development of vacuum diecasting process

Document Sample
development of vacuum diecasting process Powered By Docstoc
					Development of vacuum diecasting process

Masashi Uchida
Technical Development Center Ube Machinery Corporation, Ltd.

1. Introduction 2. Circumstances of developing GF process and its constitution 3. Know-how to use GF process effectively 4. Effect and some disadvantages of GF process 5. Evolution of GF valve 6. Relationship between high-speed shot and vacuum diecasting 7. Ultra-high-vacuum diecasting process 8. Comparison of GF process and ultra-high-vacuum diecasting process 9. Comparison of die tooling between GF process and ultra-high-vacuum diecasting process 10. Summary


1. Introduction Vacuum diecasting process was introduced into mass production in 1983, after GF (Gas free) process had been invented. Just around the same time, OPTI-VAC process was advocated in Europe, but it did not spread so much. On the other hand, GF process was introduced to all major diecaster and automobile maker in Japan and 52,690 ton (7.7% of total yearly diecasting production) was cast with this process in 1992. After few years, GF process started introducing into European market and became popular vacuum diecasting process in the world. In these days, ultra-high-vacuum diecasting process is developed in Europe and starts to be applied for production of large and thin parts that usually should be welded for automobile chassis like space-frames and pillars. This report describes history of developing vacuum diecasting process and know-how to use it.

2. Circumstances of developing GF process and its constitution Major diecasting defects are listed as blowhole, shrinkage, cold shut, flow line and misrun. Except shrinkage, all defects are related with trapped air during filling into the die cavity. This phenomena had already been well known in 1980’s, and also well known that it was good to evacuate air in die cavity to eliminate these defects after many experiences. According to these experiences, many vacuum processes were advocated and tested, but no one bore for mass production. The reason is explained using Fig.1.
Vacuum switching valve

Vacuum pump Shut off cylinder

Cover die Ejector die Melt Al

Vacuum tank Shot cylinder Plunger rod

Sleeve Tip A B C

Shot off cylinder control valve

A: Shut off close command LS B: High speed command LS C: Vacuum start command LS

Fig.1 Example of early developed vacuum process The process of Fig.1 starts evacuation depending on the position signal C during low speed filling, -2-

change shot speed to high speed mode with the position signal then close shut off valve by the position signal A that is set near shot end to avoid melt aluminum entering into vacuum line. But pouring weight of aluminum and also movement time of the shut off valve had some error (computer technology was poor at that age and AC solenoid valve needs 8ms error to be ON in the relay circuit), it was difficult to determine A position, and it had to be adjusted if high speed setting was changed. If the setting was not correct, melt aluminum flew into vacuum line and it stopped production. It took 30 minutes to 1 hour to remove intruded aluminum. According to strong demand for more reliable shut off valve, GF process was invented in 1980. Fig.2 shows GF process schematically. Whole process is similar with Fig.1 but shut off valve is completely different. It closes passively using inertia force of metal flow without being affected by porting weight error and shot conditions. Here the new vacuum process was established showing good reliability and applicability to mass production.
GF open cylinder

