CHUM. Cross Hole Ultrasonic Monitor. User Manual. All rights reserved Piletest.com

Transcription

1 CHUM Cross Hole Ultrasonic Monitor User Manual All rights reserved Piletest.com

2 Table of contents Top Installation Getting familiar Getting started Common Tasks Options Reporting Advanced options Behind the scenes Tomography 3D tomography Appendices A. Frequently asked questions B. Troubleshooting C. Definitions 1. Installation Windows XP/2000/2003/Vista: The user you log into the system must be a member of the administrators group. After you have completed the setup, the software is available for execution to all normal users.

3 Note: Before you can start using CHUM, you must fully charge the internal battery using the provided AC adapter until the orange LED is dimmed. Minimal computer requirements: Pentium III 500 MHz Windows 2000/XP Recommended For intensive field work: 900MHz, Tablet PC or touch-screen with outdoor visibility Software Setup Insert the provided installation CD into the CD drive of your computer. The CHUM setup will start automatically. Follow the setup defaults to install CHUM (Usually <1 minute). In case the setup does not start automatically, double click the SETUP icon on the CD root folder. CHUM software is installed to [Start]-[Programs]-[Pile Testing] Hardware drivers setup You only need to install the hardware drivers on the computer that is used for testing. You may, however install it to the office PC, and connect the CHUM system to it. Plug the CHUM USB connector into the computer The green LED on the CHUM panel will blink ten times and then stay on. On your computer screen you will see the "found new hardware" balloon, and the "Found new hardware" Wizard should open automatically Follow the wizard s instructions: Insert the CHUM CD to the CD drive of your computer and press [next] The Wizard will start copying the driver files from the CD On Windows XP, a security warning appears. You may safely press [Continue Anyway] CHUM s driver is fully tested and will not impair your system s stability! After the files are copied, close the wizard by clicking [Finish] 2. GETTING FAMILIAR The CHUM system consists of the following components:

4 CHUM instrument A pair of depth meters pulleys Two depth cables: A simple ("I") type for normal CSL use and a double ("Y") cable for tomography Two ultrasonic transceivers (each is emitter and receiver) Computer (Not provided) AC power adapter 12 V DC car battery adapter (Optional) CHUM instrument contains the following inside the rugged housing : Battery: The CHUM uses a Lithium-Ion battery, 11.1V 4.4 Ah that offers high capacity and improved performance enabling up to two days of typical use. The battery has almost no "memory" problems, and does not need to be discharged periodically. The battery has almost no loss of power while in storage, and requires no maintenance. A microcontroller, data conditioning, processing and acquisition circuits. CHUM connector panel. The respective connectors and controls on this panel are from top to bottom: 1. Power (from either AC or car battery source) 2. USB When plugging the USB connector to a computer, the system is turned on. The USB cable can be pushed in or pulled out in order to best fit the computer's USB port location. 3. Green LED ON to indicate all internal power supplies are OK Blinks 10 times at 2Hz when the unit is started (USB plugged into a computer) indicating the firmware is OK A faint blink every ~3sec when the system enters suspend mode (after 30 seconds of inactivity) the suspend mode saves battery life and the system resumes automatically when needed. 4. Orange LED A short blink for each transmitted pulse ON to indicate charging circuit is OK

5 5. Emitter socket 6. Receiver socket 7. Depth meter socket Hardware features Blinks to indicate intermittent charging (empty battery) Lights bright to indicate continuous charging Dims down when battery is fully charged 1. Automatic gain control: CHUM has 8 gain levels, from X1 (5v), to X256 (~20mv). The gain is adjusted automatically to acquire the maximal number of bits without saturating the A/D acquisition card 2. Excellent signal to noise ratio: 3. Low power consumption 4. Automatic Suspend mode Depth meters: Both contain a bi-directional rotary encoder. The encoder generates an exact number of pulses per revolution and should be self-calibrated. Two Transducers and Cables: The dual-purpose transducers contain piezo-electric ceramic elements that generate signals of 50 khz (nominal frequency). Each transducer can function as either emitter or receiver, depending on the socket it is plugged into Each transducer is attached via a pressure-proof connector to 50 meter polyurethanecoated cables wound on a reel. The cables may also be ordered in lengths of 100 or 150 meter. Other lengths are available upon request. 3. Getting Started Switch on the computer, and wait for the Windows main screen to appear. Make sure that the emitter, receiver and depth cables are connected to the panel. Plug the USB cable into the computer. Start CHUM software. The CHUM main screen will appear

6 Click [FILE] and then [NEW PROJECT] buttons (or the blue "Start a new project" link, or the left icon) link, then choose the name (or number) of a new project. For this lesson you may use the name "TEST". Optionally enter any suitable text on the TITLE and SUBTITLE lines. Important: Before testing your first pile, you must calibrate the depth meter See Depth Calibration Press the big plus [+] button to start testing a pile 3.1. Testing From CHUM's main screen, Press the big plus [+] button to activate the pile screen

7 The test wizard consists of several stages: Pile details and geometry Define_tube_distances Leveling the transducers Start pulling Logging Analysis To move between the test phases, click the [Next] and [Back] buttons. If a button is grayed out, CHUM needs more information to move ahead, or does not expect the button to be pressed at this stage The pile name/number is displayed on the window caption Pile details and geometry

