Geokon Model 4900 4900 Vibrating Wire Load Cell, GK-403, GK-404, GK-405 Readout Box, Load Cell Module Instruction Manual

Instruction Manual
Model 4900
Vibrating Wire Load Cell
No part of this instruction manual may be reproduced, by any means, without the written consent of Geokon, Inc.
The information contained herein is believed to be accurate and reliable. However, Geokon, Inc. assumes
no responsibility for errors, omissions or misinterpretation. The information herein is subject to change
without notification.
Copyright © 1981-2016 by Geokon, Inc.
(Doc Rev V, 3/24/16)
Warranty Statement
Geokon, Inc. warrants its products to be free of defects in materials and workmanship, under normal
use and service for a period of 13 months from date of purchase. If the unit should malfunction, it must
be returned to the factory for evaluation, freight prepaid. Upon examination by Geokon, if the unit is
found to be defective, it will be repaired or replaced at no charge. However, the WARRANTY is VOID
if the unit shows evidence of having been tampered with or shows evidence of being damaged as a
result of excessive corrosion or current, heat, moisture or vibration, improper specification,
misapplication, misuse or other operating conditions outside of Geokon's control. Components which
wear or which are damaged by misuse are not warranted. This includes fuses and batteries.
Geokon manufactures scientific instruments whose misuse is potentially dangerous. The instruments
are intended to be installed and used only by qualified personnel. There are no warranties except as
stated herein. There are no other warranties, expressed or implied, including but not limited to the
implied warranties of merchantability and of fitness for a particular purpose. Geokon, Inc. is not
responsible for any damages or losses caused to other equipment, whether direct, indirect, incidental,
special or consequential which the purchaser may experience as a result of the installation or use of the
product. The buyer's sole remedy for any breach of this agreement by Geokon, Inc. or any breach of any
warranty by Geokon, Inc. shall not exceed the purchase price paid by the purchaser to Geokon, Inc. for
the unit or units, or equipment directly affected by such breach. Under no circumstances will Geokon
reimburse the claimant for loss incurred in removing and/or reinstalling equipment.
Every precaution for accuracy has been taken in the preparation of manuals and/or software, however,
Geokon, Inc. neither assumes responsibility for any omissions or errors that may appear nor assumes
liability for any damages or losses that result from the use of the products in accordance with the
information contained in the manual or software.
TABLE of CONTENTS
1. INTRODUCTION ............................................................................................................................................. 1 1.1. THEORY OF OPERATION ............................................................................................................................ 1 1.2. LOAD CELL DESIGN AND CONSTRUCTION ................................................................................................ 1 1.2.1. Warping of the Bearing Plates and Bearing Plate Design ........................................................ 4 1.2.2. Bearing Plate Friction ..................................................................................................................... 5 1.2.3. Eccentric Loading............................................................................................................................ 5 1.2.4. Elastic Behavior ............................................................................................................................... 5 1.2.5. Temperature Effects ....................................................................................................................... 6 2. INSTALLATION ............................................................................................................................................. 6 2.1. PRELIMINARY TESTS ................................................................................................................................. 6 2.2. LOAD CELL INSTALLATION ........................................................................................................................ 6 2.2.1. Transportation.................................................................................................................................. 6 2.2.2. Initial No-Load Reading.................................................................................................................. 6 2.2.3. Installation on Tie-Backs and Rockbolts...................................................................................... 6 3. TAKING READINGS ..................................................................................................................................... 7 3.1. OPERATION OF THE GK-403 READOUT BOX .......................................................................................... 7 3.1,1. Using the Flying Leads ................................................................................................................... 7 3.1.2. Using a Load Cell Module .............................................................................................................. 7 3.1.3. Terminal Emulation Enhancements ............................................................................................. 9 3.2 OPERATION OF THE GK404 READOUT BOX ........................................................................................... 10 3.3 OPERATION OF THE GK405 READOUT BOX ........................................................................................... 10 3.4. MEASURING TEMPERATURES ................................................................................................................. 13 4. DATA REDUCTION .................................................................................................................................... 13 4.1. LOAD CALCULATION................................................................................................................................ 13 4.2. TEMPERATURE CORRECTION FACTOR ................................................................................................... 15 4.3. ENVIRONMENTAL FACTORS.................................................................................................................... 15 5. TROUBLESHOOTING ................................................................................................................................ 17 APPENDIX A - SPECIFICATIONS ............................................................................................................... 18 A.1. MODEL 4900 LOAD CELL SPECIFICATIONS ........................................................................................... 18 A.2. THERMISTOR ........................................................................................................................................... 18 APPENDIX B - THERMISTOR TEMPERATURE DERIVATION............................................................... 19 APPENDIX C - WIRING AND CONNECTOR PINOUTS ............................................................................ 20 C.1. LOAD CELL CONNECTOR AND CABLE (STANDARD WIRING) ................................................................. 20 C.2. GK-403 TO MODULE CONNECTOR ........................................................................................................ 20 APPENDIX D - VIBRATING WIRE LOAD CELL GAGE FACTOR RECALCULATION ............................................ 21 APPENDIX E - LOAD CELL CALIBRATIONS - EFFECTS OF BEARING PLATE WARPING .......... 22 INTRODUCTION ................................................................................................................................................ 22 LOAD CELL CALIBRATION PROCEDURES....................................................................................................... 22 FIELD ARRANGEMENT ..................................................................................................................................... 22 REPORT ON RECENT TESTING ........................................................................................................................ 23 CONCLUSION ................................................................................................................................................... 24 APPENDIX F – THE USE OF THE REGRESSION ZERO WHEN USING THE LINEAR GAGE
FACTOR ........................................................................................................................................................ 25 LIST of FIGURES, TABLES and EQUATIONS
FIGURE 1 - LOAD CELL ON TIEBACK FOR THE PERMANENT MONITORING OF LOAD ............................................ 2 FIGURE 2 - LOAD CELL ON TIEBACK FOR PROOF TESTING ONLY ........................................................................ 2 FIGURE 3 - LOAD CELLS FOR LOAD MONITORING DURING PILE LOAD TEST ....................................................... 3 FIGURE 4 - MODEL 4900 (3 GAGE) VIBRATING WIRE LOAD CELL....................................................................... 3 FIGURE 5 - TYPICAL LOAD CELL SYSTEM ............................................................................................................. 4 FIGURE 6 - MODE G WITH LOAD CELL MODULE ................................................................................................... 8 FIGURE 7 GK405 READOUT UNIT ....................................................................................................................... 11 EQUATION 1 - DIGITS CALCULATION ................................................................................................................... 13 EQUATION 2 - LOAD CALCULATION USING LINEAR REGRESSION ...................................................................... 13 TABLE 1 - ENGINEERING UNITS CONVERSION MULTIPLIERS ............................................................................. 14 EQUATION 3 - LOAD CALCULATION USING POLYNOMIAL ................................................................................... 14 FIGURE 8 - TYPICAL MODEL 4900 CALIBRATION SHEET.................................................................................... 16 TABLE A-1 MODEL 4900 LOAD CELL SPECIFICATIONS ..................................................................................... 18 TABLE B-1 THERMISTOR RESISTANCE VERSUS TEMPERATURE ....................................................................... 19 1
1. INTRODUCTION
1.1. Theory of Operation
Geokon load cells are of an annular design primarily for use on tiebacks and rockbolts. They
may also be used during pile load tests and for monitoring loads in struts, tunnel supports, etc.
