Search

LHD short pulse experiments are executed every three minutes. After the end of the discharge, the scientists must collect, analyze, visualize the last acquired data of the discharge, and prepare for the next discharge. From the beginning, the computer environment of the LHD (Large Helical Device) experiment has been built as a network distributed system, and various computers have been used for data acquisition or physical analysis. When one program is finished on one computer, that computer must send the results in order to the other computers to run programs. Smooth communication is required in order to finish all the tasks before the next discharge. To exchange the information among the applications running on the different computers, the authors have tried various methods, such as a commercial software to share the memory over the network, simple network file sharing method, IP multicast, web interfaces, and others. The purpose of this paper is to share our experiences of trial and error to build the network distributed systems for the consecutive plasma discharge experiments.

The Large Helical Device (LHD) successfully started the deuterium experiment in March 2017, in which further plasma performance improvement is envisaged to provide a firm basis for the helical reactor design. Some major upgrades of facilities have been made for safe and productive deuterium experiments. For radiation safety, the tritium removal system, the integrated radiation monitoring system, and the access control system have been newly installed. Each system has new interlock signals that will prevent any unsafe plasma operation or plant condition. Major interlock extensions have been implemented as a part of the integrated radiation monitoring system, which also has an inter-connection to the LHD central operation and control system. The radiation monitoring system RMSAFE (Radiation Monitoring System Applicable to Fusion Experiments) is already operating for monitoring γ(X)-rays in LHD. Some neutron measurements have been additionally applied for the deuterium experiments. The LHD data acquisition system LABCOM can acquire and process 24 h every day continuous data streams. Since γ(X)-ray and neutron measurements require higher availability, the sensors, controllers, data acquisition computers, network connections, and visualization servers have been designed to be duplicated or multiplexed for redundancy. The radiation monitoring displays in the LHD control room have been carefully designed to have excellent visual recognition, and to make users immediately aware of several alerts regarding the dose limits. The radiation safety web pages have been also upgraded to always show both dose rates of γ(X)-rays and neutrons in real time.

The Remote Experimentation Centre (REC) in Japan has been preparing to replicate the full dataset of ITER over 10 000 km distance. In such a multi-site data repository environment, the data location informing service will be essential to find and retrieve the data efficiently. Considering the long latency time and the self sustainability of remote sites, the data location database should be running at each repository site. Multi-master asynchronous replication between cooperating databases will be essential to realize the remote experimental collaborations in fusion research. This study has investigated the functional differences of some relational databases and found that Postgres BDR has the expected database replication capabilities. Bi-directional replication (BDR) tests by using the LHD database and SNET revealed that the throughputs are sufficient for remote collaborations in fusion experiments.

Technical verification has been progressing for high efficiency data replication between ITER and the Remote Experimentation Centre (REC) in Japan. Transferring a huge amount of data simultaneously to multiple destinations might cause excessive loads and network bandwidth on the sender so that daisy-chained relay transfer would be a considerable solution. This study demonstrates how efficiently the replication relay could be realized for the next-generation fusion experiments, such as ITER and JT-60SA. All the LHD data were consecutively sent to the REC through the global loop path (GLP: Toki - Gifu - Tokyo - Amsterdam - New York - Los Angeles - Tokyo - Aomori - Rokkasho) on SINET5 L2VPN, whose round-trip time is almost 400 ms. MMCFTP was used for the data transferring application. In both the Japan domestic path and the GLP cases, every transfer shows a very stable flattop speed as the preset 8 Gbps. However, longer gap times were needed in MMCFTP initial negotiation to establish numerous sessions. The performance optimized NVMe and iSCSI striped storages have shown higher throughputs than the ITER estimated initial data rate of 2 GB/s. Those knowledge enable the design optimization of not only the sender/receiver servers with their storages but also the intermediate relay server system.

During the Large Helical Device (LHD) experiments, many scientists and technical staff are working in the central control room to operate the experiment. They must manage the diagnostics and controlling devices referring to the results of the last plasma shot. Also, the experiment coordinator must decide the conditions for the subsequent experiments using the results. Furthermore, many scientists are participating in the experiment from remote sites. Therefore, it is important to share the information in the control room quickly, such as the results of the last plasma discharge, with the remote user as well as with the staff in the room. In this paper, the data monitoring environment in the LHD central control room is introduced.

