• The Journal of the Korea Contents Association
• /
• v.8 no.11
• /
• pp.57-64
• /
• 2008

The data grid computing provides geographically distributed storage resources to solve computational problems with large-scale data. Unlike cache replacement policies in virtual memory or web-caching replacement, an optimal file replacement policy for data grids is the one of the important problems by the fact that file size is very large. The traditional file replacement policies such as LRU(Least Recently Used), LCB-K(Least Cost Beneficial based on K), EBR(Economic-based cache replacement), LVCT(Least Value-based on Caching Time) have the problem that they have to predict requests or need additional resources to file replacement. To solve theses problems, this paper propose SBR-k(Sized-based replacement-k) that replaces files based on file size. The proposed policy considers file size to reduce the number of files corresponding to a requested file rather than forecasting the uncertain future for replacement. The results of the simulation show that hit ratio was similar when the cache size was small, but the proposed policy was superior to traditional policies when the cache size was large.

• Journal of Ocean Engineering and Technology
• /
• v.20 no.5 s.72
• /
• pp.42-48
• /
• 2006

This paper shows theresults of centrifuge model experiments to investigate the behavior of a replacement method in dredged and reclaimed ground. For this experimental work, centrifuge model tests were carried out to investigate the behavior of a replacement method in soft clay ground. Basic soil property tests were performed to find the mechanical properties of clay soil sampled from the southern coast of Korea, which was used for the ground material in the centrifuge model tests. The reconstituted clay ground of the model was prepared by applying reconsolidntion pressure in a 1 g condition with a specially built model container. Centrifuge model tests were carried out under the artificially accelerated gravitational level of 50 g. Replacement material of lead with a certain degree of angularity was used and placed until the settlement of the replacement material embankment reached a state of equilibrium. Vertical displacement of the replacement material was monitored during tests. The depth and shape of the replacement, especially the slope of the penetrated material and the water content of the clay ground were measured after finishing tests. Model tests for investigating the stability of an embankment after backfilling were also performed to simulate the behavior of a dike treated with replacement and backfilled with sandy material. As a result of the centrifuge model test, the behavior of the replacement, the mechanism of the replacement material being penetrated into clay ground, and the depth of the replacement were evaluated.

• Journal of Chest Surgery
• /
• v.25 no.12
• /
• pp.1556-1562
• /
• 1992

Forty two consecutive patients who had had valve replacement with St. Jude Medical prosthesis were studied on a view point of intravascular hemolysis. Patients were consisted of 14 mitral valve replacement, and 7 aortic valve replacement, and 21 double, mitral and aortic, valve replacement. Serum LDH, indirect bilirubin, GOT, hemoglobin levels and ret-iculocyte count were pursued in postopeative 1st day, 3rd day, 7th day, 14th day and 21th day. Postoperatively, all patients were not detected paravalvular leakage on the ech-ocardiographical study. The patients with double valve replacement revealed higher levels of LDH on postopeative 14th day[P<0 05] than those with single valve replacement. Among the patients with single valve replacement, the patients with aortic valve replacement revealed slightly higher levels of entire postopeative data, but considered insignificant. There was correlation between the severity of hemolysis and the size of replaced aortic valve. In the postoperative LDH levels, the patients with small sized-aortic valve[less than 21mm in diameter] replacement revealed higher levels of postoperative 3rd day, 7th day and 14th day than those with large size[more than 23mm in diameter]. The patients with high level LDH of greater than 800 WU /L on postoperative 7th day were 61.9%[26 of 42]. The high LDH frequency of DVR was 71.4%[15 of 21], MVR 50.0%[7 of 14] and AVR, 57.1%[4 of 7]. The level of LDH declined gradualiy thereafter through postoperative 3 weeks. In conclusion, intravascular hemolysis after prosthetic valve replacement was dependent on position of valve replacement and size of valve. And this study supports the conventional valve selection and usage in our hospital. The patients with subclinical hemolysis after valve replacement should be placed on a close observation.

