Box Culvert Monitoring 1

[cazkoop]

I have a situation where a multi-cell concrete culvert is showing signs of over-stress in an exterior wall. Our analysis shows dead load alone causes overstress, thus a negative load rating. I would like to set up a monitoring program to collect displacement data at critical locations. Has anyone seen similar programs?

 

[GeoPaveTraffic]

I've never monitored the inside of a box culveret, single or multi. However, I have monitored many other structures. The question I would ask you is "what is your purpose for wanting the displacement data?" and "what do you plan to do with the data?" Without these answers there is little I can provide. Some general ideas for displacment measurement are listed below.

For long term static measurements, you could probably install tape extesometer points on the inside of the culverts. These allow you to very precisely measure the distance between two points. When the reading are compared over time, you have an idea of the rate of movement. Another option would be to profile the street grade (assume there is a street above the culvert) periodically and compare the readings.

If the goal is to determine the displacement/stress that is imposed on the structure by dynamic loading then the options are more complicated and fewer. You will need a data logger and either strain gauges or LVDTs mounted on the inside of the culvert.

Give me a better discription of the problem and what your goals of a monitoring program are and I might be able to give you a better response. One of the worst things you can do is set up a program and collect data without a clear idea of what you are looking for and how you are going to use the data.

 

[cazkoop]

Thank you very much for your reply. I have heard a little about the use of tape extensometers before, and was thinking this may be a solution to our problem. Our main goal is to obtain long term static results to serve as a warning of any imminent failure. We are concerned over the signs of overstress, but do not want to completely close the bridge without this data due to significant traffic impacts. This would allow us to keep the bridge open unless diplacements start to increase. Do you have any more information on the use of the extensometers? Frequency of readings? Placement of points? Again, thanks for the information.

 

[Focht3]

I agree that extensometers are the way to go - tape and vibrating wire can both work, depending on the structure geometry, etc. I'd check out the following link:

http://www.slopeindicator.com/instruments/contents.html

Let us know what you decide - and why. Good luck!



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[Qshake]

While I understand the urge to gather data, I have to wonder if this is the best place and time to do just that.

Older culverts (50 -60's circa) are usually overdesigned however by today's standard they are also overloaded as well.

To maximize the factor of safety of a structure like this it would be prudent to restrict the travelway to a single lane (if its not already) and possibly shore the most deteriorated elements.

Then I would strongly encourage the owner to consider replacement.

When we see overload and postings getting smaller and smaller its not the time to experiment when life safety is involved.

 

[PEinc]

cazkoop,

You still never described the "signs of over-stress" or the severity. Is the wall cracking horizontally due to bending? Is it bowing in or out? Is the wall crushing vertically? Where is the wall damage - at the top of wall where it meets the roof or lower on the inside face of the wall? How bad is the damage? Can you afford the time and money to make a "research project" out of this or, as Qshake mentions, should a travel lane be closed? If so, how soon is necessary?

Does the culvert have dirt cover (overburden) or is the roof slab the roadway surface? Can the overburden load be reduced with lightweight fill? Can the inside wall face be reinforced with steel plates or fiberglass straps? Does the culvert have enough extra flow capacity to allow building out or thickening of the wall?

I realize that gathering data is important, but so is safety. You know the wall is damaged and needs to be fixed. You obviously have performed a structural analysis. If the owner has the money, I say close the lane (or shore the wall), design a fix, and start fixing ........ or, possibly more appropriate, replace the entire culvert.

 

[cazkoop]

Thanks for all the replies. This culvert is slated for major reconstruction or replacement under a capital project. The project is fairly major, so it will not be complete for several years. This would be something to hold it over. The signs of distress are some vertical cracks near the base of the exterior wall, so the severity is minor, but it is apparent that there is overstressing. We are posting the structure for 22 tons which is the posting limit on an adjacent bridge.

 

[GeoPaveTraffic]

cazkoop,

Without seeing the layout of the culvert and the cracks, there's not much I do about giving recommendations on the measuring points. Generally, install several more points than you think you will need (some will likely be lost for reason or another during the monitoring). Also, once you are there to read the points, whether you read 4 points or 8 or 12 makes little difference.

As for locating the points, remember that an extensometer reads a distance between two points. Look at the geometry of the problem and decide where changes in distance measurements would indicate that the structure is failing. Also give some thought to how much movement will be too much, but the main thing that I would be looking for is a change in the rate of the movement. To this end, I would start off reading the points weekly for three or four weeks, that will give you an idea of the repeatability of the readings. Then you could increase the time between readings, specifics are up to you. Just remember having ten sets of data too many is better than having one set to few.

Try to read the points at the same time of the day, early in the morning is generally the best.

 

[Focht3]

Sage advice, again, from [blue]GeoPaveTraffic[/blue]. Redundancy in instruments and readings is key to your goal.

Carefully document your work, and publish it if possible. Generally speaking, we have more than enough theories - but too little failure data to use in evaluating those theories. Please add - significantly - to our body of knowledge.

And if the instrumentation effort fails, publish anyway. Show what went wrong. Provide recommendations so others can avoid your mistakes. I've read far too many studies of failed efforts where the writer was unwilling to admit their mistake(s), and drew conclusions and recommendations from garbage. Your candor will bring admiration, not ridicule. It takes courage, yet it's an important thing to do.

Good luck - and let us know what happens -



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[cazkoop]

Any thoughts out there on the acceptable displacements?

 

[Focht3]

What will it take to cause "enough" failure to pose a hazard to public health, safety and welfare? The "acceptable displacement must surely be below this threshold. Be sure that your instruments have sufficient resolution so that you don't miss predicting a critical event...



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora.

 

[bmike]

Thanks for sharing this interesting problem. A few things aren't clear from the discussion that may be relevant. These may point to the dominant forces on the culvert, which in turn would suggest how it will deform, and then fail. This will then suggest what you should be measuring from which you can find the best ways to measure it.

As others have pointed out, you first need to identify the purpose of these measurements, to provide an improved understanding of the failure mechanism, or to preserve public safety. You seem to be reasonably assured of the latter.

What kind of depth of cover have you got over the roof? How high is the entire embankment?

Is this a "rigid" (cast in place) or "flexible" (precast) culvert?

How was the culvert installed, in a trench or projecting and covered with an embankment? Do you have access to records of the backfilling sequence and methods?

Is the material beneath the culvert relatively rigid or compresible, in comparison with the fill around the culvert?

Could the culvert be bending on its long axis, i.e. could the embankment be settling and pulling the culvert down with it? The description of the vertical cracks at the base of the wall suggest this, particularly if they are more predominant near the road centreline. In this case, longitudinal profiling would be suggested.

If you think this is a "wall" problem" and you've got the money, one of those Bassett convergence systems (also described at Slope Indicator)should be able to provide you with a good cross-section of the deformed culvert. If you have access to pretty reliable records of the structure, you should be able to back calculate the pressures on it.

Last of all, you have indicated that this culvert is scheduled for removal and replacement. This suggests a rare opportunity to load the culvert to failure prior to removal. In that case, you would provide a great service to the understanding of the behaviour of buried structures if you could test it to failure and publish the results.