Solver case-study

All the solutions presented in the following case study were obtained using a 1.5 GHz Pentium M processor and GAMS 2.5/DICOPT. The NLP and MTT.P combination of solvers selected for DICOPT were CONOPT and CPLEX 7.0, respectively. 

The performance of the presented formulation was tested by applying it to a literature example and an industrial case study. All solutions were obtained using the GAMS/CPLEX solver in a 1.4 GHz Pentium M processor. 

The final step can be performed by numerical solvers, such as FITEQL 4.0 (19). If inert ions are added in the solution to represent ionic strength during the titration measurement, it is worthwhile to include the possible reactions between these ions and the ftmctional groups, which is similar to the consideration in the study of activated carbon by Chen and Lin (15). The literature arising from Westall and co-workers and Schecher and McAvoy provide many other successful case studies (19,20). 

As an example of this, Spear (2004) reported on a case study of young managers who were trained in the Toyota Education Model and became problem identifiers and solvers. Some of the key approaches for success... 

Case-Study for Solver Validation Nonequilibrium Distillation Column Model... 

The basic supply chain model was a multi-period MILP. The binary variables identified whether a certain product was produced at a given plant. The continuous variables represented the annual volume in supply, production, and distribution. The objective function included fixed capital expenditure as well as variable costs for supply, production, and distribution. An example model in the case study had 6 plants, 36 products, and 8 sales regions for a 12 year planning horizon. The MILP model had 60,000 variables (2000 binary) and 145,000 constraints. The model was solved using the ILOG/CPLEX solver on a 1.6GHz processor injust4min ... 

A case study was solved to illustrate the hybrid philosophy. An important conclusion from the case study is that although the hybrid approach is able to meet the total annual cost target criterion of 10% it still suffers the MINLP limitations in terms of the solver getting trapped in local optima. The result of this is that solutions generated tend to be similar to initial structures given to the solver. 

Our assurance cases are not for particular properties of the system they demonstrate that the software does not adversely affect the system in which it is embedded. We have applied our approach in an industrial case study, but use a quadratic-equation solver as an example here. 

We run the case study by clicking Start . At this point, the lower right comer window of the PFD will indicate that the solver runs multiple times. If the solver fails for an intermediate step in the case study, we recommend increasing the number of creep and total iterations in the Solver Options Window. Once the case study is complete, we can click on the Results button to view the results of the case study. 

It possible to manually change each WAIT and H2HC ratio and re-nm the model each time. However, given the typical run time for the Reformer solver, this quickly becomes a tedious process. It is better to use the Case Study features of Aspen... 

The computation performed in this study is based on the model equations developed in this study as presented in Sections II.A, III.A, III.B, and III.C These equations are incorporated into a 3-D hydrodynamic solver, CFDLIB, developed by the Los Alamos National Laboratory (Kashiwa et al., 1994). In what follows, simple cases including a single air bubble rising in water, and bubble formation from a single nozzle in bubble columns are first simulated. To verify the accuracy of the model, experiments are also conducted for these cases and the experimental results are compared with the simulation results. Simulations are performed to account for the bubble-rise phenomena in liquid solid suspensions with single nozzles. Finally, the interactive behavior between bubbles and solid particles is examined. The bubble formation and rise from multiple nozzles is simulated, and the limitation of the applicability of the models is discussed. 

These two methods are similar to what is found in the study of hydrodynamic instability where LES is replaced most often by DNS and Helmholtz solvers by Orr Sommerfeld solvers. Both methods have already been used and shown to be efficient more examples will be provided in Chapter 9. They were also compared in a few cases [350 321[. LES is much more expensive and cannot work if all boundary conditions are not known precisely. This means that impedances of all inlets and outlets for example are required. Helmholtz formulations also need impedances but since they run faster and can provide all modes (and not only the most excited), they are easier to use when impedances are not known precisely. 

The design of the INPUT module has been greatly influenced by a study of the systems mentioned in Section 1.2. The principal new contribution in this paper is the design of the SELECTOR and SOLVER modules. The numerical analysis modules are based on existing subroutines (in the case of the SIMULATOR and CURVEFIT modules) or on well known methods (in the case of the NONLIN module). 

It is well known that solvers that exploit the sparse structure of its Jacobian are more efficient in terms of CPU computation time than solvers that do not Unfortunately, those solvers are not always available or applicable to the specific studied case. 

Many DMFT calculations have been reported in recent years, the majority discussing 3d systems, but also several dealing with actinides. Less attention has been paid to the lanthanides, most probably because the atomic limit, as outlined Section 6.1, is sufficient for an accurate description in most cases. An important exception is the y-cc transition in cerium, which has been the subject of several studies (Held et al., 2001 Zolfl et al., 2001 McMahan et al., 2003 Haule et al., 2005 Amadon et al., 2006), all using the Hubbard model, but with various impurity solvers. Ce compoimds have also been studied based on the Hubbard model (Laegsgaard and Svane, 1998 Sakai et al., 2005 Sakai and Shimuzi, 2007). 

The simulations are carried out for the first 20 s of the first three, and the first 100 s of the last three earthquakes. The major response is seen in this time frame, and it also allows for more detail in the illustrations. A direct integration method for the solution of the equation of motion in Eq. 18.2 is used. Superposition of modal responses is not possible for systems with non-proportional damping, as is the case with the current structure with added dampers. The Newmark p method (by using the unconditionally stable average acceleration method) is used as the solver for aU simulations in this study. The function that implements this ordinary differential equation (ODE) solver makes sure that the simulation time step is 20 times smaller than the smallest period of the structure. If this is not the case, it interpolates the excitation data for a smaller time step, and later outputs the response at a 0.01 s. In this work, the building type structure has a minimum period of 0.108 s. Thus, the simulation takes place at 0.108/20 = 0.0054 s. 

The time evolution of some chosen quantities of the bubble is used as the benchmark quantities. These are the sphericity, bubble size, rising velocity, and the center of the bubble mass. In order to show the superior properties of our Level Set-based multiphase flow solver, we demonstrate here a mesh cmivergence study, which has unfortunately not been published so far by the original benchmark authors. To this end, the time evolution of the bubble shapes with respect to the given resolutions for the two test cases that are visible in Fig. 13.6 and the time evolution of the selected quantities as bubble sizes (in vertical and in the perpendicular symmetry directions) and sphericity is displayed in Fig. 13.7. As it may be seen from the mentioned graphical proofs a perfect convergence is achieved already on a relatively coarse resolution for test Case 1 and also in test Case 2 within the time period covering the initial 1.5 time units. Moreover, the mass loss in both cases... 

---