8 In this screen (Above), enter information about the whole pile (not specific to profile): the pile name, sub-site, and diameter, access tube layout and level. You may also enter pile specific notes (e.g. "pile top needs trimming") The pile (or element) name may be up to 20 characters long and may contain any character Sub-sites are a convenient way of breaking up a big project into smaller units (Read more ). Select an existing sub-site, or create a new one by pressing the [New ] button [Save] exits the wizard and saves all changes and all details. Changes are also saved automatically after each log is taken [Save compact] Only arrival time and energy are recorded in the files. Files are tiny but no FAT picking can be done after they are saved, and no waterfall presentation can be done (The ASTM standard may require it) [More ] Opens advanced pile options Cut all Cuts the all profiles top (tube-stick-up) so that profiles will have the same length Report prints a report for this pile only. Which is convenient for on-site printing. To produce a detailed report, see Reporting Edit tube distances see Define tube distances Consult with piletest.com Send this for getting a second opinion from Piletest.com 3DT Perform three-dimensional tomography (Optional CHUM component) The drop-down menu above the pile scheme enables you to choose one of many available schemes you may be going to test. You may choose among:

9 A pile with up to 4 access tubes named N, S, E and W respectively A pile with 2 to 15 access tubes numbered 1...N clockwise (1 is Northernmost by convention) A straight diaphragm wall element (barrette) with up to 6 access tubes A diaphragm wall element with 6X2 tube arrangement A pile with a single access tube (AKA single hole test) After you have selected the scheme, you can rotate using the [ ] and [ ] icons to the correct azimuth. This only effects the reporting and has no other meaning. There are several ways to specify the profiles you want to test: Fastest: Click on the [ALL] icon to define all combinations. After you have finished testing you can use the [Clear Untested] button to remove the redundant combinations only. Coolest: Using the stylus, draw a line between the tubes (You can specify more then one profile in one broken line). To specify single-hole tests, draw a circle around one of the tubes. This input is very convenient when using a stylus (as opposed to a mouse) Advanced: Click on the small [+] button below the profiles list to show the new profile screen You may enter any name, but only names made of two tube letters/numbers will appear in the scheme graphics. Other profile names will be reported, but cannot be assigned to tubes. Please note that the profile letters/numbers have a significant meaning when performing Tomography Tested profiles are painted with a thick, red line, this way you can always see which profiles are still untested. To delete a profile, either: Select it from the list and press [X] Re-draw over it

10 Press [Clear untested] (if the profile was net yet tested) If the profile was tested, you must confirm the delete TIP: Under normal conditions, piles should be tested at a concrete age of around seven days. Testing should not be performed earlier than 2 days after concrete placement, as the curing process may be slower in certain areas. Also, tube de-bonding (separation of tube from concrete, in plastic tubes) becomes more likely as time after concrete placement increases. For this reason, interpreting the data may become much more difficult if testing is performed, about two weeks or more, after concrete placement. TIP: For CSL testing, the pile to be tested has at least two embedded access tubes. The diameter of the tubes should exceed the transducer diameter by at least 10 mm, and they should be checked for clearance throughout their length. The tubes should have been filled with clean water either prior to or soon after concrete placement. Place the depth meters on top of two of these tubes, and lower both emitter and receiver all the way to the bottom. Click on the profile to be tested (or highlight it and press [Next]) and the leveling screen will appear. TIP: It is possible to press [Next] before lowering the cables, this will let you monitor the approximated tube length 3.3. Define tube distances The ASTM standard requires all tube distances to be reported. For tomography, tube distances MUST be entered To enter tube distances, select "tube distances: [EDIT]" and enter the center-to-center distance for each tube pair

11 When enough tube pairs have been defined, you may use [Auto complete] to automatically calculate the position or all the other tube pairs. This is useful for a large number of tubes For example, in a four-tube pile, 5 of the 6 combinations must be entered to calculate the sixth, but for a 10-tube pile, only 18 of the 45 possible combinations must be entered, and the rest can be calculated. TIP: it is advisable to add some redundant measurements for cross-check. The algorithm will report the maximal error in this case.

12 3.4. Leveling Testing is done by pulling from bottom to top. Before you start pulling, you should make sure that both transducers are at the initial test position: At the same level, and as close to the bottom of the access tubes as possible. On this screen you can see the following items, all designed to help you bring the transducers to the initial test position. To the left, a black strip that scrolls from the bottom to the top of the screen, showing the relative signal strength (thickness of the strip) and the first arrival time (FAT).

13 A thin line on the left is also indicating the FAT, but is heavily filtered and reacts slower to changes. The combination of the fast and slow moving indicators gives a clear sense of the direction of change. On the lower right hand side an oscilloscope windows, showing the signal shape, strength (gain) and automatically picked FAT (triangle). TIP: Double-clicking the oscilloscope window maximizes and restores it. The window can also be moved and repositioned A distance meter, giving the approximate distance between the transducers (in feet or meters). If you intend to run tomography on this profile, check the [ ] Do Tomography. You must enter the tube distance for tomography. Select the sample duration (500, 1000, 1500 or 2000 µs) which best fits the current profile distance. If FAT is not picked up correctly, click the [FAT options] button to show the FAT options page Click the [Options] button to change general options (Sample spacing, unit system and profile compression) using the options page If you are lowering the transducers while in this screen, depth from top (H) is displayed, click on the depth label to reset it. Using the controls While in the leveling screen, move any of the transducers up or down, in order to bring them to the same level. This is accomplished by observing the following indicators: The black strip being closest to the left hand side of the screen and also the widest. The signal on the Scope window is strongest (low gain). Distance on the meter is minimal (This should be roughly equal to the measured distance between the tubes). TIP: Always level the transducers in a clean concrete zone. If the pile has a soft bottom, perform the leveling a few feet (~1m) above the bottom, and only than lower both cables carefully together to the bottom before continuing

