okay, here is a hypothetical, based on the 15 second / 5 minutes or the 20 minute comment.
this is just a question, not a knock on anyone, just trying to learn as some commentators say you can always learn something.
say you have a 5 acre commercial site, completely wide open, no trees, no obstructions. it has 10 boundary corners (engineered designed) and 5 additional monuments you want to tie in to verify its location. it is a car lot that makes it completely total station obsolete. i will average the time to 10 minutes a corner at 15 is 150 minutes or 2.5 hours plus walk time to each point makes it 3.5 hours, half a day. is that the norm?
Post a list of the coordiantes for the 4 points.
cls, please don't take this the wrong way, you mentioned in your original post that the short side was too long. Was that taped measurement horizontal? It's the only thing I can think of that would give a 'correct' rectangle in plan but a longer side due to measuring the slope distance.
I have used RTK successfully (once and never again!) for setting out holding down bolts for a 180 foot by 60 foot building when the TS failed due to heavy wet snow. The technique used by the contractor was what they called 'floating in' the bolts - pushing each group of four on a plywood template into the newly poured and rapidly curing concrete. Due to the weather and a company Christmas party everyone left and I spent the next 4 hours pushing in the bolts and positioning by RTK, finishing in the dark using a head torch. Remarkably the steel frames dropped into place ok for line and spacing. No bolt cones for tolerance, only 1/4" oversize holes in the base plates.
Was your rod/bubble out of plumb? Ive had that happen before. It's something simple and may not be the cause, but worth a look. Sometimes the easiest solution is the answer.
Sent from my SM-G935V using Tapatalk
Mark Mayer, post: 426179, member: 424 wrote: Per Pythagorus, the cross ties of a 40.00' x 65.00' rectangle should measure 76.32'. If your inverses say otherwise we have narrowed the search for the source of your error.
Absolutely correct. To go a bit further, if your short sides are both 0.2 too long, your diagonals will still be equal, but they will measure 76.43', or 0.11' too long. If checking the diagonals means simply determining that they're equal, then you're not going far enough. They also have to fit Pythagoras.
As [USER=197]@Brad Ott[/USER] suggests, the coordinates make a good starting point. If you calculate 40' as the length of both short sides using the coordinates for the corners of the building, then their staked positions don't correspond to their coordinate positions. If you don't get 40' by inversing the coordinates, then there's a problem with the coordinates.
I would do those calculations by hand to eliminate or expose an incorrect setting in the equipment.
rfc, post: 426160, member: 8882 wrote: Don't the cross ties of a trapezoid match?
Good question. The diagonals of an isosceles trapezoid have the same length. If the 40.2' sides form the legs of an isosceles trapezoid, then the two long sides will have different lengths and the base angles won't be 90 degrees. If the 40.2' sides are the parallel sides, and both of the long sides are 65', then one of the short sides is not 40.2'.
That's one of the things that puzzled me. the OP said "short side" instead of "short sides."
If you keep seeing .2' and you keep checking into the points, then there is a math bust somewhere. Often house/building plans will have some type of bust in the numbers, that can push it's way into your coordinates. They can be hard to spot. I would run the numbers again and pay close attention to everything making sure you can close the layout. You may think you are correct and some little thing has entered into your calculations. I would not use GPS for layout, even the R10's have some error although they are the best I've ever seen. I wouldn't even use an EDM until we got the S6.
House/building layouts are best done by turning angles and taping as much as you can.
Whenever I do a building layout I make a point to set up on my points and start swinging angles. When using GPS/or setting them radially, there is no way for an apples to apples comparison between my layout and the contractor when they set up on my points. I usually try to replicate the contractors methodology with my points. And again, I have had a bubble out of plumb before and that can lead to the issue described. I have spent hours trying to over think these things and the simple answer usually prevails.
Sent from my SM-G935V using Tapatalk
Artie Kay, post: 426157, member: 3428 wrote: If he's set out a trapezium the cross ties will match. Measuring the corner angles of what's been set out might be a useful check.
Inasmuch as Artie is in the UK, I thought I should point out:
"In Euclidean geometry, a convex quadrilateral with at least one pair of parallel sides is referred to as a trapezoid in American and Canadian English but as a trapezium in English outside North America."
Just say'n
Loyal
cls5095, post: 426191, member: 6344 wrote: ....it is a car lot that makes it completely total station obsolete....
I question that statement. But, OK, here is my general procedure for tying control & boundary points using RTK:
- Occupy the point until the coordinate quality goes to as low as it is going to go. Hopefully that is 0.01'. I'll go as long as 3 minutes waiting for the coordinate quality to improve but typically I see it plateau by 90 seconds and often much sooner than that.
