Вы находитесь на странице: 1из 23

ChapterTwo

SURFACE DEFECTS

A. DEFINITIONS

Surface defects are considered here to be those defects that are mostly visual, which cause a rejection for some kind of surface blemish. Although there are some exceptions, this will be the largest single defect category for most custom casting operations where there is a variety of castings. The defects in this category have many names, and there are often differ- ent names used in different geographical areas to describe the same problem. Some of the typical names for this type of defect are as follows:

cold flow

non-fill or not filled out (NFO)

cold laps

swirls

cold fill

chill

cold

laps

poor fill

lines

Probably the first name is the most common, and it is also the most de- scriptive, so it is the one that will be used in this book. The type of defect this term describes (as used in this book) is shown in Fig. 2.1, 2.2 and 2.3. This defect is usually referred to as a metal flow defect because it is characterized by irregularities on the casting surface where the various metal flows apparently did not knit together properly. This kind of defect occurs because it is always a race between the time the molten metal arrives at a location in the die and the rapid solidification taking place. If the metal is partially solidified when two flows come to- gether, they form wrinkles or laps and laminations that are characteristic of surface defects. This defect is often apparent at the end of the flow pattern and/or where the die is colder, such as the ends of ribs and bosses. Many times the technician will try to solve this defect with one stan- dard correction for all castings; for example, the most typical reaction is to always make some change to the gate design, even though the gating may

13

14

NADCA DIE CASTING DEFECTS

Fig. 2.1, 2.2 and 2.3. Typical examples of cold flow surface defects.

2.2 and 2.3. Typical examples of cold flow surface defects. (Fig 2.1) (Fig 2.3) (Fig 2.2)

(Fig 2.1)

Typical examples of cold flow surface defects. (Fig 2.1) (Fig 2.3) (Fig 2.2) not have anything

(Fig 2.3)

examples of cold flow surface defects. (Fig 2.1) (Fig 2.3) (Fig 2.2) not have anything to

(Fig 2.2)

not have anything to do with the problem. Certainly the gating needs to be correct to avoid these defects, but other process adjustments can be the major cause and are often much easier to modify. The idea that this defect is always a gating problem comes from the idea that this is a metal flow defect, even though other process parameters (fill time, for example) can be much more important than the metal flow pattern. Thus the first step is a careful review of the process parameters. It must be kept in mind that the problem may be an interactive one; and may be caused by the interactions of several process factors. To quickly focus on one issue (such as the gating or die temperature) will usually severely limit the potential for a robust and easy correction. A list of the main factors involved in these types of defects is shown below, and these descriptions show that there are indeed quite a number of interactions possible. These descriptions are general, there are quite a num- ber of other variables for each of these factors. The most important of these are listed below:

Wall Thickness. The average wall thickness is used for most castings (some

Surface Defects

15

use the thinnest wall section that is critical for quality issues). However the wall thickness is qualified, it is a critical factor in surface defects.

Casting Shape. The geometry of the part; mostly the flow distance, the number of reflections before the end of flow, and whether the flow can directly reach critical areas.

Fill Time (F t ). The length of time it takes to completely fill the casting with molten metal.

Flow Pattern. This is the flow pattern of the metal as determined by the gate design, and also by how many obstructions are in the flow path.

Die Temperature (T die ). The temperature of the die surface when the metal flows over it.

Metal Temperature (T metal , or T m ). The temperature of the metal as it enters the die.

Gate Velocity (G v ). The velocity of the metal as it goes through the gate.

Metallurgy. The effect of the alloy constituents on the casting characteristics.

Venting. The efficiency of the die in releasing trapped gasses. This con- cerns porosity too, but it also has an effect on metal flow from the back pressure of the trapped gas.

While there are interactions, the effect of each of these will be considered in sequence. The above list is arranged in priority order, which means that the top items on the list often will have the most effect, however, this is not an absolute situation. For example, if the wall thickness is fairly thick (for example 0.125 in. for aluminum), then the fill time is much less important, but if the wall thickness is a minimum (0.06 in. for example), then the fill time is by far the most important variable. Also, the quality characteristics vary from casting to casting. For ex- ample, if the surface quality is critical (hardware or plated finish), then the wall thickness, the fill time and the die temperature are by far the most im- portant issues. If leakers were the important quality issue, then temperatures and pressures are important. All of this makes defect elimination much more complex, and eliminates the simple solution (like changing the gate).