In practically all cases, the load cells are used in conjunction with a hydraulic jack, which applies
the load, and with bearing plates positioned on either side of the load cell.
NOTE that to protect the lead wires routed in a groove in one face of the load cell we do
not allow the use of a washer made from lead, copper, rubber or other soft material
bearing against this surface.
If a soft washer is used be sure it is used only on the other face – the one that does not
have the annular epoxy-filled groove.
The load cell is frequently used:
 To confirm the load as determined by the hydraulic pressure applied to the jack during
proof testing on tiebacks, rockbolts, etc.
 To provide a permanent means of monitoring the load throughout the life of the tieback,
rockbolt, strut or support, etc.
 To provide an electronic output for automatic data gathering.
Load cells are positioned so that the tensile load in the tieback or rockbolt produces a
compressive load in the load cell. This is done by trapping the load cell between bearing plates
positioned between the jack and the structure, either below the anchor head for permanent
installations or above the anchor head for proof-testing. Figures 1 and 2 show the two different
installations.
Figure 3 illustrates load cells being used for load monitoring during a pile load test.
1.2. Load Cell Design and Construction
The Model 4900 Load Cell body is constructed in the form of a high strength steel cylinder in
which 3 to 6 vibrating wire strain gages are embedded to measure the change of strain in the
cylinder as it comes under load. Multiple gages are needed to account for the effects of offcenter or eccentric loading. Figures 4 shows a typical load cell.
See Appendix A for complete specifications.
The cable is attached to the cell through a waterproof gland or a connector. If the load cell is
equipped with the waterproof gland, a strain relief, in the form of a Kellem's grip, prevents the
cable from being pulled out of the cell. Cables have thick PVC jackets and can be terminated in
a connector to mate with terminal boxes or the Load Cell Multiplexer for use with the GK-403
Readout Box manufactured by Geokon. See Appendix C for cable and connector diagrams.
2
Bearing Plates
Lock Off Nut
Tendon or Rod
Load Cell
Wale
Soldier Pile
Anchor Zone
Figure 1 - Load Cell on Tieback for the Permanent Monitoring of Load
Bearing Plates
Loading Shoe
Hydraulic Jack
Load Cell
Tendon or Rod
Lock Off Nut
Wale
Soldier Pile
Anchor Zone
Figure 2 - Load Cell on Tieback for Proof Testing Only
3
Reaction Beam
Threaded Bars (6)
Reaction Pile #1
Ground Surface
Spherical Bearing
Load Cell #1
Hydraulic Jack
Spherical Bearing
Threaded Bars (6)
Load Cell #2
Hydraulic Jack
Reaction Pile #2
Load Plate
Test Pile
Figure 3 - Load Cells for Load Monitoring during Pile Load Test
Figure 4 - Model 4900 (3 Gage) Vibrating Wire Load Cell
4
Additional cable protection can be obtained by either using armored cable or by placing the
cable inside flex conduit.
Figure 5 shows a typical load cell system.
Figure 5 - Typical Load Cell System
Annular load cells, because of their design, are inherently susceptible to varying conditions of
end loading, unlike solid load cells, which can be designed with button shaped ends so that the
load always falls in a uniform, predictable fashion. Thus, the output and calibration of an
annular load cell can be affected by end effects produced by:
a) Warping of the bearing plates.
b) Friction between bearing plate and load cell.
c) Eccentric loading.
All of these effects can be accumulative so that the calibrations can vary by as much as
20%, unless special precautions are taken. Considering each effect in turn:
1.2.1. Warping of the Bearing Plates and Bearing Plate Design
Warping of the bearing plates is caused primarily by a size mismatch between the hydraulic jack
and the load cell. A jack larger than the load cell tends to wrap the intervening bearing plate
around the load cell, causing the center of the load cell to "hourglass" or pinch inwards causing
the load cell to under-register.
Conversely, a hydraulic jack, smaller than the load cell, will try to punch the intervening bearing
plate through the center of the load cell, making the center of the load cell barrel outwards
causing the load cell to over-register. Both effects are exacerbated by bearing plates that are
too thin.
For further details on this topic, the reader is referred to Appendices C and D.
5
NOTE that to protect the lead wires routed in a groove in one face of the load cell we do
not allow the use of a washer made from lead, copper, rubber or other soft material
bearing against this surface.
If a soft washer is used be sure it is used only on the other face – the one that does not
have the annular epoxy-filled groove
Minimum bearing plate thickness is one inch (25 mm) where load cell size matches hydraulic
jack size, i.e., the load bearing annulus of the load cell falls within the load bearing annulus of
the hydraulic jack. For any other condition of size mismatch, the bearing plates should be at
least two inches thick and even thicker where the size mismatch is extreme or the loads large.
Bearing plates should be flat and smooth. The normal rolled steel plate surface is adequate. It
is not necessary to have machined or ground surfaces. Where plates are cut from larger plates,
using cutting torches, the edges should be carefully cleaned to remove welding slag and
solidified molten lumps.
Consideration should be given to calibrating the load cell using the same bearing plates as will
be used in the field. Also, it is possible to simulate the size of the hydraulic jack using a suitably
sized metal donut between the upper platen of the testing machine and the upper bearing plate.
Load cells calibrated in this way, will be much more likely to agree with the hydraulic jack in the
field.
1.2.2. Bearing Plate Friction
Friction between the bearing plate and the load cell can radically affect the performance of a
load cell. Interposing deformable plates or lubricant between the bearing plates and the load
cell in the field can cause the load cell to over-register, perhaps by as much as 10%. Again, for
best results, it is important to calibrate the load cell in the laboratory under the same loading
conditions as will be used in the field.
End effects of this nature can be reduced somewhat by using tall load cells. A rough rule of
thumb for good load cell design calls for a load cell height at least 4 times the wall thickness of
the loaded annulus. On some jobs where there are space restrictions calling for a pancake
style load cell, friction between bearing plates and load cell can give rise to large hysteresis
effects between loading and unloading cycles.
