Thin Walled Investment Castings
Swerea SWECAST AB
Box 2033, 550 02 Jönköping, Sweden Tel: +46(0)36 - 30 12 00
swecast@swerea.se http://www.swecast.se
1867 LEAN - Development of light-weight steel castings for efficient aircraft engines
Author Report number: Date
Roger Svenningsson & Sten Farre 2014-002 2014-02-21
Summary
This review is presented as a part of the literature review performed in the Clean Sky project LEAN - Development of light-weight steel castings for efficient aircraft engines, coordinated by Swerea SWECAST during 2011-2013.
From a foundryman’s point of view, fluidity is related to the ability for a metal to fill a mould while fluidity is related to viscosity in physics. A distinction is made between flowability and fillability depending on which mechanism is reducing the ability to flow. Fillability is related to flow problems when surface tension is the governing mechanism and flowability is related to heat transfer. The influence of surface tension will be more important as the thickness of the casting is decreased whereas heat transfer has a more pronounced influence on thicker sections, but still thin. Several parameters related to both the process and alloy affect the ability to fill a thin section. Increasing super heat, mould temperature or pressure head will improve fluidity. Changes in alloy composition will also to some extent have impact on fluidity. Flow rate and shell permeability is also important parameters. There are also other possibilities that can im-prove fluidity. Vacuum-assisted casting is one method which has been im-proven effective. This method is used both for gravity and counter gravity casting. Vibrating is yet an other method which improves fluidity.
Sammanfattning
Denna skrift är ett utdrag ur den litteraturstudie som genomfördes i Clean Sky projektet LEAN -Development of light-weight steel castings for efficient aircraft engines, koordinerat av Swerea SWECAST under 2011-2013.
Begreppet flytbarhet inom gjutningen skiljer sig åt mot fysikens definition. För en fysiker hand-lar det om den fysikaliska egenskapen viskositet, medan det för en gjutare endast är förmågan att fylla en form under vissa givna förutsättningar. De mekanismer som kontrollerar flytbarheten går att dela in i de som är termiskt drivna samt de som är relaterade till problemen att kompensera för de som drivs av ytspänningen. Inom vaxursmälningsprocessen är det ett antal olika param-eterar relaterade till processen och legering med en direkt koppling till förmågan att fylla tunna sektioner. En ökning av gjuttemperatur, skaltemperatur eller tryckhöjd, hastighet samt skalets permeabilitet kommer att påverka flytbarheten. Också ändringar i sammansättning av materialet kan också påverka flytbarheten. Vidare finns det speciella metoder som visat sig effektiva för att kunna gjuta tunnare sektioner. Assisterad vakuumgjutning samt att använda vibrationer under gjutningen har visat sig mycket effektiva.
Table of Contens
1 The investment casting process - a brief description 1
2 Fluidity - general considerations 1
3 Parameters in Investment Casting 3
3.1 Parameters related to the shell system . . . 4
3.1.1 Mould temperature . . . 4
3.1.2 Metal head . . . 4
3.1.3 Shell thickness . . . 5
3.1.4 Miscellaneous mould variables . . . 5
3.2 Parameters related to pouring . . . 6
3.2.1 Vacuum assisted casting . . . 7
3.2.2 Vibration assisted casting . . . 9
3.3 Parameters related to the alloy composition . . . 9
3.4 Miscellaneous alloy variables . . . 10
1 The investment casting process - a brief description
Present literature survey focus on thin walled investment castings and the parameters con-trolling filling of narrow sections. A general description of the investment casting process is described in Figure 1 and consists on following sequence:
• First a metal die is manufactured, normally in aluminium, to outline the geometry of the component. Thereafter the pattern is made by injecting liquid wax into the die. • The desired numbers of patterns are assembled by attaching the patterns to the sprue,
making a replica of the casting.
• Then a continuously dipping and stuccoing is performed until the desired thickness of the shell is achieved. Different slurries and stuccoing material is used during this sequence controlling the properties of the shell. Also, in the prime layer, some addi-tional element can be added to aid nucleation.
• Next step is to cure the shell. This is done after the wax is melted out from the tree in an autoclave to secure that all wax has been removed from the mould, which is subsequent cured to create a hard ceramic shell.