14 Having satisfied the leveling criteria, click [Next], and the pulling prompt will appear Start Pulling Do not use the [Next] button, CHUM will move to the next test phase when you will start pulling TIP: Select the "use sounds and voices" option to have the "pull" command spoken to you using the default system voice. See Options At this stage start pulling both cables together, hand after hand, as smoothly as possible until both transducers reach the top. The rate of pulling should be no more then 40 samples per second, if the vertical spacing is set to 5cm, this will be up to 2.0m/s (which is very fast) 3.6. Logging

15 The data collection screen is very similar to the leveling screen except for a vertical depth axis along the signal strip on the left side of the screen, as well as no distance (between transducers) presented in the meter area. The plot will scroll upward as the cables are pulled up. In suspect areas of concrete, The transducers can be raised and lowered in suspect zones to increase the number of data points taken. The plot of the signal strength will scroll along up and down as the cables are pulled up and down. Suspect zones will be noticeable when the strip is thinner and/or when the left edge of the strip is further away from the depth axis (increase in FAT). When you have finished pulling, click the [Next] button, The profile will be automatically saved, and the Analysis screen displayed When tomography is performed, this screen has some additional controls, and pulling is performed in a special way. See Tomography

16 3.7. Analysis At this stage, the results of the test are displayed on the screen as shown below. There a few options for this display, and the "Lines Plot" mode is the most common. In this form, the first arrival time (FAT) and the attenuation registered at the receiver are shown as a function of depth. A local increase of FAT or attenuation may indicate an anomaly.

17 Controls description from top to bottom The profile name (34 above) can be changed Profile specific notes may be entered (e.g. anomaly at 3.5m) [Presentation] Change the presentation mode

18 [<] and [>] Move to next/previous profile [<<] and [>>] Move to next/previous pile [+] and [-] zoom in and out [ Filter=n ] FAT and Energy filtering (0 = no filter, 5=heavy filtering) (Read about filters) [ More ] show advanced options menu: Save as CSV (Excel data format) - Compact or detailed The profile is saved in CSV format (Plain text format with comma delimiter) and Excel (or any other associated application) is launched Detailed file contains full waveform information. Read about using this feature Cut to length - enter a length to which to cut from the top of this pile. Useful when the access tubes are sticking out of the concrete in different levels. Wave Speed Calculator A simple test method for accurately measuring the wave speed read more [ ] remove the top or bottom of the profile log Erasing the top of the log is sometimes used if the access tube are raised much higher then the pile top surface Erasing the log bottom may be used if the cables were pulled with too much slack, but should usually be avoided. [FAT] start the FAT utility form to analyze the data more thoroughly, [Clear] Delete the data you just collected (Warning: You will have to redo the test!) 4. Common Tasks 4.1. Start a new project From the main menu, select [File]-[New project] Verify that the "Home folder" points to the correct path, or click [Home Folder] to change it Enter the project's name or number and click [OK] CHUM will create a new folder under the home folder, and will create the file PISA.PROJECT

19 TIP: Use project numbers, rather than names. This makes finding and managing many projects simpler 4.2. Open an existing project If the file was created or opened recently, you will find it in the "most recently used" list under the "file" menu From the main menu, select [File]-[Open project] Verify that the "Home folder" points to the correct path, or click [Home Folder] to change it Select the project and click [OK] 4.3. Save a project There is no option or need to save a project, all changes are saved immediately Transferring project files to another computer On the first (or only) visit to the site, connect the computers, and copy the whole project's folder to the target (This can usually be done by dragging the project's folder) On future visits, sort the source folder by date, and only copy the latest added files to the target computer Send piles over If the files are small, you may send them as an attachment. Compressing the files with WinZip or a similar program is recommended. Note that many mailboxes cannot accept large attachments. The simplest and best way to send a pile to piletest is using the Consulting Wizard 4.6. Change the location where projects are saved From the open, or new project pages, select [Home folder], and select the folder where you want your projects to be stored and where the Open Project page looks for existing projects.

20 See also: Projects, Sub-sites and files 5. Options The options page [Tools]-[Options] consists of two tab sheets: Logger and General 5.1. Logger options

21 [Vertical spacing] is the spacing between samples. A typical general-purpose value is 5cm (2"). A smaller spacing will give a finer coverage of the cross profile, but will produce larger files and will limit the pulling velocity (limited to 40 samples per second). [Default Filter] defines the initial filter value for new piles. The recommended value is 1 [FAT options] opens the FAT options dialog [Signal classification] opens the signal classification dialog, where you can assign one of three categories to a pulse based on velocity and attenuation changes 5.2. General options

