JMAG Newsletter September, 2014Solutions

[Back]

Solenoid Valve Analysis Solution

JMAG has supported the solenoid valve analysis since it was first released and has been used for various applications. We introduce solutions using JMAG focusing on the response characteristics evaluation of a solenoid valve in this issue.

Overview

To accurately regulate the flow volume using a solenoid valve, it is necessary to accurately predict and control the attraction force that activates the valve. Therefore, a solenoid valve is a magnetic device that is often used as an analysis target of the electromagnetic analysis. A wide variety of analyses have been run such as evaluation of response characteristics by combining motion and drive circuit as well as thrust power of plungers. This issue introduces solutions for response characteristics evaluation of solenoid valves using JMAG.
To evaluate response characteristics of a solenoid valve, handling of attraction that works on the plunger, that is handling of electromagnetic force is important. Electromagnetic force is determined by the magnetic properties consisting of a solenoid valve, but an analysis accounting for the influence of eddy current and the remnant magnetization is necessary depending on the material characteristics or environment. This issue introduces the effect of eddy current or remnant magnetization on the response properties through some case studies.
Also, the solenoid valve is commonly used as a component in the system and the control target. In terms of the evaluation of the response properties, it is necessary to analyze the response properties of the entire system including the solenoid valves, not just a single solenoid valve. This issue also introduces some cases that analyzed response properties from the view point of system control simulation.

Impact of Eddy Current

High speed response is often required for the on-off valves which just open and close the valve, but it may result in a design issue because of eddy current occurring in plunger or stator. Since the occurred eddy current obscures the principle magnetic flux and necessary attraction force cannot be obtained after current starts flowing, it is displayed as a delay in the start point of the response waveform. Since the eddy current tends to flow concentrating on the metal surface (skin effect), a modelling accurately accounting for the skin effect is required.
Such a modelling is possible by generating mesh (skin mesh) with a sufficient resolution within the skin region where eddy current occurs. The fig.1 shows an example with slots added on the stator surface to suppress the effect of the eddy current. This example compares and evaluates the improvement of the response characteristics after the current flows between with slots and without slots, and approximately 20 % of improvement was seen by controlling the eddy current flow using the slots.

Fig.1 Stator without slots (left) and with slots
Fig.1 Stator without slots (left) and with slots

Fig.2 Comparison of response wave (displacement amount) between with slots and without slots
Fig.2 Comparison of response wave (displacement amount)
between with slots and without slots

Residual magnetization effect

As the degree of detail is improved, the effect of the remnant magnetization may cause the difference between the actual measurement and analysis. Since the remnant magnetization is caused by magnetic hysteresis characteristics and contributes to the attraction, it has an effect on the response wave such as displacement amount.
For analyses using JMAG, we continue to pursue development of new technologies regarding material modeling, such as magnetic hysteresis including the remnant magnetization and provide related products. Also for solenoid valve designs, some case studies handling the effect of the remnant magnetization on the attraction have been presented in our user's conference (1).
Examples comparing the attraction force and response characteristics (displacement amount) between with remnant magnetization and without remnant magnetization including eddy current are shown here (Fig 3, 4). In fig.3, the waveform from the start of the attraction force to the peak value (to 2 (msec)) shows the decreased attraction force because magnetization is shielded by the eddy current, but does not show the effect of the remnant magnetization. This is because the initial state of the analysis is assumed to start from the origin of the magnetization curve without remnant magnetization. The flat attraction wave shows that displacement is stabilized by the current constant control. An analysis accounting for the remnant magnetization causes larger attraction force than an analysis that does not take into account it.
Also, in the response waveform in Fig.4, response just after turning off the current shows delay in the analysis accounting for remnant magnetization. This is because plunger starts moving after the restoring force of a spring exceeds the attraction force due to remnant magnetization. You will see that the attraction force causes a significant impact on the response characteristics in the material where remnant magnetization occurs.

