Hendrix Manufacturing: a case study in converting a green sand molding system to nobake.
In the foundry industry, change is a slowly evolving process. Most foundries redefine their operations through subtle additions and alterations to address specific needs at specific times. For a foundry to make the decision to discard an existing molding process in favor of a new one, there must be new markets or an expansion of existing ones available immediately for its new capabilities. But even with these opportunities, the decision to commit to this major conversion of an operating foundry can lead to sleepless nights for even the most experienced management team.
In 1993, the prospect of insomnia confronted the management team of Hendrix Manufacturing Co., Inc., a 185-employee steel foundry in Mansfield, Louisiana (see sidebar "The History of Hendrix Manufacturing"). Hendrix believed that its molding operation, which was half green sand and half nobake producing 2900 tons of castings/year [ILLUSTRATION FOR FIGURE 1 OMITTED], was falling short of the mark in providing the casting quality it required at the cost it required. It realized that to continue to be a competitive supplier in its dragline and excavator bucket markets and also broaden its reach to new markets, it had to examine other molding opportunities.
Hendrix solicited the help of Vulcan Engineering Co., an engineering firm in Helena, Alabama, and together these companies developed a four-step plan to explore the options available to improve the quality and cost of the mold operation, and minimize any sleep deprivation.
The four steps (integral to any major conversion project) were:
1. Analyzing the existing operation. 2. Outlining a plan for change. 3. Developing a funding plan for change. 4. Implementing the plan.
Following is a look at the four steps and the results for Hendrix as it converted its foundry from a 50/50 green sand and nobake molding operation to entirely nobake.
1 Analyzing the Existing Operation
For any type of process conversion in a manufacturing operation, looking at the current setup is the critical step. You must know what you have before you can determine what you need and where you want to go.
After meetings between the project engineers and the foundry management team, the first step was to utilize a "clean-piece-of-paper approach" to study the operation, eliminating any preconceived notions of what the problems were with the existing plant or how to alter operations to improve them. This is the same approach taken when laying out the plan for a new foundry. This approach allows design engineers to search out new, creative solutions to help decrease bottlenecks and flow problems throughout the operation when buildings and existing equipment are thought to be fixed restrictions on the process.
The next step was to begin to collect information about its operation. Existing drawing and production data were gathered and photos and videos were taken of the operation in action. Time studies and flow patterns evolved.
Hendrix is a captive operation with more than 2500 patterns, nearly half of which are run within a given year. A problem it encounters is that many of these patterns are unique and must be processed individually, such as chain for dragline buckets, which requires advance planning and would slow down a molding line. During the study phase of the project, the unique requirement of assembling multipart molds for the chain was found to be overly labor-intensive due to alignment problems with the mold sections. In addition, the mix of products Hendrix was casting and the properties of their runs - mold sizes, parts per mold, production run quantities, heat sizes, alloy requirements and cooling times - were analyzed to determine if they were conducive to cost-effective green sand casting.
The analysis of this information was necessary to produce the redevelopment plan that would fit this particular operation.
2 Outlining a Plan for Change
In the planning stages of a redevelopment project for a foundry, the following items, at a minimum, must be addressed:
* goals and desired end-results;
* existing equipment/buildings to be utilized;
* pattern requirements;
* installation sequence/effect on operations;
* cost vs. benefit decisions.
The first decision in the plan was to determine if the foundry was going to remain a green sand operation. The existing green sand system was old and could not produce the quality required. Its existing jolt-squeezer molding machines weren't compacting the mold enough to provide the necessary strength. The result was castings that required extensive work in the cleaning and finishing room (creating a bottleneck there). In addition, because of the layout of the plant, the foundry workers were performing excess material handling responsibilities. These factors led to a production figure of 40 man-hours per ton of castings produced, which wasn't cost-effective for the foundry.