Upon beginning the LHD deuterium experiment, the opportunity for maintenance work in the torus hall will be conspicuously reduced such that all instruments must be controlled remotely. The LHD data acquisition (DAQ) and archiving system have been using about 110 DAQ front-end, and the DAQ central control and monitor system has been implemented for their remote management. This system is based on the “multi-agent” model whose communication protocol has been unified. Since DAQ front-end electronics would suffer from the “single-event effect” (SEE) of D-D neutrons, software-based remote operation might become ineffective, and then securely intercepting or recycling the electrical power of the device would be indispensable for recovering from a non-responding fault condition. In this study, a centralized control and monitor system has been developed for a number of power distribution units (PDUs). This system adopts the plug-in structure in which the plug-in modules can absorb the differences among the commercial products of numerous vendors. The combination of the above-mentioned functionalities has led to realizing the flexible and highly reliable remote control infrastructure for the plasma diagnostics and the device management in LHD.

The LHD data acquisition and archiving system, i.e., LABCOM system, has been fully equipped with high-speed real-time acquisition, streaming, and storage capabilities. To deal with more than 100 MB/s continuously generated data at each data acquisition (DAQ) node, DAQ tasks have been implemented as multitasking and multithreaded ones in which the shared memory plays the most important role for inter-process fast and massive data handling. By introducing a 10-second time chunk named “subshot,” endless data streams can be stored into a consecutive series of fixed length data blocks so that they will soon become readable by other processes even while the write process is continuing. Real-time device and environmental monitoring are also implemented in the same way with further sparse resampling. The central data storage has been separated into two layers to be capable of receiving multiple 100 MB/s inflows in parallel. For the frontend layer, high-speed SSD arrays are used as the GlusterFS distributed filesystem which can provide max. 2 GB/s throughput. Those design optimizations would be informative for implementing the next-generation data archiving system in big physics, such as ITER.

The NIFS (National Institute for Fusion Science) Atom and Molecular Database, which has been available online since 1997, is a numerical atomic and molecular database of collision processes that is important for fusion research. This database provides the following: (1) the cross-sections and rate coefficients for ionization, excitation, and recombination caused by electron impact, (2) the charge transfer caused by heavy particle collision and collision processes of molecules, and (3) the sputtering yields of solids and backscattering coefficients from solids. It also offers a bibliographic database. We recently reconstructed the database system. The main purpose of the reconstruction was to migrate the database into an open-source architecture to make the system more flexible and extensible. The previous system used proprietary software and was difficult to customize. The new system consists of open-source software, including PostgreSQL database and Ruby on Rails. New features were also added to the system. The most important improvement is the interface with the Virtual Atomic and Molecular Data Center (VAMDC) portal. Using this interface, researchers can search for data in the NIFS database as well as in various other online databases simultaneously.

“Fusion virtual laboratory (FVL)” is the experiments’ collaboration platform covering multiple fusion projects in Japan. Major Japanese fusion laboratories and universities are mutually connected through the dedicated virtual private network, named SNET, on SINET4. It has 3 different categories; (i) LHD remote participation, (ii) bilateral experiments’ collaboration, and (iii) remote use of supercomputer. By extending the LABCOM data system developed at LHD, FVL supports (i) and (ii) so that it can deal with not only LHD data but also the data of two remote experiments: QUEST at Kyushu University and GAMMA10 at University of Tsukuba. FVL has applied the latest “cloud” technology for both data acquisition and storage architecture. It can provide us high availability and performance scalability of the whole system. With a well optimized TCP data transferring method, the unified data access platform for both experimental data and numerical computation results could become realistic on FVL. The FVL project will continue demonstrating the ITER-era international collaboration schemes and the necessary technology.

This paper presents the control, data acquisition, data analysis and remote participation facilities of the Large Helical Device (LHD), which is designed to confine the plasma in steady state. In LHD the plasma duration exceeds 3000 s by controlling the plasma position, the density and the ICRF heating. The “LABCOM” data acquisition system takes both the short-pulse and the steady-state data. A two-layer Mass Storage System with RAIDs and Blu-ray Disk jukeboxes in a storage area network has been developed to increase capacity of storage. The steady-state data can be monitored with a Web browser in real time. A high-level data analysis system with Web interfaces is being developed in order to provide easier usage of LHD data and large FORTRAN codes in a supercomputer. A virtual laboratory system for the Japanese fusion community has been developed with Multi-protocol Label Switching Virtual Private Network Technology. Collaborators at remote sites can join the LHD experiment or use the NIFS supercomputer system as if they were working in the LHD control room.

