflection is actually small enough to be accommodated by the holes in the base plate being a little bigger than the anchor bolts/rods, then slots are unnecessary. Otherwise, I'm not seeing why you couldn't provide short slots in the base plate instead of holes for the anchor bolts. If the base plate is locked in position, or doesn't have enough displacement capacity, when the cables are tensioned, there will be shear at the base plate as it tries to move out as shown in your sketch, but is restrained, producing additional axial compression in the bottom rail and higher moments in the post.

I'm not clear on the configuration of the posts. Is the 1/2" x 3" bar part of the T shape you mentioned? How is it oriented?

 

[MrHershey]

That's a great document from PTI, but please don't treat it as gospel for what you're doing. Still good for the fundamentals, but if you're getting 250 lb tension from the manufacturer your cables are NOT designed for vehicle impact and you'll find a lot of the recommendations in that PTI document to be overkill.

 

[amain002]

@HotRod10: OOh now I see what you were trying to say. Now, it makes makes of how adding slots in the base plate eases the connection. The configuration of post is attached for ref.

@Mehershey: Yes, using it as a ref material only.

 

[NorthCivil]

I've seen these around here and there but I'm not a fan. Not sure of your jurisdiction, but most jurisdictions, there are many items that govern guardrails additional to the structural performance, one of which is climb-ability.

This design undoubtedly violates the climb-ability clauses. Its a perfect setup for kids to monkey around on.

 

[AaronMcD]

Others did a great job explaining this. I only have a couple things to add:

1) Be sure to check vertical 200 lb point load at the center of the compression struts. 200 lb point load isn't only at railings and isn't only outward. You don't want the bottom strut to buckle when someone steps on it.

2) As for the equal tension at both reactions, I find the most obvious way to visualize is a cord hanging from the ceiling with a 100 lb weight attached. Obviously the tension is 100 lbs, and obviously it's pulling at the ceiling with 100 lbs, and obviously the weight itself is 100 lbs. You wouldn't expect 200 lbs of tension in a cord hanging a 100 lb weight.

 

[amain002]

@AaronMcD : Thank you for another simple but very meaningful explanation.