System architecture, column

The amount of time involved with the separation is going to be a function of the length of the column, the linear velocity, and the capacity term we discussed above. If one wants to get the most flexible column system for a variety of separation problems, what one is going to want to do is optimize each of these terms (fluid velocity, concentration and separation time). What one is forced to consider is the ability to have a dynamic column length control built into the system architecture, which will minimize the amount of dilution, shorten the amount of time to get components out, keep the pressure at its lowest possible level for maximum operating capability, and not sacrifice the ability to separate compounds. 

Column 6 describes how this component failure mode affects the component function of the module (or sub-system). Column 7 lists how this component failure mode affects the next system sub level. In safety evaluations this column is used to indicate safe versus dangerous failures. Depending on the scope of the FMEA, it is possible to consider aU levels of the system (component, module, unit or system). Frequently a FMEA at the component level describes the effect at the module level and perhaps the effect at the unit level. Unless the FMEA is being done for a specific system architecture, another FMEA is done for higher levels. 

Table 4 2 shows the results of combining the masses in the Total column. Each set of rows corresponds to one of the system architectures described in Section 6. Each component has a column that shows the rating of that component, the number required for the specific architecture, how much one component weighs and the total mass of that specific component. 

In this illustration, a Kohonen network has a cubic structure where the neurons are columns arranged in a two-dimensional system, e.g., in a square of nx I neurons. The number of weights of each neuron corresponds to the dimension of the input data. If the input for the network is a set of m-dimensional vectors, the architecture of the network is x 1 x m-dimensional. Figure 9-18 plots the architecture of a Kohonen network. 

There are many different architectural views of a system, calling for different kinds of elements. These elements include hardware and networks, packages and their structures, object types and relationships, concurrent processes and threads, tables and columns, and the patterns that dictate how they are to be used. An architectural style, or type, defines a consistent set of elements and rules for their use. 

The significant intrinsic limitation of SEC is the dependence of retention volumes of polymer species on their molecular sizes in solution and thus only indirectly on their molar masses. As known (Sections 16.2.2 and 16.3.2), the size of macromolecnles dissolved in certain solvent depends not only on their molar masses but also on their chemical structure and physical architecture. Consequently, the Vr values of polymer species directly reflect their molar masses only for linear homopolymers and this holds only in absence of side effects within SEC column (Sections 16.4.1 and 16.4.2). In other words, macromolecnles of different molar masses, compositions and architectures may co-elute and in that case the molar mass values directly calculated from the SEC chromatograms would be wrong. This is schematically depicted in Figure 16.10. The problem of simultaneous effects of two or more molecular characteristics on the retention volumes of complex polymer systems is further amplifled by the detection problems (Section 16.9.1) the detector response may not reflect the actual sample concentration. This is the reason why the molar masses of complex polymers directly determined by SEC are only semi-quantitative, reflecting the tendencies rather than the absolute values. To obtain the quantitative molar mass data of complex polymer systems, the coupled (Section 16.5) and two (or multi-) dimensional (Section 16.7) polymer HPLC techniques must be engaged. 

FIGURE 16.11 Schematic representation of eluent gradient polymer HPLC. Two polymer species A and B are separated. They exhibit different nature and different interactivity with the column packing (e.g., adsorp-tivity) or with the mobile phase (solubility). The linear gradient from the retention promoting mobile phase to the elution promoting mobile phase is applied. The focused peaks—one for each polymer composition/ architecture—are formed in the appropriately chosen systems. Each peak contains species with different molar masses. 

The presently most popular approach to two-dimensional polymer HPLC avails partial or preferably full suppression of the molar mass effect in the Id column so that the complex polymer or complex polymer system is separated mainly or even exclusively according to chemical structure or physical architecture of macromolecules occuring in sample. Appropriate coupled methods of polymer HPLC are to be applied to this purpose (compare section 11.8). In the 2d separation column - it is usually SEC - the fractions from the Id column are further discriminated according to their molecular size. In other words, fractions obtained in the first-dimension column are separated in the self-existent second-dimension column, which applies distinct separation mechanism(s). Only exceptionally SEC... 

An additional reason for peak broadening under the critical conditions may be the temperature gradient inside of column generated by viscous heat dissipation [174] and, to a much less extent pressure gradient [ 175]. As is known, critical conditions are quite sensitive to temperature [7,38,39,79,125] and, in some systems, to pressure [7,17,92]. Furthermore, another reason for broadening could be differences in polymer architecture, because isomers may manifest itself under critical conditions (Table 1). In any case, the question as to why polymer peaks are not significantly narrower under critical conditions, remains unanswered and requires... 

Table 1 shows an example architectural attribute number of ASRs. In this table, each of the n rows ( , , n) represent a system component. The second and third columns are the output interface components h and ft-2- For example, the value of the number of ASRs from component 1 to the output interface component h. Similar to Table 1, we need to create two more tables for the length of the shortest ASR and the length of the longest ASR. Having the values for all architectural attributes, we may be able to select between architecture A and A using Algorithm 2. 

---