WO2006117431A1

WO2006117431A1 - A system for ice load monitoring

Info

Publication number: WO2006117431A1
Application number: PCT/FI2006/000141
Authority: WO
Inventors: Risto Rouvari; Kaj Riska
Original assignee: R. Rouvari Oy
Priority date: 2005-05-03
Filing date: 2006-05-02
Publication date: 2006-11-09
Also published as: FI20050476A; FI118610B; FI20050476A0; KR20080012331A; EP1877747A1

Abstract

The invention relates to providing an ice load monitoring system (500) for monitoring ice-induced loads on a structure of a vessel (501). The vessel structure comprises typically at least a shell plate (102), vertical frames (106) essentially normal to the shell plate (102) and longitudinal frames (104) essentially in the horizontal plane and normal to the shell plate (102). Further the ice load monitoring system comprises a combination of stress sensors (F, S, P) each of which is adapted to convert deformations of the structure into a signal proportional to the deformations. The combination of sensors comprises at least frame bending stress sensors (F), plate stress sensors (P) and frame force sensor (S). The ice load monitoring system (500) comprises also a data processing unit (506), and an information network (508) connecting the sensors (S, P, F) and the data processing unit (506). The data processing unit (506) is adapted to collect the signals from the sensors (S, P, F).

Description

A SYSTEM FOR ICE LOAD MONITORING

TECHNICAL FIELD OF THE INVENTION

The invention relates to an ice load monitoring system for measuring loading, stresses and/or material fatigues occurring in a structure. Especially the invention relates to the ice load monitoring system for monitoring ice- induced loads on a structure of a vessel, such as a ship. The invention also relates to a measuring sensor suitable for use in connection with said system.

BACKGROUND OF THE INVENTION

Hull stress due to loading or wave loads is a burden to all ships with a large hull. Also the ice load acts on the ship hull and the shell structure response. Bulk carriers, large tankers, LNG (Liquefied Natural Gas) carriers and FPSOs (Floating Production Storage and Offloading unit) are vessel types that are particularly susceptible to hull stress. In many cases, special purpose ships such as research vessels, military craft, heavy-lift vessels and ice-breaker work on the edge of their design stress values. Therefore load exerted on a ship hull should be consistently and continuously measured for any changes taking place in the structure, such as material fatigue, and also to detect that an allowable stress, for example, is not exceeded.

Different kinds of measurement systems are known from prior art to control and monitor the hull stress induced by the load of the ship and also the effect of waves on structural stresses on the hull of the ship. The prior art systems are implemented by pressure monitoring devices or other sensors detecting the stress or deformation of the structure. The sensors are typically connected to a computer, which is connected to a monitor screen. The computer receives the measured values from the sensors and on the basis of a specified software program calculates the desired quantity, such as condition of the vessel in question. The calculated values are generally also visualized on the monitor screen.

The prior art systems, however, concentrate to monitor the structural hull stresses of the ship induced only by the load of the ship itself or the effect of the waves. There are no practical systems for monitoring ice-induced loads on a structure of a vessel. The lack of a practical system for monitoring ice- induced load may be the consequence of the fact that the ice load as a phenomenon is largely statistical in nature and thus more difficult to monitor than the effect of waves or the load of the ship itself, for example. Because of its statistical nature it is at first very hard to set the stress sensing sensors in a right location in the structure of the vessel since one cannot predict the behaviour of a block of ice and secondly it is hard to analyse the output values of the sensors so that the desired results and the real effect of the ice load on the whole hull structure can be determined.

Moreover there are problems in the prior art to bond the sensors so that the attachment would be long-lasting and will resist the unfavourable effect of the surrounding circumstances. The prior art sensors are bonded to the measured area, which is typically metal, by adhesive and then covered by silicone, whereupon the problem is that water will penetrate gradually between the metal and silicone and thus break the bonding before long.

SUMMARY OF THE INVENTION

The object of the present invention is to avoid the prior art problems mentioned above with the aid of a new measuring solution and also by a new sensor attachment solution. Further the object of the invention is to provide an ice load monitoring system for measuring loading, stresses and/or material fatigues occurring in a structure induced by the ice load regardless of its statistical nature.

The objects of the invention are fulfilled by the features described in characterising part of the independent claim 1.

