The presented work consists of a test setup study of a new pneumatic material testing device based on double bellows (DBs) loading device and with miniature autoclaves enabling applications at temperature and pressure up to 650 °C and 35 MPa, respectively. It has been demonstrated that it is technically feasible to carry out well defined and controlled material testing in the supercritical water (SCW) environment using this testing system. By using this type of system, it makes possible to investigate the intrinsic role of the applied stress on the deformation behavior of material in light water reactor (LWR) conditions and also in other harsh environments like SCW conditions. In addition, the compactness and versatility of the setup makes this system particularly attractive for deployment in a hot-cell for testing of irradiated materials.
Modern-day high technologies such as nuclear power plant and fusion reactor technologies have strongly affected the development and testing of their construction materials. Instead of traditional material parameters, such as yield and ultimate tensile strength, designers have to know much more about other specific characteristics of the construction materials. The field of materials and their properties is very wide, including aspects such as environmentally assisted cracking (EAC) susceptibility, fatigue strength, corrosion resistance, and the effect of irradiation on material properties . In many cases, materials are used in different environments and, for example, the operation environment can affect the material's durability more than simply yield and fracture strength-based calculations can take into account. Life cycle considerations also have a marked importance on today's materials development work.
The use of fracture mechanics-based approach plays important role in the research of EAC in supercritical water (SCW) environment. The main factors of interest are the susceptibility of the material to stress corrosion cracking (SCC) and stress corrosion cracking propagation rate, (da/dt) SCC, as a function of stress intensity factor KI or J integral. Also, the threshold of KI for EAC is of interest. Furthermore, the electrical and electrochemical properties of the oxide films formed on construction material surfaces have a significant influence on the susceptibility of materials to general and localized corrosion, such as in EAC in high-temperature aqueous environments. According to the results on modeling high-temperature aqueous corrosion [2–6], the properties of oxides are influenced by temperature, potential, electrolyte composition (presence of oxidizing, reducing, and/or complexing agents) and electrolyte flow rate, in addition to the chemical composition and the initial surface finish of the material.
The pneumatically powered material testing technology can play a very important role in the current and future material testing programs focused on material qualification needed for the extension of the operation license of the current generation (Gen) II and III light water reactors (LWRs), as well as for future advanced fission Gen IV concepts like SCW reactors, and for construction of the fusion experimental reactor international thermonuclear experimental reactor and the demonstrator prototype fusion reactor. This is particularly true when materials need to be tested at very high temperatures (up to 650 °C) and pressures (35 MPa), and/or under irradiation in in-pile/in-reactor facilities. As already mentioned earlier, successful extension of the pneumatic bellows-based technology can occur via utilization in future experiments such as high-temperature liquid lead facilities or future hot-cell and in-pile facilities for which design is under way. In order to perform material testing under liquid lead recirculation loop up to 650 °C and other very demanding environments, the technological development path from single bellows pneumatic loading device toward the more demanding double bellows (DBs) system for fatigue and combined tension and compression internal pressure system is under development.
The pneumatic loading technology provides important benefits and has already been successfully applied for materials testing in different test environments . The pneumatic servocontrolled system has earlier been used to perform fracture toughness, corrosion fatigue, tensile, and electrochemical measurements in gas, high temperature aqueous, and irradiation environments. The main advantages of a pneumatic testing system are its sensitivity and the possibility to separate the material testing and control system by long (>30 m) pressure tubes without decreasing the test load accuracy. Furthermore, the system can be made compact so that several testing systems can fit simultaneously into one test chamber. Thus, the total cost of testing is decreased and the reliability of the test result is considerably increased.
The primary objective of this work is to test a developed pneumatic material crack growth test device suitable for 5 mm disk compact tension (DCT) type specimens under subcritical and SCW environments. Reliable operation of the loading module, control, and data acquisition systems are to be verified by out-of-reactor tests (in autoclave), followed by possible in-core experiments in the future. The key factors and requirements for the miniature autoclave testing system are the size of the loading frames with the pneumatic loading units and metal bellows, the size of the autoclave chambers, the accuracy of the displacement rate, and the accuracy of the load.
The strength of the test material and the cross section of the test specimen determine the amount of load needed in the tests. It was determined that a load of 500 N, producing a KI value in the range of 10 MPa m0.5 for a 5 mm DCT precracked specimen, would be sufficient for most tests on austenitic stainless steels (plane strain condition), which have been chosen as candidate materials for some components of the European type of SCW reactors concept . The requirements for the test setup for the pneumatic material crack growth test device are that it should be possible to perform constant load, constant displacement rate and fatigue (R < 1) tests, all using the same pneumatic crack growth test device.