The LABCOM data acquisition system (DAQ) has set a new world record of acquired data amount of 90 GB/shot in 2005–2006 campaign. Its CompactPCI-based new DAQ enables about 80 MB/s real-time acquisition for each plasma measurement. The basic performance of a DAQ unit has already good prospects for the next-generation fusion experiments, however, some surrounding utilities are not necessarily developed to construct 10 times larger DAQ system yet. The increasing number of parallel DAQ units has made the operational and maintenance burden very heavy. So, “more distributed acquisition and centralised operations” would become expected. Here, DAQ front-end consisting of some digitizer chassis, a timing module, and an acquisition computer has been entirely re-designed to realize a lower-cost and maintenance-free “DAQ box” for fusion plasma measurements. OS and softwares have been also remodelled to use free drivers on Linux. Network-bootable diskless computers are also desirable for reducing the possibility of hardware troubles, and it simultaneously realizes to omit first-tier data storage and reduce the migration operational cost. As a result, total hardware cost could be reduced almost one-fifth than the previous. This could be the way to realize 10 times larger fusion experiment in the near future.

To support the steady-state plasma experiments in LHD, new functions have been developed for the LABCOM data acquisition system. Using new WE7000 and PXI digitizers for continuous ultra-wideband data acquisition at 80 MB/s, the system is now able to perform data acquisition, storage, and server-to-client transfer fully in real-time. To manage the high amount of data for long-pulse experiments, the system has been modified to use 64-bit integer values for the number of samples, to segment the compressed data into multiple blocks by using the “titz” library, and to use the JPEG-LS algorithm for compressing image data. Its merit as complete solution for the next generation of fusion experiments has been demonstrated by faultlessly recording an 84 GB shot.

Control, data acquisition, plasma monitoring and remote participation for steady state operation in the large helical device (LHD) are reviewed. By controlling the impedance matching of ICH, the plasma position and the electron density, high temperature plasma is confined for 1905s. The plasma parameters are monitored in real time. Data are continuously sampled by the YOKOGAWA WE7000 system and by the NATIONAL INSTRUMENTS CompactPCI system. Those data are managed by the object-oriented database system based on ObjectStore in distributed servers with mass storage. By using the multi protocol label switching-virtual private network (MPLS-VPN) technology, the local area network of LHD is expanded to the Japanese fusion community. This provides the remote participants with the same environment of the LHD control room.

To deal with endless data streams acquired in LHD steady-state experiments, the LHD data acquisition system was designed with a simple concept that divides a long pulse into a consecutive series of 10-s “subshots”. Latest digitizers applying high-speed PCI-Express technology, however, output nonstop gigabyte per second data streams whose subshot intervals would be extremely long if 10-s rule was applied. These digitizers need shorter subshot intervals, less than 10-s long. In contrast, steady-state fusion plants need uninterrupted monitoring of the environment and device soundness. They adopt longer subshot lengths of either 10 min or 1 day. To cope with both uninterrupted monitoring and ultra-fast diagnostics, the ability to vary the subshot length according to the type of operation is required. In this study, a design modification that enables variable subshot lengths was implemented and its practical effectiveness in LHD was verified., 2012 The Japan Society of Plasma Science and Nuclear Fusion Research

A high-performance data acquisition (DAQ) system has been developed for steady-state fusion experiments at the Large Helical Device (LHD). Its significant characteristics are 110 MB s−1 continuous DAQ capability and the performance scalability using an unlimited number of DAQ units. Incoming data streams are first transferred temporarily onto the shared random access memory, and then cut into definite time chunks to be stored. They are also thinned out to 1/N to be served for the real-time monitoring clients. In LHD steady-state experiment, the DAQ cluster has established the world record for acquiring 90 GB/shot. The established technology of this steady-state acquisition and store can contribute to the ITER experiments whose data amount is estimated in the range 100 or 1000 GB/shot. This system also acquires experimental data from multiple remote sites through the fusion-dedicated virtual private network in Japan. The speed lowering problem in long-distance TCP/IP data transfer has been improved by the packet pacing optimization. The demonstrated collaboration scheme will be analogous to that of ITER and the supporting machines.

Plasma diagnostic data analysis often requires the original raw data as they are, in other words, at the same frame rate and resolution of the CCD camera sensor. As a non-interlace VGA camera typically generates over 70 MB/s video stream, usual frame grabber cards apply the lossy compression encoder, such as mpeg-1/-2 or mpeg-4, to drastically lessen the bit rate. In this study, a new approach, which makes it possible to acquire and store such the wideband video stream without any quality reduction, has been successfully achieved. Simultaneously, the real-time video streaming is even possible at the original frame rate. For minimising the exclusive access time in every data storing, it has adopted the directory structure to hold every frame files separately, instead of one long consecutive file. The popular ‘zip’ archive method improves the portability of data files, however, the JPEG-LS image compression is applied inside by replacing its intrinsic deflate/inflate algorithm that has less performances for image data.

