For real time deformation analysis of a bridge under a moving load, what is a better approach? Sensors on the moving vehicle or sensors on the supporting structure?
Looking at a few ideas, thinking a few GPS sensors in PPK mode on the axles of moving vehicle or prism tracked by a robot. There must be some instrumentation outside the survey world that can monitor displacement as well I suppose.
Of course, the goal is mm's level data. PPK wouldn't get to this level in absolute accuracy but maybe relatively, the data could be reliable, when analyzed in conjunction with total station data. I imagine two curves of different accuracy but similar in trends.
For that level of precision I would probably want to use a prism mounted on the vehicle and a robotic instrument in tracking mode at one end of the structure. This would provide information on the loaded state but you would also need a fairly detailed topo of the unloaded structure.
Thinking about this a little more, you may need to consider using a scanning instrument to capture how the entire structure is moving as the vehicle load moves across it. Remember that some parts of the bridge may be lifted up as the vehicle pushes down on other parts of the span.
You will probably need to know where the vehicle is, but that data alone doesn't tell you much about what the bridge is doing - you are going to have to measure the bridge (in whatever detail needed) with and without load.
I have no idea what I am talking about but would a good place to start be a series of static loads along with the unloaded state?
Licensed Land Surveyor
Finger Lakes Region, Upstate New York
I think a total station is only good dead load type measurements, requiring loaded and unloaded measurements on identical points. The real concern would be dynamic loading and probably require high speed photos and fixed micrometer scales next to flex points on the beams.
Agree with Hub Tack.
GPS wouldn't give accuracies.
Total Station, properly set up and calibrated.
What size bridge?
Is this night and day? Sun versus no sun.
Can you control the traffic? By that drive a load across and monitor several parts of bridge simultaneously.
I imagine there's some engineering requirements that would influence your approach.
I recommend the iMetrum Camera solution. It gives absolute displacement, dynamic measurement, measurement under loading with an accuracy of 0.1mm. A total station will not give you dynamic measurement or measurement under loading and a mere 1mm accuracy.
Bridge possession is also not needed and it can be set up away form the site for safety and no external interference.
check out http://www.imetrum.com/how-it-works/
Surveying Equipment alone won't tell you anything.
You need to supplement that
Here's our approach;
Typically you look for stress, horizontal and vertical movement, tilt, deflection and maybe vibration and acceleration. This will require the use of strain gauges on the framework. Tilt and deflection can be determined either by installing tilt sensors or by monitoring the relation between two position sensors.
You would probably need to supplement that with some sort of accelerometers.
I don't believe that a total station or GNSS alone will suffice. But you can always put together numbers, how realistic they'll be is another matter all together.
Bear in mind that any measurements to the vehicle will include the dynamics of the vehicle movement itself (i.e. weight distribution changes on the tyres as the road gradient and camber change).
A conventional survey system by itself is unlikely to provide a full dynamic measurement and for highest accuracy you will need an electronic system of some kind, but it does depend on exactly what your client needs and the situation on site. I've done tests on one-off loadings using four theodolites sighting targets as the load moved across. A T2 or equivalent is quite capable of detecting a change of 0.2mm. at a reasonable distance, very quickly (literally in a blink of your eye). If the targets are scales rather than cross hairs you can leave the VA set after the initial shot and just read the changes direct from the target scale. The instruments need to change to a new target, on a strict pre-defined sequence, as the load advances so that you don't get critical gaps in the data.
The precise coordinates of the targets aren't critical, as what you are doing in this method is simply reading relative changes. As long as each pointing has an initial VA recorded the absolute changes can be calculated afterwards. A standard survey of the target and instrument positions will suffice.
Who is going to do what with any data obtained?
I could see a graduate student in structural engineering taking this one bridge on as a two year project for his thesis. There are an infinite number of things occurring all at once as the vehicle (primary load) moves across the bridge. A full analysis would reflect a huge number of factors influencing whatever data is obtained. But, in the end, there is a massive amount of data massaged into a relatively small amount of key categories for presentation that can then be used by (who???) to do (what???).