1.2.3. Eccentric Loading
Eccentric loading of load cells is the rule rather than the exception. Rarely is the axis of the
tieback, rockbolt or strut at right angles to the surface on which the anchor plate or strut rests.
In the case of tiebacks using multiple tendons, it is quite common for loads in individual tendons
to vary markedly, one from the other, despite best efforts to avoid this happening. Also, struts
are rarely at right angles to the soldier piles they may be supporting.
These factors combine to produce conditions in which the load cell experiences higher loads on
one side than on the other. This effect is compensated for by the strain gages embedded in the
wall of the cell being individually read and the average strain calculated. Thus, the higher
strains on one side are balanced by lower strains on the other and the average strain is not
affected. Thus, even gross amounts of load eccentricity can cause only slight (< 5%)
variations in the load cell output and calibration.
Eccentric loading can be minimized by using spherical bearing plates, but this is expensive and
is rarely done. Spherical seats may be of some value during pile load testing where uniformity
of the load on the top of the pile is highly desirable.
1.2.4. Elastic Behavior
Geokon Model 4900 Load Cells are designed so as to keep the normal working stresses below
30% of the yield stress of the load cell material.
6
Load cells are over-ranged from 120% to 150% of the design load prior to calibration so that,
as long as the load cell is never overloaded above this range, the no-load reading will not
change. The normal over-range capacity for a steel load cell is 300 to 400% before the load cell
will fail.
If a load cell is over-ranged and the no-load reading is shifted due to plastic yielding of the cell,
then the cell should be returned to the factory for inspection and re-calibration. Note, however,
that while the no-load zero may shift, the calibration constant will probably not be affected.
1.2.5. Temperature Effects
Temperature compensation is achieved by using vibrating wire strain gages whose thermal
coefficient is the same as that of the load cell material. Normally, the temperature coefficient of
the load cell is insignificant being -1.5 digits per °C. (see Section 4.2). (In special cases, if
required, the coefficient can be measured at the factory). It should be remembered, however,
that temperature changes on the loaded rockbolt, tieback, or strut can produce real changes of
load and these will be recorded by the load cell.
2. INSTALLATION
2.1. Preliminary Tests
Before installing the load cell, it should be checked by connecting it to the readout box and
taking a no-load reading. This reading, when compared with that given in the calibration data
provided with the load cell, will show if the cell is functioning properly. The two readings should
agree within about 50 digits (assuming that the same readout box is used for both readings).
See section 3 for readout instructions.
2.2. Load Cell Installation
2.2.1. Transportation
When transporting load cells, do not pull on the cable and, in particular, do not carry the load
cell by the cable. On the larger load cells threaded holes are provided in the ends to allow
eyebolts to be attached for lifting purposes.
2.2.2. Initial No-Load Reading
Before installing the load cell be sure to take the no-load reading. This reading is very
important since it is the reading that will be subtracted from all subsequent readings in order to
calculate the load. Note that each load cell has a different no-load reading, which is not zero.
See Section 3 for operation of the Readout Boxes.
2.2.3. Installation on Tie-Backs and Rockbolts
NOTE that to protect the lead wires routed in a groove in one face of the load cell we do
not allow the use of a washer made from lead, copper, rubber or other soft material
bearing against this surface.
If a soft washer is used be sure it is used only on the other face – the one that does not
have the annular epoxy-filled groove
Load cells should be installed between flat steel bearing plates of sufficient thickness: 1 inch
thick where load cell and jack are about the same size and 2" to 3" thick where size mismatches
are greater. Plates should be machined flat. Make sure that the bearing plates completely
cover the load-bearing surface of the load cell. Centralize the rockbolt or tieback inside the load
cell. Where the load cell I.D. is much bigger than the rockbolt or tieback, a centralizer bushing
can be used.
7
Where the anchor block of a multi-tendon tieback bears directly on the load cell, make sure that
the load cell bearing surface is completely covered by the anchor block. If the load cell is not
completely covered, then make sure that the calibration was performed using the anchor block.
If the calibration was performed without the anchor block then for best results consideration
should be given to recalibration with the anchor block.
Shield the cable for possible damage from blasting or traffic. Protect the end of the cable or the
cable connector from dirt by either using a cap on the connector or by storing the end of the
cable and/or connector inside a small box. Figure 5 shows a typical load cell system.
3. TAKING READINGS
3.1. Operation of the GK-403 Readout Box
3.1,1. Using the Flying Leads
If the load cell cable does not have a connector the individual leads will be identified as shown
in the wiring diagrams in Appendix C.1. Each sensor is read in turn by plugging the flying leads
into the terminal box at the “TRANSDUCER” port and then clipping either the red or black clip to
the lead marked “common” and the black or red clip in turn to the leads marked #1, #2, #3, etc.
The blue clip should be connected to the cable shield and the green and white clips to the cable
leads marked “thermistor”.
Switch the GK-403 “DISPLAY” selector switch to “B”. The sensor output is displayed in digits.
Read each channel in turn and record in a field book and/or by depressing the “STORE” button.
When using the “STORE” button it will be necessary to use the joystick to set the appropriate
I.D. Marker on the display screen before the “STORE” button is depressed to distinguish
individual gages (and load cells) from each other.
When the thermistor leads are hooked up the temperature at the load cell is automatically
displayed on the display screen in C.
The GK-403 will turn itself off after about 2 minutes.
3.1.2. Using a Load Cell Module
The Model GK-403 Load Cell Module acts as a multiplexer or automatic switch that can be used
to automatically read all active sensors, calculate the average reading change, apply the gage
factor and display the load in engineering units on the display screen. The 10 pin load cell cable
connector is plugged into the module and the lead from the module is plugged into the
“TRANSDUCER” port on the GK-403.
Note that the 10 pin load cell connector should not be plugged directly into the GK-403.
The current readings on all the active channels can be viewed by switching the “Display” selector
to Channel B and by operating the joystick. If desired, these current readings or the active
channels can be recorded directly into a field book and used later to calculate the load manually.
Care must be taken in setting up the GK-403 to read loads automatically and the reader is
referred to the GK-403 manual for further details. The essential procedure is as follows:
8
The GK-403 is switched to “DISPLAY” setting “G”. When this is done a display of the type
shown in Figure 6 will be observed on the display screen.
