Jürgen Wüst 2008 年 10 月

Transcription

1

2 Injection Shot Profile Monitoring Position based vs Time based acquisition Improve accuracy and avoid missing data Paolo Catterina

3 Injection Process Monitoring The analysis of injection (shot) profiles is a traditional instrument, effective and consolidated by several decades of knowledge and monitoring of the die-casting process. The signals derived from the movement of the injection plunger: position, speed and pressure are recorded, studied and accurately analyzed.

4 Injection Process Monitoring Almost all systems on all DCM equipments are based on the acquisition and representation of the motion of the plunger on a time basis.

5 Injection Monitoring: at the Heart of Process The signals derived from the movement of the injection plunger: position, speed and pressure are recorded, studied and accurately analyzed and can be considered, in a fitting metaphor, a sort of "electrocardiographic examination" of the heart of die casting machine: the injection process.

6 Injection Process Monitoring The kinematics of the plunger action while pushing the metal into the die cavity can be effectively considered the heart of the whole production process, taking upon itself many of the critical issues on the final quality of the casting. As it is universally known and proven there is a close correlation between the injection process parameters and the quality factors or product failure pieces.

7 Injection Process Monitoring Monitoring of injection profiles and constant evaluation of calculated parameters (speed of stages, pressures, positions, times) can be considered as the most widespread and applied technique for the quality survey in the die-casting process. The automatic selection between good parts and pieces to be rejected as well as the process of statistical analysis on the die-cast parts are based almost entirely on data derived from the curves of the injection control systems. The knowledge and evaluation of the injection diagrams are now an essential tool for almost all the foundry engineers both for evaluating the ordinary, serial process, and for the diagnostic of the machine.

8 Timed based signal acquisition Almost all of the systems are based on the acquisition and representation of the motion of the plunger on a time basis: Signals are sampled from the beginning of the movement along a defined duration until the end of the cavity filling phase and of the pressure intensification. The sampling strategy is based on reading the signals at a predefined frequency, roughly between 0.5 and 3.0 khz (corresponding to 500 to 3000 values per second). The traditional diagram represents velocity, pressure and position on the same horizontal time axis.

9 Timed based signal acquisition Traditional time acquisition derives historically from the use of oscilloscopes to monitor the signals as the very first instruments for shot monitoring. Sensors installed on the DCM gave electrical output and the oscilloscope was the only common instrument capable of representing electrical signals over time. Improvement in monitoring systems has led to the optimization of devices resulting from oscilloscope evolution so that the sample rate to obtain data is usually constant time base.

10 Position based signal acquisition Only a few systems are based on the representation, moreover on an acquisition strategy, relying on the position of the plunger. Sampling signals occurs at variation of the position, and therefore profile diagram represents trace of speed and pressure on the position horizontal axis.

11 Position based signal acquisition Sampling signals at position change rate highly improves getting more values when the plunger moves at fast speed while sampling signals at a fixed frequency based on time risks sometimes to lose data during the critical stage of high speed motion. In other words and with an easy motto: the faster the plunger moves, more values are acquired! And, conversely, on a time basis, systems risk to miss information or run into calculation inaccuracy due to poor quantity of data acquired

12 Position based signal acquisition On a Position based system the profile diagram represents trace of speed and pressure on the position horizontal axis. At least until the end of the filling of the cavity (or of the impact of plunger) and then switches to a time base during the final stage of pressure intensification.

13 Position based signal acquisition This multiple Data Collection Method provides the most complete representation of injection process: A.Tracking High Speed Movement (Filling Phase) Position-based sampling - Collect data from the sensor to track position during slow shot and cavity fill short time and large distance covered B.Tracking Low Speed Movement (Intensification Phase) Time-based sampling - Dynamically switch to time based sampling during intensification and squeeze small distanced covered over long time period.

14 Sampling technique & sensors At the current state of the art there is no more accurate (and absolutely cheap) way to compute real time shot velocity A board provides a precision clock period count between digital pulses created by the sensor applied on piston movement with equal incremental position increments (typically of.0125 inches). Mounting Block Transducer.050 typical Digital Quadrature Output Chrome Coated Grooved Rod Detecting magnetic variation caused by the rod s grooved internal structure: up to 12mps at mm resolution

15 Sampling technique & sensors The counter starts from of the rising Edge of the Digital Quadrature A pulse; then the rising edge of the B pulse which lags the A pulse by 90 electrical degrees. Then the falling edge of the A pulse and then the falling edge of the B pulse. As a result 4 updates per pitch can be obtained. This is referred to as X 4 count logic. The distance being known it is easy to determine velocity by tracking the time interval between each pulse transition. The faster the rod moves the faster the system gets updated the acquisition buffer.

16 Time based vs Position based Acquisition Both shot profile measurements, time based and position based put into practice the physical law on the kinematics of the linear plunger movement v = ds / dt but in the discretization of space and time value enforces the validity of the latter one compared to the other. In the following pictures we try to highlight this concept.

