Surface Defects

A Guide to Correcting the Problems

By: William Walkington &

Dr. Stephen Midson

Publication #517
 Although great care has been taken to provide accurate and current information, neither the author(s) nor the publisher, nor anyone else
associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by
this book. The material contained herein is not intended to provide specific advice or recommendations for any specific situation. Any
opinions expressed by the author(s) are not necessarily those of NADCA.

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for identification and
explanation without intent to infringe nor endorse the product or corporation.
© 2015 by North American Die Casting Association, Arlington Heights, Illinois. All Rights Reserved.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying,
microfilming, and recording, or by any information storage and retrieval system, without permission in writing form the publisher.
 SURFACE DEFECTS:
A GUIDE TO CORRECTING PROBLEMS
By: William Walkington & Dr. Stephen Midson

Table of Contents

Introduction 3
First Things First 3
Some Basics About Metal Flow 4
Atomized Flow 7
Improving Surface Finish And Correcting Cold Flow Problems 9
Fill Time 10
Correcting Problems Associated With Fill Time 11
Die Temperature 12
Adjustments for Die Temperature 12
Metal Flow Pattern — Gating 16
Other Surface Defects 20
Other Flow Factors To Conside 21
Summary 25
References 25
INTRODUCTION
This workbook addresses one of the most common die casting defects — a surface finish that doesn’t
meet the customer’s expectations. The die casting process can inherently provide a far better finish
than any other metal casting process — but (as is case for die casting in general) good surface finish
requires far more attention to the details of the process than any other casting process. This guide is
intended to provide some help and information for the practitioner in developing and maintaining a
defect-free surface finish.

The surface finish issues discussed here will be mostly those affecting appearance. The appearance
requirements necessary for customer acceptance can range from a totally defect free surface with a
perfect mirror reflection (no wavy lines in the reflection) to a coarse finish with visible blemishes
that feels rough to the finger. Just as there are these large differences in finish requirements, there is
also a great deal of difference in the engineering of the die and in the process conditions between the
extremes of the quality requirements – and it is critical that both the die caster and the customer un-
derstand just what affect these requirements have on the effort required (and hence the cost), and that
the expected surface finish should be defined clearly very early (before the project is even quoted).

The majority of this workbook will address flow-related defects such as flow lines and cold flow. In
addition, towards the end of this workbook, other surface defects such as soldering, blisters and sur-
face sink will also be examined.

FIRST THINGS FIRST

While it may seem simple and basic, the act of carefully defining the finish requirement for a specific
part is often overlooked, and this generates a lot of irritation, confusion, and extra cost. For example,
a customer that expects a “Class A” finish on a zinc die casting will need to define just what surfaces
are expected to be “Class A”, and to define what is meant by Class “A” in their opinion, and do this
precisely and completely. This information greatly affects the die caster’s engineering and the process
conditions, and these, of course, can also greatly affect the cost of the product.

Because there is such a variation in the product requirements, it has been difficult to develop general
industry standards for surface finish, so most die casters rely on obtaining a good definition of the ex-
pected appearance standards from the customer. If the customer doesn’t have well defined standards,
it is very much to the advantage of both the customer and the die caster to define them (at least for
the project at hand), and do so early and as accurately as possible.

By far the best experiences in setting standards have been with a system where quality levels are
developed, and then samples are developed that apply to the different quality levels, and the die caster
and the customer each posses equivalent samples. (If no samples exist at the start of a project, then
this surface quality definition can be done with existing castings of a similar shape, however, eventu-
ally quality samples using the actual part must be developed and used.)

In particular, it is very important that samples be developed that show the dividing line between
good and bad. This means that a set of castings is found, one of which is judged to be just barely good
enough to accept, and the other judged to be just barely bad enough to be rejected. Just the exercise

3
of finding these kinds of samples often helps a lot in defining the actual demarcation line. Things
always change, so this exercise may need to be repeated every year or every other year.

The best definition would be one where a group of castings is obtained that is ranked, say from one to
five, with five being the best and one the worst, and with the dividing line set where level 2 is a reject
and level 3 is an acceptable casting.

Once the quality level is defined, the die caster needs to focus on what can correct and prevent sur-
face defects, and the first step is to review the causes of surface defects.

SOME BASICS ABOUT METAL FLOW

In die casting, the biggest portion of surface defects come from what may be termed “flow defects”.
This refers to a condition where different metal flows come together, but do not homogenize and mix
at the dividing line, and they have not knitted together as well as needed for a good finish.

If the metal flow gets too cold, then the metal starts to solidify and will generate a “skin” of partially
solidified metal and oxides. When these flows come together, the skins keep them from mixing ho-
mogeneously, and they will just push against each other rather than mixing. This can produce a line
(or wrinkle) on the surface of the casting.

The lines or imperfections that develop between metal flows are often called “cold flow”, and that
will be the name used here. However, these lines are called many names, depending on the customs
for that location. Some of the more typical names for these surface defects are listed below, with the
most common listed first.

Cold Flow Cold lines Cold lap Knit Lines

Flow Lines Non-fill Chill Cold Shuts
Lines Poor fill Misruns Swirls
Flow Marks

A typical set of these conditions is shown in Figures 1 through 4.

(From left) Figure 1. Typical cold flow area, Mag 20x. This shows the typical roughness visible with the
naked eye. Figure 2. A close up in the center of the rough area in Figure 1. Mag 200x. Figure 3. Same
area blown up to a magnification of 420x. The folds and wrinkles are becoming obvious.
All Photos Courtesy of Cominco Ltd.

4
 Figure 4. This shows a cross section through
the area where there is a cold flow line. Note
the lines indicating here is a flow layer on both
the left and the right, but this separation of
flow does not go very deep into the casting in
this case. Mag 400x Courtesy of Cominco Ltd

The general concept is that the die is very cold to the flowing metal, and absorbs heat very quickly, so
the leading edges of the metal flows tend to form “skins”. These skins are partially solidified material
and, especially in the case of aluminum, they may also contain oxide skin layers.

Thus, there is enough time lapsed as the flow moves across the die that when two flows come togeth-
er, they do not mix homogenously, and a cold flow line is formed as seen in Figure 4. The lapsed time
(fill time) is important, but also it may be that the die surface is too cold, or the flowing metal is too
cold, or (more likely) the problem arises from both of these factors.

