 ·Table of Contents ·Workshop - Neutron workshop | Neutron Scattering Study of Temperature Controller MembranesE. Sváb, Gy. MészárosMTA Research Institute for Solid State Physics and Optics, H-1525 Budapest, POB 49,Hungary M. Balaskó KFKI Atomic Energy Research Institute, H-1525 Budapest, POB 49, Hungary Contact |
Keywords: Stainless steel, ferrite, austenite, a -iron, g -iron, texture, recrystallization, neutron diffraction, neutron radiography
A series of thermostats operating in the temperature range from 80oC to 250oC was tested with the aim of finding the origin of their defective functioning. By means of dynamic neutron radiography the inner process was visualized at different temperatures both in the cylindrical sensors and in the corresponding membranes. The crystal structure of the stainless steel membrane was analysed by neutron diffraction. The measurements were performed at the 10 MW Budapest research reactor.
Here we describe the neutron scattering measurements and the results obtained on several series of temperature-controller membranes.
Fig 1: Visual picture of the temperature controller membrane
|
2.1 Investigated material
The chemical composition of the investigated stainless steel membranes was: Fe-74%, Ni-6.5%, Cr-17%, Mn-1.3%, Si-1.1% and C, P, S below 0.1% content.
The temperature controller membranes were produced from 0.3 mm thick sheets by forming 26 mm diameter plates. Two plates were pressed together in a special form and thereafter the edge was welded. The visual picture of the temperature controller, including the membrane, the capillary and the sensor, is shown in Fig. 1.
Three series of membrane materials were investigated by neutron diffraction selected from the mass production as representatives of "defective" (Series Nos.1 and 3) and "good" (Series No.2) material.
2.2 Neutron radiography measurements
The dynamic neutron radiography (DNR) measurements were performed on the DNR facility at the Budapest research reactor [1]. Neutrons are obtained from a thermal channel of the reactor through a pin-hole type collimator with L/D=170. The thermal neutron flux f
s at the sample position is 108 n cm-2 s-1 with a beam diameter of 150 mm. For neutron radiography imaging, an NE 426 converter screen is used, that converts neutrons into light and it is detected by a high sensitivity (10-4 lux) video camera. The imaging cycle is 40 msec, thereby enabling medium speed movements up to about 2.5 m sec-1 to be visualized inside the investigated object. Resolution of the neutron image is »
200 m
m. Radiography images are displayed on a monitor and stored by a S-VHS recorder. Image analysis programs (Sapphire 5.05, QUANTEL, UK, and Imane 1.4, KFKI, Hungary) are used (the latter contains a digitization option).
Fig 2: Neutron radiography image of sensor tubes and membranes at 25oC
|
By means of DNR the inner processes were visualized at different temperatures both in the cylindrical sensors and in the corresponding membranes. Figure 2 shows the DNR image of the sensors and that of the membranes at 25oC. It is clearly seen that some of the sensors are practically empty (Nos. 3 and 4) and the other sensors are not homogeneously filled with the propellant. The temperature was increased step by step and it was checked how the propellant reached the membranes. At 250oC the membranes were filled with the propellant and the sensors were depleted. By subtracting the intensity distribution (NR image) measured at 25oC from that measured at 250oC, it was established that no intensity change - no fluid transport - has occurred for the defective thermostats. In this way, it was clarified that the problem with the sensors originated as an error in mass production, especially in the filling process.
The inhomogeneous distribution of propellant fuid observed by neutron radiography may be caused by the inhomogeneous elastical properties of the membrane, that may lead after a longer use to fragility. In order to detect the material characteristics neutron diffraction investigations were performed.
2.3 Neutron diffraction measurements and data analysis
The neutron diffraction measurements were performed on the medium resolution PSD neutron diffractometer [2]. For monochromatising the neutron beam a Zn(002) crystal of 24' mosaicity was used and Soller-type collimation of 15' in-pile and 40' before sample was applied. The PSD diffractometer is equipped with a 3He linear position sensitive detector system that covers 25o in scattering angle, 2q
, simultaneously. The incident neutron wavelength was 1.057 Ĺ. A special sample holder was constructed making available to scan the specimen in arbitrary positions. A 2 mm×8 mm neutron beam slit was applied just before the sample. The raw data were corrected for instrumental background before evaluation.
The neutron diffraction spectra have shown that the material consists of two phases: a -iron with body-centred cubic crystal structure (BCC, ferrite) and g -iron with face-centred cubic structure (FCC, austenite). For illustration, Fig. 3 shows a neutron diffraction spectrum where the relative intensities of the Bragg peaks correspond to random distribution of crystallites (powder-like case). The Miller indices of Bragg reflections both for the FCC and for the BCC phases are indicated.
Fig 3: Neutron diffraction spectrum of a stainless steel temperature controller membrane |
The neutron diffraction patterns were refined by the Rietveld method using the program package FullProf [3]. It was established that in most cases preferred orientation (texture) is present in the material. The relative content of the BCC and FCC phases was calculated and the preferred orientation was taken into account.
Fig 4: Iron-chromium phase diagram
in dependence of Cr content [4]
| 
Fig 5: Iron-nickel phase diagram in dependence of Ni content with 18 wt% Cr [4]
|
The 17 wt% Cr content of the material stabilizes the ferrite phase, while the 6.5 wt% Ni prefers the formation of austenite. As far as the composition of the investigated material is near to the phase boundary (see Fig. 5 - about 6.3 wt% Ni), both phases are present, leading to the formation of metastable structure and thus the physical properties are sensitive to the applied treatments.
| 
| 
Fig 6: Neutron diffraction pattern of steel membrane (Series No.1 - "defective") after different technological processes: a/ base material(sheet) b/ pressed and c/ welded material
| | ||
Figures 6 and 7 show typical series of neutron diffraction spectra illustrating the changes in the structure in dependence of the applied technological processes for a "defective" (Series No.1) and for a "good" (Series No.2.) material, respectively.
| 
| 
Fig 7: Neutron diffraction pattern of steel membrane (Series No.2 - "good") after different technological processes: | a/ base material(sheet) b/ pressed and c/ welded material | ||
As far as the mechanical properties depend on the existing phases we have determined the fraction of the BCC and FCC phases in the starting base material and after the different technological steps for the three investigated series. The results are summarized in Table 1.
| Stainless steel | SHEET (base material) | Pressed | Welded |
| Series No.1. ("defective") | BCC=74% FCC=26% | BCC=83 % FCC=17% | BCC=77% FCC=23% |
| Series No.2. ("good") | BCC=60% FCC=40% | BCC=76% FCC=24% | BCC=71% FCC=29% |
| Series No.3. ("defective") | BCC=73% FCC=27% | BCC=80% FCC= 20% | BCC=76% FCC=24% |
| Table 1 : Content of BCC (ferrite, a-iron) and FCC (austenite, g-iron) phases in three series of stainless steel membrane materials after various technological processes | |||
It was established that in the base material (sheet) the content of the BCC and FCC phases are different. In the "good" material the FCC phase is about 40 % while in the "defective" materials it is less, than about 26 %. On the other hand a strong texture is present in the starting material and this preferred orientation of crystallographic directions changes during the applied technological processes. The most important structural change seems to be the recrystallization from FCC to BCC phase during pressing that may lead to formation of internal stresses, dislocations causing fragility of the membranes. The effect of welding was also analysed, and it was established that a slight recrystallization from BCC to FCC phase has been occurred.
| © AIPnD , created by NDT.net | |Home| |Top| |