• The mould is preheated to 1000-1250◦C before pouring. Pouring of steel is normally
done at extra high temperatures.
• After pouring and subsequent solidification and cooling, the shell and ingate system is removed.
• Final castings are removed to be cleaned and in most cases also heat treated in various ways to obtain proper mechanical properties.
2 Fluidity - general considerations
To a person who works in a foundry, fluidity is of an empirical nature which depends on the casting piece and casting parameter configuration and is defined as simply as the ability of molten metal to flow and fill a mould, [1,2]. This definition is not the same as in physics, where fluidity is related to the viscosity of the fluid. Ordinary fluidity tests are usually per-formed by measuring the length of metal flow in a spiral or some straight horizontal design under various predefined conditions. Even though these methods succeed in measuring flow distance under different condition like superheat, changes in metal head and mould temper-ature. These methods have its limitations as they fail in the ability to study flow problems governed by surface tension. Therefore a separation of the overall term fluidity is made based on two different aspects of fluidity, namely flowability and fillability [1]. Flowability is related to heat transfer and fillability to the ability to fill really thin narrow section, e.g. small radius [3]. Flowability and fillability both act together in thin walled investment cast-ings. The term fluidity will be used as a term incorporating both flowability and fillability effects. Flowability and fillability will be used whenever a particular parameter will have influence on either phenomenon.
Swerea SWECAST AB Report no: 2014-002
Figure 1: Principle route of production for investment casting.
Thin walled investment castings have been a research topic since the fifties and some an-alytical expressions have been derived for fluidity. Lun et al. [4] present an analytical expression of fluidity according to Equation 2.1, based on the original work by Flemings et al. [5] as: Lf = ρV0t(kHf+CpTS) 2h(Tc− Tm) " 1 +h 2 π 4X V0C 0 Pk 0 ρ0 1/2# (2.1)
Where:
ρ =⇒ is the density of liquid,
t =⇒ is the thickness (t«w, where w is the width), V0 =⇒ is the velocity of liquid at entrance to section, k =⇒ is some critical fraction of solidified alloy, Hf =⇒ is the enthalpy of fusion,
Cp =⇒ is the specific heat of the liquid, TS =⇒ is the superheat defined as (Tc− Tl), Tc =⇒ is the temperature of metal at pouring, Tl =⇒ is the temperature of liquidus,
h =⇒ is the heat transfer coefficient at the metal-mould interface, Tm =⇒ is the initial mould temperature,
4X =⇒ is the length of choking zone in alloy solidification, k0 =⇒ is the thermal conductivity of the mould,
ρ0 =⇒ is the density of the mould material, C0P =⇒ is the specific heat of the mould material.
NOTE: This expression gives some ideas only on the parameters which have a directly im-pact on flowability, not on fillability.
Equation 2.1 states that flowability is a function of properties related to both alloy and shell system. Increasing the heat content of the alloy at the same time as improving the insulation of the mould and thereby retaining the heat in the system will increase fluidity. It can also be noticed that flowability is directly proportional to the thickness of the casting and the heat transfer coefficient.
Fillability is related to the ability of filling small cross sections or small features and is due to the difference in actual metal pressure and surface tension of the metal. In [1] a sec-tion thickness of 1.0 mm is mensec-tioned as the threshold when fillability becomes important. Flemings [2], often cited in papers involving solidification and thin castings, gives an ap-proximate thickness of 2.5 mm as the limit when surface tension may become important. Even though there is no exact limit when the influence of surface tension becomes impor-tant, its contribution to a decrease in the ability to fill increases as the section thickness decreases.
