Guardrail design

I'm working on an existing precast parking garage. This garage has precast concrete guard rails that are suspended 12" above the deck and the connections are failing. Long story short, we are going to try to keep vehicles from hitting the existing rails by placing a new guardrail in front of the existing ones. These will span 28' between columns. I've got a structural tube section to handle the load, but the client is wondering about the corrugated-looking bridge guardrails.



Best I can tell from my internet searching these things are called thrie beams and nobody will tell me how to actually design one. Are these all proprietery systems? I need to check if one of these will span the 28' between column (center-to-center, 26'3 clear span) for a simple vehicle impact point load of 10k (the 6k x 1.6 ~ 10k). These will be simply for parking perpendicular to the beam. intermediate supports are not an option.

Any help?

The design is usually already done and the design level is what is important. Meaning....percentage of truck traffic , 85 percentile speed studies, design speed, etc. Bottom line is that a thrie beam guardrail is designed to be hit at an angle at higher speeds. If hit at a 90 degree angle, they fail. I'm not sure the speed that they will actually fail, but their purpose is to redirect traffic. If I'm not mistaken it is usually designed at 62 mph and about a 15 degree striking angle. The ones that I've seen are Federally tested and have federal guidelines so no designer really "designs" the guardrail, usually just the size, length, and angle to the road. But you could possibly calculate a standard vehicle striking 62 mph at a 15 degree angle and compare the force to a 90 degree angle and get a speed that it would withstand. But that is risky in my book.

Now, using them for parking to keep cars from hitting a concrete rail behind them, I would assume fairly low speed if they ever got hit. A question though, why are intermediate supports not an option? They make supports that will attach to a concrete slab as long as the slab is about 8 inches thick.

The thrie bean is not a proprietary system. The only proprietary systems that I've seen are usually end treatments that attach the the guardrail. You might consider a box beam. Probably less expensive and would serve the same purpose. Or possible a W beam guardrail. In your picture there is a thrie beam transition to a W beam. It's the one that have the V in it. The picture of the shorter sections on the top left are the attachments to attach the thrie beam to a barrier wall. The question is how much deflection are you allowed behind the guardrail before you hit the concrete barrier rail?

Most, if not all(but I'm new to the game), guardrail I've seen has maximum 6' 3" spacing on the intermediate posts and the guardrail must be spliced together. The only purpose of the posts is the hold the guardrail at the correct height for impact. I'd be willing to bet that you wouldn't find any guardrail that will span almost 30' without intermediate posts. Even if were possible, I doubt you would find a contractor willing to do it. Too much liability because it is designed with the intermediate posts. Once the guardrail is impacted, the posts are supposed to detach from the thrie beam or w-beam and allow the vehicle to run down the rail without hitting the posts.

If this is something that you want more info on I could possibly give you a few websites from some classes that I've taken but I'd have to dig. Maybe our design engineer could enlighten me more so I could get you some more info. Just drop me an IM or something and I'll see if I can get you anything else.

Hope this has helped some!

That helps... it basically sounds like these aren't going to be an option (and that's fine by us really!). Since these are (almost) empirically designed, and there is no info for a 90 degree load the only other thing I wonder about is whether or not you'd have a section modulus for one of these things. I can have a 6" deflection before we'd impact the existing guardrail, but that'd have to be an elastic deflection and I just don't see how this "wrinkly tin" would deflect elastically at all!

The reason we cannot have intermediate supports is because this is an existing parking garage. We don't have the shop drawings for the precast and those types of supports to the existing deck would dramatically change the load path that is currently there. "Cannot" might not technically be the right word, but that's a whole lot of work that won't fit into the timeframe or budget we have been given.

Oh, and I'm using ASCE 7-05 loading -- 4.4.3 says 6,000 lbs (service live) horiz @ any direction. So there is no interpolation of the load based on speed and/or angle. It's just a pure load to be applied 18" up from the ground/deck.

"T" Beam is designed to flex about 4 feet, but thats with a post about every 8 feet. "W" Beam is intended to flex about 8 feet, to decrease injuries from hitting the guardrail.

We bolt them to the top of old box culverts that dont have a parapet to bolt them onto as is the current standard, I would imagine they could also bolt through the slab of a parking garage where people shouldnt be driving more than 8 mph, but you would have to decrease the spacing between posts. There really not designed for that application, but you can make them work.

Do you have a pic of your existing situation? Can you remove the existing grail and replace with new "T" Beam?

