DEVELOPING THE UNDERGRADUATE LABORATORIES MECHANICAL ENGINEERING AT VIRGINIA TECH
ABSTRACT
During the past 10 years, technology development has significantly changed the approaches for collecting, processing and displaying data in the engineering laboratory. Computerized data acquisition systems coupled with a variety of software based instruments have fostered considerable changes in the nature of experimentation and acquisition and processing of data. These developments in technology have precipitated a reconfiguration of undergraduate lab experience in Mechanical Engineering at Virginia Tech.
The redefinition of the undergraduate laboratory program focused on two aspects of the Department's strengths: a sound research program and a commitment to undergraduate technical communication. Combining these strengths with a revised curriculum to include the changing technology in a cost effective manner was a challenge. This paper presents the development and implementation of the revised undergraduate courses incorporating the use of LabView software, extensive material on data acquisition and processing, and a background in fundamental statistics. The new lab program focuses on linking the research activity of the faculty with the laboratory experience of the undergraduate student. A common thread is the emphasis on communicating information via the laboratory report. In this regard, a number of lessons have been learned as this revision has matured and evolved. A direction for the future will be discussed.
INTRODUCTION
Engineering departments across the country have struggled during the last twenty years with their undergraduate laboratory curriculum. The cost of running and maintaining teaching labs has remained high. The evolution of computational power led by the use of personal computers has exacerbated these problems by increasing the demand for continuous software updates and faster processors. The changing technology is costly despite discounts offered for educational purposes by many of the vendors. Often, the response to these pressures has been to reduce the amount of laboratory time offered the engineering student. Many departments have only a single course that exposes the engineering student to transduction techniques and the experimental process. Contributing to the reduction in laboratories is the decrease in faculty with the background to teach these labs. This is due in part to the reward system in universities that tends to emphasize research and publication at the expense of teaching especially the time intensive courses typical of the laboratory type courses.
This paper addresses the steps that have been undertaken to strengthen the undergraduate laboratory experience in the l)department of Mechanical Engineering at Virginia Tech. The future plans for consolidating these changes into an effective educational experience in the use of instrumentation, the processing of data and the efficient communication of this data is also discussed. Directly linked to this transition is the nature of the course content, the use of computers in the laboratory, the implementation of Virtual Instruments (VI) such as those available in the software LabVIEW, and support software such as Matlab, Microsoft Excel and the like. The process of revising the laboratory curriculum gave rise to a series of discussions/meetings/arguments regarding the nature of the labs, their impact on the education of our undergraduates, the goals of these laboratory courses, the resources required to properly maintain them, the measures of effectiveness, and space allocation for these facilities.
The revision of these laboratories has been centered on a computerized data acquisition system, and support software including LabVIEW, a commercial software package, that allows the easy construction of instruments within the domain of the computer. Matlab is also used for data analysis, plotting and presentation. This approach promised to address the broad range of instruments required for the engineering laboratory environment, maintain the flexibility to adapt to technology changes, and permit the subtle experimental variations necessary for a quality laboratory experience.
After five years, and nearly one thousand students, a number of lessons have been learned. The bulk of this paper discusses the intent and the results. These labs continue to evolve to address the changing student population, technology change and the reduction of state support so crucial to the maintenance of equipment.
REVISION OF THE UNDERGRADUATE LABS
Revision of ME 4005
The revision of the undergraduate laboratories was initiated with the overhaul of the capstone laboratory sequence, ME 4005 and ME 4006. The ME 4005 laboratory course which is formed around two lecture sessions and one three hour laboratory session per week was revised to consist of the following basic topics;
These topics are supported by laboratory exercises designed to reinforce with applied experience the material developed in the lecture portion of the course. The laboratory is supported with computer based data acquisition utilizing LabVIEW as the software for developing virtual instruments compatible with each of the experiments. Sufficient equipment is available such that two students work at each lab station. Weekly individual laboratory reports are required. These reports consist of an introduction, a short discussion of the laboratory exercise and an analysis of the data with the focus being the analysis of the data.
