Resolving Quality, Cost Concerns in Aluminum Permanent Mold Casting.

Several foundrymen share solutions to porosity and design challenges, how they turned defects into opportunities for improving quality and cost reduction

Maintaining tight process controls and learning from previous experience can mean the difference between a foundry that squeaks by and one that sees a healthy margin. A panel presentation on quality and scrap reduction at AFS' 5th International Conference on Permanent Mold Casting of Aluminum, held October 22-24 in Milwaukee, provided several case studies on improving quality while reducing cost.

Permanent mold foundrymen were asked to select a casting that started with either high scrap or high cost on initial runs and then describe the steps involved in correcting the problems. The focus was on the financial savings due to process improvements. The following case studies show the efforts of five foundries to improve profit by reducing casting scrap and by implementing engineering controls to solve casting production difficulties.

This article details solutions provided by Bill Willmert and Scott Blunt, Federal Mogul Corp./Sterling Pistons; Richard Andriano, Stahl Specialty Co.; David Weiss, Eck Industries, Inc.; Peter Petto III, Arrow Aluminum Casting Co.; and Phil Wilmes and Jason Williams, Bodine Aluminum.

Piston Shrinkage Porosity

Bill Willmert and Scott Blunt, Federal Mogul Corp/Sterling Pistons, Maiden, Missouri, discussed remedying shrinkage porosity in a 332-alloy automotive piston. Sterling Pistons, which uses a gravity pour casting method, was experiencing a scrap rate of more than 5% with a 1.7-lb cast piston due to shrinkage porosity on the thrust surface (Fig. 1). After investigating, they determined that the probable cause was a hot spot caused by overheating a metal core located under the ingate. They were unable to enlarge the ingate due to a cleaning room shearing operation, and enlarging the side riser to allow additional feeding did not help. The foundry also tried various mold-coating schemes and varied pouring temperatures. Movement of the core made water chilling difficult and expensive, and air chilling was imprecise.

The foundry finally solved the problem with a unique tooling change. By machining voids in the mold to create air pockets to the 450[degrees] area of the skirt, the foundry produced a hot spot in the mold adjacent to the defect location and then controlled the shrink through mold thermal control. After the mold change, more than 93,000 pistons have been cast in the new mold design with a 2.10% scrap rate.

Flywheel Housing Porosity

Richard Andriano, Stahl Specialty Co., Kingsville, Missouri, discussed his foundry's remedy for shrinkage porosity in a 356.1-alloy T-6 large flywheel housing. Stahl's experience in tilt pouring large castings allowed it to initially pour this 65-lb flywheel without excessive foundry scrap. However, the foundry later discovered internal porosity during subsequent machining operations.

Although there was no surface evidence of shrinkage (such as a "frosty" appearance or sunken surface that might indicate sub-surface shrinkage porosity), the machined castings showed porosity pores that were large (some more than 6mm in diameter), round and located in clusters in localized areas of the casting. Due to the size of the casting, filling problems were suspected as the cause of the porosity, and Stahl's initial investigation identified the problem as air entrapment during mold filling. Corrective efforts were focused on venting the areas that exhibited the defect and enlarging the radius and smoothing the corners of the casting design and gating system for a smoother fill. Despite Stahl's best efforts, the problem persisted.

Significant amounts of "flow-lines" were visible on both the inside and outside surfaces of the casting, perhaps indicating a turbulent fill. Stahl foundry engineers changed the tilt profile and slowed the tilt rate and saw a reduction in the flow-lines, significantly reducing the defect (Fig. 2).

Andriano said that a slower cavity fill time reduces airentrainment, allowing the air in the cavity more time to vent and escape ahead of the molten metal front. Through an experiment that tested various tilt times, tilt profiles (a combination of varying tilt speeds during the tilt process), metal temperatures and metal fill back time, the foundry determined the optimal fill time and profile for this particular casting. Andriano emphasized that the crucial factor is more than just tilt rate (it is possible to tilt too slowly) and that an optimized tilt profile should be determined for each job.

The scrap rate on the flywheel varied widely and occasionally went as high as 810% prior to the optimized tilt profile, but now, it is consistently below 1%, and zero foundry-related defects had been reported for the month prior to the conference. The cost savings are more significant considering the fact that the defect was discovered after machining (the cost of the scrap casting included both casting and machining costs).