Usually the lines do not go very deep into the casting, and they do not usually cause any structural is-
sues ­— although that will depend on process factors. It is possible that the line or separation between
flows may go deep into the casting, and the resulting separation can cause a casting to be weak and
to crack along this separation, or the separation can provide a leak path if the inside of the casting is
pressurized. In general, however, the presence of cold flow will rarely produce structural problems,
but it is possible.

Also, when the separation or the line between flows is not very deep (by far the most common situ-
ation), then it is possible that disturbing the surface, such as is done with grit blasting or machining,
may cause the layers to separate, and lamination (or peel) occurs. Good process engineering design,
including simulation, is the best way to minimize these problems.

Note that in Figure 4, there is also a slight difference in microstructure between the flows on either
side of the flow line. This means that the temperature and especially the freezing time of the different
flows was not the same.

Figure 5 shows a photomicrograph of the edge of the flow from a single fan gate (used to produce a
household flat iron). This difference in microstructure between the direct gate flow and the side flow
occurred as the direct flow from the gate (course structure) was hotter, and took longer to solidify, while
the area adjacent to the direct flow path filled by back flow, but filled quicker, and hence froze faster,
producing a finer microstructure. Either structure would work just fine for the function of the part, but
the difference in appearance between the fine and course microstructures resulted in a surface finish
problem as the finishing method caused the two surfaces to look different, making the casting a reject.

5
 Figure 5. Line between flows in a flat iron base. The
continuous flow from the gate is below the line (hot-
ter and a more course structure),and the later flow is
above the line (fine structure). In this case, the surface
appearance of the two structures was sufficiently dif-
ferent after finishing (epoxy and other coatings were
used) that the appearance resulted in a reject.
Courtesy J Brevick, OSU

Figure 6 shows a still from a video demonstrating how the flow can back fill along the edge of the
main flow path. This is a typical die casting flow pattern where the flow goes to the end of the casting,
and then backfills around the direct flow from the gate.

Figure 6. Fill from a single gate with back-

flow filling in around the sides of the main
fan gate flow — giving the line between
flows shown in Figure 5. (Flow 3D picture)

The time available between flows (the time it takes to develop a thin skin on the metal at the edge of
the flow) can vary considerably, and will range from a very few milliseconds for zinc or magnesium,
to probably well over 25 milliseconds for thicker walled aluminum castings cast at higher tempera-
tures.

6
ATOMIZED FLOW
While the cold flow lines are usually formed from temperature differences and/or timing of the metal flow
paths, the situation where there are several metal flow streams usually means that there will be cold flow at
the far end of the flow paths. However, the flow should be different at the beginning of the flow path, where
the best finish is obtained when the casting alloy is “spray painted” on the die surface. The “spray painting”
effect is obtained where the metal flow is atomized into a fine spray.
This atomized flow is a large part of the reason (higher die temperature is the other part) why most cast-
ings have an excellent surface finish in the area just past the gate. This area may be short (less than 0.5
inches, or 12mm), or it may extend for several inches. This is the area where the gate flow reaches the die
steel surface without impingement on other surfaces, and the flow is generally in the atomized flow form.

The atomized flow (think of paint spray type of flow) is generated at the gate by combining the cor-
rect gate design with the proper metal flow rate. The flow conditions required for atomization can be
defined approximately with the formula provided below.

(NOTE: While this formula is an excellent guideline for developing the atomized spray condition, it
should not be regarded as the only gating criteria needed for a good surface finish.)

The formula is based on research, but the constant used is empirical, as noted below. The formula
says that the metal flow will be atomized if the following calculation yields a number above 750 (1.4
million if units in metric):
English:
(Gate velocity(in/sec))1.7 x (Metal density(lbs/in3)) x (Gate area(in2)) ≥750
Metric:
(Gate velocity(cm/sec))1.7 x (Metal density(g/cm3)) x (Gate area(cm2)) ≥1.4 million

The factor of 750 is empirical, and has not been confirmed with research in die casting alloys. It
was first developed by Russ Van Rens, and, in the authors’ experience, it fits very well with practical
results on the production floor. It is recommended that this number be used for most die casting
situations, and it is especially valuable for those situations where surface finish is an important con-
sideration. (However, process engineers can do some experimenting and may choose to develop a
different factor for special situations.)

In practice, the formula to calculate the minimum gate velocity (in ins/sec) is usually transformed
into the following format, making it is easier to use:
English:
750
Minimum gate velocity = { (Metal density x Gate area) }0.5882

Metric:
1.4 million
Minimum gate velocity = { (Metal density x Gate area) }0.5882

Where: The densities are for the liquid metal, and these values will be very close to the following
for die casting alloys:

Zinc = 0.221 lbs/in3 (7.14 g/cm3) Magnesium = 0.063 lbs/in3 (1.74 g/cm3)
Aluminum = 0.093 lbs/in3 (2.7 g/cm3)

7
It is strongly encouraged that this formula be used to define the minimum gate velocity for obtaining
a good surface finish. Note that the formula indicates that a high gate velocity is important for atomi-
zation, but that a thicker gate would also provide atomization at lower speeds.

Some representative minimum gate velocity values can be calculated using a typical gate thickness
of 0.02 inches (0.5 mm), 0.07 inches (1.8 mm) and 0.06 inches (1.5 mm) for zinc, aluminum, and
magnesium, respectively. These minimum values, along with some typical values used in practice, are
listed in Table 1.

Metal Minimum Gate Velocity Typical Velocity

ips m/sec ips m/sec
Zinc 1,150 29 1,800 45
Aluminum 910 23 1,200 30
Magnesium 1,110 28 2,600-3,600 65-90
Table 1. Approximate minimum and typical gate velocities.

While the values shown in Table 1 are approximate, they are fairly typical of values used in general
industry practice, and can be used as rough guidelines – although it is best if the die caster does the
complete fill time, PQ2, and minimum gate velocity calculations for each part. This will provide the
proper process values for plunger size, machine size, expected plunger speeds, etc. that will maintain
the desired gate velocities.

