Code validation

Illustrative results obtained for two different periods of the temperatures on different sur­faces are shown in Figures 2 and 3. Numerical results were obtained running AGLA code with the same meteorological conditions read from experimental set-up and introduced as input data in the numerical code. Data shown correspond to a situation with no consumption.

(a) (b)

Figure 2: Numerical vs experimental results for a period from March 12th to 20th for Barcelona city: a)Indoor wall surface temperature, b)Temperature at the glass surface with transparent insulation

04

Figure 2(a) shows the indoor wall surface temperature for a period of nine days (from March 12th to March 20th) in Barcelona city, corresponding to days 71 to 79 of the year. Figure 2(b) shows the temperature of the TIM glass surface for the same period.

Dashed green lines represent experimental values whereas straight red lines correspond to numerical prediction.

Figure 3 represents data for a period from September 21st to October 9th. Figure 3(a) shows the temperature at the internal surface of the indoor wall and Figure 3(b) shows the temperature at the absorber surface.

Figure 3: Numerical vs experimental results for a period from September 21st to October 9th for Barcelona city: a)Indoor wall surface temperature, b)Absorber surface temperature

(a) (b)

From the previous figures it is observed that a good degree of agreement is achieved between the experimental and numerical results. This agreement was quantified in terms of the root sum square (see Reference [6] for details), values of about 8.6 to 2.6% were obtained for different variables and different periods of analysis. The higher discrepancies
were produced at the temperatures in the indoor surfaces of the accumulator, corresponding to the difficulty of evaluating the heat transfer coefficients in these surfaces.