Fan Trim Balance

June 1, 1999

Fan Trim Balance

By Jerry Justice June 1999

The first indication of an out-of-balance fan in a business jet is, more often than not, a complaint from the guy sitting in the plush, white leather passenger seat. Although an engine meets or exceeds the standards of vibration and noise levels off the production line, the owner expects it to be as quiet as the one in his high-rise office. Don't be surprised by noise complaints even though the aircraft is brand spanking new.

During the balancing procedure, you may find that it still meets the manufacturer's vibration standards, but the boss wants it even quieter. With a little tender loving care, it can be quieted. Other indications may be in the aircraft's Engine Vibration Monitoring (EVM) system, or vibrations felt or seen in the flight controls, floor, instrument panel, etc. Engine manufacturers often dictate specific intervals for fan trim balancing, while others may tie it to an event such as overhaul or inspection cycles. As a rule of thumb, I would recommend at least every 400 hours — more frequently if you have a CEO with sensitive ears.

When an engine manufacturer assembles an engine, each disk of the compressor and turbine section is individually balanced. First, the disk is assembled using an optimizing technique to place each blade in the best possible position. A static balance is either part of the optimizing process or may be accomplished after the disk is assembled. The disk is then placed in a dynamic balancer and spun up to a balancing speed. After each disk is dynamically balanced, it is assembled with each of the other stages to form the compressor and turbine sections of an engine.

The fan is actually the first stage of compression as well as the major thrust producer. Because the fan assembly constitutes the largest rotating mass in the engine, an out of balance condition will most likely result in noise and vibration. Out of balance conditions can be caused by a variety of situations. Remember that the original assembly optimized the placement of each blade. If foreign object damage necessitates the change of a blade, the balance condition may change drastically. Blades are normally, but not always, changed in matched pairs. The damaged blade is removed along with its symmetrically opposing blade and the two are replaced with the matched pair. Blade erosion, FOD damage and repair, spinner rotation, and normal wear can also change the condition of the balance. In any of these cases, the balance of the fan will change. To top off the problem, the fan rotates at speeds that resonate in the cabin. I often compare it to attaching an engine to a tin can, then crawling inside the can for a listen.

Ease of use, size, durability, and price vary from one extreme to another when choosing the analyzer or balancer. New digital designs are more efficient and less expensive than older analog technology based instruments. Ease of use for the technician should be paramount. Make a list of what you want in a balancer. Shop around, ask for demonstrations, try them out for yourself, compare features, and make a decision based on what you need to fill your requirements.

Vibration Sensors
For most fan trim balancing, a velocity sensor is the best choice. Its response range is well within the frequency bandwidth of most fans. Your application may call for a specific sensor capable of collecting higher frequencies for vibration surveys as well as fan trim balancing. In this case, the recommended sensor type for higher frequency requirements is normally an accelerometer. Some accelerometers, especially those designed for use in high temperature areas, may require an external charge unit.

A displacement sensor is designed for the lower frequency range where few jet engines operate. This type of sensor may be used in some rare cases to balance the fan, but almost never for vibration analysis. Refer to the maintenance manual if in doubt as to your specific needs. Manufacturers go to great lengths to match a specific sensor to their engine vibration analysis and balancing requirements. Take their advice first.

Check the calibration or data sheet supplied with the sensor and make note of the sensitivity. Sensitivity is the amount of voltage output from the sensor relative to an engineering unit of measured vibration. For instance, a velocity sensor might have a sensitivity of 20 mV/IPS (20 millivolts per Inch Per Second). This means that the sensor will produce 20 millivolts for every Inch Per Second of velocity transferred to the sensor. Your equipment should have the most common sensors used with the engine preloaded in the balance program. If not, you may be required to enter the sensitivity manually. With any sensor, you must also be acutely aware that the method by which they are attached to the engine (the sensor mount) can make or break the effectiveness of an entire system.

If the mount resonates at any of the operating speeds being checked, it induces a false vibration signature into the sensor and, of course, back to the analyzer. The material, size, weight, and construction of the mount all play a role in its resonance response. Here again, if the engine or aircraft manufacturer recommends a particular mount for the sensor, use it. Don't look for the easy way out. If the sensor mount is expensive, complicated, and hard to install, it's the price you pay for accurate data.

