Search results

The effects of bending force, reduction and transverse friction on the strip profile and edge drop have been discussed in this paper.
The effects of bending force, reduction and transverse friction on the strip profile and edge drop have been analyzed in detail.
The edge drop is expressed as hIhJeh −= , where Assuming: Start Input Data 1Cvx =∆ 0=∆ yv Set for the edge row of slabs . 0=yσ )( nj = Solving Eq. 1 and Eq. 3 to obtain the pressure distribution for row( j ) obtain neutral point j<1 Calculating: by Eq. 2. )1( −jyσ No Calculating the strip velocities according to Eq. 6 and Eq. 7 Correcting the frictional coefficient by Eq. 4 and Eq. 5, and flattening roll radius by Hitchcock's formula ε<∆R Output results End Yes No Yes Fig. 4.
The edge drop increases with reduction as can be seen in Fig. 14. 16 18 20 22 24 26 28 30 6 7 8 9 10 11 12 13 µ=0.06 Fb=0 C2=0.1 Edge Drop, µm Reduction Rate, % Fig. 14.
The calculated results indicate that the edge drop decreases with increasing bending force, decreasing friction and reduction.

Because best estimators are needed in uncertainty evaluation[3], Kolmogorov and Grubbs criterion, at 0.95 confidence level, are used in the data analyze[4].
It’s obviously that tensile stress, both yield stress and broken stress, of specimens which under torsional pre-strain becomes greater, than those ones which are not in torsion or the twisting angle equals to zero, as while as percent elongation and percent reduction area become smaller.
(3) related standard uncertainty from data acquisition[7] Type B related uncertainty of a valid data acquisition computer is given.
Conclusion (1) Tensile stress, both yield stress and broken stress, of specimens which under torsional pre-strain becomes greater, than those ones which are not in torsion or the twisting angle equals to zero, as while as percent elongation and percent reduction area become smaller
JJF1103-2003, Evaluation for Computerized Data Acquisition Systems of Universal Testing Machines[S], Beijing: China Metrology publishing House, 2003,9.

When in reasoning and decision-making of the new data using the multiple level rules set, match the information of the new data with the rule of its corresponding node.
With the minimal decision rule, we can make decision to the new data.
In order to obtain classified ability to the noise and abnormal data, improve the capacity in the classification of new data, and give the maximum possible solution to problems as much as possible when all the information is not complete,this article adopts the following steps to extract multiple level rules set model: l Write original data in the form of a decision table.
After the above steps, extract the rules of various levels from the original data.
Then match the current data attribute values with rule set of the node.

Preliminary comparative tests show a significant reduction of tool wear when machining 4140 steel in minimum quantity lubrication.
Their relationship and how data compared with published information follow.
Each data block is 10x5mm, 300 kPa nozzle pressure, 128 m/s nozzle speed, max: more than 35 m/s. 5mm Fig. 5: Droplet density and flow profile by particle image velocimetry. 480 kPa input pressure, 8 strokes/min.
After 75 consecutive passes, the flank wears and wear fraction for tools machining in air, flood, and MQL are 0.167 mm, 0.094 mm (43% reduction), and 0.055 mm (67% reduction).
Although not quantifying, significant reduction in crater wear and nose wear are seen for tool machining in MQL (Fig. 6).

Incorrect cooling, however, can lead to heat unbalance in the die casting tool, thus causing the destruction of casting quality, reduction of the lifetime of die casting tool and irreversible production-reduction.
Thermal data is given in tables.
For each calculation only 1 parameter was changed, for the others general workshop data were applied. 2.
Given and calculated data are presented in table 4.
In production these are the most varying parameters and this is why it is justified to continuously monitor them with data collection systems [5-6]. 4.

There are two ways to establish a sound data base: Manual data gathering, which is always associated with great effort or the usage of “big data”.
There are a few approaches which investigate data aggregation in particular as a type of data mining.
Data model.
Figure 3: Data model Technical model.
At this point the aim is the reduction of the model’s complexity.

On the other hand, slope data slices at four locations could interpret landslide potential well.
In this study, a landslide risk was analysed using the Digital Elevation Model (DEM) data, which represents the topography, slope, hill shade, aspect, and curvature data related to landslide susceptibility  [17].
Data and Methodology The DEMNAS data is a spatial dataset that contains elevation or altitude information with a resolution of 0.27 arc-second provided by the Indonesia government, a finer resolution compared to the Shuttle Radar Topography Mission (SRTM) data, which has a three arc-second resolution as a global digital elevation data.
This study analysed the slope, elevation, aspect, and curvature data.
Result and Discussion Analysis of DEM data Current DEM data provide topographic information and hold potential for application in landslide potential analysis studies  [23].

Secondly, the two wafers demonstrate different reduction time constants.
The time constant is determined via an exponential decay fit to the data.
An exponential relation of C(t) = Co[1-exp(-t/τ)] is used to fit the data.
The solid lines are model fitting [11] to the experimental data.
The area shaded in grey represents the amount of Fe gettered by the GB during the annealing process occurred in between the red and green data.

The proposed technique is tested on UCI data sets.
In the wrapper mode, the feature selection method evaluates the candidate features using the learning algorithm that is to ultimately be applied to the data.
The objective it to compare the classification accuracy, data reduction capability and the robustness of the margin based feature selection algorithms.
In order to test the robustness of the algorithms, we generate noisy data by randomly relabeling class labels of some samples, and then the relabeled samples are considered as noisy data.
It means feature selection based on BBL using FWL-L1 is better than the others about classification ability in raw data.

Figure 1 represents the flow chart for the research methodology and the way of approach to the problem statement and how data were collected.
Current Process Chart Data collection consists of time study for each operation in the manufacturing process.
The data pertaining to part manufactured in the shop floor is given in Table 1.
The line balancing diagram for P5 model is drawn using the collected data as shown in Figure 7 with cycle time for each process and TAKT time.
Data Analysis TAKT Time Calculation: TAKT time is calculated for both P3 and P5 model as given below.