Activity Feed › Discussion Forums › Strictly Surveying › How do you guys deal with engineers challenging settlement monitoring data (simply because the they don’t like the results)
-
How do you guys deal with engineers challenging settlement monitoring data (simply because the they don’t like the results)
rover83 replied 1 year, 3 months ago 15 Members · 25 Replies
-
There is a clear cut line in the divisions of responsibilities between the two disciplines. If the Engineer is not also an LS, tell him/her to stay in their own lane.
-
Are you using trivets or “turtles” for turning points? A fixed leg tripod? Thermistors high and low? As GeeOddMike said, geodetic leveling is very exacting. An INVAR rod alone isn’t enough. For first order work here, it must also be calibrated by a certified facility (we use Stanford).
-
@bill93 I agree, it’s a bit unreasonable to have an alarm so close to measurement data third standard deviation outliers. Their alarm should be somewhere beyond 4th and 5th standard deviations I think, otherwise the 1% outliers are going to trigger it too often. I’ll make a qualification; where the standard deviation is reasonably achievable. Doesn’t take much for atmospheric refraction to kill vertical measurements, not every day will be overcast and cloudy, where you get your best measurements.
When alarms are too close to normal range data, they go off too often and they lose their impact, it becomes normal, and that’s just as bad as no alarm at all; everyone starts ignoring the alarms.
I think it’s fair to say this job is bumping into the limitations of atmospheric measurement, not to mention tectonic plate movement, Australia moves 1.3mm a week from memory (we’ve got the fastest continent).
As an engineer myself (with metrology experience), the trick is to minimise stress and understand the impacts of movement, everything is dynamic.
When we’d grout the roller circle (slew bearing) on a dragline (thousands of tons, or millions of lbs), we’d fill the dragline bucket with steel to it’s rated capacity, then we’d float the machine on jacks, during the early hours of the morning (4am), we’d do the final rail set, we’d deliberately bend the rails to allow a lead in on the rollers (BE dragline). We measured it with a first order precision level with a micrometre, mounted on a frame welded to the underside of the revolving frame and we did that in ideal conditions, someone told me recently that optical levels with micrometres are obsolete, replaced by digital levels, but I beg to differ. The whole machine was capable of large deflections, the trick was to get it into its working plane to minimise stresses.
Cheers,
Peter.
-
Depending on the configuration of the circuit I do not think you benefit from not using intermediate shots and turning through everything.
Any of you heavy hitters care to weigh in on this?
A turned though point has a good blunder check to it but every turn introduces error to the circuit, and more time to the procedure which also introduces error.
My typical approach is to set hard turning points that are not also monitor points along the route and endeavor set up in similar locations going though the task in a similar order every time and side-shoot monitor points.
As to the politics of it, an old fashioned way of smoothing this over is to start the monitor well before construction activities begin. That gives the field operations some time to refine the work, and generates some noise in the data that the team can point to in order to calibrate expectations.
Having an independent check on the work might be appropriate in some cases.
Both of these ideas involve cost and are difficult or impossible to implement in all cases.
I think that other technical or personal communication based suggestions may or may not translate from one culture to another.
Perspective of the client is often a bigger factor than it really should be.
Also, does anyone have experience using TBC to do pre-analysis of network tasks? If so I would be interested in being pointed in the right direction for learning more about that.
-
Also, does anyone have experience using TBC to do pre-analysis of network tasks? If so I would be interested in being pointed in the right direction for learning more about that.
Unfortunately, this is one of the few things that I dislike about TBC. No preanalysis option.
StarNet is the way to go for preanalysis, in my opinion.
“…people will come to love their oppression, to adore the technologies that undo their capacities to think.” -Neil Postman
Log in to reply.