16

NADCA DIE CASTING DEFECTS

B. WALL THICKNESS

Since this is controlled by the part design as set by customer, and not con- trolled on the floor, it is often eliminated from the list of potential solu- tions. This should not be the case because it is so important. If an effort is made to correct wall thickness problems it will often prove to be the most robust and lowest cost solution. The wall thickness becomes very critical when it is below about 0.090 in. (2.28 mm) in aluminum and magnesium, and below about 0.075 in. (1.9 mm) in zinc. In these thin wall parts, it is worth checking the wall thick- ness because it makes such a large difference in the flow. It will be desir- able to move the wall thickness to the maximum allowed. These numbers are not absolute, and are interactive. The wall thick- ness, the die temperature, and the fill time are interactive as shown by the approximate maximum fill time calculation as taught in the NADCA gat- ing class. This calculation will be presented later under fill time. The minimum wall thickness consideration will depend on this calcu- lation, but it will be about 0.06 in. (1.5 mm) for aluminum and magnesium, and 0.04 in. (1 mm) for zinc. These numbers are for very short fill times, short flow distances, and properly heated dies. Many times operating con- ditions will not permit walls this thin, but they can be achieved with good engineering and part shape conditions; in fact, even thinner walls can be achieved by good engineering and good machines.

A plunger speed variation that causes a fill time variation of 30% may

not be significant with a wall thickness of 0.125 in. (3.17 mm) or greater,

but a plunger speed variation of 10% can be extremely significant if the wall thickness is 0.080 in. (2 mm) or less. The troubleshooter needs to be aware of these interactive factors. The wall thickness must be uniform if surface defects are to be mini-

mized. The metal flow is always disrupted by thick and thin wall sections, and, as every design handbook will say, one critical factor in die casting is to have an even wall thickness throughout.

If the walls are thin over the whole part, it is important that it be consis-

tent at the highest value allowed. This usually means that the toolmaking tolerance is reduced. Thus a print tolerance for wall thickness of 0.07 to 0.09

in. (1.77 to 1.9 mm) should be passed on to the toolmaker as a tolerance of 0.085 to 0.090 in. (2.15 to 2.28 mm). Keeping the wall thickness at the high end of the tolerance band is important, but of equal importance is keeping the thickness variation at a minimum. Avoid solving a one time toolmaking problem that allows a long term manufacturing problem in the process. When a thin wall section has critical surface appearance requirements, it

Surface Defects

17

is best if the gate is placed as close to the critical wall section as possible. On the other hand, if the critical quality issue is porosity in the adjacent heavy section, then that section should get primary consideration for gate location. To summarize some issues on wall thickness:

• Talk to designers early

• Work to get consistent wall thickness

• Check the actual wall thickness if there are problems

• Use the wall thickness dimensional tolerance for the process, not for toolmaking

• Feed critical thin walls directly from the gate

• For thin walls, expect a much narrower process window

– Use very short fill times

– Use high die temperatures

– Use high gate velocities (but not high enough to cause erosion)

C. PART SHAPE

Part shape is a very important factor when troubleshooting surface defects and often can be the most important; but unfortunately, it is the most diffi- cult to change. Important factors in the shape are:

• Flow distance (distance as the metal has to travel from the gate to the furthest point to fill)

• Complexity of the metal flow path (how many reflections are required for the metal to reach it’s final destination)

• Blind fill areas (cores, fins, etc.)

• Shaded areas (areas that are directly behind an object that divides the metal flow)

• Draft and radii allowed

• Allowable gate locations

• The shape also causes hot or cold spots in the die, which in turn affects surface defects

18

NADCA DIE CASTING DEFECTS

Part shape makes each part unique, and it is the biggest variable in the whole process. Though most of these factors are not under the control of the person who solves the defect problem, every effort should be made by the process engineer or technician to emphasize the importance of these factors to those who can change them. An example is shown in Fig. 2.4, which shows a surface defect that only appeared after a large radius was changed to a sharp radius. The next four factors are those that can be most easily controlled on the floor, and these are what concern most troubleshooters. These are: fill time, flow patterns, die temperature, and metal temperature. We will focus on these four things as the most important factors for correcting surface defects because they are the most readily adjustable for an existing die; this will be the most important concern for most readers.

D. FILL TIME

Fill time is one of the most important factors in surface finish control, and quite often the most important by far of those items that can be controlled on the floor. The fill time is defined as the time beginning when the metal arrives at the gate and ending when the cavity is full (if they are small compared to the casting volume, the overflows can be included). A good rule is that the faster the fill time, the better the surface finish.

