Recently, I have been asked to investigate the possibility of creating a system that allows for the tracking of the weld seam through the seam annealer. But first, a little background:

My current employer is a manufacturer of welded pipe. The mill receives flat rolled steel and converts that steel into pipe. The steel is cold rolled and then the seam welded to form the pipe. During the welding process, the grain structure of the steel is changed, making the weld seam brittle, with respect to the other areas of the pipe. In order to correct this, the weld seam is passed under an annealer that heats the weld area of the pipe to between 1600 and 1800 degrees. This heating effect changes the grain structure, reduces the hardness and makes the weld area less brittle. Our customers have asked us how we ensure that the weld seam and the annealer stay in-line. To date, we have assigned mill personnel to monitor and adjust the annealer if they become mis-aligned.

In order to meet the ever-growing demands of our customers, I have developed a system that automatically monitors the alignment and alerts the mill operator, should the weld seam drift from under the annealer. This system utilizes a vision sensor to detect the weld seam, an HMI workstation and a PLC to produce the tracking and alarming functionality.

The vision sensor has been programmed to track and determine the position of the weld as the seam passes under it. The operator, once the mill is started, teaches the system the position of the weld with respect to the annealer. They also are required to enter a maximum deviation that the weld is permitted to drift. This information is stored in the PLC as the learned position and deviation parameters. When the mill is operating, the actual position of the seam is sent to the PLC and the learned and actual positions compared. The actual deviation is computed and compared to the maximum deviation permitted. If the actual devotion exceeds the limit for more than 12 consecutive readings (6 seconds) and alarm is sounded and the operators must visually inspect the alignment and perform any adjustments.

In the near future, once we have base-lined the current configuration, we are looking to use the deviation values to automatically send a signal to the annealer. The data sent to the annealer will command it to move east and west, maintaining alignment to the weld seam.

Sorry, but due to the confidential nature of the project, I am unable to name vendors or provide pictures. If you have any questions, please feel free to contact me to discuss this or any of the projects I have listed on the website.

Steel Yield Calculation

I have just completed a program that computes the amount of pipe that is obtained from a coil of steel. The process of forming flat steel into pipe requires that the steel be trimmed on both ends as well as the sides. This program reads the information related to the weight, gauge and width of the coil steel and computes a theoretical length. The forming process has sensors that determine the actual length of the steel after the head and tail are trimmed. Using that information as well as knowing the required width based upon the pipe size, we are able to calculate the amount of steel that is actually consumed in the forming process. The difference between the coil weight as it was received and the steel weight that is consumed is used to calculate the yield.

This program is developed in a combination of SQL Server, WonderWare Archestra and Intouch and .NET programming. Wireless PC’s are used in the cranes to allow the crane operator to denote which coil is actually being placed into the mill for processing.

In addition, we have installed ControlLogix PLC’s to monitor the coils being placed into the mill and to sequence the coils through the process.

The entire system was successfully deployed, on time, in January 09.

Package Handling and Sorting System

I was responsible for the design and deployment of a multi-million dollar control system to handle the sorting of overnight packages. This system involved the integration of over 13,000 I/O points and 28 PLC’s. The design was made up of several redundant processors and I/O as well as redundant networking rings and multiple control room interfaces.

I was responsible for managing the PLC programming effort of 5 PLC programmers and the control system design. I also oversaw the installation of all the sensors throughout the conveyor systems, MCC and drives. I developed custom algorithms to properly distribute the packages evenly throughout the system and then managed the programmers who were responsible for the coding and integration of the system.

This system was successfully deployed in June of 1997

Package Handling / Load Distribution System

I was responsible for the design and initial integration of a vision sensor based control system that evenly distributed packages across a package slide. This system used a combination of vision sensors, lasers and filters to determine the level of packages at various points across the slide and then control the diverters responsible for the distribution to ensure and even distribution.

Once the initial system was deployed, I managed the installation of the system in several additional facilities throughout the US.

This system was successfully deployed in several locations in 2005.

Automated Circuit Breaker Test Stand

I was responsible for the design, programming and deployment of a two axis, Cartesian robot that utilized a vision sensor to perform the final inspection of circuit breakers.

This system was comprised of a two-axis gantry type robot upon which a vision sensor was mounted. The axis was programmed with the various inspection points for each type of breaker to be inspected. The vision sensor was also programmed with the inspection criteria to be used at each point.

The operator would place the unit to be tested into the fixture and select which tests were to be performed. The system would download the test criteria from the server and then begin sequencing through the various inspection points, verifying the proper criteria at each point. Upon successful completion, the system would dispense the PASSED sticker to be placed ion the unit. If the unit failed, the operator would have the chance, if possible, to correct and retest the unit. In the event the unit was not repairable, the database would store that failed units serial number in the database in the event that unit should ever accidentally be sent to a customer.