Current Ramp-down Research Articles

An overview of the STEP divertor design and the simple models driving the plasma exhaust scenario

Abstract This paper presents a comprehensive overview of the preliminary divertor design and plasma exhaust scenario for the reactor-class Spherical Tokamak for Energy Production (STEP) project. Due to the smaller size of the machine, with a major radius less than half that of most DEMO concepts, the current design features a double-null divertor geometry, comprising tightly baffled extended outer legs and shorter inner legs approaching an X-divertor. Leveraging a significant database of SOLPS-ITER simulations, the exhaust operational space is mapped out, offering valuable insights into the plasma exhaust dynamics. An approach involving the validation of simple, yet robust models capable of accurately predicting key exhaust parameters is detailed, thereby streamlining the design process. The simple models are used to simulate the entire plasma scenario from the plasma current ramp-up, through the burning phase, to the plasma current ramp-down. Notably, the findings suggest that pronounced detachment, with peak heat loads below engineering limits and electron temperatures below 5 eV, is achievable with a divertor neutral pressure between 10 - 15 Pa during
the burning phase, and pressures below 5 Pa during the ramp-up to maximise the auxiliary current-drive efficiency. Throughout the scenario, an Ar concentration of ~3% in the scrape-off layer (SOL) is required, in combination with a core radiation fraction of 70% driven by intrinsic emission and extrinsic injection of Xe seeded fuelling pellets. However, significant uncertainties remain regarding key parameters such as the SOL heat flux width, Ar screening, and plasma kinetic effects.

Simulations of the stationary Q = 10 and the exit phase from the flat-top of an ITER 15MA baseline scenario: predictive JINTRAC simulation with a consistent treatment of D and T in the whole plasma

Abstract Designing a robust termination scenario for a burning ITER plasma is a challenge that requires extensive core plasma and divertor modelling. The presented work consists of coupled core/edge/SOL/divertor simulations, performed with the JINTRAC code, to study the Q = 10 flat-top phase and exit phase of the ITER 15 MA/5.3 T DT scenario. The modelling utilizes the recently implemented option to treat deuterium and tritium separately in the SOL/divertor, enabling a consistent treatment of deuterium and tritium in the whole plasma volume, which is a unique capability of JINTRAC. In addition, these are the first JINTRAC simulations of this scenario that use a first-principles transport model to self-consistently model the ECRH power deposition and to include tungsten while keeping track of tungsten sputtering and accumulation. The flat-top simulations demonstrate the possibility of sustaining a steady state fusion Q of 10 using pure deuterium gas puffs together with DT mixed pellets, which is an option to make a more effective use of tritium. Simulations of the exit phase are set up sequentially, with each phase providing initial conditions for the next, starting with a density decay at full current and auxiliary power, and demonstrate the possibility of reducing the density robustly within a few seconds. Following the density decay, a subsequent auxiliary power ramp-down in H-mode is performed with a late H–L transition at low auxiliary power, which may provide an option for the optimization of the plasma termination. The final ramp-down phase consists of a current ramp-down in L-mode to 3.75 MA.

Density limits as disruption forecasters for spherical tokamaks

Fusion power output from spherical tokamaks would benefit from increased confined plasma density, but there exists a limit on the density before confinement is lost and the plasma current is disrupted. This density limit has long been characterized by a simple, global Greenwald limit proportional to the plasma current and inversely proportional to the cross sectional area of the plasma. It is shown that in the database of discharges from the National Spherical Tokamak Experiment (NSTX) and Mega Ampere Spherical Tokamak (MAST) spherical tokamaks, the likelihood of disruption does increase above the Greenwald limit, and especially in the plasma current rampdown phase. The physics of the density limit has been recently theoretically explored through local criteria. Several of these are tested using the disruption event characterization and forecasting (DECAFTM) code for their potential effectiveness as disruption warning signals. For a limited set of NSTX discharges, a local island power balance criteria was found to be less reliable, presently, than the Greenwald limit. An empirical critical edge line density and a boundary turbulent transport limit were both tested for MAST-U, which has an electron density profile measurement with high spatial resolution in the outer part of the plasma. Both were found to have similar dependencies on key plasma parameters. In a limited set of MAST-U discharges that appear to disrupt due to rising density at values under the Greenwald limit, crossing of the boundary turbulent transport limit occurred close to the time of disruption. Finally, these limits were evaluated for their potential use in real-time, and it was found that with the necessary real-time inputs and with refinement through further testing, these limits could be implemented in a real-time disruption forecasting system.

