A good article for who is concerned with the inclinometer probe's data quality. Much of this info is published by SINCO (Slope Indicator Company), but in this article it is presented in a slightly different way.
BIS, 2006. Conformity Assessment Vocabulary and General Principles (IS/ISO/IEC 17000:2004) - BUREAU OF INDIAN STANDARDS, MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG, NEW DELHI 110002.
A good article for who is concerned with the inclinometer probe's data quality. Much of this info is published by SINCO (Slope Indicator Company), but in this article it is presented in a slightly different way.
A good article for who is concerned with the inclinometer probe's data quality. Much of this info is published by SINCO (Slope Indicator Company), but in this article it is presented in a slightly different way.
A good article for who is concerned with the inclinometer probe's data quality. Much of this info is published by SINCO (Slope Indicator Company), but in this article it is presented in a slightly different way.
FF Saleave
Advances in Inclinometer Data Analysis
P, Brik Mikkelsen
‘Consulting Engineer, GeoMeron Bellvue, WA, USA
‘Simpostun on Field Measurements in Geomechanics, EMG 2003, Oslo, Norway, Septamber,
ABSTRACT: Tho probe inclinometer is deoeptively easy to use. Its use has become increas:
ingly commonplace, but what influences accuracy is less known, This paper dessries the chan,
scterstis of for systematic erors that can ooeur and provides some puldane for fecoeniing
and correcting the erors, Tho strongest eset ofthe traversing probe inclinometer is ie ability
to detect “spike” inthe data comesponding to shear displacement in ho ground, Over tig
or distributed displacement of the inclinometer casing is more dificult fo evauate beceuos of
‘he higher potential for systematic errors dominating theres, Reeding mars and occasional
inistakes aso produce apparet isplacements. Therefor, the data analyst must evaluae the ee
or potential and sereen the results for ero. Where potable, the errors shouldbe elimineted
prot to presenting the resus to others wio may take further engineering aston based on what
‘ay be errors instead of actual displacements.
1 INTRODUCTION
‘This pape deals with probe inlinometer accuracy issues that have not been fully addressed by
the manufacturers of inclnometas or elsewhere. Iisa subject largely ignored by our profes,
‘and was not discussed in J. Dunniclif's book (1988). Pantculely challenging te data
where even the smallest indication of displacement is a matter of concer, and from desper int
clinometers where systematic errors can accumulate significantly. Systematic errors Ir il
‘nometer surveys have been recognized sine the mid-1970' (Cornforh 1973; Wilson and Mik:
kelsen 1978; Mikkelsen and Wilson 1983; Green and Mikkelsen 1988; Mikkeln 1996), bet
very few practitioners recognize the need for error diagnostics and correction procedures The
intent ofthis paper isto explain to inclinometer users how to recognize four main ypes of sy,
‘ematio error and how to dea wit them,
‘The main innovetions in the last 20,
Feadings expedite the surveys and minimize reading Blunders and manual transposition errors A
PC spreadsheet program, such as Excel®, may be used to reduce the dit to pled form, but
ate inefficient for data screening, eror diagnostics and correction, Tho moat efficient commer
al programs are DigiPro® and Gelk@® rom the US and Canade respectively,
2 INCLINOMETER ACCURACY
2 General fauee
Inclinemeter surveys ten with ordinary caro will produce relatively good results, However, or-
‘inary care does not produce results free from systematic errors. Depending on the cit:stances, several types of systematic ero can influence the readings and produce fale indication
of aisplacement. In addition, users that are unaware of the accuracy oftheir surveys often ten 0
‘ete! the diaement dia wan exeivelyexaggened wl, aui pig ls res.
Sion of sign 7
‘The relative simplicity of taking the date, autometed recording and data reduction by
“canned” software ean produce good results fst, but It hus also invited sub-professionals (2
‘produce unscreened deta reports. Thia trend is lest desirable wherever theo is low bid pro.
‘reatent of instramentation services, Low-level technical personnel are often put in the for
tion of both taking and reporting data ia terms of displacement plots. Field work and fice
Satan sf done by diferent gaps of peopl. This pers lad fc of unde
standing and allows unchecked eros to slip through, All enors look like displacement, and cal
lead to costly, false engineering conclusions by the unaware,
22 Feld accuracy
‘The product iterture by Slope Indicator Company inthe US for example states thet the systom
field accuracy is +1-7.8 mm in 30 m. It was derived empirically from «lange numberof project
ata ses in the mide1970s. This total error is considered tobe conservative number ais an
sggrogate of both random and systematic errors of 60 reading Increments with a O.Sameter
probe as shown on Figure 1. Random oror is typically no more than +/0.16 mm for a single
Feading increment based on checksum performance. However, random eror accumulates only
‘at a rate equal to the square root ofthe number of reading intervals. Systomatic eror on the
‘other han, is aritunetically accumulative. Ifthe systematic eror for one increment is essumed
to be 0.11 mm per reading increment, the total exor por 30-metere equels the empirially de
‘ved number as follows:
‘Total error™ Random eror+ Syatematic error
= (016% V60)+(0.11 x60)
= 266.60 = 7.8mm
TASS Ti)
"igure 1, Tota and random ncliomter rors, Random errs epee he bes system prelin.
