Real-Time Safety Estimation Method and Experimental Validation for Deep Sea Pressure Structural Health Monitoring

YANG Hua-wei,WAN Zheng-quan,HUANG Jin-hao,XIANG Yang-jun

(China Ship Scientific Research Center,Wuxi 214082,China)

Real-Time Safety Estimation Method and Experimental Validation for Deep Sea Pressure Structural Health Monitoring

YANG Hua-wei,WAN Zheng-quan,HUANG Jin-hao,XIANG Yang-jun

(China Ship Scientific Research Center,Wuxi 214082,China)

Abstract:A structural strength safety estimation method of structural health monitoring(SHM)system for deep sea pressure structure is presented,based on the real-time monitoring data,structure baseline condition,and monitoring system measurement error,without structural model.The proposed method is validated by pressure structure model experiment in a pressure cabin.The experiment result shows that the safety condition of deep sea pressure structure could be estimated by proposed method without building a model.The strength condition alarm value with proposed method is 0.76 MPa more than alarm pressure value with the model-based standard method.The presented method can be applied in SHM system for deep sea pressure structure.

Key words:structural health monitoring;structure safety estimation;pressure structure;real-time sensors data

0 Introduction

The interest in structural health monitoring(SHM)of offshore structures has increased greatly in the last two decades especially due to the need to objectively manage civil infrastructure systems all around the world[1-4].A SHM is the process of comparing the current state of a structure’s condition,relative to a baseline state and determining the existence,location,and degree of damage that may exist,particularly after a damaging input[2].A SHM system is employed automatically to carry out the data acquisition and structure safety estimation process,which is an electronic measuring and processing system.The role of the SHM system is to acquire data from sensors and process these data,gives the estimation result.Typically,a model-based approach is used to estimate the structure’s condition with reasonable point due to its reliability and simplicity[5-6].However,this approach requires a model to be built completely.In fact,models for practical offshore structure are usually difficult to be built.Seraphim,et al[5]used the fuzzy-logic-based network to estimate the structure’s condition,but the method was offline and not real-time processing.Due to this fault,it is still difficult for SHM system researchers to estimate the more common offshore structures,for example deep sea pressurestructure.Up to now,there had been few reports in the literature about using real-time sensors data to estimate the offshore structure’s safety condition without the structure model[7-9].In this paper,a novel kind of structure safety estimation is presented for SHM system on deep sea pressure structure which used real-time sensors data without the need of building models.Particularly,the approach using real-time sensors data can estimate the deep sea pressure structure safety condition,the models of which are unnecessary to be unknown,and will improve the universality of the SHM system.

1 Principle and formula

The strength of deep sea pressure structure is the most important factor which resists structure destruction and should be sufficient above all.In theory,for structure’s condition estimation,if the real-time monitoring stress value of every sensor position in SHM system isthen the structure strength condition should be safe when it meets the following equation:

For example,the pressure spherical shell structure of manned submersible,the shell stress value under calculating pressure should be satisfied with the following equation:

where σsis yield strength of material,units MPa.

In fact,the SHM system for deep sea pressure spherical shell is a long term measuring instrument,the measurement results of which include multiple errors.So,the estimation result may be incorrect in most cases by using Eq.(2)to estimate the structure strength condition independently.In the next text,a novel estimation method for the deep sea pressure hull strength which could give a convincing evaluation result,is given out.

(1)The pressure structure baseline condition function

Assumed the strength condition of deep sea pressure structure is in a good condition when it has been installed with SHM system.The functions on strain and depth parameter can be obtained by 3 or more times diving experiment data,which are regarded as pressure structure baseline condition response functions.

If the deep sea pressure structure has j times diving experiments,and k ranks strain data are acquired at every experiment,then the following matrix M for every one strength condition monitoring sensor position of pressure structure can be obtained.

in which

k:the load ranks;j:the diving times;hn:the depth at nth rank of every diving experiment,units m(MPa);εmn:the strain data of one strain direction at mth time diving experiment and nth rank depth,units με;H0:the maximum diving depth of deep pressure structure.

The matrix A can be obtained from matrix M.

The pressure structure baseline condition functionfor different strength monitoring position can be obtained by using matrix N with least square method,such as the following equation:

where htis the diving depth of deep sea pressure structure at t time,a is proportionality factor,b is offset factor.

(2)The errors of the SHM system due to uncertainty

In general,assumed the measuring data errors are Gaussian distribution in a measurement system,so the mean value of the data errors for the same measurement value should be zero when the measurement times are enough large.However,for the SHM system for deep sea pressure structure,the measurement value varies at different diving depth,so the relative measurement errors are used to calculate the value of uncertainty of the SHM system.The relative errors could be obtained from the matrix M,and it could be expressed by matrix N with element ξm×n.

The strength measurement uncertainty θ of SHM system could be calculated from Eq.(7),the expand value of uncertainty is expressed by 3θ with probability level 99.87%.