The typical shell structure of the vessel comprises of at least a shell plate, vertical frames normal to the shell plate and longitudinal frames essentially in the horizontal plane and normal to the shell plate. The idea of the invention is to provide a combination of stress sensors to specified locations in the structure in order to detect a comprehensive loading on the ship hull and the shell structure response induced by the ice load, wherein each stress sensor is adapted to convert deformations of the structure into an output signal proportional to the deformations. The combination of sensors comprises of at least frame bending stress sensors (F), plate stress sensors (P) and frame force sensor (S). According to the first embodiment of the invention the bending stress sensors (F) are located in a frame flange, more advantageously in mid-span of the frame. The frame flange is advantageous location for the bending stress sensors, because the deformation caused by the ice load is considerable and thus also the measuring accuracy is excellent.

According to the second embodiment of the invention the plate stress sensors (P) are located in a shell plate, more advantageously between frames and in mid-span of longitudinal frames.

According to the third embodiment of the invention the frame force sensors (S) are located in a frame web, more advantageously in a longitudinal frame, and even more advantageously in the neutral axis of the longitudinal frame.

The stress sensors are advantageously located in a most critical area, where expected ice load or expected ice load versus structure capacity is maximum.

According to the fourth embodiment of the invention the measuring of the ice load is implemented by a sensor panel in the shell structure, where the sensor panel comprises the combination of stress sensors and is located in the critical area of the ship bow. Advantageously the loads and stresses are measured with several sensor panels, of which one is located in the load waterline height (LWL) and one in the ballast waterline height (BWL). According to an advantageous embodiment of the invention only one of these panels is active at a time depending on the vessel's loading condition.

The monitoring system of the invention comprises also a data processing unit, and an information network connecting the sensors and the data processing unit. The data processing unit is advantageously adapted to collect signals from the sensors and calculate desired results. The signals are advantageously in electric form, either converted to electric form by a converter (when the sensor is an optical gauge, for example) or directly outputted in electric form (when the sensor is a strain gauge, for example).

According to the fifth embodiment of the invention the external ice load acting on the structure is calculated (advantageously by the data processing unit) on the basis of the outputs of the stress sensors, where the outputs of the sensors are proportional to strain of the structure. According to the invention the loads and stresses acting on the structure are calculated by multiplying the outputs (strains or strain difference proportional to the output signal) of the stress sensors with influence coefficient matrix (using FEM, Finite Element Method).

Furthermore according to the invention in the ice load monitoring system load level and stress indication can be provided advantageously in real time, assisting a vessel's master in decision making. The system is also capable to make short-term load predictions based on the data collected by the sensors over the last minutes, for example. The measured loads and stresses are compared with the yield limit of the structure and the system is adapted to warn if the permissible limit of ice loading is exceeded. According to an advantageous method of the invention also maximum stress is compared with the permissible stress level and displayed with color-coded area graph with the ship profile.

The system is also adapted to store the measured load data with time and position data into long-term storage for later analysis. This data can be used for example to asses the ice navigability in the operation area, to estimate long term damage probability of the ship, in fatigue analysis and for other research purposes like development of ice class rules. By comparing the different measurements recorded in the course of time, the evolution of the structure over its life span can be monitored.

The invention relates also to a stress sensor comprising a stress sensing element, which can be implemented by a strain gauge which resistance and output voltage is proportional to strain of the structure, for example, or alternatively by an optical gauge. The stress sensing element can also be implemented by some other gauge known by a skilled person, such as by an acoustical or pneumatic gauge, which output signal can be converted to an electric signal proportional to measured strain of the structure.

According to an embodiment of the invention the stress sensing element is bonded to the structure to be measured. The stress sensor comprises also a casing covering the stress sensing element, and a reference sensing plate adjacent to the stress sensing element. The stress sensor includes in addition support members to be welded to the structure to be measured so that the support members support the casing and the reference sensing plate. According to the invention the stress sensing element and its wire connections are covered with silicon for mechanical protection, and the casing is sealed for the potting mass and filled with polyurethane potting mass, epoxy or the like to protect the stress sensor against humidity and corrosion. Advantageously oxsilane is applied on the surface, where the stress sensing element is to be attached, in order to prevent penetration of water or humidity between the surface and the stress sensing element or the potting mass. Oxsilane advantageously forms chemical compound with a metal surface and also a potting mass, such as polyurethane or epoxy.

The present invention offers also other advantageous features. It is, for example, possible to improve the safety and economy of shipping, minimize the risks for hull structure failures and thus prevent costly casualties, and save the operational lifetime of the vessel. Furthermore it is also possible to provide a crew of the vessel by information of ice load, trend, severity measure and prediction.

