Improving Steel Operations--From Riser Design to Finishing.

This year's "T&O" offered insight into the use of spherical risers to improve yield, more efficient ladle practices, and grinding methods to reduce vibration issues.

Marking another year of increased technical insight, the Steel Founders' Society of America (SFSA) played host to 215 attendees from both the U.S. and abroad at its 54th Technical & Operating Conference, held Nov. 2-4 at Chicago's Hotel Inter-Continental.

The conference's 37 speakers discussed a variety of topics, including gating and risering practices, re-examining common industry standards for X-ray analysis and feeding rules, effective use of refractories and slag techniques, and the metallurgy of cast steel, as well as ergonomics and environmental concerns. This article covers three presentations from the conference that run the gamut of the casting process from improving riser design and melt quality to bettering grinding operations.

Risering Practices

"It's often difficult to design a riser system that will produce a casting free of internal defects while using the least amount of feed metal," said Tim Hays, Harrison Steel Casting Co., Attica, Indiana, in his discussion of spherical risers. "The riser must have a greater freezing time than the section to be fed, and sufficient liquid metal must be available to assure that the last area to solidify is within the riser. It's also important that the shrinkage pipe does not extend into the casting."

In an attempt to improve yield as a result of a change in riser design, a ring casting was submitted to the Univ. of Iowa as a part of SFSA's Yield Improvement Program. This particular casting has had four different riser styles and two different gating systems since it was first manufactured, according to Hays, who said that the sand riser was eventually replaced with a feeding aid. In order to address the problem of under-riser cracking, a bell shaped spherical riser was tested.

Independent of the shape of a steel casting, bodies with equal moduli (ratios of volume to surface area) will not solidify at the same time, Hays said, stressing that computer modeling of a 6-in. sphere and a 6-in, cube showed different results. "A 6-in, sphere and a 6-in, cube have the same modulus," he said. "However, the cube has approximately twice as much volume and surface area. The cube takes about 30% longer to solidify when modeled on three different finite difference systems using independent databases for thermal properties."

The entire ring casting has a modulus of 1.08 in., while one section of the design is 1.14 in. The modulus would he 1.29 in. if engineers considered only the area directly beneath the riser. The original 7.5-in.-diameter riser had a modulus 20% greater than the worst case modulus of 1.29 in.

The riser pipe depth for the 7.5in.-diameter riser was 7.74 in. Computer modeling agreed with the calculated pipe depth, however the casting had open shrinkage. The riser diameter was increased by 1 in., and no open shrinkage existed.

The riser was changed to a 6-in.-diameter sleeved riser with a head core to eliminate "dirt" created by poor ramming of the green sand mold at the base of the riser. The 6-in.-diameter sleeved riser did not have open shrinkage, and the yield was improved, but a new problem arose when under-riser cracks appeared, Hays said. A6-in. bell spherical riser was added, which eliminated under-riser cracks as well as under-riser shrinkage.

The Univ. of Iowa concluded that by reducing the size of the sleeved cylindrical riser, the yield would approach that of the bell riser without significantly increasing carbon or sulfur under the risers (Table 1), Hays said. "The results from a computer model indicated that the part would be sound, while under-riser macro-segregation would be no different that the two cases modeled previously," he said.

Two castings were poured to validate the Univ. of Iowa's conclusion. One casting was re-rigged with the phenolic urethane nobake head core and insulating sleeve poured 9.63 full, while the second was poured with a bell mini-riser. Both castings were poured from the same heat to eliminate variation, and both were X-rayed and sliced through the center of the riser and casting without heat treatment. Slices were macro-etched to enhance shrinkage and segregation.

The sliced sections indicated no evidence of chemical segregation in the 400-lb casting, agreeing with the computer simulation.

"The outcome was definitely an improvement in yield," Hays said. "The bell mini-risers are rather expensive and actually exceed the savings in hot metal at the tap, but they did eliminate under-riser cracking, and there was savings in the cleaning room."

Increasing Ladle Refractory Life

"One of the main reasons that foundries tap at elevated temperatures relates to the nature of a typical foundry operation," said Kent Peaslee, Univ. of Missouri-Rolla, Missouri, in his discussion of ladle practices. "Most foundries require a considerable amount of time to complete the number of castings in sequence, which results in large heat losses from the ladles. However, in many cases, these heat losses are much greater because of inefficient ladle designs and practices."

