Hey Boss,
Need some education about monitoring. Not my expertise (yet) but it's on my desk because I am the only Swiss watch maker in-house. This is for high-end residential (lakefront houses). They are driving piles for the new house and monitoring the adjoining houses and concrete/masonry walls.
I am uncertain if anyone will actually read these reports, but I want to do it right regardless. The last time around we learned that the geotechs think they want thousandths of an inch and we are hard pressed to report thousandths of a foot. The crew was quite proud of closing to 0.02' and the geotechs were sure the building was falling down. In the end the geotech admitted they just needed a report, the contract had no accuracy specification, and there were no signs of movement (i.e. cracks in drywall, etc.)
This time I'm using star*net to get a pre-pile-driving control net, and we are using reflectorless whenever possible. Went to the baseline and checked the EDMs, found those two out-of-spec prisms and checked all the offsets. Adjusted all the bubbles, etc. We got the control net down to about 0.015' 95% confidence in H and 0.008' 95% confidence in vertical.
Got some control out away from the pile driving & had the crew free-station from those points to shoot the monitor points. That way they are using (hopefully) undisturbed control to shoot reflectorless points on the walls as perpendicular as they can.
How do you prove your measurements? How do you reports your findings? What are most geotechs happy with? What do you look for in a contract as far as realistic accuracy specs? What would you advise a client who wants monitoring but has no idea what they want?
Set control coordinates on reflective sheet targets set on stable locations. Use the same type of targets on your monitoring points if possible. Use the same control points for every setup. Shoot all from one setup if possible. For concrete, I used to smear a tiny dab of contact cement before applying target. Your first readings will be your baseline. This takes out all prism error, backsite error,etc. Let instrument acclimate before shooting. Keep control as close as possible to work area. For various reasons, I had to monitor highway piers and this method worked for me. FYI, it's my opinion that NOTHING stays EXACTLY in the same place.
Frequent site visits, very frequent... Star*Net is a great tool, but the reports need to be super clear: make a good clear spreadsheet of everything.
I am guessing that your Hz work will be just fine, but they probably really want some real tight Vertical. Make sure your level is fine tunned and pegged (recorded) frequently.
I was lucky at one point and had a number of structural engineers that mentored me with what they really needed.
Ask around, there must be some folks around that do dam deformation work and/or large commercial machine installs ... ask them how they handle it.
In future, do not sign contracts that don't give proper, rational specs. "Per specifications of State of Confusion statutes" is better than unspecified until some point after the contract is signed. It's the only way you can prove that you supplied what was contracted for.
I'm just pricing a job where we will be reading levels to 0.01mm, probably achieving an accuracy of 0.03mm, using invar staves and an N3.
As others have replied to you, design the observation network to be as accurate as you possibly can: if you observe to an order of magnitude better than the client needs then your results will always be absolutely right for his needs.
For both plan and height observations always set in the same positions (and instrument height) and observe the same points - that way any systematic errors will generally cancel out when you compare one set of readings with the next.
Even if you are using free-station set-ups always set on a reference mark. That way you have a double check on any ground movements outside the monitored area, and a warning if your instrument/tribrach gets a problem. Don't forget to double check all the screws and bolts on the tripod - surprising how often there is a little play in a leg fitting.
Take multiple readings to each point: with a reflectorless reading also take a check reading just off the side of the target to see if the difference you get is what you would expect - if it isn't there may be a problem with the target surface. (I've occasionally had problems with the steel face of crack monitoring studs).
I've done quite a bit of monitoring work on dams over the last few years. Shoot me an email and I will explain the process we use and send you a copy of a spreadsheet I use to do the calcs.
hey bubble,
good to hear from you again. i am not claiming to be an expert, however i have done a fair share of monitoring. knowing you and your Swiss tendencies, you are the surveyor for that project.
some other things you may want to consider (imho):
differential leveling on the building corners and some marks on existing seawall. consider using Rawl bolts in the mortar joints of the buildings. they are sort of like a 16p duplex nail with a bend at the point. predrill and hammer them in. the bend wedges them in tight. i would seriously consider a 1D network to enhance your 3D network. at least on some of the monitoring visits.
see if dewatering is going to be done on site. hydrostatic changes can do some strange things with structures.