This initial concept for the laboratory has been revised in content and format during subsequent sessions. The following lessons have been learned during the development of this course.
a) The initial concept for introducing computerized data acquisition was to provide the student with an introduction to the LabVIEW software during the initial laboratory period. Subsequent VIs necessary to support the week's experiment would be 'modified' by the student to optimize the data acquisition for the experiment. This modification would be designed into the experiment to encourage the student to put some thought into the experiment being considered and thus addressing such fundamental aspects of experimentation as data acquisition and analysis. To support this 'encouragement' was a pre-laboratory handout detailing the experiment that was to be considered and the data that would be required. The student would then, in concept, come to the lab prepared to make the required changes in the general VI and proceed to complete the laboratory assignment. Both DOS based machines and Macs serve as data acquisition platforms. The students almost exclusively use PCS in other course work. This was initially thought to be a source of problems but it never materialized. The students easily adjusted to the variance without a problem. The initial training in the LabVIEW software proved inadequate, creating the situation where many of the students spent as much time modifying their VI's as they did on the experiment. This realization was caused by several factors. One factor is the complexity of LabVIEW. Another factor, however, was the lack of background by the students to understand the basic concepts that they were to modify. The pre-labs were often left to the night before the lab due to the scheduling of second semester junior year, which consists of five Mechanical Engineering courses creating an extremely heavy workload. Another aspect that thwarted this concept was the lack of access to the software other than in the laboratory and the general lack of laboratory experience in other courses. The modification of the data acquisition VI was eliminated, creating a VI that the students could use directly for performing the required experiment. The development of the VIs has created additional problems for the laboratory. The students tend to acquire the data without paying attention to the data that they have collected. This has lead to the collection of faulty data that has not been detected during the lab. With the number of students participating in the course, makeup labs are difficult to schedule and more frustrating for the student. The solution currently employed is to mix the laboratory exercises with data collected by hand, such as obtaining information from a multi meter or and oscilloscope, and the computer. The VIs currently in use are seamless with the experiment requiring no modification and little or no knowledge of LabVIEW. With the reduction of emphasis on the LabVIEW approach, the first laboratory exercise became a measurements lab rather than an introduction to the software. We had found that many of the students were naive in the fundamentals of simple measurements such as micrometers, multi meters and oscilloscopes. Increasing the emphasis on conventional data acquisition, forcing the students to write data as opposed to simply storing on a disk appears to reduce some of the undetected acquisition errors. For example, a group of students were acquiring temperature data from a thermocouple. The output voltages were negative as being measured by the computer system. Despite this being displayed, the acquisition process continued throughout the lab exercise. A GTA noticed the negative temperatures and questioned the students as to the temperature that they were monitoring. When it was pointed out that a negative voltage for this type of thermocouple indicated a temperatures well below freezing, their response was that they assumed the computer was correct and were going to analyze it at home after they had collected the data. The fact that the water bath was boiling had not factored into the experiment.
b) The lack of laboratory skills has become apparent in this course. Courses prior to this sequence have in the past included laboratories, providing some basic skills in data collection and more importantly, recognizing errors in the data. This observation has forced changes in the handouts provided to the students and the training of the teaching assistants(TA) that support the course. Many of these TAs have little or no experience in the lab and require considerable training to be an asset to the undergraduate students. Handouts for the students that discuss the experiment to be conducted include considerable background initially thought to be unnecessary; a series of questions are also provided that are intended to stimulate some thought regarding the lab and the data that will be collected.
c) A major problem that has been faced by the revisions made in this course is the compatibility of the prerequisites. The initial concept for the revised laboratory course was to focus the student on digital techniques, under the assumption that their analog background was sufficient to meet the demands of the laboratory and their careers. This proved to be totally false. The skills at the most basic level were questionable, and demanded course content to supplement their analog electronic background. The result is the removal of nearly all the digital content originally planned and the inclusion of topics dealing with more basic topics such operational amplifiers, voltage dividers, grounding needs etc. It's acknowledged that in-depth knowledge of op-amps is probably overkill for today's mechanical engineer, however, some of the basic techniques for transducer signal conditioning and transducer function are deemed important.