JINTRAC integrated simulations of ITER scenarios including fuelling and divertor power flux control for H, He and DT plasmas

We have modelled self-consistently how to most efficiently fuel ITER hydrogen (H), helium (He) and deuterium–tritium (DT) plasmas with gas and/or pellets with the integrated core and 2D SOL/divertor suite of codes JINTRAC. This paper presents the first overview of full integrated simulations from core to divertor of ITER scenarios following their evolution from X-point formation, through L-mode, L–H transition, steady-state H-mode, H–L transition and current ramp-down. Our simulations respect all ITER operational limits, maintaining the target power loads below 10 MW m−2 by timely gas fuelling or Ne seeding. For the pre-fusion plasma operation (PFPO) phase our aim was to develop robust scenarios and our simulations show that commissioning and operation of the ITER neutral beam (NB) to full power should be possible in 15 MA/5.3 T L-mode H plasmas with pellet fuelling and 20 MW of ECRH. For He plasmas gas fuelling alone allows access to H-mode at 7.5 MA/2.65 T with 53–73 MW of additional heating, since after application of NB and during the L–H transition, the modelled density build-up quickly reduces the NB shine-through losses to acceptable levels. This should allow the characterisation of ITER H-mode plasmas and the demonstration of ELM control schemes in PFPO-2. In ITER DT plasmas we varied the fuelling and heating schemes to achieve a target fusion gain of Q = 10 and to exit the plasma from such conditions with acceptable divertor loads. The use of pellets in DT can provide a faster increase of the density in L-modes, but it is not essential for unrestricted NB operation due to the lower shine-through losses compared to H. During the H–L transition and current ramp-down, gas fuelling and Ne seeding are required to keep the divertor power loads under the engineering limits but accurate control over radiation is crucial to prevent the plasma becoming thermally unstable.

Beta-induced Alfvén eigenmodes and geodesic acoustic modes in the presence of strong tearing activity during the current ramp-down on JET

The analysis of the current ramp-down phase of JET plasmas has revealed the occurrence of additional magnetic oscillations in pulses characterized by large magnetic islands. The frequencies of these oscillations range from 5 to , being well below the toroidal gap in the Alfvén continuum and of the same order as the low-frequency gap opened by plasma compressibility. The additional oscillations only appear when the magnetic island width exceeds a critical threshold, suggesting that the oscillations could tap their energy from the tearing mode (TM) by a non-linear coupling mechanism. A possible role of fast ions in the excitation process can be excluded, being the pulse phase considered in the observations characterized by very low additional heating. The calculation of the coupled Alfvén–acoustic continuum in toroidal geometry suggests the possibility of beta-induced Alfvén eigenmodes (BAEs) rather than beta-induced Alfvén–acoustic eigenmodes. As a main novelty compared to previous work, the analysis of the electron temperature profiles from electron cyclotron emission has shown the simultaneous presence of magnetic islands on different rational surfaces in pulses with multiple magnetic oscillations in the low-frequency gap of the Alfvén continuum. This observation supports the hypothesis of different BAE with toroidal mode number n = 1 associated with different magnetic islands. As another novelty, the observation of magnetic oscillations with n = 2 in the BAE range is reported for the first time in this work. Some pulses, characterized by slowly rotating magnetic islands, exhibit additional oscillations with n = 0, likely associated with geodesic acoustic modes (GAMs), and a cross-spectral bicoherence analysis has confirmed a non-linear interaction between TM, BAE and GAM, with the novelty of the observation of multiple triplets (twin BAEs plus GAM), due to the simultaneous presence of several magnetic islands in the plasma.