(3)The estimation algorithm for structure strength safety

Based on above,the real-time structure strength safety estimation algorithm for deep sea pressure structure can be determined by Eq.(8).If the value of K is equal to 1,then the evaluation result for the structure strength is in risk;else if the value of K is equal to 0,then the estimation result for the structure strength is safe.

εt:measured strain of the monitoring structure position at time t;ht:diving depth of the deep sea pressure structure at time t;the basal condition function of strength monitoring position;3θ:the expand value of uncertainty of strength monitoring position;σt:the stress of strength monitoring position;the permissible stress for strength monitoring position;HP:the preset starting evaluation depth.

2 Materials and experiments

2.1 The model

The experiments of estimation algorithm for pressure structure strength safety were conducted by using a pressure spherical shell model.As shown in Fig.1,the prototype pressure structure was a closed spherical shell with an opening hole.The parameters of the prototype pressure structure are showed in Tab.1.

Tab.1 Parameters of pressure spherical hull

Fig.1 The model of pressure hull

Fig.2 Curve of thickness measurement of pressure hull

2.2 The sensor layout for structure strength safety estimation

To determine the sensors mounted position,the thickness of the pressure spherical shell was measured with ultrasonic thickness meter.The pressure spherical shell was respectively divided to 16 equal portions in latitude and longitude with different numbers.The measurement results of thickness are showed by latitude number versus longitude number diagram,as Fig.2.According to the measurement thickness of pressure spherical shell model,the sensors assembly position was determined.The strength monitoring sensors position in the surface of pressurespherical shell model was marked,as(M1,M2),(S1,S2),(S3,S4),(S5,S6),(S7,S8),(S9,S10),and the type of all sensors is biaxial strain gauge,see Fig.3.

Fig.3 Sensors layout scheme of SHM sytem

2.3 The test facilities and SHM system

After the strain gauges were mounted and protected in pressure spherical shell model surface,the spherical shell model was placed in a big and high pressure cabin which was full of water and closed.The high water pressure cabin can provide a pressure environment like deep sea,maximum pressure 30 MPa,inner diameter 4.5 meters(see Fig.4).

Fig.4 The spherical shell mounted sensors was placed into water pressure cabin

Fig.5 Structural health monitoring and estimation system

Fig.6 The block diagram of the structure safety experiment

The SHM system for deep sea pressure structure shell is a distribution acquisition system,which included one host computer,many data acquisition equipments,many stain gauges,one pressure sensor and real-time monitoring&estimation software(see Fig.5).Strain gauges were distributed in surface of spherical shell in accordance with sensors layout scheme.The pressure sensor was installed in the water input pressure pipe of high water pressure cabin.The pressure structure model was loaded by external water pressure with trapezoidal wave shapeproduced by high pressure-pump.The Fig.6 shows the block diagram of the experiment of the SHM safety estimation with real-time sensors data.

2.4 The load and structure safety estimation parameters

The pressure load was implemented according to following procedures:

First time:0 MPa→5 MPa→8 MPa→10 MPa→12 MPa→0 MPa

Second time:0 MPa→5 MPa→8 MPa→10 MPa→12 MPa→0 MPa

Third time:0 MPa→5 MPa→8 MPa→10 MPa→12 MPa→0 MPa

Last time:0 MPa→5 MPa→8 MPa→10 MPa→12 MPa→broken

At every pressure stage,the pressure held steady at the same pressure value for 5 minutes.

During the first 3 times pressure loading,the SHM system collected data continuously only.After 3 times pressure loading were finished,the structure strength safety estimation parameters using the collected strain and pressure data were calculated according to the method in Chap.1.

For example,the maximum strain data sensor position,the structure baseline condition functionand expand uncertainty 3θ of the position(M1,M2)were obtained by baseline condition strain data at different pressure values which were obtained during 3 times loading procedures(see Tab.2).

Tab.2 Structure baseline condition function for every sensor and expand uncertainty value

3 Validation result and analysis

Added the structure strength safety estimation parameters to the SHM system software(see Fig.7),the SHM system had a real-time safety estimation result using the real-time measuring data automatic without manual operation during the pressure collapse experiment of pressure spherical shell model.The SHM system acquired strain data and pressure data real time every 1 second.

Fig.7 Estimation parameters input and SHM system software

During the collapse experiments,the SHM system had a strength safety condition alarm when the pressure of high pressure cabin was 13.76 MPa,at the longitude direction of(M1,M2)sensors position for the pressure spherical shell model.The strength condition evaluation for pressure spherical shell model is represented above 13 MPa in Tab.3.The course of strength condition estimation was showed in Fig.8.

