WSE - BASB™ Functional Specification Summary

Bell Annealing Stack Builder - BASB™
Functional Specification Summary

Overview
The WSE Bell Annealing Stack Builder (BASB™) program is designed to create optimal stacks from a working inventory of cold rolled sheet coils based upon a set of tuned parameters. BASB™ can be run as a standalone utility or can be integrated with the WSE Bell Annealing Shop Scheduler (BASS™) and Shop Manager (BASM™) products. BASB™ runs on a Microsoft Windows 95/NT platform and comes complete with a graphic user interface. BASB™ uses a relational database management system to warehouse all parameters and system data. BASB™ comes standard with a single user Borland InterBase RDBMS but can be configured to work with MS SQL, Oracle, Informix or Sybase. In forming coils into stacks, BASB™ takes many items into consideration. These consist of coil properties, resultant stack properties, and various system tuning parameters all of which are outlined below.

Coil Properties
Coil data can be entered into BASB™ manually or can be imported via a custom interface to an existing plant system containing the information. BASB™ recognizes the following coil properties:
C_NUM - The unique coil ID number.
C_W - The width of the strip.
C_WGT - The weight of the coil.
C_OD - The outside diameter of the coil.
C_ID - The diameter of the ”eye” of the coil.
C_Q - The quality rating of the coil, i.e. CQ, DQ, etc.
C_GAG - The gauge of the strip.
C_P1 and C_P2 - Level 1 and 2 priorities of the coil. This is used to force BASB™ to choose a given coil over another in spite of another property that ordinarily would result in a different outcome.
C_DI - The date and time the coil arrived in inventory ahead of annealing.
C_DD - The latest date and time the coil must be annealed and ready to leave the annealing area for its next process.
C_IM - The maximum allowable time that the coil may sit in inventory ahead of annealing.
C_LOC_X, C_LOC_Y, C_LOC_Z - The x,y,z location of the coil in inventory. When this information is available, BASB™ will attempt to form stacks from coils which are more easily accessible in the inventory area.
C_OQn - A list of other quality coils that this coil can be combined with to form stacks. BASB™ can be configured to use this list or to build stacks exclusively with coils of the same quality.
C_ANT - Anticipated anneal time for this coil. This can be calculated by a customer supplied heating model or can be determined by BASB™ empirically. BASB™ accumulates empirical heating and cooling data over time both for coils and stacks.

Stack Properties
Stacks built by BASB™ have the following properties:
S_NUM - Unique identifier for the created stack.
S_WGT - Net weight of the stack not including base, separator or convector plates.
S_H - Total height of the stack including base, separator or convector plates.
S_DD - Latest date that the stack must be finished annealing and ready for any follow on processing.
S_DI - Date the stack was placed in inventory, equal to the least recent C_DI in the stack.
S_Q - Quality of the stack, equivalent to the highest C_Q in the stack.
S_P1 and S_P2 - Priority level 1 and 2 equal to the highest C_P1 and C_P2 in the stack.
S_ODD - The cumulative outside diameter difference of all coils in the stack.
S_HTD - The cumulative heating time difference of all coils in the stack based upon the C_ANT property of each coil in the stack.
S_HM_AT2 - Predicted anneal time for the stack based upon empirical data accumulated over time by BASB™.
S_HM_AT3 - Refined predicted anneal time for the stack based upon a customer supplied heating model.

Stack Building Algorithm Considerations
In general, all other things being equal, BASB™ will attempt to build the fewest stacks possible from a given inventory of coils in order to increase annealing shop throughput. While this sounds simple enough on the surface, this is an extremely complex task which exceeds the capability of even the most experienced and intelligent person even with an inventory of as few as 15 or 20 coils. BASB™ considers many variables while building stacks including:

Minimize number of charges (stacks)
Do not allow stacks to be built which exceed the maximum allowable height.
Do not allow stacks to be built which exceed the maximum allowable weight.
Avoid coils not being delivered on or before their respective C_DD property.
Avoid coils remaining in inventory longer than allowed as given by their respective C_DI and C_IM properties.
Avoid stacking coils with dissimilar expected heating times/qualities.
Do not stack heavier coils on top of smaller and lighter ones.
Avoid stacking heavier gauge coils on top of lighter gauge ones.
Attempt to stack coils with similar priorities thus avoiding a situation where several high priority stacks have been built when fewer could have been created.
Avoid building stacks from coils which might be “buried” in inventory or otherwise inconveniently accessible.
Attempt to maximize the volume of a stack in order to get as much capacity as possible out of the annealing furnace.
Avoid stacking coils with dissimilar outside or inside diameters as this negatively affects convection air flow thus producing extended heating and cooling times.

System Tuning
Because BASB™ considers so many variables, it is necessary to tune the system by assigning weighting factors to the various components of the BASB™ algorithm. This is done using simple slide bars and check boxes via the BASB™ graphic user interface. When building stacks, BASB™ does not simply consider the properties of the resultant stacks as singular objects. Instead, BASB™ considers the entire “Stack Pool” created. It is not difficult to find one or more very good stacks from an inventory of 100 coils. However, it is extremely difficult to measure the negative effect which will result in the long run if a given stack is built. For example, if “odd ball” coils are left behind in favor of more uniform coils, eventually the coil inventory will get out of balance which will result in the necessity of several very poor stacks having to be built. BASB™ avoids this “coil orphan” phenomenon by looking at the entire coil pool and the resultant stack pool. BASB™ builds hundreds or even thousands of resultant stack pools before finally choosing the one that best fits the given parameters. In this way, BASB™ finds a good fit (if not a perfect fit) for orphan coils continuously, thus keeping the inventory in balance. Through tuning, WSE and the customer can force BASB™ to place more importance on certain criteria. For example, in the extreme, BASB™ can be forced to choose only coils with similar delivery dates or priorities regardless of how many stacks would be required to accomplish this. On the other extreme, BASB™ can ignore all factors in favor of building stacks with no fewer than three or four coils. Neither scenario is appropriate in practice, but through tuning, BASB™ can be set up to consider all criteria with some receiving slightly more importance than others. Regardless of the tuning parameters, BASB™ will never build stacks which would violate hard and fast rules like maximum stack weight or height.

Reporting and Analysis
BASB™ includes several standard reports for viewing and analyzing coil and stack data. The coil inventory can be depicted graphically or in tabular format to see the distribution of coil weights, physical sizes, qualities, delivery times, time spent in inventory, etc. By analyzing these reports over time, one can determine if BASB™ is properly tuned and is thus maintaining the inventory in balance and not producing would be orphan coils. The coil inventory reports can also be used by a planner who, in a rare instance, wants to manually override BASB™ to build his own stack(s). In addition, resultant stack pools can be analyzed to help determine which stacks should be taken next for charging. BASB™ can rank the stacks by priority, shortest time before required delivery, longest time already spent in inventory, etc. to assist the planner in choosing the next one or more stacks for annealing. BASB™ can even be configured to “red flag” certain stacks to let the planner know that he must choose them to be annealed lest he miss an important deadline. Because all information is maintained in an open architecture, relational database, data analysis with BASB™ is limited only by the imagination of the user. BASB™ can export coil and stack data into a tab or space delimited ASCII file which can then be imported into popular spreadsheet programs like Lotus, Excel and Quattro Pro.

For more information, follow one of the links below.

Download Complete BASS™/BASB™/BASM™/BACT™ Specification PDF file (700 KB)

Overview of the BASS™ Component Topology

BASS™ System Data Flow Diagram

BASS™ Bell Annealing Shop Scheduler

BASB™ Sample Screen Captures