Fig. 3 Comparing attraction force between with and without eddy current / remnant magnetization
Fig. 3 Comparing attraction force between with
and without eddy current / remnant magnetization

Fig.4 Comparison of response wave (displacement amount) between with eddy current / remnant magnetization
Fig.4 Comparison of response wave (displacement amount)
between with eddy current / remnant magnetization

System control simulation

A solenoid valve in the system is a control target (plant) from the system point of view and treated as an equation model written in the motion equation or voltage equation. Although feature of each product is shown as a coefficient of equation such as mass and inductance, all features of the product cannot be applied to an equation with only limited number of items. For example, inductance for a solenoid valve cannot be displayed as a constant coefficient value since the magnetic circuit considerably varies depending on the displacement of the plunger or drive current. Therefore, technology incorporating a detailed plant model into the control simulation as a substitute of the equation model is required.
JMAG allows you to try the following two approaches.
First, there is a method to use an infinite element model as is (hereafter called direct coupling). An analysis is proceeded exchanging the data of voltage, current, position information and attraction force bi-directionally between JMAG and a control simulator. Since both simulators are directly connected, phenomena of eddy current occurring in the controlled plant model can be analyzed together with the detailed system control simulation. Therefore, the direct coupling is an appropriate method to look at the detailed electromagnetic phenomena occurring in the plant model during the control.
The second method extracts the necessary plant model information with a control simulator from the infinite element model and refers to it as a table data. JMAG provides a solution called JMAG-RT, which creates table data after extracting the plant model information from an infinite element model. The method to perform control coupling using this plant model is hereafter called JMAG-RT coupling. By providing the required motion region as a table instead of a plant model using an equation composed of a constant coefficient value which was utilized within the conventional control simulator, a control simulation reflecting material and geometry properties for each moving point can be performed.
Unlike in the case of the two-way direct coupling, this method operates only the control simulator and behavior of the eddy current occurring in the plant model during control cannot be confirmed at the same time, but it can realize faster analysis than the direct coupling. This is an appropriate method to check the attraction force and response of the solenoid valves from the system control point of view. Also, taking advantage of its high-speed characteristics, search and adjustment of gain parameters suitable for the system control are possible (Fig.5 and 6).

Fig.5 Control circuit for the system control simulation of a solenoid valve using the JMAG-RT coupling
Fig.5 Control circuit for the system control simulation
of a solenoid valve using the JMAG-RT coupling

* RT in the figure indicates a plant model generated from JMAG-RT which is a tool to generate a plant model for table coupling provided by JMAG

Fig.6  Command value and response value based on the control circuit in Fig.5 (MATLAB/Simulink)
Fig.6 Command value and response value based
on the control circuit in Fig.5 (MATLAB/Simulink)

Conclusion

This issue introduced solutions achieved by using JMAG focusing on the response properties of solenoid valves.
JMAG will continue to add new functions for solenoid valve analysis. An analysis accounting for the collision in motion has been supported in JMAG-Designer Ver.13.1 released in June. Using this function enables simulation of a case where a plunger comes in contact with the stator during a high speed drive and causes vibration in the response characteristics. In addition, an analysis function incorporating the outer force such as fluid force is now being developed using user subroutines.
We hope the contents descried in this issue are helpful both for those who are considering to start solenoid analysis and those who has already started.

(Takayuki Nishio)

Reference:
[1] Yoshiaki Shinozaki: Attraction Force Analysis due to Remnant Magnetization JMAG Users Conference 2012 (for users only)

Contents

1. Solutions   - JMAG Multiphysics Analysis -
2. Solutions   - Solenoid Valve Analysis Solution -
3. WEB   - JMAG Tutorial Videos -
4. Fully Mastering JMAG   - Common Questions for JMAG -
5. Event Information


Top of Page

Contact US

Free Trial

Latest Issue
NewsLetter
January, 2016
Back Issue
Back Issue