Research On Failure Prediction Method Of Boiler Welded Joint Based On Grey Markov Chain

Boiler equipment served in the bad condition of environment for a long time,in the process of operation,equipment potential safety hazard mainly in welding parts,weld defects may lead to the failure of welded structure,fracture,serious accidents such as leak accident and explosion.The safety and reliability of boiler pressure vessel welded joint is the important research direction.Cold crack in welded joint failure events,occur in the weld cooling process or after a period of time after cooled to room temperature,cold crack can occur immediately after welding,some can also be extended to a couple of hours,days,weeks or even longer to occur,due to the delay,delayed exacerbated the leak,has increased the risk of this kind of crack,because its mechanism is very complicated,difficult to qualitative quantitative analysis.In this paper,based on the prediction of the failure probability of welded joint caused by cold crack,the unbiased grey markov chain model of particle swarm is applied to carry out a deeper research around the prediction method.Kunming boiler co.,LTD as the background,in this paper first expounds the boiler welded joint failure of the seriousness of consequences,to briefly described the current main methods of prediction,and butt joint failure cause,manufacturing defects and other problems are expounded,and through to the welding specimen NDT,mechanical property test and electron microscope scanning,analyzed the welding failure mechanism of the pressure parts and many factors.Secondly,through the boiler pressure vessel welding interface of fuzzy grey relational FTA,combined with the mechanical properties test and NDT testing,widely used in boiler set of 12 cr1 movg corner joint with welding process parameters on welding quality reliability evaluation,get cold crack in welded joint failure probability statistics of the incident,because this method needs a lot of experimental data,more difficult to operate.Considering the characteristics of welding joint failure caused by cold crack,there is a large deviation when using grey prediction model tofit the exponential sequence.When the markov chain prediction model whitens the grey interval,there exists the problem of whitening coefficient selection.Therefore,based on this theory,the unbiased grey model is established to solve the problem of large deviation in the traditional grey prediction model.At the same time,the introduction of particle swarm unbiased gray markov chain model,through calculating the optimum bleaching coefficient values,solve the problem of markov chain prediction model albino coefficient selection,the establishment of a new prediction model,to overcome the limitations of traditional methods rely on a large number of experimental testing data,to simplify the process of the safety assessment of welded joint,proactive,not only improves the prediction precision,but also to analyze this kind of welding components failure events forecast.