Cyclic thermal and mechanical loads on shot sleeves in service can lead to a variety of failures including washout under the pour hole, longitudinal and radial deformation that cause plunger tip sticking and wear, gross cracking and heat checking. However, difficulties are encountered in measuring shot sleeve temperatures and distortions in an actual diecasting machine because the sleeve operates in a thermally and mechanically hostile and cramped environment. This makes computer simulations an attractive alternative to predict distortions and temperatures in sleeves.



A current trend in the industry is the use of larger inner diameter (ID) sleeves to produce large castings. To aid in the production of diecastings and the prediction of thermal and mechanical failures in the sleeves, experimental research was performed to develop computer models to predict thermal gradients and distortions for large ID shot sleeves (7 in.) as a function of sleeve wall thickness. In the following article, this research, including actual and simulated sleeve performance, is reviewed and the results are summarized.

Experimental Data

An H13 tool steel shot sleeve with a 1.98-in. ID and 0,5-in. thickness was instrumented with thermocouple probes along the longitudinal and radial direction to collect temperature data at the molten metal--sleeve interface, A380 aluminum was poured into the sleeve and allowed to cool into a "log." This log was pushed out using a hydraulically actuated plunger tip, and the sleeve was allowed to cool for a specified period of time.

Temperature data was collected through 44 consecutive cycles during which the amounts of lubricant, type of lubricant and the fill percentage were varied. After every cycle, the outside diameter of the sleeve was measured using a micrometer in both vertical and horizontal directions. The temperature data collected then was used to calculate various thermal variables.

The time averaged heat flux (Table 1) and the time averaged heat transfer coefficient (Table 2) were calculated along the length of the sleeve at different locations for two different kinds of lubricants. These measurements were made directly below the pour hole at a distance of 5 in. (127 mm) and 13.5 in. (343 mm) from the pour hole. Measurements were made at angles of 0[degrees], 45[degrees] and 315[degrees] to the axis of symmetry along the vertical section of the sleeve.

Simulation Experiment

A 2-D plain strain finite element model (Fig. 1) was constructed based on data collected from the shot sleeve experiments. A section of the sleeve halfway between the pour hole and the platen was chosen for analysis. The model had the same dimensions as that of the sleeve used to collect data. In the model, the sleeve was assigned the material properties of H-13 tool steel.

The simulation was conducted in two stages. The first stage consisted of performing a thermal analysis on the sleeve to obtain the temperature distribution. This was performed using a 4-node linear quadrilateral element for heat transfer analysis. Based on the thermal analysis results, deformation analysis was performed using a 4-node bilinear plane strain quadrilateral element.

The simulation cycle was split into two steps, the filling phase and the empty phase. The filling period was the time in which molten metal was poured into the shot sleeve and remained there. The empty period was when there was no metal in the sleeve. The average fill period was 94 sec and the average empty time was 63 seconds,

In this model, a convection heat transfer coefficient was applied to the shot sleeve outside surface and inside non-contact surface. An interface heat transfer coefficient at the molten metal-- sleeve interface was used. The heat transfer coefficient value at the interface was critical to the model and was taken from the time-averaged values derived experimentally from the small shot sleeve. This was used as a starting point to compare temperatures between the simulated sleeve and the experimental sleeve, Tables 3 and 4 show the heat transfer coefficients at the interfaces and the corresponding temperatures for the cycle.

The simulation was performed for a section of the sleeve halfway between the platen and the pour hole and corresponded to the section of the sleeve in the experiment that was 5 in. (127mm) from the pour hole. The model had a fill percentage of 33%. The sleeve was assumed to start from room temperature [86F (30C)].

Simulation Results

The stress simulation provided the deformation data on the sleeve. In order to have a better understanding of the sleeve deformations, the maximum sleeve deformation was analyzed.

The maximum expansion of the sleeve took place at its bottom node. The results indicated that the maximum displacement of the bottom node-along the GD of the sleeve was 0.017 in. (0.43mm), which was close to the experimentally measured GD displacement of 0.019 in. (0.48mm) at

Steady state

As an MBS pool ages, or four to six months after component mortgages have passed at least once the threshold for refinancing, the prepayment speed tends to stabilize within a fairly steady range. condition. This clearly reinforces the fact that the computer model provided results that were similar to the experimental results and hence could be used to model sleeves.

Case Study: Predicting Sleeve Distortions in Production

In an effort to gauge the success of the experimental results, a case study was performed.

A large aluminum diecaster using H-13 steel shot sleeves with an inside diameter of 6.693 in. (170mm) and an outside diameter of 9.974 in. (253mm) was interested in predicting sleeve distortions when a sleeve of the same inside diameter and an outside diameter of 11.974 in. (304mm) (thicker wall) was put into service to cast aluminum automotive components. The clearance between the sleeve and the tip was 0.006 in. (0.152mm) at room temperature. The sleeve was unconstrained and allowed to freely expand on its OD.

The models developed were similar to the ones used for verification. The boundary conditions also were similar to those applied on the previous model. In order to validate the use of the correct heat transfer coefficient for large diameter sleeves, the thicker sleeve in use was instrumented with surface thermocouples on its outside diameter and the temperature readings were recorded. These readings were used for comparison with the simulation model in order to estimate the heat transfer coefficient. Based on what was learned from modeling the small shot sleeve, an initial heat transfer coefficient of 1400W/sq m x K was selected.

The process time was obtained from the process sheet provided by the diecaster. The simulation was conducted in 2 steps. The first was the dwell period when the sleeve was in contact with molten metal that lasted for 10.5 sec and the second was when the sleeve was exposed to warm air for a period of 80.5 sec.

The simulation was conducted until the sleeve reached quasi steady state condition so that comparisons could be made between the two sleeves. The 2 sleeves reached quasi steady state condition in approximately 275 cycles. The initial temperature of the shot sleeve was assumed to be 30C and the fill percentage used in the sleeve was 33%,

The simulation was conducted first for the thicker sleeve whose temperatures were compared with the experimental sleeve. This helped in estimating the correct value of the heat transfer coefficient between the sleeve and the molten metal, which was then applied to the thinner sleeve.

In order to achieve agreement between the computer model and the measured surface temperatures on the case study shot sleeve, the heat transfer coefficient between the sleeve and molten metal for the thicker sleeve had to be reduced to 275W/sq m x K. The same value of heat transfer coefficient at the interface was used to conduct simulations on the thinner sleeve, It was assumed that the heat transfer coefficients in the thin and the thick sleeve would be similar since they had the same ID, surface finish, were of the same material (H13), and consisted of the same aluminum alloy at the same initial pour temperature.