3 Parameters in Investment Casting
In this review an overall description of the different parameters related to thin walled invest-ment castings will be given without considering different material groups or alloys. Also, different process variables that effect both fillability and flowability in investment casting will follow the same classification as in [1] and are according to:
• Parameters related to the shell system • Parameters related to the pouring
Swerea SWECAST AB Report no: 2014-002
3.1 Parameters related to the shell system 3.1.1 Mould temperature
In investment casting pre-heated moulds are used and it is generally considered that an increase in mould temperature will have a positive impact on fluidity. As can be seen in Equation 2.1 flowability is dependent on both metal temperature (TC) and the mould
tem-perature (TM) as well as many additional parameters. In investment casting the full range
of mould temperature can be used. But for fillability an increase in mould temperature will hardly have any effect [1]. Walker [6] and Ollif et al. [7] found an increase in fluidity with an increase of mould temperature when casting an austenitic stainless steel blade in air. Ca-padona’s review on fluidity [1] presents an interesting work by J. Campell et al. for a nickel based alloy. They showed that when using the same geometry as previous authors, casting in vacuum a mould temperature of 400◦C or higher did not affect the fluidity. The maximum thickness that was used was 1.3 mm and the minimum was 0.6 mm. One explanation might be that fillability is the controlling mechanism for the fluidity. At lower temperatures heat is removed quickly because of the temperature difference resulting in very rapid freezing and at higher mould temperatures surface tension is the main reason for fluididty. This can also be confirmed by the work done by [7] in which the effect of mould temperature becomes lesser at the same time as the thickness gets smaller, especially when the flow rate is low. Chandraseckariah et al. reports in [8], when casting a thin blade like geometry in a Nickel based alloy. The influence of the mould temperature on fluidity showed that the fluidity decreased in the temperature interval 600-750 ◦C. A further increase in mould tempera-ture increased the fluidity. However, the maximum temperatempera-ture of the mould was 900◦C. Maguire et al. [9], performed experiments to find out how super heat of the melt and mould temperature influenced the filling length of horizontal oriented thin wide plates in 17-4PH stainless steel. They found out, surprisingly, that the mould temperature had no effect on the fluidity. They varied the mould temperature between 870-1150◦C, a temperature range that must be regarded as relevant for investment casting of stainless steel. They explain that this result probably was related to the in-gate dimensions, which was held constant during all experiments to a thickness of 1.5 mm. When pouring at low temperature there was not sufficient heat left in the metal to enter the gate and freezing began all ready at the in-gate. But as pouring temperature was increased the heat in the system was sufficient to enter the in-gate and mould filling was observed.
3.1.2 Metal head
Both flowability and fillability is affected by the metal head, or pressure of metal column. (PM), defined in Equation 3.1. The metal head is defined as the product of density of the
Flowability and fillability can be improved due to an increased velocity and the effect of increased pressure height respectively. Following Campbell [3] and assuming a non-wetting interface between the melt and the mould the expression according to Equation 3.2 needs to be fulfilled if the liquid metal should have a chance to enter a thin section.
ρ gh − PMould >
γ
t (3.2)
Where:
ρ =⇒ is the density of liquid,
g =⇒ is the acceleration due to gravity,
h =⇒ is the pressure height,
PMould =⇒ is the back pressure of gases inside the mould,
γ =⇒ is the surface tension,
t= 2r =⇒ is the thickness of the plate.
As can be seen in Equation 3.2 the back pressure in the mould will reduce the effective metal head. This implies that venting or using a mould with good permeability is vital for an improved filling. Casting in vacuum chamber will help to some extent and may increase fluidity because of a reduction in back pressure due to evacuating gasses before pouring [8]. A linear relation between metal head and fluidity is reported in literature [1, 4] and references within.
3.1.3 Shell thickness
Chandraseckariah et al. [8] included in their work a study on the influence of the number of shell layers of the mould in relation to fluidity. A blade like geometry was cast under vacuum with a Nickel based alloy. A non-linear variation of fluidity was observed with the number of shell layers. With an increasing number of layers (7-8) the fluidity decreased, but with further increase of the shell layers (8-9) the fluidity increased again. This behaviour was repeated regardless of mould and pouring temperature at a vacuum level of 0.1 Torr. 3.1.4 Miscellaneous mould variables
As flowability is governed by heat transfer the interfacial heat transfer coefficient (IHTC) will have impact on the fluidity as seen in Equation 2.1. In [1] it is reported that the flowabil-ity will increase in proportion to 1/IHTC. The interfacial heat transfer coefficient is affected by the surface roughness of the mould. If a flat smooth surface is considered, the entire metal surface will be in contact with the mould surface and a large IHTC is to be expected. But if the surface is rougher then contact between the surfaces only take place at the asper-ities tops, leaving air in the valleys and a lower IHTC is expected. The drawback of using this as a method in improving fluidity is of course the rougher surface of the casting [10,11]. This behaviour is also noticed by Campell for casting horizontal placed plates [12].