Double Bellows Loading Device
To manage the high pressure of the SCW environment (25–35 MPa), the DB loading device is equipped with additional secondary bellows, as shown in Fig. 1. The primary bellows (working bellows) is installed into the pressure chamber and generates the needed test load. The secondary bellows eliminates the effect of the environmental pressure of 25 MPa in the case of SCW environment. The operational principle and load generation of the DB loading device with double bellows are described below.
Theoretical Load Calculations of the DB Loading Device at Room Temperature.
The DB loading device together with the miniature size autoclave consists of the working and secondary bellows, inner piston, pressure chamber, and autoclave as shown in Fig. 2. The system has two different pressure boundaries, i.e., (A) between working bellows (p1) and chamber (p2) and (B) between secondary bellows (p2) and environment (p3), see Eq. (1). Inner pistons are needed for the following reasons; to act as a support element for the corrugated bellows elements, to connect the two bellows together and also to minimize the gas volume of the bellows. The test load can be calculated by using the following equation [9,10]:
where F (N) is load, Δpb(A) (MPa) is the pressure difference on boundary (A), Aeffwb (mm2) is the effective cross section of the working bellows, Δpb(B) (MPa) is the pressure difference on boundary (B), Aeffsb (mm2) is the effective cross section of the secondary bellows, and cδ (N/mm) is the intrinsic stiffness values of the primary and the secondary bellows.
The load generation of the DB loading device is based on the operation of pressure boundaries between working bellows (p1), chamber, secondary bellows (p2), and environment pressure (p3). In the starting position the chamber, working bellows, secondary bellows, and environment are pressurized up to 20 MPa pressure level by using pressure pipes p1 and p2. The pressure of the chamber is the same as the secondary bellows pressure. Because of zero pressure difference for both boundaries (primary bellows and secondary bellows), the test load is zero. After that the environmental pressure can be increased up to 25 MPa. To reach the zero load level at 25 MPa (environmental pressure) the primary bellows pressure (p1) can be set as 12.6 MPa and the chamber pressure (p2) as 9 MPa as shown in Table 1.
Correspondingly to reach the desired test load level the primary (p1) or secondary bellows (p2) pressures can be adjusted according to the needed load level and direction. For example, to reach −468 N load the primary bellows pressure (p1) can be set as 11.3 MPa and chamber pressure (p2) as 9.0 MPa as shown in Table 1.
It should be noted that all three pressures (p1, p2, and p3) were connected into the programmable logic control (PLC) system. To adjust the needed test load and avoid the pressure fluctuation of the autoclave, the primary bellows pressure (p1) and chamber pressure (p2) shall be synchronized together with the autoclave pressure by using their ratio. To reach the zero load situation, the ratio for primary and the autoclave pressure was 0.504 and for secondary and autoclave pressure 0.360 as shown in Table 1. On the other hand, when the primary and autoclave pressure ratio was reduced to 0.452, the load level of −468 N was achieved.
Double Bellows Loading Device With Miniature Autoclave.
The following basic requirements were determined for the SCW autoclave which needs to be fulfilled. The autoclave capable of working at SCW conditions, the maximum pressure and temperature were setup as 35 MPa and 650 °C, respectively. The construction materials are heat resistance Nimonic 80 A and Inconel 625. The maximum outer diameter of the miniature autoclave is 64 mm and inner diameter 32 mm. Conax/Novaswiss type feed throughs are needed for the linear variable differential transformer (LVDT), pressure, electrodes, and potential drop (PD) wires, four places on the lid of the autoclave. The lid is fixed to the autoclave with the bolts and the sealing element is double conical metal ring made of Hastelloy C276. Inductive heating system and water inlet pipe is installed into the outlet pipe. The fulfillment of these requirements is possible through a setup as shown in Fig. 2.
Load Calibration of DB Loading Device.
Commercial load sensors are typically designed for low temperature gas environments, and they cannot be installed for high temperature water conditions. The load determination of the pneumatic loading device is based on developed calibration methods where the metal bellows' intrinsic stiffness and effective cross section are determined for the true load calculations. For this purpose, high temperature calibration of the pneumatic loading device was performed in a furnace with gas environment.
The bellows (working and secondary bellows) intrinsic stiffness and the effective cross section are needed for calculation of the load acting on the test specimen, see Eq. (1). The bellows intrinsic stiffness value reported by the bellows manufacturer, e.g., spring rate, cannot be applied directly and the true intrinsic stiffness of the complete pneumatic loading device has to be determined experimentally. The simplest method to measure the true intrinsic stiffness and the friction factor of the pneumatic loading device is to perform a test with steadily increasing load using the loading device without any test specimen. The inside pressure of the bellows, the true intrinsic stiffness and the friction factor of the pneumatic loading unit with double bellows are needed to calculate for the force acting on the test specimen.
