Cost-Reduction Via Gating at Wisconsin Aluminum Foundry.

In its efforts to reduce costs, this foundry has successfully used combinations of filters, pouring devices and chill fins to reduce cycle times, scrap and lead time.

When it comes to dealing with today's business climate, Wisconsin Aluminum Foundry Co., Inc., Manitowoc, Wisconsin, is in the same boat as many other foundries--lead times are being shortened, margins are being squeezed, and the market requires that the foundry ask more of its people and processes than at any time in the past. Experimentation with gating techniques has helped the 500-employee nonferrous foundry reach its cost-reduction and efficiency goals.

Wisconsin Aluminum, which casts green sand, nobake and permanent mold components up to 1500 lb, has benefited from a combination of past experience and an openness to new ideas and techniques in its gating and feeding designs. The focus of ongoing efforts has been optimizing gating and cooling in aluminum castings, and the continuous improvements at the foundry have led to greater efficiencies per man-hour and faster throughput.

Following are the results from several case studies that involve innovative gating design and the use of chill fins. By sharing in the foundry's gating solutions, readers can glean practical techniques to bring to their own operations, perhaps leading to some success stories of their own.

Marine Valve Housing

Prototype tooling often takes the form of a flat-back pattern and cover cores produced in dump boxes, as in the case of a four-cavity, 0.0187-lb valve housing on a 12 x 14-in, board for a marine products manufacturer.

Due to the size of the flask and configuration of the component, designing the gating down and around the core and into the part would have been difficult and time-consuming. In addition, the design of the part doesn't offer good flat areas at the parting line on which to locate the ingate (any ingates located on this parting line would have made for difficult clean-up).

A0.875-in.-diameterplugwas mounted in the corebox to make a hole in the cover core opening into the part. In the molding process, a pipe core is set in the drag mold. The cover core is set over the top of the pipe core, leaving the 0.875-in. hole exposed at the parting line. A 1 x 2in. fiber screen is placed over the top of each hole, through which the runners drop the metal directly into the casting.

The gating consists of two 1 x 1-in. runners at a right angle to and underneath two feeder bars (Fig. 1). The integrity of the final cast part was not compromised at the parting line, leaving only one flat 0.875-in.diameter ingate surface that was relatively easy to saw and sand flush.

Upper Housing Casting

The foundry encountered a similar situation with an upper housing casting for the same customer. Once again the tooling was a flat-back pattern and a dump box for a cover core. Again, rather than gate down around the core and into the part, the metal was dropped directly through the core and into the casting.

Unlike the previous example, instead of using a fiber screen, a pouring cup and filter were pasted directly onto the top of the cover core set in the drag. The finished part has only one ingate area, a 2 x 0.313-in, crescent shape (Fig. 2a). Minimal gating time spent on the pattern and minimal clean-up time on the finished part allowed the foundry to supply castings to the customer quickly.

As sometimes happens, the customer's usage projections changed, and the foundry was asked to produce more than 1500 parts using the prototype tooling. To accomplish this, the impression was removed from its board and remounted to an offset plate for an automatic molding machine. A block at the base of the sprue was sized to fit a 2 x 2-in. 20 ppi ceramic foam filter. In the draw station, the cover core was set in the drag, and a filter was inserted into the opening under the sprue until it fit tightly. The final part cast on the automatic high-pressure molding machine used the same ingate as the original prototype with a sprue and filter, rather than a pouring cup and filter (Fig. 2b).

Engine Mounting Plate

The foundry has successfully used this filter and sprue combination on many other castings, such as this 4.75-lb engine mounting plate in 319 aluminum (Fig. 3). In the initial design stages of this job, the customer was asked to enlarge an already existing boss to 2 in. around--large enough for a ceramic filter. A 20-ppi round ceramic foam filter is shoved into the enlarged base post for a 1-in. sprue until it is lodged tight, and the casting is poured directly through the sprue and filter, resulting in a casting with excellent yield.

Alternator Bracket

A pouring device located on a 2-in.-diameter pad allowed for quick turnaround for samples on a 4.3-lb 356-T6 alternator bracket. However, the foundry's cleaning department objected to the pouring device being located between two raised features, which made for difficult clean-up. Before proceeding to production, the department asked that alternatives be explored.

No other feature on the casting was suitable for a pouring device, so a new device, a "filter post," was located on a heavy section at the opposite end of the casting (Fig. 4). This feature is a tapered riser with the bottom matched to the configuration of the feature being gated and with the top squared to match the filter shape. When the mold is rammed and the tapered riser removed, a filter is inserted into this opening (where it is positioned I in. above the casting), and metal is poured through this riser and filter.

CHILL FINS

The foundry was having problems with chill-related defects on a high-production green sand job on one of its automatic molding lines. Based on a reference from the AFS 101St Casting Congress on the use of cooling fins to enhance solidification rates (modern casting, June 1997), the foundry experimented with chill fins on this job, with dramatically positive results, leading to greater experimentation with chill fins on other difficult-to-cast jobs.

While chill fins must be removed at finishing, they increase production efficiency because chills do not have to be inserted manually into each mold. In addition, they control hot spots in the same way as chill blocks--but without the added expense.

Hydraulic Solenoid Housing

The chill fins were first tested on a 14.25-lb 319-alloy hydraulic solenoid housing. The original process required a 2 x 4 x 1.375-in, block chill on the heavy front section of the casting. The foundry was experiencing a 4.8% scrap rate for various problems, including sand inclusion defects around the chill and blows on the chill face. Setting the chill also was labor-intensive, and the mold cycle time was averaging just over 52 molds/hr.