Control Arm Porosity

David Weiss, Eck Industries, Inc., Manitowoc, Wisconsin, discussed the shrinkage porosity and blistering his foundry was experiencing with a B356-alloy automotive control arm. Eck was awarded an order for the lower control arms (Fig. 3a) for the Plymouth Prowler. which is safety-critical with tight appearance guidelines. The casting requirements included class "A" surface finish, ASTM E-155 frame 2 or better soundness, and a minimum of 35 ksi tensile strength, 25 ksi yield strength and 7% elongation. Computer modeling in the design stage showed potential shrinkage, but design and aesthetic requirements prohibited the use of risers in the needed areas. The initial casting problems included two areas of shrinkage and some localized blistering near the metal entry point and misruns.

Using thermocouples that were installed at various points in the mold, Eck tested the mold to determine mold temperature at these locations. Through the use of X-rays to help locate problem areas in the casting and by carefully examining the thermocouple results, foundry personnel discovered that they had a reversed thermal gradient in the mold, and the areas of the casting that should solidify first were the last to solidify. So the focus turned to modifying the thermal and solidification sequence through increased mold temperature control during tilt pouring and solidification.

The foundry personnel could not develop a steep enough thermal gradient to correct the problem strictly through the manipulation of mold coating, so they began to look at other areas. They were able to control the mold temperature and promote directional solidification by developing cooling zones by air-cooling localized sections of the mold. Tubes were installed into the mold in a "U" shape (both in and out), and air was used to cool selected areas (Fig. 3b). The cooling is computer-controlled by thermocouple temperature readings so that the air-cooling is used only when mold temperature exceeds the designated levels (Fig. 3b). Eck was able to reduce the mold temperature on the lower part of the mold by 120F (48.9C) and, through tight process thermal control, showed a 50% reduction in scrap rate. The job, which is checked using 100% real time X-ray, is now running at about 5% scrap, and almost all of that is mold start-up related.

Breather Misrun, Shrink

Peter Petto III, Arrow Aluminum Casting Co., Avon Lake, Ohio, discussed his foundry's efforts to fix misrun and shrink problems in a 356-alloy T-6 breather casting. This component originally was set for production as a die casting, but the customer needed production to begin sooner than die casting tooling could be contructed. The casting is a canister-shaped piece with a thin-walled tube running inside. The inside diameters were close-tolerance to accomodate a metal mesh filter and rubber hosing on the outside (added later in assembly).

By working with the customer, Arrow was able to redesign the casting to eliminate the thin wall sections by incorporating low-cost steel tubing. The design change, however, created a heavy boss inside that was bound to create shrinkage problems, especially troublesome considering that the boss would have to be machined and the casting must be leak-free.

The foundry considered using a composite core that includes iron chills in the core sand to help chill the center boss--a system that has been successful in similar gravity pour situations. However, the foundry was able to remove enough mass (by reducing the internal boss with an internal core pin to draw heat away) that it was successful in running the casting without risers or shrinkage.

A secondary issue the foundry faced was the customer's cost expectation. Because the casting was intended to be diecast, Petto knew that the casting selling price would be a critical issue and that the foundry would have to minimize costs to be competitive. Because finishing room costs compose a major portion of the casting selling price, a reduction in cleaning time was an area that received considerable attention in the design stage. The desired cost reduction was accomplished by using an unconventional "one-point" gating system that attached to the machined flange surface (Fig. 4).

Although this casting was eventually converted to plastic, Arrow ran the part for almost 4 years. More importantly, by controlling the selling cost through process controls and through design and engineering innovations Arrow was able to develop a trust with the customer that resulted in cooperative engineering and the acquisition of other casting jobs.

Cylinder Head Porosity

Phil Wilmes and Jason Williams, Bodine Aluminum, Troy, Missouri, discussed their foundry's remedy for porosity on the exhaust port flange of a AC2C-alloy (319) T-6 cylinder head casting (2.2 liter, 16-valve). Bodine, a high-production automotive foundry producing parts for Toyota, was required to produce 320,000 of these low-pressure permanent mold parts annually. The economics of foundry scrap are highly multiplied during production volume at these levels, so scrap reduction and process controls become even more critical. Bodine was producing this 17.1-kg, four-cylinder manifold (for the Toyota Camry and RAV-4) at 1-2% foundry scrap when the levels suddenly started to increase. Defects were discovered only after machining at the customer's facility. The porosity was localized on the flange of the exhaust port (Fig. 5), a section of the casting that is machined and assembled with a gasket.

