Experiments in steel gating three foundries report: this article takes a look at three different foundries that have experimented in gating and or risering, and the benefits that they have garnered.
This article takes a look at three different foundries who have experimented in gating and/or risering, and the benefits that they have garnered.
Flat Runner Gating
When Harrison Steel Castings Co., Attica, Indiana, began looking for ways to improve its casting cleanliness, it turned to flat runners, an effective and practical method for separating nonmetallic inclusions. Harrison began experimenting with flat runners in 1999, realizing improvements in casting cleanliness. Harrison considered expanding the use of flat gate technology to its daily green sand production for castings having cleanliness problems.
A flat gate runner is designed to trap foreign particles from the metal once they leave the funnel or down sprue and enter the runner bar. The particles must travel slow enough while in the runner to float and stick to the topside of the runner before the metal enters the mold cavity. The flat gate helps to constrain the metal with little exposure to air.
Harrison tested the flat gate technology on a 12-in, by 34-in, test plate that was 3 in. thick and had 408-sq-in. of cope surface area, which represents 41% of the total casting surface. Four sets of five plates were cast in nobake molds. The firm went through four different trials to see how effective a change to flat runners would be (reference Table 1 for data corresponding to each trial).
Trial 1--The first set of five plate castings were poured through a 13-in. shell funnel for 14 sec into a 3/8-in, flat runner system shown in Fig. la. The cope surface of the flat runner is clean, but the castings had severe cope surface defects, averaging 19-defects equaling 31-in, of dirt as shown in Fig. 2.
Trial 2--After viewing a casting process model of the previous gating system, Harrison concluded that the velocity was nearly 1 in/sec. In an effort to reduce the cope surface defects, Harrison decided to slow down the flow of metal in the gating system by increasing the height of the runner and ingates from 3/8-in, to 3/4-in. The results of the second trial continued to have a cope surface with defects similar to the first samples.
Trial 3--The pattern length was increased to 36-in, and the runner and ingates were increased to 1 1/2-in, to make the metal flow slower than in the previous trials. While the castings improved over the previous two trials, the test proved that castings with a high percentage of flat cope surface area are difficult to produce without cope surface defects. Two of the castings were rated with class 2 defects while the other 3 castings were rated either as class 4 or 5.
Trial 4--Harrison took a different approach on casting the final trial, using a traditional 2-in, diameter, ceramic tile bottom gate system (Fig. 1b) and a pour basin to reduce the velocity. These castings appeared cleaner than the castings with the flat gate system.
Conclusion--The flat runner concept is such that the turbulence created from the higher velocity allows the particles to bounce around and stick to the cope side of the runner. However, the exit velocity from the ingate is so great that the metal splashes around and is severely damaged in the mold cavity, resulting in poor test results. If the metal can enter the cavity as low as possible with a low velocity, Harrison believes that the flat gate can work to reduce the cope surface defects.
Saw Cutting Large Diameter Risers
Controlling casting concentricity and overall length is critical when producing parts that will be run on CNC machining centers, In rapid traverse mode these machine tools can move at up to 400 in/mm. If the casting sizes are out of tolerance as prescribed on the casting drawing (which is used to set-up the machining programs), a crash will occur, resulting in downtime and damage to multimillion dollar machines.
For the last 20 years, The Falk Corp., Milwaukee, accomplished casting length control in a two-step process. First, the risers were removed by flame cutting. Secondly, the cope surfaces were milled to within specified tolerance, This process insured that the parts were within the tolerances specified on the casting drawing.
Although double processing worked well, it was clearly no longer cost effective, The 42-year-old milling machine, located in the foundry's cleaning room, was incurring high maintenance costs. In order to address the problem, Falk initiated a project to find an alternative that included reductions in material handling, set-up, lead-time and energy costs.
The choice was made to complete both riser removal and CNC qualification in one step. The obvious solution was to cut the riser away, at a highly controllable size, with an industrial saw, The search began for manufacturers and users of industrial saws.
Finding a Manufacturer--Difficulties arose after saw cutting was chosen as the method for meeting the goals. The largest hurdle was finding a saw manufacturer. The saw needed to be designed and built capable of cutting full contact risers up to 30 in. in diameter, at a feed rate of 1 in./min. Most manufacturers did not have the equipment to satisfy Falk's requirements, and it was cost-prohibitive to design a custom saw.
Falk found a manufacturer and had the saw delivered in February 2002. The machine was installed in one of Falk's machine shops, so that upon completion of the sawing operation, the part is already near the next scheduled operation. The existing milling machine foundation needed little retrofitting to be adapted for the new saw.
