Opening up to robotic change: by deciding to automate its cleaning and finishing room with robot, Pacific Steel reduced its costs and opened up a common bottleneck in its casting process.
At Pacific Steel, a suspension component liar a heavy-duty track was a prime example of the cleaning room blues. Each time the high-volume casting reached the cleaning and finishing room, a hold up occurred because of the required levels of labor. Four operators spent 70 sec. working in a parallel process to make a single cast component ready for the customer. Four of Pacific Steel's 10 cleaning and finishing operators found their time was quickly monopolized by the suspension component.
There had to be a better way. And, as it turns out, there was.
Deciding to Automate
Greater numbers of metalcasting facilities are attempting to robotically automate a variety of applications within their operations. Pouring systems, mold and core handling operations, machining and assembly are all incorporating robotic technology to eliminate labor and make the process more consistent. From job shops to high production operations and from steel shops to aluminum, the opportunities for the technology exists--it is just up to the operations to lay out the justification for them.
In most job shops, the cleaning room is usually the main bottleneck in casting production, As a result, studies indicate that cleaning costs can reach 30% of the total casting production cost. This turns opportunities for cost savings in cleaning and finishing into life rafts for ailing bottom lines. With that in mind, Pacific Steel was open to the idea of attempting to automate the cleaning and finishing process--especially if it could free up the time of some of its operators and open up the bottleneck.
In order to justify the investment in robotic automation, the firm looked specifically at the suspension component. Because of its high volume, there was tremendous price pressure on the casting. Implementation of the robot wasn't the issue; it was the initial cost and making sure there was enough volume to keep it busy that was causing concern. Including training, the robot cost about $1 million. After doing a cost analysis based on time studies, throughput using different processes and material costs, it was decided that if Pacific Steel didn't automate. there was a very good chance it couldn't have kept producing the suspension component. Executives at Pacific Steel had toured another plant using the robot to see it at work. After witnessing the demonstration, and knowing that options lot the steel industry were limited, the decision was made.
The robotic cell that is in rise at Pacific Steel consists of a robot, a cutoff saw, a compliant cup grinder and a hydraulic trim press with customized controls and programming. The robot has a six-axis manipulator and a reach of 8 ft while being accurate to within 0.1 ram. In addition, the robot can handle parts weighing up to 200 lb. The entire robotic cell is 20 x 20 ft. A touchpad serves as the communication device that controls the movements of the robot and other equipment in the cell. Programming also can be done offline and loaded into the robot using a floppy drive.
The Traditional & Serial Processes
Before moving to automation, Pacific Steel was using a traditional process that involved four work stations and four operators working in parallel. The first operation was removing the risers and gating system using a cutoff saw. The next operator opened holes by removing fins. Then the cast components endured grinding operations to remove riser contacts and parting line fins. The process time for each of these operations ranged from 35-40 sec. After including the queue time and effects of operator fatigue, the average cycle time was 70 sec.
Once automated, the casting went through a similar set of operations in the robotic cell, except only one operator was needed, The process also is different from traditional cleaning in that it is a serial process, so only one cast component could be worked on at a time. However, after taking into account the time for changing the grinding stones, cutoff wheels and calibration time, the average cycle time is still 70 sec.
The castings are loaded onto a fixture by the operator, and then the robot arm picks it up by gripping a window, which is part of the casting, and uses a cutoff saw to remove the riser and gates. The riser contacts and the fin are ground using a cup grinder. The casting is then placed in a trim press to punch holes and remove the fin on the outside of the casting. The component is then sent down a conveyor, where an operator removes it from the cell. Each of these operations is assured by sensors in the system.
While Pacific Steel didn't increase production speeds in cleaning and finishing through the robotic automation, the firm did achieve the ultimate benefit of reduced costs Three operators are now free to work on other cast components. In addition, the cleaning and finishing cost for the suspension component plummeted from $5.64 for a single casting to $2.64. The cost benefit is expected to be even higher after the depreciation period for the robot is completed.
Automation Doesn't Automatically Solve Problems
While the implementation of robotic automation at Pacific Steel has been positive, it was not without its problems.
The 3-6 month ramp-up period required familiarization between the robot and castings and between Pacific Steel workers and the robot. During ramp-up, a set of 30 sample castings with risers and gating systems were used to teach the robot movements and write the program. Customized tooling and fixtures for the casting also were built. Three operators, a programmer and maintenance personnel, were trained onsite over a period of two weeks.
The initial program was modified substantially to take into account the process variation of the cast components, keeping an eye on the cost of the spent materials like the cutoff saws and grinding wheels. Multiple grinding stones and cutoff wheels were tried with the robot with the intent of increasing their longevity and reducing the lost time in changeovers.
Once the robot was in full automation, minor modifications were made to the program on an as-needed basis to take into account the variations in the castings from different hatches. Multiple changes were made to the risering and gating of the casting, which necessitated changes to the program. The operators, who worked with the robot on a regular basis, are not computer literate and are only given the authority to make minor modifications to the position of the robot and not the program itself. The program was locked and password protected in order to limit the possibility of unintentional changes to the program.
Initially, a lot of program modifications were performed to increase fine acceptance rate. The robot does not expect to receive a cast component that is bent, has too much fin or has holes that aren't properly aligned. The program was modified to be more receptive. Initially, 40% of the castings produced could be sent through the robotic cell. After a year, that figure rose to about 95%.
Enjoying the Results
After the initial hiccups, the results Pacific Steel has seen with the robotic cell have been gratifying. After a careful measurement and analysis of the costs before and after the robot, Pacific Steel found that by automating, costs were reduced by more than 50%. Most of the cost savings occurred because of fewer operators and reduced cleaning and finishing material usage. Advantages also were seen in improvements in the quality and consistency of the processed casting. In addition, there was a higher throughput per square foot of cleaning room area, which opens up the cleaning room for other work and reduces the lead time on parts.
An intriguing benefit up the production line from the robot was increased attention to detail by molding and coremaking because the robot required a consistent part with little to no variation.
This article was adapted from a paper presented at the 2003 Steel Founders' Society of America Technical and Operating Conference.
For More Information
"Cleaning Room Automation for Higher Quality, More Consistent, Lower Cost Investment Castings," Orf, N., Incast, Nov. 2000, p. 12-15.
Ravi Anaparti is the computer-aided engineering manager at Pacific Steel Casting Co., Berkeley, Calif.
|Printer friendly Cite/link Email Feedback|
|Date:||Jan 1, 2004|
|Previous Article:||Building better molds: can nobake & green sand work together?|