11/22/91 15:43
NOW 23.7 C
6547.3digLDCB
ROW: 1 COL: 1
ID:Load Cell 1
11/22/91 15:42
MEM 23.6 C
6547.1digPOSB
Figure 6 - Mode G with Load Cell Module
 From the Main Menu select Option 3 and set the correct date and time
 From the Main Menu select Option 5 “Gage Params” and then Option 6, “Switch Position” and
set the switch position to B
 Escape to the main reading screen and use the joystick to set the “Row” to “1” and “Col” to
“1”. The displayed date should be correct and the displayed “LDC” should be at “B”. The
temperature should be indicated in the top right hand corner.
 The easiest way to distinguish between load cells is to use the column (COL) number. It will
be seen that the columns run from 1 to 7 and then jumps to column 11 to 17 and then jumps to
21 to 27 etc. all the way to 241 to 247. Columns 8, 9 and 0 are never displayed. Load cells with
6 sensors will use all columns 1 through 6 (or 11 through 16 etc.) while load cells with 3 or 4
sensors will use only this number of channels and channels 4, 5, 6 or 5 and 6 will remain blank.
The first load cell can be read on COL 1 to7, the second on COL 11 to 17, the third on COL 21
to 27 etc.
 With the COL number set to 1 and the load cell at zero load it is now necessary to set the
zero readings. This is done by selecting the “ZERO” option from the “GAGE PARAMS” menu.
Now select option 2 “Use current reading as F (zero)”. This automatically sets the zero reading
to the current reading on channel 1 and also performs the same function on all the other active
channels from 1 to 6 so it is not necessary to repeat this process on the other active channels.
Now when the COL number is set to 7 the displayed load number should be close to zero
corresponding to zero load.
(Note that if the load cell has a load on it the above procedure cannot be used. Instead Option 3
on the Zero Factor menu must be selected and then the zero numbers entered manually from
the calibration sheet for each of the active channels. There is a shortcut to this procedure: the
average zero reading can be taken from the calibration sheet and input directly into channel 7
(17, 27, etc.). The drawback to this is that only channel 7 (17, 27, etc.) will show the load in
engineering units while the rest of the active channels will show only the readout digits. This
would make it more difficult to see the change in digits displayed on each active channel and to
appreciate the degree of uniformity or non-uniformity of the load applied to the load cell.)
9
 The gage factor shown on the calibration sheet, usually in lbs./digit or kgm/digit, must be
entered only on channel 7 (or 17, 27, etc.). It should not be entered on any of the active
channels 1 through 6 etc. (The gage factor on channels 1 thru 6 should be set to 1.000). So
return to the main menu and set COL reading to 7 (or 17, 27, etc.). However there are two
important provisos, which is that the gage factor on the calibration sheet must first be
converted to Kips or metric tons before it is entered. This is necessary in order not to over
range the limited capacity of the GK-403 readout screen. So a calibration factor of say 152.4
lbs./digit must be converted to 0.1524 Kips/digit. Secondly, it is necessary to put a negative
sign in front of the factor so +0.1524 becomes –0.1524.
 The correct units must now be set on channel 7 (17, 27, etc.) from the “Units” option on the
“Gage Params” screen. Set to either “Kip” or “mtn”.
 It will be wise to check the validity of the readings by comparing the displayed load readings
on channel 7 with the load calculated from the readings on the active channels taken with the
display switch on the GK-403 readout box set to position B.
 Readings on all channels can be stored in the GK-403 memory at any time by depressing the
“SELECT/STORE” button. To distinguish sets of readings taken at different times use the ROW
number, by advancing the row number with every set of readings. Any sets of readings at any
particular time can be accessed and inspected by scrolling through the ROW numbers. Note
that storing data on any ROW number will erase and write over any data already stored on that
ROW.
 A useful feature of the GK-403 is its ability to display the previous readings taken on any
channel. On the main screen the reading is at the bottom of the screen. Thus any sudden
changes of load from one time to the next are immediately apparent.
3.1.3. Terminal Emulation Enhancements
Pressing <ENTER> on the host computer while in terminal emulation with the GK-403 will
transmit the status of the Readout. If the Display Mode selected is G, the individual readings for
the Load Cell as well as the calculated average (or load) will display. The following illustrates;
<<gk403 STATUS>>
Date & Time: 09/11/92 11:42
Switch Pos: G
REF/COL#: 11 Load Cell Module attached.
Temperature: 22.3
xx1
xx2
xx3
xx4
xx5
6543.3
6554.3
6654.1
6589.0
6521.2
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
Approximately every 5 seconds the display will be updated.
xx6 xx7(avg)
6522.2 6566.3
·
·
·
·
·
·
10
3.2 Operation of the GK404 Readout Box
The GK404 is a palm sized readout box which displays the Vibrating wire value and the
temperature in degrees centigrade.
The GK-404 Vibrating Wire Readout arrives with a patch cord for connecting to the vibrating
wire gages. One end will consist of a 5-pin plug for connecting to the respective socket on the
bottom of the GK-404 enclosure. The other end will consist of 5 leads terminated with alligator
clips. Note the colors of the alligator clips are red, black, green, white and blue. The colors
represent the positive vibrating wire gage lead (red), negative vibrating wire gage lead (black),
positive thermistor lead (green), negative thermistor lead (white) and transducer cable drain wire
(blue). The clips should be connected to their respectively colored leads from the vibrating wire
gage cable.
Use the POS (Position) button to select position B and the MODE button to select Dg (digits).
Other functions can be selected as described in the GK404 Manual.
The GK-404 will continue to take measurements and display the readings until the OFF button
is pushed, or if enabled, when the automatic Power-Off timer shuts the GK-404 off.
The GK-404 continuously monitors the status of the (2) 1.5V AA cells, and when their combined
voltage drops to 2V, the message Batteries Low is displayed on the screen. A fresh set of 1.5V
AA batteries should be installed at this point
3.3 Operation of the GK405 Readout Box
The GK-405 Vibrating Wire Readout is made up of two components:

the Readout Unit, consisting of a Windows Mobile handheld PC running the GK-405
Vibrating Wire Readout Application
 the GK-405 Remote Module which is housed in a weather-proof enclosure and
connects to the vibrating wire sensor by means of:
1) Flying leads with alligator type clips when the sensor cable terminates in bare wires
or,
2) connection by means of a 10 pin connector..
The two components communicate wirelessly using Bluetooth®, a reliable digital
communications protocol. The Readout Unit can operate from the cradle of the Remote
Module (see Figure 7) or, if more convenient, can be removed and operated up to 20
meters from the Remote Module
11
Figure 7 GK405 Readout Unit
3.3.1 Readout in digits
1. Switch on the Handheld Field PC by depressing the power button at the lower right.
2. Wait 70 seconds for the Handheld to warm up. After the Windows Mobile Screen
appears wait another 20 second. (This wait is only when the box is first switched on.