Avoidance of impurity-induced current quench using lower hybrid current drive

This work reports observations of a tokamak plasma that experienced a thermal quench due to a large, transient high-Z influx but avoided a current quench. This is argued to be caused by the presence of lower hybrid range of frequency (LHRF) waves that sustain a non-thermal, current-carrying electron population. In Alcator C-Mod L-mode plasmas at kA, m−3, nearly all of the current can be sustained non-inductively by injecting 700 kW of LHRF power at 4.6 GHz and . A sudden influx of a large amount of tungsten, , triggers a cooling wave that propagates at 2–3 m s−1 all the way into the core, dropping on-axis Te from 3 keV to temperatures less than measurement floor of 50 eV. An off-axis reheat begins after 100 ms, but Te profiles remain hollow for 300–350 ms after the injection. Throughout this temperature evolution, the plasma density, current and shape remain unchanged to within 10%. Following the expulsion of the tungsten, the plasma returns to its baseline conditions and the plasma ends as planned with a controlled current ramp-down. Energy balance analysis shows the LHRF power continues to be absorbed in the plasma after the thermal quench, as a significant fraction of it is needed to be consistent with radiated power measurements. Examination of current relaxation time, , and fast-electron slowing down time, , indicate the LHRF must contribute to driving current, despite the low temperatures, as the current remains nominally stationary despite ms and ms for relativistic electrons. These measurements represent an important existence proof of a possible technique for avoidance of disruptions caused by sudden, unplanned influx of impurities in the form of dust or flakes of high-Z wall material. Implications and suggestions for future experimental and modeling and simulation work are summarized.

Estimate of 3D power wall loads due to Neutral Beam Injection in EU DEMO ramp-up phase

Heating and current drive systems such as high energy Neutral Beam Injection (NBI) are being considered for pulsed EU DEMO (“DEMO1”) pre-conceptual design. Their aim is to provide auxiliary power, not only during flat-top, but also during transient phases (i.e. plasma current ramp-up and ramp-down).In this work, NBI fast particle power loads on DEMO1 first wall, due to shine-through and orbit losses, are calculated for the diverted plasma ramp-up phase. Numerical simulations are performed using BBNBI and ASCOT Monte Carlo codes. The simulations have been done using a complete 3D wall geometry, and implementing the latest DEMO NBI design, which foresees NBI at 800 keV particle energy. Location and power density of NBI-related power loads at different ramp-up time steps are evaluated and compared with the maximum tolerable heat flux taken from ITER case. Since NBI shine-through losses (dominant during low density phases) depend mainly on the beam energy, plasma density and volume, DEMO has a more favourable situation than ITER, enlarging NBI operational window. Using ITER criteria, DEMO NBI at full energy and power could be switched on during ramp-up at <ne> ~ 1.3 × 1019 m-3. This increases the appeal of neutral beam injectors as auxiliary power systems for DEMO.

Radiation diagnostics for plasma current ramp-up and ramp-down research.

The plasma current ramp-up and ramp-down are the basic processes in the tokamak operation. In order to research these processes in SUNIST (Sino-UNIted Spherical Tokamak), some diagnostic systems that detect the plasma radiation ranging from hard X-rays to visible light are developed. CdZnTe and silicon drift detectors measure the energy spectrum of hard X-rays and soft X-rays coming from the plasma. A pinhole camera equipped with absolute extended ultraviolet array photodiodes has been installed on the top of SUNIST to observe the radiation power loss and the magneto-hydrodynamic activities with high temporal and spatial resolution. The spectrum of vacuum ultraviolet is acquired by using a CCD camera, and the intensity of the lines can be measured by using a photomultiplier tube with a scintillator. The full spectrum of the visible light can be acquired in every 3 ms, and the intensity of some lines, such as Hα, Hγ, can be measured by filter scopes with high time response. Additionally, a Doppler broadening measurement system is developed to measure the ion temperature of edge plasma.

Multi-machine analysis of termination scenarios with comparison to simulations of controlled shutdown of ITER discharges

To improve our understanding of the dynamics and control of ITER terminations, a study has been carried out on data from existing tokamaks. The aim of this joint analysis is to compare the assumptions for ITER terminations with the present experience basis. The study examined the parameter ranges in which present day devices operated during their terminations, as well as the dynamics of these parameters. The analysis of a database, built using a selected set of experimental termination cases, showed that, the H-mode density decays slower than the plasma current ramp-down. The consequential increase in fGW limits the duration of the H-mode phase or result in disruptions. The lower temperatures after the drop out of H-mode will allow the plasma internal inductance to increase. But vertical stability control remains manageable in ITER at high internal inductance when accompanied by a strong elongation reduction. This will result in ITER terminations remaining longer at low q (q95 ~ 3) than most present-day devices during the current ramp-down. A fast power ramp-down leads to a larger change in βp at the H–L transition, but the experimental data showed that these are manageable for the ITER radial position control. The analysis of JET data shows that radiation and impurity levels significantly alter the H–L transition dynamics. Self-consistent calculations of the impurity content and resulting radiation should be taken into account when modelling ITER termination scenarios. The results from this analysis can be used to better prescribe the inputs for the detailed modelling and preparation of ITER termination scenarios.