Speed (Phase Reference) Equipment
When it comes to collecting phase data there are four basic methods being used today:

A strobe light gives a visual relationship of where the out-of-balance mass is located on propellers and some turbofan engines. You'll need two technicians to do the balance with the strobe light — one to run the engines from the cockpit, and the other to stand in front of the running engine to operate the strobe. Collecting accurate phase data by this method is a hit and miss prospect. Remember that it's a visual reference only and a difference of ten degrees is hard to estimate at a distance of thirty or forty feet. The strobe cannot be used with equipment that calculates the balance solution automatically because the strobe is not capable of supplying a phase reference to the analyzer. This means you, the operator, must calculate the phase angle for adding weight.

A more advanced method of collecting speed and phase data is the photoelectric cell. It uses a strip of reflective tape placed on the spinner as a trigger. It is usually mounted somewhere in the intake of the engine by opening the cowling, removing an inlet sensor, and placing the photoelectric cell in the hole vacated by the sensor. It must be firmly positioned to project its beam on the reflective tape as it passes the photocell's position. The reflected light then triggers a signal as it is received back at the photocell. The photocell sends the once-per-revolution signal to the balancing equipment, which is then used to automatically calculate the phase angle for adding weight. The time required in opening and closing the cowling, removing the inlet sensor, and mounting the photocell can add an hour or more to an otherwise short job. Another limitation of the photocell is its range. Optimum range for the photocell is 12 to 18 inches from the reflective target.

A large step up from the photocell is the laser tachometer. It works on the same principal as the photocell with a strip of reflective tape but proximity to the tape is extended out to 30 feet. The laser is attached to a locking swivel head, which is then mounted on top of the wing or on the side of the fuselage with duct, or speed tape. The laser beam is aimed at the reflective tape attached to the spinner and the swivel mount locked in that position. This eliminates the need to open cowling or remove sensors. As with the photocell, the Lasertach sends the once-per-revolution signal back to the analyzer. The analyzer uses the input to calculate a phase angle and RPM. The photo on page 48 shows the reflective tape positioned for both a laser tachometer (tape near the tip of the spinner) and the photo electric cell (located nearest the fan blades). The eight bolts shown in the spinner assembly hold the spinner in place. When trim balance weights are required, the designated bolt is removed and a specific class weight similar to a washer is placed on the bolt. The bolt is then reinstalled.

Newer engine designs are using variations in a single tooth design on a phonic wheel as a once-per-revolution signal generator. The single tooth may be missing, shorter, longer, or offset when compared to the other teeth on the wheel.

This photo shows an accelerometer mounted on a 731 turbine engine.

Regardless of which of these systems you use, the speed and phase angle source provides several pieces of the balancing puzzle. The speed portion is obvious. It provides an indication of the rotation frequency (RPM) of the fan so that you may adjust controls to match the frequency of interest, or required balancing speed. The balancing equipment also uses this input to synchronize the out of balance mass passage with the actual speed of the fan.

An "overall" or "composite" vibration reading is the sum total of all vibrations being generated. The oscillation event can only occur once per each revolution of the fan. When the vibration sensor measures that oscillation event and compares it to the input of the tachometer, all non-synchronous vibration events can be electronically filtered out leaving only the vibration attributed to the condition of the fan. And finally, the phase angle information provided to the analyzer is used to calculate the radial location of the dominating mass (the heavy spot on the fan causing the oscillation) relative to a known fixed point, usually the vibration sensor.

The balance procedure
Equipment Setup
As stated before, equipment varies in many aspects. Refer to the maintenance manual for recommendations on balancing equipment. Be sure you have a sufficient supply of the class weight set for the engine. Most newer engine models incorporate a dual-purpose bolt that attaches both the spinner and the balance weights to the engine. There are design variations in engines and engine sub-models that require the spinner to be removed to add or remove trim balance weights.

Weights may also be attached with rivets or may necessitate changing a null weighted screw with a specifically weighted one. Later models such as the TFE731-60 require only the bolt to be removed when adding weights. Some engine manufacturers number the weight locations relative to an indexed point. In the case of the TFE731, the holes are not numbered.