When in the sleep mode the box will wake up instantly). Once the box has been
switched on it can be left in sleep mode.
3. Plug the flying leads into the 10 pin connector marked “Sensor” on the side of the box.
Connect the flying leads color for color to the leads on the load cell. If the load cell
has a 10 pin connector connect it directly to the 10 pin connector marked “Load Cell”
on the side of the box.
4. Now Press ‘Start’ {if it fails to respond press ‘Unlock’)
5. Select ‘GK405 VW Readout’ application
6. Press the Blue tooth button on the base of the GK405 the blue light will start blinking
7. After a few seconds the Bluetooth button on the readout box should now show a
steady blue light indicating that a connection has been made between Handheld and
base.
8. On the Live Reading Screen ‘Display Mode’ will show ‘B’ and the ‘Sensor Index’ ‘Avg’
if connected to multi-sensor load cell via the “Load Cell” connection.
9. The screen now displays the average digit reading of all the strain gage in the load
cell.
10. Record this reading in a field book.
12
Due to the lengthy warm-up period when first switching on the box it is best to leave the
readout in a sleep mode if more than one reading are to be taken during the course of the day.
This is done by first pressing ‘Menu’ and then select ‘Close GK405’ then pressing the Handheld
power button momentarily. To switch the readout box off completely hold the power button down
for few seconds until a ring tone is heard.
Writing data into a field book for later transferring to a spreadsheet is in many ways the best
approach to take, For further details concerning storage of the readings in the readout box
consult the GK405 Instruction Manual. If the readout is required in units of load then use the
following instructions:
3.3.2 Readout in Load units
Before you can store readings in the readout box you need to create a workspace where you
can identify the load cell and configure the load cell parameters - gage factor , zero reading,
units, etc., You need to do this for each load cell.
For each load cell a sensor file must be created. On the Workspace screen press and hold
‘Project Sensors then hit ‘Add Sensor’.
.
11. Hit the keyboard icon at the bottom center of the screen and use the keyboard to
enter the Load cell identification name and/or number
12. Select the sensor model number, 4900, from the drop-down menu
13. From the drop down menu insert the number of gages in the load cell.
14. Select the ► arrow to reach the next page
15. Select ‘Linear’ (If selecting ‘Polynomial’ follow the instructions for entering the A,B and
C coefficients).
16. Select ‘Output Calculation’ to G(R1-R0) in the dropdown menu,
17. Using the keyboard set the zero reading equal to the initial Regression Zero shown on
the calibration sheet. (See Appendix G for further explanation)
18. Using the keyboard set the ‘Gage Factor’ equal to the Gage Factor given on the
calibration sheet. Don’t forget the negative sign.
19. Leave “Gage Offset ‘ at zero
20. Select the ► arrow to reach the next page,
21. Select ‘Measure’ to ‘Load’.
22. Select the input units equal to the gage factor units.
23. Select the desired output units that will show on the display.
24. Select the ► arrow to reach the next page
25. For most application the temperature coefficient is not important in which case leave
the ‘Sensor Correction’ at ‘disabled’
13
26. Select ‘Menu’ and ‘Save the Settings’.
27. Select ‘Applications’
28. Select ‘Live Readings’
29. Select ‘With Selected Sensor’ (the same one created in Step 11)
30. The readings should now be in units of load.
31. To store data hit the ‘Store’ icon on the screen. The readings will be stored under that
particular load cell file, the one created in Step 11.
Note that if one of the strain gages fails a red warning will appear on the display. The readout
box will continue to display the average reading of the remaining strain gages.
At any time in the Live Reading screen, when ‘Raw Readings’ is selected it is possible to display
the reading on each individual strain gage in the load cell. This can be useful in assessing the
degree of eccentricity of the applied load,
3.4. Measuring Temperatures
Each Vibrating Wire Load Cell is equipped with a thermistor for reading temperature. The
thermistor gives a varying resistance output as the temperature changes. See the wiring chart
in Appendix C.1. for the appropriate connections.
1. If an ohmmeter is used, connect the ohmmeter to the two thermistor leads coming from the
load cell. (Since the resistance changes with temperature are so large, the effect of cable
resistance is usually insignificant.)
2. Look up the temperature for the measured resistance in Table B-1. Alternately the
temperature could be calculated using Equation B-1.
Note: The GK-403, GK404 and GK-405 readout boxes will read the thermistor and display
temperature in C automatically.
4. DATA REDUCTION
4.1. Load Calculation
The basic units utilized by Geokon for measurement and reduction of data from Vibrating Wire
Load Cells are "digits". Calculation of digits is based on the following equation;

2
1
3
Digits  
or
  10
 Period (sec onds ) 
Digits 
Hz
2
1000
Equation 1 - Digits Calculation
To convert the digits readings to load, the gage readings for each cell must be averaged, then
the change in reading average multiplied by the gage factor supplied with the load cell.
L = (R1 – R0)  G  K
Equation 2 - Load Calculation Using Linear Regression
14
Where;
L is the load in lbs. or kg. etc.
R0 is the regression no-load reading in digits (average of all gages).
R1 is the current reading in digits (average of all gages).
G is the gage factor as supplied on the Calibration Sheet (Figure8). It has a negative
sign since as the load increases the reading decreases
K is the conversion factor (optional) as listed in Table 1.
From
To
Lbs.
Kg.
Kips
Tons
Metric
Tonnes
Lbs.
1
0.4535
0.001
0.0005
0.0004535
Kg.
2.205
1
0.002205
0.0011025
0.001
Kips
1000
453.5
1
2.0
0.4535
Tons
2000
907.0
2.0
1
0.907
Metric
Tonnes
2205
1000
2.205
1.1025
1
Table 1 - Engineering Units Conversion Multipliers
For example, a Model 4900 has a regression no-load reading (R0) of 7309 (see Figure 8) and a
current average reading (R1) of 5497. The Calibration Factor is -397 lbs. per digit.
L = (6408  7309)  -397 = 357,700 lbs.
Note that the equations assume a linear relationship between load and gage readings over the
full load range, and the linear coefficient is obtained using regression techniques. Note that
when using the Calibration Factor obtained from the regression formula it is necessary
to use the regression zero. This may introduce substantial errors at very low loads. A
measure of the amount of non-linearity is shown on the Calibration Sheet in the column entitled
“Linearity”. (See Figure 8). See Also Appendix
For greater accuracy, the data given can be represented by a polynomial or can be treated as a
series of segments over the entire load range.
For instance, in the example of Figure 8, the load between 0 and 180,000 lbs. could be
represented by the following equation;
L = ((7304  6860)  -405 = 179,820 tonnes.