The following notions are used in this application:

FEM Finite elements method, which is particularly useful when a robust approximation is sought to solve structural response described by partial differential equations.

Loading history The system can show time histories of different measuring channels. The suitable time window length can be 2 hours, for example, and loads can be shown as percentages of the permissible loads.

Load histograms The system can show load histograms of the different measuring channels. The histogram is based on the detected load amplitudes. Loads can be shown as a percentages of the allowed loads. Suitable time period for the histograms can be 2 hours, for example.

Neutral axis Geometric centroid of the cross section or transition between compression and tension.

Permissible loads For example for a typical ship construction steel the yield limit is 235 MPa. This is the permissible limit for frame and plate stresses if this kind of material is used . The yield limit can, however, be greater (such as even 600 MPa) or lower depending on a material. For partial and total load for one frame the limit load values are calculated based on the FE-model of the instrumented area, or the area, where the sensor panels are located. The limit value for ice load corresponds the case where the frame bending stress exceeds the yield limit of the material. There exist two limit values for the panel total load. One is the case that the limit load value on any frame is exceeded and the other is the case where limit load value of the main frame is exceeded.

Frame load Is the ice load acting over a longitudinal frame span. Frame load is proportional to a shear strain difference between stress sensors mounted on the both ends of the frame span.

Partial frame load Is the ice load acting over a half-length of longitudinal frame span. Partial frame load is proportional to the shear strain difference between stress sensors mounted on one end and middle of the frame span. The local ice pressure is the partial frame load divided by the nominal contact area A_nOm (⁼ frame spacing * half length of the frame span). It should be clear to a skilled person that the half-length used here is only exemplary length.

Severity Can be determined by the ratio between measured and permissible load, for example. According to an embodiment of the invention ice forces and ice-induced stresses are measured close to a bow of a ship at most loaded areas and then the forces and stresses are extrapolated to the other areas in order to determine a 'single measure of severity for the ice loading'. Ice load monitoring display shows the measured data advantageously either in severity units (percentages of permissible loads, using a colour code for different ranges of the measure of severity) or in engineering units (kN, MPa).

Preferred embodiments of the invention are described in the dependent claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Next the invention will be described in greater detail with reference to exemplary embodiments in accordance with the accompanying drawings, in which Figure 1 illustrates a side view of an exemplary arrangement, where the sensors are provided in certain locations according to an advantageous embodiment of the invention,

Figure 2 illustrates a top view of an exemplary arrangement, where the sensors are provided in certain locations according to an advantageous embodiment of the invention,

Figure 3 illustrates an exemplary method of determination of a neutral axis where the frame force sensors are located according to an advantageous embodiment of the invention,

Figure 4 illustrates an exemplary arrangement of internal strain measurements according to an advantageous embodiment of the invention,

Figure 5a illustrates an exemplary ice load monitoring arrangement with two sensor panels on the vessel according to an advantageous embodiment of the invention,

Figure 5b illustrates an exemplary arrangement of two plate stress sensors on a panel on the vessel according to an advantageous embodiment of the invention,

Figure 6 illustrates an exemplary influence coefficient matrix a for the sensor panels according to an advantageous embodiment of the present invention,

Figure 7 illustrates an exemplary chart of time and event windows according to an advantageous embodiment of the present invention,

Figure 8 illustrates an exemplary chart of monitoring and comparing windows in trend evaluation according to an advantageous embodiment of the present invention,

Figure 9 illustrates an exemplary chart of load prediction according to an advantageous embodiment of the present invention, Figure 10 illustrates an exemplary exploded view of a stress sensor used in the ice load measuring system according to an advantageous embodiment of the invention, and

Figure 11 illustrates an exemplary stress sensor used in the ice load measuring system as attached into the structure to be monitored according to an advantageous embodiment of the invention.

DETAILED DESCRIPTION

Figure 1 illustrates a side view of an exemplary arrangement, where the sensors (F, P, S) are provided in certain locations in a structure of the vessel according to an advantageous embodiment of the invention. The typical structure 100 of the vessel comprises at least a shell plate 102, vertical frames (as 106 shown in Figure 2) normal to the shell plate 102 and longitudinal frames 104 essentially in the horizontal plane and normal to the shell plate 102.

According to the invention the bending stress sensors (F) are located in a frame flange 104A, the plate stress sensors (P) are located in a shell plate 102, more advantageously between or in mid-span of the longitudinal frames 104. Further the frame force sensor or sensors (S) are located in the longitudinal frame 104.