The first option considered for Hendrix was an upgrade of the green sand molding system to some type of high-pressure, semiautomatic green sand line, which would provide the mold quality necessary to achieve its goals. The problem was that the foundry couldn't just add the molding equipment because it would need a new green sand preparation system and a new mold conveyor system (because the current pallet lines couldn't be retrofitted). In addition, such a line wouldn't fit well with Hendrix's short-run, job-shop nature to its production and the pattern modification cost would be high. This option also didn't take into account its existing nobake production, which would have to be modified to be incorporated into new mold conveyors and sand delivery systems.
The next option explored was the conversion to nobake molding. Since Hendrix already was producing nobake cores and half of its sand was being used in nobake molds, there wasn't going to be as great a learning curve for employees if this operation was expanded. In addition, the nobake process presented Hendrix with an opportunity to achieve the higher quality requirements it needed (tighter dimensional tolerances and lower cleaning and finishing costs), but still maintain a cost-effective system with small production runs and lower pattern modification costs. Although binder costs for the nobake sand were going to rise, this would be more than offset by the reduced man-hours per ton in the molding and cleaning and finishing areas.
Based on these parameters, a new phenolic urethane nobake molding plan was drawn for Hendrix with an automated nobake molding system as the centerpiece [ILLUSTRATION FOR FIGURE 2 OMITTED].
The new system incorporates a molding loop with powered roller conveyor, rollover/draw and transfer cars, and a coresetting and closing area. A key feature in the new system is the staging area where the molds are sorted and stored until enough molds are ready for pouring each heat [ILLUSTRATION FOR FIGURE 3 OMITTED]. To prepare the molds for pouring, a staging line is selected, and its molds are automatically moved via a transfer car to the loader and then moved onto a loop conveyor for pouring. This system allows all the pouring to be accomplished in one area of the melt bay to minimize hot metal handling, and cooling on the same line without transfers [ILLUSTRATION FOR FIGURE 4 OMITTED]. Molds are pushed off automatically into an existing mold dump conveyor that is centrally located from the old green sand pallet system.
To optimize the performance of the system, a decision had to be made on the optimum mold size based on pattern conversion costs and product quantity requirements. The result was a mold size of 38 in. x 40 in., which required the remounting of patterns (some with multiple impressions to maximize efficiency in castings produced per hour and sand-to-metal ratio). In addition, some of the green sand patterns required rework for nobake molding. With a forgiving green sand system, cuts, gouges, undercuts or missing fillets may draw, but not with chemically bonded sand.
Once the main production system was outlined and optimized, the other remaining systems were addressed. Short-run, larger molds would be made with a separate, larger mixing station located in the main bay for crane access. Patterns for this molding line are powered under the mixer on a roller conveyor, and then the crane hoists them to the pouring floor for drawing and closing. A separate shakeout was installed for these tight flask molds with the sand going into the same existing conveyor system with the automated nobake molds.
The multiple section chain mold would require a separate handling system. Although some of these mold sections could be made on the new molding loop while others could be made on the old nobake core loops, the decision was made that the assembly of chain molds would delay the new automated molding system and require larger bottom boards than the pouring loop. The solution was a separate pallet car system with 18 in. x 8 ft machined cast iron tops. This would provide a precise, solid base for assembling the overlapping mold segments and reduce shimming, run-outs and casting finishing. These molds are poured and cooled on the pallet cars and dumped into the same existing mold dump conveyor as the other molds.
Sand from all the molds, along with castings from the chain system and the automated loop, are transferred by the old green sand vibratory conveyor to a new shakeout/sand reclamation system. Reclamation of the nobake sand must be part of any large nobake system to offset the cost of new sand and disposal. For Hendrix, mechanical reclamation was deemed the best option initially because of the cost of equipment, but an allowance was made for the future addition of thermal reclamation.
Due to the elevation of the existing vibratory conveyor, the shakeout was placed below floor level. A casting handling unit was installed to remove castings from the shakeout and deliver them to an existing conveyor for degating and finishing [ILLUSTRATION FOR FIGURE 5 OMITTED].
In determining the final layout, the sequence of installation was important. The first step was extending the existing building's main crane bay to install the large mold area. This cleared a side bay for the installation of the molding loop and staging portions of the automated system with minimal interference to production. The large shakeout, outside sand silos and pneumatic transporters were installed with the foundry operating. The old green sand system with pallet lines and shakeout were then removed, with the pouring loop, pitwork and shakeout/reclamation system installed last.