22 Units: Select if the system is using Metric (m) or English (Feet and Inch) units Software keyboard: Check this option to have automatic screen keyboard popup when an alpha-numeric input is required. This option is only useful for keyboard-less computers Unit Code: Changes the first character prefix of new generated pile files (e.g. the 'A' in A _2.PILE) If you have more than one CHUM unit, assign each unit a unique code ('A', 'B', etc.) to avoid name collisions which might otherwise happen if the same project is being tested by several CHUM units. (See also Projects, Sub-sites and files) Use Sounds and Voices: Disable or enable voice instructions. CHUM is using the default Windows voice, accessible via [Control panel]-[speech]-[text to speech]. You can control the voice personality and speed Depth Calibration CHUM is using a digital depth meter producing an exact number of pulses per pulley rotation. Depth calibration is determining the number of those pulses per length pulled. This number is internally stored in the test computer. You should perform depth calibration: Once before starting to test your first pile When starting to use a new computer Once a year Whenever there is a doubt regarding the depth reading To perform depth calibration, simply select [Tools]-[Depth calibration] and follow the wizard, Verify repeatability by re-running the wizard. For your reference, here is an extract from ASTM (Standard Test Method for Integrity Testing of Concrete Deep Foundations by Ultrasonic Crosshole Testing) Transducer Depth-Measuring Device... The depth-measuring device shall be

23 accurate to within 1 % of the access duct length, or 0.25 m, whichever is larger 5.4. Signal classification options From the Options Logger tab, select signal classification, to open this page The signal classifications are used to color-code a pulse based on velocity and attenuation changes. This is also used for real-time and fuzzy-logic tomography, control the values that classify a pulse into one of three color-marked categories. Those can be called "good", "questionable" and "poor", but they are really just two logic criteria options, as depicted below

24 You can control the area of the three zones, and change the logic operator ("AND" / "OR") as well as to ignore attenuation altogether, and use just velocities FAT Picking options The FAT (First Arrival Time) picking options can be accessed from the main menu (Tools- Options), from the leveling page, or from the FAT utility Four options for picking the FAT are available, each method uses different settings: Piletest highly recommends the use of the "Automatic" algorithm. Dynamic threshold: A Piletest.com algorithm which uses a threshold level based on the pulse amplitude Minimum Time: You can set this limit so that the signal picking routine will not look for a shorter FAT. For example, a delay of 200 µsec will cause all FATs to be 200µsec or higher. Please note that this time should be sufficiently lower than the minimum wave travel time between the tubes for normal quality concrete. Example: 250µsec for tubes 3 feet apart and a typical 12,000 ft/sec concrete wave velocity (t=l/c). The minimum time is represented as the horizontal position of a small red triangle. Threshold Ratio: This is a limit that is referenced to the pulse amplitude. In general, a higher ratio means that the program will pick FATs sooner, but is also more sensitive to noise. The signal is typically strongest in the earlier portion of the record after the first few peaks. A ratio, for example, of 10 means that the first occurrence of a signal amplitude 1/10 of the maximum is defined as the FAT. Typical value is around 10 to 20, but under poor test conditions, lower values may be required. Minimum Level: This is an absolute limit that does not depend on the background noise level. If the threshold is lower than say 10 mv, then anything lower before the first 10mv peak will be ignored. This option is rarely needed and may usually kept 0, it comes into effect in very week signals and is mainly kept for backward compatibility Fixed threshold: Old-Fashioned picking, threshold value is fixed. Automatic: A Piletest.com proprietary algorithm, no additional settings are needed.

25 Read more here for algorithm explanation. STA/LTA Stands for "Short Term Average / Long Term Average" Description: A window of defined size is moved along the time axis and measures the average of samples in this window. Another window of larger size follows the first one FAT is defined as the earliest point where the ratio between the two averages exceeds a user-defined ratio Typical values are: Window size: 6 Ratio: 1.6 ref: Caltech Earthquake Detection and Recording (CEDAR) system [Johnson, 1979]. To learn more - Enter "STA/LTA" to your web search engine Presentation The presentation window is accessible from the Analysis page, or from the Report page.

26 You can select the presentation mode among the currently available methods: Line Plot, with a combination of the following curves o FAT [micro seconds] o A colored flag at FAT+20% o Either Relative energy [unit-less] or Attenuation [db] - but not both o Apparent velocity [m/s of feet/s] o washed-out Waterfall background (Note the ASTM standard requires that waterfall presentation will be added if any filtering is used) Dual Waterfall Fuzzy logic tomography (*) Parametric tomography (*) * - Only if the data was collected with diagonal readings using two depth encoders (Tomography)

27 Color - select the color for the presentation. When selecting a color, consider your final report capabilities: Can your printer print in color or in gray shade Is the report going to be faxed? most faxes are B&W only 6. Reporting The reporting options are divided between three tab sheets Report contents Filtering Layout options Once you click [OK], a report file called report1.rtf is generated and your assigned word processor is started on it. You can then edit the report, merge it with other documents, e- mail it, etc. See also: [Page setup dialog] [Report style] Check [X] Project Totals to produce a table specifying the number of piles and profiles and the total length of piles and profiles for each sub-site

28 Check [X] Project Summary to produce a table stating the measured length of each profile, and related comments recorded with it. [ ] Site Plan is not yet implemented Check [X] Detailed report to produce a table for each pile, including graphs and all recorded details, the [Options] tab is only available if this option is checked You may select to report all piles, piles from the latest month or from the latest visit. If a pile is selected in the main page, a fourth option is displayed for reporting only the selected pile Check [X] Include non-tested piles will include piles which have no recorded logs in the final report.