StructuralPoke said:

Oh, and I'm using ASCE 7-05 loading -- 4.4.3 says 6,000 lbs (service live) horiz @ any direction. So there is no interpolation of the load based on speed and/or angle. It's just a pure load to be applied 18" up from the ground/deck.

Click to expand...

for a private garage ASCE 7 might be acceptable. We typically use AASHTO as a reference for vehicle traffic. The latest LRFD code Section 13 is for railings and there are empirical equations to get the loads. I believe that you will find that in addition to the rail section itself, it is equally, if not more important, to get that force into the supporting posts and through the connection to the slab.

You might want to take a look at AASHTO.

You need to check the stability of your ciolumns to take that lateral load/moment.

Taking the existing rail off and replacing something doesn't work architecturally with this building. (saying "Architect" and "Parking Garage" in the same sentence makes me throw up in my mouth a little -- most times, but this Arch is a good one). I've attached some photos -- it's the 2nd and 3rd floors that are in question here. Higher than that, the perimeter L-beams make up the guard rail. On the lower floors, the embed into the column has only two 5/8" headed studs holding the plate on. The 30 year old garage has got some spalling problems around the embed plates and the bolts are rusting off. There has been multiple instances where someone pulls too far forward, hits the guardrail and the connection fails with the guardrail falling onto the perimeter L-beam.

The columns are already (theoretically) designed for the impact load since there are already guardrails connected to them. Code prescribed loading has increased since it was designed, but my numbers show that they are still A-OK.

why not come up with a new connection detail, reattach the existing "guard rails" and install parking bumpers on the slab to prevent cars from pulling in too far and hitting the perimeter rails.

FWIW: I don't believe that that "guard rail" would prevent a car idling forward from going over. That existing connection looks insufficient from the day it was installed.

Adding the tube like I described in the OP doesn't involve removing each guardrail, fiddling with the existing connection (replacing/strengthening/whatever), or dealing with handling the guardrails on the interior of the garage. These things weigh 10 kips each, with 40 psf for parking structures, that's 250 sq ft, not accounting for the fork truck's point loads, point loads for the shoring, etc.

The client doesn't want to use the parking bumpers because they "don't know where to put them". (Notice the quotes!) They said this is because there are many elderly patients (read - patients driving land yachts) that don't know to stop unless they hit something. That's where the initial failures have come from. The difficulty they have with locating the bumpers is the the drive lanes are a bit narrow and vehicles will protrude into them w/ bumpers that are located far enough away from the existing beams to accommodate the land yachts.

And yes, the original connection is a bit suspect...

How many of these things you got? The rails are already hanging from the columns. How about a collar around the column to or some other jacking contraption to temporarily support the end while the connection is changed out/retrofitted. you may not need to remove them at all.

If you're just looking for bump strip, then your initial proposal is probably right on. You need something that makes noise to let the careless driver know he's hit something, kind of like curb-feelers. If this is the case and you can justify not having to meet any guard rail "design codes" by restoring the capacity of the original design you might be all done with a simple bump stopper.

I think you're going to run into problems designing something to withstand "real" automoblie guard rail criteria. Sounds like fun.

I can get a HSS12x8x1/4 to work elastically in a 2-span condition -- it's a little beefier for a simple span. All-in-all there's around 84 of these guardrails. There are interior drains and other things that will have to be rearranged to accomodate all of this, but the client was really leery about pulling the guardrails off and fiddling with them. We're convinced that the existing connection is inadequate and almost not really salvageable and we've said it often enough that the client is ready to "sacrifice" the existing guardrails for the supplemental beams. As for a temporary support -- you can't really tell, but in that 2nd picture, there is a recess that these existing connections are setting in, and they are small enough so that there is no room to get in to work. The recess is 6" across and 6" deep...

The bump stoppers were my first proposal, but they've been tried in other parts of the garage and didn't perform/the client thought they were ugly.

[how the hell do I turn off the bump graphic in favor of just having the word b u m p?]

'bump

StructuralPoke said:

The difficulty they have with locating the bumpers is the the drive lanes are a bit narrow and vehicles will protrude into them w/ bumpers that are located far enough away from the existing beams to accommodate the land yachts.

Click to expand...

Just install the 'bump strips and let the cars hang out.

And yes, the original connection is a bit suspect...

Click to expand...

Yeah, I thought it was lame when I saw it and I'm not a Civil. A little corrosion, and 'Geronimo!'