Figure 2 illustrates an exemplary arrangement, where the typical structure 100 of the vessel, illustrated as a top view, comprises at least a shell plate 102, vertical frames 106 normal to the shell plate 102 and longitudinal frames 104 essentially in the horizontal plane and normal to the shell plate 102. In Figure 2 the precise locations of the bending stress sensors (F) and the frame force sensor (S) can be seen according to an embodiment of the invention, where the bending stress sensors (F) are located in a frame flange 104A and more advantageously in mid-span of the frame 106, and the frame force sensors (S) are located in the longitudinal frame 104, and even more advantageously in the neutral axis 202 of the longitudinal frame 104. Now it should be noted that the location of the neutral axis may be different in different longitudinal frames 104.

Figure 3 illustrates an exemplary method of determination of a neutral axis, where the frame force sensors (S) are located according to an advantageous embodiment of the invention. The location of especially the frame force sensors (S) is essential for obtaining accurate readings according to the invention, especially in order to gain shear strains in the longitudinal frames (104). Thus the frame force sensors are mounted on the neutral axis e* (or 202 in Figure 2) of the longitudinal frame 104. In this exemplary Figure only 3 sensors are used, but it should be clear to a skilled person when reading this document, that the number -of the sensors may vary, and typically so that the more the sensors, the greater the resolution.

The neutral axis is determined according to the invention by the following formula: _e* _{= .} S₁ + S_{2 +} S₃

A₁ + A₂ + A₃

where S₁, S₂ and S₃ are static moments calculated by the following manner:

5₂ = y₂ ^χ A₂, and

5₃ = V₃X A₃,

and where A₁, A₂ and A₃ are areas calculated by the following manner:

A₁ = bi X h₁,

A₂ = b₂ x h₂, and

A₃ = b₃ x h₃,

Figure 4 illustrates an advantageous arrangement of internal strain measurements according to an advantageous embodiment of the invention.

An internal strain measurement is used to find out an external force F caused from ice impact pressure. Strain readings from strain measurement can be converted into ice loads. As strain is assumed to be affected linearly by structural response then the relationship between the external load F and the measured strain difference Δγ is determined using the influence coefficient a

F = a • Δγ .

The shear strain is measured on the neutral axis 202 of the longitudinal frame 104 by the frame force sensors (S), as illustrated in Figure 4. According to the measured shear strains, γ¹ and γ², shear difference Δγ is

Δγ = γ' -γ² ,

where superscripts 1 and 2 refer to the strain measuring points. When structure under consideration includes n number of frames, the relationship is determined by influence coefficient matrix. The relationship between a force, Fj, on a frame i and a shear difference Δγ_j, on a frame j is calculated, and as a results, an element of influence coefficient matrix, a^ is obtained. Calculating the shear difference for each frame j the results can be given in the following vector form:

Further it should be noted in relation to Figure 4, that to avoid stress concentration spots near the frames 106 that could cause undesirable distributions, the frame force sensors (S) are spaced advantageously in a distance from the frames 106, advantageously 150 mm from the frames 106, in the longitudinal direction. However, the 150 mm is an advantageous example of the invention, and it does not be considered as a strict limitation in the spirit of the claims. In this exemplary Figure only 3 sensors (S) are used, but it should be clear to a skilled person when reading this document, that the number of the sensors may vary, and typically so that the more the sensors, the greater the resolution.

Figure 5a illustrates an exemplary ice load monitoring system 500 with two sensor panels, namely the lower sensor panel 502 and the upper sensor panel 504 on the vessel 501 according to an advantageous embodiment of the invention, where the plate stress sensors (P) are located in a shell plate 102, and in mid-span of the longitudinal frames. The bending stress sensors (F) are located in a frame flange and more advantageously in mid-span of frames 106 (for sake of clarity only part of the frames 106 is illustrated in Figure 5a by solid lines), and the frame force sensors (S) are located in the longitudinal frame 104.

In addition the ice load monitoring system 500 comprises also a data processing unit 506, and an information network 508 connecting the sensors (S, P, F) and the data processing unit 506. According to an exemplary embodiment of the invention the analog output signals are taken directly from each stress sensor and the summing/subtraction is made after A/D -conversion. Then measured strains are converted to engineer units (MPa, kN), for example. The conversion from measured structural strains into ice loads relies on a technique using shear strains. The loads are obtained using an influence coefficient matrix estimated initially with FEM, and if necessary confirmed by experiments. As an example, the measuring quantities from one sensor panel may be for example: plate stress (3 sensors), frame normal stress (6 sensors), force on frame (5 strain sensors on the web of each frame, together 15 sensors), and the total force on the area (sum of the forces on the frames).