Simulation of profile evolution from ramp-up to ramp-down and optimization of tokamak plasma termination with the RAPTOR code

The present work demonstrates the capabilities of the transport code RAPTOR as a fast and reliable simulator of plasma profiles for the entire plasma discharge, i.e. from ramp-up to ramp-down. This code focuses, at this stage, on the simulation of electron temperature and poloidal flux profiles using prescribed equilibrium and some kinetic profiles. In this work we extend the RAPTOR transport model to include a time-varying plasma equilibrium geometry and verify the changes via comparison with ATSRA code simulations. In addition a new ad hoc transport model based on constant gradients and suitable for simulations of L–H and H–L mode transitions has been incorporated into the RAPTOR code and validated with rapid simulations of the time evolution of the safety factor and the electron temperature over the entire AUG and TCV discharges. An optimization procedure for the plasma termination phase has also been developed during this work. We define the goal of the optimization as ramping down the plasma current as fast as possible while avoiding any disruptions caused by reaching physical or technical limits. Our numerical study of this problem shows that a fast decrease of plasma elongation during current ramp-down can help in reducing plasma internal inductance. An early transition from H- to L-mode allows us to reduce the drop in poloidal beta, which is also important for plasma MHD stability and control. This work shows how these complex nonlinear interactions can be optimized automatically using relevant cost functions and constraints. Preliminary experimental results for TCV are demonstrated.

Heating & current drive efficiencies, TBR and RAMI considerations for DEMO

The heating & current drive (H&CD) systems in a DEMOnstration fusion power plant are one of the major energy consumers. Due to its high demand in electrical energy the H&CD efficiency optimization is an important goal in the DEMO development.The H&CD power for DEMO, based on physics scenarios for the different plasma phases, is needed for plasma initiation phases (incl. breakdown), current ramp-up, heating to H-mode, burn control, controlled current ramp-down, MHD control and other functions. Plasma control will need significant installed H&CD power, though not continuously used.Previously, in the DEMO1 2015 baseline definitions, optimistic forecasted H&CD efficiencies had been assumed in the corresponding system code (i.e. PROCESS) module. Realizing that there is a high uncertainty in the assumptions the efficiencies have been modified and the impact on the DEMO power plant and basic tokamak configuration are discussed in this article.A comparison of the various H&CD systems NBI (Neutral Beam Injection), Electron Cyclotron (EC), Ion Cyclotron (IC) in terms of impact on Tritium Breeding Ratio (TBR) due to various openings for the H&CD front end components in the breeding blanket (BB) is presented.For increasing the reliability as major features the power per system unit and the redundancy are identified leading to a new proposal for clusters for EC and modular ion-sources for NB.

Joule Losses in the ITER Cold Structures During Plasma Transients

The ITER magnet system, which is composed of toroidal field (TF) and poloidal field (PF) coils, central solenoid (CS), correction coils, and all their structural supports, is cooled by supercritical helium flow at 4.5 K. Eddy currents are induced in the metallic components during normal operation, due to the variation of PF and CS coils current, and to the plasma ramp-up and ramp-down. In addition, during disruptions, as well as other fast plasma transient events, the eddy currents circulating in these structures reach very high values due to the high induced electric fields. An effective cooling is therefore needed to limit the temperature increase of the magnets and their supports. The Joule energy dissipated in the cold structures of the ITER magnet system has been computed by means of the electromagnetic finite-element code CARIDDI. A model of a 40 degree sector of the ITER magnet has been built in order to represent in detail the connections between the CS and the TF coils, PF and CS supports, and all the intercoil structures and the metallic part of the TF coils (radial plates, case, etc.). Vacuum vessel, thermal shield, and cryostat have also been modeled. The reference 15-MA inductive scenario (with plasma current ramp-up duration of 80 s and ramp-down of 200 s) and a 15-MA fast inductive scenario characterized by the fastest plasma current ramp-up (50 s) and the fastest plasma current ramp-down (65 s) have been analyzed. In addition, several plasma instabilities have been simulated considering the reference electrical connections between the metallic parts.

