Search results

Such through cracks were monitored during their propagation along the specimen width, in order to have available for the simulation a realistic initial scenario and experimental propagation data useful for the correlation with the simulated crack path and growth rates.
In order to develop an effective riveted reinforcement methodology it is important to be able to accurately determine the complex stress fields created by the doublers as well as the resulting reduction in the stress intensity factors (SIFs) of a crack on the skin.

Most recently, the technique has been developed to measure the SIFs during the fatigue crack growth [3] and new techniques for data noise reduction in this application have been suggested [4].
The X data image is the data that is in-phase with the loading cycle and the Y data image is the data that is out-of-phase with the loading cycle.
If Eq. 4 is valid then the Y data should be zero.
The phase angle data shown in Fig. 2 gives the phase shift from the reference datum for the Y-data.
However, the out-of phase data given in Fig. 3 shows a marked change in the shape of the Y-data depending on the notch depth.

On the contrary there are a huge amount of Unlabeled Data.
While the second stage is the iterative convergence stage, during this stage, SUCS will continuously classifies the unlabeled data, and using the classified unlabeled data and the original labeled data to train itself again.
Most of the traditional machine learning methods rely on the labeled data excessively, and neglect the effect of the using of the unlabeled data.
But they just proposed a framework which integrates the semi-supervised learning and UCS, its semi-supervised learning component learns using both the labeled data and the unlabeled data, once the labeled data reaches enough confidential level, they will be remove from unlabeled data to the labeled data for the training of UCS component.
The major difference is that SUCS is an iterative algorithm, and it can work by use of both the Labeled Data and the Unlabeled Data.

When this is achieved, there will be reduction in defect, which will result in more profits.
In case of the latter, a new design concept needs to be developed; this new design could be design from scratch (Creative Design) or re-design from a datum design (Incremental Design).
The bench test and field data (design-test-fix-retest) also result in high development cost, longer time to market, lower product quality and marginal competitive edge.
Thus, DFSS strategy subject designs to noise factors through appropriate design parameters; this helps to attack design vulnerabilities and integrate tools and methods for their elimination and reduction.

Storage and use of specific data are concentrating in databases.
Data collection and administration is concentrating in databases.
Manufacturing data, structures, relations between components are centralized in data types that describe product configuration, using configuration management in a product data management system, PDM.
The problem of arriving at the optimum levels is approached by researchers and engineers with the aim of cost reduction.
Data type for inserts is presented in Table 3.

Langmuir and Freundlich isotherms were studied and the equilibrium adsorption data was found to fit well with the Langmuir isotherm model.
The experimental data fitted better to pseudo-second-order kinetic model than pseudo-first-order kinetic model, based on high R2 for pseudo-second-order kinetic model.
The Langmuir isotherm model gave better fit to the experimental data than Freundlich isotherm model.
Langmuir isotherm model fitted the experimental data better than Freundlich isotherm model.
The pseudo-second-order kinetic model fitted the experimental data well, indicating chemical adsorption of Ni(II).

(iii) The phase noise power spectrum can be calculated from the delta-sigma TDC output data using FFT.
Delay τ in the delta-sigma TDC is set to 200ns, and the number of the delta-sigma TDC data we use is 4096.
b) Phase variation of 50kHz Fig.7 shows the FFT analysis results of the delta-sigma TDC output data with phase noise of 50kHz (in eq.(4), ω1/(2π)=50kHz).
It takes 66ms measurement time to obtain 65,536 TDC output data with 1MHz clock.
Frobenius, RF Measurements for Cellular Phones and Wireless Data Systems, Jon Wiley & Son.

Data was collected through questionnaire in tsunami-impacted coastal communities within the area of Kuala Muda in the state of Kedah, Malaysia.
Data was collected for this study through survey questionnaire, interviews, documents analysis and field observations.
For this paper, the focus is more on the data collected through the questionnaire.
A total of 211 respondents from eight villages devastated by the 26th December 2004 tsunami took part in the data collection exercise.
With reference to the availability of data, this relatively low score may be attributed to the preference of the public in receiving information that is simple and easy to digest as compared to data that is complicated to understand.

The results show a good ability of the model to fit the measured data.
Initially, the modeling approach is based on empirical relationships usually elaborate on experimental data in order to identify any significant correlation between the subsurface outflow and the rainfall depth and duration [10].
In all distributions the modeled data are close to the observed data.
Thetemporal distribution obtained for observed and modeled data is fairly close.
In particular, the results showed that the model over-estimated the initial runoff with respect to observed data.

Third, GRNN can handle linear and nonlinear data.
, have a forecasting effect when be short of sample data.
The input layer merely serves as an input data buffer and does not perform any processing.
Hence, there are I pattern neurons running in parallel if the training data set consists of a total of i = 1, 2, … , I samples.
This paper choose 11 parameters illustrated as Table 1 according to the really data.