The gathered data is then transmitted to the central data processing unit, via an information network for further analysing, displaying and datalogging, and the loads and responses of the ship hull structures outside the panel area can thus be estimated based on the measured ice loads. Moreover the personnel on the bridge can be alarmed to take actions in case the measured values indicate high stress levels.

As an example, the response the different stress sensors detect, can be derived following way:

1. The frame bending stress is proportional to the strain with following equation: σ = ε*E where ε = measured strain reading (//Strain)

E = Young's modulus (206 GPa)

2. The relation between measured strain (meassured by the plate normal stress sensors) and stresses in the plating:

E σ_v = ^■

2(l -v) ^■•fo+O where \ε_φ + s_^) = measured strain reading (//Strain) v = Poisson coefficient (0.29)

Relation 2 is valid only in a situation, where the plate stress sensors (P) are mounted in an angle of 56° in relation to each other (Figure 5b), whereupon the Poisson coefficient for steel is compensated. It is, however, clear to a skilled person to modify the relation 2 for situations, where the angle between the sensors differs from 56°.

3. The relation between measured shear strain difference (measured by the frame force sensors) and force on frame i:

3 =∑a_rΔ_γj

where Δ/_j- = measured shear strain difference (//Strain) ay = elements of influence coefficient matrix

The measuring resolution will be more accurate, when more sensors are located between the frames.

Figure 6 illustrates an exemplary influence coefficient matrix a for the sensor panels according to an advantageous embodiment of the present invention, where there are 12 sensors in the each panel. The influence coefficient matrix a is made considering the results of the preliminary sensitivity analysis by FEM. In this exemplary case the simplified structure model and the load case of the line load, were used.

Figure 7 illustrates an exemplary chart of time and event windows according to an advantageous embodiment of the present invention. The essential information in the time histories is contained to load peaks, which correspond to loading events by ice features. The ice feature takes contact with the ship hull and either bounces back or slides along the hull. The loading events are local, with load heights typically corresponding to the thickness of loading ice features and with load lengths of at most few metres. The duration of the events is typically from 10 milliseconds to one second in a typical ice-going speed.

Statistical models that are used to describe the loading process are principally based on the amplitude of the load peaks. In the present invention two methods to select the load peaks from the data are considered. In the first method such sections of time history are selected that are located above some threshold value. The method for peak amplitude identification used here is the Rayleigh criterion. It states that if local minimum between two local maxima is lower than r times the shallower maximum, then there exists two different load peaks. The standard value for parameter r is 0.5.

The other method used in the invention divides the time history into time windows of specified length. The maximum value from each time window is selected. The minimum time window length used is 1 second.

Figure 8 illustrates an exemplary chart of monitoring and comparing windows in trend evaluation according to an advantageous embodiment of the present invention.

The forecast of maximum ice load on certain time period is one essential function in operative ice load monitoring system, but also the trend of the forecast is important. Changing ice conditions and changes on ship's speed and course have effect on ice load level and significance of these changes can be evaluated by observing the ice load trend.

In the ice load monitoring program the trend is evaluated by using monitoring and comparing windows, as described in Figure 8. Earlier forecasts are kept in computer's memory and the forecasts are divided into two windows as shown in Figure 8. Then the mean of the forecasts is calculated in both windows. The trend is then determined by comparing calculated medians. First the system calculates the mean values of the load forecasts in comparing and monitoring windows. The length of comparing window can be 110 minutes and the length of monitoring window 10 minutes, for example. The ratio of these means is the measure for ice load trend:

„ , (mean of forecasts in monitoring window) _Λ

Trend = - 1

(mean of forecasts in comparing window)

If the trend is positive, the load level is increasing and if the trend is negative the load level is decreasing.

Figure 9 illustrates an exemplary chart of a load forecast according to an advantageous embodiment of the present invention.

Acceptable level of ice induced loads means that the ship can continue navigating in its present mode with negligible damage risk. Thus the load level is not a descriptor of experienced loads but is of predictive nature. It is also not related to average loads but to extremal loads. The determination of the load level is based on measured loads. The observed loads have certain statistics which can be expressed by a distribution function. A distribution model is fitted to the observed histogram and used to calculate the probability of occurrence of high loads that have not been observed. Below two example techniques for prediction are described.