- Rotate the rod 180 degrees and repeat step 1. Occupy until the coordinate quality plateaus. The duration of the second session is based on my experience with the first. Often time the second occupation is no more than 30 seconds. Average the results of these 2 ties. The purpose of this second tie is mainly to prove that the rod is plumb, and to partially compensate for the inevitable imperfection.
- Move on to other points, repeating steps 1 and 2.
- I will repeat steps 1 & 2 on control points a second time after a passage of time - at least a half hour, preferably 4 hours. Average the ties. If the "splits" are large (+0.02') I'll repeat ties on the boundary points also.
Even more important than observation time is the PDOP (lower is better, look for under 2, quit tying control & boundary if it's over 3), and the multipath conditions.
I adjust all this in StarNet, but if you don't have that the data collector averaging is nearly as good.
If your cash register does not zero out, somebody goofed....
No matter what.
N
The logic of using RTK followed by check taping for construction layout is flawed. Just use a total station and achieve better results faster. The other alternative I suppose is to call all the corners approximate and let the builders square it up.
The rod bubble suggestion is a good one, by the time all the crap gets attached to the rod and without looking down line at one it can be a easy problem to miss when working robotic or GPSy.
party chef, post: 426211, member: 98 wrote: The logic of using RTK followed by check taping for construction layout is flawed. Just use a total station and achieve better results faster. The other alternative I suppose is to call all the corners approximate and let the builders square it up.
The rod bubble suggestion is a good one, by the time all the crap gets attached to the rod and without looking down line at one it can be a easy problem to miss when working robotic or GPSy.
I recently did a building survey and my bubble was out 0.04' when I would rotate it 180. Luckily it wasn't an issue (I always face rod the same way, so it cancelled itself out), but during my weekly checks I caught it and adjusted the screws. It's an easy thing to miss.
Sent from my SM-G935V using Tapatalk
The other part of the equation here we haven't mentioned yet is your localization. A general rule of thumb. Take the absolute value of your error shown per point, then sum them all. This is the total error in the project. I bet you're sum is close 0.2'.
The localization creates a scale factor that may not be in harmony with your tape or EDM.
I do set out (that's stake out Loyal!) for a handful of builders, that's in spite of me encouraging them to buy their own 'construction' total stations. The most efficient method is to locate the house footprint on the site using RTK so the soil can be cleared, then RTK again to mark the edges/corners of the concrete strip foundation dig, typically 2' wide. After the foundations are cast locate the end points of the longest side on the concrete, approx. with RTK, then accurately for length with the TS and mark with nails. Then, using the total station check any offsets/ties to boundaries after setting up over one of the nails. Meanwhile the builder has put in pairs of horizontal 'profile boards' on heavy posts forming an L shape round the outside of each corner. The top of these boards are at the critical level of the interface between the 'underbuild' built off the foundation and the factory made insulated wall panels. The most important item comes last - locating nails on the top of the profile boards for string lines that the mason builds to, they're looking for and can build to -0, +3mm, and he will check and double check everything with a good quality steel tape. Co-ordinates are easiest for locating a mini prism, with a short spike below for marking the nail positions on the profiles. If you can set up the TS on an arbitrary SW corner everything is then positive.
Artie Kay, post: 426217, member: 3428 wrote: The most important item comes last - locating nails on the top of the profile boards for string lines that the mason builds to, they're looking for and can build to -0, +3mm, .
Would the string line and nail thickness not eat up that 3mm fairly easily?
Why localize?
Shawn Billings, post: 426222, member: 6521 wrote: Why localize?
Good question for original poster (OP)?
Totalsurv, post: 426220, member: 8202 wrote: Would the string line and nail thickness not eat up that 3mm fairly easily?
I didn't appreciate at first how much the masons know their stuff, they want the nail centres on the dimensions and they knot the string lines so they are centred on the nails and about 1/8" above the board and they build up to that. They're also now using factory made blocks which have sharp, hard edges. Not just any old string line either, multi strand nylon which can be pulled tight.
Artie Kay, post: 426237, member: 3428 wrote: I didn't appreciate at first how much the masons know their stuff, they want the nail centres on the dimensions and they knot the string lines so they are centred on the nails and about 1/8" above the board and they build up to that. They're also now using factory made blocks which have sharp, hard edges. Not just any old string line either, multi strand nylon which can be pulled tight.
I normally tell guys my stuff is +/- 0.02'. I have an S6 and in a perfect world I am looking at 0.01' (3mm), but with all the sources of error I figure the best I can do is 0.02'-0.03' single point accuracy when setting points radially...anything tighter and I'm fooling myself. I would never tell anyone that I can stake to 3mm (1/8"), but that's just me. It's only after when I set on my own points and start replicating the contractors methods that I can make the correct adjustments to make everything tighter. As I say in the field, "we aren't building a piano here."
Sent from my SM-G935V using Tapatalk