Fig. 2.4. The arrows point to corners that were made sharp by the customer after the part had been run without problems for several years. The sharp radii caused the surface defects shown here to appear.

the part had been run without problems for several years. The sharp radii caused the surface

Surface Defects

19

It should be observed that no surface defect problems arise from a very short fill time. (Note: unless the gate area is changed, changing fill time will change gate velocity at the same time, and excessive gate velocities can cause problems; but a quick fill time with the appropriate gate velocity and proper gate design will never by itself contribute to surface defects). The actual fill time must be measured with an injection system moni- toring device, and will range from less than 5 msec for smaller zinc and magnesium parts to more than 200 msec for a large transmission housing in the 40 to 50 lb. (88 to 110 kg) range. The typical range for a 600 ton machine part (3 to 7 lb., or 6.6 kg to 15.4 kg of aluminum) would be in the range of 40 to 100 msec. The important relationships between fill time, die temperature, metal temperature, and wall thickness can be approximated with the formula re- ferred to earlier, and is the one used in the NADCA gating course. This is:

Max fill time = Kx (T i (

T

f

T

f

+

d

)

SZ )

T

Where:

xT

K

= a constant

T

= average casting wall thickness

T i = metal injection temperature T f = metal flow temp (solidus) T d = die temperature

S

= percent solids at cavity full

Z

= conversion for latent heat

This formula is best used to define a maximum fill time when the formula is used for process calculations. It is not exact, but provides a very good start point for a maximum fill time estimate (and consequently plunger speed requirements). It should be modified by any previous experience with this part or this type of part. Note also that since the die temperatures and the metal temperatures are included in the calculations, the formula takes these factors into ac- count. The metal temperature is the temperature at the gate, which is typi- cally 70°F (21°C) to 100°F (55°C) lower than the furnace temperature for cold chamber machines, and about 40°F (22°C) lower than the pot tem- perature for hot chamber. The die temperature is the surface temperature of the die when the metal is injected, which can be measured best with a surface pyrometer.

20

NADCA DIE CASTING DEFECTS

The fill time should be considered one of the primary tools for the troubleshooter when reducing surface defects. A maximum fill time should be determined using the formula (or from other sources), with anything less than the maximum being desirable. The best approach to finding the estimated maximum fill time is to use the formula as a starting point, and modify the results if there is solid experience to support a change. For some rough guidelines in estimating the fill time, the following are some maximum fill times based on calculations and experience and will be reasonable for most castings:

Table 2.1. Approximate fill times for average surface finish.

 

thin wall

average wall

 

<.09 in (2.3mm)

>.09 inches(2.3mm)

Al, approx

…5lb (11kg)

.09 sec

.1 sec

Zinc, approx…3lb (6.6kg)

.03 sec

.05 sec

Mg, approx.….2lb (4.4kg)

.02 sec

.03 sec

Note: these numbers are approximate, and should be used for an aver- age functional casting. For high quality surface finish, these numbers should be reduced by as much as 50%. Note also that the fill time may not be the dominant factor for other defects; other factors may be more important if surface defects are not being considered.

E. CONTROLLING FILL TIME

The following list shows some of the things that can affect the plunger speed, which in turn changes the fill time and the casting surface finish. Some of these are:

Changing the plunger speed control. Faster is lower fill time, and better finish.

A gate size change. A smaller gate will generally cause the plunger to slow down because of the extra resistance at the gate (if no other changes are made). This increases the fill time and causes a worse finish. It also af- fects gate velocity, but the main effect on surface finish is on the fill time.

Surface Defects

21

A larger gate will generally reduce the fill time and give better finish, but

the PQ 2 calculations must be run to know the results before making the

change.

Changing the shot system hydraulic pressure (if no other changes are made). Increasing pressure increases plunger speed and reduces fill time (it may also contribute to flashing). Lowering pressure reduces plunger speed and increases fill time.

(Note: PQ 2 calculations are a must for gate size, plunger size, or pres- sure changes because these factors affect each other, plus they are in- teractive to other factors.)

Dragging tip. A dragging tip will cause the plunger speed to change and hence the fill time to change, which can make the casting look different every shot, and will cause surface defect problems. You must use a monitoring system to measure and control speed. Causes of drag- ging tips include:

• Plunger lubrication

• Poor sleeve condition

• Poor plunger condition

• Poor cooling water flow to the plunger

• Sleeve deflection

Gooseneck and plunger ring conditions. A hot chamber injection system that is leaking metal because of poor sleeve or poor plunger rings will have an extended fill time, and as a result can have surface defect problems. It

is imperative that the gooseneck, plunger, and plunger rings all be kept in

good shape.

