You are on page 1of 6

A Single Plane Balancing Method Using Amplitude Measurement Only

Larry A. Meidell Applied Vibration Consulting


The force, generating a vibration due to mass unbalance, is a constant amplitude force in the rotating frame and a periodic amplitude force in the fixed frame. Imagine that you are attached to the end of a rope and being swung in a circular horizontal plane. You would experience only a constant force equal to the centrifugal force, Fc = mass (you) x angular velocity squared x R, the length of the rope between you and the center of rotation. Now imagine that someone else or a thing is attached to the rope and is rotating on that same circular, horizontal plane. Place your hand on the axis of rotation. Here you would feel a periodic force toward your hand each time the mass was in at the same azimuth position as your hand. This force and motion describes a rotating vector whose amplitude can be measured in terms of (a) Displacement, (b) Acceleration, or (c) Velocity. Although all of these parameters are mathematically related, the easiest to visualize is displacement, as might be measured with a dial indicator. Each time the heavy spot is in line with the dial indicator probe, the maximum displacement amplitude, high spot toward the probe is observed and the dial indicator will register its maximum positive reading. When the machine is rotating below its first critical speed the high spot and the heavy Spot are coincident. We all know that vibration is a vector quantity, that is, it is represented by amplitude and direction. The direction is simply referenced to some arbitrary point in the fixed plane. In the example above, your hand (no matter where in azimuth you placed it) became that reference point. A simple way to demonstrate this is with your top loading washing machine. Place a few wet towels concentrated on one side of the tub. Advance the cycle setting to SPIN. Rig the lid switch closed with the lid open so you can observe the tub spin. Start the SPIN CYCLE. Using a grease pencil or similar marking device, hold it against the fixed housing with its marking end close to the spinning tub. Move the grease pencil toward the spinning tub edge until it just leaves a mark each time the high Spot passes the grease pencil. When you turn off the machine and the tub stops youll note that the grease pencil has marked the high Spot of the tub, the location of the towels, which is the heavy spot. If the heavy spot (wet towels in this example) were caused by unequal mass distribution of the tub itself, which you couldnt see, you could still balance it by removing weight at the grease pencil mark or adding weight 180-degrees from mark. This technique only works when the tub is spinning below its first critical speed, that is to say more than 20% below its resonant frequency. This is the dynamic condition where the heavy spot and high Spot are located at the same azimuth position. As a machine attains its resonance speed there is a 90-degree difference between high spot and heavy spot. As the speed increases further, to well above resonance, the high spot and heavy spot are 180-degrees apart. There have been numerous mass balancing machines invented over the ages, each applying their own techniques for deriving mass balancing solutions. They all used vector methods to solve for correction weight amount and attachment location. Most of these machines were not portable. Like the common automotive dynamic wheel/ tire balancer, the component needing to be balanced had to be brought to the balancing machine. In the aircraft business, helicopters in particular, use of hand held balancing instruments began in 1964 with the introduction of Chadwick-Helmuth Model 170 for dynamic balancing of the Hughes 369 (OH-6A) tail rotor while installed on the helicopter. This instrument used a fixed frequency band-pass filter tuned to the 369s tail rotor frequency, 3100 rpm, and displayed the vibration peak amplitude on an analog meter. The phase measurement was determined using a strobe light to illuminate a strip of retro reflective tape on one of the tail rotor blades. The strobe would fire each time the heavy spot was aligned with the accelerometer. This produced a stopped image of the retro-reflective tape at some clock (azimuth) position, therefore a phase measurement. The balance solution for correction weight and azimuth location was obtained by plotting the amplitude and phase on a Chadwick-Helmuth Polar chart containing the 369s tail rotor characterization data. Theoretically this should only take two runs. However not all 369 tail rotors responded the same due largely to manufacturing tolerances and normal service wear. Three or four runs were sometimes required for satisfactory balance. Chadwick-Helmuth understood this and