The presentation, which was co-authored by Jeffrey Smith, Univ. of Missouri-Rolla, covered a research program funded by the Dept. of Energy and SFSA to determine the best combinations of refractory materials, slag compositions and melting practices to increase refractory life in electric arc furnaces.

Essentially, after steel is tapped into the ladle, heat is lost as steel "soaks" the refractory, bringing it closer in temperature to itself; via conduction through the walls of the ladle; and by radiation and convection losses at the top surface of the slag/steel to the atmosphere, Peaslee said.

Two foundries, one using an acid practice and one using a basic practice, were studied to develop a heat balance and heat transfer model. This model could be used to evaluate preheating practices with ladle designs, and model predictions were compared with actual measurements in the foundries to validate the operation of the model, Peaslee said.

In the acid foundry, 10-ton teapot ladles lined with 80% alumina rammed monolithic refractory are preheated vertically by placing a natural gas burner in the pouring spout prior to tap. Because no cover is used, much of the heat can escape through the top of the ladle, resulting in non-uniform and inefficient preheating, Peaslee said.

At this foundry, plain carbon and low-allow steels are tapped at 3150F (1732C), and after tap, a carbon-based insulating topping is added to the top of the ladle. The ladle then is transported with an overhead crane to a cradle, where the ladle is tilted and steel is poured into smaller teapot pouring ladles at 2900F (1593C). The smaller ladles are lined with disposable silica-based refractories.

The steel temperature decreased from 3150F (1732C) to 3010F (1754C) when it was transferred to the ladle. Temperature then was measured with thermo-couples every 2-3 min after tap and during transfer to the pouring ladles. During the first 5-10 min in the ladle, steel temperature dropped more than 100F due to the soaking effect (an indication of inadequate preheating). The steel temperature dropped again during the last 15 min of the heat due to temperature stratification from radiation and convection heat transfer, Peaslee said.

With a magnetic thermometer attached to an outside wall, ladle shell temperature was measured. In the first 10 min after tap, temperature increased 80F, correlating with the loss to soaking. "If the ladle had been adequately preheated, the shell temperature would remain unchanged after tap rather than rapidly increasing," Peaslee said.

The model derived from the acid foundry predicts that at least 125F could be saved in the tap temperature by improving the preheating system and ladle cover, Peaslee said. "A 100F decrease in the tap temperature should be achievable for this foundry."

In the basic foundry, ladles are preheated using a gas burner and lid to aim for an average inside surface temperature of 1510F (821C). After removing the ladle from the preheater, the ladle typically is left uncovered for several minutes waiting for tap, and the inside ladle temperature dropped from 180-200F. At this point, 15 tons of plain carbon or low-alloy steel at 3050F (1677C) are tapped into the preheated ladle.

Based on the model developed for this furnace, ladle preheating should be improved to prevent the large temperature drop, Peaslee said. The model shows that soaking loss accounted for a 91F drop in steel temperature, and the second largest path of heat loss was radiation/convection loss at the top of the ladle. "If the ladle was preheated to 2200F (1204C), a large part of the energy lost due to soaking would be saved," he said.

The most effective method to reduce radiation at the top of the ladle would be to place a covering on top of the ladle as either a lid, insulating slag topping or a combination of both. However, a cover is more expensive to construct and maintain, and by adding an extra inch of synthetic slag to the top of the ladle, the radiation and convection loss would be decreased by just over 50%, he said.

By improving preheat practices and adding a lid or topping to the ladle, these foundries could avoid l00F in heat loss. In both foundries, a decrease in the required tap temperature of 100F could save $3.50-5/ton of cast steel based on estimated energy, refractory and electrode savings, Peaslee said. He stressed that additional savings were possible in cleanersteel based on lower dissolved oxygen content, less skull formation in the ladle, a decrease in scrapped heats due to a clogged spout or cold metal, and possible reduced manhours due to more efficient and therefore quicker production.

Vibration Issues

"The adverse effects of exposure to excessive grinding-induced vibration in the foundry have been identified as vibration white finger, hand-arm vibration syndrome and, perhaps, carpal tunnel syndrome," said Robert Beebe, Acme Abrasive Co., Warren, Michigan, in his presentation on decreasing vibration problems arising from the use of abrasive wheels. "The areas involving grinding in the foundry are generally portable grinders, swing frame and offhand grinding on floorstands or pedestal machines--at these locations foundry workers could develop vibration-induced problems."

