Investigation of Wear Behavior of 1060 and 1095 Steels using Regression Analysis

The dry adhesive wear behavior of 1 060 and 1095 steels using pin on wheel wear test rig at room temperature under different wear conditions has been studied. The prepared specimens were normalized to m ake sure that all specimens are in the same conditions. 'The objective of this rese arch is to study the effect of operating parameters such a s rotating s peed , normal load and sliding time and their interaction on the wear behavior using regression analysis. Detailed data obtained was used to develop equations to describe the wear rate of steels. The wear losses of the specimens were expressed in term s of simultaneous contribution from the effects of rotating speed , normal load , sliding time and t heir interaction. By applying the parameters such as rotating speed , normal load , and sliding time,observed a certain effects on the wear behavior of two types steel , the 1060 steel , sliding time is the m a in factor, followed by normal load and rotating speed . However ,for the 1095 steel , rotating s peed i s the main factor, followed by normal load and sliding time. From results, it was concluded that the wear behavior of these steel s and the effects of this parameters on the wear depend on the physical and mechanical properties of this types of steels.


INTRODUCTION
LL wear analysis are very complex since there are many interdependent factor affecting the final removal of wear debris and formation of worn surface (1)(2).It was \Nell established that .sometimessmall changing in some variable totally change the wear rarer and wear mechanism (3).Therefore, it requires a technique that can predict the mechanism and wear rate under all parameters affecting the wear phenomena (4).A statistic analysis namely regression analysis based on modeling can analysis all the• variables in concern (5).Regression analysis i s focused on the effect of dependent variable and the related independent variables control the process.All information regarding the regression analysis can be found in the article of Prof. Sykes (6).
Wear is one of common failure mechanisms in machine parts.It has become an important subject in Tribology [7].Steels are the most widely used and least expensive metallic on earth [8].As well -known, the wear behavior of steel s can be influenced by many factors [9].The factors which are studied rotating speed, normal load, and sliding time are the common factors.Many researches on wear behavior of steels have beendone and reported on the effect of these factors on wear [1O] .The effect of operating parameters and their interactions on wear behavior of steel s has not yet been studied.The aim of this paper was to investigate the effect of rotating speeds, applied loads, sliding time and their interactions on wear behavior of steels using regression analysis• in order to fully understand the effect of these factors on wear behaviour of materials.The wear tests were carried o u t using pin on w heel wear test rig at room temperature in dry sliding.

Materials and Experimental Procedures Materials and preparations
The material s used in this paper were 1060, and 1095 steels.Their compositions andmechanical properties are given in Tables 1 and 2. The specimens prepared ascylindrical pins.The pins were 20mm in length and 6mm in diameter.All pins were treated as normalizing by heating to austenite temperature for 10 min then air cooled to room temperature.The aim of normalizing treatment was obtaining a homogeneous structure.

Steels using Regression Analysis pin on Wheel
The wear tests were carried out u s m g pm on wheel wear test ri g in production engineering and metallurgy laboratories a t room temperature in dry sliding, as shown in Fig .l .The tool steel was used as the wheel material.W heel of 60mrn in diameter and 20mm in thickness, was fitted on shah that was driven by a n electrical motor.Loa d was applied by a dead weight on the pin holder.Three rotating speeds were fed from a n electrical motor and reduced by stepped pulley.Prior to the test, the pins were weighed using a balance with a precision of ±0.l mg.The wear amount of pins was defined as wear loss in (mg) with respect to the initial weight.

Results and Discussion Test Arrangement and Results
The effects of rotating speed ( V, rpm), normal load ( P , N ), sliding time ( t, min) and their interactions on the wear loss were investigated .The investigated parameters and their test level the range of normal load s was 50, 100, 150, and 200 N and the range of Speeds was 500, 850, and 1 200 rpm.Sliding times were differing in range of 30, 60 and 90 min as shown in Table 3.The test arrangement and results are listed in Table 4.