Low (or high) nitrogen charge in the shot accumulator. The nitrogen charge affects the speed of the plunger, especially at the end of the stroke. The nitrogen charge should be checked frequently, and always when unexpected surface defect problems occur.

When working in the design stage, or when trouble shooting a problem, the fill time can be estimated with good accuracy by using the PQ 2 calcu- lations. This will allow estimating the fill time even though there is no monitoring system in place.

22

NADCA DIE CASTING DEFECTS

Fig. 2.5. Typical PQ 2 printout using the computer program from NADCA.

PQ 2 printout using the computer program from NADCA. These calculations match the machine to the

These calculations match the machine to the die, and will save a lot of time and money over the trial and error method of changing sleeves, gate areas, and machine settings on the floor. The PQ 2 calculation works and can be done in a few minutes with a computer – it should be done for every die. Fig. 2.5 shows a typical PQ 2 calculation by the computer. A summary of fill time management would include these main points:

1. Set fill time maximum values with calculations supplemented by ex- perience, then use disciplined process control to keep it there.

2. Use PQ 2 to predict the right values for gate size, plunger size, ma- chine pressure, and machine speed settings - eliminate costly trial and error.

3. Measure and control process variables with monitor system.

Surface Defects

23

F. FLOW PATTERN

The flow pattern is at least as important as the fill time in correcting sur- face defects, however, it is not an adjustment that can be made easily on the floor. Getting the best flow pattern is an engineering design issue, and

it should be done correctly at the beginning.

A troubleshooter needs to have a basic understanding of the way metal flows into the die. This flow is different than many think, and these con- cepts will be briefly reviewed here because they are key to analyzing sur- face defects. One important step in developing the correct flow pattern is obtaining the correct gate velocity. The actual gate velocity is either measured with

a monitoring system or predicted by the PQ 2 calculation. The metal flow for die castings works far better (especially for good surface finish) if it is kept in the atomized flow range. While this is a very turbulent flow region, research has demonstrated that this type of flow is required for a good surface finish. It appears that the good surface is “spray painted” on the steel, and metal flow streams not in the atomized flow region do not do as well. The minimum velocity that will generate atomized flow is roughly defined by the formula:

[G v ] 1.7 x [gate thickness] x [weight density ( )] > 750

This says the minimum G v (gate velocity) will be defined by the gate thick- ness and G v to the 1.7 power. Thus a thinner gate will require a higher velocity to maintain atomized flow. The maximum G v will be determined by the onset of erosion and sol- dering in the die, and the velocity where this occurs for a particular situa- tion should be a judgment of the gate designer. This judgment includes consideration of the type of cast material, the type of die material, shape of the cavity, and the fill time. For 380 aluminum, the maximum velocity will typically be about 1600 to 1800 ips (40 to 45 m/s), and for zinc it will be about 2200 to 2400 ips (55 to 60 m/s). For magnesium it is typically higher, possibly up to 3000 ips (75 m/s) or even higher. For 390 (or other high silicon alloys) it is generally kept as low as possible to avoid gate erosion – usually by in- creasing the gate thickness as much as possible. In addition to the gate velocity, the gate design that develops the metal flow pattern should receive careful attention. This design should be done systematically, and is best done with computer aided tools. The

24

NADCA DIE CASTING DEFECTS

flow pattern should be developed with careful consideration of the fol- lowing design rules:

1. Use PQ 2 to size the gate and the plunger, using the appropriate gate velocity, fill time, and cavity pressure criteria. The fill time is calcu- lated by the formula given earlier, the gate velocity is set above the minimums and below the maximums given above. The metal pres- sure is discussed at length in the discussion on porosity, but in gen- eral should be above about 3000 psi static pressure for aluminum and 2000 psi for zinc.

2. Divide the casting into zones that represent different areas of concern for quality issues or for metal flow. These will be areas that are thicker or thinner than the rest of the casting, or that have special finish re- quirements, or that for some other reason need special attention. This is one of the most important actions in the design process. See the NADCA design rules for more details.

3. Proportion gates so as to fill each zone at the same time. Each gate area needs to be proportioned to the volume for the zone that will be fed by that gate.

4. Flow the short way across the casting if at all possible, depending on the directions of ribs, openings, and the location of critical areas.

5. Avoid mixing flows if possible (unless close to the gate). Metal flow paths that mix together far from the gate will likely cause surface de- fects.

