Wave_logger_processing
Contents
1. Em o 1000 0 77 0 87 1 2 6 1 mean H m acc BG 3 Bed slope 1 1000 obs Rayleigh BG m 156 1 72 1 73 Ho m 1 61 1 84 1 88 m 227 23 Average depht m 8 5 water density 1018 1500 time sec 0 0005 2000 Wave height distribution 2500 0 001 0 002 0 005 00 0 0 0 05 0 0 250 5 0 751 1 251 5 1 752 energy spectrum 2 292 E f 2 705 9 d 0 15 0 2 frequency Hz 0 25 0 3 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft Calibration The pressure sensor of the wave logger has been calibrated in a deep water pit From this calibration was the following relation found p 254V 43250 Pa 16 In this formula a correction is included for the atmospheric pressure Variations in the atmospheric pressure have influence on the results but this effect 1s very small V is the reading from the wavelogger The wavelogger has been applied also during one of the Bardex experiments at the large wave flume of Deltares Netherlands the DeltafIume In this flume prototype size w2. 1 f a oum EP GP e e mw lt lt lt i m SSS The measurements by EMS wave logger and Deltares Equipment were not at exactly the same location n Wave surface elevation measured with Deltares equipment 19 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd UDelft Technische Universiteit Delft Appendix 3 Matlab scripts Wavelogger m Script for processing pressure data from the EMS wavelogger Script developed by H J Verhagen Delft University of Technology February 2013 clear 211 define basic variables Rho 1030 q 9 915 Avolt 254 1 6Calibration value A of the pressure meter BvOLt 43250 6Calibration value B of the pressure meter TanAlfa 0 001 bedslope NofFiles 1 number of files to be processed StartFile 1 first file to be processed for ii StartFile NofFiles if ii StartFile FileName PathName uigetfile txt Select the wavelogger file end if NofFiles 1 fid fopen PathName FileName pathstr heading ext fileparts FileName else if ii StartFile pathstr heading ext fileparts FileName heading heading 1 length heading 1 end tellerstring int2str ii FileName heading tellerstring txt fid fopen PathName FileName end o o heading is caption of figures default headin
3. cepere mti 14 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft This annex outlines the basic do s and don ts when operating the unit It assumes that all necessary care will be taken with the electronics It must be noted that the system has been designed to be as simple and cost effective as possible and for this reason some of the user error protections that may be found on other devices don t exist here 1 The Pressure Sensor The pressure sensor outputs 0 5 to 4 5 volts full scale 0 50 PSI At the moment the micro controller uses a reference voltage of 3 3v It 1s therefore not possible to submerge the unit past 15m water depth on the safe side The pressure sensor has been soldered directly to the controller board If for any reason the logger is taken apart make note of the pin outs reversing the voltage to the sensor will destroy it The sensor has a stainless steel membrane which is exposed to the water This hole should be kept clean and no tools should be used remove debris as this will damage the membrane Wave logger has a Honeywell MLH pressure sensor See http sensing honeywell com index php ci 1d23108 amp la id 1 amp pr_id 31550 MLH 050 PGP 06 A 50 psi psi gage Flying leads 20 AWG 6 in 1 6 in 27 NPT 0 5 Vdc to 4 5 Vdc ratiometric from 5 Vdc exication 2 The SD Card SD cards used should be formatted FAT16 Some SD cards may
4. 6Time domain analysis counter 0 0 21 Communications on Hydraulic and Geotechnical Engineering 2013 01 5 TUDelft Technische Universiteit Delft Pmax 0 period 0 start 1 if method 1 loop to find individual pressure waves needed to find number of waves using mean period for k 1 n 1 if Pwave k 1 Pwave k lt 0O amp amp Pwave k 1 lt 0O Snew wave starts at 1 counter countertl 5 counter is number of found waves Pmax 0 Pmin 0 end if Pwave gt Pmax Pmax Pwave k end if Pwave k lt Pmin Pmin Pwave k end end Scounter number of waves found Tmean n interval counter Sloop to estimate water level variations for k 1 n period Tmean LO g 2 pixperiod period Lt Depth k LO 0 356 L sqrt g Depth k 1 Depth k L0 periog else L L0 end eta Pwave Rho g cosh 2 pi L Depth k end Sloop to find individual waves from waterlevel record eta counter2 0 etamax 0 etamin 0 for k 1 n 1 if eta k 1 eta k O0 amp amp eta k 1 lt 0O Snew wave starts at 1 counter2 counter2 1 H counter2 etamax etamin etamax 0 etamin 0 end if eta k etamax etamax eta k end if eta k etamin etamin eta k end end vector H contains all individual waves 5 sort to make exceedance graph else eta 1 n 0 Method 2 starts
5. loop to find individual pressure waves needed to find number of waves using mean period lor k l n 1 if Pwave k 1 0 amp amp Pwave k 1 lt 0 Snew wave starts at 1 counter countert 5 counter is number of found waves TT counter period LO g 2 pixperiod period if Depth k LO lt 0 36 L sqrt g Depth 1 Depth k 10 period else 22 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft L L0 end H counter Pmax Pmin Rho g cosh 2 pi L Depth k Hunreduced H counter Hreduced H counter if H counter gt steepness L H counter steepness L Hreduced H counter end if H counter lt Hminimum Hreduced Hminimum counter counter 1 end reduction Hreduced Hunreduced Lor jjJ SLart ik 33 39 Rho g cosh 2 pi L Depth jj reduction Start k 1 remember startpoint for next wave Pmax 0 Pmin 0 period 0 end period periodtinterval if Pwave gt Pmax Pmax Pwave k end if Pwave k Pmin Pmin Pwave k end end end counter number of waves found o vector H contains all individual waves 5 sort to make exceedance graph if extraplot gt 0 figure subplot 4 1 1 plot time eta xlabel time sec vlabel waterlevel im title heading date tijd end Send waterlevel loop counter length H for k 1 counter prob k tc