A high gate velocity will tend to make certain the flow is atomized, especially if the gate and runner
design are not proportioned properly, and this will tend to make the part look somewhat better. If
the operator is allowed complete freedom to adjust, it is likely that the resulting gate velocity will be
high. However, this often comes at a price – especially in aluminum – because the high gate velocity
will generate gate erosion, solder, and sticking problems for aluminum castings – and, to a lesser
extent, it can also happen in zinc.

Magnesium seems to be quite a bit less sensitive to the velocity issues, and the use of very high veloci-
ties is often done — although a well-designed gate, using reasonable velocities, also seems to work
well for some magnesium die casters. A good engineering design effort will almost always result in
a reasonable gate velocity (with low erosion) and a part appearance that is as good or better than one
depending on extra high gate velocity to get a good appearance.

A factor in the atomized flow situation is that the “spray“ coming from the gate does not last long be-
cause of the complexity of most die castings. Given this complexity, the spray will not go far before it
encounters a wall or feature that deflects the flow. After one or two of these deflections, the flow is not
a spray anymore, and becomes a very turbulent flow with some jets and squirts imbedded in it. This
kind of flow is generally is what will reach the end of the casting, and is the kind of flow that gener-
ates cold flow.

This condition is shown in Figures 7 and 8 . These pictures show how the flow slows down and
becomes turbulent after a one or more bends or obstacles. Once the flow becomes turbulent (not a
spray), there is usually some jetting (or squirting) type of flow mixed in with slower flow.

8
 Figure 7. The coloring is by velocity, and
the yellow coloring shows a velocity fast
enough to maintain atomized flow, the
green coloration appears after the second
90° bend in the flow path, and atomiza-
tion has stopped after the second bend
for this casting.

Figure 8. Cross-section of the gate and

part for the flow shown in Figure 7. Note
the atomization flow at the arrows, and
that the flow is slower and the metal
flow is quite different after the flow slows
(after two 90° bends). Novacast Simulations

IMPROVING SURFACE FINISH

AND CORRECTING COLD FLOW PROBLEMS
After reviewing some of the basics, it should be clear that cold flow can be eliminated by keeping the
metal more liquid and trying to keep skins from developing. This is the basic goal of all the corrective
efforts. Several options are available for the engineer, with other choices being available within each
of these options. The three basic approaches are:

1. Do everything faster (basically, reduce the cavity fill time) – if the die is filled quick enough,
there is not time for the skins to form, and the metal flows can mix without cold flow.
2. Increase the heat content of the die or the metal, with the same result as #1, except more flow
(fill) time is available with higher temperatures.
3. Optimize the gating to improve the metal flow characteristics, thereby minimizing the type
of paths that allow cold flow to develop between different metal flows.

9
Correcting an existing cold flow problem usually involves working on one or all of these three areas.
These are all interactive — meaning that a very fast fill time will substitute for a poor die temperature
conditions (up to a point), or that a very hot die will correct for a slow and poor performing machine,
but again only up to a point. Usually, the trouble shooter will need to review each of these three fac-
tors to do a good job of solving the problem.

FILL TIME
Selecting the correct fill time is achieved either from experience with similar castings, or by calculat-
ing the fill time using the NADCA formula. The NADCA gating design was initially based on the
concept of first developing a fill time that will provide a good finish, then calculating all the other
process parameters that are necessary to sustain that fill time. Thus, the NADCA gating design is
intended primarily to provide a good finish and a casting that looks good — and it does work quite
well in this regard.

The first step in the NADCA system is to calculate (or estimate) a reasonable fill time that includes
allowances for the part thickness, expected die and metal temperature, and for the quality of the
desired surface finish. The formula to identify the maximum fill time allowable to get a good surface
finish is:
(T - T + SZ)
t = k{ i(T -f T ) }T
f d

Where:
t = Maximum fill time allowable to provide a good finish
Ti = temperature of the molten metal as it enters the die
Tf = Minimum flow temperature of the metal
Td = Temperature of the die surface when the metal enters
T = casting wall thickness
S = Percent solids fraction allowable for the finish quality that is desired
Z = Units conversion factor, different for each alloy involved
k = Empirical constant = 0.866 sec/in (or 0.0346 sec/mm)

For any given situation, the user of this formula must make judgments about S (the percent solids),
the casting wall thickness, the die temperature, and the metal temperature. The average casting thick-
ness is usually taken as the thinnest wall section that has a visual requirement, or else is taken as the
average wall thickness for the most critical part of the casting. Detailed instructions on using this
formula are given in the NADCA book on gating design.

It should be noted that the value “T” should be regarded as the maximum fill time that will assure a
good surface finish. A very short fill time will always improve the appearance of the casting, so there
is no minimum fill time. Choosing a lower value for S, using a lower die temperature or a lower
metal temperature will result in a shorter fill time being required by this formula. It is quite possible
that selecting conservative values will result in a fill time that is impossible to achieve, so performing
the PQ2 calculation for all castings is also important.

10
Using this formula, therefore, should only be considered the first step and only predicts the maxi-
mum fill time that can be used for the conditions selected. The next step is to calculate the minimum
gate velocity, and then to select a target gate velocity to be used for this casting. The PQ2 calculations
can be run next, and they will tell whether or not the calculated maximum fill time can be achieved
for the machine that has been selected. Thus, they tell if the machine is suitable, although these cal-
culations should always be run (even for new machines that are known to be suitable) because they
will also provide information on the proper plunger size and machine setting for optimum perfor-
mance.

When trouble shooting a problem in surface finish, measuring the fill time is a high priority, and
this requires a shot monitoring system. Eyeball estimates are not adequate. When comparing the fill
time from a shot monitoring system to the calculated value, a good comparison comes from calculat-
ing the time from the average speed during fill — the fill time reported on most monitoring systems
doesn’t compare directly to the fill time predicted with the calculations.

If measurements are not available, it will be difficult to get a good assessment of whether the fill time
is an issue for a surface defect problem. One way of estimating the fill time is to use the PQ2 calcula-
tions (if the dry shot speed is known). This will usually be fairly accurate, but assumes everything is
working as it should.

