Below there was a thread about surveying with transit & cloth tape. I'd like to change that to transit & steel tape, tell some stories (no lies) and encourage others to chime in.
When I first started surveying in 1970 a well equipped crew would have a 1 minute transit and a 100' tape. A year later I got a real job and the company was using 300' tapes and with 13 crews we had instruments from 1 minute transits up to Wild T2's. We were taught to do things right. Chaining was always slope chaining and temperature correction. Tension - we just know how much to pull.
When I started my own business I had a 20" transit and 300' tape.
Our goal was to always strive for 1/10,000. And I think we made it most of the time based on visits to old jobs after we moved on up to total stations.
My best story involves a job I did in 1976 - my first summer. I was instrument man and back chain and I had a helper that was BS & front chain. Move ahead to just a couple of years ago. Sent the crew out to the same job. Not only did they find my old traverse stations, but every time they set up and checked a backsight distance they were either flat or +/- 0.01. Not too bad.
I am convinced based on my experiences that 99% of the time we were actually achieving the magic 1/10,000 precision.
What are your experiences.
My experience is error of closure has little real meaning. You can catch gross error when closure is bad. Other than that, its usefulness is very limited.
Larry P
We did the same in the Army when I was a young 82C. That was with a 30 meter steel tape, using knuckles, temperature correction and a tension scale. Anyone out there remember I-See-O Hall?
We easily got 1:20,000 back and forth over the taping range.
The Gentleman I spoke about had his cloth tape up a slope, over a fence and bent around whatever hedges happened to be in the way. It was quite a sight. I seriously doubt that "breaking chain" is in his vernacular, let alone possible. I didn't really lose it until I watched him set a missing corner with the same care. That corner will be held by those who come behind him.
They will have no clue as to how carelessly that corner was set.
In the late sixties and early seventies, with 30" K&E and 100' babbit 'cut' chains we targeted 1:5000 to 1:7500 for a typical 'rural' section boundary, and usually got there. If we didn't, we were "back out there". 1:10000 wasn't impossible, but required a lot more attention and good field conditions. It wasn't uncommon to be on a boundary for a week or more. We use to chain traverse legs from both directions. We did the very best we could.
1974 (75?) we got a HP3800 distance meter and a 10" Nikon theodolite. The first traverse we did with that equipment was almost 30k feet. Raw closure was 1:75000. The "Space Age" had arrived...thank gawd!
I started in 1973. I think we always easily attained 1/10,000, usually much better.
We used to "double chain".
i.e. each chain length had two pencil points on the ground maybe 0.3-0.5 apart and were carried or tabulated separately. Were supposed to match at the end.
Used an "Add Chain"
i.e. the only 0.01 graduations were at the head of the chain.
Often used 300' chains. Even on a slope w/ vertical angle.
That usually required two guys at end. One guy anchoring the guy holding the spring balance and plumb bob. Maybe about 35-40 lbs tension if I remember right.
Instrument was a 20" transit. Angle accumulated six times and averaged.
It's funny that people now make a biiiiig deal about having to correct distances grid/ground.
Want the old simple days.
Sheesh, we had to correct all distances back then with spring balance and especially temperature making sure you went the right way. Usually had to correct for slope too. (no breaking chain when really steep)
Everybody carried a thermometer and a laminated temperature correction chart along with a hand level.
What about correcting distances for curve chords especially on an offset?
The most notable survey I can remember from the 70's took several months to complete between it and a few other jobs.
Using a 1min Dietzen transit and Lufkin 100ft chrome super highway chain we measured the 12± mile of perimeter on a 1,400 acre tract that snaked in and out of Black Cypress Bayou and closed 2.4N and 2.7W and had an angular closure of 1min. One line was across a deep pool we could not wade and we shot stadia for 400+ft
B-)
Anyone ever run a "chain of triangles" or shoot a subtense bar?
Haven't used a subtense bar for real stuff, but I do own 3 and have played with them some. If you take your measurements at appropriate distances they can be quite good. But the longer the distance the weaker they become.
> I am convinced based on my experiences that 99% of the time we were actually achieving the magic 1/10,000 precision.
>
> What are your experiences.
My experiences in retracing surveys made with transit and (steel) tape in Central Texas have been that very few surveyors either applied temperature corrections or used a tension handle to standardize tension on the tape. While they may have had hand levels and chainmen who were adept with plumb bobs, the habit of pulling tension by "feel" and not worrying about temperature corrections tend to leave me unsurprised by errors as large as 1:2,500 in distances taped by those methods.
Because of computational time, it was also a more common practice to break chain on slopes than to slope chain and reduce the measurements from measured vertical angles, which is by far the better method for accurate results.
A surveyor who was applying temperature corrections, pulling tension with a tension handle, and slope chaining as necessary should have been able to get accuracies of 1:10,000 or better on distances, particularly on lines longer than 100 ft. as the propagation of plumbing errors tended to make a diminished contribution to the overall result.
Kent and Larry P have it right. We got closures of 1/10,000 regularly if not 95% of the time, but that didn't account for chain calibration, temperature, sag and sometimes slope errors. Keeping those errors that low was a tremendous amount of work. Not many of us did it.
Did you
find any points on those retracements? If you did, did you hold them?
I'm always amazed by the work of Milard H. Hackney, a surveyor in East Texas from the 1930s-1970s. He listed his bearings to the arc second and distances to the thousandth of a foot. I'm following a 1965 survey of his which we've worked on before.