29 This tab is only available if [X] Detailed report is checked Vertical scale can be set to a specific value, or set to [Automatic] to best fit the page height In "Automatic", the scale is calculated to fit the longest profile in the report to one column. Smaller scales may cause some profiles to be printed over several columns. Number of columns (one graph per column) can be set. It is recommended to switch to landscape print orientation when using more then 3 or 4 columns Presentation mode can be selected for the whole report "Distortion" is only enabled for a tomography report, and will improve the visual presentation of piles by distorting their width relative to height. TIP: You cannot mix several presentation modes, scales, etc in one report. You may, however, achieve this by cutting and pasting two separate reports 6.1. Controlling the report style From the main menu, select [Tools] [Style] to show the style screen Select an item from the list on the left hand side and change its appearance using the controls on the right side. [x] Use: Include or exclude items from your report. [Default]: Restore the style for all items to the factory settings [Close]: Done. Try your new style by producing a report

30 TIP: You may select several items from the list on the left by pressing the [shift] or [control] keys while pointing on items. This way you can change the appearance of several items together. TIP: underscores (_) are converted to spaces. Append a few underscores to your prompts to unsure proper tabulation of titles labels and values. TIP: Avoid using too many font styles in your report; a good-looking report usually uses two to three font styles at the most Page setup From the main menu, select [File] [Page Setup] to show the following screen Your currently selected printer name is displayed on top, and you can control the margins size on each side of the report page (mm) 7.Advanced options 7.1. Enable/Disable features From CHUM's main menu select [Tools]-[Enable/Disable features] to activate the features manager.

31 The checklist represents the enabled features. By default, all features are enabled, and you may choose to disable some for one or more of the following reasons: To simplify the use of CHUM by hiding features you never use. To limit the amount of control you give the field technician. To prevent possible operation errors in the field To enforce a specific usage pattern on your team. If you wish to protect (to some degree) the changes, you may select a password. After you select a non-empty password, you will be requested to re-enter it each time you start the feature manager. Tip: If you forgot your password - look it up in the file CHUM.INI, stored under your user s "application data" folder Note: There is no way to really protect this password from malicious users. Tip: To remove the password, set it to empty First Arrival Time (FAT) Picking From the Analysis screen, press [FAT] This dialog is split into two panes: Left pane: the FAT vs. depth plot. Right pane: the actual signals.

32 Note that the vertical position is referenced from the bottom of the tubes rather than the top of the tubes. This is because the data is collected from the bottom up. When you leave the FAT signal picking screen, the software re-plots the depth from the top. The vertical separator between the panes can be dragged right or left The horizontal separator between the signals on the right pane can be dragged up and down In some cases, the FAT will not be selected correctly by the software. This tool is powerful for manually adjusting the automated FAT signal picking routine performed by the software. There are two major methods available for picking the FAT signals by yourself; manually or automatic. Manually: Click on the signal trace on the right hand side of the screen. A highlighted signal will be shaded. Please note that a small circle will appear on the left side highlighting where the particular signal is in the plot. Remember that in this screen, vertical position is from the bottom. You can directly pick the FAT in the signal by moving the mouse to the point where you think it should be and then click the left mouse button. The FAT will be delineated by the two diagonals intersecting at the FAT on the horizontal (time) axis. The time of the FAT (in Microseconds) will be shown in the data area. The highlighted circle in the depth plot on the left will move accordingly depending

33 on where you selected the FAT on the signal plot on the right. Additional data such the depth (in meters/inches) and scale are also shown in this area. Automatically: This option is very powerful as it can change the FAT for all applicable signals along the pile. Click on [Calc] to enter the automatic FAT picking menu. Once you have selected the values, click on [Close]. The program will then recalculate all of the FATs for the entire profile. You will see a change in the plot of the individual FATs. If you agree with the FATs, than click accept. Click [Cancel] to go back and ignore any changes. The last button in the FAT signal picking screen is the [Actions ] button. Use this to: Change the view of the signals, You may view the raw data and/or the filtered data and/or the signal envelope Delete / undelete any signals that you have highlighted on the right. Restore all previously deleted signals (as long as you have not left the FAT signal picking screen). Copy the highlighted pulse shape to the Windows clipboard. You may than paste it into your report 7.3. Consulting with Piletest.com From the Pile details page, select [More ] and [Consult] Piletest.com offers limited consulting services in order to: Help novice testers with analysis Help advanced users with difficult cases Collect interesting test cases The "Consulting Wizard" opens

34 The consulting wizard lets you select the profiles of the pile that will be compressed. The tradeoff is yours: Compressed profiles contain no waveform data and no FAT picking can be done, but the profiles are very small in size and can be transferred quickly Non-Compressed profiles contain full waveform data and Piletest.com can observe and modify the FAT pickings. But such profiles are much larger and take much longer to transfer. If the whole file size is below 20Mb, and you do not have any bandwidth limitations, send the file uncompressed The wizard provides an estimate of the file size and the expected transfer time. Files will first be transferred to Piletest.com's web storage (by FTP). Next your default editor will be opened with a message to consult@piletest.com - please provide clear information about the consulting you require Data Sources CHUM can accept data from several sources. This allows the same software to run in the field and on the desktop computer for training and demonstration purposes. From the main menu, select [tools] [Data source] (or the wrench icon) to show the following screen