CORRECTING PROBLEMS ASSOCIATED WITH FILL TIME

One problem might be that the gate may be too small for the machine, causing back pressure from
the gate to slow the machine down enough that the fill time is too long, and the casting appearance is
terrible. This is a very common occurrence in some shops. Again, it should be emphasized that the
plunger speed cannot be gauged by eye, so it is impossible to visually check if this is happening.

A long fill time often occurs where the tendency is to use the old philosophy of “make the gate small,
then we can always open it up” prevails. In this case, often the initial gate is small enough that it cre-
ates back pressure, slows down the machine, and causes a long fill time. In this case, any gate change
that opens the gate area will cause a reduced back pressure and a shorter fill time, which will cause
the casting to look better. Thus, the trouble shooter is almost guaranteed to be right in making a cor-
rection that opens up the gate.

Another clue that the gate may be too small (which probably makes the fill time too long) would be a
situation where there doesn’t seem to be any change (or very little change) in the casting appearance
(or the plunger speed if it is being measured) when the plunger speed control valve is opened up.

In general, it is cheaper and better to design the gate properly from the start. If this is performed us-
ing the NADCA procedures and done correctly, very seldom will much be required in the way of gate
changes.

If no data or calculations are available, a good step is to compare the gate area of a similar casting that
has a good finish and (hopefully) that has been run on the same machine. A second step would be to
just assume a reasonable plunger speed, and do the calculations to get a rough idea if the design value
is close to the right range.

11
It is also possible that the gate area on the problem casting may be too large, which would not cause
a long fill time, but rather could result in low gate velocity. This will also be a problem, although of a
little different nature. (In this case, there will probably be evidence of metal flow pattern problems,
such as streaks, cracks, cold flow between good flows, etc.)

Too small a plunger or gooseneck will cause a long fill time. For a good finish, the plunger size HAS
to be selected based on the fill time requirements of the casting. In cold chamber machines, some-
times the plunger size is chosen based on a desired fill percentage, and if the plunger size is inap-
propriate it will result in a poor surface finish arising from a low flow rate and long fill time. In this
case, it is often much better to compromise and use a larger plunger that will give the appropriate fill
time. While it is true that this could possibly trap some additional air, it is likely that this is the only
solution that will provide the correct fill time and a suitable surface finish – and often the additional
trapped air will not be noticed.

Other operational things may also cause a long fill time and a poor finish. One check is to find out if
the problem is consistent, or if it comes and goes. If it is inconsistent, it could still be a long fill time
caused by some problems in the injection system. For example, in a cold chamber machine, if the
plunger is worn and dragging, this will slow down the shot just when it needs full pressure and speed
to get the fill time required for a good finish. Look for chattering movement in the plunger, or for
flash around the plunger tip.

Another cause could be the accumulator pressure falling off, usually from a low nitrogen change —
although an excessively high nitrogen charge can give the same results. This will cause the pressure to
drop and the speed to slow just at the critical time when the cavity is filling.

DIE TEMPERATURE
Along with optimizing fill time, having the die temperature in the right range is critical for obtain-
ing a good finish. Once the fill time is adjusted to the best setting available (proper plunger size and
speed), then further adjustments to improve surface finish will be normally be done with the die tem-
perature. Die surface temperature is often controlled by the operator — and it is also something that
is not measured as much as it should be, which makes it difficult to control in some shops.

While most temperature adjustments are controlled by the operator, one thing that is controlled by
the engineer or toolmaker is the placement of the heating/cooling channels — and some general
guidelines are provided below for locating heating/cooling lines — but most of the following discus-
sion for die temperature control is about operational adjustments of the process.

ADJUSTMENTS FOR DIE TEMPERATURE

General practice suggests that the minimum die temperature should be above 350°F (176°C) for zinc,
and over 400°F (205°C) for aluminum and magnesium — and such temperatures are necessary just
to obtain an average or typical surface finish. For a good surface finish, however, zinc dies should be
run at die temperatures over 400°F (205°C), and aluminum and magnesium dies should be run at
least at 450°F (230°C). However, aluminum dies will often run at 500°F (260°C) or higher — which is
a suitable die temperature to obtain a good surface finish.

12
Cycle Time
Cycle time affects the die temperature because it controls the heat input to the die. For steady state
conditions, the instantaneous die temperature depends entirely on how many pounds of metal were
solidified in the die in the last 1 to 2 hours. Thus under steady state conditions, the die temperature
will be controlled by the recent cycle rate and changes in cycle rates will not cause instantaneous
changes to the die temperature. Therefore, as is true for most die temperature situations, the trouble
shooter needs to make adjustments and wait at least 15 minutes (and maybe even several hours for
big dies) to see the results of a die temperature adjustment.

For a good surface finish, a steady cycle rate is critical. A very fast rate that cannot be sustained
(running 10 or 20 shots, then stopping to clean flash, for example) will obviously give bad results for
finish. It is better to run at a slower rate, but steadily, even if this results in a colder die than would be
desired. One the process is stable, then the die temperature can be increased if necessary using other
adjustments.

These other adjustments include the addition of hot oil, cartridge heaters or torches. The ability to
set and hold a temperature on hot oil systems is a great advantage for getting good surface finish, and
this system should be used whenever possible — even to the point where oil is used where cooling
water would work or where normally no lines would be installed. The oil system temperature control
is automatic, and even though it is limited on its heat removal capability, it tends to result in a much
more stable process. Thus it is desirable to use hot oil as much as possible for surface finish concerns,
even where it may not be regarded as a necessity.

Spray Conditions
For most operators, die spray is the first choice of the tools available for controlling die temperature,
especially for aluminum castings. Of course, spray can only add extra cooling, and so it is used to cool
hot spots — so an increase in temperature (as would typically be needed for improving surface finish)
using die spray is basically a just a reduction in the cooling effort. Thus, die spray is most useful in
trying to make the die temperature more even, which is valuable of course, although the die must be
heated by other means for this to be effective.

In some aluminum dies, spray is the primary (or even the only) method of die cooling — this can
work fairly well if the operators are experienced and jobs are long running (although this approach
can result in very short die life). For zinc and magnesium, the spray function is often used just to
provide a little die release agent, but also can provide a little extra cooling in some spots.