The (r,Xτ-)-plotting technique. A time history pertaining to some measuring channel and a time period T₀ are considered. The period is divided into shorter windows of length T and the maximum x,- value is selected from the time history for each of the N windows. The average value is then calculated as

¹ \?

The graph of *y against 7^" is called an (7,Xr)-plot. The time window approach is based on the insight that the relevant information about the loading process can be obtained from the (7,x_τ)-plot. Here a certain range of T is considered. A (7>r)-plot is very simple to generate. The comparison of two time series can be in terms of the (7,x_r)-plots. The plots can also be used predictively: the function x_τ [T) allows the estimation of expected maximum for some long time window if the function can be generated during a shorter period.

Scaling ranges. The basic question in the time window approach is thus how the average XT changes with T. Another time window λj is selected and the average x_λT. Relations that allow the derivation of x_λT from x_τ are called scaling relations. There are two principal types of scaling relations applicable here. The first has power law form:

The other scaling relation is x_λT = x_τ + H'1n(λ)

The analysis of these scaling relations is done by simple plotting.

First the scaling relation can be written in the form ln(x_T) = H\n(T) + c

Thus if ln(x_r) is plotted against In(T) and a linear slope is found for a certain range of T, then the scaling law applies in the range. The value c can be expressed in terms of some fixed T₀ in the range. This is called Gumbel Il range. On the other hand, the second scaling relation can be written in the form

;c_r = H'in(r)+ c'

If in a plot of Xτ against In(T) a linear range is found, it is called Gumbel I range and the scaling law applies in the range. The value c¹ can be expressed in terms of some fixed T₀' in the range. Thus in both cases the (7,X₇)-plot is characterised by two parameters.

The practise for load forecasts in the system is next described. The length of the observation window is 1024 seconds and the length of the prediction window is 7200 seconds (i.e. 2 hours). First the observation window is divided into following period lengths:

T = 2^Λ[0:1 :10] = [1 , 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024] seconds

so the ratio of successive period length is always 2. Then the average load maxima x_T for the periods are calculated. Using Matlab commands the average maximums for these periods are:

for J = 1:11,

x = ones(T(j) , 2M0/T(j));

x(:) = LOAD_DATA;

avermaxφ = mean(max(x));

end; x_τ=avermax;

Then the regression function between logarithm of periods and the average maximums (In(T), x_r) is found. Using Matlab curve fitting function the regression function is solved:

p = polyfit(log(T), X₇-, 2); This is second order polynomial: x_τ (T) = p, [log(T)]² + p₂ logCO + p₃

with three coefficients P₁, p₂, and p₃. The ice load forecast for the next 2 hours is made using this polynomial and setting T = 8224 seconds.

The ice monitoring system according to the invention is capable to estimate all potentially critical areas of the bow structure, not only the spots where the sensors are placed. This involves the calculation of the load carrying capacity of structural members and the estimation of the ice loads for the whole bow area. The task is to generalise the loads measured in the panel area for the whole bow area.

According to an embodiment the invention, there are at least two advantageous and different ways to develop a procedure that deduces, based on the measured quantities from the panels at the bow area, the stresses in the framing and plating of the bow structure. The stresses must be processed to obtain a Single Measure of Severity. The assumption is that the ice load is very local and acts in the ice belt area where also the stresses are highest in the shell structure i.e. frames and plating.

One way is to extrapolate the measured stresses to other areas and the other to extrapolate the measured forces and then use these forces to calculate the stresses in the structural elements of other areas. The methods are described below.

Method 1 Extrapolation of the forces

The monitoring system is able to measure the ice induced force F acting on the area bounded by the frame spacing s and the distance between the shear gauges on the frames, L. Thus the force F can be given as an average pressure on the area p_PL = F/(s-L). If the distance between gauges is more than two frame spacings (L>2s), then the maximum stress of the plating in any loading event can be assumed to caused by the maximum measured single force F_MAX. For the frames the maximum stress is caused by the maximum of the sum of all component forces on one frame F_Toτ (=F1+F2), if there are only two forces measured on a single frame. The average pressure for frames is p_FR=F_Toτ/(L1⁺L2), if the component forces are measured from one frame. The maximum stress in the plating is now

where the constants relating stress to force are obtained for different locations from FEM analysis. Similarly for the frames the maximum stress is σ_FR=C_FR-F_TOτ. Again the relationship between the stress and the load described by the constant C_FR is obtained from FEM calculations. The load in the FEM calculations must be the same as that used in the measurements i.e. pressure p_PL on plating on an area of s x L and pressure p_FR for frames on an area of s x (L1+L2).

