Page 417 - Design and Operation of Heat Exchangers and their Networks
P. 417

400   Design and operation of heat exchangers and their networks


             With known values of k and α o , the heat transfer coefficient at the tube
          inside can be evaluated by
                                                        1
                                  1   d i  d i ln d o =d i Þ
                                               ð
                             α i ¼                                    (8.31)
                                  k  α o d o   2λ t
             Eq. (8.31) is meaningful only when α o ≫ α i . The relative deviation of α i
          due to the relative deviation of α o can be expressed as

                                 Δα i       Δα o  α i d i
                                                                      (8.32)
                                  α i        α o  α o d o
          That means, if Eq. (8.29) has a maximum deviation of 30% in α o , to limit the
          relative deviation of α i less than 3%, we shall ensure that α o  10α i . Under
          this condition, Eq. (8.30) reveals t s  t w  0.1(t s  t). It means that the wall
          temperature t w would be close to the saturation temperature t s .

          8.2.3 Modified Wilson plot method

          In 1915, Wilson (1915) has proposed a graphical technique to determining
          heat transfer coefficients from a number of measured overall heat transfer
          coefficients without knowing the temperatures of the heat transfer surfaces.
          Since then, the so-called Wilson plot method has been widely applied and
          further developed into a computational technique. A detailed description of
          the use of the Wilson plot method and meaningful discussion can be found
          in Shah and Zhou (1997). Roetzel and Na Ranong (2002, 2003) considered
          a few simple examples and their evaluation using the Wilson plot technique.
          It is shown that, in spite of good results of the data regression, undesirable
          systematic errors in the obtained heat transfer coefficients may occur. The
          experiments with different thermal conductivities of the fluids would reduce
          such an error. An overview of the Wilson plot method was given by Fer-
          nandez-Seara et al. (2007), which reveals numerous modifications and
          would be useful for experimental investigation of convective heat transfer.

          8.2.3.1 Original Wilson plot method
          In the original Wilson plot method, the test fluid (fluid 1) flows on the tested
          side (side 1) for which the heat transfer coefficient vs flow rate or Nusselt
          number vs Reynolds number shall be determined; however, the heat trans-
          fer characteristics of the other side (side 2, fluid 2) are unknown or cannot be
          exactly evaluated. If the flow in side 1 is in the fully turbulent region, the
          Nusselt number can be expressed as


                                            m
                                         n
                             Nu 1 ¼ C 1 Re Pr μ =μ w,1   0:14         (8.33)
                                         1
                                               1
                                            1
   412   413   414   415   416   417   418   419   420   421   422