A pattern was made for a chill fin section to be inserted into the pattern. The dimensions for the three chill fins (2 x 3 x 0.0625-in., plus draft) were calculated according to the formula in the original modern casting article (0.1 x half the plate thickness wide and 4 x half the plate thickness tall for the 1.5-in. cross-section to be chilled) (Fig. 5).

After one mold with the chill fins was poured, the casting was sectioned and X-rayed. Since the casting appeared sound, the foundry notified the customer of its suggested revision to the gating system and requested feedback as the project progressed. A trial run of 25 castings was produced and shipped to the customer for testing.

From that run, 24 out of 25 castings machined cleanly and passed pressure testing. One casting was rejected in machining for dross and not tested. Upon return of this casting, it was determined that the defect was minor and would never have been cause for a reject in a regular production shipment. This was seen as an indication that these castings were examined very critically--and served to increase the foundry's confidence in the process change.

A 200-piece run was segregated, flagged as a "Process Change Trial Run," and shipped to the customer for further evaluation. The customer informed the foundry that no significant problems were noted on the trial lot of castings, and the change was approved.

As of July 1999, the foundry had produced 28,532 castings with the chill fins, gaining an average of more than 5.5 molds/hr (up to 57.97-an 11% increase) and realized a 30% reduction in scrap.

Rectangular Bracket

The foundry was experiencing low productivity and high scrap and rework due to dross and shrink on a 9.5-lb 319-alloy rectangular bracket, This was an inherited job with minimal room in the flask for gating. Gating was a spine in the open center of the casting with four runners extending from the center directly into the casting, which made for difficult clean-up.

To improve the design, the conventional gating was eliminated, the pattern was flipped cope-for-drag, and the metal was poured through a sprue and filter directly into the casting. The castings exhibited no dross and reduced clean-uptime, but there were two areas of shrink in the short sides of the frame. Two 2 x 4-in, chills in the cope and two 1 x 1-in, chills in the drag were added to produce a sound casting.

To improve productivity, two chill fin sections were added at the location of the cope chills (Fig. 6). A block was inserted into the pattern with three fins 2 x 3 x 0.0625-in. at the top (plus draft). This fin section in the cope eliminated all four chills, increased molds per hour 18% (to 50.3 molds/hr), improved yield and decreased clean-up time.

Water Inlet Elbow

The quoted rate of 45 molds/hr was not being met on a 319-alloy, four-cavity, 1.6-lb water inlet elbow pattern. The foundry was averaging 27.8 molds/hrdue to the labor involved in setting a rib chill on each cavity. The boss being chilled was angled slightly from the horizontal, requiring three studs protruding from the pattern to hold the chill in position and making for a messy clean-up.

The foundry experimented by eliminating the chill on two cavities and inserting chill fins instead (Fig. 7). On the sample run, the castings that were conventionally chilled were indistinguishable on X-ray examination from the castings with the chill fins.

Water-Cooled Manifold 

The foundry was experiencing a reject rate of more than 30% for shrink-related leaker defects on a 15.3-lb 319-alloy water-cooled exhaust manifold. The inherited job had three large cold risers on the top of the casting meant to feed eight internal bolt mounts through the casting from cope to drag.

The job went through a series of improvements including the elimination of cold risers and chilling of the bolt mounts in the cope and drag. The foundry eventually eliminated the chills in the cope by mounting a chill fin on top of each of the bolt mounts (Fig. 8). The reject rate has been reduced to below 2% and molds per hour increased 10% (to 22.8 molds/hr).

Water Box

A 13.6-lb 319-alloy "water box" casting required what seemed to be impossible gating. The customer wanted no ingates located at any area near the parting line around the outside of the casing.

This made the casting an immediate candidate for a pouring device. The device was located on the top center of the part as posts hold it upright. Initial samples showed the pouring device was able to feed the entire casting, except the two sides of the diagonally oriented flange face.

This area required placing a chill, which was labor-intensive. The 1 x 3-in. rib chill had to be hand-tucked on the angled flange face and occasionally fell out of the mold wall and into the casting during pouring.

A chill fin application directly abutting the coreprint was attempted. Each fin is 5.75 x 1.25 x 0.094-in.-thick at the top, tapering to 0.375 in. at the parting line (Fig. 9). Average molds per hour increased 12% (from 34.6 to 39.1), and the foundry realized a 25% reduction in scrap (from 2.1% with the chill to 1.6% with the chill fin).

Ongoing Improvements

Chill fins continue to be a major cost saver for the foundry. The most successful applications seem to be small-to-medium, isolated heavy sections located on the horizontal. The foundry has had less success with larger, more massive sections and on bosses or heavy sections located out of the horizontal.

Initially the foundry was making up the chill fins one at a time in its machine shop, doing the section thickness calculations specifically for each casting tried. Conventional chills were replaced with chill fins on so many jobs that the in-house machine shop began complaining about the amount of time put into manufacturing them.

It was noticed that, though the chill fins varied dimensionally somewhat when the section thickness calculations were followed, many were similar in size and shape. It was suggested that a couple standard size chill fins would suffice--maybe the chill fin dimensions didn't have to be calculated exactly to be effective.

The engineering department worked with a pattern shop and made two standard chill fin sizes out of a flexible vinyl material, corresponding roughly to two of the most commonly used conventional chills. This made chill fin implementation even easier. For sample purposes, a chill fin now can be glued to a pattern, and if proven effective, the machine shop insets the vinyl fin permanently into the pattern. An important side benefit is that these flexible fins don't snap off if the pattern is mishandled (which happened in the past).

This article was adapted from a presentation at the AFS 1st International Conference on Gating, Filling & Feeding of Aluminum Castings. Proceedings are available from AFS Publications at 800/537-4237.