Predicting residence time distributions in a single screw extruder from operating conditions

Modeling residence time distribution in a twin-screw extruder as a series of ideal steady-state flow reactors

“…However, the change in d value was low within the range of input variables. The reason for the small range in d values (0-0.161) may have been, as found by Yeh and Jaw (1999), that the d value was influenced primarily by screw configuration and during this investigations, the screw configuration was kept constant. The other reason may be that for all the runs, the feed rate was maintained at a maximum level that corresponded to the screw speed resulting in similar dead volume fractions.…”

“…Mean residence time (t ‾) and fraction of plug flow (p) were calculated as (5) (6) A dimensionless parameter, normalized time (θ), was used to compare the RTD curves (7) Thus, the dimensionless plots of E(t) and F(t) became (8) (9) Abdelrahim et al (1993), Altomare and Ghossi (1986), and Yeh and Jaw (1999) and others have used similar mathematical functions (Equations (1), (2), (5)- (9)) to represent the RTD for different food processing applications. However, Equations (3) and (4) were developed by simplifying the model presented by Levenspiel (1972) for this application.…”

“…For most conditions, n was found to be either 2 or 3, which may imply that 2 or 3 CSTRs in the model were adequate, and optimal for the model. The simplified model proposed by Yeh and Jaw (1999) contained only one CSTR. Figure 3 shows that though the simplified model was able to describe the tail portion of the curve, the initial portion did not fit with the experimental RTD curve.…”

“…In this paper, a model of plug flow with a finite number of CSTRs having dead volume fractions in series is referred to as a "complete model." Yeh and Jaw (1999) proposed a "simplified model" assuming plug flow in series with one CSTR with a dead volume fraction. They found that the simplified model and complete model represented the RTD better than other general conceptual models.…”

Modeling of the residence time distribution in a buss kneader with a back‐propagation neural network

“…Thus, once V, MRT, Q m , and r are available, ADF for a specified operating condition can be calculated. The three main approaches to measuring the RTD are the offline method, [15][16][17][18][19][20][21][22][23] inline method, [24][25][26][27][28][29][30] and online method. 31,32 It is well known that the traditional offline method is quite time-consuming and yields few data points, which are not enough to allow for a detailed analysis at the tail region of the RTD curves.…”