Automobile brake drum is part of the braking system that make contact with other parts for as many times as the brake is applied when the automobile is in motion. These contacts results in noise (squeals), frictional wear and fatigue failure of the drum. A lot of research and development have gone into the production of the brake drum in order to ensure that they perform optimally in service and that they not fail prematurely. Work is still going on in order to identify and reduce the noise level generated during braking (Lee et al., 2001; Fieldhouse et al. (2002a and b, 2003). Thermal fatigue cracks are also found to occur frequently on brake drums surface, a phenomenon that leads to premature failure (Zhou et al., 2006; Heidrov, 2002; Buni et al., 2004; Cho et al., 2003).
The automobile brake drum is normally made from GL 250 cast iron on ASTM A-247
specification or the SAE standards. This specification is similar to DIN 1671
specification as presented in Table 1.
Under the standard charts available from American society for testing and materials (ASTM) specification A-247, the form of graphite should be I while the order and size should be A4-A7. Flake graphite. The graphite distribution should be uniform with an apparent random orientation. Type C is totally prohibited because of the coarse Kish graphite (Walton and Opar, 1981), which is undesirable for auto-mechanical application.
It is also specified by Peugeot Automobile Nigeria Limited (PAN) that the matrix
must be lamellar pearlite with ferrite content in the core be less or equal
to 10% and the percentage of free carbide be less or equal to 2%.
Mechanical properties specifications for the brake Drum are
||Hardness 220±95 BHN
||Tensile strength greater or equal to 250 N/mm2.
||Cracks of any sorts must be avoided.
Obtaining a brake drum conforming to above standards is however dependent on the following:
||Composition of the molten metal and hence on the scrap metals
from which the melt is produced.
||Inoculating technique used during casting
||Mould material properties
||Excellent casting practice.
Of highest importance in the composition of the molten metal are carbon and
silicon. Silicon is necessary for the formation of graphite.It also imparts
corrosion and elevated temperature oxidation resistance to the drum.
A large proportion of carbon in the automobile brake drum is present as graphite,
which has little strength or hardness.
||Specification of brake drum (Courtesy of PAN*)
|*PAN stands for peugeot automoblie of Nigeria
||Typical properties of moulding sand for casting brake drum
(courtesy of PAN)
The effect of carbon, silicon and phosphorus on the tensile properties of brake
drum combined into a number called Carbon Equivalent Value (CEV). This CEV is
expressed mathematically as:
C.E.V. = total % carbon+(Si %+P %)/3.
Inoculating reduces the size of the eutectic cell and the span of the graphite
within each cell, thereby improving the strength of the casting (Krause, 1969).
Rare earths, calcium, aluminum, barium and strontium are often used as active
elements inoculants. Most commonly used is ferro-silicon. They have an important
effect on the formation of nuclei of graphite during solidification, thus influencing
the resulting structure (Ziegler, 1964) and strength of casting. Inoculating
also promotes the formation of type A graphite. Mould material are also very
important in obtaining the correct microstructure and hence the right properties
in a casting. Mould materials affect the heat transfer characteristics during
solidification. Sand composition requirements to produce defect-free casting
of the brake drum as recommend by PAN are as shown in Table 2.
Shake-out time is of very great importance in determining the pearlite-ferrite distribution and carbide formation. The longer the shakeout time, the more the pearlite-ferrite distribution and the fewer the carbide formation (Jain, 1997). The shorter the shake-out time the finer the grain size for the grains will not have enough time to distribute themselves and consequently the harder the casting (Flinn, 1963).
The objective of this study is to develop a procedure for obtaining the properties
in a brake drum (as specified by PAN) from grey cast iron using sand casting.
MATERIALS AND METHODS
Moulding sand preparation and testing: In order to meet PAN specification
for sand mould properties (Table 2), sands from two sources
were used. The sources of the sands and the proportion in which they were mixed
are as shown in Table 3.
The proportion of 70% Igbokoda sand and 30% Bacita sand was found suitable and used throughout for the experiment. The choice of this mixture was based on the various test measured properties shown in the table.
Charge calculation and melting of charge: In order to meet the specified
composition of metal from which automobile brake drum is produced, analysis
of various charge materials were done using spectrometer and typical charge
composition of charge materials is shown in Table 4.
