# Question 5

Urban and Civil

Nov 15, 2013 (4 years and 4 months ago)

76 views

Question 5

Generally the required accuracy for subgrade construction is ±25mm. Machine
operators will use the dumpies as a guide during the placement and compaction of
material. Surveyors should aim for a much higher accuracy when placing the
centreline

dumpies. Certainly ±10mm would be a minimum and ±5mm might be
preferable

(10)

What accuracy is achievable by spirit levelling techniques.

“Survey Standards, Setting Out and Earthworks Measurements” Report of a joint
Institution of Highway Engineers an
d Institution of Civil Engineers Working Party
Published by the Institution of Highway Engineers, 3 Lygon Place, London, SW1.
Tel: 01
-
730
-
5245 2nd Edition September, 1982

The order of error due to setting out was quoted as;

Horizontal

(a)

Marker coordinate
s

±10mm

(b)

Control line pegs (main)

±25mm

Vertical

(a)

Accuracy of TBMs

±2mm

(b)

Comparison of levels direct from same TBM

±2.5mm

(c)

Ditto, but from two different TBMs

±4mm

Assuming the nearby benchmarks were reliable and using Class LD survey techniques
(Refer to
St
http://www.anzlic.org.au/icsm/publications/sp1/sp1.htm) the expected accuracy of the
RLs measured to the top of the dumpies would be ±4mm. Of course it would be
expected that as the dumpy is actual
ly being driven to a certain RL that there would
also be a further mm or more left because the dumpy is close enough to the correct
value not to risk a further hit with the hammer and possibly driving it in too far. Lets
say ±5mm which satisfies the requi
rements stated in the first paragraph.

Of course the ±5mm mentioned above is for an ~66% confidence interval and equates
roughly to ±12mm for a 98% confidence interval.

(40)

What accuracy is achievable by trig heighting techniques.

RL
B

= RL
A

+ Inst Ht +

Ht Diff

Targ Ht

Assume that the RL
A

has no error.

Inst and Targ heights are often quoted as ±5mm

Ht Diff = Slope Dist * cos ZD

An error of ±20” in ZD results in an error of ±9.7mm

Summating the variances;

mm
B
A
B
RL
Ht
T
Diff
Ht
Ht
Inst
RL
RL
0
.
12
5
7
.
9
5
0
2
2
2
2
2
arg
2
2
2
2

This method might
just satisfy the requirements of paragraph one.

Of course the ±12mm mentioned above is for an ~66% confidence interval and
equates roughly to ±30mm for a 98% confidence interval.

Note that vertical collimation would have to be removed or corrected for
observations
would be on one face only.

OR (40)

2 peg test

acceptable collimation error <4mm

level from PM#27931

Include a second benchmark (possibly BM13k) in level run to confirm
that PM#27931 has not moved

Set level in convenient locations to sight
as many dumpies as possible,
restrict sight distance to between 50
-
85m depending on atmospheric
conditions

As most of the dumpies would only have been levelled as intermediates a
second set of readings is required to check the first set. Move level and
re
peat observations to all dumpies

Close back to sterting benchmark or a second benchmark if not yet done

Check misclose between two control stations. If in doubt go to a third.

Adjust misclose and determine dumpy RLs for both forward and reverse
runs

Comp
are RLs and average if necessary

(50)