## Die Compaction

The major difficulty with die compaction traces to die wall friction. This friction inhibits ejection and, more importantly, causes density gradients in the green compact. Punch motion against the powder is similar to plowing snow. Close to the punch, the packing is dense, but far removed from the punch, the powder is unaffected. This pressure decay with distance is because the powder spreads load to the die wall in the form of friction.

Green density increases with compaction pressure. Consequently, density declines with distance from the punch. When the compact length is approximately six times the width, then the powder is poorly compressed. This is among the reasons why most die compacted parts are squat, to minimize density gradients. Also, density gradients are reduced when both the upper and the lower punch move toward the compact center. Double action pressing is a term used to describe the application of stress from both the top and bottom punches. In one variant, the die floats with respect to the lower punch, giving double action even though the lower punch remains stationary. Positive die motion ensures uniformity in green density. Inside the compact, the point where the lowest density occurs is the neutral pressure plane, and in most instances, the desire is for the neutral plane to be in the center of the compact.

As the compaction pressure and density increase, the die wall friction also increases. The die wall friction is proportional to the radial pressure when the powder is compressed uniaxially. By the end of the compaction stroke, wall pressures exceed 50% of the applied pressure. Thus, the higher the applied pressure is, the higher the radial pressure and concomitant die wall friction.

If pressure decays with depth in a compact due to wall friction and green density depends on local pressure, then die compaction results in green density gradients. Besides friction with the die wall, an additional factor is friction on the punches and core rods. Consequently, an objective of computer simulations is to calculate the green density contours in a multiple level compact pressed under a variety of press or tool options. If there are large density gradients, then the compact will exhibit nonuniform shrinkage or swelling in sintering, resulting in component distortion and loss of precision. Low density regions tend to change dimensions more in sintering as compared to high density regions.

Minimized density gradients are important to final dimension control. Accordingly, much study has been focused on reducing density gradients. Unfortunately, most cures possible by changing the powder or alloy are ineffective, but a few concepts are important. However, computer simulation techniques show some important options. These include keeping the compact thin in the pressing direction (thickness), lubricating the die wall, and double action pressing. One alternative being studied is to move the top punch in a circular motion, which causes powder flow in a sideways motion to help eliminate low density regions. The need for minimum density gradients increases as the component becomes more complex. Since die wall friction is the culprit, the more die wall there is per unit volume, the greater the problem. This is one of the major barriers to the production of precise gears to P/M, since gear teeth have a large surface area leading to more die wall friction and larger density gradients.

Because of density gradients, problems occur with local weak spots, cracking on ejection, and subsequent component warpage in sintering. Thus, much effort is being applied to calculation of density gradients in pressing. Usually the friction between the powder, die, and punches is the main cause of density contours, but there is possibly a large effect from the die filling operation, especially if powder separation or nonuniform flow occurs. Most of the calculations are performed using a concept of finite element analysis. In this approach, the component is numerically subdivided into hundreds of tiny computation cells. The placement of these calculation cells is visible on the component as a fine mesh. Each cell compresses and deforms based on the forces from the neighboring cells using equations that relate pressure to density. Computer calculation of what happens in each cell then determines the flow of mass and stress through the compact, resulting in the plotted density contours. Soon, most tooling design and compaction setups will be guided by such computer simulations, with the goal being reduction of density gradients and weak spots in the compact. The aim is to make the tooling correct with minimal trial and error. Most tool design is assisted by computer-aided engineering packages. These packages allow accurate drawings and analysis of form, fit, and function. Also, finite element analysis is used to assess the stresses on the press and in the tool sets. Accordingly, the leading press manufacturers provide codes for analysis of press frame deflection. Of course, such efforts require knowledge on the part size and shape, and powder characteristics.

Most presses operate in an open-loop control mode. The position of the various tool members is adjusted at the beginning of a production run, but no monitor is provided for following operation conditions. Parts are selectively checked for weight and dimensions, but the preservation of the desired tool motions, pressures, fills, and other characteristics of compaction is not ensured.

A few advanced presses include pressure sensors, displacement transducers, strain gages, temperature sensors, position sensors, and other monitors needed to watch for consistent production. Rarely does this advance to in situ control where instantaneous corrections are made; although a few hybrid computer controlled presses are now installed. If sensors are used, they usually measure the die temperature and strain, upper and lower punch loads (via in-line transducers), and relative displacement of the tool components. Hydraulic and computer numerical controlled presses prove easier to instrument and control. For example, in-line load cells can be used to adjust pressure on the hydraulic cylinder via a servo-hydraulic valve. This results in one form of a hydraulic computer numerically controlled press, best suited to complex, multiple level part. The result is more consistent densities. Likewise, a few presses have multiple step feed shoes. The first step weighs the powder charge to ensure the proper amount is used in filling.

Computerization of die compaction has two objectives. One objective is to record the operating data, including set-up parameters and production data. This allows initialization at the same conditions each time the tooling is installed. In one variant, the tool setup is fully automated using computerized electric motors. The second variant is to monitor trends in parts. Part-to-part variations and differences between production lots are both concerns, which require monitoring systems and logic systems. Robots might be coupled to the press for component extraction, placement on sintering trays, or inspection. These too are coupled into the logic and control system. Pick and place robots are used on large volume production items, but much of the P/M industry relies on human intervention should there be a problem.

Monitoring is largely through load measurements in the pressing direction. Logic systems then decide what changes are appropriate when drifts in product quality are detected. Most important, these systems can stop production in cases where tool damage might occur. For example, if the powder fill changes, then excessive powder in the cavity might deform a punch. A load monitor during compaction would detect such a change and halt production. Generally only three monitors are used to detect product quality drift at the press: changes in weight, changes in compaction load, or changes in dimensions. The latter parameter might be manually input to the control computer via an operator.

0 0