Industry Today - Are you into it?
New Users - Register click here
An in depth look at the impact of simulation on automotive development
Posted via Industry Today. Are you into it? Follow us on Twitter @IndustryToday
By Lingmin Shao, Technical Specialist, and Ryan Elliot, Sensor Team Leader
KSR International, Ridgetown, Canada
Electronic throttle control works behind the scenes to provide a number of benefits in automotive applications. It facilitates the integration of features such as stability control — which stops the vehicle from skidding — automatic braking systems, cruise control, traction control and pre-crash systems that require the throttle to be moved
independently of the position of the accelerator pedal. At the same time, electronic throttle control lowers emissions and improves fuel economy.
The technology removes the mechanical link between the accelerator and throttle and instead controls the throttle with an electric motor. An electronic control unit (ECU) determines the correct throttle position based on data collected from two sensors that track the position of the gas pedal. The motor that controls the throttle is then driven to the required position via closed-loop control from the ECU.
KSR International is a leading supplier of inductance sensors used to determine gas pedal position, which allows the electric motor to control the throttle. Most inductance sensor applications can be addressed by a common design; however, as KSR expands the technology to different applications with different sensing ranges and packages, a
custom design is required.
Initially, KSR used trial-and-error methods, which took approximately three months to develop custom inductance sensor designs. More recently, the company has been using
Nexxim circuit simulation and HFSS 3-D full-wave finite element electromagnetic simulation software from ANSYS to evaluate the complete operation of its inductance sensors.
Because evaluating software prototypes is faster than building and testing hardware prototypes, KSR has reduced the time required to engineer a custom inductance sensor application to only two weeks.
A critical requirement of electronic throttle control is that a sensor accurately and reliably determines gas pedal position as the driver moves it with his or her foot. Hall-effect
sensors provide one option; however, they rely on a permanent magnet whose magnetism might be reduced over its lifetime — which might cause the sensor to produce inaccurate
readings. Inductance sensors, on the other hand, send alternating current through fixed transmission coils.
These devices produce an electromagnetic field that generates eddy currents in a metal rotor connected to the gas pedal or other device. The eddy currents produce an alternating
current in receiver coils. The magnitude and phase of this alternating current depend on the position of the rotor. A single application-specific integrated circuit (ASIC) excites the transmission coil and interprets the signal from the receiving coil to determine gas pedal position.
One concern with conventional inductance sensors is that inevitable manufacturing tolerances, such as variations in the air gap between the rotor and transmitter and receiver
coils, affect the sensor's transfer function, which, in turn, may generate inaccurate readings.
KSR's innovative design overcomes this problem by incorporating a reference coil
designed so that the mutual inductance between the transmitter coil and the reference coil is affected by the air gap to the same degree. Then, when processing the signal from the receiving coil, the mutual inductance between the transmitter and receiver coils is divided by the mutual inductance between the transmitter and reference coils.
This provides a rotational angle measurement that is independent of the air gap. KSR faces a significant design challenge in designing custom sensor applications that vary significantly from the baseline. These applications typically have a different range of
motion, which can require new coils and an application-specific integrated circuit to drive the transmitter coil, measure the voltage generated in the receiver coil, and calculate the
position of the gas pedal or other device whose rotational position is being measured. In many cases, inductance sensor designers must address interference from other components in the customer's product. For example, nearby metal parts might affect the magnetic field generated by the transmitter coil and rotor.
In the past, KSR engineers developed an initial concept design based on their experience and intuition and then built a prototype. In some cases, performance fell short of requirements. Physical testing provides only a very limited number of data points, so it was often difficult to diagnose the root cause. As a result, the process called for a considerable number of design-build-test cycles to meet customer requirements and
reach an optimized design. The time consuming approach was dependent on the engineer's skill and dedication.
More recently, KSR's use of HFSS and Nexxim software has reduced the need for physical prototyping. The circuit simulation setup process begins by importing initial
design geometry into HFSS from a computer-aided design (CAD) file. The engineer defines the electrical material properties — such as permittivity and dielectric loss tangent, permeability and magnetic loss tangent, bulk electrical conductivity and magnetic saturation — along with any boundary conditions necessary to specify field behavior on surfaces.
HFSS computes the full electromagnetic field pattern inside the structure, calculating all modes and ports simultaneously for the 3-D field solution. It then computes the generalized S-matrix, reducing the full 3-D electromagnetic behavior of the structure to a set of high-frequency circuit parameters.
The HFSS model is inserted as a block into a Nexxim full-circuit simulation, which combines the accuracy of electromagnetics with the simulation speed of circuit simulation. Nexxim provides transistor-level accuracy required for simulating
sensitivity analog and wireless frontend circuits as well as robustness and
capacity needed to address the diversity and complexity of modern mixed-signal integrated circuit designs.
Inductive sensor performance depends on a good oscillator. Full circuit simulation makes it easy to determine if the oscillator has a clean wave form and how much current it
consumes. Circuit simulation output provides other important design information, such as the signal at various rotor angles. Full ASIC models can be imported directly into Nexxim in protected or unprotected versions.
Parametric simulations are often performed to characterize the model in terms of physical dimensions, material properties or various states. The HFSS model is parameterized and characterized over user-defined ranges in parameter space and frequency. During circuit simulation, a multidimensional interpolation is applied to obtain S-parameter data required by the circuit simulator for particular parametric instances of the model.
For example, KSR engineers typically perform a parametric study of the air gap between the rotor and transmitter and receiver coils to ensure that inevitable manufacturing variations will not affect the sensor accuracy. Different air gaps require different tank current levels to maintain the same raw signal strength. The output should be independent of air gaps because of the reference coil's role (described earlier).
The full-circuit model helps KSR engineers to address inevitable issues that arise in real-world applications. For example, steel brackets located near the sensor act as a secondary
eddy plate and can have a significant impact on output. In one application, adding a steel bracket, shaft and bolt reduced output voltage by 11.6 percent, an unacceptably high amount. KSR engineers created a series of HFSS models that isolated each of the components. This showed that the bracket alone reduced the output by 10.2 percent, the shaft by only 1.3 percent, and the bolt by 0.1 percent. KSR engineers focused their attention on the bracket and attempted several design changes.
The team discovered that adding a 20 mm hole to the bracket reduced its impact to an increase of 0.7 percent of output. When added to the effects of the bolt and shaft, the net result was a negligible impact on output voltages, which can be corrected for with standard final programming adjustments.
In another application, parametric analysis revealed that changing the temperature of the inductance sensor caused the output to drift. KSR engineers analyzed the root cause by
separately analyzing changes in coil temperature and ASIC. This study determined that the coil was the major contributor to sensor temperature drift. The team developed a new coil that reduced the output variation from 2 percent to 0.5 percent.
Simulation has enabled KSR engineers to optimize inductance sensor designs prior to building a prototype. The new approach substantially reduces the time required to
engineer its products to fit customer applications. The net result is a substantial decrease in engineering costs and a reduction in time to market.
Contact Information
Media Queries
Press office
Communications Department
UK
NN4 0RQ
BACK TO TOP