A DESCRIPTION OF THE LOGGING SOFTWARE
Cambridge Insitu manufacture a number of different types of pressuremeter, but the system by which their outputs are captured, stored and displayed is common to all. The system is made up of a suite of programs.
Our pressuremeters contain strain gauged transducers, the outputs of which are analogue voltages describing the displacements of the membrane and the stress being applied to the soil. These voltages are amplified, multiplexed, converted to digital code and eventually are output as an RS232 signal of ASCII encoded numbers for connecting to a computer serial port. The programs run under MSDOS (version 3.2 or later) on an IBM PC or similar. They will run on computers that use the simple 8088 microprocessor, but for best results the computer should use an 80836SX microprocessor or better. We ourselves use 386 based laptop computers on site.
The programs obtain their input either from the serial port of the computer or from data already stored on disc. The main program is called LOGGER; the remainder are utility programs. LOGGER collects the data coming in on the serial port, stores it, converts the readings to engineering units and displays the data as raw numbers, calibrated numbers or as a graph such as pressure versus displacement in real time. It writes a text file of calibrated data that can be input to a spreadsheet or our own analysis program. Once stored, tests can be replayed at any time, calibration factors can be amended and a range of graphs plotted.
Each program can be used independently of the other programs in the suite; however it is more convenient to call them from a menu, and the program called LOGMENU provides such a menu. Because most users will also have the analysis program called INSITU there is an option on the menu for this purpose. Working instructions for INSITU are issued separately.
Most of the time the user will be using the logging programs to collect and store data from real tests. Occasionally it is convenient to look at the output of the pressuremeter directly, that is as lines of readings in volts. This is the purpose of the program called DISPLAY . It is particularly useful for checking an instrument is working correctly and as an aid when calibrating the transducers of the instrument.
EDITREC is a program for modifying various settings that the LOGGER program requires. For example it is possible to change the serial port that the program is looking at, or change the disk drive that the program uses to write data.
How the logging software works
The logging software is called either by typing LOGGER or by selecting
the first option from the menu offered by LOGMENU.
On being called, the program searches for a file called SETUP.REC. If found, it uses the information in the file to set up various fundamental options. These include
a) Deciding what pressuremeter to use.
b) Identifying the drive and directory to use for storing data
c) Selecting the serial port for incoming data
There are other options in the file but these are the most important.
If the file SETUP.REC cannot be found then the program prompts the user for the type of pressuremeter being used. The user selects from the following table:
SELECT INSTRUMENT TYPE
2. Digital 6 Arm Self Boring Pressuremeter
3. Analogue 3 Arm Self Boring Pressuremeter with MDCU
4. Analogue High Pressure Dilatometer with MDCU
5. Digital High Pressure Dilatometer with compass
6. Cone Pressuremeter (via special A-to-D box)
7. Digital 3 Arm Rock Self Boring Pressuremeter
Pressing the appropriate number will select the instrument, and the program will then write a SETUP.REC file with default settings for all the other options, and will then load.
If the SETUP.REC file is found the user never sees the SELECT INSTRUMENT TYPE menu described above. Users with more than one type of pressuremeter can force the program to ask for the instrument by deleting the SETUP.REC file. Again, this is an option offered by LOGMENU.
LOGGER needs to read and write to the data disk immediately. If it cannot, possibly because there is no disk in the nominated disk drive, then the program will prompt the user and wait.
Assuming that it can read and write to the nominated data disk drive then the program will complete its loading, and the user will be presented with the first screen. The program is then menu driven, the result being a graphical display of the pressuremeter test in real time. Scaling is automatic and a range of graphs are available.
The settings chosen to run the test can all be modified subsequently by replaying the test.
The program is arranged to store data from the moment the program starts until it is quit. The only way of stopping this is to stop the instrument itself transmitting data (by switching it off or by disconnecting the serial link). Data is placed in the volume files even if the test is not started. This can be irritating, but is failsafe. It follows that the program is not used to monitor the output of the pressuremeter except when necessary to carry out a test. Monitoring is better done using the utility program DISPLAY. The data are stored several times in different forms to safeguard against the loss or disk corruption.
How the files are arranged
The arrangement of files that the logging program writes is not immediately obvious. The first file it writes to disk is a very short one called VOLUME.ID. This file merely holds a number that the program uses to ensure that before subsequent reads and writes the disk has not been changed (potentially disastrous).
The logging program then looks for three files, VOLUME.IDX, VOLUME.A and VOLUME.B. If it cannot find them (for example if the disk is a new one) then it makes these files.
VOLUME.A and VOLUME.B are text files where the incoming data are written. The program writes a few lines to these files in an overlapping manner so that in the event of a system failure one of the files is certain to be properly closed, restricting the amount of data that can be lost. If all is well, the two files are exact copies of each other.
VOLUME.IDX is a special library file which the program uses to keep a note of any bad lines of data.
When making a test, information special to that test is stored in a file of the form B*T*.TST (where * stands for numbers entered by the user as references to a borehole and test). This file does not contain the actual lines of data that constitute the test. These are contained only in the VOLUME.A and VOLUME.B files. The *.TST files merely contain a reference to the start and finish points of where in the volume files the test data is stored. It follows that when making backup copies of tests, all the files on the disk must be kept together.
The program can make a file of the test data in the form of engineering units called B*T*.PRN but it does not use the file itself. When the program wants to display test data as engineering units it calculates the values every time using data in the VOLUME.A file and applying calibration information stored in the B*T*.TST file.
The *.PRN file is only made after a test is completed. The file is a text file, consisting only of ASCII characters. It can be read by a number of spreadsheet programs, by a text editor or by the Cambridge Insitu analysis package. In general all the corrections necessary to convert the instrument output in volts to a description of the stresses and displacements occurring on the outside of the membrane where it bears against the soil are applied.
It happens that the program makes an additional file called B*T*.RAW after completing a test. This file does contain all the lines of data that constitute a test, but is not at the present time used by the program. However it is a useful backup in the event of corruption to the volume files; should this happen, the *.RAW file could be renamed VOLUME.A and the program would accept it.
The disk on which data is being written must not be removed from the disk drive whilst the test is running or substituted for another. The program (via the VOLUME.ID file) normally can detect when this has happened but there are certain operations where the check is not carried out.
Written by R.Whittle October 1999