d) The original two-lecture format coupled with the three-hour lab was modified to having two lectures, a two-hour lab session and a one-hour recitation class. The addition of the recitation period served to deal with homework, pre-lab discussions, question and answer on lecture material and in general, more direct contact by the faculty with the students. This has proven to be quite successful, with the faculty member meeting the lectures ( 175 students) and 5 recitations section ( 35 students) a week and GAs supporting 9 to 10 lab sections. The maximum lab section allowed is 20 students. Our culture has treated an environment that does little to encourage 'hands-on' skills. As a result, the students are required to mount and test their own strain gage, wire and test an amplifier, and make and test a resistance thermometer. These innovations have been successful with one of the students commenting, " Mounting the strain gage was the most exciting thing I did in engineering school." That statement says wonders about the courses that are provided.
Revision of ME 4006
The second course in the capstone sequence, ME 4006, follows directly in the curriculum. This course attempts to focus the educational process on the communication skills of the students. The objective of the course is to teach technical writing in the format of the laboratory experience. A series of five experiments are developed by the faculty based on the research interest in the department. These experiments tend to be state-of-the-art exercises providing the student with exposure to the research of the faculty and the latest in techniques for acquiring and processing data. The five experiments that are currently being used in this laboratory course are:
The five experiments are performed by each student and a full write-up of the experiment is required. The experiments are formatted such that it requires the student to author their report toward specific goals. For example, the experiment is the result of a request from a customer for technical information about the device to be tested.
The format of this course includes a lecture by the individual faculty outlining the theory of the experiment and an overview of the procedure to be used. Care is taken avoid giving detailed procedures. The experiment is performed by the students in groups of five or less and supported by a graduate teaching assistant. A follow-up meeting with the faculty is provided to discuss the results and answer questions that arose during the experiment. An additional session is also provided for the students with the GT. to discuss the writing of the report. Lectures on report writing are provided in addition to the technical session$ for the specific laboratories. Feedback is provided to the students in terms of their writing skills by allowing the student to revise and resubmit the write up after the initial grading.
VIs have been developed to support each of the experiments. Multiple channels of data acquisition are typical for these experiments and the post processing includes the development of analytical models to explain the data and least squares techniques to extract or estimate certain parameters. Statistical analysis of the data is required where applicable and considered an important part of the reporting process.
One example of the incorporation of research into ME 4006 projects is the "Measurement of Surface Heat Flux and Temperature" which was a laboratory exercise used between 1991 and 1994. The experiment demonstrates a new heat ~lux sensor technology which was recently developed and patented [ 1] by Virginia Tech. The students used the sensor to observe details of a combustion flame and performed the data processing with the use of LabVIEW. It also demonstrated the different effects of integrating versus differentiating data.
The Heat Flux Microsensor is a new sensor made using thin-film microfabrication techniques. It consists of several thin-film layers forming a differential thermopile across a thermal resistance layer. The combination of series thermocouple junctions and thin-film design creates a gage with very attractive characteristics. It is not only physically nonintrusive to the allow, but also causes minimal disruption of the surface temperature. Because it is so thin, the response time is less than 20 µsec [2]. Consequently. the frequency response is flat from 0 to over 50 kHz. Moreover, the signal of the Heat Flux Microsensor is directly proportional to the heat flux. Therefore, it can easily he used in both steady and transient flows, and it measures both the steady and unsteady components of the surface heat flux. In addition to the heat flux measurement, the surface temperature is measured with a platinum resistance layer (RTS). Therefore, these gages simultaneously measure the surface temperature and heat flux.
The laboratory involved using one of these sensors in a combustion flame provided by a propane torch. After the torch was lit and the flame adjusted to the desired level the test was started hy opening a shutter to allow the flame to impinge on the heat flux gage. An amplifier box was provided with the gage to boost the heat flux signal and convert the resistance of the surface RTS to a voltage.
The fast time response of the temperature and heat flux sensor combination allows comparison of the time-resolved measurements. Conduction modeling of the heat transfer gives a prediction of the temperature rise from the measured heat flux [3].