Thermal diffusivity, α, is a property which is defined according to Equation 3.3.
α = k
ρCP
Swerea SWECAST AB Report no: 2014-002
Where:
k =⇒ is the conductivity, ρ =⇒ is the density, Cp =⇒ is the specific heat.
Thermal diffusivity describes the ability of a material to transfer heat. Therefore reducing the diffusivity and thereby increasing the insulation effect of the mould material will in-crease the flowability [1]. In [11] it is argued that fused silica is better for thin castings than a shell made up of alumino-silicate due to better thermal properties. The heat capacity of fused silica is 78% of zircon, which is often used in the prime layers. Snow also refers to a work done at Rolls Royce in which results showed that filling a thin section was best achieved by using fused silica as prime layer, followed by alumino-silicate, zircon and alu-mina [10]. It is important for any gas (air) in the mould to be evacuated during the fill to ensure a complete filling. The permeability of the shell controls the ability of the mould to evacuate the gas. The prime coating in investment mould is much denser than the middle and outer layers and has therefore lower permeability. Brezian and Kondic [13] showed that the time to evacuate a given amount of air was doubled when using two prime layers instead of one. Capadona [1] reports that the back pressure of the mould is proportional to 1/K, where K is the permeability. In [11] the permeability is presented for fused silica and alumino-silicate and it can be concluded that fused silica has higher permeability than alumino-silicate.
3.2 Parameters related to pouring
By increasing the flow rate during filling extended flowability can be achieved. Capadona reports in [1] that some researches achieved a 70 % improved fluidity by increasing the flow rate from 100 g/s to 300 g/s using a preheated mould of 970 ◦C when casting a thin blade geometry. The blade was 2.5mm in its thickest part. When the mould temperature was increased to 1150◦C they only achieved an increase in fluidity of 12%. This range of flow rate is however probably low comparing to a real casting situation. I order to be able to fill a thin section the speed at the entrance needs to be high. If the speed is low premature freezing may occur. Therefore the design of the entrance is very important. Studying the parameters for horizontal placed castings in relation to the sprue according to Equation 3.4 and 3.5 derived by Flemings [14] give us:
V = 2g(Z − y) 1 + K00
1/2
(3.4) Where:
As can be seen, in order to maximize the velocity, K00 needs to be small, i.e. the radius should be large. Also we have from Flemings [14],
y= 4γ Rρg (3.5) Where:
y =⇒ is the decrease in effective metal head due to surface tension, γ =⇒ is surface tension,
R =⇒ is the curvatures radius, ρ =⇒ is is the density,
g =⇒ is the acceleration due to gravity.
This relation implies that surface tension, γ, is an important factor affecting the velocity. The drawback when increasing the velocity is that turbulence is expected to be higher and caus-ing defects in the castcaus-ing. Therefore it is of great importance that the turbulence is kept to a minimum.
3.2.1 Vacuum assisted casting
By vacuum assisted casting, which is not to be confused with casting in a vacuum chamber, the intention is to create a pressure difference that will aid the metal flow by removing the gases which is inside the mould and creating a “drawing effect” of the melt. There are two different methods involved depending on if the melt is poured in the direction of gravity or drawn upwards in the opposite direction of gravity. Use of assisted vacuum or partial vacuum as it is in reality, is used by both methods, which have proven increase the fluidity of thin walled castings. Assisted vacuum casting in the direction of gravity can be illustrated by Figure 2. The casting can also be inside a perforated flask or similar as in [4]. This package is lowered into a sealed vacuum chamber. The air is withdrawn as the metal is being poured. Using this method the atmospheric pressure can act on the liquid metal via the runner system with zero (or little) resistant from the gases in the mould which will increase flowability. Fillability is also improved by the effect of suction [1]. The partial vacuum is held during the entire solidification. Sin et al. [4] demon-strated the effect of vacuum assist casting magnesium spirals with a thickness of 1.6 mm. The casting without assistant from vacuum had non complete fill, whilst the casting with assistant is completely filled. Vainer et al., [15], used the aid of partial vacuum during casting of an Al-Ni-Bronze alloy. In this work they used a perforated flask in which they varied the open area. They also varied the vacuum pressure in order to investigate how these parameters influenced fluidity. They used spirals with different thickness (2 and 3.8 mm). The findings from this investigation are that an increase in fluidity could be observed if the open area of the perforated shield was increased. Three times better fluidity was observed when the active vacuum area was increased five times. Also an improved fluidity for the test spirals was observed, 1.2 to 1.5 times, when the pump efficiency was increased 1.4 times. Matsuna et al. [16] choose assisted vacuum to cast a thin walled impeller with the alloy 17-4PH. In this work they report a castable thickness of 0.9 mm for the thin blades of the impeller using this method. In their work they de-pressurized the mould to 200 mm Hg during casting. The total weight of the casting was approximately 22 kg.