35 Data sources that are built-into CHUM are Name USB Demo Replay Description Connects to the CHUM hardware via USB, in order to test piles. In the field CHUM should only select this data source Uses simulated pulses with one simple defect Replays a stored profile from any pile (Advanced) The currently selected data source is displayed in the window title. To change a data source - select one from the drop-down list. To test the currently connected data source, press [Test], the dialog will expand to show the data source diagnostics In this mode, CHUM triggers the emitter at 10Hz and displays the following parameters: The data source description Tomography supported / not supported

36 Hardware status and error codes Depth encoders raw counters readings for both counters (ignore the second reading if you only have one depth encoder) Sample rate (Should be 500 KHz) Battery voltage. The automatic gain picked for the displayed signal Attenuation for the received pulses Press the [Stop] button to exit the self-diagnostics mode 7.5. Exporting data You may export profiles stored in CHUM internal format to a CSV (Comma delimited) format. CSV is a text format recognized by most spreadsheets (such as Microsoft Excel). To generate a CSV file, start CHUM and select the pile and profile you want to export, then click [More] and select either a detailed, or a compact file formats Compact CSV contains the profile details, and the individual pulses FAT and Energy Detailed CSV contains, in addition to the compact data, each of the pulses shape. CHUM now asks you for a filename and location to store the CSV file, the default file name is <pilename>.<profile name>.csv, for example if you save profile NS of pile 3-1, the default filename suggested is 3-1.NS.CSV: Once you saved the file, CHUM will try to launch the associated application (such as Excel), if this does not work, start your spreadsheet application and open the saved CSV file.

37 NOTE: If you get a "file not loaded completely" message, simply ignore it. Cells description A2..F2 Project name, Pile name, Profile name, Filter, tubes distance (cm), profile notes Row 3 - Samples headings: Column Title Description Units A D1 Master (or single) depth encoder readings cm from bottom B D2 Secondary depth encoder readings (identical to D1 if not tomography was done) cm from bottom C T0 FAT microseconds D Energy Total energy of the pulse mv. µs E SampleRate A/D sampling rate (samples / second) Hz F Gain AGC (Automatic Gain Control) value * - G Number of samples how many samples are taken for one pulse H... Value(n) Pulse shape data, 12-bit value ( )* * To convert the pulse value to volts, use the following formula Volts = Value * 5.0 / (2048 * gain) Samples Bits

38 Generating a curves chart (Specific to MS Excel) select B4..D4 Press CTRL+Shift+Down to select 3 columns of data Press the chart wizard icon Select X-Y scatter without markers (Optionally) swap X and Y axis and add titles and labels press [Finish] Extracting a single pulse shape chart (Specific to MS Excel ) Only available if a detailed CSV file was saved select Hn (where n stands any row, for example H4) Press CTRL+Shift+Right to select the whole row Press the chart wizard icon Select X-Y scatter without markers (Optional) specify titles and labels press [Finish]

39 7.6. Wave speed Calculator The velocity calculator uses the standard test wizard, but testing is done as follows: Lower both transducers together to exactly the same level. This level should be in the middle of the pile, at the bottom of a 4-5m (10-15') zone of uniform concrete. This is the tested sample. Define a profile in the pile, and press [Next] to start testing it, just like in a standard test Set sample size to maximal (2000us) Leave the passive transducer stationary, and raise only the transducer connected to the depth pulley. Watch the scope window and keep pulling until the signal becomes too weak and FAT picking is no longer possible Press [Next] to show the profile, this should show something similar to this:

40 Use the [ ] (scissors) icon to trim the top or bottom part of the profile, keeping only the part where the FAT pickings are reliable. Enter the exact distance between the test tubes. Select [ More ]-[Wave Velocity Calculator] The velocity and R 2 (coefficient of determination) are replacing the profile notes, e.g. "Velocity=4200m/s, R^2=0.998" Internally, CHUM is using a linear regression on the distance vs. FAT vector; the slope of the trend line is the wave speed. Advantages over laboratory test of a concrete cylinder sample: Intuitive Large sample size: on samples close in size to the wavelength, the wave speed is different than the one in a real pile Averaging many samples, eliminating intermittent noises Affordable: No special equipment or test method is needed Independent of constant delays Good indication of the test reliability (R 2 ) 8. Behind the scenes 8.1. Projects, Sub-sites and files Background CHUM uses one dedicated folder to store all projects, this folder is called "Home folder", by default this is "C:\Pile testing" Projects are stored as sub folders of the home folder Each pile (All profiles) is stored as a separate file (with the extension.pile) under the project folder. A project folder also contains a file called "CHUM.PROJECT", which is a text file storing the project's titles and other miscellaneous settings.

41 A pile's file is consisted of a unit code, a system-wide unique number, the digested pile's name and the extension ".pile". For example, the pile 123** will be stored under B PILE B is the unit code (Only relevant for users with more than one CHUM unit) means this is the 59th pile stored on this system 123 is the pile's name, with underscore (_) replacing characters that cannot be used in a filename.pile is the file extension This naming scheme makes sure that no two piles will ever get the same file name, and no test work will get overridden and lost For example: C:\PISA_PROJECTS 1003 <-- Home folder <-- A project A00059.A100.PILE <-- A pile A00060.A101.PILE <-- Another pile CHUM.PROJECT <-- Project's settings : 1004 <-- Another project : Sub-Sites Sub-sites are a way to "break up" a project into smaller parts, which are easier to manage and can produce a clearer report. For example: if the project contains several buildings, each building can be used as a sub-site. Piles on different buildings can have the same names. When a project is opened, piles are grouped by the sub-site stored within each pile. When the last pile of a sub-site is deleted or moved to a different sub-site the sub-site is removed. Sub-sites are never empty.

