Sensors & Data Acquisition
Table of Contents
EngineThe engine used by RIT Baja SAE Team is the Briggs & Stratton 10 Hp OHV Intek Model 205432 engine. General specifications are:
- About 10 hp
- 14-18 lb-ft Torque Output
- Average intake air approximately 20 cfm
- Idle RPM of 1750, Max RPM of 3900
Variable of InterestParameters to be measured using the test platform, as requested by the customer include:
- Oil Temperature
- Carburetor Bowl Temperature
- Intake Air Temperature
- Exhaust Temperature
- Cylinder Pressure
- Intake Air Flow Rate
- Fuel flow rate
- Exhaust O2/Air-to-fuel ratio (optional)
- Output Torque (and hence horsepower)
Sensors Choice & PlacementThe resolution of the sensors to be used on the test platform, were not specified by customer. Sensor specifications were chosen based on engine's operating conditions, budget and availability. The chosen sensors were placed in appropriate locations to accurately measure engine parameters.
Thermocouples are widely used temperature sensors which are comprised of junction between two dissimilar metals that provide voltage relative to the temperature difference. This junction is commonly referred to as the "hot" junction. The potential difference (voltage), typically in millivolts, resulting from the temperature difference at the junction is measured from the positive and negative leg of the device. The readout terminal of the device is referred to as the "cold" junction. By measuring the output voltage and comparing it to a reference junction, the temperature at the "hot" junction can be determined. Since thermocouples output voltages in millivolt (mV) range, additional amplification and signal conditioning are required to obtain a useful output.
There are various types of thermocouples, usually selected based on the temperature range and sensitivity needed for its application. To determine the type of thermocouples applicable to the test platform, the team ran the test engine (Briggs & Stratton 10 Hp OHV Intek Model 205432 engine) for about 10-20 minutes; and using an IR thermometer, measured the temperature of certain points of interest such as exhaust, engine oil and muffler. Based on this experiment, worst case estimates were made and the required operating range and type of each thermocouple was chosen as listed:
|Parameter||Measured Value||Worst-case estimate*||Unit||Type of Thermocouple|
|Intake Air Temperature||73.94||130||Fahrenheit||K|
- Worst-case values generated by adding +/-(50 to 70) degrees Fahrenheit to the measured values. For the muffler region, the high limit was chosen because temperatures up to 700 degrees Fahrenheit has been reportedly measured.
- The K-type thermocouples used are unsheathed with ungrounded junction.
As mentioned earlier, thermocouples output very low voltages (in mV range) relative to temperature. To verify performance of chosen thermocouples, a TC-2095 terminal block from National Instruments, supplied by the M.E. Department was used for this purpose. Thermocouples were connected to National Instrument's SCXI-1102 Thermocouple Input Module (with multiplexing) for signal amplification as well as data acquisition.
Load cells are used to directly measure force applied by engine. The output torque of the engine is computed as the product of the measured force and distance in the direction of the applied force. At the moment, two strain-gauge load cell sensors from Interface Inc. are available for the project: SM-500 ( a 500 lbf load cell used by Formula SAE team on their test stand ) and smaller-scale SM-150 ( 150 lbf load cell ).
To measure torque, the load cell is mounted on a lever arm placed at a known distance (16 inches) from the rotor, directly underneath the DC motor dyno. The lateral force (with respect to the gear ratio) measured by the load cell is multiplied by this known distance to compute the output torque of the engine.
The SM-150 load cell was used for this setup instead of the SM-500 load cell previously used by Formula SAE. The basis of this decision was that the output torque of the engine is only about 1/10th of that from the Honda CBR600 engine used by Formula SAE.
A new MS connector had to be ordered for this project. Lead wires from the SM-150 were soldered to the MS connector. The load cell was connected to the analog RPM input on DynesSystems Dyno-Loc IV controller using the MS connector. Since the SM-150 is slightly smaller than the SM-500, new rod ends were devised to mount the sensor to the dyno frame.
A hall effect (magnetic) encoder with 60 gear-tooth wheel is used to measure speed of the DC motor. The output is connected to the Dyno-Loc IV controller for feedback/control and also coupled to the PCI-6034E DAQ card using BNC cables via the BNC-2095 Rack. This encoder, however, measures speed of the DC motor not the engine speed. The engine speed could be determined by multiplying the dyno speed with the drive ratio.