Discriminating the trapped electron modes contribution in density fluctuation spectra

Quasi-coherent (QC) modes have been reported for more than 10 years in reflectometry fluctuations spectra in the core region of fusion plasmas. They have characteristics in-between coherent and broadband fluctuations as they oscillate at a marked frequency but have a wide spectrum. This work presents further evidences of the link recently established between QC modes and the trapped electron modes (TEM) instabilities (Arnichand et al 2014 Nucl. Fusion 54 123017). In electron cyclotron resonance heated discharges of Tore Supra, an enhancement of QC modes amplitude is observed in a region where TEM cause impurity transport and turbulence. In JET Ohmic plasmas, QC modes disappear during density ramp-up and current ramp-down. This is reminiscent of Tore Supra and TEXTOR observations during transitions from the linear Ohmic confinement (LOC) to the saturated Ohmic confinement (SOC) regimes. Evidencing TEM activity then becomes experimentally possible via analysis of fluctuation spectra.

Plasma vertical stabilisation in ITER

This paper describes the progress in analysis of the ITER plasma vertical stabilisation (VS) system since its design review in 2007–2008. Two indices characterising plasma VS were studied. These are (1) the maximum value of plasma vertical displacement due to free drift that can be stopped by the VS system and (2) the maximum root mean square value of low frequency noise in the dZ/dt measurement signal used in the VS feedback loop. The first VS index was calculated using the PET code for 15 MA plasmas with the nominal position and shape. The second VS index was studied with the DINA code in the most demanding simulations for plasma magnetic control of 15 MA scenarios with the fastest plasma current ramp-up and early X-point formation, the fastest plasma current ramp-down in a divertor configuration, and an H to L mode transition at the current flattop. The studies performed demonstrate that the VS in-vessel coils, adopted recently in the baseline design, significantly increase the range of plasma controllability in comparison with the stabilising systems VS1 and VS2, providing operating margins sufficient to achieve ITER's goals specified in the project requirements. Additionally two sets of the DINA code simulations were performed with the goal of assessment of the capability of the PF system with the VS in-vessel coils: (i) to control the position of runaway electrons generated during disruptions in 15 MA scenarios and (ii) to trigger ELMs in H-mode plasmas of 7.5 MA/2.65 T scenarios planned for the early phase of ITER operation. It was also shown that ferromagnetic structures of the vacuum vessel (ferromagnetic inserts) and test blanket modules insignificantly affect the plasma VS.

ITER-like current ramps in JET with ILW: experiments, modelling and consequences for ITER

Since the ITER-like wall in JET (JET-ILW) came into operation, dedicated ITER-like plasma current (Ip) ramp-up (RU) and ramp-down (RD) experiments have been performed and matched to similar discharges with the carbon wall (JET-C). The experiments show that access to H-mode early in the Ip RU phase and maintaining H-mode in the Ip RD as long as possible are instrumental to achieve low internal plasma inductance (li) and to minimize flux consumption. In JET-ILW, at a given current rise rate similar variations in li (0.7–0.9) are obtained as in JET-C. In most discharges no strong W accumulation is observed. However, in some low density cases during the early phase of the Ip strong core radiation due to W influx led to hollow electron temperature (Te) profiles. In JET-ILW Zeff is significantly lower than in JET-C. W significantly disturbs the discharge evolution when the W concentration approaches 10−4; this threshold is confirmed by predictive transport modelling using the CRONOS code. Ip RD experiments in JET-ILW confirm the result of JET-C that sustained H-mode and elongation reduction are both instrumental in controlling li.

Shape Control with the eXtreme Shape Controller During Plasma Current Ramp-Up and Ramp-Down at the JET Tokamak

The eXtreme Shape Controller (XSC) has been originally designed to control the plasma shape at JET during the flat-top phase, when the plasma current has a constant value. During the JET 2012 experimental campaigns, the XSC has been used to improve the shape control during the transient phases of plasma current ramp-up and ramp-down. In order to avoid the saturation of the actuators with these transient phases, a current limit avoidance system has been designed and implemented. This paper presents the experimental results achieved at JET during the 2012 campaigns using the XSC.

