Abstract
Introduction
A.C. contactor is a common mechanical switch, which is widely used in motor control, power switching, and transformer control. Traditional A.C. contactors have disadvantages such as uncontrollable working processes, slow response, and large contact bounce.1,2 Some new technical proposals have been proposed to solve these problems, including permanent magnet contactor,3,4 semiconductor contactors, 5 and composite contactors.6,7
Although the permanent magnet contactor has great advantages in energy consumption and noise, the working mode of dynamic and static core suction makes it difficult to completely eliminate the contact bounce phenomenon in the closing process, and when the control circuit failure or demagnetization coil failure occurs, the permanent magnet contactor can not be disconnected normally. 8 Because semiconductor contactors need to support a wide range of current, this current tends to reach 5–10 times the nominal current in the duration of milliseconds to seconds, during which the conduction loss and the temperature of the semiconductor die increase very rapidly. This may lead to overheating, damage, or premature aging of semiconductor devices, and because it is made up entirely of power electronic devices, the semiconductor contactor cannot be reset manually in the event of failure.9,10 As for the compound contactor, if the thyristor breakdown occurs, it will not be able to break the load or cause the load to run out of phase, with serious consequences. At the same time, because the compound contactor has higher requirements for control technology, the complexity of the system increases and the reliability decreases.
Smart material direct-drive contactors, which is a new research direction of intelligent contactor. Wang et al. 11 designed a novel type of drop-out fuse that can be repeatedly operated without the need for fuse replacement utilizing the deformation function of shape memory alloy with temperature. This design enhances both safety and convenience. The German company ETO proposed a conceptual product MAGNETOSHAPE® Circuit Breaker, 12 which can achieve a response speed of 400 µs. ABB’s Tüysüz et al. 13 developed a small magnetic shape memory alloy linear actuator based on the design requirements of miniature circuit breakers. Linnebach et al. 14 designed a contactor with a dielectric elastomer as the drive element. The operating mechanism can reach 524.2 mm/s, the displacement can reach 1.8 mm, and the power dissipation of maintaining switching-in is only 0.015 W under high voltage.
The MSMA is a late-emerging smart metal material. Its dependent variable is up to 6% compared to giant magnetostrictive material (GMM) and piezoelectric materials (PZT). It has an ms-level response speed15,16 compared with traditional shape memory alloys (SMAs). Therefore, the MSMA is an excellent driver component. It is very suitable to be used as the driving element of contactor. The current research on MSMA drives primarily focuses on implementation methods and feasibility verification, as well as the exploration of various control algorithms to optimize output performance. 17 There is relatively little research on actual loads and specific application objects.
An A.C. contactor for low-voltage electrical control driven by the MSMA was designed in the work. First, the material strain mechanism of the MSMA was introduced, and an experimental platform was built to test the relevant characteristics of the material. The overall structure of the contactor was designed based on the output characteristics of the MSMA and the design requirements of the contactor. The switching-in characteristics of the contactor under the drive signal were analyzed through simulation. Finally, an experimental platform was built to test the performance of the contactor prototype during separating brake and switching in.
MSMA material characteristics
Deformation mechanism
The magnetic-field-induced strain of the MSMA is caused by the glide reflection of the martensite twin and the reorientation of the martensitic phase under the external magnetic field at the microscopic scale. The easy axis of magnetization (small arrow in Figure 1(a)) of the material deflects in alignment with the direction of the external magnetic field. The easy axis of magnetization (

MSMA deformation mechanism: (a)
Characteristic test
The material tested in this section is a sample with a size of 2 × 3 × 30 mm obtained by bonding two 2 × 3 × 15 mm MSMA rods produced by the German company ETO. The rated working magnetic induction intensity is 0.6 T and the Austenite transformation temperature is 60°C. Figure 2(a) lists the material sample, and the material test is carried out in the test device (Figure 2(b)).

