scholarly journals Neural Networks Applied to the Wave-Induced Fatigue Analysis of Steel Risers

2018 ◽  
Vol 2018 ◽  
pp. 1-16 ◽  
Author(s):  
João P. R. Cortina ◽  
Fernando J. M. de Sousa ◽  
Luis V. S. Sagrilo
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Cross Sections ◽  
Computational Cost ◽  
Offshore Structures ◽  
Fatigue Analysis ◽  
Element Analysis ◽  
Long Time ◽  
Wave Induced ◽  
Hybrid Methodology

Time domain stochastic wave dynamic analyses of offshore structures are computationally expensive. Considering the wave-induced fatigue assessment for such structures, the combination of many environmental loading cases and the need of long time-series responses make the computational cost even more critical. In order to reduce the computational burden related to the wave-induced fatigue analysis of Steel Catenary Risers (SCRs), this work presents the application of a recently developed hybrid methodology that combines dynamic Finite Element Analysis (FEA) and Artificial Neural Networks (ANN). The methodology is named hybrid once it requires short time series of structure responses (obtained by FEA) and imposed motions (evaluated analytically) to train an ANN. Subsequently, the ANN is employed to predict the remaining response time series using the prescribed motions imposed at the top of the structure by the floater unit. In this particular work, the methodology is applied aiming to predict the tension and bending moments’ time series at structural elements located at the top region and at the touchdown zone (TDZ) of a metallic riser. With the predicted responses (tensions and moments), the stress time series are determined for eight points along the pipe cross sections, and stress cycles are identified using a Rainflow algorithm. Fatigue damage is then evaluated using SN curves and the Miner-Palmgren damage accumulation rule. The methodology is applied to a SCR connected to a semisubmersible platform in a water depth of 910 m. The obtained results are compared to those from a full FEA in order to evaluate the accuracy and computer efficiency of the hybrid methodology.

Time Series Simulations of Wide-Band Spectra for Fatigue Tests of Offshore Structures

1984 ◽  
Vol 106 (4) ◽  
pp. 466-470 ◽  
Author(s):  
N. K. Lin ◽  
W. H. Hartt
Keyword(s):  
Time Series ◽  
Fatigue Testing ◽  
Wide Band ◽  
Simulation Method ◽  
Offshore Structures ◽  
Fatigue Tests ◽  
Time Series Model ◽  
Long Time ◽  
Band Spectra ◽  
Time Series Simulation

A time-series simulation method, based on the principle of time series modeling for dynamic systems, is used to reproduce a wide-band stress history from a prescribed stress spectral model for fatigue testing of offshore structures. The optimization procedures and stability of the time series model for the prescribed spectrum are presented and discussed. The optimization procedures are developed on the basis of the Levison-Durbin algorithm, which usually produces a stable time series model if the order of the time series model is even. An example is presented to demonstrate the applicability of the proposed method to long-time, high-cycle fatigue testing.

Closed-Form Spectral Fatigue Analysis for Compliant Offshore Structures

1988 ◽  
Vol 32 (04) ◽  
pp. 297-304
Author(s):  
Y. N. Chen ◽  
S. A. Mavrakis
Keyword(s):  
Closed Form ◽  
Fatigue Damage ◽  
Computational Cost ◽  
Power Spectra ◽  
Offshore Structures ◽  
Fatigue Analysis ◽  
Cumulative Fatigue Damage ◽  
High Computational Cost

Spectral fatigue analysis frequently has been applied to welded joints in steel offshore structures. Although, on the theoretical basis, the spectral formulation holds certain advantages over other formulations such as the discrete, design wave type of analysis, numerical methods developed on that basis generally suffer from the shortcomings of lack of precision and high computational cost. This paper synthesizes the uncertainties resulting from modeling errors that are regarded heretofore as unavoidable in an analysis. Such errors are traced to the approximations introduced in handling of wave data, in numerical integration of the response power spectra, and in the integration that leads to the determination of cumulative fatigue damage. To each of these sources of modeling error, a transparent, closed-form method is proposed which not only eliminates the potential errors but, surprisingly, improves the computational efficiency many times. The sensitivity of fatigue damage upon the variability of the shape parameter due to variability of wave environment for the so-called simplified analysis utilizing an idealized mathematical long-term probability density function (for example, the Weibull distribution) is also discussed.

Efficient Mooring Line Fatigue Analysis Using a Hybrid Method Time Domain Simulation Scheme

2013 ◽  
Author(s):  
Niels Hørbye Christiansen ◽  
Per Erlend Torbergsen Voie ◽  
Jan Høgsberg ◽  
Nils Sødahl
Keyword(s):  
Hybrid Method ◽  
High Speed ◽  
Computational Cost ◽  
Response Prediction ◽  
Fatigue Analysis ◽  
Training Data ◽  
Mooring Line ◽  
Element Analysis ◽  
Line System ◽  
Drift Motion

Dynamic analyses of mooring line systems are computationally expensive. Over the last decades an extensive variety of methods to reduce this computational cost have been suggested. One method that has shown promising preliminary results is a hybrid method which combines finite element analysis and artificial neural networks (ANN). The present study presents a novel strategy for selecting, arranging and normalizing training data for an ANN. With this approach one ANN can be trained to perform high speed dynamic response prediction for all fatigue relevant sea states and cover both wave frequency motion and slow drift motion. The method is tested on a mooring line system of a floating offshore platform. After training a full fatigue analysis is carried out. The results show that the ANN with high precision provides top tension force histories two orders of magnitude faster than a full dynamic analysis.