Now the influence of the hull shape on the pressures should be taken into account. The waterline entrance angle α influences the indentation speed into ice and the higher this speed, the higher are the ice loads. The frame angle β influences the vertical bending component and again the more vertical the frame is, the higher are the loads. If it would be agreed that the ice pressure is proportional to some constant C(α, β), then the measured forces could be corrected to all other locations using this constant. The problem is to find this constant. The new (1999) Russian ice class rules use one possible shape constant which is proportional to

The theory developed in Finland about dynamic bending of ice cover would suggest a form factor to be of the following form

C(α,β)= - β-βo

Where the constants are Ci = 4.0 and β₀ = 8° (for smaller frame angles than about β=15° the dynamic bending is not valid and thus the proportionality to β is not to be used any more). Now it is suggested that the latter formulation based on dynamic bending of ice cover is used.

Method 2 Extrapolation of the stresses

The monitoring system measures the stresses on the plating and on framing at several places. It would also be possible to extrapolate these measured stresses to other areas. The method is based on the relationship between the scantlings, loading and the stress. For plating the stress is proportional to

where t is plate thickness and p the ice pressure. Similar proportionality relationship for the frames is

s² - 1² ^FR ^ P —

where I is the frame span and Z elastic section modulus for the frame. These relationships can be used to convert the measured stresses at the location where the scantlings s, I, t and Z (also α and β if the shape. is taken into account similarly as in the first method) are known to other locations where scantlings have slightly different values.

In addition a single measure of severity is required to be displayed on the ice load monitoring system according to an embodiment of the invention. The most clear measure would be the measured or deduced i.e. calculated stress on other areas than the measuring panel stress divided by the yield stress σ_γ. The measuring software could for each ice load event seek the highest stress among the plate and frame stresses and then display this (along with information at which location the maximum occurred). Thus the measure of severity S is (if all the N_PL plate stresses measured or calculated are labeled as σ_PLιi, i=1 N_PL and all the N_FR frame stresses measured or calculated are σ_FRii, i=1 N_FR),

MAX ^σPL,i ^σFR,j S = i = l,...,N_PL,j = l,...₃N_FR ^σY,i ' ^σY,j

The unit of the measure of severity is advantageously [S] = % and according to an embodiment of the invention the display uses a colour code for different ranges of the measure of severity.

The fatigue analysis for the frames where the sensors locate will be done using Rain flow method. Rain flow counting routine is implemented to the monitoring program. The Rain flow calculation is made according to the standard E 1049-85 (Re-approved 1990) (The American Society of Testing and Materials). Rain flow analysis is made for the measured stresses. Frame stresses are measured with 4 stress sensors and stresses in the plating with 2 stress sensors on each panel, for example. Data from Rain flow analysis is advantageously written in a log file for later analysis.

Figure 10 illustrates an exemplary exploded view of a stress sensor 1000 used in the ice load measuring system according to an advantageous embodiment of the invention, where the stress sensor 1000 comprises a stress sensing element 1002, which can be is implemented for example by a strain gauge, which resistance and output voltage is proportional to strain of the structure 1004. However, it should be clear to a skilled person that also other kinds of stress sensing elements known from prior art could be used, such as an acoustical, optical or pneumatic gauges, which output signal can be converted to an electric signal proportional to measured strain of the structure.

According to an embodiment of the invention the stress sensing element 1002 is adapted to be faced with the structure to be measured 1004. The stress sensor 1000 comprises also a casing 1006 covering the stress sensing element 1002, and a reference sensing plate 1008 adjacent to the stress sensing element. The reference sensing plate is used for temperature compensating. The stress sensor 1000 includes in addition support members 1010 to be welded to the structure to be measured 1004 so that the support members 1010 support at least the casing 1006 and the reference sensing plate 1008.

Figure 11 illustrates an exemplary stress sensor 1000 used in the ice load measuring system as attached into the structure 1004 to be monitored according to an advantageous embodiment of the invention, where the stress sensing element and its wire connections (advantageously inside the stress sensor casing) are covered with silicon for mechanical protection, and the casing is sealed for the potting mass and filled with polyurethane potting mass, epoxy or the like to protect the stress sensor against humidity and corrosion.