Material test: (a) test sample and (b) test device.
Figure 3 shows the test platform. The material elongation process applies varying degrees of prepressure to the MSMA by adjusting the pressure adjusting nut, while the precision DC power supply outputs different sizes of drive currents to regulate the size of the excitation magnetic field. The pressure is controlled by a digital push and pull tester in the process of material strain recovery.

Material-performance testing platform.
Figure 4 illustrates the test results of the material elongation process. The strain of the MSMA increases with the increase of magnetic induction intensity before maximum strain, and the maximum strain of MSMA decreases with the increased load. The strain of the MSMA varies less between 0 and 200 mT and 400 and 600 mT under 0 N load, and it varies more between 200 and 400 mT, which is close to linear variation. It does not change after 600 mT, and the elongation of the sample is 5.9% at this time. Meanwhile, the increase in pre-pressure inhibits the maximum strain of the material, and the material exhibits a low-strain recovery.

MSMA strain under different magnetic induction intensity and pre-pressure.
Figure 5 presents the recovery process of material strain. The pressure stress grows faster at the initial stage of strain recovery. That is, the MSMA sample unit has a strong ability to maintain recovery, and the unit pressure stress can only cause a smaller strain recovery. The situation is the opposite in the middle stage of strain recovery. The sample’s retention capacity is inadequate. A small increase in pressure stress can lead to a larger strain recovery of the sample when the pressure stress exceeds 6 N. The retention capacity at the end of the strain recovery is stronger than the initial stage, and the strain recovery caused by unit pressure stress is subtle.

Relationship between the strain recovery and pressure stress.
Design of contactors
Switching principle
Contactor, a device for circuit on-off control, has the function to close and separate the moving and static contacts. The length of the MSMA that has been strained under a magnetic does not spontaneously recover since the MSMA has a magnetic shape memory effect. There are two ways to achieve deformation recovery: by force or magnetism. It requires a huge magnetic field and a complex magnetic structure if by magnetism. Therefore, force is often used to achieve material strain recovery in the practical application of MSMA.
Spring is used to provide force to simplify the structure of most MSMA drives. However, the prepressure of the spring is not conducive to the strain and response speed of the MSMA material. Therefore, MSMA is applied to the operating mechanism of contactors, and electromagnetic force is selected to provide the recovery force of material strain.
Figure 6 presents the strain generation and strain recovery of the material. Figure 6(a) to (c) present the material elongation stage, and Figure 6(c) to (e) show the material recovery stage. The reciprocating adjustment of the deformation of the MSMA can be achieved by applying different pulsed magnetic fields, and then the movement can achieve the separation and closing of the contactor. The mechanism responsible for generating the drive magnetic field that induces strain in the MSMA is mentioned as the driving mechanism, while the mechanism enabling recovery of the MSMA from such strain is known as the recovery mechanism.

Material deformation and recovery process: (a) initial state, (b) magnetic-field-induced strains, (c) strain retention, (d) strain recovery by magnetism, and (e) strain recovery.
Design of driving mechanism
A strong magnetic field is required to produce obvious strain known from material testing. Unnecessary losses such as magnetic flux leakage, eddy current loss, and magnetic hysteresis loss need to be reduced to minimize the energy loss during excitation. Therefore, DT4C with high magnetic permeability and low coercivity is selected as the core material. The magnetic circuit structure excited by MSMA in the driving mechanism is designed as a double “E”-shaped structure (Figure 7).