A logical progression of the study would be to analyze what happens as varying loads approach on adjacent traffic lanes from opposite directions, then expand to two loads from each direction at varying spacing, then to the maximum number of potential loads that could occur on the full span simultaneously traveling at a variety of vehicle speeds. It is mind blowing to determine the number of iterations required to expand this data to a bridge with four lanes or so in each direction with curving lanes and superelevation laterally over varying spacing between the primary support structures and with varying component designs and materials.
Look up fellow beerlegger John Hamilton, national seminar guru, and pretty much an expert on such matters as bridge deformation analysis and tracking.
He is the "go to guy" here with some significant national and international credentials. You ever get a chance to go to one of his classes - go.
You might be able to collect some useful information using scales/rulers attached to the bridge and viewed from a distance using an astronomical telescope with a crosshair in the eyepiece and a video camera. Way cruder that accelerometers and strain guages but maybe enough for your purposes. I don't see how anyinformation on the moving vehicle would be useful.
Thank you for replies. We were just chit-chatting about this type of work the other day, I find deformation surveying very interesting although I know very little about it.
> For real time deformation analysis of a bridge under a moving load, what is a better approach? Sensors on the moving vehicle or sensors on the supporting structure?
>
> Looking at a few ideas, thinking a few GPS sensors in PPK mode on the axles of moving vehicle or prism tracked by a robot. There must be some instrumentation outside the survey world that can monitor displacement as well I suppose.
>
> Of course, the goal is mm's level data. PPK wouldn't get to this level in absolute accuracy but maybe relatively, the data could be reliable, when analyzed in conjunction with total station data. I imagine two curves of different accuracy but similar in trends.
Hello Mr James Johnson
Im Andy I work for Imetrum and this is a task that we look at every day
yes as "mtbscott" from Brisbane, says Videogauge !
Our Imetrum Video Gauge is the perfect solution, suited to solve your monitoring problem,
Real time dynamic displacment under real loads especially of bridges amongst a multitude of other structures.
Please have a look at our youtube channel "imetrum" or please do get intouch if you would like more information,
www.imetrum.com
( a steel bridge) and
(a masonry bridge)
Imetrum Technology
Imetrum's Video Gauge system is a precision measurement device used for measurement of displacement, strain, extension and rotation of materials and structures, both under test and in-service.
Imetrum systems use pattern recognition technology, robust industrial grade cameras and the world's most accurate image correlation algorithms. These enable the Video Gauge to take measurements similar to those generated by a strain gauge, LVDT, tilt meter or robotic total station to obtain non-contact measurements with a resolution of around 1/100,000 of the area being measured. The system is easily scalable, and has been used for testing individual structures from 1 km long. It can monitor movements in real time, or record a video for off-line analysis.
> For real time deformation analysis of a bridge under a moving load, what is a better approach? Sensors on the moving vehicle or sensors on the supporting structure?
>
> Looking at a few ideas, thinking a few GPS sensors in PPK mode on the axles of moving vehicle or prism tracked by a robot. There must be some instrumentation outside the survey world that can monitor displacement as well I suppose.
>
> Of course, the goal is mm's level data. PPK wouldn't get to this level in absolute accuracy but maybe relatively, the data could be reliable, when analyzed in conjunction with total station data. I imagine two curves of different accuracy but similar in trends.
Hi James,
I'm not sure if it is called the same thing over in the USA, but it may be possible to collect deflection results by carrying out a "Benkleman Beam" test. This is a method used alot by geotechnical engineers to gauge pavement deflection to an accuracy of 0.1mm. We use it before placing asphalt as any deflection more than 1mm in the supporting pavement will reflect to the surface in the future and result in what engineers refer to as crocodile cracking.
As a check for any permanent deformation, maybe carry out a deformation survey with a total station both before and after the Benkleman Beam test to confirm the results. Setting target points underneath the bridge deck would be the most reliable way of confirming this.
Before making a final solution choice, you should *really* check these folks out:
They make a full range of wireless, data logging, long term/short term, high frequency or low, sensors. Inertial, displacement, the whole nine yards; they have very competent engineers who probably have a ton of customers already monitoring bridges, exactly what you're trying to do.