Larry Scott, post: 409266, member: 8766 wrote: Tunnel, tram, bridge? This probably ain't sea level.
We are approx. 600m asl here but this bridge and tunnel are in a river valley for the subway system here. more like light rail than a tram.
MightyMoe, post: 409257, member: 700 wrote: If you do it with GPS? I know it's probably not helpful but I would want R-10's even if I have to rent some.
Well adjusted tribrachs, spin them anyway, get lots of redundant measurements, and let the least squares program tell you what you get, you won't be able to constrain it outside the 300x500 area and get the 2mm; that is a given with the primary control not up to the standard.
Just curious because I don't know, what benefit do the R-10's have for control surveying? better sensors?
MightyMoe, post: 409257, member: 700 wrote: You don't mention if this is 3d, is the 2mm also for elevations? That is a whole nuther issue, particularly with a river running through the site.
The project does require 2mm verticle but this does not need to be real world, just relative to itself. I was just going to find the closest Benchmark to it's design coordinate and hold that. Use it as my datum.
Just a few ideas. Get about 10 L1-L2 modern GPS units.
Set up concrete pillars, in locations, where control is needed. Set 5/8"x11" all thread in the tops.
Cut down trees, or remove obstructions, around these Pillars.
Set these L1-L2 units out, all at the same time, for 24 hrs. Then, SHIFT and swap the receivers, and repeat. After about 10 x like this, you should have pretty tight control.
However, IF you also did this with Total Station, it would give you a very good look at things.
And, did a Least Sq adj of the whole network.
Just thinking out loud.
N
You are looking to produce new control monuments with relative accuracies of 2mm. That is, relative to one another, and not relative to the remote pre-existing control. By all means, you should tie these new monuments to the overall project control, but to whatever degree of precision that is reasonably attainable, not necessarily the 2mm spec.
I will also say that I worked with a company that built permanent control points, these were for GPS. They set on concrete pillars and had a threaded bolt to put the antenna on, it works great. I would say that something like that would allow you to set control that might have a chance to
mrewenmacdonald, post: 409329, member: 11707 wrote: Just curious because I don't know, what benefit do the R-10's have for control surveying? better sensors?
The project does require 2mm verticle but this does not need to be real world, just relative to itself. I was just going to find the closest Benchmark to it's design coordinate and hold that. Use it as my datum.
I was being a bit of a smart-a$$ suggesting you get your hands on some R-10s, however, if you can get some they are the best I've ever used. Everything has tightened up since getting them for us.
That being said the specs for the R-10 in static mode is 3mm+0.1pmm RMS. That doesn't meet the 2mm spec. I don't know about the Lieca units you have, maybe they are actually better. The 2mm spec is very restrictive, I am dubious that it can be met without extraordinary measures. Check out the specs for your GPS and your instrument and be sure they are even capable of meeting the stated specifications.
And vertically, you will also need a level that will accomplish the task, I would also check the specs for what you have....
The fact is that you are building a bridge that has been precisely fabricated. That 0.002m precision may be necessary for all pieces to fit together.
Once complete that bridge may be a few tenths off alignment to the approach roads, which will not be noticeable.
Paul in PA
Nate The Surveyor, post: 409330, member: 291 wrote: Just a few ideas. Get about 10 L1-L2 modern GPS units.
Set up concrete pillars, in locations, where control is needed. Set 5/8"x11" all thread in the tops.
Cut down trees, or remove obstructions, around these Pillars.
Set these L1-L2 units out, all at the same time, for 24 hrs. Then, SHIFT and swap the receivers, and repeat. After about 10 x like this, you should have pretty tight control.
However, IF you also did this with Total Station, it would give you a very good look at things.
And, did a Least Sq adj of the whole network.
Just thinking out loud.
N
Time out please, for a newbie question: If these new control points are created using "modern GPS", does an inverse between any such two points represent the ground point to point slope distance? That is, will they match distances observed using EDM?
rfc, post: 409371, member: 8882 wrote: Time out please, for a newbie question: If these new control points are created using "modern GPS", does an inverse between any such two points represent the ground point to point slope distance? That is, will they match distances observed using EDM?
Excellent question. I should post some screenshots from my javad, to answer your question. But, I'm not close to it.
Basicly, you can set any job up, with any scale factor, rotation, ew tilt, ns tilt, and local coord values.
You can pick one, or a selected group of coords, to derive the csf, (csf = combined scale factor).
This gives you full, and absolute control over your network.
Keep in mind, that this CAN be happening either in the field, or in the office, in the post processing software.
For this project, I'd do it in the office, with network post processed static.
This discussion is very important. If this is new to you, (it was to me, and all of us, at one time), then it's preferable to get an initiate, to go it with you, and not to try it solo, on your first go. Do you know anybody who can do this? Even a data consultant, for this project, can be invaluable.
Nate
GPS results are first calculated in XYZ IGS08 coordinates. Inversing those gets you mark-to-mark distance, i.e. slope distance on the ground.