6. 5 z is scaled probability for Rayleigh paper plotting data points convert probabilities for plot zi sqrt log yi hold on piob xci 21 4 oN o to draw regression line o coef polyfit xi zi 1 0 5 min x1 xb 1 5 maxixi xx xa xi xb yy polyval coef xx plot xx yy H1 BG Htr Hrms 1 H2 BG Htr Hrms 2 xxx xmark 1 01 xmark m lxxx length xxx calculate plot with CWD of Battjes Groenendijk Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft for 1 1 1 18 xXx J Htr yyy j l exp power xxx j Hl Hrms 2 0 5862 else j 1 power j H2 Hrms 3 6 580 3 6 end end convert probabilities to scale for plot zzz sqrt log l vyy hold on Communications on Hydraulic and Geotechnical Engineering 2013 01 rd U D elft Technische Universiteit Delft crosgk m function P E dof crosgk X Y N M DT DW Stats CROSGK Power cross spectrum computation with smoothing in the frequency domain Usage P F CROSGK X Y N M DT DW stats oP oe Inputs X contains the data of series 1l Y contains the data of series 2 N is the number of samples per data segment power of 2 M is the number of frequency bins over which is smoothed optional oP no smoothing for M 1 default DT is the tim
7. set hlines 7 Displayname T_ m0 2 set hlines 8 Displayname T_ m 1 0 set hlines 9 Displayname T_ peak legend Location WestOutside xlabel days ylabel wve period s 22 TUDelft Technische Universiteit Delft 35
8. varw df Pee gt Pyy 2 ns N 2 varw df Pyy 2 ns N 2 varw df Pay j smoothing if M51 w w hamming M w sum w w w ceil M 1 2 M zeros N M 1 w 1 ceil M 1 2 1 w fft w Pxx ELC Px Pyy tft Pxy IIL PXY Pxx ifft w Pxx Pyy ifft w Pyy PRY 1 end Pxx l N 2 3 Pyy 12 7232 Pxy 1 N 2 frequencies 112 Y signal variance if DW 1 nn ns 1 N 2 else 22 TUDelft Technische Universiteit Delft 3l Communications on Hydraulic and Geotechnical Engineering 2013 01 rd UDelft Technische Universiteit Delft nn ns N end avgx sum X 1 nn nn varx sum X linn 9 avgx 2 7 nn 1 avgy sum Y 1 nn nn vary sum Y 1 nn avgy 2 nn 1 sum X l nn avgx Y l nn avgy nn 1 mOxx 0 5 1 sum Pxx 2 N 2 1 0 5 Pxx N 2 df mOyy 0 5 stm Pyy Z2N 72 L 0 5 22 df mOxy 0 5 Pxy 1 sum Pxy 2 N 2 1 0 5 Pxy N 2 dtf disp mO0x varx num2str mOxx varx mOy vary num2striumDyy vary mOxy varxy num2str real mOxy covxy o Pxx Pxx varx Pyy Pyy vary mOyy Pxy Pxy covxy real 0 P pps Pyy Pxyl output spectrum
9. 5 example of the output of a wave height distribution In the output of the script the blue line 1s the Rayleigh fit the red line the Battjes Groenendijk fit and the blue crosses the observed data This dataset was observed on a natural beach in very shallow water 1 9 m The extreme wave heights in the record are relevant for the design of coastal structures The stability of armour units depends on the highest values in the record Assuming a Rayleigh Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft distribution usually gives a significant overestimation in relatively shallow water leading to a too heavy structure Note that with a frequency domain analysis see below one cannot determine these highest waves in the record Irregular waves frequency domain The time domain analysis is not very accurate in determining the periods of the wave Also it does not give information on the shape of the wave spectrum Therefore a frequency domain analysis 1s needed In principle two methods are possible l Determine the surface elevation of the water and apply spectral analysis on this surface elevation 2 Determine the spectrum of the pressure and calculate from the pressure spectrum the surface elevation spectrum wave energy spectrum Method 1 is standard in case the surface elevation was measured directly e g with wave buoys or step gauges However when applying this method
10. 595 startpunt 10 and eindpunt 5 In summary the following input values can be given in the Matlab script Line Variable Default description nr name value 1030 Density of the water kg m 8 lg 981 Acceleration of gravity m s 9__ Avolt__ 254 1 _ Calibration value A of the pressure meter PV 00008 Bvolt 43250 Calibration value B of the pressure meter Pa 58 Smooth factor for the wave spectrum a low value M 25 high value M 100 wee MM _ 62 0 08 parameter to eliminate noise from the tail of the spectrum when the spectral value is less than a cutoff fraction of the maximum value then the spectral value is set to exact zero 0 gives no plots 1 only summary 2 include pressure 3 include depth 4 include pressure 113 methods 2 uses eal period per wave Maximum allowed wave steepness for individual waves 72 interval 0 25 interval is the sample frequency interval of the sensor do not change this value unless another wavelogger is used References BATTJES J A GROENENDIJK H W 2000 Wave height distribution on shallow foreshores Coastal Eng Vol 40 pp 161 182 13 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd U D elft Technische Universiteit Delft Appendix 1 Description of the wave logger The Philosophy A need for a cost effective wave recording instrument A need ror an adaplabdie Shallow walor wave rec
11. Blinking rapidly Orange OFF Red OFF ERROR SD Card Green OFF Orange OFF Red ON 6 USB Cable The supplied USB cable has a built in FTDI chip a driver may need to be installed and can be found here http www ftdichip com FTDrivers htm The cable is energized with 5V from the computer USB port it 1s therefore necessary to connect the cable to the logger in the correct orientation or damage will result The colour of the corresponding wire has been written next to the FTDI pins on the logger 7 Configuring the Unit The device may be configured via the supplied USB cable Any terminal program may be used to communicate with the device Hyper Terminal 1s recommended The date and time of the unit can be checked and changed and the sampling duration may be changed Sample intervals are 15 min 30 min 60 min and 180 min So if 15min chosen log times will be 0 00 0 15 0 30 0 45 When 180min chosen log times will be 0 00 03 00 06 00 09 00 etc Method Remove power to the unit Plug in the USB cable to the computer Start Hyper Terminal or some other terminal program gt Select to COM port that corresponds with the FDTI USB cable Com port settings Baud Rate 9600 Data Bits 8 Parity None Stop Bits 1 Flow Control None Open the port Connect the USB FTDI cable to the device take note of orientation Connect power to the unit All three LED should be lit In the terminal window text should appear promp