At this point, the insomnia factor really kicked in because, for the new systems to be installed, the plant had to shut down for a week (with weekends, 9 total days). By working the installation crews extra shifts and good planning, this was determined to be feasible. When Hendrix's employees left on Friday for vacation, they would see the last of the green sand system.
3 Developing a Funding Plan for Change
The key to funding a project such as this is to emphasize during decisions the cost vs. benefit of retaining existing equipment, building additions and modifications, and possible interruptions to production. These decisions must be made while weighing the effects on the long-term cost of operations. By considering this influence, an accurate total cost projection for the project can be made to determine funding requirements.
Although the method of funding a major project varies from foundry to foundry, one fact is certain - funding requirements must be determined from a studied and planned program. Accurate total cost estimates are the only sure way to achieve the original goals. Projects that use budgets conceived before the study and planning are completed restrict the final performance of the project.
4 Implementing the Plan
Once the project was approved, funded and started,communication between the project engineers, foundry management and production was critical as all operational factors must be considered during the implementation stage to minimize disruption during start-up.
Pattern changes must start while equipment is being built and installed because simple things, such as core fit, may cause problems during start-up because the nobake molds will not experience as much mold wall movement as the green sand molds did. The delivery of riser sleeves, chills and cores to the molding system must be planned because it may require changes to the scheduling system. Purchased sand may vary in specification from the previous requirements, so up-front communication must exist with the sand supplier. In addition, personnel may need to be retrained and reassigned to positions with the new system.
Then the question is: What happens after the project start-up? The answer to this is how the operation has improved.
The overall cost of the conversion for Hendrix was $3.5 million. In the 4-year period prior to the conversion to nobake molding, man-hours/ton averaged 39.9. Since the conversion, this figure has dropped 31% to 27.51. In addition, the tonnage of castings poured is up 32% since the conversion to 3800 tons/year, and new sand purchased is down 48%.
It is important to note that major conversion decisions must be made on a foundry-by-foundry basis. What Hendrix accomplished may not be suitable for every plant. These results have been achieved through analysis, planning, funding and implementation of the new design. A cooperative team approach smoothed the transition of Hendrix to a modern, efficient nobake foundry, and a leading supplier to the construction and mining equipment markets...hopefully, without losing too much sleep.
The History of Hendrix Manufacturing
The forerunner to Hendrix Manufacturing Co. was the Desoto Foundry and Machine Co., which was established in Mansfield in 1906 as a gray iron foundry serving the sawmill industry in Louisiana and surrounding states. In 1909, Desoto turned to the production of commercial steel castings and became the first steel foundry west of the Mississippi River.
In 1940, Desoto was acquired by Joe W. Hendrix, the patent-owner of Hendrix Dragline Buckets, Guy B. Hall, Sr., L.C. Hall, St., and George E. Trippe. To better identify the foundry with its new product line, its name was changed to Hendrix Manufacturing Co. in 1946. A year later, production of commercial castings ceased, and the foundry focused solely on dragline buckets and related components. Since that time, Hendrix has manufactured more than 53,000 dragline buckets for a worldwide customer base.
In 1985, to augment its line of dragline buckets, Hendrix diversified its product mix to include excavator buckets. In 1989, Hendrix entered into a license agreement with J.B. Sales International of New Zealand to build a hydraulic coupler to use to connect and disconnect attachments to excavators.
Today, Hendrix Manufacturing still is owned and operated by the descendants of the original founders, and the foundry continues to produce dragline and excavator buckets and hydraulic couplers for both the domestic and international markets.
|Printer friendly Cite/link Email Feedback|
|Comment:||Hendrix Manufacturing: a case study in converting a green sand molding system to nobake.|
|Date:||Sep 1, 1999|
|Previous Article:||Identifying casting's challenges.|
|Next Article:||Understanding agglomeration behavior in green sand.|