Open lock ball Main valve

GF shut spring

Vacuum switch valve

Valve shut

Finish filling

GF valve

Vacuum pump

Cover die Ejector die Melt Al

Vacuum tank Shot cylinder Plunger rod

Sleeve Tip B C

B: High speed command LS C: Vacuum start command LS

Fig.2 Schematic drawing of GF process -3-

3. Know-how to use GF process effectively In this chapter, know-how of effective usage of GF process that has been established through a lot of casting trials is described. (1) Vacuum time should be within 0.3s (maximum 0.5s). If vacuum time (from vacuum start command C to high speed command B) is more than 0.5s, blowhole starts to increase. The reasons are follows; Die spraying must be applied to die surface to prevent adhesion of aluminum. Almost all water base solvent is vaporized by heat of the die and solute makes coating film at the die surface. But very few amount of liquid may intrude into gap of dies and cores, and those liquid will remain without drying especially in the case of die tooling employs a lot of slide cores. If die cavity is evacuated for more than 0.5s in such a condition, remaining liquid is sucked into the cavity surface and vaporized rapidly with touching to the melt. These gas can be easily entrapped into the casting and be blowholes. Spray liquid is diluted over 40 times with water and vaporized water inflates 1240 times in volume. Even if very small amount of spray liquid, it must be paid attention about inflation. According to more than several hundreds experiences about this item, vacuum within 0.3s is the right way to prevent above mentioned problem. (2) Good vacuum system and well designed die tooling need no sealing at the parting line of die or slide cores. 20-27kPa vacuum could be obtained optimum conditions without sealing and it is good enough to get good casting quality.

Fig.3 Partial shots showing wrong vent runner layout




Fig.4 Modification of runner layout -4-

(3) Die tooling for ventilation should be designed to have enough evacuating ability as 20-27kPa could be established within 0.3s.
Calculate air volume to be evacuated Assume inspired air volume Calculate capacity of required evacuation Calculate vent runner area

(4) Vent runner should be adopted at the final filling point determined by metal flow analysis. One fifth of air is remaining after evacuate to 20kPa in the die cavity, so remaining air should be pushed out by metal flow itself. Fig.3 is an example that expected final filling point was wrong from actual one after casting trial. Last remaining air could not be pushed out through air vent because it was already entrapped by melt. Modifying vent runner from side to bottom and blowhole could be remarkably reduced. This product is tray of grill and original die layout was like Fig.4(a). This product is polished inside of tray portion after deburring and then applied PTFE coating with burning treatment. Blistering may occur at this treatment if blowhole exists near surface of the casting. Fig.4(b) is first die layout for GF process with wrong vent runner as described above. Fig.4(c) is modified tooling and 0.5% rejection rate can be obtained with this layout against 30% for the first one. This is just one successful experience, and we know hundreds cases same as this example. Trial for this product was made 19 years ago with procedures as checking metal flow by partial shots (stop the shot during high speed filling at desired stroke then solidified for a while, stopping stroke may be changed step by step), to confirm final filling point and then correct position of over flow and vacuum vent runner. Final filling point is easily presumed with CAE in these days, so optimum die tooling can be obtained without any casting trial. Fig.5 is an example of flow analysis result showing final filling point Fig.5 Flow analysis result (ADSTEFAN)

4. Effect and some disadvantages of GF process Effects of GF process was regognized as eliminating defects like blowhole, cold shut, flow line and misrun during casting trials. But also unexpected effect appears in case of almost all casting. That is, good parts can be obtained with lower shot speed and/or lower metal pressure. For instance, shot speed can be reduced from 2.5m/s to 1.5m/s, or metal pressure can be reduced from 80MPa to 50MPa without sacrificing casting quality. This is because excess power was needed to cancel backpressure of remaining gas in the cavity without GF process. At casting site, reduction of shot speed and metal pressure prevents burrs remarkably, and also it prolongs die life, eliminates deburring operation and increases up-time of casting machine. In starting up new parts, good parts can be easily obtained with GF process so modification process of die tooling can be reduced, and totally new product can be introduced in mass production in -5-

short time. All these words are come from GF users. But there is some disadvantage. One of them is intruding aluminum into vacuum line when aluminum dreg that stuck at valve sheet at last shot interferes shutting valve. The story of trouble shooting of this problem is describing in next chapter.

5. Evolution of GF valve Fig,6 shows mechanical type GF valve. Main valve
GF open cylinder Open lock ball Main valve

is locked by ‘open lock ball’ that holds valve in open position against shut force of ‘shot spring’ and vacuum sucktion until melt front hits the valve. Open lock ball must be adjusted to loose against shut off force just after metal front hitting. If locking force is too weak, main valve shuts with vacuum force before metal comes to the valve portion remaining large amount of air in the cavity. On the other hands, valve does not shut at proper moment if locking force is too strong and aluminum may intrude into the vacuum line. So it is very