Metallic oxides and sand inclusions were shown to be the source of the porosity during investigation by scanning electronic microscope and standard microscopic magnification. Because the foundry uses low-pressure mold filling, personnel knew that this defect was probably not gating and filling-related--the oxides must come from the melt itself. Examination of the melting and holding process showed that the holding furnace under the mold was not properly cleaned; there was excessive oxide build-up on the walls. Molten metal inclusion sampling confirmed poor metal quality. When personnel re-examined foundry metal handling and melting procedures, they concluded that their procedures were too restrictive and rigid and immediately changed them to allow more flexibility for melting personnel to make on-line changes as needed, such as more flux for a certain batch of metal. They also experimented with different fill rates. Improved metal processing and a slightly slower fill time improved the scrap rate and these c hanges were maintained as new shop processes.

Although these new procedures reduced the scrap rate, the inclusion defects continued at an unacceptable rate. Foundry personnel turned their attention to their filter practice. Although Bodine does not use filters in the stalk tube, as do many low-pressure permanent mold foundries, a small mesh filter screen (55.9% open area) is inserted at the gate entrance to the casting. Following an experiment with perforated tin plate screen (50.9% open area), the inclusion defects almost immediately diminished. After running with the tin plate screen for 9 months, the foundry achieved a consistent scrap rate below its target rate of 0.75%.

Putting the High Cost of Scrap in Perspective

According to Dan Groteke, QC Designs, reduction of foundry scrap is perhaps the greatest opportunity for profit enhancement in most foundry operations and can be controlled on the shop floor by the foundry operating personnel. He discussed the high cost of scrap castings and the financial incentive to reduce scrap.

Based on an imaginary casting (with a 10% scrap rate) weighing 10 lb and selling at a price of $22.87 each, Groteke defined the high cost penalty and the erosion of net profit due to excessive scrap (Table 1). The full production costs ($2062) include the cost of the alloy, the energy costs to melt, the pattern setup cost, labor to run the castings, general overhead, and the "metal credit" cost that reflects cost of new metal saved due to remelt of the scrap castings.

With 10% production scrap, a run of 110 castings would be required to produce the 100 units the customer ordered. This will generate revenue of $2287, yielding a net profit of $301 (revenue of $2287 minus a production cost of $2062, plus the metal credit for scrap retained in the system), according to Groteke. "The financial incentive to reduce scrap can be easily seen when it is demonstrated that a 5% reduction in scrap rate would net a 25% increase in bottom line net profit," he said. "Conversely, the cost penalty of excessive production scrap is also evident, as an additional 15% increase in scrap (to 15% total) cuts profit margin on the job by 25% (or 40% from the 5% scrap level). The excess scrap rate of 15% also can compound the financial loss if the customer asks the foundry for a rerun to make up for the short, shipment. Including the cost of rerunning the castings, the foundry suffers 56% reduction in net profit due to the excessive foundry scrap."

                          Breakdown of the Costs
                      Associated with Scrap Castings
Cost                   Cost per    Cost    Cost       Cost of
Center               100 Castings  @ 5%    @ 15%       Rerun
                     (@10% Scrap)  Scrap   Scrap
Lot Size                 110        110     110          15
Alloy Cost ($.80/lb)     $880      $880    $880          $6
Melt Cost ($.10/lb)      $110      $110    $110         $15
Set-up Cost              $50        $50     $50         $50
Labor                    $235      $235    $235         $64
General Overhead         $787      $787    $787         $258
Castings Sold            100        105     95           12
Revenue                 $2287      $2401   $2172        $274
Metal Credit             $76        $38    $114         $22
Production Cost        ($2062)    ($2062) ($2062)      ($387)
Net Profit               $301      $377    $224        ($91)
Margin                   Base      + 25%   - 25%  Total of Rerun &
                                                    15% = - 56%