Those coordinates can be transformed into anything else you want, such as lat-lon and plane grid projections. If you use a UTM, State Plane, or similar grid projection then an inverse gets you grid distances and you need to apply the combined scale factor to get back to ground distances. A Low Distortion Projection is similar to the other grid projections except that it is chosen so the combined scale factor is negligible for many applications and grid is close enough to ground distance. For this project's tolerance, an LDP would need to be customized to a small area around the project.
For instance the OPUS re[prt gives you IGS08 XYZ, IGS08 lat-lon-ellipsoidal height, NAD83(2011) XYZ, NAD:83(2011) lat-lon-ellip ht, UTM meters, SPC in meters, SPC feet, and also ortho height (within the accuracy of the geoid model).
rfc, post: 409371, member: 8882 wrote: Time out please, for a newbie question: If these new control points are created using "modern GPS", does an inverse between any such two points represent the ground point to point slope distance? That is, will they match distances observed using EDM?
Static Gps, using 2 or more receivers, long obsevation, and baseline (phase differential) processing, returns a distance: mark-to-mark distance, notmak section azimuth, and ellipsoidal hgt difference. All relative mark to mark. The coordinates may be approximate and then the baseline vectors are adjusted in a bundle adjustment. You can do anything you want to.
Something like opus or nrc-ppp, is way different. No short, mark to mark measurement. So a much larger mark to mark uncertainty.
And distance is distance. No different than EDM. Baseline phase differential at 500 m works great. Anyone remember Trimvec and GeoLab?
Larry Scott, post: 409415, member: 8766 wrote: Static Gps, using 2 or more receivers, long obsevation, and baseline (phase differential) processing, returns a distance: mark-to-mark distance, notmak section azimuth, and ellipsoidal hgt difference. All relative mark to mark. The coordinates may be approximate and then the baseline vectors are adjusted in a bundle adjustment. You can do anything you want to.
Something like opus or nrc-ppp, is way different. No short, mark to mark measurement. So a much larger mark to mark uncertainty.
And distance is distance. No different than EDM. Baseline phase differential at 500 m works great. Anyone remember Trimvec and GeoLab?
If the OP ends up using Static GPS in his network, and he's also going to tie into (but allow to float as you suggest), points from either the primary control or adjacent secondary control, doesn't this adjustment workflow get complicated? Assuming transformations as Bill93 are "errorless", that is, carried out to the umpteenth decimal, is it better to do adjustments with the raw data, (and translate everyone-who's-come-before-you's work accordingly), then transform to whatever the project requires, or is it better to get everyone's work to the local LDP, then adjust? Does it matter? These questions are inching away from my grasp of what's involved, so don't want to risk hijacking the thread, by "dumbing it down".
rfc, post: 409418, member: 8882 wrote: is it better to do adjustments with the raw data
Using the raw data would be preferred, but raw data for the primary control may be hard to obtain. Next best thing would be to determine or carefully estimate the standard positional errors in the primary control and use those errors in a unified adjustment.
rfc, post: 409418, member: 8882 wrote: If the OP ends up using Static GPS in his network, and he's also going to tie into (but allow to float as you suggest), points from either the primary control or adjacent secondary control, doesn't this adjustment workflow get complicated? Assuming transformations as Bill93 are "errorless", that is, carried out to the umpteenth decimal, is it better to do adjustments with the raw data, (and translate everyone-who's-come-before-you's work accordingly), then transform to whatever the project requires, or is it better to get everyone's work to the local LDP, then adjust? Does it matter? These questions are inching away from my grasp of what's involved, so don't want to risk hijacking the thread, by "dumbing it down".
What would I do ...
I would make a network, including mandatory control. Adjust my network, minimally constrained to be as pure a reflection of my network's integrity. Editing as necessary. Then, (as is required) re-adjust (second network really) holding project control. And assuming project control fits well enough, or identifying any errant point(s), assuming I included as many project pts as possible, go with the fully constrained network.
And that would be whether I used, phase differential (baseline) gps and EDM and triangulation, or just EDM and trig. They're all tools, and not fundamentally different. The ground v grid is a big deal. And if the existing control isn't as good as assumed, you have to evaluate your own sub network first. It's an important project and control is worth a serious effort.
That's sort of what I would do. But, the real situation may bend some of that.
mrewenmacdonald, post: 409036, member: 11707 wrote:
I was wondering how you might tackle such a problem as this? Static in primary control points to view how good they are and rotate the secondary control that I will place to that control network using StarNet or some other software?
Cheers,
Ewen
Please let us know what you end up doing, and how it worked.
Thanks.
Hey Guys, Sorry for the radio silence on this thread.. Just finished my adjustment on Leica Infinity (still working on understanding StarNet) things are looking good, I ran static from 2 different base setups (shooting each of the points for 30-40mins during good GPS windows)
After that I came in with Digital levels, my 2" Leica with some high precision tribrach carriers (good to .5mm) and closed out a traverse, tying into as many relevant control points as necessary.
Now I am feeling confident and I owe a lot of it to you guys! cheers!
