I'm looking for some good rules that I can give field crews to govern the use of resection control (also known as free stationing).?ÿ One rule would be to always use at least 3 points. And I know that they need to be well spread out relative to the work area. But just how well, stated in a way that can be applied in the field, is what I'm looking for. Any ideas??ÿ ?ÿ?ÿ
Rule 1:?ÿ Make sure your resection points are good.
?ÿ
Rule 1:?ÿ Make sure your resection points are good.
?ÿ
Agreed. Kent is still King StarNet around here, but while he is off crusading I rule as Prince Regent.?ÿ
Remember the mechanics of the triangle.
Remember the mechanics of the triangle.
I will meditate on that.?ÿ?ÿ
Strive for strong geometry between points, close to 90 is best. Have them tie in and store side shots on the points used after the resection is complete, that way you'll know and have a record in the office of how well things worked and can adjust everything later if needed.?ÿ
Use good geometry for better results
Give your backsite a new number along with all control points shot (rotate/transpose/corrections/analyze/records)
Remember that things move for many reasons and near fits are the best results at the end of some days and on others it is practically spot on.
Shoot 3 and more known locations, control points, when possible, calcuate multiple solutions and average.
For the days you want top precision, let every on the crew have their own repetition.
Cross check distances between control newest data and check for possible one or more disturbs.
Most every time I've had a problem with free stationing was from using wrong point number or sitting on a point that belonged to someone else that I had no data on.
Strive for strong geometry between points, close to 90 is best.
This seems intuitive to many of us, but it's not true.?ÿ I think the intuition probably harks back to the days when triangulation was the only practical means of resecting a position, but when using modern instruments that measure both angle and distance, you're actually going to get a better result from points that are nearly in a line than you will with points spread apart.
Consider the two scenarios below, in which I used the standard errors I use for my GeoMax Zoom80 2-second gun and the Star*Net preanalysis function.?ÿ Points 1, 2 and 4 are each about 500 feet from point 3; I set point 3 about a foot offline from 1 and 2 just to eliminate any suspicion that something magical happens when the points are all on the same line.
In the first scenario, points 1, 2 and 4 are fixed in location, and point 3 is the resection station.?ÿ?ÿ?ÿ The angles at point 3 are measured point 1 to point 4 and from point1 to point 2, and distances are measured from point 3 to points 1, 2 and 4.?ÿ The 2-sigma errors at point 3 are 0.020'N, 0.012'E and 0.012' in elevation.
In the second scenario, points 1, 2 and 3 are fixed in location, and point 4 is the resection station.?ÿ Angles are measured at point 4 from point 1 to point 2, and from point 1 to point 3.?ÿ Distances are measured from point 4 to points 1, 2 and 3.?ÿ The 2-sigma errors at point 4 are 0.028'N, 0.014'E and 0.013' in elevation.
?ÿ
I guess that I should get busy with that Preanalysis function and test several different scenarios
I'm not sure that you're sigma error numbers have any significance as they are so similar. My general comment had to do with avoiding situations like your scenario 1, where you might be setting up on point 3 and resection using just points 1 & 2. These points would be in a line and I'd doubt would provide very good results. In any case, scenario 1 is getting better residuals which is what you'd expect because the angles are closer to 90's.?ÿ
Worse case: setting up on a circle that all the points are on.
These points would be in a line and I'd doubt would provide very good results.
Better check your doubts at the door.?ÿ Using only points 1 and 2 from point 3, the 2-sigma errors are 0.025'N, 0.016'E and 0.017' elevation.?ÿ Using only points 1 and 2 from point 4, the errors are 0.044'N, 0.017'E and 0.019' elevation.
I think what we tend to ignore is that the distances from point 4 to points 1 and 2 are about 40% longer than they are from point 3, which magnifies both the angular and range errors.
Worse case: setting up on a circle that all the points are on.
