 |
|
|
|
|
||||||
|
|
|
||||||||
|
|
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
6.1 Influence of humidity and temperature on polyimide-coated fiber Bragg gratings6.1.4 Results and discussion
Fig. 6-4 Bragg wavelength of polyimide recoated in-fiber Bragg grating for different relative humidity and temperature Fig. 6-4 shows the Bragg center wavelength of the polyimide recoated FBG as a function of time. At constant temperature an increase in humidity shifts the Bragg wavelength to higher values. The RH influence on the polyimide seems to be reversible, as the Bragg center wavelength is the same at the beginning and the end of the RH-cycle for constant temperature. Previous experiments have shown a non-reversible component depending on the maximum temperature that the FBG has experienced in the past. It may be due to a thermal curing process of the polyimide. This effect is not well understood and will be studied in the future. For bulk polyimide, the volume variation for a RH change is isotropic in all directions. Since the polyimide is tightly attached to the fiber, polyimide longitudinal strains are transferred to the fiber. A volume change induced by the water content inside the polyimide matrix will lead to a fiber elongation or retraction.
Fig. 6-5 Normalized FBG time response at 28 °C and 50 °C is compared to the normalized response of the RH sensor. At each step the saturation level is obtained after several tens of minutes. The time constant of the process depends on temperature (Fig. 6-5). At low temperature the polyimide coated FBG responds much slower than the climate chamber RH evolution measured by the reference gauge ("Rotronic" sensor). With increasing temperature the response accelerates. Diffusion of water molecules through the coating determines probably the time constant [4]. Fig. 6-6 shows the Bragg wavelength shift as a function of relative humidity (steady state average values) for the different temperature cycles. For each temperature we obtain a linear function for the Bragg wavelength shift vs. relative humidity. Small deviations from linearity are within the measurement errors. We can describe the relative wavelength shift with temperature and relative humidity as:
where AT and BRH% are the respective T and RH sensitivities of the polyimide recoated FBG. A two dimensional regression to the temperature and relative humidity data leads to
AT = 1.06×10-5 ± 1×10-7 K-1 BRH% = 4.36×10-6 ± 5×10-8 RH%-1,
where the errors are obtained as the respective standard deviations from the fit. A relative humidity variation of 80% leads to a maximum wavelength shift of 0.54 nm at 1550 nm. This corresponds to a 33 °C temperature variation. This wavelength shift is more than 2/3 of the channel spacing in 100 GHz DWDM systems and may cause system failure. Polyimide re-coatings are generally used in high temperature environments (T > 120 °C). In such environments where relative humidity and temperature can change, temperature sensing needs the measurement of relative humidity by an additional grating. The observed linearity of the Bragg wavelength with relative humidity may find use in an all-fiber RH sensor [5].
Fig. 6-6 Bragg wavelength shift of polyimide recoated FBG as a function of relative humidity for different temperatures. |
|
|
|
|
||||||||
|
|
|
|
|
||||||