42 A new pile in a new project is created in the "default" sub-site. This name will not appear in the report. If you create a new pile in a new sub-site, CHUM will suggest you to rename the "Default" sub-site to a more meaningful name, to avoid confusion CHUM filters Signal Filter Signal filter is a hardware band-pass filter applied to each collected signal before it reached the A/D/ Signal filter is hard-wired with no user control. The picture below shows a signal before (gray) and after (green) the filter is applied to it. It is clearly visible that the lower frequency noise at the left side of the signal was significantly reduced allowing an accurate FAT picking Profile Filters Profile filters are three different algorithms applied to the whole profiles of picked data points (FAT and Energy) before they are presented. The user can control these filters by selecting a single filter value F of 0 to 5, where 0 denotes no filter and 5 denotes maximum filtering. The three filters are: Median filter, Running averages filter and delay filter, the first two are based on the fact that the theoretical vector of measurements is convoluted by relatively large kernel (exact size is determined by the pulse shape and transducer geometry), the delay filter is based on an a-priori knowledge stating that arrival times significantly lower then the expected arrival time are impossible.

43 Median filter: Replace each value in a vector with the median value of its neighbors: A i <-- median of (A i-k, A i-k+1, A i-1, A i, A i+1,, A i+k-1, A i+k ) Where k equals the filter value F. Running averages filter Replace each value in a vector with the average of its neighbors: A i <-- (A i-k + A i-k+1 + A i-1 + A i + A i A i+k-1 + A i+k ) / (2k+1) Where k equals the filter value F. Delay filter: Find the N percentile arrival time T when N=F * 5 ("5" was found experimentally) Set all arrival times lower then T to T Examples: If filter=0, N=0, thus T is the 0 percentile (Minimum) arrival time and the filter does nothing since no time is lower then the minimal time If filter=5, N=25, thus T is the 25 percentile arrival time, which is usually the normal arrival time of the profile. The following table shows the save arrival times vector with three levels of processing: without any filter, with a low filter (F=1) and with the maximum filter (F=5): 1 - Noise removed by the median filter 2 - Noise removed by running averages filter 3 - Noise removed by the delay filter

44 F=0 F=1 F= What is "Attenuation"? Energy is the amount of energy that arrives to the receiver for each transmitted pulse. Since the transmitted energy of the signals is approximately constant, it is also expected to be approximately constant at the receiver. When a defect in the tested medium blocks the signal, it absorbs some (or all) of the transmitted energy, and a lower value is read at the receiver. Some defects (such as necking) can significantly reduce the received energy but not to change the first arrival time (FAT) since a direct path between the transmitter and the receiver still exists. Energy is calculated as the sum of the absolute voltage values along the received pulse:

45 E = sum ( V i ) Attenuation is calculated at A = 20 * Log 10 (C / E) Where: A - Attenuation [db] E - The received energy C -A constant value representing the maximal possible value of received energy Hence, Attenuation of ~6db means the received energy is about half the maximal possible value. Since the scale is logarithmic. A shift of ~6db from the normal attenuation value recorded within a profiles means the energy drops by ~50% at this point and a shift of ~12db means the energy drops to about a quarter. The Chinese standard defined a suspected anomaly as a 6db attenuation increase relative to the normal received attenuation. This assessment must be backed up by some FAT increase How does the energy change with regard to distance? In a homogeneous medium, the signal attenuation can be calculated according to the following formula: Where: E r = E t * exp(-k * x *f) E r - The received energy E t - The transmitted energy k - Attenuation factor of the medium x - Distance between the transmitter and receiver f - Signal base frequency (Hz) Sometimes, it is interesting to find out the change in distance needed to half the energy of the signal (6db drop):

46 X = ln(e r /E t )/-kf = -ln(0.5)/kf Since k and f are constants, x is also constant with a typical value (in good concrete) of 60cm (2 feet) and much higher values in water. The following graph was created by holding one sensor at the bottom, and pulling only the other sensor, hence gradually changing the distance. The data was exported to CSV, and plotted using a spreadsheet. The maximum distance logged was 4.5m (!) and the linear correlation is The slope indicates exactly 10db/m (=6db/60cm) - so the energy drops to half every 60cm. The total energy drop is 40db = 2 orders of magnitude. 9. Tomography The equipment To perform tomography, you should have two instrumented depth meters, and a suitable "Y" depth cable. Overview In CHUM, perform tomography by pulling both transducers together, normally, and when passing a suspected zone, raise and lower each of the transducers in a special way to collect diagonal readings. Once the suspected zone is "covered" from all angles, level

47 the transducers and keep pulling normally up to the top of the pile or the next suspected zone. See here a general discussion about tomography Know your left and right! It is critical to assign the correct depth meter to the correct access tube. Failing to do so will produce mirror images in 2D tomography, and will produce meaningless junk in 3D tomography! The assignment convention is by color codes. Green=Primary. Red=Secondary Older model "Y" cables are connected directly to one depth meter called "primary" and then branches to the other meter, called "Secondary". The primary depth encoder is always assigned to the first profile letter and the secondary, to the second letter. For example: When logging profile "34", place the primary encoder on tube "3" and the secondary on tube "4".