The gage factor -405 lbs./digit is calculated from the slope of the line between a load of 0 and
180,000 lb., i.e.,
(0 – 180,000)/(7304– 6860) = -405 lbs./digit
A polynomial expression to fit the data would be:
L = A*R12 + B*R1 + C
Equation 3 - Load Calculation Using Polynomial
Where;
L is the load in lbs., kgms., etc.
R1 is the current reading (average of all gages).
A, B and C are the coefficients derived from the calibration data.
15
First Calculate C from the initial average field zero reading, Suppose this is 7305
Then,
0 = -0,00247* 73052 – 367*7305 + C, from which C = +2,812,740
So when the applied load is 360,000, R1 = 6409 and
The calculated Load = -0,00247* 64092 – 367*6409 +2,812740 = 359,180 lbs.
4.2. Temperature Correction Factor
There is a small correction that can be made for temperatures. As the temperature goes up the
average reading of all the sensors will go down by about one digit per °C So the load, corrected
for temperature, would be
L = G [ (R1 – R0) + (T1 – T0) ]
The temperature effect shown above is for a load cell that has not been installed yet and is very
minor. There is no telling what the actual temperature effect will be on a load cell that is installed
on a tensioned bar or cable. This depends on the length of the bar or cable and on the
properties of the surrounding ground. The actual temperature effect can only be arrived at
empirically by simultaneous measurements of load and temperature over a short period of time.
4.3. Environmental Factors
Since the purpose of the load cell installation is to monitor site conditions, factors which may
affect these conditions should be observed and recorded. Seemingly minor effects may have a
real influence on the behavior of the structure being monitored and may give an early indication
of potential problems. Some of these factors include, but are not limited to: blasting, rainfall,
tidal or reservoir levels, excavation and fill levels and sequences, traffic, temperature and
barometric changes, changes in personnel, nearby construction activities, seasonal changes,
etc.
16
Figure 8 - Typical Model 4900 Calibration Sheet
17
5. TROUBLESHOOTING
Problems with the load cell are usually associated with cable damage or moisture getting into
the system. Both problems can be minimized by protecting the cable from damage, by visual
inspection of the cable in the event that problems arise and by keeping the plug clean and dry at
all times. Avoid carrying the load cell by the cable.
Check the cable for damage such as pulling out of the load cell or connector, crushed spots,
cuts or kinks. If there is cable damage, the cable should be repaired by cutting and splicing. All
splices should be mechanically strong (soldering connections is usually best), well insulated and
protected from dirt and moisture with an epoxy based splice kit such as the such the 3M

Scotchcast , model 82-A1. These kits are available from the factory. Alternately, a mastic
type sealant, such as AquaSeal (Kearney), and vinyl tape may be used to cover a splice.
Consult the following list of problems and possible solutions should difficulties arise. Consult
the factory for additional troubleshooting help.
Symptom: Load Cell Gage Readings are Unstable
 Is the readout box position set correctly? If using a datalogger to record readings
automatically are the swept frequency excitation settings correct? Channel A of the GK-401
and GK-403 can be used to read the strain meter. To convert the Channel A period display
to digits use Equation 1.
 Is there a source of electrical noise nearby? Most probable sources of electrical noise are
motors, generators and antennas. Make sure the shield drain wire is connected to ground
whether using a portable readout or datalogger. If using the GK-401 Readout connect the
clip with the green boot to the bare shield drain wire of the load cell cable. If using the GK403 connect the clip with the blue boot to the shield drain wire.
 Does the readout work with another load cell? If not, the readout may have a low battery or
be malfunctioning.
Symptom: Load Cell Gage Fails to Read
 Is the cable cut or crushed? This can be checked with an ohmmeter. Nominal resistance
between the two gage leads is 45 to 50. (75, 90 or 180, 10 on some older models)
Remember to add cable resistance when checking (22 AWG stranded copper leads are
approximately 14.7/1000' or 48.5/km, multiply by 2 for both directions). If the resistance
reads infinite, or very high (megohms), a cut wire must be suspected. If the resistance
reads very low (20) a short in the cable is likely.
 Does the readout or datalogger work with another load cell? If not, the readout or
datalogger may be malfunctioning.
Symptom: Thermistor resistance is too high.
 Is there an open circuit? Check all connections, terminals and plugs. If a cut is located in
the cable, splice according to instructions above.
Symptom: Thermistor resistance is too low.
 Is there a short? Check all connections, terminals and plugs. If a short is located in the
cable, splice according to instructions above.
 Water may have penetrated the interior of the load cell. There is no remedial action.
18
APPENDIX A - SPECIFICATIONS
A.1. Model 4900 Load Cell Specifications
Available Ranges:¹
Accuracy:
Linearity:
Resolution:²
Repeatability:
Temperature Effect:
Temperature Range:
Frequency Range
Over range:
Coil Resistance:
Cable Type (3 Gage):³
Cable Type (4 Gage):³
Cable Type (6 Gage):³
100, 150, 200, 300, 500, 600, 1000, 1500, 2000 kips
1.0% FSR (or better)
0.5% FSR
0.02% FSR
0.1% FSR
0.02% FSR/°C
40 to +80° C
-40 to 110° F
1400-3500Hz
150%
45 to 50 (70, 90, or 180  on some older models)
4 twisted pair (6 conductor) 22 AWG, Purple jacket
Foil shield, PVC jacket, nominal OD=9.5 mm (0.375")
4 twisted pair (8 conductor) 22 AWG, Purple jacket
Foil shield, PVC jacket, nominal OD=9.5 mm (0.375")
6 twisted pair (12 conductor) 22 AWG, Orange jacket
Foil shield, PVC jacket, nominal OD=12.7 mm (0.5")
Table A-1 Model 4900 Load Cell Specifications
Notes:
¹ Other ranges available.
² Minimum, depends on the readout instrument and technique.
³ Other cable types, i.e. armored, are available.
A.2. Thermistor
Range: -80 to +150° C
Accuracy: ±0.5° C
19
APPENDIX B - THERMISTOR TEMPERATURE DERIVATION
Thermistor Type: YSI 44005, Dale #1C3001-B3, Alpha #13A3001-B3
Resistance to Temperature Equation:
T
1
 273.2
A  B( LnR )  C( LnR )3
Equation B-1 Convert Thermistor Resistance to Temperature
Where;
Ohms
201.1K
187.3K
174.5K
162.7K
151.7K
141.6K
132.2K
123.5K
115.4K
107.9K
101.0K
94.48K
88.46K
82.87K
77.66K
72.81K
68.30K
64.09K
60.17K
56.51K
53.10K
49.91K
46.94K
44.16K
41.56K
39.13K
36.86K
34.73K
32.74K
30.87K
29.13K
27.49K
25.95K
24.51K
23.16K
21.89K
20.70K
19.58K
18.52K
17.53K
T  Temperature in C.