Regressive Equation between Mass Loss and its Affecting Parameters
A regression equation was established with the wear loss as the dependent variable and wit h a quadratic dependence on the three factors as independent variables.The significant coefficients of this equation were identified by regression analysis, using computer program Minitab 12 [11].The coefficient describes the contribution of parameters to the amount of wear.

Steels using Regression Analysis
of the squared differences between y predicated and y observed.The absolute value of a coefficient, relative to its standard error, suggests the relative degree to which thecorresponding variable contributes to the &mount of wear.
A positive sig n for a give n coefficient suggests that higher contribution of the parameter increase the amount of wear, whereas a negative suggests that lower contribution provides higher wear amount.Variables are considered in order of their relative importance as indicated by tests studied based on the magnitude of the coefficient, relative to the standard error.Coefficients whose given tests have a small probability (P < 0.05) under the null hypothesis indicate significant contributions to the dependent variable.
Based on the output results in Tables 5 and 6, by using regress ion equations between wear loss y (mg) and each individual parameter as well as their interactions can be obtained.For 1060 steel, the regression equation y 1 (mg ) was expresses as:

Checking the Adequacy of the Developed Equations
The adequacy of fit estimated regression equations was tested by applying Analysis of Variance (ANOVA), as shown in Tables 5 and 6, using computer software SPSS 9 [1 2].For significance level α = 0. 05 , and based on p-value, it can be concluded that the developed equations of 1060 and 1 095 steels f it the experimental wear data adequacy.

Regression Analysis
The standard error (Se) to estimate i s one of a good indication of the fit in regression analysis method.If there are 68 % of the observations will falls within ± 1 Se, if there are 95% of the observations will fall s within ±2Se , and 99.7% of the observations will falls within ± 3 Se, then the regression equation s will be fit the experimental wear data [13].The bellow results show good fit of the developed equations to the experimental wear data shown in table 7.
Calculated R2 values (0.924) for ]• 060 steel and (0.887) for 1095 steel as shown in Tables 5 and 6, it can be indicated that 92.4(% of the variation in the response (wear) was attributed the selected operating parameters, for 1 060 steel , and 88. 7% of the variation in the response (wear) was attributed the selected operating parameters, for 1095 steel, during wear process.The calculated adujsted-R2values (0.9 11 ) for 1060 steel and (0.8 68) for 1095 steel as shown in Tables 5 and 6, it can be indicated that the samples size and terms of the equations was satisfactory.

DISCUSSION
The regression coefficient can be taken as a measure of the role of parameter on the wear, the stronger the role of the parameter, the higher the coefficient.From the wear regress ion Equations, ( 1) and ( 2), it was see n that, during the wear process, the parameters such as rotating speed, normal load, and sliding time have certain effects on the wear behavior of two steels.Among these parameters, sliding time is the main factor, followed by normal load and rotating speed for the wear of 1060 steel.However, rotating speed is the main factor, followed by normal load and sliding time for the wear of 1095 steel.
The relative role of operating parameters have been varied due to change in the physical and mechanical properties such as surface hardness, toughness, microstructure,… etc. resulted in the variations of the relative role of the parameters during the wear processes.

CONCLUSIONS
The main aim of this research was to investigate the effects of three operating parameters and their interactions on the wear of 1060 and 1095 steels, and to correlate the wear loss with these operating parametars.In particular, the following observations and conclusions wear made: 1.
A successful attempt has been made to describe the wear behavior of 1060 and 1095 steels using regression equations.The tegression equations between operating parameters and wear loss exhibit a good correlation with experimental measurements within the range of investigation.

2.
Slibing time is the main parameter, followed by normal load, while rotating speed ptays less effect on the wear of 1060 steel then the other two parameters.

3.
Rotating speed is the main parameter, followed by normal load, while sliding time plays less effect on the wear of 1095 steel then the other two parameters.

4.
The individual operating parameters exerted a different on the two steels.The value of the relative degree of effect was dependent on the physical and mechanical properties of the materials.

&Tech.Journal, Vol.
The coefficients are estimated by minimizing the s um