TPE331 Strain Gage Torque System
Verification and calibration
By Greg Napert
July / August 1998
One of the most important items to monitor on a turboprop engine is torque. Many engines are torque limited due to limitations in the airframe. Although temperature is also a limiting factor, turboprops are often de-rated, and are capable of operating at a higher torque setting than the airframe allows before they reach their temperature limitations. Because of this, the torque gages need to be verified. In addition, the torque system needs to occasionally be calibrated in order to reflect an accurate indication of actual torque so the pilot knows how the engine is performing relative to each other, and relative to aircraft limitations.
Duane Moore, turboprop crew chief in charge of the TPE331 program at Garrett Aviation in Springfield, IL, explains that in order to verify torque, you must measure the torque directly from the engine using a special instrument called the Lebow Load Cell.
"There are two different types of torque sensing systems on the TPE331s. Depending on when they were built, the TPE331-1 to 10s have the hydraulic torque sensing systems, and on some -10s through all -14s, an electronic torque sensing system is used," says Moore.
The electronic system was an upgrade that increased the reliability of and simplicity of the torque sensing system. The unit consists of some electronic strain gages mounted on a torque ring which twists as torque on the engine increases.
According to an AlliedSignal training manual TSG-112, the effect of having dissimilar torque sensor signals from engine to engine can be confusing and dangerous. Assume the left engine were removed from the aircraft, mounted in a test cell, and connected to a dynamometer. The dynamometer is a very precise load-measuring device that measures exactly the power being produced by the engine and absorbed by the propeller. Notice the numbers below the left engine in Figure 1. We are running the engine in the test cell at precisely 35,650 inch-pounds of torque, which equals 900 horsepower in this case. If the torque sensor raw signal coming from that engine is put directly to a torquemeter, it reflects some reading.
Now assume that the right engine from your aircraft is mounted in the test cell right next to the other one. It is also connected to a dynamometer and runs at exactly the same power level of 35,650 inch-pounds or 900 horsepower. In this case, however, the torque sensor raw signal coming to the torquemeter is different.
Obviously, the pilot cannot tolerate this situation. His flight manual dictates a set torque limit. The manual does not identify this by specific engine serial number, it only says that on your airplane, both engines shall be limited to the same value. These raw torque signals, therefore, must be calibrated and compensated to reflect accurately the horsepower being produced, so that when the engines are producing identical power, the torquemeters in the cockpit will register identically.
Adjusting the signal conditioner
It should be emphasized that the calibration of the signal conditioner is required to ensure that the instruments reflect the true horsepower
being produced and is not done just to match torquemeters.
The raw torque signal from the bridge circuit cannot be changed once the engine is assembled. So, the raw torque signal from the torque ring is adjusted through calibration of the signal conditioner so that the horsepower reflected on the gage corresponds to the horsepower actually being produced by the engine.
One important point to note about the signal conditioner is its relationship to the cockpit torque indicator. Although the torque strain gage produces an increased voltage as the engine torque is increased, the signal conditioner reduces the voltage to the cockpit torque indicator. In other words, the relationship between the cockpit indicator and the signal conditioner is inversely proportional. In fact, in some aircraft, when no electrical power is applied, the torquemeter in the cockpit would actually reflect maximum torque limit.
When you turn on the electrical power on the system, the gage will then drop back to zero horsepower because it is being supplied with approximately five volts of input.
The signal conditioner control box is remotely located from the engine in various places per the aircraft manufacturer. The signal conditioner will permit the engine's raw bridge signal from the strain gage to be calibrated so that it accurately indicates the torque being produced.
One of the most desirable characteristics of the electronic torque indicating system is the ease with which the system can be calibrated. It is necessary, however, to run the engine before starting the procedure in order to heat the oil to the normal green range. This will warm the metal parts to their normal operating temperature and will minimize erroneous strain gage signals from tensions due to the dissimilar metals.
In order to calibrate the strain gage bridge, three adjustments are necessary. These are: zero out the signal conditioner with a zero adjustment, adjust the signal conditioner to the specific raw torque output or Data Sheet (DSC) signal from the bridge, and make the span adjustment, which involves calibrating the signal conditioner to the cockpit torque indication.
Some of the strain gages have dual bridges as a backup so that disassembly of the engine is not necessary in the event the first bridge fails. The second bridge can only be calibrated at the time you decide to use it, however. It is not possible to calibrate the signal conditioner to both bridges ahead of time because they have slightly different signals. Once the system has been calibrated for a particular bridge, calibration may be checked operationally by the use of a torque test kit.
The torque test kit is optional under normal circumstances, but is required if replacement of the strain gage torque ring is required. The test kit, referred to as a Lebow Load Cell, is installed at the prop shaft and measures actual torque between the prop and the engine. The torque test involves a check at five load points during an engine run.