GF shut spring

Fig.6 Mechanical type GF valve

important , and also defficult, to adjust locking force of

‘open lock valve’. It must be adjusted once a day with exclusive jig that mesasures locking force. Early GF valve adopts mechanical lock system as shown in Fig.6, so locking force reduces as locking groove of the main valve is worn. There are screws behind open lock balls to adjust locking force but high skill is needed adjusting two spring forces even. If this adjustment is not enough, shut off timing may be inaccurate and insufficient evacuation (shut off faster with too small locking force by groove wear) or aluminum intruding into the vacuum line and jam (shut off slower with too much locking force) may occur. Needs to adjust is one of issues of mechanical GF valve but it is not immediate reason of aluminum intruding. Countermeasures at the casting site, replace valve set to new one if there is a sign of jam like error of shutting the vavle detected by electric signal. Valve unit is formed as cassete unit to be replaced easily within 10 minutes. Some of the customers replaces valve unit every 4 hours even if there is no trouble. Because of large benefit in casting quality, many customers had been using GF valve with these know-how against difficulties. Recently the reason of valve jam are going to be clear. According to the investigation of collision force applied to the main valve, sometimes it comes intermittently against supposition. This phenomena is so called “flying melt” as indicated in Fig.6 and thought to be the reason of valve jam. If the vent runner is streight and short, it is easily occur. Be the vent runner long and add more than two of 90 degrees turn are the very effective coutermeasures. At the same time, structure of GF valve itself is going to evolve as shown in Fig.7. Fig.7(a) shows the moment just before GF valve opens. As introducing air pressure into Room-A, the main valve is pushed downward and the piston meets with o-ring. Under this condition, bring air pressure into Room-B continuously and exhaust air in Room-A, air pressure force in Room-B pushs the piston -6-

downdard and keep the main valve open shown in Fig.7(b). Vacuum evacuation is done under this state. Air Exhaust Room-A Rod-B Air Piston Room-B Exhaust O-ring Exhaust Room-C Rod-A Main valve Vacuum Exhaust Exhaust Exhaust Air Exhaust Air

(a) Just before valve open

(b) During evacuation

(c) Shut off valve

(d) Shot end

Fig.7 Air drive type GF valve Shot velocity shall be changed to high speed mode at the moment that vacuum pressure reaches to the set value. Fig.7(c) shows the moment that main valve shuts off. If front of the melt jumps to the main valve and push the main valve up a little, o-ring seal is break and air comes through the gap. Then air pressure starts to act not only upper side but also bottom side of the piston, and upward force will conquer because Rod-A is thinner than Rod-B and bottom side area of the piston is larger than upper side. The main valve goes up and shuts vacuum line, and keep its position. Characteristics of this air drive type GF valve is easiness of adjusting locking force by just only controling air pressure introduced into Room-B. As a result of this, the locking force can be adjusted during continuous casting operation without using special jig by automatic controller. And there is no mechanical lock, so wear is negligible for long time run. The other benefit of air drive type GF valve is sensitivity for starting shut off movement. It is needed for mechanical type valve to be be lifted up for 0.6mm by melt to release locking force. On the other hands, just only 0.2mm lift is needed for air drive type GF valve with one third aluminum inertia force energy of mechanical type. This type valve recorded 92,000 shots run without a jam. And many user add thier own know-how into die tooling for stable production. Details of die tooling may not be described in this paper for paying respects to engineers who made a lot of effort to establish their own know-how. The GF valve works without jam with proper die tooling and spreads all over Japan as if it is a standard to use GF valve from the start of designing new die. As the inventor of GF valve, UBE is continuously studying for more strong defence against jam, for example, change sequence to get more sensitive reaction against melt collision by reducing air pressure in Room-B and open-holding power then add pressure in Room-A in starting vacuum evacuation not to be aspirated by air flow.