In an investigation by Service Engineering, more than 180 4.5-in.-and 7-in.-diameter Type 27 wheels (from a number of manufacturers) were tested using Third Octave Band vibration analysis with filters that have upper and lower passband limit ratios of 2.33. To determine how grinding wheels contributed to overall vibration experienced by foundry operators, vibration was measured in three orthogonal directions using a triaxial transducer so that vibration changes could be determined in all directions. The 4.5-in wheels were tested using 10,000-rpm and 8000-rpm electric grinders. The 7-in, wheels were analyzed using an electric drive 8000-rpm angle grinder, a 7000-rpm conventional vertical spindle pneumatic grinder and a 6000-rpm pneumatic turbine drive antivibration angle grinder.

Experimental results showed that for wheels with nominal imbalance and bore gaps, the overall vibration is influenced most by the machine on which they are mounted and how they are used. The air turbine-driven grinder provided significantly lower vibration levels, but in all cases, the measured vibration was lower than the theoretical predicted vibration.

Based on the testing as well as his experience with foundry grinders, Beebe had several recommendations for improving grinding operations. "Use better portable wheel grinders, which present lower vibration to users," he said. Beebe stressed that pneumatic geared turbine-driven machines with vibration damping devices, die grinder designs with a spindle suspended in soft elements, electric grinders with computer-developed optimized mass distribution, floorstand grinders with a single-speed spindle, and both robotic and new swing frame cup grinders all reduce or eliminate vibration problems.

"Vibration damped portable grinder handles, while nowhere near as effective as the turbine grinders, do provide some help," Beebe said. "Unfortunately, they are often abused and poorly maintained in the foundry environment, and their damping value is lost." In addition, a new work glove incorporates a pocket of anti-vibration lead shot surrounding the area of contact within the grinder handle, however it is heavy and expensive, he said.

Beebe also recommended that foundries match the abrasive wheel to the job at hand. "Wheels that are of a specification unsuited to the job will wear unevenly and cause increased vibration in the grinding system."

In addition, foundries should purchase only quality wheels with controlled bore and imbalance, he said. "A sloppy fit results in increased grinder vibration. You should know a lot about the wheels you are buying and understand that most grinding wheel brands are not interchangeable."

By developing a formal written grinder-monitoring program, foundries can help control vibration and increase safety, Beebe said. This program should cover: air pressure control and filtering devices; spindle speed (the free running speed must be less than the rated speed of the wheels); spindle diameter of the machine and bore diameter of the wheels (fit must be close but not result in forcing a wheel onto a machine spindle); flange condition; functioning of air lubrication devices (turbine drive grinders do not require oil in the air for lubrication); speed regulating the governor of conventional air grinders and spindle bearing for all machines; use of guards; and function and adjustment of speed control devices for variable-speed floorstand grinders. "Provide logbooks for recording and verifying activity, and develop an effective means to correct problems."

                    Summary Results of Testing on Sand,
                    Sleeved and Spherical Riser Systems
        Riser          Riser       Riser       Riser   Segregation
        Size          Modulus  Solidification   Pipe      Pipe
                                    Time       Depth      Depth
   7.5 x 9.63-in.
     Sand Riser       1.54 in.   46.71 in.    7.74 in.  12.9 in.
   8.5 x 9.63-in.
     Sand Riser       1.72 in.   57.94 in.    6.09 in.  10.15 in.
    6 x 9.63-in.
    Sleeved Riser     1.68 in.   59.68 in.    6.89 in.  11.48 in.
6-in. Spherical Riser 2.14 in.   68.88 in.    3.92 in.  6.51 in.
        Riser            Casting       Riser   Percent
        Size             Results      Weight    Yield
   7.5 x 9.63-in.
     Sand Riser       Open Shrinkage 240.31 lb  62.5%
   8.5 x 9.63-in.
     Sand Riser        No Shrinkage  308.6 lb   56.5%
    6 x 9.63-in.
    Sleeved Riser      No Shrinkage  155.21 lb   72%
6-in. Spherical Riser  No Shrinkage  80.59 lb   83.2%