Search results

Because this device has a thicker bottom oxide layer, it should be capable of achieving a smaller gate-drain charge (Qgd) as well as further reductions in switching loss compared to a conventional MOSFET.
The data in Table 1 help to compare the simulated Qgd of the conventional trench SiC-MOSFET and the SiC-MOSFET with the thick bottom oxide trench gate structure.

This leads to higher hardness and strength of the parts, i. e. reduction of the parts weight.
Change of the edge thickness, r = 4 mm, s = 1.8 mm, R = 150 mm, Δh = 1.7 mm: а) f = 0.1; b) f = 0.2; c) f = 0.4 Conclusion The obtained data in combination with the earlier results of the theoretical studies [6, 7, 8] have provided for elaboration of the method for designing the curved edges constrained bend process with elastomer that minimizes the bend radius, increases the part walls thickness, and decreases the edge springing angle.

The results of current simulation were compared with the experimental data those obtained by forging of hollow spur gear forms.
Results and discussion In order to verify the FEM simulation results of DEFORM-3D software for the forming of the hollow spur gears, the theoretical results and experimental data of Choi et. al [10] are compared with the results of the current simulation.
The load predicted by the current FEM simulation is closer to the experimental data than the prediction by upper-bond method [10] for the hollow spur gear forging.
The maximum punch load is almost the same for different initial billet's height. 0.0 10.0 20.0 30.0 40.0 50.0 Height Reduction [ % ] 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 Load [ KN * 10 ] Current simulation Experimental data of Choi et. al [10] Theoretical results of Choi et. al [10] 0.00 50.00 100.00 150.00 200.00 Punch Load (KN) 0.00 2.00 4.00 6.00 8.00 10.00 Punch Travel (mm) Di = 0 Do Di = 0.2 Do Di = 0.4 Do N = Number of teeth M = Module u = Friction factor Di = Inner diameter of billet Do = Outer diameter of billet H = Initial height of billet Current simulation N = 15, u = 0.1 M = 2 Do / H = 1.0 0.00 4.00 8.00 12.00 16.00 Punch Travel (mm) 0.00 200.00 400.00 Punch Load (KN) M = 2 M = 3 M = 4 Current simulation N = 15, u = 0.1 Di / Do = 0.2 H / Do = 1.0 N = Number of teeth M = Module u = Frictoin factor Di = Inner diameter of billet Do = Outer diameter of billet H = Initial height of billet 0.00 3.00 6.00 9.00 12.00 15.00
The load predicted by the current FEM simulation is closer to the experimental data than the prediction by the upper-bond method.

Its core lies in fitting evaluation the implicit prices of each property through the market transaction data and establish the response function model of price and the relationship between each attribute.
In model (1) and (2), eta direct responses the size of house price appreciation when distance nearly one meter, and in model (3), the added value when distance nearly one meter is multiply house prices by eta [4]. 3 An Empirical Study 3.1 Data Acquisition.
Sample size amounted to 50, and ultimately into the model sample for 40, and the data are basically the same point, time sensitive. 3.2 Quantitative Data.
Because the distance here is the actual data, the distance here was each reduced by 1 meter; its value can be increased by 0.1%.If other characteristics are constant, the closer distance the higher housing price.
The impact of the distance loss(y) to the house prices can be expressed by, in the formula, x stand for the original distance; y is the distance reduction.

The mathematical formula description of model GM(1,1) is as follows: Suppose there is a non-negative original data sequence with the variable .
The smaller C is, the smaller S2 is, the greater S1 is, that means the original data variance is large, the residual is concentrated, the swing range is small.
The more dispersive the original data is, the greater the swing range is.
Chose the previous duration of similar projects as references and make them a data sequence.
Use the data sequence as the original data sequence of predictive coupled task set T.

In contrast with PCA, independent component analysis (ICA) employs higher-order statistics (HOS) and exploits inherently non-Gaussian features of the data.
Kurtosis is a common and widely used to estimate the PDF of EMG which is based on the fourth moment of the data (i.e., HOS).
All EMG data were measured by an EMG measurement system (Mobi6-6b, TMS International B.V.) with a gain of 19.5x and a sampling rate of 1024 Hz.
Normalized EMG data (x) with zero mean and unit variance were used for all calculations.
Limsakul, Feature reduction and selection for EMG signal classification, Expert Sys.

All the experimentally obtained data has been used to build up numerical models for adjacent welding simulation.
The material data deposited in the used software do not take into account temperature dependent mechanical and thermal properties.
The data set is bounded below to room temperature.
Measurement data obtained from experiments have been used for numerical model validation.
The calculated welding residual stresses have been compared to experimentally obtained data from X-ray diffraction.

(2) Let r(i) to express the No.i receiving data block vector,v= [v0,v1,…,v(Nb-1)]Tshowed noise vector, then after passing the channels, we haveri=H0si+H1si-1+v LetN×1 dimensional vectory(i) to express the No. i data block after canceling CP, that is yi=RCPri=RCPH0TCPxi+RCPH1TCPxi-1+v
(4) According to the theoretical knowledge of matrix, cyclic matrix can be diagonal zed with the Fourier transformation matrix, that was H=FH⋀F. (5) Did FFT operation of N points with data block after canceling CP, which is equivalent to multiply F in the left of the both ends of formula (4), which is Yi=Fy(i). (6) WhereYi=[YiN,YiN+1,…,Y(iN+N-1)]Tis the No. idimensional vector of the output of FFT module, plug formula (4) and (5) into formula (6), we have: Yi=FHxi+Fv=⋀Fxi+Fv. (7) LetXi≝FxiXiN,XiN+1,…,XiN+N-1T. (8) to be thedimensional frequency vector after N point FFT transformation from No.i data symbol vector.
Hereinto cyclic prefixis added in single carrier modulation as the previous section described, and theauxiliary data block is introduced for channel estimation in receiver to get excitation functionΦn.
We extract auxiliary data block to do channel estimation and obtain excitation function set[Φ1,…,Φn]T.
This system optimizes the existing single carrier system from the two aspects of signal transfer principle as well as math model reduction, we also overcome the multipath fading effectively and improves transferring efficiency.

Many data need to be processed before the pictures are displayed.
Source code of the program is the input of the profiler with the input data.
Profiler is used to analyze the source code to find out the use frequency of the different parts of the data and code.
Before they are in scratchpads, they are in main memory and taken as the common data or code.
Generally, the data and code are placed in main memory.

The diffusivity data from this work are lower than those obtained from the extrapolation of Fe/Pt bulk diffusion couples experiment [3] and those from Fe/Pt bilayers [14].
Figure 5 (b) shows comparison of our data with others' work.
It was found that our diffusivity data are very similar to the Co/Pt epitaxial thin films data.
Temperature dependence of interdiffusion coefficients in Fe-Pt alloy systems in comparison with previous data in (a) Fe-Pt and (b) other thin film systems.
Q and D0 in Fe/Pt thin film in comparison with previous data in other systems.