How About ZERO Tolerance?
This is from the plans for a parking garage my buddy Carlos is laying out in California:
When RFI'd about the impossible-to-meet tolerance, General Contractor said Well, OK, how about 1/16".
Dave
Accuracy seems good. Honestly I haven't checked against any sort of baseline yet. Been too busy learnin' the new gun and collector.
How About ZERO Tolerance?
At a 1/16" (0.005'), it might as well be zero. I have only staked bolts for one metal building and I staked them as tight as I could (probably around 0.02'). Tell me if this is different than normal. They had me just stake out the center of the plate where the 4 bolt were going to be (i.e. I didn't stake each individual bolt location). After I staked the center, they centered a jig plate with the bolts that would go into the concrete. The jig plate insured the beam plate would fit the bolts precisely, so all you needed to worry about was the relative difference between the beams. I'm not sure how much play they had with that, but it apparently worked.
I really like the S6, we use it a lot for high accuracy work, as well as for other day-to-day work like topo, etc. That said, if starting over I might be inclined to go with Leica just because they seem to be more open with their protocols. I do not like Trimble's "everything is proprietary" approach, which also seems to apply to their GPS line as well.
How tall is the building going to be? If it's only 1 story, I would think up to and including 0.03' or 3/8" would be good enough. However, I think it is possible to stake them to 0.01' relative to each other provided you can do it all from one setup and use a peanut or short prism pole to reduce error. It won't go that fast though.
I think bowtie is right about the template being used to set the actual bolts. What are you staking just the cl of the pattern or cl of the column line? That's the way I've seen it done around here most often. Even when we conduct the as-built it's the same way. We locate the center of the pattern not the individual bolts. Most times the bolts are beat up and bent by the time we come out to check them, so shooting the top would not give an accurate interpretation of where it enters the concrete.
How About ZERO Tolerance?
BTS,
Yup, that's normal. I've staked tons of bolts on casino projects in Las Vegas, and unless one gets misplaced, there's a 1/8" thick template provided for each set of bolts. The holes in the template are drilled 1/16" larger than the bolts, so there's that. We usually scratch or mark centerlines on the template and then lay out offsets on line each side of the footing. Bolt heights are usually set with a laser. As soon as concrete is dumped in the footing we check the center of template and rotation. I call it good if we're within a hundredth. Anything tighter would be too time-comsuming. Adjustments are difficult because of rebar in the footing.
Dave
Ok my 2 cents on the subject, remember it is a system…
I worked for a construction company that specialized in concrete foundation work for about 10 years. As the assistant superintendent I probably spent 85% of my non-paperwork time laying out footings, piers, slab pours, and anchor bolts. Of course the bolts took the most time and care. Primary reason for this has already been hit on by many in this thread…tolerance for the mating item be it a metal column base plate, a pre-stressed bearing plate mount, a light pole break-away mount, etc.
[sarcasm]Pardon the inches in the following but we all know architects and structural types cant work in a base 10 system, that would make all their calculations a lot easier![/sarcasm]
Lets look at the mating that traditionally has the most slop allowed that would be the metal column base plate. I have always been told (and learned the hard way) that tolerance is 1/4 inch but 3/8 inch will work. This tolerance is the total allowed for any connection in the system. If we layout column line “A” 1/8 inch left and column line “B” 1/8 inch right both are in tolerance but we have used 0.67% of the workable tolerance for the system. Now for the curveball: the system is also vertical if we are out 3/8” at the base what does that do to the beam connection 10 feet up the column. Is it plumb; is it rigid, have we weakened the connection by increasing the shear value in the bolts too much?
Having said that lets look at who all shares in this tolerance pool:
- Us :stakeout: (the only one on this list who’s job is precise measurements)
- Concrete guy who digs the footer and sets the bolts per our layout
- Structural fabricator who produces the columns and beams
- Erector that has to make it all fit
- Interior wall guys who have to cover everyone’s mistakes with a plumb wall
- Brick masons exterior walls that have to be plumb but still attach to the skeleton
- On and on and on…
So remember that your anchor bolt layout is the base in a system, and therefore the most critical for the system. Personally I strive to use as little of the tolerance as possible on the layout so the rest of the process that follows me has room to work.