scholarly journals PREDICTION OF WAVE-INDUCED SEAFLOOR MOVEMENTS

1976 ◽  
Vol 1 (15) ◽  
pp. 93 ◽  
Author(s):  
Leland Milo Kraft ◽  
David James Watkins
Keyword(s):  
Cyclic Load ◽  
Offshore Structures ◽  
Foundation Design ◽  
Element Analysis ◽  
Wave Statistics ◽  
Strain Behavior ◽  
Potential Depth ◽  
The Stability ◽  
Wave Induced ◽  
Soil Movements

Foundation design for offshore structures in areas where wave-induced bottom pressures cause submarine mud slides requires a knowledge of the potential depth of slide and of the magnitude and distribution of soil movements below the slide. Several methods have been developed to evaluate the stability of the seafloor due to wave-induced bottom pressures. These methods are reviewed and an improved procedure is presented. This procedure makes use of finite element analysis and combines in a rational manner oceanographic information on wave statistics with stress-strain behavior of soils under cyclic load conditions in order to evaluate the effects of a given storm history on the behavior of submarine sediments.

LEARNING THE RULE OF A TIME SERIES

1992 ◽  
Vol 03 (02) ◽  
pp. 167-177 ◽  
Author(s):  
I. Ginzberg ◽  
D. Horn
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Random Noise ◽  
Noise Amplitude ◽  
Quadratic Map ◽  
Correct Rule ◽  
The Future ◽  
Sunspot Data ◽  
Long Time ◽  
The Difference

Neural networks can be trained to predict the next value of a time series on the basis of its preceding values. We try to find out how well such a network approximates the rule which underlies the series. For this purpose, we study the net-sequence, which is a long time series generated iteratively by the network. We introduce a new measure: the difference between the distributions of function values in the data and the net-sequence. We demonstrate its usefulness on the problem of the chaotic quadratic map. Adding random noise to the series we find, by using this tool, that the networks can approximate well the correct rule only if the noise amplitude is very small. Applying the new measure as a weak constraint in the problem of sunspot data, we see that it correlates well with the ability of the network to predict several time steps into the future.

scholarly journals Multi-Step Solar Irradiance Forecasting and Domain Adaptation of Deep Neural Networks

Energies ◽  
2020 ◽  
Vol 13 (15) ◽  
pp. 3987
Author(s):  
Giorgio Guariso ◽  
Giuseppe Nunnari ◽  
Matteo Sangiorgio
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Solar Irradiance ◽  
Short Term Memory ◽  
Domain Adaptation ◽  
Model Parameters ◽  
Network Architectures ◽  
Long Time ◽  
Unique Predictor ◽  
Hidden Neurons

The problem of forecasting hourly solar irradiance over a multi-step horizon is dealt with by using three kinds of predictor structures. Two approaches are introduced: Multi-Model (MM) and Multi-Output (MO). Model parameters are identified for two kinds of neural networks, namely the traditional feed-forward (FF) and a class of recurrent networks, those with long short-term memory (LSTM) hidden neurons, which is relatively new for solar radiation forecasting. The performances of the considered approaches are rigorously assessed by appropriate indices and compared with standard benchmarks: the clear sky irradiance and two persistent predictors. Experimental results on a relatively long time series of global solar irradiance show that all the networks architectures perform in a similar way, guaranteeing a slower decrease of forecasting ability on horizons up to several hours, in comparison to the benchmark predictors. The domain adaptation of the neural predictors is investigated evaluating their accuracy on other irradiance time series, with different geographical conditions. The performances of FF and LSTM models are still good and similar between them, suggesting the possibility of adopting a unique predictor at the regional level. Some conceptual and computational differences between the network architectures are also discussed.

Artificial Neural Networks Applied to Flexible Pipes Fatigue Calculations

2015 ◽  
Author(s):  
Victor Chaves ◽  
Luis V. S. Sagrilo ◽  
Vinícius Ribeiro Machado da Silva ◽  
Mario Alfredo Vignoles
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Computational Cost ◽  
Local Analysis ◽  
Flexible Pipe ◽  
Training Time ◽  
Flexible Pipes ◽  
Dynamic Analyses ◽  
Artificial Neural ◽  
Hybrid Methodology

Flexible pipes play an important role in offshore oil exploitation activities nowadays. However, time-domain flexible pipe irregular wave dynamic analyses are extremely computational expensive. One of the various existing methods to reduce computational costs in dynamic analyses is the hybrid methodology that combines dynamic Finite Element Analyses (FEA) and Artificial Neural Networks (ANN). This paper presents a novel application of this methodology for flexible pipes fatigue calculations. In order to decrease computational cost involved in these analyses the proposed hybrid methodology aims to predict tension and curvatures in the bend stiffener region. Firstly using short FEA simulations to train the ANN, and then using only the ANN and the prescribed floater motions to get the rest of the response histories. With the predicted tension and curvatures, a local analysis is applied to calculate stresses in tensile armour wires and the corresponding fatigue lives. To evaluate the optimal ANN a sensibility study is developed for some key parameters as: training time length, neurons on hidden layer and delay length. A full FEA is also performed in order to evaluate the accuracy of the proposed hybrid methodology, comparing both full FEA flexible pipe fatigue results and those obtained using the hybrid methodology.

Sign in / Sign up

Export Citation Format

Share Document