Thermal Modeling and Design of an

Experiment to Measure the Thermal Behavior of a

Linear Motor for a Precision Motion Control System

 

Spring 2005

 

 

Sponsor:             

 

         

 

 

Instructor:           Savas Yavuzkurt 

 

Completed:          April 20, 2005

 

Team Members:  Jake Conley           jjc310@psu.edu

                             Will Gross             whg107@psu.edu

                             Rich Heibel           rsh903@psu.edu

                             Pete Matthews       pjm281@psu.edu

                             Surya Seiple          sjs304@psu.edu

 

 

Problem Statement:  To predict the temperature at various nodes on the linear motor and u-channel slide to allow for better performance from the system.

 

Overall Objectives of the Project:

 

Executive Summary:

 

Little research has been done on the heat transfer of linear motors, although heat greatly affects the performance and capabilities of them.  As such, Aerotech requested a computer model which would predict the heat transfer through one of its linear motors.  With such a model, measures could be made to increase performance, reliability, and overall quality of the motors.

 

To provide initial heat transfer analysis, a one dimensional thermal resistance model was created.  Through simplifying assumptions, the model predicts the temperature at different nodes on the forcer based on a given power input and environment conditions.  For a more detailed analysis, ANSYS software was used to create a finite element model of the heat transfer in the forcer.   This model uses experimentation inputs to provide representation of the heat transfer throughout the forcer showing hot spots and temperature distribution.  

 

To validate the thermal models, a test stand was constructed with which the temperature of the forcer could be measured.  The test stand is designed to model the linear motor as it would appear in application, but the motor remained stationary throughout testing.  Two forms of temperature measurement were used during testing.  First, a thermal imaging camera was used to view the temperature of the linear motor.  The thermal camera allowed for the identification of hot spots on the forcer and provided preliminary temperature.  Further experimentation was done using thermocouples attached at the hotspots of the forcer determined by the thermal camera.

 

The temperature data obtained in the experimentation was compared to both the linear resistance model and the ANSYS models.  Close representation of the forcer temperatures was achieved using the one dimensional thermal resistance model.  This model predicted the average temperature of the forcer side to within around 2.9% at various power inputs.  The model did not give an accurate representation of the top plate temperatures.  Both models can be refined to produce more accurate results.

 

Experimentation:

 

To verify the one dimensional thermal resistance model and the finite element analysis model, experimentation needed to take place.  To perform experiments, a test stand was constructed.  The purpose of the test stand was to simulate operating conditions of an actual motor.  This way, the above mentioned models would be tested against conditions that the linear motor might see in actual application.  With the results from experimentation, the analytical models can be modified to make them more applicable to Aerotech’s needs.  Two methods of measurement were used: thermal image camera and thermocouples.

 

       

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

One Dimensional Thermal Modeling:

 

A simple linear thermal resistance model was chosen for analysis and a simplified geometry used.  Several assumptions were made in regards to the model including steady state, symmetry along the y-axis, stationary forcer, and one dimensional conduction through the walls.  Nodes were placed at material interfaces.  The model solved fro average surface temperatures using assumed material properties, simplified geometry, and heat transfer equations.  These equations were plugged into MathCAD which can solve the equations numerically given the electrical input voltage and resistance.

 

 

 

Finite Element Analysis Modeling:

 

Finite element analysis modeling was carried out using ANSYS.  Unfortunately, the ANSYS program used was an older version making the model very difficult to construct.  Using ANSYS, a two dimensional model was created.  Similar assumptions were made as with the one dimensional thermal resistance model including half of the forcer symmetry, a stationary forcer, and ignoring the track forcer interaction.  The methodology behind the model was to create areas with proper dimensions relative to the vertical and central axis of the forcer.  Once accomplished, the forcer material properties were set.  Using known surface temperature values obtained from experimentation as well as the power input in the experiments, data was collected when steady state was reached and a plot contour map of the nodal solution created.

 

 

Comparing and Modifying:

 

The one dimensional thermal resistance model and experimental temperatures correlated well for the side of the forcer yielding around a 2.9 percent error when the model was compared to the average temperature and a 5.1 percent error when the model was compared to the peak temperature.  Sources of error include the one dimensional simplification and thermal resistance values.  Further development of the one dimensional thermal resistance model includes adjusting the thermal resistances to bring the data closer to the experimental results. Because the ANSYS software was very difficult to work with, the ANSYS model’s results were inconclusive.  ANSYS has the capability to produce a more accurate model, but more work needs to be done to achieve this.