As is the case with many new installations, there was one major obstacle that placed Falk's project in jeopardy--poor blade performance. There was a major drop in performance when compared to the preshipment acceptance tests. The firm found that the blades supplied with the saw were not the same as the blade used in the acceptance testing.
Blade life when compared to the originally specified feed and speed rates had to be reduced by 50-70% of the specified rate to achieve blade life of a minimum of 150 castings per blade. At the reduced levels the process proved too costly to justify continued use. A manufacturer was located that could provide a blade that would cut at the prescribed feed and speed rates and with acceptable blade life.
Evaluation--The question remained, "Does the new process meet the original goals?" Evaluation first centered on the primary goals, gas cut versus saw cut cost, material handling, set-up and lead-time.
During August 2001 through December 2001, 1320 castings from 58 different patterns had the risers removed by the saw cut method. The result was a 5.0% total cost savings over the gas cut method. Castings are shipped in three conditions: annealed, normalized and tempered, and quenched and tempered. For castings that are shipped in the annealed condition, lead-time was reduced by five days, set-ups reduced by one and material handling reduced by seven occurrences.
There also is energy savings when the castings are saw cut due to the removal of the pre-heat operation that was required for gas cutting. The blade life was recently improved to 375 castings/blade after an upgrade to the blade coolant system. The percentage of water-soluble lubricant also was increased to extend the blade life. The improvement lowered the blade cost per casting to less than one dollar, with overall manning reduced by one.
The range of casting styles now routed to the saw rather than to gas cutting has increased, allowing Falk to take advantage of the full capabilities of the machine, including both captive and contract castings. The process adds value and reduces cost and lead time.
This article was adapted from presentations at the 2001 and 2002 Steel Founders' Society T&O conferences.
For More Information
"Cost Reduction Via Bating at Wisconsin Aluminum Foundry," T. Dimmick, MODERN CASTING, March 2000, p.31-33.
Table 1 Harrison Steel Casting Co. Trial Data Runner Runner Horizontal Pour Dwell Thickness Area Runner Time Time (in.) (sq. in.) Velocity (in./sec) (sec) (sec) Trial 1 0.375 2.25 34.70 14 0.26 Trial 2 0.750 4.50 22.08 11 0.40 Trial 3 1.500 9.00 12.55 10 0.91 Trial 4 2.0 Diameter Tile 3.14 20.92 18 NA
RELATED ARTICLE: Evaluating Cating & Risering at Talladega Castings
In 1992, Talladega Castings and Machine Co., Inc., Talladega, Alabama, began testing the principles of its gating and risering systems for steel castings. The foundry used two basic runner systems: a non-pressurized 1:4:4 ratio, on castings less then 300 lb, and a 1:6:6 ratio on castings 300 lb and above.
These two basic runner system ratios were arrived at by tracking ceroxides. Talladega determined that ceroxides in the surface finish were more prominent on castings using smaller runner systems that caused longer pour times and increased turbulence. To compensate and improve quality, the size of the runner and ingate were increased.
Round tapered sprues were used with a choke area of 1-in. or 0.7854 sq. in. for most of the 1:4:4 ratio gating systems and 1.25-in, or 1227 sq. in. choke area for the 1:6:6 ratio gating systems. A 4-in, diameter by 4-in, deep well base was generally used on everything, never going smaller than a 3-in, diameter by 3-in, well to give the metal falling down the sprue a good cushion. A drag runner and cope in-gate also were used in conjunction with as wide and thin a runner and in-gate system as the board, flask and casting configuration, and wall thickness' allowed. Talladega found that with a wide flat runner system more impurities and inclusions (such as slag) sticks to the larger surface area on the cope. This larger surface area allows this to happen without slowing the metal flow.
Talladega tracked the success of its new gating methods by measuring the usage levels of welding rods. In 1992, the foundry determined that 1060 tons of castings were shipped. During this time, 600 lb of welding wire and rods were used per month. This includes steel and stainless steel wire.
In the year 2000, 1521 tons of castings were shipped, with only 800 lb of welding wire and rods being purchased. This was a reduction of 90% just in welding rods, not to mention time spent welding and grinding. The improvement in surface finish also cut down on cleaning room labor.
About the Authors
Tim Hays is the superintendent of casting engineering and design at Harrison Steel Castings Co. George A. Hartay II is the manager of technical services at Falk Corp. Chuck Ball is the Quality assurance manager at Talladega Castings and Machine Co.
|Printer friendly Cite/link Email Feedback|
|Date:||Mar 1, 2003|
|Previous Article:||Developing a 'feact-based' marketing plan: this article provides a framework for preparing a "grounded in the facts" strategy to mobilize your...|
|Next Article:||6 Steps to reducing inclusion defects.|