Session: 03-13-01: Manufacturing: General I

Paper Number: 166363

Automation of a Bandsaw Operation to Improve Manufacturing Performance With Nonuniform Part Feature Geometry 

Manufacture of nonlinear part geometries presents challenges in maintaining uniform feature characteristics when the manufacturing process conditions are not constant. One such manufacturing process that is challenged by varying part conditions (e.g., geometry changes or material characteristics) is bandsaw operations, since constant feed rates of the saw blade can result in different force profiles through the cutting operation, which can result in undesired part performance (e.g., part features violate tolerance requirements).  To overcome this issue, bandsaw operations are performed manually to maintain required part features.  However, the manual process can be more labor intensive. Therefore, it is desirable to automate the process to improve production and part performance. There are several related works on sawing of square and rectangular shapes, as well as wood sawing. However, publications pertaining to sawing of polymer round solids and pipes, such as those being studied in this application, are sparse. One such part is a Pre-Pour sleeve, which is an ABS polymer pipe with fixed lengths used as a form when pouring concrete floors for industrial and commercial structures. It is costly to adapt the extruder to change the Pre-Pour sleeve length, so an adapter piece is cut from a pipe and epoxied to a standard-length sleeve, where the part is currently cut to length manually.

 

The work in this paper investigates a proposed autonomous process to produce the adapter sleeves on a vertical bandsaw. In the proposed process, an operator loads parts in a fixture, actuates the process, and removes and packs the finished part. A semi-skilled assembler can be used to facilitate this operation. A proposed control system architecture (e.g., closed-loop PID) is used along with PLC-controlled electric actuators. The saw blade speed is fixed, but the feed rate is varied (i.e., manipulated parameter) as the instantaneous part thickness changes through the operation, so a constant force (i.e., controlled parameter) is maintained on the saw blade to improve overall performance (e.g., excessive forces can result in the saw blade not producing acceptable performance). Algorithms to facilitate a constant force are developed that can be implemented in the PLC, where the objective is to keep the cutting force consistent, which will help maintain a desired straightness tolerance (e.g., ±0.050 in). Monitoring of the constant force is accomplished by closed loop feedback data to the PLC controller from a piezo dynamometer mounted on the movable fixture base.

 

The economics of adapting this system are studied. Process analysis of the manual actuation and the automated actuation are compared to see the cost per part produced using each method, including identification of the break-even point for how many parts need to be produced to support automating the process. Savings are also realized using semi-skilled assemblers at a low labor rate rather than a higher skilled/higher paid employee required for manual implementation. Cost savings is also realized by lower cycle time as well as scrap reduction. Empirical data for the manual process shows that a minimum of about 3.5 minutes is the total cycle time, with about 20% scrap rate. The semi-autonomous cycle is estimated to take approximately 2.5 minutes and scrap can be held to around 2%. Additional work will investigate a data-based transfer function model of the process and a simulation-based case study to validate performance per representative real-world data.

Presenting Author: Michael Smith University of North Carolina at Charlotte

Presenting Author Biography: Michael Smith (Member: ANS, ASME, IEEE Senior Member, SME) received a B.S. in Mechanical Engineering Technology (2005), M.S. in Mechanical Engineering (2008), M.S. in Electrical Engineering (2012), and Ph.D. in Electrical Engineering (2015) from the University of North Carolina at Charlotte, Charlotte, NC, USA. He is currently an Assistant Professor in the Department of Engineering Technology and Construction Management (within the William States Lee College of Engineering) at the University of North Carolina at Charlotte, Charlotte, NC, USA. His background includes instrumentation-based process control, process modeling, data analytics, data-driven decision making, and software development, with over 10 years of industrial experience in the energy industry and more than 15 years of teaching experience. With a particular interest in industry applications, his research focus areas include: (a) control systems (e.g., adaptive control, optimal control, system dynamics, and stability), (b) process modeling and data analytics (e.g., physics-based and data-driven methods, including machine learning), and (c) monitoring/instrumentation. His research interests include applied energy, process automation and optimization, electromechanical systems, and manufacturing.

Authors:

Brian Barringer UNC Charlotte
Michael Smith University of North Carolina at Charlotte