The use of spray benefits surface finish mostly by making the surface temperature more uniform, and
this can only be effective if the operator can tell where the hot and cold spots are located. Thus adjust-
ments by operators are often done fairly slowly as the operator tries to determine the actual tempera-
ture balance that works the best. Obviously, the operator needs a way to measure die surface temper-
ature if he is to do a decent job in adjusting die spray, and a reasonably priced hand held pyrometer
should be available for all operators and set up personnel. Without it, time is wasted as the operators
try to guess at the die temperature by using small clues in the casting appearance.

Die spray can also affect the surface finish with stains from the release agent if applied incorrectly.
The spray mixture should be measured and monitored carefully, and it is best if the application is
automatic to minimize variations.

13
There are many other issues about die spray, however, for good surface finish, the consistent applica-
tion of a small amount of lube is usually the best.

Water or Oil Lines

The location of water or oil lines is best determined by the use of thermal simulation. However, there
are many valuable empirical rules for determining the best location for cooling lines in the die, and
while there isn’t room for all of them here, some of the important ones for good die temperature con-
trol (and hence good surface finish control) would be as follows:
1. Don’t use the same water line for controlling the hot spots and the cold spots (see Figure 9).

Figure 9. Left sketch shows an insert with one line that cools a hot area (under the runner), and also
goes past a cold area (top of the die). This means the water line is trying to cool both a hot spot and a
cold spot. A better design is shown on the right.

2. Run the lines under the cavity if at all possible — compromise on ejector pin locations before
compromising on water lines. Heating and cooling are much more effective if under the cav-
ity — don’t put the lines in as the last thing and then just put them around the outer edge of
the insert. Back drill so the cooling line starts and ends underneath the casting.
3. If possible, it is better to run the lines perpendicular to the metal flow direction.
4. Vary the depth as needed — for example, if the line goes past a thin section of the die that
doesn’t need cooling but this path is necessary for the water line, then change depth to reduce
the impact of the line. Cooling lines should be deeper where less cooling is needed (such as
lines at the far end of the casting). If possible, use calculations to set the depth.
5. The use of lots of fountains and baffles from the back of the die is a good technique that al-
lows for much better thermal control, and is especially valuable for a die with uneven wall
sections and for the best management of surface finish.
6. Always number lines and stamp IN and OUT on the die.
7. Pay attention to the depth of water line beneath the surface of the casting — typical mini-
mum for aluminum is 0.75 inches (19mm), although this can be reduced if new, tougher die
steels are used. Zinc and magnesium can be less, approximately 0.6 inches (15mm) for most
situations. Minimum depth for hot oil lines is typically about 0.5 inches (12.5 mm).

Guidelines for proper water and hot oil connections

In addition to these design concepts that the die designer or engineer must have in mind, the proper
use of water and hot oil connections in the shop deserves a lot of attention. Some guidelines:

14
 1. Looping (series connection) will at least insure the same flow rate through all lines in the
loop, although the last connections can get relatively hot water or oil. Parallel (manifold)
connections do not always divide flow evenly, and some lines may only get a very small
amount of flow. This is most important for hot oil connections.
2. Make sure you have turbulent flow rates in the water lines — a flow rate of about 1 gpm (3.8
liters/min) or above for a 0.44 inch (11 mm) diameter line will usually assure turbulent flow.
(Heat transfer drops very significantly below this velocity — low flow rates often occur when
lines are connected in parallel and one line has low resistance and one has high resistance.)
3. Do not allow the lines to develop scale — even 0.005 inch (0.12 mm) thick scale will cut the
heat transfer by 40% or more.
4. If the designer has increased the line size to remove or add more heat, the shop connections
must increase the supply line size, otherwise the increase in heat transfer from a bigger size
channel will be negated by the reduced flow velocity.
5. Measure the openings in fittings, particularly quick connect fittings. Small openings can have
a severe effect on cooling, and hence on cycle time and productivity.
6. For dies subject to critical surface finish requirements, the flow rates usually need very sensi-
tive adjustments, and the usual pet cocks are not sensitive enough. The best systems are
devices that will measure the actual flow rate, which allows precise adjustments.
7. It is best to control pressure at the machine with a pressure regulator so the pressure at the
die is constant.
8. For good surface finish, the goal is to have an evenly heated die with the die surface tempera-
ture above 400°F (205°C) over the area where good finish is required.

Overflows
Another way to control the die temperature is to add overflows. The overflow helps improve surface
finish in two ways. First, it will provide a place for the first metal through the die to go – this metal
will contain some extra trapped air (and perhaps some lube and flash washed off the die surface), and
so it is desirable that this metal (if possible) be moved into an overflow. Second, and certainly much
more importantly from a good finish point of view, using overflows is an excellent method of spot
heating a die.

Other methods of heating a die are not as specific. Adjusting the cycle time, for example, will adjust
the heat applied, but with a heat distribution determined by the shape of the part. Usually die tem-
perature problems are caused by uneven heating of the die, with the end of flow being a colder area.
A faster cycle time will add heat, but it will be uneven. Very often, there may not be an oil line located
correctly, and it is normally difficult to add one. The solution may be to add large overflows to pro-
vide spot heating just where it is needed on the outer edge of the part.

The size and shape of the overflows are important. For example, it is usually desirable to add heat to
the stationary side of the die, but the tendency is to keep the overflows in the ejector side because of
ejector pin locations. The amount of heat transferred to the stationary die half depends on the part-
ing line surface area and on whether any of the overflow volume is in the stationary half. Thus, adding
surface area at the parting line, or adding an extension (say a half round section) into the stationary
side is often very important in getting real benefits from overflows.

15
Of course, the heat from overflows is not controlled, as with hot oil. Thus, if the overflows are set for a
specific cycle rate, they may add more heat than is desired if the cycle time is reduced, and may even
limit the cycle rate.

The thickness of the gate into the overflow should reasonably thin. Usually it is desirable that the
overflow gate breaks off easily and smoothly, which means it is desirable to keep the gate thin. Also,
this will make it so the metal doesn’t flow heavily into the overflow until the cavity is pressurized,
which is desirable if the intent of the overflow is to remove some flow porosity. However, often the
major function of the overflow is to provide heat, and the time of filling is not as important for this
function. Typical gate opening values would be 0.02 to 0.03 inches (0.5 to 0.75 mm) for zinc and
magnesium, and 0.03 to 0.07 inches (0.75 to 1.75mm) for aluminum.