12. and Geotechnical Engineering 2013 01 PlotSeries m FileName PathName uigetfile mat Select the datafile to be plotted pathstr heading ext fileparts FileName load FileName k length uitvoer for i 1 k day i uitvoer i 1 1 724 end startpunt 55 eindpunt 15 datum StUitvoer startpunt 1 tijd StUitvoer startpunt 2 subplot 3 1 1 hlines 3 plot day startpunt k eindpunt uitvoer startpunt k exndbunt o323 T59 5 xlabel days ylabel waterlevel m 1 tijd TFI set hlines 3 Displayname tide legend Location WestOutside subDplobti 35 12 35 hlines 4 plot day startpunt k eindpunt uitvoer startpunt k eindpunt 4 4 b hold on hlines 5 plot day startpunt k eindpunt uitvoer startpunt k exrndpunt 5 5 r 5 xlabel days ylabel wave height m set hlines 4 Displayname H 1 3 set hlines 5 Displayname H m0 legend Location WestOutside subDplob 3 1L 2 5 hlines 9 plot day startpunt k eindpunt uitvoer startpunt k exndpunt 9 9 Tb hold on hlines 8 plot day startpunt k eindpunt uitvoer startpunt k eindpunt 8 8 r hold 6n hlines 7 plot day startpunt k eindpunt uitvoer startpunt k eindpunt 73575 9953 hold on hlines 6 plot day startpunt k eindpunt uitvoer startpunt k eindpunt 676 8777 set hlines 6 Displayname T_ mean
13. for pressure records inaccuracies are included because of the fact that the calculated surface elevation is smoothed and does not contain all period information any more Therefore for pressure data method 2 is to be preferred In the script the pressure record is processed by the script crosgk This script developed by Klopman of Deltares determines the power cross spectrum using a Fast Fourier Method This results in a pressure spectrum x 10 pressure spectrum Asparuhovo pressure Pa Hz 0 A ro 0 05 0 1 0 15 0 2 0 25 0 3 0 35 0 4 frequency Hz Figure 6 A pressure spectrum derived from the pressure data A certain degree of smoothing of the spectrum is recommended This value can be set by changing the parameter M in the script 0 is no smoothing Practical values for M are between 20 and 100 For each frequency bin one can calculate from the pressure value the elevation value on the vertical axis is the pressure in Pa Hz in the wave energy spectrum we need m Hz This means that the same formula can be used as was applied in the time domain analysis eq 5 Now the exact value for the period 18 known because that is given by the number of the frequency bin This process results in the energy spectrum Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft energy spectrum 4 3 T T t 2 E E 1 0 1 0 2 0 3 04 freq
14. spectrum is the product of arm and area For higher order moments one raises the arm to a certain power The zero th order moment is the area multiplied with the arm to the power zero which is in fact only the Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft surface area The zero th order moment gives the total energy in the spectrum The general equation of the moment is m f SG df 12 0 The zero th and the 1 order moment are sow DX j 13 m ESES e The script calculates 71 1 m and From these moments the following values are calculated H E 8 14 H 6 T 7 RE 15 Ma One should expect that Hmo and are equal If this is not the case one should verify if the settings of the cutoff parameters An overview of all standard output of the script is given on the next page Communications on Hydraulic and Geotechnical Engineering 2013 01 waterlevel m Time series Asparuhovo 22 TUDelft Technische Universiteit Delft 1 WNT AE 111111 1 I N ALI 11111 1 N N 111 1 1 I
15. stringB num2str HlpermilleRayleigh 3 stringC num2str H01 3 Strill e T H 10 15 0 String PS EEIDnGgB 114 set gcf CurrentAxes h texti l 55 S5tr FontSize 0 end Simple script utilizing crosgk by Klopman to obtain Spectral estimate data contains the data N is the number of samples per data segment power of 2 5 M is the number of frequency bins over which is smoothed optional no smoothing for M 1 default DT is the time step optional default DT 1 5 DW is the data window type optional DW 1 for Hann window default DW 2 for rectangular window stats display resolution degrees of freedom optimal YES 1 NO 0 5 P contains the cross spectral estimates column l 25 Pyy Pxy 5 F contains the frequencies at which P is given load time series DT interval data Pwave P F dof crosgk data data length data M DT 1 0 srecalcultate pressure spectrum to energy spectrum energy 1 length F 5 length is number of frequency bins for i 1 length F energy 1 0 end m020 6zero th moment m120 tfirst moment m2 20 second moment 101 0 first negative moment m 1 0 deltaF F 31 F 30 emax 0 prmax 0 maximum value of pressure energy claculation loop to transform pressure spectrum to energy spectrum and to calculate the moments of the spectrum low and high frequencies are deleted range from 200 to 0 2 max
16. to 1 466 1 549 1 520 Leolo ime 1 936 1 061 2 280 1 467 14 549 Le 15 2 1 1 914 0529 1 796 L061 1 2501 1 468 1 990 1 622 Leolo 0 6 lad 9 120052 duos 1 469 Ia 1 624 1 817 0 65 1 578 1 064 1 284 1 471 1 554 1 626 15920 Bu y 1 492 1 066 12 250 1 474 1991 1 629 1 823 Dera 1 419 1 063 1 290 1 478 iuo 1 634 i n29 0 8 12 21 1 205953 1 294 1 483 135907 1 0599 14059 0 85 14 302 14077 1 300 1 490 l5 79 1 646 1 843 0 9 1 296 1 083 1 307 1 498 12 992 T655 14992 0 95 122216 1 090 1 34 9 1 9507 1 591 1 4055 1 864 1 Let SZ 1 324 LL 1 603 14 677 1 0447 12 12153 L106 INS 1 2530 1 616 1 690 1 092 Lal 1 128 1116 141245 14543 1 630 1 05 1 909 Las 1 108 1 31 2 109 1 1 0 1 645 1121 14927 loud 1 090 1 138 5 30 Lg 1 662 Letop 1 946 1 249 i 0795 1 120 1 3941 1990 1 078 12757 1 967 12 1 2005 Lelo 12304 1 607 1 099 1 776 1 998 1 092 1 395 1 626 lag 1 7 945 220121 1 4 1 043 1 51159 1 2599 1 644 1434 1 017 2 034 1 45 1 036 1 209 1 403 1 664 Le 1 939 2 058 EC 1 020 Legs 1 406 140599 Led fo 1 860 2a 082 Ll 1 024 1 408 15203 1 799 1 882 2 106 14 020 1 226 1 410 Letal 1 820 1 904 Ze doa T65 1 016 1 261 1 411 la 196 1 841 1 947 2 156 1 7 12015 lex d L412 1 749 1 949 2 180 12022 2 290 1 5 1 3 759 1 884 l9 2 2 207 Len 14009 JE 1 413 la 19 1 906 1 994 Zs 23 Teco L007 1 320 1 414 1 773 143214 2 017 22404 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universite
