Haddeleme işleminin silindirik parçalarda uygulaması, deney sonuçlarının yapay sinir ağları ile modellenmesi
Öz
Haddeleme işlemi talaşlı imalatta son işlem operasyonu olarak kullanılmakta ve uygulandığı iş parçası
yüzeylerinde bazı iyileşmeler (yüzey sertliğinin artması, yüzey pürüzlülüğünün iyileştirilmesi, yorulma ve
aşınma ile ilgili iyileşmeler, yüzeydeki çatlak oluşturabilecek derinliklerin azaltılması vb.) meydana
getirmektedir. Bu çalışmada haddeleme yapabilen bir aparat yardımıyla silindirik Al 6061–T6 malzemesi
üzerinde farklı parametreler (Kuvvet (N), İlerleme (mm/dev), Bilye Çapı (mm), Paso sayısı) kullanılarak
haddeleme işlemi gerçekleştirilmiştir. Haddelemede kullanılan parametrelerin yüzey pürüzlülüğü ve yüzey
sertliğine etkileri incelenmiştir. Son olarak deney sonuçları Yapay sinir ağları ile modellenerek elde edilen
sonuçlarla deney sonuçları karşılaştırılmıştır.
Anahtar Kelimeler
Application Of The Burnishing Process On Cylindrical Parts, Modelling Experimental Data With Artificial Neural Network
Abstract
Burnishing process a sensitive processing
technique being applied in super final of
functional surfaces of machine elements without
machining is being used as the finishing operation
in machining, and brings along reward (increase
of surface hardness, improved surface roughness
improvement on fatigue and wear, and so
reducing the depth of the surface cracks can be
created.) work piece on which it is applied. In his
final operation technique, it can also be used in
geometrical improvement processes such as
removing the ovalness as well as mechanical
improvement processes. While performing
burnishing operation, the use of lubrication
assists the improvement of mechanical features.
As many processing methods (milling, stoning,
polishing, honing, adapting) are being used
frequency improving the surface roughness value,
burnishing operation is also a processing method
being used for this operation. The purpose of this
final operation can be explained as not obtaining
a dimensional correctness but as composing a
appropriate surface roughness value on the piece,
increasing the abrasion and corrosion resistances
of the work piece, enabling improvements in the
fatigue and tensile strengths of the piece and
reaching enhanced micro hardness values on the
surface layers of the piece.
In this study, application of burnishing operation
on cylindrical pieces and the affects of parameters
used in burnishing operation (pressure force,
radius of ball and rotation speed) on the surface
roughness and surface hardness was discussed.
For the tests, Al 6061 –T6 alloy was used as work
piece material. Finally, experimental results was
modelled with artificial neural network and The
experimental results were compared with the
results achieved from ANN. The tests were
realized by using CNC Taksan lathe machine.
Coolant was not used in the tests. Average surface
roughness of the work pieces prepared for the
tests was measured as 2,6μm, and normal surface
hardness was measured as 65 RB. So as to apply
burnishing operation, an apparatus was used
which can be used for the burnishing of
cylindrical pieces on the lathe and which has
ballon its end as roller set. The hardness value of
balls, which are taken as ready and which are
used as crusher, was 58-65RC, and surface
roughness was 0.10-0.15μm. Spring system was
used in order to adjust the pressure force during
burnishing. Three different head design was made
to hold the ends of apparatus (balls) procured as
ready and used as roller. These heads are being
used as per the diameter value of the balls used.
The force required for the performance of
burnishing operation had been obtained by spring
pressure. Consequently, different forces had been
obtained at different spring tightenin grates. 108
tests were realized in order to examine the affects
of parameters used in burnishing operation. As
the result of tests, the measurements were realized
with Mitutoyo Surftest Sj-201 device for surface
roughness and with Digirock-RBOV device for
surface hardness. As the result of the tests made,
it was determined that the burnishing operation
positively affects the surface hardness and surface
roughness of the material. The highest surface
hardness was measured as 81.49, and hardness
increased by 25% compared to raw piece. The
lowest surface hardness was measured as 77.6
RB, and hardness increased by 19% compared to
raw piece. In the same manner, the best surface
roughness value was measured as 0,12μm, and
roughness improved by about 20 times compared
to raw piece. The highest surface roughness value
was measured as 1,23 μm, and about 2 times
improvement realized in respect of roughness
value compared to raw piece. Under the
conditions of current test;
The negative effect of increase of feed rate on
the surface roughness and surface hardness
was observed.
The positive effect of increase of pressure force
on the surface roughness and surface hardness
was observed.
The positive effect of increase of pas number
on the surface roughness and surface hardness
was observed.
The positive effect of increase of ball diameter
on the surface roughness and surface hardness
was observed.
The optimum model is obtained within the
4×24×2 configuration that consists of a single
hidden layer with 24 neurons in addition to the
4 input sand 2 output with the hyperbolic
tangent activation function for the ANN
intelligent model.