If the amount of die heating needed for a good finish is difficult of achieve — such as when a small
casting has to be run in a fairly large die — then it is possible to heat the die primarily by overflows.
In this case, the overflows would be likely much larger than the part. This approach needs to be
planned and engineered into the die from the start so there is enough room for the proper size of
overflows. Using hot oil is probably the best method but there may be areas difficult to access for oil
lines and, in this case, overflows can provide a simple and direct heating method.

A die that is too hot will generate other problems, such as heat pits in zinc, and sinks in all metals.
Usually this problem arises from uneven heating, with one hot spot generating sinks, drags, soldering,
or heat spots, while the rest of the die may still be below the ideal temperature and need heat. Adding
heat by reducing cycle time or adding general hot oil will not help in this situation (and may hurt)
because heat is added everywhere, not just where it is needed. Any added heating or cooling must be
done carefully in just the right spots (usually with overflows, spot oil heaters, or via die spray).
Engineering solutions for thermal problems such as hot spots are best achieved through simulation.
Hot areas can be predicted, and the simulation can show what will happen if cooling lines (or heating
lines) are applied at those locations. It is much easier, and much cheaper in the long run, to do the
trial and error for water/oil line locations on a computer before the die is built, rather than suffer slow
cycle rates and high scrap on the shop floor.

METAL FLOW PATTERN — GATING

Many die casters spend most of their process adjustment money and time on changing the gating —
usually this is done to get a better looking casting. There is no doubt that the metal flow pattern is
critical for obtaining castings with good surface finish, but the factors discussed previously (fill time
and die temperature) are equally important, and sometimes a correctly calculated small change (such
as increasing the plunger size) can have a larger effect than a lot of expensive gating changes.

The correct design of gating systems should be done with reference to the correct manuals that de-
scribe that effort, but some factors on the metal flow pattern development will be reviewed here.

Direct Flow
The concept of atomized flow has already been discussed, and it is clear that castings will have a good
finish in the direct flow path in front of the gate, while metal flow that is reflected off a wall will have

16
a poorer finish than that in the direct flow path. Thus, the gate should be located so the direct flow
path is as long as possible.

Figure 10. Very short direct flow path, metal has to reflect several times from the cavity walls to reach
the critical area on top of the casting.

Figure 11. Long direct flow path right in front of gate, the visible area on top should look quite good –
but a disadvantage is that the casting has to have some shape change to get the draft angles needed for
the gate location.

While surface finish is influenced by other things besides direct flow, it is one of the important issues,
and it will be valuable (and worth a fight) if the finish is critical.

Distributed Flow
The best finish comes from a distributed flow. This is true in all metals, and is one of the most impor-
tant factors in developing successful gating that is intended to provide a good surface finish. Distrib-
uted flow means the metal flow should be spread out as much as possible, and especially that it should
cover the critical areas. Distributed flow concepts apply to aluminum and zinc, although it is a more
common practice in zinc, probably because zinc often has higher surface finish requirements.

It is much better to have several gates that spread the flow, rather than a few small gates (often called
chisel gates). The chisel gate does have its place, for example, where it is necessary to feed small po-
rosity sensitive locations, but it is a poor gating design if surface finish is important.

The narrow chisel gate generates jet flow, which causes swirls — and in the center of the swirl it is
common to see cold flow, blisters, and non-fills. The jet usually circles the casting, sealing off the
overflows and trapping extra gas in the casting — most of it in the center of the swirl pattern. The
center of the swirl will also have surface defects like cold flow, non-fills, etc. A chisel gate arrangement
is shown in Figure 12 for a flat plate casting.

17
 Figure 12. “Chisel gate” arrangement.

Figure 13 shows the jetting flow with the swirling that takes place with the small “chisel” gate design.
Each of the swirls captures air, and can be a source of blisters, cold flow, and porosity. It is better for
good surface finish to have a more distributed flow — one that spreads the flow as evenly as possible
over the surface of concern.

Figure 13. Flow in a thin flat plate

with two chisel gates. Note the high-speed jets and
the circular pattern. Flow 3D Simulation.

A distributed flow pattern would be more like that shown in Figures 14 and 15. Figure 14 shows the
gating design, while Figure 15 shows how the flow covers the surface evenly, which is much better for
optimizing surface finish than the jet flow shown in Figure 13.

Figure 14. Distributed gate flow — in the form of

two fan gates.

18
 Figure 15. Distributed flow for a
thin plate using two wide fan gates.
Nova Flow Simulation

The distributed flow can be attained with several gating techniques. One is the fan gate arrangement
shown in Figure 16, the other is the tangential gate arrangement shown in Figure 17.

Figure 16. Multiple fans for a distributed gate.

Figure 17. Tangential gate does a good job of

distributing the flow.

19
Some advantages of tangential gates include:

 It does not require the large steel area outside the cavity that a series of fan gates requires,
because the tapered tangential runner can be run close to the cavities.
 It is easier to use for irregular shapes, as shown in Figure 18. The fan gate can also be used this
way, but it is easier with the tapered tangential. The ramp used to feed the irregular casting
shape can be tapered by 3° towards the runner. The width of this ramp can be approximately
0.5 inches (12.5 mm) for gates in the 0.04 inches (1 mm ) thickness range (or even a little
more), and up to 0.9 inches (23 mm) for a gate in the 0.08 inches (2 mm) thickness range.

Figure 18. Sections in red are “gate ramps” that are equal to the gate thickness at the casting, but ramp
up to the tapered tangential runner at a 3° to 5° slope. The arrows show the approximate flow pattern —
which can be calculated — but it is better to get an accurate flow direction from a flow simulation.

 When using the tangential gate, it is very important that the runner shape be tapered, and it
is far better to calculate this shape using the NADCA procedures and/or software. A tangen-
tial runner without the proper shape can provide some unexpected flow directions and poor
results.
 This kind of gate takes up less room, and can be directed towards a particular area that needs
flow to get a good surface finish. The gate can be segmented, and does not have to continue
unbroken along the edge of the casting. This is necessary when the PQ2 calculations indicate
that the gate area should be limited to a certain amount.