17. 02 HlIeSHrms BG Htr 0 01 HOlI Hrms BG Htr 0 00 1 splot Rayleigh graph if extraplot gt 0 h axes Position 0 0 1 11 Visible axes Position L 45 40 2 45 30 Lor i l counter probsorted2 i 1 i counter 1 end maxh 1 3 Hsorted counter step 1 if maxh 5 1 step 0 5 end if maxh lt 2 6 step 0 25 end xmark 0 step maxh coef Rayleigh Hsorted probsorted2 xmark Hrms Htr title Wave height distribution m string num2str Hmean 2 str 1 2 H mean m rob st string num2str Hrms 2 str 2 H rms m oer Ling i string num2stri Hls 2z s str 3 H 5 m string num2str Tmean 2 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd U D elft Technische Universiteit Delft str 4 4 8 LS i string num2str T3 2 stro PILA UD String j string num2str Htr 2 str 6 H tr m BG j string num2str 1 TanAlfa str 7 Bed slope 1 string str 8 obs Rayleigh BG string num2str H2percentMeasured 3 stringB num2str H2percRayleigh 3 stringC num2str H2 3 str 9 H 2 m 8LPIDmg tetrxngcl J 3 string num2str HlpercentMeasured 3 stringB num2str HlpercRayleigh 3 stringC num2str H1 3 str 10 2 H 1 m setrangol l ys string num2str HlpermilleMeasured 3
18. Communications on Hydraulic and Geotechnical Engineering Gy 2013 01 TU Delft Technische Universiteit Delft ISSN 0169 6548 EMS wave logger data processing Processing data from a pressure gauge Henk Verhagen March 1 2013 Associate Professor Section of Hydraulic Engineering Faculty of Civil Engineering and Geosciences Delft University of Technology P O Box 5048 2600 GA Delft The Netherlands Tel 31 15 27 85067 Fax 31 15 27 85124 e mail H Verhagen tudelft nl Delft University of Technology Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft Communications on Hydraulic and Geotechnical Engineering 2013 01 ISSN 0169 6548 The communications on Hydraulic an Geotechnical Engineering have been published by the Department of Hydraulic Engineering at the Faculty of Civil Engineering of Delft University of Technology In the first years mainly research reports were published in the later years the main focus was republishing Ph D theses from this Department The function of the paper version of the Communications was to disseminate information mainly to other libraries and research institutes Note that not all Ph D theses of the department were published in this series For a full overview is referred to www hydraulicengineering tudelft nl gt research gt dissertations At this moment this series is mainly used to disseminate back
19. In line 12 and 13 one should NofFiles and StartFile to 1 Then in lines 58 to 72 the other input paramters can be set In case of processing multiple observations set NofFiles to the number of observations If one intends to skip the first files because they do not contain information set the StartFile to a higher number This might be useful when the Wavelogger already started to observe before it was placed on its final location In case of multiple observations no plots of individual observations are made like spectra and exceedance lines In the command window the number of processed files 1s shown so one can follow the progress of the calculation After completion of all processing the data are written to an outputfile This file has the same name as the inputfiles of the observations but without the sequence number and the extension mat e g Logger2 012Part mat Note that this version of the matlab script requires a heading with the following form S 1 12 0 3 63 6 7 12 In newer versions of the Wavelogger this header 15 S BurstIndex 2012 11 01T00 00 00Z Temp 12 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft One can plot the results with the Matlab script PlotSeries m In this script one can set the first point to be plotted startpunt and the last point eindpunt This means that 1f one has a record of 600 points and on intends to plot the series from 10 to
20. ater depth and wave period The accuracy of the wave logger is in the order of 250 Pa In fig 2 the accuracy of the calculated water elevation is plotted as a function of the water depth and the wave period It is clear from the figure that in larger water depths the accuracy decreases significantly for the shorter wave periods In general one may conclude that up to a water depth of 10 waves with a period of 5 seconds and larger are quite accurate Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft Irregular waves time domain In case of irregular waves the above theory 18 still valid but cannot be applied without restrictions A real surface record may look like figure 3 0 8 0 6 021 41 ENIM water level m 0 4 0 6 0 8 0 50 100 150 time s Figure 3 example of a surface water level record As can be seen all individual waves have different periods and heights To calculate back from a pressure record to a water surface level the individual wave periods should be known In theory one could split a pressure record into individual waves and then calculate the resulting surface elevation for each individual wave However this usually leads to large errors with shorter waves Therefore one can better first calculate the average wave period in the whole record 1 the total observation pe
21. aves can be generated The flume 1s equipped with standard high performance wave gauges with a very high sampling interval 20 Hz Appendix 2 shows a comparison of the data from the EMS wavelogger processed with the TU Delft script and the results from Deltares equipment in the flume User Manual The data are processed by the Matlab script Wavelogger m This script processes files with one observation only The Wavelogger produces one single datafile with all observations in sequence The first step in processing is splitting the data from the logger in separate files This can be done by any program A nice tool is the program TextSplitter exe Load the output file from the Wavelogger into this program enter the number of observations and let te program slit the datafile into separate chunck of file When the file is large this may take several minutes be patient The number of observations can be found at the end of the file The observation number is the number after D It is recommended to avoiud filenames with an underscore _ because this symbol makes that letters are printed as subscript 1n the Matlab script observations from one ensemble should have the same name and a different number e g Logger2 012Part1 txt Logger2 012Part2 txt Logger2 012Part3 txt etc Both individual observations as well as a batch of observations can be processed by the Matlab scripts wavelogger m In case of a single observation
