Analysis of the Surface of Parts Injected with ABS and Treated with Acetic Acid: Influence of Process Settings - Visual Computing: Scientific Visualization and Imaging Systems - page 148

8.3.2 Statistical Analysis
8.3.2.1 DOE1 Experiment
Main parameters of the injection molding process are as follows: Injection time,
Packing time, Injection pressure, Packing pressure, Injection temperature, Mold
temperature, Injection speed, and Flow rate. Some of these parameters can be
correlated through rheological equations of the material. Therefore, their use in
this chapter would be redundant. Thus, in order to assess the influence of injection
process settings upon residual stress in regions close to the surface of ABS
injection molded specimens, a composite experiment named DOE1 was designed.
Parameters chosen were parameters as given in Table 8.1 . The limits of process
settings were set based upon equipment injection capacity. Also, temperature
range is in accordance with industrial standards, both for injection and for mold
temperature. Packing time is a function of attack gate thickness.
Experiment DOE1 is a fractional factorial experimental design 2(5-1) totaling
27 settings (S). This experiment has 16 settings plus two axial points for each of
the 5 parameters, resulting in 10 additional settings and one central point also
considered a setting. These settings can be seen in Table 8.2 . Response variables
are residual tension values shown in the last column, which were determined by
simulation and obtained as indicated in Fig. 8.1 .
In order to identify the most relevant parameters for the DOE1 experiment, a
variance analysis with the aid of STATISTICA 9.0 [ 14 ] was carried out. Results
are given in Table 8.3 .
The first column shows the parameters being assessed, where numbers stand for
the following parameters: (1) Injection Time, (2) Injection Temperature, (3) Mold
temperature, (4) Packing Pressure, and (5) Packing Time. The letters indicate the
linear (L) and square (Q) (nonlinear) effect of the parameter being assessed upon
the variance of the response variable Surface Residual Stresses. The other columns
show values of sums of squares (SS), degree of freedom (df), mean square (MS),
F0 distribution, and p-value, respectively. The parameters Injection Time (Q),
Mold Temperature (Q), and Packing Time (Q) do not show a relevant influence on
the response variable and thus may be disregarded. A new variance analysis can be
found in Table 8.4 .
Based on the results above and on a significance level of 0.05 (chance of error
smaller than or equal to 5 % [ 1 , 2 ]) and based on the values of sums of squares of
SS parameters can be classified by degree of influence on Surface Residual
Stresses. Injection Time, and Packing Pressure were shown to account for 76.5 %
of the variance of Surface Residual Stresses. Other settings did not show influence
greater than 3 % each, as can be seen in Fig. 8.5 .
Upon the use of the statistical software, a response surface correlating Surface
Residual Stresses with the two main parameters identified in the DOE1 experiment
(Injection Time and Packing Pressure) was obtained.