6. Relationship between high-speed shot and vacuum diecasting Effect of vacuum process was described in former chapters. But sometimes a claim like “There is no need of vacuum process for ultra-high-speed diecasting.” comes to be announced. It must be clear

whether it is true or not in this chapter. Fig.8 shows density of plate type castings made with normal diecasting and GF ultra-high-vacuum process as changing shot velocity. Fig.9 also shows gas content of both process as Fig.8 using same die. Dimensions of the plate castings are 200mm width, 270mm height and 4mm thick. Gate thickness is 2mm, metal pressure is 50MPa, and alloy is ADC12(equivalent to AA383). According to these figures, the faster shot velocity shows more gas contents in normal diecasting process.
Cavity:200x270x4, Gate:2, Alloy:ADC12

2.76 2.75 2.74 2.73 2.72 0


GF ultra-vacuum diecast Normal diecast



3 4 Shot velocity(m/s)



Fig.8 Density of plate type castings at various shot velocity

25 Gas content(cc/100gAl) 20 15 10 5 0 0

Cavity:200x270x4, Gate:2, Alloy:ADC12

GF ultra-vacuum diecast Normal diecast







Shot velocity(m/s)

Fig.9 Gas volume involves in plate type castings at various shot velocity

In some cases blowhole defects seem to disappear with increasing shot velocity over 4m/s especially on the machined surface. But entrapped gas still remains, never disappears. It is just scattered into small pieces and hardly can be seen with naked eyes. These very small blowholes also cause leakage problem if they are linked together and/or blistering problem as PTFE coating. The comment “There is no need of vacuum process for ultra-high-speed diecasting.” can be said only in case of discussing about defects on machined surface inspected by visual inspection. It is believed that vacuum process is necessary to get high quality castings combined with ultra-high-speed process. Data of vacuum process in Fig.8 and Fig.9 are made with GF ultra-vacuum process that will be described in next chapter, but usual GF process also can get such a good data if desirable conditions,

especially die tooling, are achieved, and the casting made under such conditions involves gas content lower than 4cc/100gAl that is within the limit of castings for welding. The reason why ultra-high-vacuum process is becoming a subject recently will be described in next chapter with comparing to normal GF process.
7. Ultra-high-vacuum diecasting process Ultra-high-vacuum diecasting process was born in Europe around 1995 to produce large and thin automobile parts like pillars and space frames. These parts need to be welded so gas containing must be lower than 4cc/100gAl. Ultra-vacuum as 5-10kPa is generally achieved by sealing between dies and cores (sometimes at ejector pin hole) with o-ring and with piston ring around plunger tip. By the way, normal GF process achieves 20-27kPa vacuum without any sealing to simplify die tooling.


Main valve


Valve shut Vacuum switch valve

Finish filling

GF valve O-ring


Vacuum pump

Cover die Vacuum tank Shot cylinder Melt Al Plunger rod

Ejector die

Sleeve O-ring Piston ring Tip B C

B: High speed command LS C: Vacuum command LS

Fig.10 Ultra-high-vacuum 20kPa but one fifth of GF valve Normal GF process evacuates die cavity to aboutdiecasting system with atmosphere still remains. So -9-

remnant gas must be pushed out to air venting slot by metal flow with adjusting shot conditions. To do so, die tooling may have long flow length that is not reasonable for thin and big castings because long metal flow reduces metal temperature so much as to cause defects like misrun or cold shut. By increasing vacuum up to 5kPa, remnant air is reduced one twentieth of atmospheric pressure, so flow route can be minimized without strict consideration to exhaust air from venting slot by metal flow. Schematic diagram of ultra-vacuum system with GF valve is shown in Fig.10. The system is almost same as normal GF system except sealing between dies and cores and piston ring around plunger tip. Through casting trails using this system, it comes to be clear that time for evacuation is important. It is better switching from slow speed to high speed with in 0.5s to prevent “flying metal” that overhang on the distributor and inspired by vacuum flow into the cavity, but it must be changed after cavity pressure reaches target value. There is only UBE’s GF ultra-vacuum process that can be evacuate to 5kPa within 0.5s (UBE’s process can do that within 0.32s). Flow capacity of vacuum switching valve must be doubled and also same as vacuum tank capacity compared with normal GF system, most important change is location of the vacuum switching valve. It used to be located on tank side with the vacuum pump as one unit to be portable easily, and connected with GF valve by long hose. To explain with Fig.10, former system had 5-7m hose at part A and 0.5m at part B. So the system must evacuate two to three times much air in the hose than cavity volume at the same time. It is not problem in case of normal vacuum as 20kPa because it can be reached within 0.3s by normal GF system with former system. But it is difficult to do so in case of ultra-high-vacuum as 5kPa with this layout. To overcome this problem, the vacuum switching valve was moved to top of the ejector side platen. To explain with Fig.10, new system has 1m hose at part A and 5-7m at part B, and the hose at part B is double thick and continuously evacuated so as to be a part of vacuum tank. Combination of large exhaust volume of GF valve and this unit makes fast vacuum as 5kPa within 0.32s.