Charge calculation was done based on the above composition. After melting,
spectrometric analysis was done and the two results are nearly the same as shown
in Table 5.
Melting is done in a coreless induction furnace of 250 kg capacity and power rating of 250 kw/1000 Hz.
Innoculating techniques: The inoculant used was ferrosilicon. This was introduced into the ladle just before tapping of the molten metal. One Hundred gram inoculant was used on every 250 kg ladle.
Post-casting treatment: The castings were left in the mould for various times before they were removed. These constituted the shake-out time. It varied from 12 min to 10 h.
||Sources of sand and their mixing proportion
||Specified composition of metals
||Spectrometric analysis of charge
Metallographic examination: Specimens were prepared for micro-structural
examination using standard method of preparation. Polished and etched specimens
were observed under the microscope at x100 magnification to assess the following:
||The microstructure developed in the casting under various
||The type of carbide and graphite flakes formed.
||The grain size of the carbide. The linear intercept method of grain size
measurement was used.
Mechanical properties: Hardness measurement were carried out on the specimens from which the tensile strength were calculated using Hardness conversion Table (DeGarmo et al., 1997).
The result of the experiment and various measurements carried out on the casting
are as shown in Fig. 1 to 2.
|| Graph of shake-out time vs grain size of carbide
|| Graph of shake-out time vs hardness
Multiple regression analysis was carried on the parameters that were varied
(shake-out time and silicon content) and those measured (carbide grain size
Optical X 100 A-E: Optical micrographs of automobile brake
drums cast in sand mould containing 1.7% Si, pouring temperature of 1400oC
and shakeout times of (A = 12 min, B= 30 min, C =1 h, D = 5 h, E = 10 h),
F-J: Optical micrographs of automobile brake drums cast in sand mould containing
(2.2%) Si, pouring temperature of 1400oC and shakeout times of (F = 12 min,
G = 30 min, H =1h, I = 5 h, J = 10 h), K-O: Optical micrographs of automobile
brake drums cast in sand mould containing 2.7% Si, pouring temperature of
1400oC and shakeout times of (K = 12 min, L= 30 min, M = 1 h, N = 5 h, O
= 10 h)
The result obtained with shake-out time as the dependent variable is given
Shake-out time = -13.8+0.0799Si+6.954GS
Micrographs of the samples produced are shown in Fig. 3.
The parameters that have varied in this study are shake-out time and silicon content and those that were measured, which resulted from these variations are carbide grain size and hardness. As reported earlier in the results, multiple regression analysis has shown that:
Shake-out-tine = -13.895+0.799 Si+6.954
GS + 0.0467 Hardness
Hence, shakeout time can be predicted and or specified using this equation. Also from the results of this work, grain size of carbide and hardness are both dependent on Silicon content of melt. Therefore, a particular shake-out time and silicon content can be chosen that will give the correct carbide grain size and hardness and hence the correct mechanical strength desired in the brake drum.
As the micrographs have also revealed (Fig. 3), choice of
appropriate shake-out time is also necessary in order to develop the right microstructure.
As shown in this figure, the finest microstructure are obtained in the shortest
shakeout times at different levels of silicon (Fig. 3A, F
and K). This also shows that the silicon levels are immaterial
as long as the shakeout times are short. This is also evident in Fig.
1 and 2 where the hardness and grain size values almost
coincide at short shake-out times.
It must be noted that the results which led to the above equation were obtained using a mould of known sand properties. Any change in the properties of the mould will definitely affect the microstructure of the brake drum, most especially the carbide grain size. It is mould properties that will determine the cooling rate of a casting and consequently how long the casting to remain in the mold before shake-out to develop the desired microstructure. It is therefore necessary in any foundry producing brake drums to determine their own relationship between shake-out time and other parameters based on their own mould properties and other raw materials for production of brake drum.
The control of processing parameters is of vital importance in the casting of automobile brake to required specifications and a predictive equation can be found linking all the processing parameters using multiple regression analysis. This equation can be used to predict shakeout time when all other parameters relating to mould properties are kept constant within a particular foundry.