Tab.3 Real time strength condition evaluation result for spherical shell model

Fig.8 The course of strength condition estimation

Fig.9 The shape of spherical shell broken

In pressure spherical shell model experiments,the real-time structure strength condition estimation method gives a good prediction for the strength condition of spherical shell model and had an alarm at 13.76 MPa which before the collapse pressure of 17.68 MPa.The change curve of strength condition for the pressure structure was revealed very clear in Fig.8.FromTab.3,the difference value 0.76 MPa of strength condition alarm pressure between the model-based method and the presented method can be calculated,it shows that the presented method can improve the accuracy of structure strength condition estimation.From the Figs.8 and 9,the structure strength of individual position with opening hole is insufficient,when pressure increased and became bigger,the different position of pressure spherical shell model deformed inconsistent obviously,the deep sea pressure spherical shell model was broken because of deforming inconsistent largely at last.

4 Conclusions

In this paper,the real time pressure structure strength condition estimation method for SHM system is presented,which is based on the real time monitoring data,structure baseline condition,and monitoring system measurement errors.The capacity of proposed method was validated by the pressure test of the pressure spherical shell model in a high pressure cabin,which had a strength condition alarm starting from 13.76 MPa,and the pressure spherical shell model collapsed at 17.68 MPa.The test result indicates that the proposed method can be used in SHM system for deep sea pressure structure and has an important role on forecasting structure safety.

[1]Necati Catbas F,Susoy M,Frangopol D M.Structural health monitoring and reliability estimation:Long span truss bridge application with environmental monitoring data[J].Engineering Structures,2008(30):2347-2359.

[2]Geoffey Chase J,Begoc V,Barroso L R.Efficient structural health monitoring for a benchmark structure using adaptive RLS filters[J].Computer and Structures,2005(83):639-647.

[3]Park H W,Sohn H.Parameter estimation of generalized extreme value distribution for structural health monitoring[J].Probabilistic Engineering Mechanics,2006(21):366-376.

[4]Ko J M,Ni Y Q.Technology developments in structural health monitoring of large-scale brides[J].Engineering Structures,2005(27):1715-1725.

[5]Abou S C.Fuzzy-logic-based network for complex systems risk assessment:Application to ship performance analysis[J].Accident Analysis and Prevention,2012(5):305-316.

[6]Stull C J,Earls C J.Phaedon-stelios koutsourelakis.Model-based structural health monitoring of naval ship hulls[J].Computer Methods Appl.Mech.Engrg,2011(200):1137-1149.

[7]Lin Tian Ran,Pan Jie,O’Shea P J,Mechefske C K.A study of vibration and vibration control of ship structures[J].Marine Structures,2009(22):730-743.

[8]Elshafey A A,Haddara M R,Marzouk H.Damage detection in offshore structures using neural networks[J].Marine Structures,2010(23):131-145.

[9]Flynn E B,Todd M D.A bayesian approach to optimal sensor placement for structural health monitoring with application to active sensing[J].Mechanical Systems and Signal Processing,2010(24):891-903.

[10]Yang Huawei.The report on estimation method for pressure structure safety and software implementation technique[R].Wuxi:CSSRC,2016.

深海耐壓結(jié)構(gòu)健康監(jiān)測實時安全評估方法和試驗驗證

楊華偉，萬正權(quán)，黃進浩，向楊君
（中國船舶科學(xué)研究中心，江蘇 無錫 214082）

文中提出了一種可直接用于深海耐壓結(jié)構(gòu)健康監(jiān)測系統(tǒng)的結(jié)構(gòu)強度安全性評估算法，該算法基于實時監(jiān)測數(shù)據(jù)、結(jié)構(gòu)基準狀態(tài)和系統(tǒng)測量誤差，與常規(guī)方法相比不需要建立監(jiān)測對象結(jié)構(gòu)模型。研究出的結(jié)構(gòu)安全性實時評估方法在耐壓結(jié)構(gòu)模型壓力筒破壞試驗中進行了驗證，結(jié)果表明，結(jié)構(gòu)健康監(jiān)測系統(tǒng)在耐壓結(jié)構(gòu)破壞前發(fā)出了報警信息，與常規(guī)基于模型的結(jié)構(gòu)安全性評估方法相比，報警壓力值提高了0.76 MPa，對深海耐壓結(jié)構(gòu)的使用具有重要的工程價值。

結(jié)構(gòu)健康監(jiān)測；結(jié)構(gòu)安全性評估；耐壓結(jié)構(gòu)；實時監(jiān)測數(shù)據(jù)

U661.4

A

楊華偉（1981-），男，中國船舶科學(xué)研究中心博士研究生，高級工程師；萬正權(quán)（1962-），男，中國船舶科學(xué)研究中心研究員，博士生導(dǎo)師；黃進浩（1975-），男，中國船舶科學(xué)研究中心研究員；向楊君（1986-），男，中國船舶科學(xué)研究中心工程師。

10.3969/j.issn.1007-7294.2017.09.008

Article ID： 1007-7294（2017）09-1136-09

Received date：2017-05-22

Biography：YANG Hua-wei(1981-),female,Ph.D.,student of CSSRC,senior engineer,E-mail:yhw9832@126.com;WAN Zheng-quan(1962-),male,researcher/tutor of CSSRC.