Turbofan instrument calibration for the TFE731
Calibration of instruments on turbofan engines doesn't involve torque. Instead, turbofans are limited by temperature and speed. With this being the case, three instruments are critical to insuring the safe operation of the engine on the TFE731. These instruments are N1 speed, N2 speed, and T5 temperature. Calibrating these instruments for accuracy is a normal part of maintenance of these engines.
The methods used for the calibration of the N1, N2, and T5 have improved and have been simplified over the years, but the basic principle behind verifying the cockpit gages against what's actually happening at the engine is still the same.
Keith Crain, crew chief for turbine engines at Garrett in Springfield, IL, says that its old system was a bit more cumbersome than what they are using today. "With our old system, we would hand enter our data which we would monitor during an engine run with our multimeters directly into a desktop computer, and analyze the data."
Essentially, the process involves tapping into a harness between the engine and the electronic engine control unit (EEC) or digital electronic engine control unit (DEEC). These are the computers that control the engine, and provide information to the cockpit indicators. We tap into these wires between the engine and the EEC/DEEC with digital meters so that we're receiving raw data directly from the engine.
Crain explains, "The engine is still running its information through the computer and to the cockpit the way it normally does, but we're tapping in to get the same information the computer is seeing. A separate set of wiring goes to the cockpit. If the computer says it's seeing 500 degrees, and the cockpit says it's seeing 520 degrees, we know we have a problem such as a gage error."
"In practice, we go out to the aircraft with our calibrated equipment, locate the EEC or DEEC box, plug into our cannon plug. One technician stays in that location and another goes into the cockpit and performs an engine run. The technician at the engine has a communications set with which he talks with the technician in the cockpit. At specific points during the engine run (usually takeoff power, and then stepping down three percentage points at a time), the technician at the engine reads the information from the monitoring equipment to the technician in the cockpit."
"If the data that is being observed at the engine matches the data in the cockpit, the aircraft instruments are OK," he says.
One of the problems with this system, however, is that the accuracy of the tests are based on how well the technician at the EEC is interpolating the data on his instruments. Crain says, "The data is changing rapidly and the time delay between the moment he is asked for the data and his response can mean a few degrees difference. The net effect is that you may think the gages are off by 20 degrees, when in actually the temperature increased by 20 degrees between the time he was asked and the time he responded. Additionally, some of the indications fluctuate somewhat and so you have to manually average the temperature or other turbine indication. It takes some experience to do this. In fact, the gage in the cockpit is actually averaging the temperature continuously. We are getting away from this system because of these discrepancies."
To get away from these problems, Garrett adopted a new system called JEDA (Jet Engine Data Analysis). The equipment is installed in the same way, however, a long cable allows for the monitoring to be brought back to the cockpit with the technician. The data from the JEDA monitoring equipment is then observed at the same time the cockpit instruments are observed. The readings are then automatically averaged and recorded electronically, and the cockpit indications are entered by the technician. The unit is then brought to a personal computer and downloaded automatically to AlliedSignal's MEDRA software for analysis. For the most part the human factors errors are eliminated with this system.
Crain says, "These instrument verifications are performed as standard procedure at the Garrett facility for any incoming runs, or before any MPI, CZI, or performance troubleshooting, or any outgoing MPI, CZI or troubleshooting repairs are done."
Crain explains that they still keep the older equipment around, however. "Interestingly, for troubleshooting purposes, it can be more advantageous to use the older equipment. The older gages have the ability to monitor specific isolated indications and because they give you instantaneous actual readings, they can give you very good troubleshooting data. JEDA is strictly a five-point run instrument. Other than for determining gage error, it's not really useful for troubleshooting. So, if we're just looking at one of the isolated signals, we want to use simple equipment.
Crain reiterates, "All you're doing with any of these systems is seeing if your indications agree with what the cockpit is seeing. If it doesn't, you start to troubleshoot. You may have a number of items that might be bad — the gages in the cockpit, the wires, connectors, thermocouples, etc."
Breakout boxes for these engines are either available from AlliedSignal and other companies manufacture variations on them to include Howell's Jetcal®, TEC/Aces JEDA system, and many maintenance facilities have even custom designed their own break-out boxes. The TEC unit is one example of a vibration analyzer computer that was modified to provide a five point run instrument. Howell's Jetcal unit is custom designed to function as a calibration instrument for multiple engine parameters to include speed/frequency, temperature, pressure/EPR systems, etc.
These units are typically designed to collect data only. The data is the downloaded to a computer program which reduces the data so that it can be analyzed in a logical fashion.
Crain says, "Analyzing the data with the program can take some experience as well. The program analyzes the data at each one of the five points and you can get clues, depending on how the data matches up, as to what problems, if any, exist."