Note that this is similar to integrating the heat flux measurements. In similar fashion the solution for determining the heat flux from a given set of surface temperatures is [3]
Note that this is similar to differentiating the surface temperature measurements. Both of these expressions are in the form needed for digital signal processing. They have been programmed on a computer disc given to each student for subsequent processing of the data. By calculating the corresponding surface temperature response to the measured heat flux and comparing with the experimental temperatures, additional confidence in the measurements can be established.
Each student was to plot comparisons of their measured and calculated data versus llme Ior both temperature and heat flux and the resulting residuals from these comparisons. Sample results are illustrated in Figures 1 and 2. Students were expected to complete a statistical analysis of the several hundred data points tor each test and present it in tabulated form.
The course concept for ME 4006 has required less modification. There has been a consistent attempt to rotate at least one new experiment into the lab and frequently two new lab exercises are produced each time the course is offered. The governing requirement for these experiments is a direct relationship with current research being undertaken within the department. The approach serves several benefits. First, it introduces the undergraduates to the research efforts of the faculty, exhibiting the connection between the student's educational experience and the forefront of the technology of their chosen profession. Secondly, it encourages the faculty to participate in the laboratory course by showcasing their research and often providing interesting feedback for future research paths.
ADDITIONAL LABORATORY CHANGES
To compliment the previously discussed laboratory changes, we have begun to take advantage of the Web to communicate with the participating students. Lab information, including the lab manual, prelabs, and scheduling changes, are placed on the Web for student access. Homework answers, lecture handouts and other lab materials are also placed on the Net such that students can have these materials available to them. This also saves time reducing by reducing the shuffle of paper at the beginning of class. We have a long way to go to take advantage of the Web for enhancement of the laboratory education. Using LabVIEW type software for developing virtual instruments is certainly the direction that the laboratory experience is heading. A student version, well supported by the manufacturer, is readily available and the original concept of including some software modification needs to be revisited. This format (LabVIEW) offers many possibilities to address the technology these graduates will encounter during their engineering careers. The perception that virtual instruments reduces the cost of running and maintaining a laboratory experience is greatly exaggerated. Keeping computer systems upgraded, software maintained, and personnel trained, is a costly endeavor. Our experience has been that the cost of running the lab has not decreased significantly over the last five years.
An advanced laboratory course series is also planned at the graduate level. This series can effectively utilize the breadth of the software and the computer capability available. At this level the student would be expected to develop his or her own VI for application to the experiment. This would not only include the data acquisition function but the post processing as well Coupled into these experiments would be a control function taking advantage of the output channels or the D/A capabilities.
CONCLUSIONS
The Mechanical Engineering Department at Virginia Tech. is dedicated to providing a comprehensive laboratory experience to its undergraduate student population. Historically, the foundation of the engineering laboratory has been its hardware. That era has changed producing a laboratory that features data acquisition systems, computational capability and its software suite. The success of our laboratories will depend on our ability to perceive the future needs of our undergraduates and adapt the lab experience to meet those needs. We have made a start in that direction. The most important lesson learned to date in the development of this lab is the ability to change and adapt to the changing technology is more important than the few mistakes that are discovered during the changing process.
The information technology that is available far exceeds our ability to utilize it. The Internet is a resource that we as experimentalist must access to the benefit of our students. Performing experiments on line from remote locations is feasible. Virtual experiments, visualization tools, and post processing software are topics for considerable exploration for direct inclusion into the laboratory experience.
REFERENCES
1. Diller, T. E., and Onishi, S.,"Heat Flux Gage," U.S. Patent No. 4,779,994, Issued 2 October 1988.
2. Hager, J. M., Simmons, S., Smith, D., Onishi, S., Langley, L. W., and Diller, T. E., "Experimental Performance of a Heat Flux Microsensor," ASME Journal of Turbomachinery, Vol. 113, 1991, pp. 246-250.
3. Baker, K I. and Diller, T. E., "Unsteady Surface Heat Flux and Temperature Measurements," ASME Paper No. 93-HT-33, 1993. '
Back to Table of
Contents