Advantageously oxsilane is applied on the surface, where the stress sensing element is to be attached, in order to prevent penetration of water or moisture between the surface and the stress sensing element or the potting mass. Oxsilane advantageously forms chemical combound with a metal surface and also a potting mass, such as polyurethane or epoxy. The invention has been explained above with reference to the aforementioned embodiments, and several advantages of the invention have been demonstrated. It is clear that the invention is not only restricted to these embodiments, but comprises all possible embodiments within the spirit and scope of the inventive thought and the following patent claims.

Claims

1. Ice load monitoring system (500) for monitoring ice-induced loads on a structure of a vessel (501), wherein the vessel structure comprises at least a shell plate (102), vertical frames (106) essentially normal to the shell plate (102) and longitudinal frames (104) essentially in the horizontal plane and normal to the shell plate (102), characterized in that the ice load monitoring system (500) comprises a combination of stress sensors (S, P, F) each of which is adapted to convert deformations of the structure into an signal proportional to the deformations, a data processing unit (506), and an information network (508) connecting the sensors (S, P, F) and the data processing unit (506), and that the combination of sensors comprises at least frame bending stress sensors (F), plate stress sensors (P) and frame force sensor (S), and that the data processing unit (506) is adapted to collect the signals from the sensors (S, P, F) for a later analysis.

2. Ice load monitoring system according to claim 1 , wherein the sensors are arranged into several sensor panel (502, 504) of which at least one (504) is located at the height of the load water line and at least one (502) is located at the height of ballast water line.

3. Ice load monitoring system according to claim 2, wherein only one of the sensor panel(s) (502, 504) is active at a time.

4. Ice load monitoring system according to claim 3, wherein the selection of the active sensor (502, 504) panel depends on the vessel's loading condition.

5. Ice load monitoring system according to any of the preceding claims, wherein the sensors are arranged on the vessel's bow area, and especially in the area, where ice load or ice load in respect to a structure capacity is expected to be at maximum.

6. Ice load monitoring system according to any of the preceding claims, wherein the frame force sensors (S) are mounted in the longitudinal frame (104) and on the neutral axis (202) of the longitudinal frame (104).

7. Ice load monitoring system according to any of the preceding claims, wherein the plate stress sensors (P) are mounted in the shell plate (102) between the frames and in the mid-span of the longitudinal frames (104).

8. Ice load monitoring system according to any of the preceding claims, wherein the frame bending stress sensors (F) are mounted in a frame flange (104a) and in the mid-span of the frame.

9. Ice load monitoring system according to any of the preceding claims, wherein the data processing unit (506) is adapted to determine the external load acting on the structure by multiplying measured chair strain differences proportional to the signal of each of the stress sensor with an influence coefficient matrix.

10. Ice load monitoring system according to any of the preceding claims, wherein the ice load monitoring system is adapted to estimate the load in all potentially critical areas of the structure in addition to the sports where the sensors (S, P, F) are located.

11. Ice load monitoring system according to any of the preceding claims, wherein the ice load monitoring system is adapted to process the signals to obtain a single measure of severity.

12. Ice load monitoring system according to any of the preceding claims, wherein the ice load monitoring system is adapted to forecast a maximum ice load on a certain time period using a fitted distribution model to an observed loads.

13. Ice load monitoring system according claim 12, wherein the forecast is divided into a first and second window, where the forecast in the first window is for later period that the forecast in the second window, and wherein the ice load monitoring system is adapted to evaluate a trend of a changing ice load level by comparing a mean of a forecast of a first window to a mean of a forecast of a second window.

14. Ice load monitoring system according to claim 1-8, wherein the stress sensor (1000) comprises a stress sensing element (1002) adapted to be bonded to the structure (1004) to be measured, a casing (1006) covering the stress sensing element (1002), a reference sensing plate (1008) adjacent to the stress sensing element, and support members (1010) to be welded to the structure to be measured and the support members (1010) supporting the casing (1006) and the reference sensing plate (1008).

15. Ice load monitoring system according to claim 14, wherein the stress sensing element and its wire connections are covered with silicon for mechanical protection, the casing is sealed for the potting mass and filled with polyurethane potting mass, epoxy or the like to protect the stress sensor against humidity and corrosion.

16. Ice load monitoring system according to claim 14 or 15, wherein oxsilane is applied on a surface of the structure, where the stress sensing element is to be attached, in order to prevent penetration of water or humidity between the surface and the stress sensing element or the potting mass.

17. Ice load monitoring system according to claim 14-16, wherein the stress sensing element is implemented by a strain gauge, which resistance and output voltage is proportional to strain of the structure or optical gauge with output signal proportional to stain of the structure.