6. Gate directly into areas that are of concern for surface finish if pos- sible.

7. Avoid jet flows at all times, the narrow chisel gates and widely sepa- rated gating plan will cause problems with surface finish. Always use distributed flow as much as possible.

8. If possible, set gate location so venting can be used opposite the gate.

Surface Defects

25

10. Keep runner lengths equal – the distance between the biscuit and the cavity should be as equal for all cavities as possible. Avoid “tree” type runners.

11. Eddies in the flow path (from cores or openings) will cause swirls; gate to avoid this if possible.

12. Gate to allow for high momentum – the velocities necessary will cause the metal to by-pass pockets, fins, and cores. Plan for this in the de- sign of the gates.

This list of rules is not intended to be complete. The point of listing them is to illustrate some of the things that a troubleshooter needs to consider; and to show that it is important to have a systematic procedure and defined rules. These are taught in the NADCA gating classes, and are well estab- lished procedures. They do work and the best chance of success is to use these rules. The engineer or technician who does the gate design should leave a trail of drawings and calculations to show the reasoning. This will allow an intelligent correction if things don’t turn out as expected, and prevent du- plicating efforts later. The guess and grind approach is outdated and much too expensive. While the trial and error approach is popular, it just isn’t accurate enough for modern quality requirements. It is very common practice for the first correction for a surface defect to be a gate change. While it may be desirable eventually, gate changes should not be done until the rest of the process is set within the desired guidelines. The best approach is to bring the fill time, the gate velocity, and the die temperature into normal operating windows, then do a gate change if necessary. Optimizing the rest of the process parameters, then doing any gate changes will provide a more robust solution.

G. DIE TEMPERATURE

Die temperature is another of the important factors that control surface defects. Like other factors, there are times when process conditions and part quality requirements make die temperature the primary adjustment for surface defects, and there are times when other process factors should be primary. Generally, a minimum die temperature of about 400 °F (204 °C) is required.

26

NADCA DIE CASTING DEFECTS

A too-low die temperature cools the metal flow and increases the per-

cent of solidified metal in the metal stream. If the percent solidified metal is high, it may not knit together well and the flow forms “wrinkles,” or cold flow. This is the commonest type of surface defect. Other important factors are a long fill time, a long flow path, a thin wall, and a low metal temperature. A short fill time can compensate (up to a point) for long fill time and thin wall; higher gate velocity and short fill time can compensate for a long flow distance.

Controlling die temperature is vital, yet it is not measured regularly in many shops; although those most concerned with this defect will have ther- mocouple or hot oil control in all the dies. Since the control of die tem-

perature must start with measurements, they will be discussed briefly here.

In general, measuring die temperature can be done three ways:

• Hand held pyrometer

• Thermocouple in the die

• Infrared device

Each has advantages and disadvantages:

Hand held pyrometer – very accurate and easy to use, but must stop the machine to get a reading.

Thermocouple in die – continuous reading (necessary for good process control), but does not measure surface temperature and is generally hard to maintain and awkward to install.

Infrared – easy to use, but not accurate because the die surface changes color over time (emmissivity value changes continually).

Other factors are timing during the cycle when the reading is taken and the reading location. Procedures must be established for these factors if mea- surements are to be compared from one setup to another. Following are some typical temperature range minimums needed to avoid surface defects: (these are approximate surface temperatures, as measured with a hand held pyrometer just after the casting ejects):

Aluminum

good finish 475 - 600

average finish 375 - 600

Zinc

450 - 550

375 - 550

Magnesium

450 - 550

400 - 550

Surface Defects

27

The surface temperature is not measured as often as it should be, although it is one process factor that is almost totally under the operator’s control. The temperature is often adjusted by the “look” of the casting, which is a very crude guide. The following parameters control die temperature, all of which are primarily adjusted by the operator.