LnR  Natural Log of Thermistor Resistance
A  1.4051  10-3 (coefficients calculated over the 50 to +150 C. span)
B  2.369  10-4
C  1.019  10-7
Temp
-50
-49
-48
-47
-46
-45
-44
-43
-42
-41
-40
-39
-38
-37
-36
-35
-34
-33
-32
-31
-30
-29
-28
-27
-26
-25
-24
-23
-22
-21
-20
-19
-18
-17
-16
-15
-14
-13
-12
-11
Ohms
16.60K
15.72K
14.90K
14.12K
13.39K
12.70K
12.05K
11.44K
10.86K
10.31K
9796
9310
8851
8417
8006
7618
7252
6905
6576
6265
5971
5692
5427
5177
4939
4714
4500
4297
4105
3922
3748
3583
3426
3277
3135
3000
2872
2750
2633
2523
Temp
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Ohms
2417
2317
2221
2130
2042
1959
1880
1805
1733
1664
1598
1535
1475
1418
1363
1310
1260
1212
1167
1123
1081
1040
1002
965.0
929.6
895.8
863.3
832.2
802.3
773.7
746.3
719.9
694.7
670.4
647.1
624.7
603.3
582.6
562.8
543.7
Temp
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
Ohms
525.4
507.8
490.9
474.7
459.0
444.0
429.5
415.6
402.2
389.3
376.9
364.9
353.4
342.2
331.5
321.2
311.3
301.7
292.4
283.5
274.9
266.6
258.6
250.9
243.4
236.2
229.3
222.6
216.1
209.8
203.8
197.9
192.2
186.8
181.5
176.4
171.4
166.7
162.0
157.6
Temp
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
Ohms
153.2
149.0
145.0
141.1
137.2
133.6
130.0
126.5
123.2
119.9
116.8
113.8
110.8
107.9
105.2
102.5
99.9
97.3
94.9
92.5
90.2
87.9
85.7
83.6
81.6
79.6
77.6
75.8
73.9
72.2
70.4
68.8
67.1
65.5
64.0
62.5
61.1
59.6
58.3
56.8
55.6
Table B-1 Thermistor Resistance versus Temperature
Temp
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
20
APPENDIX C - WIRING AND CONNECTOR PINOUTS
C.1. Load Cell Connector and Cable (standard wiring)
10 pin
Bendix
PT06A-1210P
A
B
C
D
E
F
G
H
J
K
Function
Gage #1
Gage #2
Gage #3
Gage #4
Gage #5
Gage #6
Shield
Common
Thermistor
Thermistor
3 Gage VW Load
Cell
Geokon Purple
Cable
Red
Red's Black
White
NC
NC
NC
All Shields
White's Black¹
Green¹
Green's Black
4 Gage VW Load
Cell
Geokon Purple
Cable
Red
Red's Black
White
White's Black
NC
NC
All Shields
Green
Blue
Blue's Black
6 Gage VW Load
Cell
Geokon Orange
Cable
Red
Red's Black
White
White's Black
Green
Green's Black
All Shields
Blue
Yellow
Yellow's Black
Notes:
¹ White's black and Green wires are switched on Geokon 3 gage VW load cells prior to serial
number 3313.
C.2. GK-403 to Module Connector
Module
10-pin Bendix Plug
(PT06F-12-10P)
A
B
C
D
E
F
G
H
J
K
Interconnect
Wire Color
(6 Pair)
Brown
Brown's Black
Red
Red's Black
Yellow
Yellow's Black
Green
Green's Black
Blue
Blue's Black
Interconnect
Wire Color
(Belden)
Description
Brown
VW Gage
Red
VW Gage Ground
Orange
Thermistor
Yellow
Thermistor Ground
Green
Shield
Blue
+12 VDC
Violet
Ground
Grey
Mux Sense
White
Mux Clock
Black
Mux Type
Module
Board
Connection
JP1-2
JP1-1
JP1-3
JP1-1
JP1-1
JP1-4
JP1-9
JP1-9
JP1-8
JP1-9
21
APPENDIX D - Vibrating Wire Load Cell Gage Factor Recalculation
Note, the following applies only to manual readout. When using the GK403 with the Load
Cell Module the absence of any gage is automatically compensated for.
Overview
This appendix describes how to recalculate the gage factor for a Load Cell and then
approximate the load where one or more strain gages in the cell have failed after installation.
Procedure
If the load is applied uniformly to the load cell then, as the load changes the change in reading
on each gage will be the same and, should one gage fail, the gage factor given on the
calibration sheet can be applied to the average change of the remaining gages
Note the following example where #3 gage in a six-gage load cell has failed. The load cell gage
factor for the six gages is 0.2439 tonnes./digit. If the load is uniformly applied to the load cell,
then, to calculate the load we can simply apply this gage factor to the average reading change
of the remaining five active gages: in the example below the load on 7/1/02 would be calculated
to be 0.2439( 7298-6139) = 282.7 tonnes.
In the field however it is a rare condition to have the cell uniformly stressed. So, it may be more
accurate to calculate a new gage factor using just the active gages.
In cases where the load is eccentric (in the present example the reading change on gage #3
was higher than the other five gages), we can calculate the new gage factor for the remaining
five active gages as follows:
Date
Gage #1 Gage #2 Gage #3
Gage
Gage
Gage
Avg
Load
#4
#5
#6
Initial
7318
7363
7247
7448
7222
7191
7298
0
6/1/02
6485
6363
6220
6618
6362
6331
6396
220.2 tonnes
7/1/02
6202
6034
No
6324
6075
6058
6139 293.8 tonnes.
Reading
1)
Calculate a new zero load average using only the initial readings of the five remaining
active gages = 7308
2)
Using only the readings of the active gages #1, #2, #4, #5 and #6 from the time of the last
readings, (6/1/02), when all six gages were active, calculate the average reading: = 6432
3)
Calculate the new gage factor for the remaining five active gages by dividing the
calculated load at the last time when all gages were active, (6/1/02), by the change in the
five gage average readings calculated in steps 1 and 2, = 220.2 /(7308-6432) = 0.2514.
This is the new gage factor to be applied to all subsequent changes of the remaining five
active gages
4)
Using the averages of the current and initial five-gage readings, calculate the load on
7/1/02 by using the new gage factor. Thus on 7/1/02: (7308 – 6139) x 0.2514 = 293.9
tonnes . As will be seen this gives a better result than applying the old gage factor for the
six gages to the average reading of the five active gages. (The applied load was 291
tonnes)
5)
Repeat step 4 for subsequent readings or repeat all steps if more gages in the load cell
fail.