Note that the best design will be one that follows the seven steps listed in the NADCA gating design
handbook. These include defining the quality requirements, defining the process limits, doing the
PQ2 calculations to match the gate and the machine, define the flow pattern, then define the gate
shapes, runner system and finally, the vents and overflows.

OTHER FLOW FACTORS TO CONSIDER

Improving the surface finish could also include adding ribs or wide bands through openings to get
flow to the backside of a window or other opening in the casting. These feeders (or flow bars or
bands) need to be fairly thick to guarantee flow. A thin opening into a feeder will cause the metal to
bypass the feeder until later in the fill sequence. This is because the flow area is not pressurized until
late in the flow sequence, thus the easiest path is for the metal to go around rather than through a

20
feeder with a thin gate. The best approach is to make the feeder as thick as possible – probably a goal
for minimum thickness would be about 0.08 inches (2 mm) inches thick (or more) for aluminum,
and about 0.05 inches (1.25 mm) or 0.06 inches (1.5 mm) thick for zinc or magnesium. Additionally,
overflows in the openings can help mostly by adding heat to a spot that needs it.

Even when the gate cannot be located just where it is desired, there are a number of techniques
that will allow at least some flow to be directed at critical areas. A number of these are shown in the
NADCA Gating Manual, and some others are listed below:

 Design the metal flow path to cross ribs at an angle for a little better flow into the rib section.
 Add small feeders for bosses or extensions that can be reached with a small gate and that
need a good finish.
 Follow the NADCA rules about the area reductions in the runner path, and balance the run-
ners so the gates get the flow they should at the time they should.
 Use shock absorbers for tangential runners.
 If necessary to spread the flow to a critical area, use reflections off a cavity feature to get flow
to the critical area in one or two reflections.
 In zinc, use stippling or roughing on inside die surfaces to get better flow on the critical fin-
ish area

It is worth the extra engineering effort to use these kinds of gate design techniques if surface finish is
an issue. Modern customers will not tolerate the old fashioned trial and error approach (that costs ex-
tra and also shortens die life) — they expect first shot success and rightly believe that it is achievable.

OTHER SURFACE DEFECTS

Finally, it is worth discussing other defects in die castings that can reduce surface finish. These have
been introduced earlier in this workbook and include solder, blistering and surface sinks.

Soldering
Soldering is a mechanical or chemical interaction between the die casting and the tool steel where a
portion of the cast metal adheres to the steel die surface and remains there after ejection. It normally
produces a “scab-like” defect on the surface of the die casting (Figure 19a), along with build-up of cast
metal on the steel die (Figure 19b).

a) b)

Soldered face on Solder build up on a long core(2)

an aluminum die
casting(1)
Figure 19. Photographs of soldering

21
A number of researchers have examined the origin of soldering(3-5), and the mechanism is as follows.
When the casting die is cool, soldering occurs by a mechanical interaction between the casting and
the die steel, primarily in the form of an undercut or some other defect in the shape of the tooling.
However, when the die surface is heated above a critical temperature, chemical interaction occurs
between the cast metal and the die steel, and this is the mechanism more commonly thought of as
soldering. When the liquid aluminum contacts the steel casting mold, the aluminum will diffuse
into the die steel, and the iron from the casting mold will dissolve into the aluminum(3,4), causing the
aluminum and the iron to react with each other. Figure 20 shows a micrograph of a section through
solder, showing that it consists of several layers — the base die steel, an FeAl3 intermetallic phase, and
a layer of aluminum enriched in iron that reaches the Al-FeAl3 eutectic composition. Soldering oc-
curs when the aluminum eutectic cools and solidifies, causing the aluminum casting to stick (solder)
to the casting mold.

Eutectic
Aluminum Figure 20. Cross-section showing the differ-
ent layers present when an aluminum casting
solders to a steel die(4)
FeAl3

Steel

A number of methods have been identified to minimize soldering:

 It is well know that soldering usually occurs when the steel die become too hot, and a critical
temperature has identified below which chemical soldering of an aluminum casting to a steel
die will not occur(4). The magnitude of this critical temperature is dependent upon a number
factors, but essentially anything that will cool the die will help prevent soldering. Therefore,
improved application of internal water cooling, external application of cooling via die spray
(or just simply water) can be used to minimize soldering.
 Other factors, such a reduced injection velocity and change of lubricant, can also reduce sol-
dering.
 Alternate die materials, such as tungsten alloys, slow lower propensity for soldering, and so
replacement of die components in troublesome areas (such as core pins) with tungsten-alloy
parts can also reduce soldering(6).
 Nitriding of the die surface has also been shown to reduce soldering(7,8).
 But probably the most effective method to reduce soldering is to apply a coating to the sur-
face of the die. Normally this is not applied to the entire cavity, but just to the troublesome
components such as core pins or die inserts that get very hot and exhibit soldering. Exten-
sive research on coatings has been performed over the past 20 years(9-11), but coatings today

22
 normally are applied by Physical Vapor Deposition (PVD) techniques, and commonly used
coatings include CrN(6), TiAlN(12) and AlCrN(13).

More information about avoiding solder can be found in NADCA’s textbook on the subject(1).

Blisters
Surface blistering is a problem unique to casting processes that utilize high pressures, such as die,
squeeze and semi-solid casting(14). With conventional die casting blistering most commonly occurs
directly after ejection from the die casting die, while with squeeze and semi-solid casting blistering
is normally only a problem during subsequent T6 heat treatment (although as the heat treatment of
conventional die castings is becoming more common, blistering of conventional die castings during
T6 heat treatment may also be something the reader should be aware of). Photographs of blisters are
shown in Figure 21(15).

a) b)

Typical appearance on as-cast surface Cross-section through a blister

Figure 21. Photographs of blisters(15)

The mechanism of blister formation is as follows(14). During casting it is possible to trap defects such
as air or lubricants below the casting’s surface (Figure 22a). During the die casting process, high
intensification pressures are applied to the liquid metal as it is solidifying, and so the defects are also
compressed to these high pressures. While the casting is within the steel die, the die surface resists
the pressure from the gas defect, but if the casting is ejected from the die at a high temperature where
the cast metal is still soft, the strength of the casting may not be sufficient to resist the pressure from
the internal pressure from the defect, and a surface blister is formed (Figure 22b).