As for “learned the hard way” at this company the rule was you screw it up you fix it. 4 hours of jack hammering anchor bolts out of a dug up pier footer to reset them :excruciating: will leave you with a memory that screams “Get it right the first time @#$%^@#” it only took once.;-)
Back in the 1990's I staked the entire site for an ethanol plant that was getting torn down in Tennessee and moved to Nebraska. They wanted the 0.01' for horizontal and vertical on all bolt spacing. The first major problem was that they had hired an architectural firm to go to the existing plant and get the measurements before they tore it down and shipped the pieces by truck. The next problem was building it during the winter months. Then no matter how hard we tried, we could never achieve that accuracy due to equipment and personal errors. I had to assume that the measurements given to me by the architectural firm were precise, which were full of errors. I had to sign and date all of my work each day and I was just a party chief at the time. Then there was a squabble between union and non-union construction workers and someone sabotaged the bolts in one of the piers after I checked and signed off on them and before the concrete was poured. I had many long nights in the motel revising messed up architectural plans and trying to make things work. Then back to the job site with very little sleep. If I die early, that job can be attributed to taking many years off my life.
Definitely depends on the contractor as to how accurately they are built, but they may still need accurate as-built to provide to steel contractor to assure everything is within tolerance.
I have as-built bolts on high tolerance projects where the bolts were out of plumb well over the tolerance limit! That makes for a time-consuming headache to ascertain the center of the bottom of the bolt. It's easily done, just takes a lot longer.
Most contractors have no idea what it takes to actually as-built work to 0.01' max error. It's definitely doable, but your control and setup errors (resections are best) have to be less than 0.005' and the subsequent locations of bolts need to be less than 0.005'.
Not to mention if the position of any two bolts sampled needs to have error less than 0.01' the numbers above go down even more.
The next hurdle is competition. I have dealt with surveyors that take the project, and just try their best. Their reasoning is "you can't get it that tight, anyway, and who's going to check me". So: they don't know that they don't know how to do it...and their price is half of yours. That has always been my biggest hurdle. The time required to complete the task rises sharply as accuracy requirements increase, doubly so when you get near the operational limits of the equipment.
John Hamilton mentioned above that he uses an S6 High precision. The key to doing these projects as quickly and accurately as possible is to combine the right equipment with the right techniques. The S6 HP is 1mm + 1ppm EDM and 1" accuracy compared to 0.01'+2ppm (there's your whole error budget right there) EDM and 2" or 3" accuracy for the regular version.
Combine the S6 HP with a high grade mini prism and resections (or even better, have a ZNL plummet) and you can do the work a lot faster and better.
"... but your control and setup errors (resections are best) have to be less than 0.005'"
I would maintain that is not possible in a real world scenario.
I doubt I'm going to change your mind without taking you through a project example step-by-step. It is dang hard to do, and on some project sites it IS impossible.
It is doable, though. He said that his site is 100 feet from end to end.
1" in 100' is 0.0004848'. 1" pointing is rather hard to do in outdoor environments, especially on construction sites. If you used robotic aiming, though, and took a lot of wrapped shots, say 10 per interior angle, you could achieve this easily.
Remember, robotics can wrap 10 angles about as fast as you can wrap one BD,FD,FR,BR.
So, if you can measure distance to 1mm angles to half a thousandth, and can achieve centering error for your mini prism to 1mm what's the problem?
http://www.leica-geosystems.us/en/Mini-Reflectors_84906.htm
The only varialbes left are your tripod and tribrach, and if you have good ones the error they introduce is negligible.
Forgot to mention you should double your centering error if you are using minis for backsight as well. Like I said, though, if you do any amount of this work at all invest in either a ZNL plummet, or one of the new ones I linked below and your setup error goes to virtually zero (1mm at 30m)when compared to measuring ability:
http://www.fpm.de/lang_en/index.php?c=1&s=plumb
It also helps to use better points for setup than mag nails. I usually use an ultra-fine Sharpie and hit it with clear enamel. I like to setup the control a day before I observe it. Water tables on buildings make great out-of-the-way locations for control. Another great method, especially if the building will go up vertically, is to surround the site with reflective targets stuck on surrounding buildings. You'll need to oberserve them angles only, though. Like I said, it's doable, didn't say it was easy...
Edit: these are nice, too: http://www.allenprecision.com/supplies/equipment-accessories/prism-poles/accessories/seco-monopod-kit-for-mini-prisms/
I think they're what John Hamilton uses for dam monitoring. I've used them for resectioning off of known control.