Swerea SWECAST AB Report no: 2014-002
Vacuum Sealing
Figure 2: Schematic principle of assisted vacuum casting.
Capadona et al. reports in [1] that the position of the vacuum channel should be placed approx-imately two thirds up the side of the flask. If the vacuum channel is placed at the bottom of the chamber there is a risk that the effect of vacuum will be at the place where it is not needed, i.e. at the bottom of the mould where the pressure head is working at its maximum effect. By placing the channel at a higher point in relation to the flask the vacuum will have a greater effect where it is most needed.
Counter gravity processes are frequently used in sand, permanent mould and investment casting using either a pressure or vacuum driven filling of the cavity. The difference in counter gravity processes is that the melt is drawn upwards due to an external applied pressure difference. All counter gravity methods have the advantage that a controlled filling can be achieved, i.e. min-imizing turbulence, minmin-imizing defects caused by oxides and a cost save due to a better yield because the liquid metal in the filling pipe will return back to the metal basin after the vacuum is released. Also very thin shells can be used if the shell is supported with dry permeable sand in the shell container.
The Hitchiner process is a method for counter gravity casting for the investment casting industry, patented by Hitchiner Manufacturing Co., Inc. Milford [17]. In this method the alloy is melted
3.2.2 Vibration assisted casting
Using vibration has proven to be effective to increase fluidity of thin walled investment casting. A. Karem et al. [19] investigated the influence of vibration casting Aluminium alloy A356. Figure 63 shows that vibration has a positive effect off the fluidity, especially at higher pouring temperatures. Flemings [5] explains the increase in fluidity as a result of an increase in metal head due to an extra acceleration.
600 620 640 660 680 700 720 740 760 780 800 40 50 60 70 80 90 100 110 120 130 140 Temperature,◦C Fluidit y, mm Vibration No vibration
Figure 3: Illustrates the effect off using vibration during casting, reproduced from [19].
Note: Sten Farre do not agree with the theory of acceleration, but rather the fact that vibration act as a divider of initial cells that have started to link together in the melt. During vibration the bond between to neighbouring nucleation is broken and the melt will contain a lower viscosity. At lower temperatures the bond becomes stronger and thus the vibration will not be able to break the cells into smaller parts. To some extent it is believed that the surface of the oxide is broken and therefore gives the melt a lower viscosity.
3.3 Parameters related to the alloy composition
The influence on fluidity with an increase in pouring temperature is well documented in liter-ature. Equation 2.1 states that there is a direct relation between fluidity and super heat. Most authors see a linear relation between fluidity and an increase in pouring temperature. But in [1] it is stated that over a certain critical temperature, an increased superheat of the melt will not improve fillability. Above the critical temperature the interaction between metal head and surface tension will determine how the section will be filled. This behaviour is also illustrated in Figure 4. As can be seen at small thicknesses the influence of super heat is negligible at any given superheat indicating that surface tension is the limiting factor. These tests were performed in sand moulds but the arguments are probably also valid for investment castings.
Swerea SWECAST AB Report no: 2014-002 0 1 2 3 4 5 6 7 8 9 10 0 100 200 300 400 500 600 Thickness, mm Fluidit y, mm 1470◦C 1520◦C 1580◦C
Figure 4: Illustrates the effect off fluidity for sand cast 17-4PH, reproduced from [3].