22. characteristics dof floor 2 ns M 1 2 3 DW if stats fprintf number of samples used 8 0f n nn fprintf degrees of freedom 5 0fin floor 2 ns Md 3 DW fprintf resolution 2 2 end 32 Communications on Hydraulic and Geotechnical Engineering 2013 01 5 TUDelft Technische Universiteit Delft BG m function Hnorm BG Htr Proba Function developed by H J Verhagen oo Delft University of Technology oo January 2013 oo Inputs oo oo Htr H transition acc Battjes Groenendi Jjk Proba code for required exceendandce eq 8 in paper Proba 3 gt H1 3 Proba 10 gt 11 10 Proba 0 02 gt H2 Proba 0 01 gt H1 Proba 0 001 gt 15 oo Ourpur Hnorm oo source Wave height distributions on shallow foreshores Coastal Engineering Volume 40 Issue 3 June 2000 Jurjen A Battjes Heiko W Groenendijk oo Pages 161 182 oo stable from Battjes groenendijk 5 Htr H1 H2 H173 1710 25 H1 HO ls BGtable 12 193 415 000 Lom 1 466 1 548 1 620 Leolo Jel T4003 1 060 Ladro 1 466 1 548 1 620 1 813 Oa S063 1 060 1 279 1 466 1 548 1 620 1 813 Usa 4 022 1 060 L279 1 466 1 548 1 5620 1 913 Uu so Da 1 060 Lae 719 1 466 1 548 1 620 1 013 ine 2 908 1 060 1 279 1 466 1 548 1 620 1394 3 0 59 1 060 1 279 1 466 1 548 1 620 1 813 0 4 23994 1 060 1 219 1 466 1 548 1 620 1 0 45 2 104 1 060 Lee
23. e step optional default DT 1 DW is the data window type optional DW 1 for Hann window default DW 2 for rectangular window stats display resolution degrees of freedom optimal YES 1 NO 0 o9 contains the cross spectral estimates column 1 Pxx 2 Pyy 3 F contains the frequencies at which is given Gert Klopman Delft Hydraulics 1995 o if nargin 4 M 1 end IL margin lt 3 DT 1 end LR nargin lt Gy DW 1 end Heroin 7 stats 1 end gr 1 data window w 1 if DW 1 5 Hann w hanning N dj N 2 else rectangle w ones N 1 dj N end varw sum w 2 summation over segments nx max size X ny max size Y avgx sum X nx sum Y ny px zeros size w py zeros size w Communications on Hydraulic and Geotechnical Engineering 2013 01 PXX Zeros size w zeros size w Pyy Zeros ns 0 ror J lt djinx Nt 1 ns ms Ls compute FFT of signals px 115 evox WwW px LU px 7 py Y j j N 1 py py IItiby y compute periodogram Pxx Pxx px con px Pyy py Gon py Ex Pxy Q7 eon px 3 end Pex 2 7 9 1122
24. ferred method Also waves which are smaller than a given value Hminimum are considered not to be real The value of Hminimum depends on the accuracy to the pressure sensor and the waterdepth In shallow water the value of Hminimum can be very small In deep water Hminimum is approx 10 cm Subsequently the waves are sorted on individual height In the Matlab script the array with sorted wave heights is called Hsorted The number of waves found is called counter In this report the following notation is used N Number of waves individual wave average wave height Hiis root mean square wave height significant wave height In the next step the time domain variables are determined 1 nen Da 6 _ 1 2 H ms N 7 Y 8 1 3 2 m Also are determined the wave height exceeded by 2 1 and 0 1 of the waves These wave heights are determined in three ways a Directly from the observations b Using a Rayleigh distribution C Using the Composite Weibull Distribution according to Battjes and Groenendijk 2000 With method b a Rayleigh distribution of the waves is assumed This assumption is usually correct in deep water For the Rayleigh distribution is valid H 140 H 1 50H 9 1 85 The equation for the Rayleigh distribution 1s 2 P H 10 In shallow water the highest waves in the record are already broken Therefore the Rayleigh dis
25. frequency bin 200 means f 200 deltaF which is approx 30 seconds 0 2 length F deltaF 0 4 so Tmin 2 5 seconds o o9 o 25 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd UDelft Technische Universiteit Delft loop to determine maximum value of pressure spectrum for 1 50 2000 if sqrt P i l prmax prmax sqrt 1 1 end end for 150 2000 1 115 pr sqrt P 1 1 pr pressure value in pressure spectrum 110251 20592191 if AverageDepth L0 0 36 L sqrt g AverageDepth 1 AverageDepth L0 T else L L0 end energy i pr Rho g cosh 2 pi L AverageDepth 2 e energy density in Hz m2 lf pr cULtorr prmax energy 1 0 end if energy i gt emax emax energy 1 Tpeak T end mO m0 energy 1 deltaF ml ml energy 1 deltaF F 1 m2 m2 tenergy 1 deltaF F 1 2 if E 1 20 mO1l m01 energy i deltaF F i end end Hm0 4 sqrt m0 sone may assume Hs Hm0 Hrmss sqrt 8 m0 root mean square height from spectrum Tm sqrt m0 m2 spectral approximation of mean period TOI SmO mls SPeriod based on first moment T102m01 m0 SPeriod based on first negative moment plot spectrum and print output if extraplot gt 0 h axes Position 0 0 1 1l Visible off axes Position 45 05 45 20 plot F energy axis low high O 10000 axis auto xlabel frequency Hz ylabel Energy m 2 Hz title energy spectrum string num2str Hr
26. g FileName if needed assign other value to heading here heading fill in your heading HDRS textscan fid s d jJG HDRS temp HDRS 4 date HDRS 5 1 0 s 90d 87 1 delimiter while ccc DATA textscan fid s 54 54 d delimiter ccc strcmpi DATA 1 SD 1 transformation from Volts to pressure using calibration constants Avolt Bvolt SAvolt has dimension Pa V Bvolt has dimension if ccc 1 g n Avolt DATA 3 Bvolt 0 1 end end n n 1 count number of samples fclose irid P double q lt additional input data spectrum 20 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd U D elft Technische Universiteit Delft M 50 higher values of M give more smoothing of the spectrum low value for M is 20 high value for M is 100 high 0 25 highest frequency for plots default high 0 4 low 0 05 lowest frequency for plots default low 20 05 cutoff 0 10 parameter to eliminate noise from the tail of the spectrum 5 when the spectral value is less than cutoff of the max 5 value then the spectral value is set to exact zero default 0 08 extraplot 2 extraplot 0 gives no plots 1 only summary 2 include prssure 3 include depth 4 include pressure method 2 method 1 uses Tm for calculating waveheight from pressure method 2 uses rea period per wave Stee
