Characteristics Of Steam Condensation Flow And Heat Transfer In Multi-Channel Cylinder Dryers

The drying section of the paper machine is the section with the highest energy consumption in paper production,due to the heat transfer efficiency of the core component,the drying cylinder,is difficult to improve.And the accumulation of steam condensate in the drying cylinder seriously hinders the effective transfer of heat.As an innovative type of cylinder dryer,multi-channel cylinder dryers(MCD)can effectively solve the problem of severe accumulation of condensate inner dryer by continuously pushing out condensate with the help of steam limited in small channels,thus significantly improving the heat transfer efficiency.At present,some problems have not been solved thoroughly,including the unclear key factors belonging to numerous influencing factors affecting the heat transfer characteristics of MCD,the incomplete characterization of the total heat transfer coefficient(HTC),the inaccurate two-phase flow resistance model,the unquantified key parameters of two-phase flow and the non-optimal geometric shape of the channel section.Therefore,an experimental visualization platform with replaceable channels for steam condensation flow and heat transfer for MCD’s research was designed and built,based on a macro-scale theory of two-phase fluid flow and heat transfer.Firstly,the characteristics of steam condensation heat transfer and flow in different cross-sectional channels were studied through visualization experiments.Then the prediction method of total HTC and the influence of critical parameters on total HTC were studied based on machine learning technology.Finally,the theoretical prediction model of two-phase flow pressure drop was studied based on two-phase fluid dynamics.The key parameters and prediction models of flow patterns were analyzed by digital image processing technology.The main research contents are as follows:(1)Based on a macro-scale theory of two-phase fluid flow and heat transfer and the multi-channel dryer as the basic model,an experimental visualization platform with replaceable channels for steam condensation flow and heat transfer was designed and built.The three basic systems of the experimental platform provided a reliable experimental basis for visualization research,due to accurately collecting the data of condensation flow and heat transfer characteristics in the channel in real-time,.(2)Based on machine learning technology,the influence of different parameters on the total heat transfer coefficient was discussed,and the total HTC was comprehensively characterized.Firstly,the influence of many factors on the steam condensation temperature distribution,heat load parameters,and heat exchange in the horizontal channel was studied experimentally.Then,four machine learning models were established by machine learning technology,and the ANN model with better prediction results of HTC was selected.Four parameters(the convection HTC hw/mass flow rate m/Nusselt number Nu/Reynolds number Re of cooling water)with the greatest impact on the total HTC were further optimized.Moreover,the importance of the relevant parameters of the cooling medium on the characterization of the total HTC was clarified.Finally,the Shah correlation,with its mean absolute percentage error(MAPE)is 28.6%,was determined as the prediction model of steam condensation HTC of horizontal channels of multi-channel dryers,by comparison of the experimental value of steam condensation HTC with the predicted value of four correlations suitable for this study.The total HTC calculated based on the Shah condensation heat transfer correlation was further compared with the total HTC predicted by the machine learning model.It was found that the total HTC can be quickly and accurately predicted by the supervised machine learning model based on the Bayesian optimization algorithm.The MAPE between SVM,ANN,linear SVM,RF and experimental data was 0.013%,1.16%,0.012%and 6.02%,respectively.(3)In order to study the steam condensation heat transfer characteristics in channels with different cross sections,the heat transfer characteristics of steam flow and condensation in horizontal rectangular,U-shaped,and triangular cross-sectional channels under non-adiabatic conditions were conducted experimentally,By analyzing the influence of different parameters of heat transfer and cooling medium on the condensation HTC of channels with three cross-sectional shapes,it was found that the increase of steam mass flux and cooling water mass flow rate led to better heat transfer performance for all three shape channels.Furthermore,The heat transfer performance of steam condensation in channels with different cross-sectional channels was evaluated,the best matching relationship between heat transfer and cooling medium in channels with different cross-sections was proposed,and the better heat transfer performance of triangular cross-sectional channels was determined under the appropriate steam mass flux(35 kg·m-2·s-1)and cooling load.(4)Based on the basic principle of two-phase fluid mechanics,the flow resistance characteristics of steam inside different cross-sectional channels were studied.Firstly,the theoretical models of separated phase flow and homogeneous phase flow,which are suitable for studying the pressure drop of steam condensation two-phase flow,were established.Then,the pressure drop of steam condensation flow in the channel was studied experimentally in order to evaluate the two theoretical models,and it was clear that the Chisholm B correlation of the separated phase flow model was more suitable as a prediction model when the steam mass flux was 20～40kg/(m2·s).The influence of the parameters of heat transfer and cooling medium on the pressure drop of two-phase flow in horizontal non-rectangular channels was studied experimentally.At last,the two-phase flow resistance characteristics of steam condensation flow in horizontal non-rectangular channels were evaluated,and it was determined that the steam condensation flow in triangular cross-section channels could provide the most stable annular flow.(5)In order to study visually the flow pattern of steam condensation two-phase flow in different channels,the critical parameters of two-phase flow were quantified,and a prediction model of stable flow pattern was proposed.The distribution of gas-liquid two-phase interface in horizontal channels with different cross section shapes under non-adiabatic conditions was observed visually.The flow patterns of two-phase flow in U-shaped and triangular cross-sectional channels were obtained accurately,and the influence of flow and heat transfer parameters on the flow pattern transition was analyzed,finally,it is found that the triangular channel shows a more stable flow state.In quantifying the section void fraction,with the help of digital image processing technology,the image binarization for the annular flow pattern of steam condensation in the triangular-sectional channel was completed,and the void fraction α was measured online,and the prediction model was determined as the R-A model(MAPE was 11.9%).Furthermore,the two-phase flow pattern diagram of the triangular cross-sectional channel was evaluated,and the Cavallini flow pattern diagram model was proposed to predict the annular flow pattern.Finally,it was found that the flow law of the triangular channel was different from that of other channels,and the flow and heat transfer advantages of the triangular channel were determined.This research provided not only a theoretical basis based on heat transfer and fluid mechanics for optimizing the cross-section parameters design of MCD,but an idea for improving heat transfer efficiency,energy saving and emission reduction in the fields of paper making,petrochemical industry,energy power and other fields involving steam condensation heat transfer process.Meanwhile,a research basis for the development of multidisciplinary interdisciplinary research technologies such as heat flow testing,machine learning and two-phase flow detection technology was also provided.