Analysis of the results

The first set of experimental data for the distribution tube was analyzed by the factorial design analysis using MINITAB version 15.

Factorial design analysis is a statistical method in which every level of one factor is tested in combination with every level of another factor. In general, in a factorial analysis, all possible combinations of factor levels are tested [5].

Four variables (factors) were considered to have the highest influence on the distribution test. The first factor is the LiCl volume flow rate with three treatments (levels) with a range between 0.3 l/min-1.2 l/min. The second factor is the bore size with five levels, ranges between 0.5 mm – 0.9 mm. The third factor is the distance for each throttling-bore from the LD feed-box entrance with ten levels, ranges between 4 cm-58 cm. The forth factor is the run time for each test with two levels (10 min and 15 min). Two replicates were taken for each trial in order to get an unbiased estimation. Since the number of levels for the studied factors is not equal, general full factorial design was used to analyze the measured data. The significance level a of the test is 0.05.

Подпись: Fig.4 depicts the normal probability plot of LiCl mass throttled through the perforated plastic tubes. The normal probability plot is a graphical technique for assessing whether or not a data set is approximately normally distributed. The data are plotted against a theoretical normal distribution in such a way that the points should form an approximate straight line. Departures from this straight line indicate departures from normality. The figure shows a little bit lower tail that do not fall exactly along a straight line passing through the centre of the plot. This indicates some potential problem with the normality assumption. However, no severe deviations from normality are obvious apparent. It is clearly shown that the minimum mass flow rate that could be used to ensure normal probability assumption corresponds to a bottle content with a total mass of 230 g.

From the output a large F-statistic and a low p-value is seen, indicating strong evidence of a correlation between the four factors and the LiCl throttled through each discharge-point. It is also shown that p-values for all terms in the test are less than the significance level. This confirms the quality of the test.

Fig. 4. Normal probability plot of the response (LiCl mass)

Fig. 5. Residual versus the response (LiCl mass)

Подпись: Fig. 6. LiCl adsorption-speed test for different textiles

The second set of tests focused on the absorption capacity and diffusivity of different textiles. A probe of lithium chloride solution (45% salt concentration by weight) was poured on a taut piece of textile. Then the needed time for the LiCl droplets to be completely adsorbed by the textile fibers were measured. The tests were carried out for the textiles in dry and wet state. Fig. 6 shows the adsorption speed for probes of one of the promising wood-pulp based textiles, polyamide, viscose, cotton, polyester and polystyrene fibers.

The best results were obtained for pure wood-pulp, polyamide and polyester fibres. Different probes of polyester and polyamide fibres reveal essentially inferior adsorption speed. Therefore the material itself is not the only factor influencing the adsorption speed. During the tests a strong dependency was detected between the sorption speed and the material thickness, the pore size and the weaving structure. Further tests must be carried out to clarify these influences. A good LiCl diffusion behaviour through the pure wood-pulp fiber was proved in the experiments.

2. Conclusion and Outlook

This paper presents experimental tests and a factorial design analysis of a desiccant distributor system which will be used in the prototype of a LD regenerator. It shows the optimal conditions that will ensure even distribution and the applicability to serve a wide range of desiccant flow rates. Furthermore, it presents experimental investigations for different textiles in order to choose the best textile with the highest adsorption speed and the finest desiccant diffusion through the textile fibers.

The factorial design analysis shows the optimal diameter of the throttling-holes, and the optimal spacing between those throttling diameters. Those optimal values will ensure a fairly even distribution and it will serve a wide range of desiccant volume flow rates, which will make the LD distributor flexible to serve the requested volume flow rate requested by the air conditioning cooling load.

The experimental investigations related to textiles approved that new textiles such as pure wood – pulp fibers and polyamide perform much better than traditional textiles currently in use such as cotton.

Further investigations must be carried out to clarify the influence of the textile thickness, the pore size, the desiccant type and the weaving structure on the adsorption behaviour of textiles. Experimental investigations might be considered concerning: the durability of the throttling-bores against the desiccants corrosivity, the influence of the regenerator operating conditions (high temperatures and corrosivity) on the textile performance, the effect of using higher desiccant concentration e. g. 40% on the distribution through the perforated tubes (in case if the LD distributor will be used in a conditioner instead of a regenerator), the necessity to flush the distributor and the solution that might be used.


[1] Kathabar technical information, www. kathabar. com.

[2] Krause M., Saman W., Vajen K. (2005), Regenerator Design for open Cycle Liquid Desiccant Systems – The-oretical and Experimental In-vestigations, Proc. Interna-tional Conference Solar Air-Conditioning, Staffelstein, 6. – 7.10.2005.

[3] Lowenstein A., Slayzak S., Kozubal E. (2006), A Zero Carryover Liquid Desiccant Air Conditioner for Solar Applications, ASME/Solar06, Denver, USA.

[4] Lavemann E., Peltzer M. (2005), Solar Air Condition-ing of an Office Building in Singapore using Open Cycle Liquid Desiccant Techno-logy, Proceeding of the International Conference on Solar Air Condi-tioning, Staffelstein, 06.-07.10.2005

[5] Montgomery, D., Runger G. (1999), Applied statistics and probability for engineers, second edition, John Wiley and Sons, Inc. New York.

Leave a reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>