27. ground information related to other publications e g data reports with data underlying journal papers and Ph D theses Recent issues of the Communications are only available in digital format A notification will be sent to interested readers when new issues are released For placement on the notification list please send an e mail to h j verhagen tudelft nl Older versions before 1986 were published as Communications on Hydraulic Engineering A number of internal reports were not published in this series but are available via this website Postal address for the Communications is TU Delft Faculty of Civil Engineering and Geosciences department of Hydraulic Engineering Stevinweg 1 2628CN Delft Netherlands Permissions for republishing parts figures data can be obtained from the responsible publisher ir H J Verhagen This publication has been produced in cooperation with the Water Resources University Hanoi Viet Nam This study is part of the project Technical Assistance for Sea Dike Research financed by the Government of the Netherlands is acknowledged for funding to build the Wave Overtopping Simulator and to perform all the destructive tests in Viet Nam Tests were performed by the Faculty of Marine and Coastal Engineering Water Resources University Ha Noi Viet Nam 2011 TU Delft Section Hydraulic Engineering Henk Jan Verhagen Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Tech
28. it Delft 1 9 1 006 1 334 1 414 Lerro 1 949 240 39 2 282 1 004 1 349 1 415 1 783 1 9710 241162 2343207 1 004 LAIS 1 786 1 985 2 084 2 2 205 1 003 1 378 1 4105 1 789 1983 2 106 2 301 Zal 12 002 1 392 14415 l 791 1 982 24128 Zu 04 222228 L002 1407 1 4109 la 0193 l 991 232121 2 406 CERE 1 001 142421 12415 14795 1 981 27 149 214230 22 1 001 1 555 1 22 105 1 796 1 980 2 148 dod ou 212 15001 1 449 14415 1 797 1 979 2 148 2 478 115001 1 462 1 415 1797 1 979 25147 2 502 2 4 T DO 1 476 1 416 1 798 1 979 2 147 2 0223 205 1 0006 1 490 1 416 1 798 1 979 24147 2 948 213 1 000 1 503 12416 1 799 1 978 END M 22 2409 1 000 1 5145 1 416 1 799 1 978 2 146 2 4093 2 6 1 000 1 529 1 416 1 799 12 978 2 146 2 616 2 69 T4000 14 547 1 416 Le 199 T1978 2 146 2 629 221 1000 12 22 1 416 lw 99 1 976 2 146 2 629 LINE ES 1 000 1 OS 1 416 1 900 L9 358 2 146 2 628 2 49 1 000 1 580 1 416 14 9000 L978 2 146 2 628 240 1 000 1999 1 416 1 900 1 978 2 146 24028 2 9 1 000 1 605 1 416 1 800 1 918 2 146 2 628 2 909 1 000 1 617 1 416 1 800 1 979 2 146 2 628 3 1 000 1 620 1 416 1 800 15 21 18 2 146 2 042043 j70 if Proba 1 J 2 end if Proba 2 J 3 end if Proba 3 1 4 end if Proba 10 J 5 if Proba 0 02 156 if Proba 0 01 1 7 end if Proba 0 001 j 8 end i rouncd Her 0 05 if 1 60 end Htir 05 i 1 end 2 0 Hnorm NaN else Hnorm BGtable i Jj end end Communications on Hydraulic
29. mss 2 str 1 T H rms m stringl string num2str i Hm0 2 DITE m stringi Struingsnum2stri rm z s str 3 T m s String strring num2str r0l 2 str 4 T imD 1 string strringenum2strirlo0 2 I BS 1 00118 1 172 string num2str Tpeak 2 str 6 Z peakb s string yal 7 7 etr o er 26 Communications on Hydraulic and Geotechnical Engineering 2013 01 string num2str AverageDepth 2 str 9 Average depht m stringl string num2str Rho str 10 water density kgm 3 ySEEIDgIJIS str ll set gcf CurrentAxes h text el 15 str FontSize lt 3 end SOptional plot pressure spectrum if extraplot gt l1 figure plot FE P rf l P is pressure 2 Hz axis low high O 1500 axis auto y xlabel frequency Hz ylabel pressure Pa 2 Hz title pressure spectrum heading end StUitvoer 1i 1 date StUitvoer ii 2 tijd uitvoer ii l ii uitvoer ii 2 AverageDepth uitvoer 11 4 uitvoer ii 5 Hm0 uitvoer ii 6 Tmean uitvoer ii 7 Tm uitvoer ii 8 T10 uitvoer ii 9 Tpeak MeanDepth 0 for 11 StartFile NofFiles MeanDepth MeanDepthtuitvoer 11 2 end MeanDepth MeanDepth NofFiles StartFile for ii StartFile NofFiles uitvoer ii 3 uitvoer ii 2 MeanDepth end savefile heading mat save sa
30. nische Universiteit De EMS wave logger data processing Introduction Waves can be measured in several ways One way of measuring waves is by measuring the wave pressure at a certain depth using a pressure sensor and calculate the wave information from the pressure record The EMS wave logger uses a Honeywell MLH 050 PGP 06A pressure sensor The information is stored by the logger on a SD card The software in the logger controls the sample durations from 1 to 30 minutes and the sample intervals from 15 min to 3 hours The sampling rate is fixed to 4 Hz Wave pressure In a regular small amplitude wave the instantaneous water level is given by 7 asinQ 1 Under a regular small amplitude wave the pressure 1s given by coshk h z p DESTERU o BU 2 in which instantaneous water level m pressure at the requested location Pa density of the water kg m acceleration of gravity m s depth of the pressure note under water z is negative m amplitude of the wave m wave number 27 7 phase angle of the wave rad local wave length m 7 4 0 0 1 X Figure 1 definitions in the wave pressure equation When inserting eq 1 into eq 2 leads to cosh k h z _ 24 3 G cosh kh The wave logger is produced by Environmental Mapping amp Surveying Durban South Africa see also Appendix 1 Communications on Hydraulic and Ge
