Optimization of Process Parameters for Minimizing Surface Roughness in Abrasive Water Jet Machining

Dalkılıç Ö. F., Aydın L., Gültürk E.

International Cappadocia Scientific Research Congress, Nevşehir, Türkiye, 15 - 17 Aralık 2021, ss.78

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: Nevşehir
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.78
  • Kocaeli Üniversitesi Adresli: Evet

Özet

Abrasive water jet (AWJ) processing technology is one of the advanced non-traditional methods, widely used in the processing of different materials such as titanium, steel, brass, aluminum, stone, all kinds of glass and composites. The efficiency of the AWJ machining process is highly influenced by machining parameters, which are generally classified as hydraulic, abrasive, work material and cutting parameters. As well as, surface roughness is a measure of the quality of a technological product and greatly affects the production cost. In this study, Simulated Annealing (SA) algorithm and Neuro-regression modeling were used to find optimal machining process parameters for good surface finish in Abrasive Water jet (AWJ) machining. AWJ process input parameters were taken as traverse speed (V), waterjet pressure (P), standoff distance (h), abrasive grit size (d), and abrasive flow rate (m). The objective function was to minimize average Surface Roughness (Ra).  The experimental data randomly divided into 80%, 15% and 5% groups, respectively, as training, testing and validation. A verified non-linear mathematical model was determined to be used in the neuro-regression approach by computing and  values ​​(>0.90). The calculated values of    ,  and    were within the desired limits (>0.85), confirming that the proposed model is suitable for the neuro-regression approach and estimation of the AWJ process. A stochastic optimization method (SA) was used to obtain optimal AWJ process parameters that provide minimum surface roughness. It was found that the optimum values required to obtain minimum Ra=1.69721 µm were determined that correspond to V=50 mm/min, P=125 MPa, h=1mm, d=94.0135 µm, m=0.5g/s. As a result of these calculations, it has been determined that the surface roughness is inversely related to the abrasive grit size. This study suggests that the neuro-regression used in determining surface roughness in cutting with AWJ can be a new approach, but more experimental studies are needed.

Keywords: Abrasive water jet (AWJ) machining, Neuro-regression approach, Surface roughness, Stochastic Optimization