KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Settings

Author: search.filters.author.Lim, Chin HockSubject: search.filters.subject.Partial least squares regression

Search Results

Now showing 1 - 2 of 2
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Item,
    Near infrared spectroscopy as an alternative method for rapid evaluation of toluene swell of natural rubber latex and its products
    (2018-06-01)
    Lim, Chin Hock
    ;
    Sirisomboon, Panmanas
    Toluene swell or equilibrium swelling is universally used by rubber factories to measure the degree of crosslink of their compounded or prevulcanized latices at different stages of production. To apply near infrared spectroscopy for rapid and accurate quality control, spectral acquisition of prevulcanized latex, thin film and thick film was performed using a Fourier transform near infrared spectrometer in diffuse reflection mode across the wavenumber range of 12,500–3600 cm<sup>1</sup>. For prevulcanized latex an effective model was developed using partial least squares regression with preprocessing (first derivative + straight line subtraction method). The coefficient of determination (r<sup>2</sup>), root mean square error of cross validation and bias of the validation set were 0.71, 3.93% and 0.005%, respectively. For the thin film model the r<sup>2</sup>, root mean square error of cross validation and bias were 0.65, 4.01% and 0.028%, respectively. Whereas for the thick film model the r<sup>2</sup>, root mean square error of cross validation and bias were 0.70, 4.00% and 0.006%, respectively. Three models including prevulcanized latex, thin film and thick film were validated by 23 unknown samples, providing standard error of prediction and bias of 5.357 and 2.494, 4.565 and 1.001 and 3.641 and 0.961%, respectively, for prevulcanized latex, thin film and thick film. The model developed for the thick film spectra gave the best results.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Item,
    Evaluation of prevulcanisate relaxed modulus of prevulcanised natural rubber latex using Fourier transform near infrared spectroscopy
    (2017-01-01)
    Lim, Chin Hock
    ;
    Sirisomboon, Panmanas
    The analysis of the cross-link density of prevulcanised natural rubber latex using near infrared spectroscopy was conducted using a Fourier transform near infrared spectrometer in diffuse reflection mode over the wavenumber range of 12500- 3600 cm<sup>-1</sup>. As the density of cross-link is an indication of the degree of cure, hence the properties of the latex products, the proposed method is useful for industrial purposes. For samples of prevulcanised latex of 50% total solids content (i.e. PV 50%) at 100% extension (prevulcanisate relaxed modulus 100%), the best model was developed using the partial least squares regression from the spectra, which were pre-treated using the first derivative method, where the coefficient of determination (r<sup>2</sup>), root mean square error of prediction and bias were 0.66, 6.06×10<sup>4</sup> Nm<sup>-2</sup> and 1.63×10<sup>4</sup> Nm<sup>-2</sup>, respectively. The ratio of standard error of prediction to the standard deviation of the reference data in the prediction sample set was 1.8. This model could be used for screening. For samples at 300% extension (prevulcanisate relaxed modulus 300%) for PV 50%, the best model was developed using spectra pre-treated for scattering correction: r<sup>2</sup>, root mean square error of prediction and bias were 0.88, 6.74×10<sup>4</sup> Nm<sup>-2</sup> and 1.35×10<sup>4</sup> Nm<sup>-2</sup>, respectively and the ratio of prediction to deviation was 3.0. Hence, the near infrared spectroscopy technique can be utilised as a rapid screening method for estimating the cross-link densities of prevulcanised natural rubber latex.