31. not work due to speeds and or capacity 3 Data Logging The unit will log data to a file called datalog txt This file is created by the unit if it does not exist If a file 1s present it will be appended to If you would like to deploy the instrument it is recommended that you remove the existing datalog txt from the SD card so that only new data will be written The data is only saved to the file when the SD card is closed by the microcontroller at the end of each burst period Therefore if the green LED 1s still flashing and power is removed the last burst will NOT be written to the card There is a push button to close the file at any moment When pushing the button all data will be saved and both LEDs go solid The device is setup to log once per hour when the minute variable is 0 The user may only change the number of samples to be recorded done the menu The following file format is used to store data Data File Structure S header Num burst number Time Temp Date This 1s followed by the data at a rate of 4 Hz D header Num burst number Raw ADC Reading E header Num burst number End Time 15 Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft 4 Indictor LED s On startup for 10 seconds Green ON Orange ON Red ON In sleep mode between bursts Green OFF Orange Blink every 4 seconds Red OFF Recording measurements Burst Green
32. ordar Super insiru emu recovery Increase apah al resolubon of wave dynamics through multiple depleyment One sinn shoan anpnmach instfamanf data collaction Physical Properties data processing software final usable results Sua 500mm X 155mm OD Weignt 3 kg Marci ial PY C u ylic Wight Nae Direstional Wave Logger Operabng Depth Um with acrylic end plate and rugged Buret Pregsure 500 psi PVC housing Electrical Features Batteries 3 X 4 5Ahr Operatng Voltage Sv Eiu arme up lu 30 Days Operabng Ourrent 20 mA Slandby Current 16 mA Sampling Properties Sampling Period 20min Programmable 5ampiing Interval 1 Hr Programmable Burst Sampang Raw 4riz Programmabie 9D Card up to ZOB Left Turic al deployment mooring with stabil ing Pressure Transducer li a es i T ype Honeywell MHL Series STEP i ocean ng Parameters 0 3 FSS a on Resulultiur Tui T Aca rmm PO anm Tempoeratiire Seneor one 11 Interna UTEP 2 Dua Fie Temp Range 50 to 150 Deg C Wave Parameters re of Waves per sampling period Right Srreenetint nt riata H 3 processing software M aximum Wave Amplitude Average Wave Amplitude wave Energy Eos 1 fas L 100 420 5 ample No 604 i40 1064 khoe Sangre dala ert
33. otechnical Engineering 2013 01 UDelft Technische Universiteit Delft In order to do this correct the density of water needs to be known This parameter Rho has to be set in the script For fresh water 1000 for sea water one should raise p up to 1030 kg m Subtracting the hydrostatic pressure leads to coshk h z cosh kh 4 _ coshkh pg cosh k h z Ap And in case the pressure is measured at the bottom this reduces to n cosh kh 5 Eq 5 can be used to calculate the surface elevation from the pressure record However one should realise that this derivation is only valid for regular small amplitude waves In shallow water usually waves cannot be considered small amplitude waves and second order theory should be applied to describe the water surface elevation The main difference between first order and higher order waves is the exact shape of the surface Wave height and wave period are not very different Because of this first order wave theory 1s acceptable for this purpose In important drawback of determining wave heights from pressure records 18 that small short waves cannot be measured at larger water depths 0 1 0 09 0 08 0 07 E 0 06 d a 0 05 eta m 0 08 M 0 02 _ 0 01 _ 0 2 3 4 O N Co col e depth m Figure 2 accuracy of the calculated water elevation as a function of w
34. ounter ktl coounter end Hsorted sort H Hmean mean Hsorted Hrms sqrt mean Hsorted 2 5 rms wave height from round 2 counter 3 to round counter Hs mean Hsorted from to H 1 3 or significant wave height Tsorted sort Tmean mean TT from rourndi z counter 3a3 s to round counter 25 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd UDelft Technische Universiteit Delft T32mean Tsorted from to T 1 3 or significant wave period calculation of the H2 Twopercent round counter counter 50 if Twopercent counter H2percentMeasured NaN else H2percentMeasured Hsorted Twopercent end H2percRayleigh 1 4 Hs eCaculation or the H1 Onepercent round counter counter 100 if Onepercent counter HlpercentMeasured NaN else HlpercentMeasured Hsorted Onepercent end HlpercRayleigh 1 5 Hs calculation of the 15 Onepermille round counter counter 1000 if Onepermille counter HlpermilleMeasured NaN else HlpermilleMeasured Hsorted Onepermille end HlpermilleRayleigh 1 85 Hs Compositie weibull ditribution for shallow water source Wave height distributions on shallow foreshores Coastal Engineering Volume 40 Issue 3 June 2000 Pages 161 182 Jurjen A Battjes Heiko W Groenendijk o o Htr 0 3545 8 TanAlfa AverageDepth Transistion wave height acc to BG eq 38 H3 Hrms BG HEr 3 H10 Hrms BG Htr 10 HZeHrms BG Htry0
35. pness 0 03 max allowed wave steepness individual waves default 0 05 Hminimum 0 10 minimum waveheigth to be considered default 10 m interval 25 5 interval is the sample frequency interval of the sensor do not change this value unless another wavelogger is used if NofFiles gt 1 extraplot 0 end in case of multiple files plotting ihdivadual files is blocked tottime n interval Scalculate total duration of observation in sec ttime interval interval tottime create array with time time ttime 6change orientation of matrix calculate regression coefficients to compensate for change in waterlevel during the observations if extraplot gt 3 figure plot time P xlabel time sec ylebel pressure Paj title observed total pressure heading end BB polyfit time P 1 regression analysis to determine real waterdepth at any moment sand correct for hydrostatic pressure Intercept BB 2 Slope BB 1 Pwave P Intercept time Slope Pstatic P Pwave SHydrostatic pressure Depth Pstatic Rho g AverageDepth mean Depth SOptional figure to plot waterdepth as function of time if extraplot gt 2 figure plot time Depth ylabel depth m xlabel time sec title average waterdepth heading SOptional figure to plot wave pressure as function of time figure plot time Pwave xlabel time sec ylabel pressure title presure variations heading end
