Headwall+Guardrail to Box Culvert Connection 2

[TehMightyEngineer]

I've been tasked with designing a culvert headwall to box culvert connection. The headwall supports a traffic railing (guardrail) to be designed for TL-3 loads (54 kips over a 4 ft length of rail).

Designing the rebar to connect the headwall to the end of the box culvert is simple, I should have no trouble with that. Where I'm hesitant is the rebar in the culvert to take this load. I imagine that in addition to a tensile breakout of the culvert top slab there is going to be significant local bending in the top slab as well. Is there any design guides or reference documents anyone is aware of for this type of detail?

If not I suppose I can throw up a finite element model of this but that seems like overkill for this.

See attached sketch for details.

EIT with BS in Civil/Structural engineering.

 

[Teguci]

I'd design the headwall as a torsional grade beam to get the overturning moment to the sidewalls of the culvert. I'd use the top of the culvert to take the horizontal shear but, for analysis, not put the bending in the top slab.

 

[TehMightyEngineer]

Ahhhh! That's the key I was missing. That makes perfect sense to take the headwall as a torsional member. So simple yet so effective, I knew I was over-thinking this. You probably just saved me about 4 hours of work, +star for you.

EIT with BS in Civil/Structural engineering.

 

[huntley101888]

I had to design a moment slab to support a concrete barrier on top of an MSES wall, and I found a helpful NCHRP report (NCHRP 663, 20MB pdf) located here:

http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_rpt_663.pdf

That report gave an equivalent "static" load that you can use instead of the dynamic load that AASHTO gives you. While this is not the exact situation that you are dealing with, you may be able to use it to form the basis for a more economical (and at the same time safe) design. Good luck.

 

[TehMightyEngineer]

Thanks, I didn't read the article in full but it seemed to confirm that 54 kips was the load I should be designing for (see conclusions 3.4 on page 45).

EIT with BS in Civil/Structural engineering.

 

[huntley101888]

The equivalent load is actually much less than that (10kip per 7.1.1, page 152).

 

[TehMightyEngineer]

Hmmmm, my understanding (after brief reading) is that the 10 kip load is a static load for elastic failure modes and the 54 kip load is a dynamic impact load for in-elastic (rupture) failure modes.

So, from that it seems that the dynamic load should be used (54 kips, see page 155, section 7.2.2) as I'm checking for rupture of the rebar in shear and tension.

Hmmmm, then again, after reading the conclusion (page 163) I see that they're finding that the barriers can indeed withstand either a 9.1 kip static load or a 47.2 kip dynamic load.

So, it seems that I should perform two design checks. One to check yielding of the rebar with a 10 kip load using the AASHTO phi factor of 0.85. The other check should be tensile rupture of the bars with a 54 kip load using the AASHTO phi factor of 1.0. (Both with an AASHTO load factor of 1.0 for the extreme event load combination.)

EIT with BS in Civil/Structural engineering.

 

[ACtrafficengr]

Rather than re-engineering the wheel, check your DOT's website for standard drawing sheet for culvert rail.

FHWA has a page with links to most DOT's standard sheets:

http://fhwapap04.fhwa.dot.gov/nhswp/standardDrawings.jsp

They also have a page with standard specifications.

Maybe the tyranny of Murphy is the penalty for hubris. -

http://xkcd.com/319/

 

[ACtrafficengr]

PS: If you use a standard design, local precasters may have an easier time fabricating it.

Maybe the tyranny of Murphy is the penalty for hubris. -

http://xkcd.com/319/

 

[ztengguy]

What is your span of the culvert? If designed as a torsional member, it also has to take the out of plane bending correct?

 

[TehMightyEngineer]

ACtraffic: I'm in an interesting position where the DOT reviewer asked me to submit calculations for this detail. I actually inquired about such a detail and they sent me one and asked to design for the capacity of that detail. This caused some obvious confusion and, after discussing it with the reviewer, he basically insisted that we re-invent the wheel.

In addition, the height of the headwall is significantly larger than the height used in the standard detail so I suspect that invalidates it.

Thanks for that website, though. That's extremely useful!

ztengguy: 8'-6". I would think that the headwall wouldn't see lateral flexure since we're transferring the shear into the top slab of the culvert. Thus, we just have to anchor the bars taking the shear load into the top slab far enough back to ensure that we don't pull off the end of the culvert top slab. The only way the headwall will be able to move is rotation around a longitudinal axis.

EIT with BS in Civil/Structural engineering.

 

[TehMightyEngineer]

I've attached a sketch roughly showing my intended design:



EIT with BS in Civil/Structural engineering.

 

[Teguci]

You could consider including the top of the culvert as part of the torsional beam. Keeps everythig nice and tight, you have to do two pours anyhow, gives you more perimeter on the torsional beam, just some thoughts.

 

[TehMightyEngineer]

Teguci: Yeah, the idea behind the original way of thinking was that if a large impact (we have a lot of heavy logging trucks around here) it wont destroy the culvert as the headwall will "break away" if it's capacity is exceeded.

I agree, that would be more efficient.

EIT with BS in Civil/Structural engineering.