• Die spray

• Water/oil flow rates

• Cycle time

Thus the usual situation is that the operating temperature of the die is con- trolled directly by the operator; and this temperature management is prob- ably the most important activity of the operator/technician at the machine. The action of these three die temperature controlling activities will be reviewed here because it is so important to controlling surface defects. Die spray is about 99% water. When sprayed on the die, the evapora- tion (boiling) of the water on the hot die extracts large amounts of heat from the die. Thus, the water controls the temperature – not the die lubri- cant material. The spray cools the die quickly because of the large amount of heat quickly pulled out of the die as the water in the spray boils away. The type of lubricant does have a major effect on the cooling action, however. The wetting action of the lubricant can cause the liquid to spread out and have a lot of contact with the die at certain temperatures. If the wetting action of the lubricant works well at the die temperature it encoun- ters, then the amount of heat extracted can be increased by at least 50% according to research done at The Ohio State University. Long spray times in aluminum die casting can cause over-cooling, which can easily cause surface defect problems and may cause water to remain on the die, which causes porosity problems. This is not common in zinc or magnesium. Application of die spray is left to the skill of the operators or techni- cians. Type of spray used, mixture ratio, spray volume, spray location, drop size, and amount of air pressure used, are a few of the important parameters. These must all be considered important if there is a surface defect prob- lem. Like everything in die casting, it is far more important to be consis- tent than anything else. If the consistency is there, corrections can be made if needed, but if the application is not consistent, it is likely that problems will never be solved.

28

NADCA DIE CASTING DEFECTS

The operator’s skill is usually an inconsistent factor; a new operator or one with a different skill level will change results quickly. In order to minimize the dependency on operator skill, it is best to do a thermal analy- sis in the design phase, and engineer the cooling so as to obtain even and reasonable die temperatures at the expected operating conditions. This up-front engineering effort will put die thermal conditions in a more robust operating window that can tolerate more variations in spray conditions, and minimize dependency on the operator’s skill. Experience has shown that a good computer thermal analysis can increase production rates by 10% – this is worth a lot of money. Some good techniques for managing die spray include:

• Spray in 2-3 sec increments with careful adjustment of the spray pattern

• Spray the areas that need cooling, not just where the casting might stick

• Automatic sprayers with a set manifold for each job are a good way to keep spray to a minimum

• Keep spray equipment in good shape – the most important factor is consistency

• Document pressures, nozzles sizes, flow adjustments, and spray times in detail, and use this data every set up

• Undocumented changes should not be allowed

• As the ends of bosses or fins will always be cold, avoid spraying the bottom of bosses or fins with direct spray because of overcooling

Factors in spray actions that are very important in controlling die tem- perature:

• Length of time of spray

• Spray nozzle adjustment, or spray pattern

• Distance from nozzle to the die

• Balance between air pressure and lubricant velocity)

• Minimize over spraying

• Document everything - spray time, pressures, spray location, mixtures, etc.

• Be consistent

pressure (drop size and

Surface Defects

29

The second die temperature control factor is control of water/hot oil flow. It is important to understand that flow rate is the critical parameter for controlling cooling or heating the die. Flow rate is generally more important than line size or even the inlet temperature of the fluid. This is especially true for water cooling where there is little control available on the inlet temperature, but less true for hot oil units where inlet temperature is more controllable. Measuring flow rate is a very desirable way to improve process con- trol and will be a very good way to maintain consistency. Another method is to maintain consistent water pressure at each machine with pressure re- ducing valves. Flow rate is determined by the smallest opening in the supply line; this is usually the quick connect fitting, but it can be other connectors. The water (or oil) manifolds must be designed properly, otherwise there will be more area in the outlets than in the inlet pipe. This can make for reduced flow in some lines. Flow adjustment valves should be easy to see, easy to use, and have settings that can be repeated (or the flow rate measured). Low flow rates may over-cool and cause cold flow defects, when this happens, consider using air instead of water or use solenoids to turn the water on and off each cycle. Hot oil systems will often make a significant difference in the surface defect rate for two reasons:

• It keeps the die hot during stoppages – this is where hot oil can easily pay for itself and

• It can add heat or cooling to the die as needed to maintain consistent die surface temperature

Many surface defects come during start up situations. Often it is hard to determine the last bad casting and the first good one and frequently there are some bad castings in with the good, or vice versa. External heating units (hot oil or cartridge heaters) can eliminate this major source of sur- face defects. Hot oil units cool about half as effectively as water, so thermal design must account for this to get the cycle times desired (use higher flow rates, move lines closer, make larger, etc.) It is difficult to get the full benefit from hot oil when using dies de- signed for water cooling as high flow rates usually are needed. A pumping unit that can supply high flow rates is desirable.