An additional optical encoder (with its own trigger wheel) is mounted directly on the output shaft of the engine to measure engine speed and trigger cylinder pressure measurements in the DBS 5 Data Logger from TFX Technology. The unit connected to the DBS 5 Data Logger and is mostly used to correlate measured cylinder pressure with engine speed, as well as for cross-reference.
KISTLER piezoelectric pressure sensor Type 6125C was chosen to measure the cylinder pressure in the test engine. The sensor boosts measuring range of 0-300 bars and outputs varying capacitance with respect to pressure. The sensor is connected using BNC cable to KISTLER Dual Mode Charge Amplifier Type 5010B. This converts the measured charge to corresponding user-scalable voltage. The output of the amplifier is connected to DBS Basic 5 data logger from TFX Technology. All measurements are analyzed using TFX combustion pressure analysis software. This setup is independent of the LabVIEW GUI. An optical crankshaft encoder compatible with TFX units will be used to trigger pressure measurements as well as correlate measured pressure with engine RPM. In addition, a spark plug cylinder pressure transducer from PCB Piezotronics, also integrated with DBS Basic 5 will also be used. Note that a trigger wheel had to be constructed for this setup.
Due to the small-scale of the engine, the MAF sensor requirement ranges from 0 to 20 cfm. MAF Sensors at this range are expensive and out of budget. We will not be able to integrate an MAF sensor into this setup. However, we will from time to time borrow a air velocity transmitter from Mr. John Wellin in the ME Department. The sensor, mounted in a snorkel of known area, measures air velocity which could be used to determine mass flow rate. Recall, mass flow rate is simply product of air density and volumetric flow rate. Assuming dry air, density could be determined from intake air temperature. The volumetric flow rate is known area of snorkel times air velocity. The sensor has user-selectable range and is configured to output 1-5V depending on selected scale, with 1V corresponding to 0FPM or 0MPS. This analog output is coupled to the DAQ card via BNC-2095 rack.
An Innovate! LM-2 Kit provided by the M.E. Department will be used for this project. The kit contains a universal Bosch Wideband Lambda sensor for measuring air/fuel ration, data logger unit with 4 fully-differential analog inputs for 0-5VDC & tach signals, 2 programmable analog output channels, USB connection to PC, LogWorks software for configuring the device, SD card for data storage. Since the kit comes with an oxygen sensor, no additional purchase was required.The team only had to purchase AC-DC car cigarette socket adapter to power the kit. The kit will be programmed to output linear analog signals from the lambda sensor. These signals will be sent via a BNC connector (linked to BNC-2095 on SCXI chassis) to the PCI-6034E DAQ module.
Throttle adjustment during runs, is achieved using a stepper motor and a cable/lever arm setup. The cable, which is connected to an Anaheim Automation stepper motor, is tied to the throttle of the engine and can be controlled via software in the monitoring room. The stepper motor control is connected the PC via USB cable.
LabVIEW Graphical User Interface / MATLAB
Data acquisition and engine throttle control was implemented through National Instrument's LabVIEW program. The DC motor was controlled using Dyne Systems Dyn-Loc IV dynamometer controller. For safety, all control and data acquisition modules were located in a monitoring room away from the testing room, giving the user remote access during testing.
As shown in Figure below, the LabVIEW Graphical User Interface (GUI) featured multiple tabs, each for different tasks. Post-analysis to generate plots was handled using MATLAB script.
- Prior to each run, the user can specify test configurations using the 'Settings' tab or use default values.
- During each run, the GUI displays real-time engine speed, torque, horsepower, temperatures, fuel-flow rate, and AFR under the 'Acquisition' tab.
- Data logs of each parameter are created in a time-stamped CSV file located in dated-folder at 'C:/P10221/Data Files'.
- Real-time plots of engine speed and other debug parameters, such as output voltages of the velocity transmitter and LM-2 data logger, are available under the 'Analysis' tab.
- User can enter comments in the 'Observations' tab during runs. At the end of each run, comments were processed and logged in time-stamped text files.
Post-analysis of acquired data in the CSV file were performed using MATLAB script that plots the acquired parameters. The script is available in 'C:/P10221/Data Files' and gives the user the ability to plot one run, compare two runs from the same engine, or compare two different engines.
A Simple Moving Average (SMA) could be applied to the data, by asserting "EnableAVERAGING" in the script, to smooth out short-term fluctuations and highlight long-term trends. The period of the SMA is simply the ratio of the sampling rate to the number of samples collected by the data acquisition hardware i.e. 1000/100 = 10.