3.4 Miscellaneous alloy variables
Changing the alloy composition has an impact on fluidity. The most obvious variables in the thermo-physical properties of the alloy to change are to increase specific heat, density, and heat of fusion according to Equation 2.1. Also surface tension or viscosity can change by chang-ing the alloy composition. As the temperature of the alloy goes towards liquidus the viscosity increases, therefore it is important to have enough super-heating before pouring.
4 Test Geometries used in literature
Different test geometries to measure fluidity are presented in the literature. Following the dis-tinction stated before test geometries needs to be able to fill the requirements for both flowability and fillability. The most important aspect when choosing test geometry to study fluidity is if the test piece will give the expected answers. Therefore choosing test geometry is not a trivial task depending on what to be studied. In figure 5 is a test geometry showed for the aero industry. This casting resembles to a very high degree a turbine blade. The geometry is designed to fulfil the requirements for both fillability and flowability. The outer radius approaches zero and fulfils the requirements for fillability. Because of the curvature the middle section is wider and at the widest point (in the middle) the thickness is defined. This part of the geometry is used to study flowability.
Figure 5: Illustrates a geometry used for fluidity tests in the areo industry, reproduced from [8].
Ordinary spiral, horizontal, or vertical plates have also been used for fluidity tests. Special ar-rangements where vacuum is used to “suck” up metal in a tube and measure the flow distance. Examples on geometries are presented in [1]. But if both fillability and flowability is to be stud-ied at the same time care must be taken when choosing the test geometry so both requirements are fulfilled.
Swerea SWECAST AB Report no: 2014-002
References
[1] J. Capadona, D. Albright, Review of Fluidity Testing as Applied to Lost-Polystyrene In-vestment Castings, AFS Transactions 78-12 (1978) 43–54.
[2] M. C. Flemmings, Solidification Processing, McGRAW - HILL, 1974. [3] J. Campell, Castings, Butterworth-Heinemann, 1993.
[4] S. Lun, Sin, D. Dubé, Influence of process parameters on fluidity of investment-cast AZ91D magnesium alloy, Material Science and Engineering A386 (2004) 34–42.
[5] M. Flemings, F. Mollard, E. Niiyama, Fluidity of Aluminium, AFS Transactions 70 (1962) 1029–1039.
[6] B. Walker, New type of Fluidity Test Piece, Foundry Trade Journal 115 (1963) 713–721. [7] I. Oliff, R. Lumby, V. Kondic, Some new ideas on fluidity test mold design, Foundry 93
(1965) 80–85.
[8] H. Chandraseckariah, S. Seshan, Effect of Foundry Variables on Fluidity of Investment Cast Nickel-Base Super alloys, Transactions of the American Foundry Society 110 (2002) 681–695.
[9] M. Magurie, F. Zanner, Gating Geometry Studies of Thin-Walled 17-4PH Investment Cast-ings, in: 40th annual Investment Casting Institute (ICI) technical meeting, 1992.
[10] J. Snow, Why Fused Silica?, in: 57th Technical Conference & Expo, Investment Casting Institute, 2010.
[11] J. Snow, D. Scott, Comparing Fused Silica and Alumino-Silicate Investment Refractories, Modern Castings jan (2001) 45–47.
[12] J. Campell, Thin walled castings, Material Science and Technology 4 (1998) 194–204. [13] M. Brezina, V. Kondic, Flow phenomena in investment casting, The British Foundryman
66 (1973) 337–348.
[14] M. Flemings, Fluidity of Metals-Techniques for Producing Ultra-Thin Sections Castings, 30th International Foundry Congress - (1963) 61–81.
[15] M. Vainer, Y. Lerner, Vacuum-Assisted Investment Casting of Al-Ni-Bronze, Transactions of the American Foundrymen’s Society 107 (1999) 35–42.
[16] K. Matsuna, S. Ohama, Application of the Investment Casting Method for Extremely Thin Blade Impellers, Ishikawajima-Harima Engineering 21 (1981) 45–50.
Transac-Acknowledgement