The call between the Southwest corner and Northwest corner of the tract was:
N 09°19'30" E, 2743.32'
We measured:
N 09°18'47" E, 2743.21' (Geodetic North)
There is a 50 foot rise between the two corners. He monumented them well too. "Auto axles set in concrete".
I've seen this level of accuracy in other surveys of his we've followed as well.
On the other side of the county, another surveyor - an engineer in Kilgore named A. P. Noyes did a subdivision around 1954. Even though he didn't subdivide the entire 102 acre tract, he referenced an original corner in the subject tract.
The call was:
N 88°57' E, 2613.90
We recovered the monuments in good condition and measured:
N 89°36'03" E, 2613.53
This also had a rise of probably 50 feet and crossed a creek (probably 30 feet between high banks).
I have no idea what it took for these guys to produce this kind of work. I've also followed some great work by a guy named John Klotz. His chaining was not this accurate (more like that 1:2500) in rural areas, but he left good monuments and witness trees.
It saddens me to see crappy surveyors today with all of the modern advantages we have producing results that are less accurate than these guys did 50+ years ago with archaic equipment.
As to the 1:10,000 standard. Up until the new rules were adopted, Texas had a standard of 1:10,000 + 0.10' (similar to the ALTA/ACSM standard of 1:50k + 0.07'). This makes a lot of sense considering that 1:10,000 on shot lines is impractical(50 feet would give an error of 0.005'). So I think that a blanket 1:10,000 standard would really depend on an appropriate baseline length.
Did you
> find any points on those retracements? If you did, did you hold them?
I'll bite. How can you discover scale errors in a prior survey without finding marks to which the prior survey connected and that definitely fix certain positions on the ground?
1:5,000 was about the WORST we normally did. We never temperature corrected. Just pulled a little harder, when cold. This would be a traverse around 40 acres, (1/4 mile to the side) and that was often closing with 0.20' per 1/4 mile, in rough country. Most of the time, when we ran around a section, 1 mile to the side, we closed with about a foot of error.
We used abney levels. 200' and 300' tapes. Slope chained.
I think I wore meself out doing it.
🙂
N
Slope Chaining?
What is meant by "slope chaining"? Was the instrument vertical angle used? If so, was it set up at the top and bottom of hills while chaining ahead?
Did you
> > find any points on those retracements? If you did, did you hold them?
>
> I'll bite. How can you discover scale errors in a prior survey without finding marks to which the prior survey connected and that definitely fix certain positions on the ground?
LOL! I was thinking the same thing. 🙂
Slope Chaining?
Measure the slope distance, and vertical angle. You can measure this from the axis of the inst, (point cross hairs at chain man's hands, 300' or so away) then multiply the cosine of the angle, (angle from horizontal) by the distance, to get the actual dist.) Dad used to carry a book of natural tables, and an circular slide rule, to do this in the field, in the 1970's. Or, the Sin of the zenith, does the same ting.
Am I expressing my age?) I just had a brifday! (My new Indian name is "Many Candles""
Nate
Chaining
It can a chore to get 1:10,000 chaining. Assuming you have a 100ft chain that had been calibrated recently against a CBL, which I suspect most had not. As noted on page 16 of the ACSM 1973 "Horizontal Control As Applied To Local Surveying Needs":
Possible Errors using Common Procedures
Source Error Error2
Temp (10d uncertainty) 0.006 0.000036
Tension (5 lb error) 0.009 0.000081
Tape Not Level (0.5 ft) 0.001 0.000001
Plumbing (0.005 ft error) 0.005 0.000025
Marking (0.001 ft error) 0.001 0.000001
Interpolation (0.001 ft error)0.001 0.000001
Total 0.023 0.000145
SqrtRt of Error2 = 0.012 ft
or 1:8,333
This also assumes you had calibrated tension handles and thermometers.
If your chain had not been calibrated you could probably add upwards of 0.01 ft.
Chaining
Well that didn't line up the way I had hoped it would.
Chaining
> It can a chore to get 1:10,000 chaining. Assuming you have a 100ft chain that had been calibrated recently against a CBL, which I suspect most had not. As noted on page 16 of the ACSM 1973 "Horizontal Control As Applied To Local Surveying Needs":
[pre]
Possible Errors using Common Procedures
Source Error Error2
Temp (10d uncertainty) 0.006 0.000036
Tension (5 lb error) 0.009 0.000081
Tape Not Level (0.5 ft) 0.001 0.000001
Plumbing (0.005 ft error) 0.005 0.000025
Marking (0.001 ft error) 0.001 0.000001
Interpolation (0.001 ft error)0.001 0.000001
Total 0.023 0.000145
SqrtRt of Error2 = 0.012 ft
or 1:8,333
[/pre]
> This also assumes you had calibrated tension handles and thermometers.
> If your chain had not been calibrated you could probably add upwards of 0.01 ft.
Well, for starters, if you were using a tension handle and were measuring the temperature of the tape with a thermometer, the standard errors quoted above wouldn't be realistic. Naturally if the tape is laid flat on a heated surface like asphalt in bright Sun in the warm season, correctly measuring the temperature of the tape would be problematic, but if suspended in air, why would we think the temperature of the tape could not be estimated much better than +/-10degF ?