Magnetic circuit structure.
Gap reluctance is considered exclusively in magnetic circuits since the relative permeability of DT4C is much larger than that of air. The total reluctance of the magnetic circuit can be expressed as follows.
where
The air-gap clearance flowing through the magnetic circuit in the equivalent magnetic circuit is related to the magnetic flux, magnetomotive force, and the reluctance of the magnetic core according to magnetic ohm’s law.
Where
The coil can be regarded as an equivalent voltage source in Kirchhoff’s voltage law. Therefore, there is the following relationship between the magnetomotive force in the equivalent magnetic circuit and the turns per coil and coil current analogous to law:
where
The magnetic induction at the magnetic air gap is as follows according to the equation of the magnetic induction.
The turns per coil is
where
Software COMSOL is utilized to simulate the magnetic field intensity to ensure the reliability of the design. Figure 8 shows the simulation results. The magnetic field intensity vector lines generated by the two drive coils form a closed-loop circuit and are concentrated in the middle of the iron core. Therefore, the magnetic flux in the magnetic circuit gradually decreases from the inside to the outside, but the overall distribution is uniform. Besides, the magnetic leakage is less, and the magnetic induction in the area where MSMA is located can reach 0.63 T, which meets the design requirements.

Distribution of magnetic induction.
Recovery mechanism design
Figure 9 presents the magnetic circuit composed of moving and static iron cores in the recovery structure. A setup spring with an elasticity coefficient of

Recovery magnetic circuit structure.
The relationship between iron core displacement
Electromagnetic force
where
The stress surges with the increase of the strain recovery in the last short period of strain recovery from the stress-strain curve. The strain of 0.3 mm is reserved to avoid unnecessary energy consumption. That is, the strain recovery is between 0.3 and 1.5 mm in Figure 5. The MSMA produces 1.2 mm strain during the switching-in process, and air gap
The magnetic induction at the air gap in the magnetic circuit is as follows referring to the calculation in Section “Characteristic test.”
where
Resultant force
The downward resultant force of the moving iron core in the initial state is 1.93 N when
Overall structure of the contactor
The mode of motion of contact is designed as a direct-drive type by considering the deformation characteristics of MSMA. Meanwhile, one set of contacts is designed and the arc-control device is omitted to simplify the structure since this design mainly explores the application of the MSMA in contactors. Figure 10 lists the overall structure of the final MSMA A.C. contactor.

Overall structure of the contactor.
Contactor dynamics simulation
Mathematical modeling
The following assumptions are made before establishing the model: (1) Strain

Equivalent mechanical model.
The MSMA is subject to the pre-pressure
Output stress
where
Kiefer and Lagoudas proposed the evolution of MSMA reorientation strain and the function hypothesis that controls reorientation or the activation of the domain wall motion process. 21 The relationship between strain and magnetic field intensity at the elongation stage of MSMA is as follows.
where
The transfer function from the displacement
where
Based on the relationship between the drive current and the magnetic field described in Section “Characteristic test,”
Model simulation
A simulation module (Figure 12) of the displacement response of the moving contact is built in Simulink based on the above model under the step signal during the switching-in of the contactor.

Simulation module of displacement response.
The step response of the model can be obtained (Figure 13) when the drive coil inputs 6-A currents. The step curve of displacement is stable without oscillation from the simulation results. The displacement process of the moving contact is stable under the MSMA drive and no vibration occurs for the contactor. Meanwhile, the entire displacement time of the moving contact is about 5 ms, which can realize the high-speed switching-in.

Step response of moving contact displacement.
Experimental measurement
Experimental system construction
An experimental test platform (Figure 14) is built to perform experimental tests on the performance of the contactor. It measures the response speed of the contactor, the displacement curve of the moving contact, and whether the contact bounce will occur during the switching-closing when the moving and static contacts are in contact. Response time and energy consumption during the switching-closing process, as well as the contact effect of moving and static contacts during switching-in, can be analyzed by measuring changes in voltage signals (major loop signals) at both ends of the load in the major loop of the contactor. Changes in voltage signals (drive signal and reply signal) at both ends of the coil can also be measured. The displacement curve of the moving contact in the process of switching-closing is obtained by a laser displacement sensor.

