Effect Of Surface Roughness Of The Dielectric On DBD Discharge Characteristics And Ozone Synthesis And Multi-parameter Optimization

As a strong oxidant,powerful disinfectant,catalyst,decolorizer,deodorant and environmentally friendly material,ozone is widely applied in many fields.The most common method for ozone synthesis in industry is dielectric barrier discharge(DBD)firstly proposed by Siemens in 1857.The research on the DBD discharge characteristics is quite sufficient,but the current spikes in current waveform,that is,the phenomenon of the current instantaneous increase,are relatively rare.The cause and the influencing factors of current spikes still have not been reported.At the same time,the effect of dielectric surface morphology and structure on the discharge characteristics,ozone synthesis and current spikes should to be further studied.In this paper,though the point of the dielectric surface roughness,a parallel DBD ozone generator was used to study the current spikes,equivalent capacitance,discharge power and ozone concentration and ozone yield by changing the surface roughness,peak voltage,discharge frequency and gas flow rate.At the same time,the effects of surface roughness,peak voltage,discharge frequency and gas flow rate on ozone concentration were analyzed with response surface methodology,and then to obtain the optimum experimental conditions for ozone concentration.In the scope of this experiment,the main conclusions are as follows:(1)During the procedure of the discharge,the current spikes in the positive half cycle of the current waveform are caused by the local strong discharge by the charge accumulation on the surface of the dielectric.The increase of surface roughness will aggravate the effect on the current spikes.The number of the current spikes in an AC cycle increases with the increasing surface roughness,peak voltage and frequency,and the peak voltage is obvious.And the number of the current spikes is almost independent of gas flow rate.With the increasing of the surface roughness,the initial discharge voltage also increases.When the surface roughness is 0.2?m,0.8?m,1.6?m,2.4?m,3.2?m,the initiation discharge voltage is 4.3 kV,4.5 kV,.4.6 kV,5.0 kV,5.1 kV,respectively.(2)Under different conditions,the Q-V Lissajous figure appears to be a regular parallelogram.The area of the Q-V Lissajous figure decreases with the increasing of surface roughness,increases with the increasing of the peak voltage,and almost independent of the frequency and the gas flow rate.The dielectric layer equivalent capacitance decreases with the increasing of surface roughness and increases with the increasing of peak voltage.The discharge power is almost linear with the increase of peak voltage and frequency,and independent with gas flow rate.At a smaller peak voltage,the discharge power increases with the decrease of surface roughness.(3)With the surface roughness changes from 0.2?m to 3.2?m,ozone concentration and yield decrease at first and then increase.With the peak voltage increases from 5.5 kV to 8 kV,ozone concentration and ozone yield first increases then decreases.With the frequency increases from 5 kHz to 7.5 kHz,ozone concentration first increases then decreases,ozone yield continued to decrease.With the gas flow rate changes from 1 L/min to 3 L/min,ozone concentration decreases,and ozone yield increases first and then decreased.(4)The results of response surface methodology show that the influence of four factors on ozone concentration is very significant.According to the F value of each factor,the order of influence is gas flow rate C>peak voltage D>surface roughness A>frequency B.At the same time,the interaction between surface roughness and peak voltage,the interaction between peak voltage and frequency,and the interaction between peak voltage and gas flow rate on ozone concentration are also significant.The optimum experimental factors in this experimental area are surface roughness Ra3.2,discharge frequency 5.90 kHz,gas flow rate 1.0 L/min,peak voltage 7.20 kV.Under these conditions,the ozone concentration is 91.0(+0.5)g/m~3.The relative error between the regression model and the predicted value is small,which verified the accuracy and reliability of the regression model.