30

NADCA DIE CASTING DEFECTS

Some things to consider in design of water/oil heating/cooling:

• Do not place water lines around the outside edge of the cavities (this cools the cold areas)

• Give priority to cooling/heating lines, even if this means moving ejec- tor pins or other changes

• Do not use the same line to control temperatures in both a hot area and a cold area

• Depth of the line is critical, set depths carefully

• Size of line must match the flow rate, do not use a big line in the die and use the same connections (the heat transfer will go down instead of up)

• Provide for thermocouples in the die, this will not cost anything in the die build, but will provide the basis for thermal control if needed later

• Treat cooling water because deposits only 0.005 in. thick (0.125 mm) cut heat transfer by about 40%

• Review the connections to be sure the lines aren’t too large for the connection method

• Use constant pressure (pressure reducing controls)

• Measure flow and use flow controls

• Do use computer thermal analysis for optimum cycle time

• Use hot oil to keep the die hot for pre-heat and short stoppages

• Maintain cooling water and oil temperature at a set point as much as possible

Adding overflows is a common method of reducing cold flow defects. The biggest benefit is usually the increased die temperature from the added metal volume. This heat can be added at spot locations, and directed at either die half, so it is a good way to add spot heating as needed. This can make a much more noticeable difference if the casting has thin walls in the area of concern. A good thermal and flow analysis can show the need for the spot heating and help select the best method for adding the needed heat. Recycling metal does have some costs. Overflows do change the metal flow pattern to some extent, and some of the leading edge of the flow (with it’s attendant porosity, oxides, and

Surface Defects

31

Fig. 2.6. Porosity indicates flow pattern into an overflow. Probably indicates poor venting, but also could be influenced by restrictions at the overflow gate, slow fill time, poor flow pattern, low die temperature, and long flow path in combination with a thin wall section

(50x).

flow path in combination with a thin wall section (50x). Courtesy Prof. J Brevick, The Ohio

Courtesy Prof. J Brevick, The Ohio State Univerisity

solidified metal) may find it’s way into the overflow, but this is not always the case because the small gates required inhibit free flow. Fig. 2.6 shows some of this.

A well designed thermal management system and well engineered flow

system will greatly reduce the number of overflows required. If these are done in the up-front engineering, then a minimum number of overflows should be used; frequently they are not needed at all.

Another important control for the die temperature is the cycle time. Again, the cycle time is essentially under the operator’s control, even in an automatic system. The die temperature at any given time is the direct result of the number of pounds of metal that went through the die in the last one to two hours.

A consistent cycle time is one of the most important factors in good

defect control. The cycle time should be measured and displayed to the operator or technician in some way to get good control (if it isn’t mea- sured, it isn’t controlled). Die temperature changes slowly, which can cause some delayed ef- fects, and perhaps some confusion.

32

NADCA DIE CASTING DEFECTS

Example: What can be expected from a change that adds nozzles to the spray manifold and shortens the spray time, but supplies much more cooling than before and shortens the cycle time? There will be two effects, one short term and one long term. The short term effect will come from the increase in spray cooling, and the long term effect from the change in cycle time. Increasing the cooling from spray will reduce the die temperature quickly (short term), and reducing the cycle time will increase the die temperature (long term). While these two effects may possibly balance out in the long run, the first change the operator notices may not show the total result of the changes made because of the delayed effect of cycle time on die temperature. Part of controlling the cycle time is keeping the die hot during short stoppages and pre-heating the die before all startups. If record keeping is done as suggested earlier, it will become apparent to management that the startup scrap is a very significant cost, and also that much of startup scrap is surface defect scrap. It is not unusual to find that the startup scrap is 30% to 50% of the total scrap, which makes it worth a considerable engineering effort. Much of this kind of scrap comes from short stoppages during production. The die isn’t usually pre-heated after these stoppages, and the first few pieces are scrap or at least marginal. The primary actions that will make a difference are: using hot oil or electric cartridge heaters that will automatically add heat as soon as the die stops; and not starting the die until a pre-set temperature is reached. Summary of correcting surface defects with die temperature:

• Minimize start up scrap and marginal production situations with die pre-heating; keep die hot during short stops

• Measure die temperature to know where to change and how much

• Establish temperature goals for minimum defects

• Increase die temperature in the area of the defect by:

– Reducing spray

– Reducing water flow rate

– Adding overflows

• Increase overall die temperature by:

– Reducing cycle time

– Increasing hot oil temperature and flow rate

Surface Defects

33

• Use good spray practices and keep consistent

• Use good computer aided thermal analysis for cooling /heating line design

• Measure and control flow rates and establish process discipline

• Use good engineering to provide quality cooling water at consistent pressures and flow rates