Experimental test platform.
The current output from the precision DC power supply is connected to both ends of the drive and recovery coil through the ship-type switch to realize the switching-closing of the contactor in the measurement of the drive signal. The experiment uses a function generator as the AC power supply in the major loop to output 5-V 50-Hz AC signals for the measurement of the major loop signal. The oscilloscope detects and obtains the voltage signal across the load in the major loop where the moving and static contacts of the contactor are in a separated and closed state in real-time. The major loop is filtered to avoid interference with the clutter signal since a higher resolution is required to determine the bounce of the moving contact through the voltage across the load.
Experimental results and analysis
An excitation current of 6 A is introduced into the drive coil and recovery coil to drive the contactor to complete the switching-closing in the experiment.
Figure 15 shows the experimental results obtained from the measurement of the response time of the switching-closing. The voltage of the drive signal is attenuated to 1/5 of the true value in the figure for easy observation. The contactor starts to switch in when the current starts to pass into the contactor coil from the working principle of the contactor. That is, the drive signal begins to show step changes, and moving contacts move under the operating mechanism. The major loop is turned on when the moving contact is in contact with the static contact. The switching-in is finished when stable electrical signals (5-V sinusoidal signals) appear in the circuit. This period is the response time of the contactor switching in. The response time of separating brake is between the release of the separating brake command to the disappearance of the stable electrical signal in the major loop. Therefore, the response time of the contactor switching-in and separating brake is 9.8 and 3.2 ms, respectively.

Contactor response during switching-closing: (a) switching in and (b) separating brake.
The period from the beginning of a non-zero electrical signal in the major loop to the appearance of a stable electrical signal is time when the moving contact bounces during the switching-in of the contactor. Whether the contact bounces when the moving and static contacts are in contact during the switching-in and the duration of the contact bounce can be learned by observing the voltage signal curve across the load of the major loop. Figure 16 shows the major loop signal when the moving and static contacts of the MSMA A.C. contactor are in contact under 6-A excitation. The change of electrical signal in this process occurs in an instant, and there is no sporadic phenomenon subsequently. No bounce occurs during the contact of the moving and static contacts.

Contact bounce test.
Figure 17 lists the displacement curve of the moving contacts of the contactor during the switching-closing. The actual displacement of the moving contacts of the MSMA A.C. contactor is about 1.16 mm during the switching-closing due to assembly errors under different current excitation. The displacement time of the moving contacts during the switching-in and separating brake is 5.8 and 1.2 ms, respectively.

Displacement curve of the moving contact: (a) switching in and (b) separating brake.
The moving contacts of the MSMA A.C. contactor have a relatively smooth displacement curve under different excitation currents whether it is during switching-in or separating brake. Besides, there is no obvious vibration such as that in the NXC-06 contactor. The perspective of displacement confirms that the MSMA contactor does not bounce when the moving contact is in contact with the static contact during the switching-in process. The displacement curve during the switching-in shows that collision occurs when the pull-in displacement of the moving iron core is cut off. It has no significant effect on the state of the moving contact.
The energy consumption of the contactor during the switching-in and separating brake is 0.83 and 0.25 J from the electrical work equation, respectively. Contactor power during switching-in and separating brake is 85 and 56 W, respectively, by referring to time spent in the switching-closing. Energy consumption of the contactor switching-in is maintained at 0 due to the memory effect of MSMA.
Conclusions
Material characteristic experiments were performed to obtain the output strain of the material under different magnetic fields and pre-pressure as well as the relationship between block force and the strain recovery. An A.C. contactor driven by MSMA was designed based on the magnetic shape memory effect of MSMA. Its basic structure and working principle were expounded in the work. Finally, the simulation analysis and experimental verification of the contactor were carried out, and the main conclusions are as follows.
The deformation of the MSMA sample during elongation and recovery was approximately linear within a certain range. The increased load gradually decreased the strain of the MSMA, and the maximum block force of the MSMA sample was 19.82 N.
The MSMA contactor achieved a fast switching-in response driven by a saturation magnetic field of 0.6 T, and response time was only 9.8 ms. Meanwhile, contact was avoided through MSMA during switching-in. The energy consumption of the contactor during switching-in was only 0.83 J due to the rapid response of MSMA.