223. Random error
‘The less influntal random error tends to reretin constant whereas the systematic error tends to
‘ery with each survey. The random error of += 1.24 mim pet 30 m remains afer ll systematic
ceors are removed and isthe limit of precision as shown in Figure 1. Accuracy may be im
Proved by repeating the surveys and using the mean results. However, ifenty a singe fnterval,
‘such asa shear zone, is of most interest, the error would only be #/- 0:3 mim,
24 Systematic error
Generally, one or s combination ofthe folowing factors generates systematic errors: Sensor
bins shi sensiivty dei, sensor alignment shift (rotation), depth postoning ero, casing iar
clinaton and curvature. Carectons fr these systemati errors can be made ifthe nature of he
‘rors is understood through the analysis ofthe dat. The methods aalabl for detection and
correotion are relatively unknown and wil be defined and discussed below,
naa ng ete hs hanes of rs nd oy inn oer em at
is point, one can senge tat accuracy is a commodity, Is guaity can be speci I
‘mented, Balanced with what is stritly needed and What somecoo is ‘wiling to pay fori As
demonstrated in Figo | there i signifeant room for improving the accuracy from +78 am
er 30'm to +/1,24:mm per 30 m by minimizing or coreding the systematic error. tis usually
tchievable on exceptionally good installations that are vertical aod fee from excesive curva
fare using careful fold procedures The alternative isto use corection procedures described
below.
3. SYSTEMATIC ERROR DETECTION AND CORRECTION
‘Systematic errors can be corrected asing strict mathematical procedures. Rendom errors cannot
be corrected, but can be minimized in beter installations and with more prcite reading prooe-
dures, Mistakes or gross ecrors are ouside the scope ofthis discussion, but these ere normally
‘detected in the checksum ealclations and plot that shouldbe a routine check for ell datasets.
‘Once detected, the date set can be rejeced or errors corrected based on the readings in the op
posite direction as well s historical dats.
3.1 Definition of systematic error pes
‘Commercial software for reducing and graphing inlinometer data such a DiiPro and Git
Ihave the option to insert correction factors for common errors. However, identifying the error
‘ype and quantifying it requires experience and judgement The four, most commen systematic
errors that can be comocted are:
‘+A single calfration blas-shift berween opposite reuding traverses in a data st isthe most
‘common systematic eror of all. I isa sell error within one dataset and is caused by a
shift inthe sensor calibration value "b” (6c Figure 2) between opposite traverses. With &
Title practice, tis simple to cores
‘+ Sensitivity drift i the last common error, but is often the most devious eror to netic.
‘Once recognized, it is eagy to correct in most cases, Factory ealibetion of the instrament
constant *K" (ee Figure 2) and repair of the probe is usually required
‘+The combination of casing inclination and seasor axis alignment shift produces rotation
‘error. Any small amount of shif in the probe or senor alignment since th initial dataset
produces this eror wien to casing has bees installed with an inclination in the erss-a
‘Traland error corrections are simple to make with sme pectic.
‘+ The combination of significant casing curvature and vertical plement eror ofthe probe,
potenti foe real displacement.
I Jecmenta should be questioned and chee. Systematic eors should be a
ossible, corrected a suggested above
pone come ely us 5 detect ad comet wt a sof
rin eal ial o det, to
setae reg cul be overcoreted Wy inex ee
pil iss eombinan wih checksum end nals.
ae hn pia nie ofthe sensor and wheel cai be
probably the on itv del wit, sino installed in
regressive deplaemens els jnformatin as
‘this way hs in appar po iemetay ses ad remeia action wee 7
1h eas rand. As sim, S08 2 of error an prefersbly correct:
aly soe sso be part oth engine enn Are nclinometers ee used.
REFERENCES
‘Coraferh, D. H. 1973 eras Chanson lope net Sec 0-8 neo
ie Gece Se, nce Cari Geta Engrg 126135 Now YORK
ant 988, Getcha! Isrumentaion for Motori eld Peformance. 256281, Now
"york Jain Wiley.
a a eB. 1988, Deena sn Traperta
ean es shan DC: Trsperos Reesch ‘Beer, National Re
tL
ad Wit, 8: D. 1983. nl retain: Aseray Pefonans ABET
Procurene. sa Mi grote minal Sporn Fl Mace
Procure Te Volt, Zari: 251-272, Ronda: Bale
enh ee, ld seen, eh), Lond
Hao, Seca prt 247: TENS, NaS ‘Dies Transporation
wi . Fie Tsuen in LL Sct and
ikon, a eed ad Contra Spl Report 176: Tizi3e, Wasingion, DC
a
BIS, 2006. Conformity Assessment Vocabulary and General Principles (IS/ISO/IEC 17000:2004) - BUREAU OF INDIAN STANDARDS, MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG, NEW DELHI 110002.