48 When logging profile "NS", place the primary encoder on tube "N" and the secondary on tube "S". If you have mistakenly logged the data in the wrong way, you should re-log the profile, or simply rename it to "43" or "SN". On the tomography plot, the primary channel is always on the left. How to collect the data? To collect good data, the following conditions must be satisfied: 1. All the profile area should be "covered" by horizontal readings (The software can extract those readings only to produce a 1D CSL plot) 2. Every pixel in the suspect zone should be covered by at least two forward (//) and two backwards diagonal (\\) readings. The more, the better. Mathematically, each pulse is a row (Linear equation) in the matrix. To get consistent results, you must get an over-determined matrix. (More pulses than pixels). The way the data is collected in the CHUM system generates a good distribution of information (matrix very-over-determined on suspect zones and under-determined in good pile zones) In order to get consistent results, the data must be collected in a consistent way, for this, a data collection procedure is defined. The procedure collects the data in "fans". As demonstrated below:

49 The right transducer is lowered in small step and for each step, a "fan" of samples is performed by the second transducer A schematic way to describe the procedure is by plotting D1 against D2 (It does not matter which of D1 and D2 is primary or secondary) In words: 1. Pull both transducers horizontally until above and out of the defective zone (Point A) 2. Keep #1 steady, and lower #2 until the signal is almost lost (You can go over +/- 45deg) 3. Keep #2 steady and lower #1 by a small increment (about 2-4 inch)

50 4. Keep #1 steady and raise #2 until the signal is almost lost, or both #1 and #2 are above the suspected zone 5. Again, Keep #2 steady and lower #1 by a small increment (about 2-4 inch) 6. go to step 2 until both #1 and #2 are below the defect zone (Point B) 7. Level #1 and #2 to horizontal (+/- 2deg) and continue pulling upwards till the top, or the next suspected zone And in an animation of the process This might look complicated at first, but it is actually simple, and once you practice it, it is very easy to use in the field, and will give you good, repetitive results each time. Take your time, and practice this several times in a non-stressing environment. While the data is logged, watch the following indications Primary encoder pulled up ( ) down ( ) or steady ( ) Secondary encoder pulled up ( ) down ( ) or steady ( ) Average length pulled from bottom (H:nnn.nn) Angle between transducers (0 is horizontal) FAT and Gain on the oscilloscope window

51 Once the data points are collected, they can be viewed and reported in all presentation modes. The data can also be presented in the one-dimensional lines plot or waterfall presentation. For this, the software filters-out the diagonal readings and uses only the horizontal ones. TIP: When doing tomography, change the vertical spacing in the options logger tab to a smaller value, such as 2-3cm (0.75" 1.0") See also: Signal classification options Types of tomography CHUM supports the following tomography methods Real time (Based on simplified fuzzy-logic) Fuzzy-logic Parametric Three Dimensional (3DT)

52 The data points for all types of tomography are collected in the same way. For 3DT, enough profiles must be collected to cover the pile s cross volume. Principle of Fuzzy-Logic (and real time) tomography In one sentence: "A concrete pixel is painted by the highest category of the rays passing through it." In detail: The profile is broken into pixels All the rays passing through a pixel are collected, and assigned a color category according to the logical classification conditions you specified. The classification counts are sorted, and the final color is the Nth order-statistics, where N=100-2*Filter (No filter N=100 = Maximum). Using order-statistics instead of just Maximum compensates noisy readings. In real-time tomography, N always=100 To speed things up, CHUM is using a variable pixel size. Initially, the whole pile is considered as one pixel. If all rays going through a pixel agree on color (or the pixel is small enough) it is painted, otherwise it is cut to two smaller pixels, and so on. Limitations of Fuzzy-Logic tomography The calculations are done on apparent velocity and attenuation. Those are an average of values along the ray path. The apparent value is therefore always higher than (or at best equal to) the value inside the poor-quality pixel. As a result, the contrast of fuzzy-logic tomography is lower than matrix-based solutions. Because of angle limitations, the plotted areas always show "ghost" shadows 3D tomography 3DT is an optional component of CHUM and is being sold separately. 3DT is the next step in tomography, and its success depends on the success of the 2D tomography phases. Do not try to "force-feed" the 3DT algorithm with poor 2D results hoping that it will somehow turn it into good 3D result.

53 Before performing 3DT, verify the following: FAT pickings have been performed and reviewed. All profile names map to the pile schema those are used to locate the profiles in the X-Y plane Profile names are using the encoders conventions (See Know your left and right) All tube distances are correct Open the Pile details page and select [More ] and [3DT] (Or run the CHUM3DT directly from the start menu) The 3DT wizard opens Follow the wizard to perform the matrix inversion (can take some minutes, or significantly more, depending on the data and computer power)

54 The calculations results are kept in a file, and in the next time, you can just view the results instead of re-calculating them. Once the calculations are done, the following screen appears: 3D Image pane: Moving, tilting and zooming the pile. Action Mouse Other Rotate the camera around the pile axis Drag Left-Right: - Move the camera AND the view point up or Drag Up-Down -