8. Comparison of GF process and ultra-vacuum diecasting process Ultra-high-vacuum UBE GF process 5-10kPa 0.32s O-ring --------Piston ring 2-10cc/100gAl General diecast Piston ring Special vacuum sealing 2-4cc/100gAl Pillar, Sub-flame, Node, Bracket, Upper-arm 1.0-1.5s O-ring Others

Item Vacuum pressure Time to be evacuated Die parting Core Seals Ejector pin Plunger tip Gas content in products Usage

GF process 20-27kPa 0.3s

9. Comparison of die tooling between GF process and ultra-high-vacuum diecasting process -10-

Metal flow length LA Gate GF valve Vacuum vent runner Metal flow length LA GF valve Vacuum vent runner

Product Runner Biscuit

Product Gate Runner Biscuit

(a) GF normal vacuum

(b) GF ultra-vacuum

Fig.11 Die tooling for normal and ultra-vacuum diecasting system with GF valve As described in chapter 7, die tooling for normal and ultra-high-vacuum GF diecasting are different. Schematic drawings shown in Fig.11 explain basic concept of both processes. The keys for both processes are directional filling of normal vacuum diecast and shorten flow length of ultra-high-vacuum, respectively. Trend nowadays is ultra-vacuum is best process to get weldable products. But normal GF process can get such weldable quality with proper conditions as shown in Fig.9. Most reasonable selection is that apply ultra-high-vacuum process for thinner less than 4mm and larger as flow length over 700mm.

10. Summary GF process made public in 1983 and this was the start of, for vacuum processes, introduction into mass production field. After GF process, several vacuum processes were advocated. In 1995, ultra-high-vacuum process first appeared and several ultra-vacuum processes followed in this decade. But application of these ultra-vacuum processes is limited in automobile body parts. All products are very large and need huge size diecasting machine like 2500-4000 ton of die clamping force, so these processes did not become popular as limited among car makers and large diecasters. General diecast parts do not need ultra-vacuum process. Normal vacuum process is good enough to eliminate defects except shrinkage for them. It is reported that combination of ultra-high-vacuum and ultra-high-speed shot process can produce high quality castings having 8-15% elongation with lower metal pressure. This indicates high quality castings may be produced with the casting machine now in use with some modification. Technologies needed to achieve this statement are almost already developed as ultra-vacuum process and multi-metal pressure (of course including low pressure) casting system. Remaining theme to be developed by machine maker is the machine works stably without burr even if shot velocity is high as 6-7m/s actual. Recent automobile parts, especially safety suspension parts that needs high strength or rigidity, are converted from ductile cast iron to heat-treated gravity aluminum casting, and then to heat-treated aluminum diecasting, so cast wall is very thick as 20-30mm. To get lighter parts, material conversion from cast iron to aluminum acted as main role, but it may change to conversion from thick aluminum casting to -11-

lighter rib-structured thin aluminum casting. This tendency seems to appear in Europe since more than 10 years before. As described above, diecasting machine and casting process are now in progress to seek new technologies into the future, and most important one among them is vacuum casting process.


Shared By:
Description: development of vacuum diecasting process