The technician conducting the balance job picks any one of the holes and designates it as No. 1. The placement of the reflective tape along an imaginary line between the leading edge of the hole and the tip of the spinner insures that the optical tachometer will trigger as the hole, now designated as No. 1, passes the light beam.

If holes on the engine you are balancing are numbered by the manufacturer, review the maintenance manual for specific guidelines on tape placement. Since the laser can be positioned or aimed to trigger at any desired point in the engine rotation, we will define its point of trigger at the six o'clock position as viewed from the front of the engine looking into the intake.

When the position for the tape is defined, clean the area with a quality degreaser, then dry the area thoroughly before applying the tape. Make sure there are no bubbles or loose edges in the tape. These act as airfoils at high speed and will cause the tape to come loose which results in the loss of tach signal.

When the tape is attached, rotate the fan until the tape is at the six o'clock position. Mount the laser tachometer on top of the wing. Power the laser from the analyzer and aim the laser beam to strike the reflective tape midway in its length. Tighten the mount and recheck the alignment. Adjust as necessary.

For this example, we will place the vibration sensor at the twelve o'clock position. Again, be sure to use the recommended sensor mount. Cables from the tachometer and the vibration sensor are routed along the fuselage to the cockpit and secured about every two to three feet with duct tape. The cables are connected to the analyzer then the analyzer is placed in the cockpit where it will be operated by the technician running the engines.

Fan Trim Balance

By Jerry Justice June 1999

At what speed do you bal-ance?
If you are attempting to stop a vibration or noise at a reported speed (RPM), balance at the speed where the problem is reported. Fan-generated vibration and noise usually occurs at a specific power setting, but may continue to varying degrees throughout a power range.

For routine balancing, engine vendors will usually specify a speed, RPM, or power setting at which the balance should be performed. This isn't always the best path to a solution for the problem. Vibration changes in amplitude as speed changes. It may even reoccur at multiples of the problem speed.

Vibration sensors vary in size weight and shape.

You may have had the same sort of experience with your car. Suppose you drive to work every day at 60 mph. You begin to notice an increasing vibration in the steering wheel and you suspect that one of your front wheels is out of balance. You have the wheels rotated and balanced at your local garage. The next morning, you notice a slight vibration while accelerating through 37 mph that wasn't there before but everything seems fine at 60 mph. When you speed up to 74 mph to pass a truck, there's the out of balance vibration again. As you slow back to 60, the vibration goes away. What's going on here?

The influence of a mass on the balance of a disk changes with speed. You can't always get rid of a vibration at one speed without moving it to another speed — a process my boss calls "pushing down on a balloon." You push it down at one point and it pops up and grows at another.

Because of this phenomenon, it is best to balance at several speeds that encompass the normal range of in-flight power settings. Balancing at several speeds in the flight range doesn't lower the vibration amplitude and noise to its minimum achievable level at each speed, but rather brings each to an acceptable level without pushing the amplitude of any one too high.

What about speeds outside the normal flight range? The noise may be a little higher at takeoff power — but how much time will the power remain at this setting in a five-hour cross-country flight? On the other side of the coin, how much time will the power be at a cruise setting? Are you getting the picture here? If you must sacrifice "quiet" at one power setting to achieve it at another, consider which scenario will be more advantageous in the operation of your particular aircraft. Don't forget that you can also include maximum power as one of the multiple balance speeds if you wish.

Collecting the data
Now that we have a tach input, a sensor input, and a target speed, let's collect the data we need to arrive at a solution. Make sure the aircraft is positioned into the wind and that the wind is below 10 knots if possible. Crosswinds and wind gusts are not acceptable for balancing because they cause the fan speed to fluctuate. This in turn causes conditions to change. Reference the previous section on balancing speeds.

The analyzer is setup with the program for balancing the specific engine. The sensor type and speed are selected and the analyzer instructs you to start the engine. Before you begin collecting data, make sure the engine is at normal operating temperature. The analyzer will instruct you to accelerate to the first speed. When the speed is stabilized, begin the collection of data as instructed by the analyzer.

When data collection begins, a sequence of events occurs. As shown in the illustration on page 52, the vibration sensor is mounted at twelve o'clock and the laser tachometer is positioned to trigger at the six o'clock position as the reflective tape passes through its beam. In Figure A, on the next page, the mass causing the out of balance condition is located at the three o'clock position.