The best method of eliminating blistering during conventional die casting is minimize the amount
of trapped air, lubricants, or water (see NADCA’s book on gas porosity(16)). However, the blisters can
also be hidden by increasing the solidification time, to allow the castings to cool to a lower tempera-
ture while still supported by the die casting die, thereby strengthening the cast metal and allowing it
to better resist the high internal pressure from sub-surface defects.

23
Figure 22. Schematic showing the mechanism for blister formation

Surface Sinks
Sinks are surface defects that are related to sub-surface shrinkage porosity. Sinks occur when the
interior of the metal shrinks during solidification, and the surface layer is unable to withstand the
stresses developed in the interior of the casting, causing the surface to deform (sink) towards the inte-
rior. Sink is generally most visible on large, flat surfaces, and examples of sink are shown in Figure 23(15).

a) b)

Surface appearance (Mag ~ 1x) Fracture through a surface sink – note that
the shrinkage porosity under the skin is
visible (Mag ~ 5x)

Figure 23. Photographs of surface sink in die castings(15)

Methods for preventing or correcting sink are listed in Reference 15, but essentially consist of mini-
mizing (or moving) the underlying shrinkage porosity.

24
SUMMARY
In summary, a designer should think about the following concepts when surfaces finish is paramount:

1. Do the fill time calculation, or use good experience to develop a reliable fill time estimate.
2. Do the PQ2 calculations to develop the proper gate area.
3. Make sure the selected machine will give the fill time required.
4. Early in the design process, establish an estimated die temperature with calculations or simu-
lation, and make provisions (hot oil, cartridge heaters, etc.) to make sure the die temperature
is above the minimum.
5. Review the process operating conditions to make sure there are adequate heating/cooling
fluid flow controls, good spray equipment and methodology, and good operating procedures
— these are needed to run a process that will generate good surface finish.
6. Use distributed gate techniques to design a gate that will provide as much direct flow to the
critical areas as possible.
7. The designer and trouble-shooter should always keep track of the 3 basic areas — proper fill
time, good die temperatures, and good distributed gating flow patterns — and success will be
as easy as it can get.
8. Address areas of the casting where solder, blistering or sinks are a problem.

REFERENCES
1. John F. Wallace, A Guide to Correcting Solder, NADCA publication number 527, 2006
2. John Moore, Sterling Myers, Jianliang Lin, Brajendra Mishra, Fengli Wang & Peter Ried,
“Effect of Nitriding and PVD Coating on Soldering, Washout and Thermal Fatigue Cracking
Behavior”, NADCA 2008 Congress, Paper number T08-092
3. Sumanth Shankar and Diran Apelian, “Die Soldering: Mechanism of the Interface Reaction
between Molten Aluminum Alloy and Tool Steel”, Metallurgical and Materials Transactions
B Vol. 33B, June 2002, page 465
4. Q. Han & S. Viswanathan, “Analysis of the Mechanism of Die Soldering in Aluminum Die
Casting”, Metallurgical and Materials Transactions A Vol. 34A, January 2003, page 139
5. Hans I. Laukli, Alessandro Graciotti, Otto Lohne, Haavard Gjestland & Stian Sannes, “Mech-
anism of Die Soldering During Aluminum Die Casting”, NADCA 2002 Congress, Paper
number T02-034
6. Eduardo K. Tentardini, Augusto O. Kunrath, Cesar Aguzzoli, Maria Castro, John J. Moore &
Israel J.R. Baumvol, “Soldering mechanisms in materials and coatings for aluminum die cast-
ing”, Surface & Coatings Technology Vol. 202 (2008) page 3764
7. K.A. Nazari & S.G. Shabestari, “Effect of micro alloying elements on the interfacial reactions
between molten aluminum alloy and tool steel’, Journal of Alloys and Compounds Vol. 478
(2009) page 523
8. Y. Zhu, D. Schwam & J.F. Wallace, “Effect of Nitriding and PVD Coating on Soldering,
Washout and Thermal Fatigue Cracking Behavior”, NADCA 2008 Congress, Paper number
08-092
9. V. Joshi, K. Kulkarni, R. Shivpuri, R.S. Bhattacharya, S.J. Dikshit & D. Bhat, ”Dissolution
and soldering behavior of nitrided hot working steel with multilayer LAFAD PVD coatings”,
Surface and Coatings Technology Vols. 146–147 (2001) page 338

25
 10. S. Gulizia, M.Z. Jahedi & E.D. Doyle, “Performance Evaluation of PVD Coatings for High
Pressure Die Casting”, Surface and Coatings Tech., Vol. 140, (2001), page 200
11. Mulong Yu, Rajiv Shivpuri, Robert A. Rapp, “A Diffusion Chromium Carbide Coating for
H-13 Die Steels in Aluminum Die Casting: Coating Development and Evaluation”, Ohio
State University ERC for Net-Shape Manufacturing report number ERC-95-02
12. Q. Han, H. Xu, P.P. Ried & P. Olson, “Accelerated Method for Testing Soldering Tendency of
Core Pins“, International Journal of Cast Metals Research
13. Z. Gay, S. Knickel, R. Donahue, V. Khominich, D. Bell, “Case Study of Structural Aluminum
Casting Operations Using CertiPhy Plus Coatings”, NADCA 2014 Congress, Paper number
T14-061
14. Steve Midson, “Minimizing Blistering During T6 Heat Treating of Semi-Solid Castings,” Die
Casting Engineer, 2011
15. William G. Walkington, Die Casting Defects: Troubleshooting Guide, Pub: North American
Die Casting Association, 2003
16. William Walkington & Steve Midson, Gas Porosity: A Guide to Correcting the Problem, Pub:
North American Die Casting Association, 2014

26
3250 N. Arlington Heights Rd., Ste. 101
Arlington Heights, IL 60004
T: 847.279.0001 • F: 847.279.0002
www.diecasting.org