Relationship Between Moisture And Strain Behavior During The Drying Process Of Ulmus Pumila

The stress strain caused by wood moisture migration is an important factor leading to macro deformation and even cracking of wood,and how to measure wood moisture and strain accurately,in real time,non-destructively,and in situ is an important research content in the field of wood science.In this thesis,stereoscopic microscopy,nuclear magnetic resonance(TD-NMR),digital image correlation(DIC)and hyperspectral imaging(HSI)techniques were used to determine the tangential/radial distribution of water and strain in white elm and its changes in the drying and hygroscopicity processes in situ,the tangential/radial strain interaction of earlywood/latewood during the drying process of white elm was revealed,the relationship between the overall shrinkage of white elm and the surface and internal moisture were clarified,and the relationship between strain and moisture content in various states were discussed.The main conclusions are as follows:1.Based on stereoscopic microscopy,DIC,TD-NMR combined technology,the moisture and strain behavior of white elm during the drying process were measured and analyzed.The results show that the tangential and radial drying shrinkage rates of earlywood/latewood gradually increase with the extension of time,and show peak-valley alternating changes in both tangential and radial directions.The maximum strain of earlywood was at the junction of earlywood and latewood,which was 0.033,while the maximum strain of latewood appeared in the middle part,which was 0.036;the obvious changes in tangential and radial dry shrinkage strain of white elm mainly occurred when the moisture content(MC)was higher than the fiber saturation point(FSP).The dry shrinkage strain increased rapidly at 0.043 in the earlywood stage of drying(4-12 h),and gradually slowed down the increase of shrinkage strain in the middle stage of drying(12-16 h),and the content of free water and bound water had an effect on drying shrinkage.In the later stage of drying(after 16 h),drying shrinkage is mainly dominated by bound water.2.Based on the full-band HSI near-infrared information,the moisture during the drying process of white elm was rapidly detected.The results show that there is a strong correlation between the reflectivity of spectral information and the moisture content.B y constructing partial least squares(PLS),support vector machine(SVM),reverse artificial neural network(BPNN)and random forest(RF)regression model,the RF algorithm regression prediction model with wood moisture content was selected,with R_c~2=0.9232,RMSEC=0.0801,R_p~2=0.9112,RMSEP=0.0853,and the model prediction ability was improved by 0.0877 and 0.1236 compared with the original Raw-RF model R_c~2 and R_p~2,respectively.At the same time,the best preprocessing method,i.e.convolutiona l smoothing(SG)combined with standard normal transformation(SNV),is screened.3.Based on the characteristic band HSI near-infrared information,the moisture during the drying process of white elm was rapidly detected.Competitive adaptive resampling(CARS),continuous projection algorithm(SPA)and non-information variable exclusion(UVE)methods were used to extract the characteristic bands,and the moisture prediction model of PLS and RF in the characteristic band was constructed.The results show that SPA-RF is the optimal model,where R_c~2=0.9564,RMSEC=0.0647,R_p~2=0.9555,RMSEP=0.0682,at the same time,according to the problems in the quantitative analysis of RF model,the algorithm variable is improved,and the accuracy of the moisture content prediction model of white elm is improved after optimization,where R_c~2=0.9968,RMSEC=0.0175,R_p~2=0.9932,RMSEP=0.0262 with predictive power close to 1;This model was applied to visualize the moisture distribution of white elm.4.Based on the joint technology of HSI and DIC,the moisture and strain behavior during the drying process of white elm were measured and analyzed in real time.The visualization data show that:(1)With the increase of drying time,the edge shrinkage strain gradually decreased less than the dry shrinkage strain of the middle part,and the dry shrinkage strain variable of the middle part gradually increased,and the maximum dry shrinkage gradually moved to the middle with the drying process.(2)When the average moisture content of white elm is 32%,the moisture content of latewood is much higher than the fiber saturation point,while the moisture content of earlywood is already lower than the fiber saturation point.The moisture content of earlywood and latewood transition area was significantly lower than that of latewood,and the moisture migration rate of earlywood was higher than that of latewood.With the increase of drying time,the tangential radial dry-shrinkage strain variable of morning and evening materials gradually increased,and the change law of peak-trough alternation appeared from the edge of the selected specimen to the center part.(3)There is obvious difference between the dry shrinkage strain of earlywood and latewood in the tangential direction,and the maximum strain ratio value of earlywood and latewood is 0.89;in the radial direction,the late wood drives the earlywood to dry shrink together,and the maximum dry shrinkage strain of earlywood always appears at the junction of morning and evening wood in the drying process,while the maximum dry shrinkage strain variable of latewood appears in the middle part of latewood.The maximum strain ratio of earlywood to latewood was 0.53,and the tangential dry shrinkage strain was about twice that of the radial dry shrinkage strain,and the radial dry shrinkage was significantly affected by earlywood and latewood.