Development of the ITER baseline inductive scenario

Sustainment of Q ∼ 10 operation with a fusion power of ∼500 MW for several hundred seconds is a key mission goal of the ITER Project. Past calculations and simulations predict that these conditions can be produced in high-confinement mode operation (H-mode) at 15 MA relying on only inductive current drive. Earlier development of 15 MA baseline inductive plasma scenarios provided a focal point for the ITER Design Review conducted in 2007–2008. In the intervening period, detailed predictive simulations, supported by experimental demonstrations in existing tokamaks, allow us to assemble an end-to-end specification of this scenario consistent with the final design of the ITER device. Simulations have encompassed plasma initiation, current ramp-up, plasma burn and current ramp-down, and have included density profiles and thermal transport models producing temperature profiles consistent with edge pedestal conditions present in current fusion experiments. These quasi-stationary conditions are maintained due to the presence of edge-localized modes that limit the edge pressure. High temperatures and densities in the pedestal region produce significant edge bootstrap current that must be considered in modelling of feedback control of shape and vertical stability. In this paper we present new results of transport simulations fully consistent with the final ITER design that remain within allowed limits for the coil system and power supplies. These self-consistent simulations increase our confidence in meeting the challenges of the ITER program.

Disruptions, disruptivity and safer operating windows in the high-β spherical torus NSTX

This paper discusses disruption rates, disruption causes and disruptivity statistics in the high-βN National Spherical Torus Experiment (NSTX) (Ono et al 2000 Nucl. Fusion 40 557). While the overall disruption rate is rather high, configurations with high βN, moderate q*, strong boundary shaping, sufficient rotation and broad pressure and current profiles are found to have the lowest disruptivity; active n = 1 control further reduces the disruptivity. The disruptivity increases rapidly for q* < 2.7, which is substantially above the ideal MHD current limit. Under quiescent conditions, qmin > 1.25 is generally acceptable for avoiding the onset of core rotating n = 1 kink/tearing modes; when EPM and ELM disturbances are present, the required qmin for avoiding those modes is raised to ∼1.5. The current ramp and early flat-top phase of the discharges are prone to n = 1 core rotating modes locking to the wall, leading to a disruption. Small changes to the discharge fuelling during this phase can often mitigate the rotation damping associated with these modes and eliminate the disruption. The largest stored-energy disruptions are those that occur at high current when a plasma current ramp-down is initiated incorrectly.

Current ramps in tokamaks: from present experiments to ITER scenarios

In order to prepare adequate current ramp-up and ramp-down scenarios for ITER, present experiments from various tokamaks have been analysed by means of integrated modelling in view of determining relevant heat transport models for these operation phases. A set of empirical heat transport models for L-mode (namely, the Bohm–gyroBohm model and scaling based models with a specific fixed radial shape and energy confinement time factors of H96−L = 0.6 or HIPB98 = 0.4) has been validated on a multi-machine experimental dataset for predicting the li dynamics within ±0.15 accuracy during current ramp-up and ramp-down phases. Simulations using the Coppi–Tang or GLF23 models (applied up to the LCFS) overestimate or underestimate the internal inductance beyond this accuracy (more than ±0.2 discrepancy in some cases). The most accurate heat transport models are then applied to projections to ITER current ramp-up, focusing on the baseline inductive scenario (main heating plateau current of Ip = 15 MA). These projections include a sensitivity study to various assumptions of the simulation. While the heat transport model is at the heart of such simulations (because of the intrinsic dependence of the plasma resistivity on electron temperature, among other parameters), more comprehensive simulations are required to test all operational aspects of the current ramp-up and ramp-down phases of ITER scenarios. Recent examples of such simulations, involving coupled core transport codes, free-boundary equilibrium solvers and a poloidal field (PF) systems controller are also described, focusing on ITER current ramp-down.

Augmentation of Current Ramp-Down Contact Pressure in C-Shaped Armature Railguns

It is necessary to maintain adequate contact pressure between a rail and an armature to avoid transition in a solid armature railgun. The contact pressure of the C-shaped armature at start-up arises from its initial mechanical interference. As armature wear grows and temperature rises, the initial mechanical interference decreases. When the current is established, the electromagnetic (EM) force provides the contact pressure. Transition usually occurs when the current ramps down, except for relatively light situations that have low currents and speeds. In this paper, launch experiments with armatures of the same geometry and material under different driving current amplitudes were performed. Based on the experimental results, the temperature rise of the armature with increasing action was analyzed. Then, the preload during the period of temperature increase was analyzed. In a simplified model, the EM force was also analyzed. The results indicate that the sum of preload and EM force has shown a decreasing trend. To reduce the consequence of EM force that decreases with current ramp-down, a partially augmented railgun, which could maintain adequate contact pressure in current ramp-down, is proposed. The feasibility of this kind of railgun is also analyzed.