The preliminary calibration tests and calculations for the DB loading device were performed at a load level of ∼500 N and without environmental pressure. The objective of this precalibration procedure was to find out how the intrinsic stiffness and the effective cross section determinations of the DB loading device should be performed. In the first calibration test, the pressure loss arising from the two metal bellows' intrinsic stiffness (working and primary bellows) and internal parts friction factor of the pneumatic loading device were determined over the working range of the bellows. The working range of the double bellows loading device depends on the used bellows size and number of the bellows corrugations. In this case, the maximum displacement for the double bellows loading device is ±1 mm and the maximum load level under SCW environment is ∼3 kN.
The intrinsic stiffness values for the DB loading device were 2.06 MPa/mm at 23 °C and 1.79 MPa/mm at 550 °C. A special calibration furnace was used to calibrate the applied gas pressure in the double bellows (secondary and working bellows) and pressure chamber with the actual load acting on the load cell. The LVDT sensor is placed into the furnace to measure the compliance of the whole system during the calibration. This is due to fact that the intrinsic stiffness of the DB loading device is relatively high and thus affects the load calibration accuracy.
The effective cross section of the secondary bellows was determined by pressurizing working bellows p1 and chamber pressure p2 at the same time. The effective cross section of the working bellows was 368 mm2. There was a small deviation between calculated and measured load curves. Most probably, this was due to the double bellows loading unit's intrinsic stiffness and the internal friction factor of the moving parts. The effective cross section for the secondary bellows was 81 mm2. As an example, Table 2 summarizes the deviation between calculated and measured load values over the estimated working range of the precracked 5 mm DCT specimen at room temperature. The accuracy of the load and pressure calibration over the tested range (0–500 N) was approximately less than 4.5% up to 200 N load level. After that the accuracy was less than 1% from measured value. The same accuracy level was achieved also at higher temperature, i.e., 550 °C.
According to these data, the biggest deviation of the calculated load was at the beginning of the calibration curve. The main reason for this deviation is mostly mathematical but also caused by compliance in the beginning of the calibration. The compliance of the calibration furnace and free clearance of the rolls slightly affects the calibration results. In any case, the calibration accuracy is dominated by the accuracy of the pneumatic sensor, which in this case is ±0.2% from 20 MPa.
Reference Tests in Air at 550 °C.
The first test with the pneumatically powered miniature autoclave testing system was a reference test at ∼550 °C in air. The 5 mm DCT specimen was installed into the double bellows loading frame with the PD wires. The noncontact type of LVDT sensor was used to measure the displacement of the main load post (connected to the top of the secondary bellows). The tensile load for the test specimen was created by changing only the chamber pressure during the test. The primary (working) bellows pressure was constant (0.7 MPa) during the test. The PD wires and LVDT sensor's wires were placed into the main feedthrough with ceramic and graphite insulators. A typical set of raw data from the constant rising load test is shown in Fig. 3. The starting point of the specimen loading was easily determined from the curve (load line slope is changing when load is applied to the specimen).
LWR Conditions at 288 °C/25 MPa.
The second test with the pneumatically powered miniature autoclave testing system was performed in LWR conditions at 288 °C/25 MPa. The autoclave was connected to the recirculation loop with the preheater as shown in Fig. 4. The target inlet water flow during the tests was ∼16 ml/min. The material used was an AISI 316 stainless steel, and the specimen was a precracked 5 mm DCT specimen.
The designed pressure synchronization program, i.e., PLC, worked well during the pressurization of the autoclave and during the test. The primary and secondary pressures followed autoclave pressure variations. The fluctuation of the autoclave pressure was around ±0.2 MPa during the test. The primary and secondary bellows pressures can automatically follow the changes of the autoclave pressure with the given pressure ratio, i.e., the test load was constant (and zero) during the pressurizing period of the autoclave.
The miniature autoclave testing system was pressurized up to 25 MPa. The dissolved oxygen content was maintained between 150 and 200 ppb (μg/kg H2O) in the inlet flow and during the experiment the inlet water conductivity was below 0.1 μs/cm. The fluctuation of the autoclave pressure was around ±0.2 MPa during the test and even then the test load accuracy was ±2 N at both test conditions, i.e., 288 °C/25 MPa and 500 °C/25 MPa. The primary and secondary bellows pressures follow automatically the changes of the autoclave pressure as well as the fatigue control signal. The maximum load for the AISI 316 stainless steel specimen was 400 N at 288 °C/25 MPa corresponding to a K-level around 8 MPa m0.5 for 5 mm DCT precracked specimen. The mean displacement was −210 μm and amplitude was ±10 μm as shown in Fig. 5. Test load was increased from the initial maximum value of 400 N by 38 N steps up to 480 N (the maximum peak value) during 7 days cycling, Fig. 5. The crack started to growth when the maximum test load reached the value of 480 N. The crack initiation took long time as shown in Fig. 5 and it was observed that it is very sensitive to load levels used in the test. Thus, it is of utmost importance to use devices with high accuracy and sensitivity like bellows based technology.