A recent paper (Imazu, 1986) purports to demonstrate that the fusion reaction rates in high temperature plasmas decrease as the strength of the applied magnetic field rises. The magnitudes of the claimed effect is significant, the B=5T rate being only half the B=0 rate. If one leaves aside the question of the reacting particles being spin polarized, then it is well known (Jancel and Kahan, 1966) that a magnetic field has no effect whatever on the velocity distribution of particles in thermal equilibrium, hence it can have no effect on the fusion reaction rate. The error in the paper by Imazu can be found in equation 7 where a geometrical model of the particle orbits has been misinterpreted.

The purpose of this paper is to indicate theoretically how much the magnetic field influences the collision frequency for the nuclear fusion reaction. Since the Larmor radius, rL, of a particle in a usual magnetic field strength is large compared to the radius of the cross section, sigma , for the nuclear fusion reaction, the magnetic field does not affect the cross section. However, due to the gyration of the particle during a free path, the relative velocity is affected. First, the flow of charged particles across an area sigma per unit time is obtained, and then the collision frequency in the presence of the magnetic field is determined. It is shown that when 2 pi rL is small compared with the mean free path lambda , the velocity component contribution to the collision is the velocity component due to the motion along the field axis and not that due to the gyration.

Deals with thermal equilibrium between the particles in argon plasma arcs. The current limit for LTE is experimentally defined as a function of the pressures. Plasma parameters were spectroscopically determined. In particular, the electron temperature and heavy particle temperature were measured. The electron temperature was determined from an absolute continuum strength. Two temperatures were obtained for the heavy particles, one of which was determined from the Doppler component of the Halpha line when argon gas seeded with hydrogen was used. The second was obtained through a kinetic formula using the measured density and temperature of the electron. The results show that the electrons did not equilibrate with the heavy particles at low currents and pressures. The limit value of the current for LTE increased with decreasing gas pressure. It is shown, furthermore, that the measured field strength of the arc was remarkably lower than those calculated at currents less than the above limits.

This paper contains experimental results on partial LTE in the lower excited levels of Ar(I). The experiments were performed in a well-confined argon arc source. The population densities were obtained from self-absorption measurements of the lines. Plasma parameters were determined spectroscopically; measurements included the electron density (from the H β line width) and the electron temperature (from the Boltzmann plot for higher excited levels of Ar(I)). The results show that the levels 1 s 5 to 1 s 5 are in Saha-Boltzmann equilibrium for electron densities above 3 × 10 16 cm -3 .

The results of a theoretical study of the effect of a magnetic field, B, on the collision frequencies, Ff, for fusion reactions at comparatively low energies are presented. In a usual magnetic field strength, both the cross-section for fusion reactions and the cross-section for the Coulomb scattering between the ion-ion collisions are unaffected by B. However, due to both the gyration of the particle and the Coulomb scatterings occurring during a free path for fusion reactions, the relative velocity between colliding particles making the fusion reaction is affected by B. If omega is a gyrofrequency and tc the deflection time for a 90 degrees deflection of multiple Coulomb scattering when omega tc>2 pi , it was determined that the velocity component contribution to the collision for fusion reactions is only the velocity component along the field axis and that Ff decreases with increasing B. However, when omega tc

The line-intensity jump in decaying argon arcs has been investigated theoretically. The population density at the peak of the jump is in Saha-Boltzmann equilibrium. Using the Saha-Boltzmann relation, the influence of the change in electron density is clarified. As a criterion above which the change in electron density does not influence the amplitude of the jump, τ N /τ T ≧ 10 3 is derived, where τ N and τ T are, respectively, the density and temperature relaxation times for the electrons. This criterion is not satisfied in most arcs. An improved method is presented for the determination of the gas temperature in stationary arcs.

The data acquisition (DAQ) and management system of the Large Helical Device (LHD), named the LABCOM system, has been in development since 1995. The recently acquired data have grown to 7 gigabytes...

Helical instability of a positive column in a longitudinal magnetic field is investigated, under the assumptions that the perturbed densities have finite amplitudes and that the densities of electrons and ions are perturbed independently. A relationship between the perturbed densities of charged particles and the corresponding electric field is determined by solving Poisson's equation. Furthermore, the equations describing the behavior of steady‐state densities are readily solved by transforming them into the integral equation of Volterra's type. With this treatment, simplified equations of the radial distribution of the steady‐state density and potential are found as functions of the amplitude of the superimposed helix for the helical mode (m = 1) and the phase shift between the perturbed densities of electrons and ions. The results are compared with experimental data and found to be in fair agreement with experimental values of the radial density and potential distribution. The measurements also show the onset of modes 2, 3, and a low‐frequency oscillation as a function of magnetic field intensity.