H. METAL TEMPERATURE

The metal temperature can affect surface defects; obviously, the hotter the metal the better the surface finish, although the metal temperature is al- ways limited by other factors. In aluminum, limits on high metal temperature come from flash, more oxide formation, greater sleeve and die erosion, soldering, shorter plunger life, difficulty in maintaining consistent temperatures and other problems. In zinc, high temperature causes flash plus erosion on the pot and the goose- neck, the loss of magnesium and other problems. In magnesium, high metal temperatures can cause flash and oxide problems. In general, adjusting metal temperature is not one of the best ways to reduce surface defects for the long term, although it can be a short term fix. This is particularly true if the temperature adjustments are not controlled, and the metal temperatures are changed frequently. In this situation, sur- face defects can be better improved by eliminating all adjustments and using one setting for all shifts. In general, the most desirable situation is to keep the metal tempera- ture at a high range, but not high enough to cause a lot of other problems. Keeping zinc at 810 °F (432°C) as a maximum, and aluminum at about 1270 °F (687 °C) maximum are reasonable limits. The most important factor is maintaining a consistent holding tem- perature; allowing more than a 10 °F (5.5 °C) variation can cause prob- lems, especially in thin wall parts with long flow distances. This should be watched carefully no matter what set point is used. The holding furnace and the furnace temperature controllers need to be kept in good condition so temperature variations can be kept within these limits. The dynamics of many metal delivery systems make it difficult to change holding temperatures when low energy holding furnaces are in- volved. This is because the delivery system will deliver at one tempera- ture, while the holding furnace has to add or lose heat to get the metal to the correct set point. This can be a slow process, and a better operating

34

NADCA DIE CASTING DEFECTS

procedure is to set the holding furnace at a reasonable temperature and maintain that setting for all dies ( a few exceptions can be permitted). In cold chamber operations, temperature loss in the ladle and sleeve must be kept to a minimum, or at the least it should be kept consistent for all situations. The ladle waiting time must be consistent, and delay in the sleeve should be carefully controlled (it may be set by other process conditions). For hot chamber, the temperature of the nozzle can be significant. The nozzle must be hotter than the molten metal so the nozzle can add heat to the metal, but if the nozzle is too hot, it will start to react with the molten metal and cause washouts and other problems. The biggest problem is leaving the nozzle on high fire for short stoppages; this causes large fluc- tuations in nozzle temperature – can be controlled automatically. In summary, the metal temperature control should use the following factors:

• Control metal temperature on holding furnaces to within ± 10 °F (5.5 °C) or less using good furnace and controller maintenance procedures

• Use consistent holding furnace set points, do not use metal tempera- ture as a variable unless absolutely necessary

• Use consistent times between ladling and shot

• Control nozzle temperatures for hot chamber machines

The type of alloy can make a quite a bit of difference in the surface finish. In zinc, Zamak 7 was designed to have the best fluidity and surface finish of the common alloys In aluminum, the alloys closer to the eutectic will be more fluid and will tend to have better surface finish (384 for example). The Zamak alloys have up to 4.2% aluminum in them and the fluidity is quite a bit higher at the high end of the aluminum range; this makes a difference in thin wall decorative castings. The silicon content is important in the common aluminum alloys, and should be maintained at the high end of the range. The variation in silicon content should be reduced from the 2% allowed in the specifications be- cause the difference in the flow and the freezing characteristics is mark- edly different at the extremes of the allowed silicon ranges. The eutectic alloys will have a smaller freezing range and will be more sensitive to variations in the holding temperature and other process varia- tions, and so are regarded as harder to cast. Venting and vacuum can be significant for surface defects in some cases. Certainly the trapped air will cause blisters and gas porosity, but

Surface Defects

35

will also cause back pressure in the cavity. This back pressure can change the flow enough to cause surface defects. This will be most noticeable in blind features (such as bosses, fins, etc.), where the back pressure from trapped gas may be enough to prevent a complete fill in these areas. Review the metal flow pattern to find out if this could be a problem. Probably the regions close to the last points to fill will be the most affected. The gating design should have been carefully done so as to set the last point to fill at a location where venting would remove the trapped gasses. To remove gases from blind features, it may be necessary to add vacuum. Vacuum will be beneficial in all situations where back pressure is sus- pected of altering the flow path and should be used wherever it is feasible. The other benefits of vacuum, such as reducing out-gassing, blisters, and gas porosity, make it desirable to use in any case. Vacuum can be added to hot chamber machines by drawing the vacuum before the shot starts, then starting the shot when the vacuum is established and the metal has been sucked almost to the sprue by the vacuum. However, just the proper engineering of the gating and venting should provide a system suitable for many die castings without the use of vacuum. This requires a well designed gating system with a planned fill pattern, and vents that are large enough to allow the air to escape during cavity fill (see section on gas porosity).