17 Time based vs Position based Acquisition

18 Time based vs Position based Acquisition

19 Comparing performances of two methods Case 1 A very short Stroke acquired on special hot chamber DCM. An extreme case study Slow velocity 0.10 m/sec Plunger displacement 4 mm Movement time sec Fast velocity 0.40 m/sec Plunger displacement 11 mm Movement time sec Time sampling at 1 KHz Total duration sec Slow Stage samples 40 Fast Stage samples 27 Total samples 67 Position sampling Movement Resolution 0.1 mm Slow Stage samples 40 Fast Stage samples 110 Total samples 150

20 Comparing performances of two methods Case 1

21 Comparing performances of two methods Case 2 A short Fast Speed Stroke, common profile on thin walled parts. Slow velocity 0.10 m/sec Plunger displacement 400 mm Movement time sec Fast velocity 3.00 m/sec Plunger displacement 50 mm Movement time sec Time sampling at 1 KHz Total duration sec Slow Stage samples 4,000 Fast Stage samples 16 Total samples 4,016 Position sampling Movement Resolution 0.1 mm Slow Stage samples 4,000 Fast Stage samples 500 Total samples 4,500

22 Comparing performances of two methods Case 2 A short Fast Speed Stroke, common profile on thin walled parts.

23 Comparing performances of two methods Case 3 A common Shot. Slow velocity 0.12 m/sec Plunger displacement 450 mm Movement time sec Fast velocity 2.50 m/sec Plunger displacement 150 mm Movement time sec Time sampling at 1 KHz Total duration sec Slow Stage samples 3,750 Fast Stage samples 60 Total samples 3,810 Position sampling Movement Resolution 0.1 mm Slow Stage samples 4,500 Fast Stage samples 1,500 Total samples 6,000

24 Comparing performances of two methods Case 3 A common Shot.

25 Comparing performances of two methods Case 4 High Velocity Shot. Example from a structural part on a 2000ton DCM Slow velocity 0.15 m/sec Plunger displacement 600 mm Movement time sec Fast velocity 5.50 m/sec Plunger displacement 250 mm Movement time sec Time sampling at 1 KHz Total duration sec Slow Stage samples 4,000 Fast Stage samples 45 Total samples 4,045 Position sampling Movement Resolution 0.1 mm Slow Stage samples 6,000 Fast Stage samples 2,500 Total samples 8,000

26 Comparing performances of two methods Case 4 High Velocity Shot. Example from a structural part on a 2000ton DCM

27 Possible loss of information and inaccuracy Essential reason of the potential loss of accuracy in the measurement for Time base Acquisition derives from the limited amount of values acquired during the fast movement. In example 1. It is quite clear that most of the velocity variations during fast velocity stage cannot be measured nor represented due to a poor resolution of the samples collected during signal acquisition

28 Possible loss of information and inaccuracy A weakness in the process control system based on time sampling could arise hide process (and part quality, of course) issues: in that case a non-optimal functioning of the DCM because of unstable fast velocity. The anomaly could have important effects on the quality of castings produced. In the below picture an excerpt from a statistical analysis over about 3,000 shots measured with a position based system presents a variation of 0.20 m/sec on a range of 3.58 to 3-78 which

29 Comparing and overlaying profiles The reference position is a most definitive stable than the time base. As an example it is sufficient to think of an injection with non-constant motion during the first part injection and this will originate the phase displacement of the entire profile along the time axis. This makes it difficult to achieve effective overlap and comparability which is, however, a useful visual and intuitive tool to represent the process and its variations.

30 Conclusione Systems working by sampling based on the position offer the most accurate results for the measurement of velocity in almost all the circumstances. The graphical representation of the injection process through the conjunction of the diagram position based until plunger impact and the subsequent switch to time base during the intensification stage appears more comprehensive and intuitive. The most significant aspects of the process and the more critical - the speed during cavity filling and the pressure intensification with consequent compression of the metal - which otherwise, in a representation over a time scale, result simplified and displayed almost in the background.

31 Thanks for your attention and for the patience that you have granted.

32 References F. Schmidt, M. Müller, U. Vroomen, A. Bührig-Polaczek, Process parameter influence on selected quality features in high-pressure die casting, in La Metallurgia Italiana, June E.Vinarcik, High Integrity Die Casting,, Addison Wiley, 2003 H. Bakemeyer, Operating the die casting machine, NADCA, J. Wilkinson, J. A. Scott, G. E. Wilson, A. Connor, Statistical Process Control in Pressure Die Casting, A.R. Adamane, L. Arnberg, E. Fiorese, G. Timelli, F.Bonollo, Influence of Injection Parameters on the Porority and Tensile properties of High-Pressure Die Cast Al-Si Alloys: a Review, Int. Journal of Metalcasting, J.S.Kirkman, Guide to Process Monitoring Control, NADCA, J.Vann, B.McClintic, Interpreting Shot Profiles, Troubleshooting & Problem Solving for increased margins and happier customers,