·Table of Contents
·Workshop - Neutron workshop
Thermostats studied by dynamic neutron radiography and vibration diagnosticsMárton Balaskó,
KFKI Atomic Energy Research Institute, H-1525 Budapest, P.O.B. 49. Hungary
MTA Research Institute for Solid State Physics and Optics,
H-1525 Budapest, P.O.B. 49. Hungary
KFKI Research Institute For Particle and Nuclear Physics
H-1525 Budapest, P.O.B. 49, Hungary
2.1. Dynamic neutron radiography
The DNR station at the 10 MW Budapest research reactor services as a high performance experimental facility for investigating model units and actual industrial objects under running conditions . The neutron beam is obtained through a horizontal thermal channel of the reactor using a pin-hole type conical collimator. The collimation ratio is 170, the beam diameter is 150 mm and the neutron flux is 108n cm-2 sec-1 at the detector position. Objects with a surface area of 700 × 1000 mm2 and weights of up 250 kg can be investigated and are manipulated into the correct position of the beam by means of a remote control mechanism. The transmitted neutron image of the object is converted into light by a scintillation converter screen (type NE 426) and the "light image" is detected by a low-light-level TV camera (10-4 lux for type TV1122). The imaging cycle is 40 msec, thereby providing the positioning for visualizing medium speed movements up to about 2.5 m sec-1 inside the investigated object. The resolution of the detected radiography image is about 150 mm. The schematic block diagram of the imaging system is shown in Fig. 1. The radiography images are displayed on a monitor, stored by an S-VHS recorder and for further analysis, an image processing system is used (type Sapphire 5.05 by Quantel). A hard copy of the picture is provided by a video printer. In addition to the radiography image other parameters characterizing the operation of the investigated object - such as operating time, pressure, temperature, flow velocity, power consumption - are measured and controlled.
|Fig 1: Schematic layout of imaging system of the DNR station|
2.2. Vibration diagnostics
The study of the thermostats were obtained in the neutron radiography laboratory. The same auxiliary equipments were applied as they were on the neutron radiography station A 4343-type accelerator sensor was used as the detector of the noise events in the thermostats. The signals were amplified by 2635-type charge amplifiers and registred by a NAGRA-IV. S-type stereo magnetofon. The sensor and amplifier were produced by Brüel and Kjaer company. The magnetofon was made by KUDELSKI SA from Switzerland. The evaluation of the noise spectra is performed by a 2033-type Dual Channel Frequency Analyser, which was made by Brüel and Kjaer company. The sensors were fixed to the surface of the examinated thermostat as it is shown in Fig. 2. One of them was placed on the sensor, while another one on the membrane.
|Fig 2: Measuring arrangement of the thermostat's noise spectra in the neutron radiography laboratory|
4.1. DNR measurements on thermostats
The goal of our study was to explain the cause of the delayed starting of the compressor in the refrigerators. The supposed cause of this defective operation was the incorrect functioning of the thermostat. First of all we designed a special small cooling tower with the aim to avoid the disturbing effect of neutron scattering of the insulation materials of the refrigerator box. Within the special tower is a double level, water cooled Peltier-block which contains a microbox, that's minimum temperature - 40 oC. Its neutron radiography picture is illustrated in Fig. 3. Since the behaviour of the propellant gas depends on the position of the thermostat sensor it was possible to investigate it by this rotatable arrangement. Figure 4 shows the long, liquid gas column at -40 oC in the vertical position of the sensor. When the compressor switched off, the temperature began to increase slowly and the length of the liquid was reduced slowly. At -5 oC the residual column suddenly collapsed generating a relatively strong pressure change in the capillary tube whereupon the membrane restarted the cooling procedure. After some seconds the liquid returned to the sensor and the cooling was stopped immediately. This finding proved to be the explanation for the delayed starting because the compressor has a contact breaker to avoid damage caused by the repeated starting. The circuit waits for five minutes after such an event before it switches on the power again.
|Fig 3: Neutron radiography picture of double level, water cooled Peltier-block||Fig 4: Liquid propellant column in the sensor at -40oC|
The above described model experiment was continued on actual refrigerators and our statement was supported by the results. We were therefore able to determine the correct mounting position for the thermostats. By adjusting the thermostat to the correct mounting position, the vertical outflow of the sensor was avoided and the problem was solved.
4.2. VD measurements on thermostats
The membrane of the thermostat was disassembled from the frame of mechanical actuator. We had to fix directly the accelerator sensor to the surface of the membrane (see Fig. 2), because the noise level of the propellant material's motion was very low.
In the first arrangement the position of the sensor was vertical in the cooling tower, where the end of the sensor was upper. Two VD detectors were used. The first one detected the noise on the capillary tube under the Peltier-block, and the second one was placed on the membrane (see Fig. 2). The temperature of both places were measured by thermocouples. The noise of the cooling procedure was not detectable. When the temperature was increased the motion of the membrane gave a well distinguishable double noise at -5 oC. The measured water-flow diagram is shown in Fig. 5/a.
|Fig 5a: Water-flow diagram of the membrane in vertical position|
|Fig 5b: Water-flow diagram of the capillary tube in vertical position|
Each spectrum was measured during one second. The running time was measured from the beginning of evaluation. It is independent on the time schedule of the worming procedure. The form of the line is depending on the detected noise. A primary noise was detected at the 28th second and at the 30th second was detected a higher intensity noise. This is a typical hazard situation for the contact breaker of the compressor to delay the restarting of the cooling procedure. Figure 5/b shows the water-flow diagram measured on the capillary tube. The switching phenomena was detected at the same time except the double wave form, and the intensity of the noise was smaller.
After the cooling tower position was turned by 90o, and the position of the thermostat sensor became horizontal, the VD measuements were repeated (Figs. 6/a,b). Figure 6/a shows a relevant sharp switching effect of the membrane at -10 oC without any secondary noise.
|Fig 6a: Water-flow diagram of the membrane in horizontal position|
|Fig 6b: Water-flow diagram of the capillary tube in horizontal position|
Later we tried to extend the VD application in the measurement of the controller system in the complete refrigerator box. It seemed this technique was not available to use because the isolation technology and the high background noise of the evaporator system prohibited the success of the experimental work.
|© AIPnD , created by NDT.net|||Home| |Top||