• Journal of Chest Surgery
• /
• v.24 no.3
• /
• pp.265-270
• /
• 1991

Since advent of the prosthetic cardiac valve replacement, much efforts for accurate assessing value function in-vivo have been attempted. To evaluate the postoperative functional and morphological status of the replaced cardiac valve prosthesis, 33 patients with valve replacement were studied by transthoracic and transesophageal 2-dimensional echocardiac imaging as well as by color Doppler flow velocity imaging. Twenty four patients had mitral valve replacement. 6 patients had aortic valve replacement and 3 patients had both mitral and aortic valve replacement. There were 34 mechanical and 2 biological prosthesis. Comparing to transthoracic echocardiography, transesophageal approach showed transvalvular regurgitant jet flow amid the prosthetic mitral valve ring during. systole and much clear visualization of cardiac chamber behind prosthesis which could give shadowing effect to ultrasound beam. According to the quantitative grading by the length and area of mitral regurgitant flow, 24 out of 27 mitral valves revealed mild degree regurgitation considered as physiological after prosthetic bileaflet valve replacement and the other 3 valves including 2 biological prosthesis had moderate degree regurgitation which was regarded as pathologic one. 2 cases of left atrial thromboses and 1 case of paravalvular leakage which were not visible by transthoracic approach were identified by transesophageal echocardiography in patients with mitral valve replacement and patients with aortic valve replacement respectively. We conclude that in patients with prosthetic mitral valve replacement, transesophageal 2-dimensional imaging with color Doppler can suggest reliable information beyond that available from the transthoracic access even though it gives patient some discomfort to proceed.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.45 no.3
• /
• pp.78-89
• /
• 2022

The purpose of this study was to propose useful suggestion by analyzing preventive replacement policy under which there are minor and major failure. Here, major failure is defined as the failure of system which causes the system to stop working, however, the minor failure is defined as the situation in which the system is working but there exists inconvenience for the user to experience the degradation of performance. For this purpose, we formulated an expected cost rate as a function of periodic replacement time and the number of system update cycles. Then, using the probability and differentiation theory, we analyzed the cost rate function to find the optimal points for periodic replacement time and the number of system update cycles. Also, we present a numerical example to show how to apply our model to the real and practical situation in which even under the minor failure, the user of system is not willing to replace or repair the system immediately, instead he/she is willing to defer the repair or replacement until the periodic or preventive replacement time. Optimal preventive replacement timing using two variables, which are periodic replacement time and the number of system update cycles, is provided and the effects of those variables on the cost are analyzed.

• International Journal of Contents
• /
• v.3 no.1
• /
• pp.14-18
• /
• 2007

General replacement policy includes document-based LRU or LFU technique and other various replacement policies are used to replace the documents within cache effectively. But, these replacement policies function only with regard to the time and frequency of document request, not considering the popularity of each web site. In this paper, we present the document replacement policies with regard to the popularity of each web site, which are suitable for modern network environments to enhance the hit-ratio and efficiently manage the contents of cache by effectively replacing documents on intermittent requests by new ones.

• Journal of Chest Surgery
• /
• v.11 no.4
• /
• pp.428-432
• /
• 1978

Between Jan. and Oct. 1978, mitral valve Replacement had been performed on eight patients using porcine aortic valve mounted on flexible polypropylene or Egyloil wire and Dacron covered stent. Mitral valve replacement was performed on five, mitral valve replacement with tricuspid valve replacement on two, and mitral valve replacement with tricuspid valve annuloplasty on 1 patient using five Carpentier and five Hancock Xenograft. Overall operative deaths were 2 cases. All other 6 cases did well after surgery and improved. No systemic thromboembolism occurred without anticoagulant-No valve failure took place. The results are satisfactory.

• Journal of the Korea Safety Management & Science
• /
• v.8 no.2
• /
• pp.139-153
• /
• 2006

This paper deals with two forms of preventive replacement policy with minimal repair at failure. Those are, 1. the replacement policy I based on the cumulative operating time. 2. the replacement policy II based on the number of failures. The basic assumptions are; (1) the cost of minimal repair at failure is increasing with the number of failures since the last replacement, (2) the equipment fails stochastically with time.

• Journal of Applied Reliability
• /
• v.11 no.4
• /
• pp.409-417
• /
• 2011

In this paper, we consider a replacement model following the expiration of warranty. In other words, this paper proposes the optimal replacement policy for a repairable system following the expiration of payable renewing replacement-non-renewing minimal repair warranty. The expected cost rate per unit time from the user's perspective is used to determine the optimality of the replacement policy. Thus, we derive the expressions for the expected cycle length and the expected total cost to obtain the expected cost rate per unit time. Finally, the numerical examples are presented for illustrative purpose.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.11a
• /
• pp.161-164
• /
• 2006

The sales of consumer durables are composed of first time purchases and replacement purchases. Since the sales for most mature durable products are dominated by replacement sales, it is necessary to develop a model incorporating replacement component of sales in order to forecast total sales accurately. Several single product diffusion models incorporating replacement demand have been developed, but research addressing the multi-product diffusion models has not considered replacement sales. In this paper, we propose a model based on consumer choice behavior that simultaneously captures the diffusion and the replacement process for multi-product relationships. The proposed model enables the division of replacement sales into repurchase by previous users and transition purchase by users of different products. As a result, the model allows the partitioning of the total sales according to the customer groups (first-time buyers, repurchase buyers, and transition buyers), which allows companies to develop their production and marketing plans based on their customer mix. We apply the proposed model to the Korean automobile market, and compare the fitting and forecasting performance with other Bass-type multi-product models.