After few days cycling, the fatigue frequency was changed from 0.1 Hz to 0.001 Hz during the test. It was observed that the crack growth rate decreased as a function of the slower fatigue frequency. The total displacement of the double bellows loading device was increased as a function of the crack growth rate. This means that the intrinsic stiffness of the double bellows loading device have decreased the maximum test load level to around 30 N. This value has been correlated during the actual test.
SCW Conditions at 500 °C/25 MPa.
After 15 days of crack growth period at 288 °C/25 MPa, the autoclave temperature was increased up to 500 °C/25 MPa, i.e., SCW conditions. The maximum test load was increased by 38 N steps for several times up to 500 N under SCW conditions without clear crack growth observation as shown in PD measurements, Fig. 6. The mean displacement was −110 μm and amplitude was ±12.5 μm as shown in Fig. 6. This interruption in crack growth is most probably a result of crack tip blunting due to high oxidation rate in SCW.
The scanning electron microscopy characterization was performed on the AISI 316 stainless steel specimen after the LWR test at 288 °C/25 MPa and SCW at 500 °C/25 MPa conditions. Based on the scanning electron microscopy characterization, the crack tip with the intergranular and transgranular crack growths at 288 °C/25 MPa and interrupted crack growth by plasticity (stripes) in SCW at 500 °C/25 MPa were observed, Fig. 7.
A prototype DB-based pneumatic loading device with miniature size of the autoclave capable of performing loading for tensile type testing in SCW environment has been designed, built, and pretested. The DB loading device was based on new technology; it operated successfully and gave reliable pretest results. The calibrations for the DB loading device were performed with the following results: intrinsic stiffness of the two bellows was 1.79 MPa/mm over the 0.3 mm bellows working range at 550 °C, the shape of the intrinsic stiffness curve was linear, and the effective cross sections of the secondary bellows and working bellows were 81 mm2 and 368 mm2, respectively. The accuracy of the load and pressure calibration over the tested range (0–500 N) was approximately less than 4.5% up to 200 N load level. After that the load accuracy was less than 1% from the measured value.
The selected material for autoclave tests was AISI 316 stainless steel which was tested at 288 °C/25 MPa and in SCW at 500 °C/25 MPa. The specimen type was 5 mm DCT precracked specimen. The designed PLC programs worked well and reliably during the pressurizing period and testing period of the miniature size of the autoclave. Clear crack growth (both intergranular and transgranular) was observed at 288 °C/25 MPa. However, very little or no crack growth (only plasticity) was observed under SCW conditions at 500 °C/25 MPa. Based on this, the crack growth rate was higher at 288 °C/25 MPa most probably due to plasticity of crack tip in SCW at 500 °C/25 MPa. The interruption may occur due to fact that the crack is blunting by high oxidation rate at SCW conditions. Further tests are needed in order to confirm this test result.
As it is shown in Fig. 5, the SCC initiation time can be very long and thus, accelerated tests are attractive by increasing, e.g., temperature or stress/strain levels in LWR environments. Supercritical water is known to be aggressive environment compared to typical LWR environment [11–14] and it can be seen as an accelerating media for studying SCC initiation similar to LWR conditions. However, according to test results achieved in this work, additionally tests are needed in order to assess the sensibility to use SCW conditions for accelerating tests.
Based on the preliminary tests, it was also demonstrated that the compactness and versatility of the device makes it particularly amenable to implementation in a hot-cell setting for testing of irradiated materials.
The Academy of Finland project MENUCHAR (Microstructure Evolution in Nuclear Fuels: Advanced Characterization).
VTT (Technical Research Centre of Finland Ltd).
- Aeff =
effective cross section of each of the bellows (mm2)
- Aeffsb =
the effective cross section of the secondary bellows (mm2)
- Aeffwb =
the effective cross section of the working bellows (mm2)
- cδ =
the intrinsic stiffness values of the primary and the secondary bellows (N/mm)
- da/dt =
crack growth rate (mm/s), a = crack length (mm), t = time (s)
- F =
- Fown =
bellows own (intrinsic) stiffness of the working and primary bellows (N)
- Fsec =
secondary bellows load (N)
- Fwork =
working bellows load (N)
- Gen =
generation, i.e., referring to Gen II–IV reactor concepts
- J =
- KI =
stress intensity factor (MPa m0.5)
- p1 =
primary/working bellows pressure (MPa)
- p2 =
chamber/secondary bellows pressure (MPa)
- p3 =
environmental pressure (MPa)
- PD =
potential drop (μV)
- R =
- Δpb(A) =
the pressure difference on boundary (A) (MPa)
- Δpb(B) =
the pressure difference on boundary (B) (MPa)