36. riod divided by the number of waves and use that wave period for the calculation When doing so the short waves in the wave field are not really accounted for But these waves are also the lower waves and therefore usually not the most relevant part of the wave height distribution For the time domain analysis therefore a surface elevation 18 generated using eq 5 and the mean period of the observation time With a counting algorithm then the individual waves are distinguished in this generated surface elevation record To separate the waves the downward crossing method is used From each wave the height 1s determined In order to calculate the water surface from the pressure record the period has be known The period is included in k in eq 2 The program has two methods to do this In method 1 the average period observation time number of waves 1s used In method 2 the period of each individual wave 1s calculated and used Method 1 goes wrong with a wide spectrum Method 2 goes wrong in deep water Small pressure variations in deep water are translated to huge ridiculously short waves This is physically not correct Therefore the wave steepness is limited to a maximum value for individual waves this max value is called steepnesss in the Matlab script Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft In general method 2 with a correct limitation of the steepness is the pre
37. ting the user to enter the M capital key to enter the setup gt Follow the instructions on screen to complete the setup Note the sequence of the above actions is relevant V V V V M Communications on Hydraulic and Geotechnical Engineering 2013 01 UDelft Technische Universiteit Delft 8 The Housing The housing is made from rigid PVC pipe The fittings have been thermo welded to each other There 1s a single o ring seal for the unit This o ring should be kept clean and silicon applied after each deployment The seal is compressed with the screw on the end cap 9 Batteries The unit has a 5v step up converter for Li batteries One D Cell 3 6v Lithium Ion 16500AHr is required The unit should operate for at least 30 40 days on all configurations 17 Communications on Hydraulic and Geotechnical Engineering 2013 01 Appendix 2 Wave logger calibration Calibration test EMS wave logger 7 June 2012 Deltaflume Deltares 22 TUDelft Technische Universiteit Delft LA 11 04 06 de voe height tr 18 TUDelft Communications on Hydraulic and Geotechnical Engineering 2013 01 Calibration test EMS wave logger 7 June 2012 Deltaflume Deltares Wave surface elevation with EMS wave logger 1 L Lob 2 p a aaa aa T sf 2
38. tribution is not valid any more Battjes and Groenendijk 2000 suggested a approximation of this distribution using 2 kan i es 2 lt H Pr H lt H 11 Ranzi 2 H gt H Below a transitional value of the wave height the Rayleigh distribution remains valid Above this value the exponent in the distribution has a different value 3 6 The values Communications on Hydraulic and Geotechnical Engineering 2013 01 U Delft Technische Universiteit Delft and were fitted from measurements and tabulated in the original paper These values depend on the water depth and the bed slope Figure 4 below shows the ratio Hyo Aims for various values of H H Compose Weibull distribution n T p 1 100 50 2010521 0 1 P H H Figure 4 Wave height distribution in shallow water Battjes Groenendijk 2000 H The parameters from the Composite Weibull Distribution are loaded into the script from the matlab script BG Note that 1n case of a limited number of waves in the record 1000 no observed value of can be determined The script gives then as output NaN Not a Number Wave height distribution EN M 0 41 s m 0 45 5 m 0 62 Tae 43 H m ace BG 0 69 slope 1 1000 Rayleigh BG 0 763 087 0708 Qm 0 787 0 932 0741 sr 0 827 1 15 0 829 Figure
39. uency Hz Figure 7 The energy spectrum calculated from the pressure spectrum of figure 6 It is clear that the peak at the right side of the spectrum is an artefact It is caused by the fact that a small value of the pressure is multiplied with an extremely high multiplier This has no physical meaning The above data were measured on a water depth of 8 5 m Small variations in the pressure less than the accuracy of the meter are interpreted by eq 5 as being caused by huge waves of very short periods While in reality these variations are only random noise Therefore the spectrum should be cut off in this case at a value of approximately 0 28 Hz see also figure 6 With the parameters cutoff and high this is realized The variable cutoff limits the analysis while high only changes the axis of the plot See figure 8 for the result energy spectrum 2 1 5 Energy m Hz 0 5 ES 0 005 01 015 02 025 03 frequency Hz Figure 8 corrected wave energy spectrum using the parameters cutoff and high J One can determine the values for cutoff and high by using the plot of the pressure spectrum This plot is not printed on default only when the parameter extraplot is set to 1 There is also a value low This value should be used for deleting long periodic waves which have nothing to do with the waves to be analysed e g seiches From the spectrum various moments are computed moment of a
40. vefile StUitvoer uitvoer 22 TUDelft Technische Universiteit Delft 21 Communications on Hydraulic and Geotechnical Engineering 2013 01 rd UDelft Technische Universiteit Delft Rayleigh m function coef rayleigh xi yi xmark Hrms Htr function coef rayleigh xi yi xmark o o Rayleigh Probability Paper The paper is marked with probability of exceedance 0 002 0 005 0 01 0 05 0 1 0 1 0 9 horizontal lines associated with these values are drawn In addition it draws the linear curve fitting for input data points o 5 horizontal input physical quantity 5 yi vertical input in probability of exceedance xmark vertical grid lines to be plotted 5 coef slope and intercept of the linear regression line xmark poe l 0 1 20 1 0 05 0 02 0 01 0 005 0 0025 0 001 0 00051 DroOD Or exceedance y Sqrt logipoe n length y m length x 5 1 Xm yl dx abs x m x 1 25 xh x 1 x m working on horizontal line Klin yh y k y k line xh yh text x 1 2 dx y k num2str poe k prob of exceedance end 5 set gca linestyle working on vertical line horizontal scale dy abs y n y 1 25 yv y 1 y n ror kel im xv x k x k line xv yv text x k dx 4 y 1 dy num2zstr xma rk k end axis orr x is wave height y is probability correct numbers
Download Pdf Manuals
Related Search