For the purpose of this example, let's assume the fan is rotating at one (1) RPM in a clockwise direction as viewed from the front of the engine looking aft (FLA). In Figure B, as the rotation progresses, approximately 15 seconds (90 degrees of rotation) have passed and the mass is now located at the six o'clock position. Accordingly, the reflective tape has rotated to the three o'clock position.

In Figure C, the mass has reached the nine o'clock position and the reflective tape has entered the laser beam. The laser beam is reflected back to the receiver of the laser tachometer and sends a pulse to the analyzer. At this point, 30 seconds (180 degrees of rotation) have passed.

In Figure D, the mass has reached the sensor position. The upward movement compresses the piezoelectric element inside the sensor and results in a voltage being produced, sent to the analyzer and measured as 10 mV. This value is then held in memory. The time is now 45 seconds or 270 degrees into the rotation. The rotation continues back to the position in Figure A and the process is repeated.

After a sufficient number of data samples are collected, an indication on the analyzer screen shows averaged vibration amplitude, angle, and speed. As the number of samples increases, the percent of error is steadily decreasing. The user is prompted by the analyzer to stop the data collection with a key press. If using multiple speeds, the analyzer will prompt you to set power for the next speed and the process is repeated.

For this example we will use a single speed. Using the one RPM scenario above, we determine that the amplitude of the vibration is 0.5 IPS because the 20 mV/IPS sensor sent a 10 mV signal to the analyzer. The vibration event occurred fifteen seconds AFTER the tach event. The analyzer converts fifteen seconds of a rotation to 90 degrees.

With the reflective tape in the position where it triggered the tach event, we can see that the heavy spot on the fan is located approximately 90 degrees opposite the direction of rotation from the vibration sensor. (Refer to Figures C and D above). Of course, in actuality, all this occurs at a rate from 5,000 to 20,000 times faster but the results are the same. The task now is to counter the effects of that mass.

Applying a solution
The first thing we need is an influence coefficient for the application. The coefficient influence is the amount of weight, corrected for lead or lag, required to counteract the mass causing the vibration.

Without getting too deep into theory, the lead or lag correction is done automatically by the analyzer. It is simply a correction to compensate for the transmission characteristics of the engine being balanced and a correction for the type of measurement being made.

Peak Velocity and Peak Acceleration physically occur at different points in the rotation of the unbalanced fan — 90 degrees to be exact. We may enter a known influence for the engine before acquiring data or calculate one by adding a "test weight" to the fan and collecting data from another run to see how the weight affected the previously recorded condition.

For this job, we will enter the influence of 20 gram IPS (20 grams of weight for each inch per second of measured velocity). Again, just to make things understandable, I'll use the example just as it was stated above in the previous paragraph and say that the mass is located at 90 degrees. The solution then should be to add weight at 270 degrees, in an amount sufficient to counteract the out of balance condition.

If the analyzer is programmed for the specific engine type, the location of the tach trigger and vibration sensor, as well as the class weight set are each a part of the program.

When using class weights, the limitation is the value and number of available class weights. In this job, we will use a class weight set as follows: There are five weights in the set; a -1, -2, -3, -4, and -5. Their values are 2.25, 4.52, 6.8, 8.69, and 11.44 grams respectively. For this balance job, the analyzer directs us to place a -1 and a -3 weight in hole number 3. Remember that for this job, the tape designates the number one hole. The holes are numbered in ascending order sequentially from that point, opposite the direction of rotation.

In the figure below you can see that the No. 3 hole is 180 degrees from the location of the mass. The -1 and -3 weights total 10.05 grams, which is near the 10 grams calculated as necessary to counter the 0.5 IPS. Remember that 20 gram IPS is the influence for this job.

After installing the weights, it is necessary to enter the actual amount of weight and the location back into the analyzer. In this case, we would use the class weights (-1 and -3) in the hole number (hole #3). A second run is then conducted to verify and refine the solution as necessary. This sequence of events is repeated until the fan is at an acceptable vibration limit.

The typical problems associated with noise and vibration from an unbalanced fan are decreased life of avionics and airframe components, as well as discomfort and stress on the crew and passengers. As the maintenance professional in charge of maintaining a smooth running aircraft, it's to your advantage to keep the air clear and quiet.