Not only do you need to have the right equipment the right personal and procedure you need to control the site and that is what I always found to be the hard part. There is very little possibility of achieving tight accuracy’s if you have no control of line of site and height of site. Meaning if you are working around an excavator compactors dirt piles and so on you will never reach those tolerances. I will guarantee that when a problem arises and the erector company is saying there is a problem the anchor bolts and you go to the site to discuss. There won’t be a piece of equipment running and the site will be clean enough to eat off. There will be no sympathy for the surveyor that was working in chaos. So you have to demand an environment conclusive to rendering the tolerances needed. My 2 cents, Jp
I like to surround the site with targets to use in angles only resections, at the top of surrounding buildings if available/accessible. Rarely do people not let you set them, though. A few near horizon level are good to mitigate vertical angle error, too. If you have four/five near 90 degrees vertical angle you don't have to worry too much about equipment.
Angles only resections are fast and seem to produce better results, too. The trick with them is get the control coordinates for them as tight as possible by observing them from multiple places within the control box, and ONLY measuring to them with reflectorless when the angle of attack to the target is near perpendicular to minimize reflectorless angular bleed. The reflective nature of the targets helps with this, too.
I like Carlson, too, but you definitely don't want Carlson with an S6. Trimble's data collection software is superior in some ways to Carlson, inferior in others.
" It is dang hard to do, and on some project sites it IS impossible."
I think it might be possible indoors in a controlled environment, but on most of the project sites I have been on I don't think so. We can easily have a 40° temp. change (sometimes 60°-70°!) from 6 A.M. to 5 P.M.. The scrapers driving around the site make sighting and measuring difficult, not to mention when the water truck drives by, followed by the vibratory roller. Then we have the concrete trucks and cranes driving past and over the control point. Some of our heavily clay-based soils are amazingly expansive. In the winter I have seen freeze/thaw push a 36" long rebar 0.3' - 0.4' vertically overnight.
Just recently observed MAG nails in asphalt drift around ±0.03' from one day to the next.
I'd think it is heavily dependent on how they plan to set them; Epoxy, wet-set, or poured-in. I find you get better results when they are set, then poured. Wet setting is always fun.. If they are epoxied in, than you can shoot the lay out, set the bolts, plumb them, then shoot the top of the bolt too.
When laying out for them to be epoxied in we usually shoot out center lines then proceed to pull tape to get the bolt layout, this is taking in account if the bolts are in a 2x2 or similar configuration. I've found .0104' is an industry standard for most anchor bolts.. That's what they'd like to see any way. Most QC's don't understand they just go per the specs, so you have to get it as close as you can, that is a general rule of thumb. as previously mentioned you are affecting everyone who works on the said building down the road.
We use either a Trimble S3 robotic or a Sokkia Set 330 to make shots like this.
That's why the references to ultra fine point sharpies and reflective targets. You introduce some error with any control point/target. PK/Mag nails have a few thousandths just in the dimple. I slightly worn mini prism will actually sit in a PK more accurately than a sharp tip. Sort of ball-and-socket.
Also why I started the thread a while back (and was flamed by some) about accurate stakeout with a rebar. It's dang hard to do, they'll walk on you, but sometimes it's all a contractor wants/understands.
About the only thing you could do is set it as close as you can, concrete it, and set a tack in the top - or use an aluminum cap/punch mark.
Yes that is what I use. Very fast to center over a point accurately. Our points are either disks (usually the control) or bolts with punch marks (the points to be monitored).
I have a ZNL but I stopped using it (except for zenith plumbing) with the S6, as the S6 has a rotating optical plummet. You can immediately tell if the OP is out, and even if it is you can still get centered to better than 0.5 mm pretty easily and quickly. I also have a rotating optical plummet that we use for setting tripods when needed, because all of our tribrachs are without OP's (by design). I do have a few tribrachs with OP's, but don't normally use them, instead preferring to be able to check with the rotating plummet.
I cannot think of anything bad to say about the S6. The one I have is 8 years old and going strong. Only had 1 problem-the dual axis compensator went bad, and it was replaced. Because of the mag drive it is light weight and batteries last a long time, and it turns very fast. Plus, you can move it with your hand and you are not fighting against gears/motor.
Of course I wish the radio range was longer, and that the auto-lock had a longer range, but not usually an issue.
I also have a Zeiss S10 that is robotic, but I don't have any data collector that will drive it anymore. It is also 1"/1 mm+ 1 ppm. The one thing about it as compared to the S6 is that it will shoot much further, and through vegetation better. Back when I was using it I could just point it in the general direction of the prism and it would lock on and shoot. Rarely use it, but it was a fine instrument in its day. Very heavy, though as compared to the S6.
>
> by the way, which island are you on?
Kauai