If it were angles only, yes.?ÿ
But as Jim noted, nowadays people have both angles and distances to work with.?ÿ Configurations that are weak with one kind of measurement tend to be stronger on the other, so they complement each other and there are a lot more workable options.
always use at least 3 points
Say you have five control point visible from the setup. ?ÿIf you use all five to establish the new position you may, or may not, have a "better" solution statistically. ?ÿIf you use three of the control point to establish the coordinates of your new point, then based on those coordinates in the field, side shot the other control and compare the coordinates of the side shots to the published values, you have some real time, on site, empirical information to evaluate the actual quality of the resection.
If it's a construction site you are on frequently, using different stations for the resection and the check shots each visit can give you an indication of disturbed control faster than, statistically washing out the "bad" point by using it with all the remaining good ones in the solution.?ÿ
I agree that all of the myths about geometry harken back to the days of angles only resections.?ÿ
I was taught never to do intersections.
I was taught never to do intersections.
I have always told field crews that resections were okay to do as long as I do them.?ÿ Very powerful tool, very prone to being poorly used. That's why I'm looking for a manner of dictating their use bty those under my supervision.
?ÿ
With the exception of TBM verification surveys and maybe a couple other tasks I have not set up over a point for any reason other than to set control in over four years. The control network I have been working with is robust and built around the concept of using resection for all purposes post control, we set threaded "drop ins" which a prism attaches to and can be pointed in any direction, or in some cases sticky targets. I have had the benefit of strong technical leadership from the office and have been really impressed with the results of this approach. Without the benefit of technical acumen in the office in the development of the network some additional caution would be well advised.
Get between your points and try to be balanced, do not shoot two targets in one direction and one in the other for example.?ÿ
Two points is fine, a check into a third can help guys get comfortable.
Three points if the angle is 120 is fine or four if two up and two down. More is not better, it is just more, although as mentioned it can be a good way to QAQC the network. Some software is better for this than others, too many vectors in the bucket can start to get overwhelming.
Stay within your points for your area of work, be reasonable if you violate this rule it can be bent but should not be shattered.
If you need to throw a temp point out to traverse to, that is fine but only once. This if for temp points, if you are launching a traverse based on resection that is a different thing.
I like to work with metric for all things control because I feel like a millimeter is real where a hundredth is too chunky and a thousandth of a foot is ephemeral.
I like to work with azimuths in general but with resection in particular, it just feels more natural, a lot of guys get freaked out and want to check the back sight, Leica has some nice functionality for that shift-user-check point, if not then a person can figure their back AZ by cogo and check that way.?ÿ?ÿ
If elevations are critical or temp control is going to be thrown from the point then do the resection in two faces.
The Sigma values E,N,Z, AZ are not pure "error" like with a BS HD delta, that is important for people to wrap their head around or they will start thinking that resection is some magic and want to shoot ten points for every set up and other such nonsense.
?ÿ
I could not imagine working as a one man crew on a construction site without mastering resections.?ÿ Funny how many of us were taught or mentored "your messing with fire doing resections...."?ÿ This mind set comes from a time of one minute instruments and the steel chains.?ÿ Like any survey function, make sure you check your work.. Jp
..This mind set comes from a time of one minute instruments and the steel chains.?ÿ Like any survey function, make sure you check your work.. Jp
I agree.
The demon we fight on "dynamic" jobsites with daily compromised stations and monuments is?ÿthe degradation of the control accuracy.?ÿ In today's surveying environment an "iffy" resection is still probably vastly better than "tape and transit" control geometry.?ÿ With all the tools we possess nowadays compounding error should be easily identified and rectified to fit the job requirements...but we still have to take the time to get this done. I don't care how many transit mix or asphalt trucks are backed up or how many union operators are waiting for stakes.
Every once in a while a zealous super will remind me how much it's costing him to wait?ÿfor me to get "comfortable" with my control.?ÿ I usually tell them "it's a helluva lot easier and cheaper to move concrete while it's still in the truck".?ÿ?ÿ ??ÿ