Limitations
This is not a foolproof method: For example, if the load distribution changes in the course of
monitoring, the calculations based on the above-described method will be in error
22
APPENDIX E - LOAD CELL CALIBRATIONS - EFFECTS OF BEARING PLATE WARPING
Introduction
Load cells used to measure loads during testing of tiebacks, driven piles and drilled shafts give
calculated loads which are frequently in disagreement with loads calculated on the basis of
hydraulic jack pressure and piston area. Because of this, there is a general lack of confidence
in load cell data and the fault is often ascribed to manufacturing defects, or to improper,
inaccurate calibration procedures. Nevertheless, it is also well known that the effects of
eccentric loading and uneven and/or warped bearing plates do have a profound effect on load
cell readings. The purpose of this technical note is to provide some insight into these effects.
Load Cell Calibration Procedures
The usual calibration procedure is to use a testing machine to apply a load to a load cell. The
measured load cell output is then correlated against the known applied load as measured by the
testing machine. Usually, the testing machine has a hydraulic pressure applied to a piston of
known cross section area. The testing machine itself is checked out periodically by running
tests on a load cell traceable to NIST and there is generally little doubt about the accuracy of the
testing machine. Accuracy's of ¼% FS ½% FS or 1% FS are normal.
Usually, the calibration tests are performed between large, flat parallel platens in the testing
machine so that there is no bending of the platens, only the elastic compression in the zone
immediately bearing against the load cell.
Field Arrangement
Such a state of affairs may not exist on the job site since the bearing surfaces next to the load
cell are usually much less rigid, and liable to bending.
This bending is particularly apparent if there is a mismatch in size between the load cell and the
hydraulic jack. If the hydraulic jack is larger than the load cell there is a tendency for it to try to
wrap the intervening bearing plate around the load cell. If the hydraulic jack is smaller than the
load cell it will try to push the intervening bearing plate through the hole in the load cell.
Thicker bearing plates will bend less, but the effect will never be entirely eliminated. The
consequence of this bending can be quite large since the effect on the load cell is to cause it to
either barrel out at its mid-section if the jack is too small, or pinch in at its mid-section if the jack
is too big. For vibrating wire load cells the gages are usually located in the center of the cell
wall, on the neutral axis, thereby minimizing these effects.
23
Report on Recent Testing
A series of tests were conducted in a testing machine to investigate the magnitude of this effect.
A load cell with a bearing surface of 4" ID, 5¾" OD was used.
Simulated jack A had a bearing surface of 2" ID, 4" OD.
Simulated jack B had a bearing surface of 4" ID, 5¾" OD.
Simulated jack C had a bearing surface of 6" ID, 8" OD.
The maximum applied load was 150 tons.
Load Cell response to applied load (100%)
Jack
1" thick plate
2" thick plate
108%
102%
100%
100%
96%
98%
J
A
(smaller)
LC
J
B
(same size)
LC
J
C
(bigger)
LC
From the results it can be seen that if the jack is smaller than the load cell, the load cell will
over-register, while a jack bigger than the load cell will cause the load cell to under-register.
The effect is bigger if the bearing plate between jack and load cell is thinner.
The correct bearing plate thickness will of course depend on the extent of the mismatch
between jack and load cell. However as a rough rule of thumb the following thickness should be
required:
100-200 kip load……………...25mm(1”) thick
Up to 400 kip load……………37mm (1.5”) thick
Up to 1000 kip load…….…….50mm (2 “) thick
Up to 2000kip load …………..75mm (3”) thick
24
Conclusion
The consequences of all this would seem to indicate that, for best results, the load cell
calibration should be performed with the actual hydraulic jack that will be used, both being
placed in the testing machine at the same time. Or failing that, the load cell should be loaded
through a ring, having the same dimensions as the hydraulic jack bearing surface, positioned
on the other side of a bearing plate of the correct thickness. In this way one of the variables
affecting the agreement between load cell readings and hydraulic jack readings can be removed
and the agreement should be that much closer.
This technical note has addressed only the subject of the size mismatch between load cells and
hydraulic jacks. Other factors affecting the agreement between load cell readings and hydraulic
jack load are important: thus frictional losses within the hydraulic jack can cause underregistering of jack load indications by as much as 15%. (Dunnicliff 1988' Section 13.2.6)
Also annular style load cells are susceptible to end effects and eccentrically applied loads. The
height of the load cell should exceed 4 times the wall thickness of the annulus and at least 3
strain gages should be used increasing to 6 as the size of the load cell increases.
References
J. Dunnicliff. 1988. Geotechnical Instrumentation for Monitoring Field Performance, John Wiley
& Sons, New York, NY: 577pp.
25
Appendix F – The Use of the Regression Zero when Using the Linear Gage Factor
It is normal for load cells, having an annular design and for solid load cells that do not have
‘button heads’ or spherical seated bearing plates, to be susceptible to irregular load distributions
at low loads. This is because there is a ‘bedding in’ process that takes place while the surfaces
at both ends of the load cell conform to the surfaces they bear against causing the load cell to
deform in an unpredictable way giving rise to strange strain patterns and faulty readings at low
loads.
This irregularity of load disappears once the load cell surfaces have ‘bedded in’ and from that
point on the load cell behaves in the a more linear fashion such that there is a constant
relationship between the applied load and the observed change in readout as quantified by the
linear gage factor shown on the calibration sheet.
Because of this the linear gage factor shown on the calibration sheet has been calculated after
excluding the often-anomalous zero reading from the data points. And this gage factor best
describes the performance of the load cell at moderate to higher loads
This linear gage factor describes the slope of the best fit line drawn through the calibration data
points and the reading where the line intersects the zero load point on the load axis is called the
‘Regression Zero’ shown on the calibration sheet.
It is important when using the linear gage factor to calculate loads that the value of R0 in the
linear equation be equal to the regression zero.
For greater accuracy a 2nd order polynomial can be used to map the data points. In this case the
regression zero is replaced by the factor C shown on the calibration sheet.
Now it may be, for a variety of reasons, (e.g., if the load cell is used repeatedly on a number of
jobs), that the no load zero reading might change significantly. Again, for greater accuracy, the
value of the Regression Zero can be adjusted by an amount equal to the observed change in
the no load zero from that shown on the calibration sheet. Similarly the C factor of the
polynomial can be adjusted by the amount of the zero load change multiplied by the linear gage
factor to convert it into the corresponding load change.