CIP Box Culvert on Heavy Skew Torsional Effects

[bridgethegap17]

Hello everyone.

I designed a buried single cell cast-in-place reinforced concrete box culvert a while ago and have been second guessing something.

This is a large (24' clear span between walls and about as tall) box culvert set on a 45 degree skew with angled edges. For the record - I was not involved in the original concept decision; in hindsight I would have recommended something else be done, but I digress.

Due to other site constraints which I won't get into for brevity, I determined that the only practical way I could reinforce it without breaking the design concept (which by the way did not include headwalls or cutoff walls) was to run the main bars parallel to traffic instead of perpendicular to the centerline of the culvert. I'm well aware this is not the typical way to design these structures and it made my design span much longer. Nevertheless, this essentially took the concept of having flared bar lines at skewed ends and applied it for the full length. During design I investigated flaring the bar lines to have perpendicular lines toward the middle of the culvert but couldn't quite make it work (setting aside constructability issues). I also considered a case where bending acted perpendicular to the centerline and reduced the bar effectiveness proportional to the angle as an additional design check.

Now I'm concerned about how this behaves in reality near the center of the culvert. As the line between the walls runs at a significant angle to the bar direction, I'm concerned about torsional effects that may result which weren't designed for.

I spoke to some colleagues about this and one wasn't concerned - thinking that the steel taking the tensile force is much stiffer than the concrete, and therefore sort of dictates the direction the bending will behave in (a bit like a deck and girder bridge on a skew). Other colleagues weren't too concerned either because of the amount of steel already in the culvert design and because load effects should redistribute substatially.

Some research I found seemed to suggest that the torsional effects may not be as significant as my paranoia may lead me to believe, but as I don't currently have access to any FEA software to validate this, I figured I'd ask if any of you fine folks had any other views on this.

Any thoughts? Am I just over-thinking this?

Thanks in advance for the responses!

 

[Settingsun]

The reinforcement direction defines the span direction after it cracks. The cracking extent is whatever is necessary to achieve this (compatibility). Not too great a concern provided you didn't go too light on the transverse reinforcement. The torsion (or twisting moment) will reduce due to cracking, again to the extent that the reinforcement can handle. Slabs are pretty robust side from punching. Since you're not subject to punching, serviceability issues are the most likely problem, if any.

 

[SlideRuleEra]

Other colleagues weren't too concerned either because of the amount of steel already in the culvert design and because load effects should redistribute substantially.
Am I just over-thinking this?

IMHO, you are overthinking the issue.
The culvert is made of exactly one material... reinforced concrete, a composite of reinforcing steel and concrete. As long as sufficient reinforcement is present (a highway box culvert meets that requirement), I concur with your "other colleagues".

Just to be clear, I have not designed bridges/culverts... I built them. This included removing part or all of existing ones. I can tell you first hand, properly reinforced concrete is a "tough cookie" in an engineering sense. It will successfully carry loads and resist forces in ways that are hard to imagine and impossible to predict.

BTW, there is nothing unusual about running main reinforcing bars parallel to traffic in a skewed superstructure.

http://www.SlideRuleEra.net

 

[r13]

From reading, I understand you have a 24' square culvert (app.) in 45° skew angle to the longitudinal center line, which is the direction of traffic, and is the direction of your main reinforcement. Please correct me, if it is not the case.

You didn't mention the length of the culvert, nor the end conditions (free opening, or built with walls that capable of providing restrain), so I couldn't comment on the logic of your arrangement of reinforcing steel, but to address, seems likely the most of, your concern - torsion. For underground culvert (I assume), torsional effect is usually negligible as it is surrounded by soil, also remember that the box shape is strong in resisting torsion, so you probably is overthinking in this regard.

Unless there are clear restrains at the ends of the span, I don't quite follow your colleagues reasoning, because the top slab can fail and the longitudinal steel still don't feel a thing. But would the top slab really fail easily? Maybe not, if the shear strength at slab-wall interface is adequate, and there are reinforcement in the short direction that may introduce a "netting effect", which in turn would strengthen the concrete locally by the confining effect of the two way steel. However, this netting effect will decrease from mid-span towards the free edges/ends. (In future design, always stiffen the free edge)

The last resort to relief your anxiety is to evaluate the margin in flexural strength provided by the actual transverse steel (in short direction) and the demand, be it around 1, you should be fine - as the dead load is certain at this stage, and the road is unlikely to be occupied in full with moving load as in the design assumptions. Wish this helps.

 

[bridgethegap17]

Thanks for the responses folks.

@steveh49
I agree that if initial cracking occurs, it will loosen in a way where the main bars would define the direction of bending. I'd just like to prevent that initial cracking if I can. That said, given the angle of the main bars and the fact that the longitudinal bars are just what is required for temperature and shrinkage reinforcement, that may have been inevitable without greatly over-reinforcing the longitudinal bars. Regardless, the main bars are, to be blunt, heavy - large diameter bars at fairly close spacing.

@retired13
The culvert is quite long (~140' along the centerline of the culvert), so end effects would trend to being negligible for most of the central areas. The free edges are not stiffened for project site-specific reasons, but this is part for the reason the main bars are run along the direction of traffic - at the end faces there is enough steel in the top and bottom slabs and the walls themselves to meet the shear and moment demands in and of themselves, generally with some buffer as well. Which is to say, the 1' design frame slice was taken at the edge face and along the direction of the main bars; not perpendicular to the centerline of the culvert.


What I was a bit concerned about was because the main bars run at a substantial angle to the way the slabs 'try' to bend in the middle regions, it could cause internal torsional effects. However from steveh49's post it seems that in the worst case there will be some cracking to allow the main bars to do its work. Not ideal, but we can live with a little hairline cracking if that is all we need to be concerned about.

When originally designing this I looked at AASHTO LRFD Bridge Design Spec Article 5.7.2.3 which discusses when transverse rebar is needed (in this case transverse would be vertical in the slabs because of the internal torsion I'm thinking about) - and it specifically excludes slabs, footings, and culverts from this consideration. So in that I should be fine.

Also, when you consider a precast box culvert with a skew, you run the main bars parallel to the skew edges and you use the design span relative to the edge face, which is essentially what I did, but for a cast-in-place application.

Thanks again, and hopefully this post better explains the full context.