Research On Polymer Film Thickness On-line Measurement Technology With Broadband Optical Absorption Method

Polymer film is an important kind of chemical material. With the development ofpetrochemical industry and polymer processing industry, polymer film production andapplication also obtain a great development. Polymer film plays an increasinglyimportant role in agriculture, packaging, chemicals, defense and people’s life. In manyindustrial performance requirements, the thickness of the polymer film is a maintechnical indicator for measuring quality. The polymer film thickness is not only relatedto the production cost, but also affects the actual application effect such as the tensilestrength, barrier properties. Therefore, polymer film thickness on-line measurement hasalways been a main concerned problem.In this dissertation, broadband optical absorption method is proposed for polymerfilm thickness on-line measurement. At present, the dual-wavelength contrast method isused in commercialize film thickness on-line measurement device, while it has theproblems of double monochromatic measurement position deviation and poor versatility.For these problems, to develop an on-line film thickness measurement system with ananti-environmental interference ability, long-term reliability, and without having toconfigure filters is a key issue of this dissertation. Applying Lambert law in broadbandand building a film thickness measurement model based on broadband opticalabsorption method is another important issue of this dissertation. In addition, for theindustrial polymer film production line actual situation of non-stop continuousproduction, compensating illuminant spectral intensity fluctuations, according to thechange of broadband intensity during the long working time is also an important issueof this dissertation. In response to these problems, carry out research work of thisdissertation. The main research contents are as follows:(1) According to the optical radiation attenuation of broadband light after passingthrough the film, the film thickness on-line measurement method based on broadbandoptical absorption is proposed. Exponential decay laws of optical radiation pass inmedium and Lorentzian lineshape of material spontaneous radiation spectrum andabsorption spectrum are analyzed. When a certain frequency range of light radiationpropagates in material, the influences of the bandwidth ratio of the material opticalthickness spectral distribution and the illuminant spectral intensity distribution and theircenter frequency drift rate on the material optical thickness measurement are discussed.The abilities of resistance to the systematic error and random error are analyzed, whenusing broadband optical absorption method to measure material thickness. Thebroadband film thickness measurement method is proposed, after the concept of band Lambert law was proposed. This model does not need to analyze the characteristicabsorption band of the film, and is suitable for the film with a wide band lightabsorption characteristics, and has a strong resistance of random error for spectraltransmittance measurement. The absolute value of simulation error is less than0.4mand the absolute value of experimental error is less than0.8m when the film thicknessis about100m.(2) For special application requirments of polymer film production in industrial site,a broadband film thickness on-line measurement system is developed. A functionbetween optical power and film thickness is built according to the broadband filmthickness measurement model. With the characteristics of discrete function, a filmthickness inversion method based on numerical analysis is adopted. A broadbandthermal radiation illuminant and a optical power meter based on thermopile detector areused in the system. The broadband optical absorption method is used for polymer filmon-line scan measurement. The influence of relative position change of illuminant anddetector to the light radiation received by the detector is analyzed. The scanning frameguide rail is calibrated with the dynamic calibration method using small-sized standardsamples to scan the whole range. After the calibration, maximum error range due toguide rail scanning is less than0.7m when measuring film at the thickness of100mor less.(3) For the long-term continuous industrial polymer film production line case,broadband illuminant spectral radiation intensity fluctuation compensation methodbased on band intensity change is proposed. The functional relationship betweenspectral radiant exitance and band radiant exitance is built by studying the dependencerelationship between spectral radiant exitance and whole radiant exitance of blackbody.When this method is applied to on-line measurement system, no additional hardwaredevice is needed. Illuminant spectral radiation intensity fluctuation can be compensatedjust by measuring the illuminant band intensity change. This method solve the problemof the influence of illuminant fluctuation to measurement results. The vast majority ofdata points compensation results are in a relative error of less than1%, when theblackbody whole radiant exitance vary double. The fitting relative error of broadbandilluminant spectral intensity is less than1.5%at different power input.(4) The experiment research and uncertainty analysis of the measurement system iscarried out. The film samples are obtained by using on-line method and theirthicknesses are measured by contourgraph. The spectral transmittances of the filmsamples and the illuminant relative spectral intensity are measured by Fourierspectrometer. The relationship curve of optical power and film thickness is obtained,and the measurement resolution of the system is0.4m for measuring film at the thickness of less than100m. Use film standards with different thicknesses to test thesystem measurement error. The maximum deviation between30thickness outputs andthe standard thickness are less than1.5m; each thickness output value is the average of5measurements. The uncertainty of the measurement system is analyzed. In themeasurement range of50m to100m, a single measurement expanded uncertainty(k2) is3.0m and the mean of5measurements expanded uncertainty () is lessthan1.7m.