the need to incorporate a correction feature for their charts for what is termed, like-aircraft differences. They developed a plastic overlay Chart Corrector for phase correction along with weight correction instructions and included them in their Vibrex kits. The phase measurement could become unreliable with this type of equipment, due to instrument errors, strobe light operation, structural resonances (creating nonlinear changes in amplitude and phase), vibration sources generating beats; all reasons other than the mechanical health of the rotating component installation itself. Should one or more of these faults arise, the instrument operator would be unable to accomplish the balance job. If he were knowledgeable about the Four Run No Phase single plane mass balance technique, he could accomplish the job with amplitude measurements only. I dont know who first originated the Four Run No Phase methodology but I know it is used currently to minimize the lateral 1P vibration of the coaxial rotors on the Russian Kamov KA-32 helicopter. I discovered this in 1996 while developing the KA-32 rotor track and balance characterization model for Chadwick-Helmuth 8500 Smart Chart. The Kamov procedure uses a spring motor driven, mechanical, strip chart recorder that draws a time history sinusoidal vibration trace on pressure sensitive paper. The vibration input is through a plunger probe hand held against a specified point on the structure. The positive peaks of this trace are measured with a millimeter scale then drawn on plain paper as shown in the example problem to follow and used to solve for correction weight amount and placement location. Its worth noting here that the classic trial weight installation method, measuring both amplitude and phase, requires two runs (if you dont count the verification run): 1. The original unbalance amplitude and phase measurement. 2. Amplitude and phase measurement following trial weight installation. If a characterization polar chart is available (ala Chadwick-Helmuth and others), theoretically only one run is necessary. Both of these techniques use a phase measurement instrument, requiring more expensive equipment due to additional circuitry, hardware and installation time, i.e. photocell, magnetic pickup or strobe light. The additional cost of a phase measurement feature is a trade off against the two additional runs, equipment simplicity plus reliability of correction weight position derivation of the Four Run No Phase method. Kamov KA-32 Example Balancing Problem Here is an example balance problem on a Kamov, KA-32 six-blade coaxial rotor system. The vibration instrument read out is in millimeters (mm) displacement. I probably should mention here that the instrumentation read out, be it displacement, acceleration or velocity, all work equally well for this kind of problem. I suspect that the Russians chose displacement because it was the simplest and least expensive alternative. The metric system lends itself to scaling for this graphical method 1. Measure the vibration amplitude with the rotor operating at normal speed. This first measurement is the original unbalance amplitude, which well label O. For the purpose of this example well assume 12mm displacement. Stop the rotor. Use a compass to draw a circle whose radius is 12 units of your choice. Mark off the circle at three points, 120-degrees apart and label them 0-degrees, 120-degrees and 240-degrees. See figure 1.

Fig 1. Run 1. Original unbalanced condition 2. Mark the rotor weight installation locations with 0-degrees, 120-degrees and 240-degrees. Select a suitable trial weight. For this example Ive chosen a 200-gram trial weight because I have experience in balancing rotors this size. Youll need to have some balancing information on your rotor. Given the maximum balance weight allowable, select about half that weight for your trial weight.

3. Add 200-grams to the rotor at 0-degrees location and run the rotor at normal operating speed. Measure and record the new vibration amplitude. Assume we recorded 17mm. Note here that the addition of 200-grams to the rotor at location 0-degrees increased the vibration from the original unbalance amplitude of 12mm to 17mm. The increase is the vector sum of the original unbalance and the effect of the 200-gram trial weight. Stop the rotor. Draw a 17-unit radius circle centered at location 0degrees. See Figure 2.

Figure 2. The effect of 200-gram added to 0 location 4. Remove the 200-gram trial weight from the 0-degree location and attach it to the 120-degree location. Run the rotor again at the normal operating speed. Measure and record the new vibration amplitude. Ill call it 8mm. Stop the rotor. Draw a circle of 8-units radius at location 120-degrees on the original unbalance amplitude circle. See Figure 3.

Figure 3. The effect of 200-grams added to location 120

5. Remove the 200-gram trial weight from the 120-degree location on the rotor and attach it to the 240degree location. Run the rotor again at normal operating speed. Measure and record the new vibration amplitude. Assume this reading was 13mm. Stop the rotor. Draw another circle of 13-unit radius at

location 240-degrees. See figure 4.

Fig 4. The effect of 200-grams added to location 240

6. The three, trial run, weight circles intersect at a common point. Draw a line from the center of the original unbalance circle to this intersection point and label it L. This line length is a vector amplitude and its direction is the angle it makes relative to the 0-degree rotor location. In this example the angle is 159-degrees. Scale the line length in units and calculate the correction weight amount. See Figure 5.

Figure 5. Intersection of arcs defines end point of vector L and its direction

Correction weight Wc=0/L x Trial Weight = 12mm/5.6mm x 200-grams = 428 grams

Correction Weight = Trial weight x Original Amplitude/line length (L) Correction weight = 200-grams x 12mm/5.6mm Correction weight = 428-grams In this rotor example there is no weight attachment point at the 159-degree location so the 428-gram correction weight must be split between location 120-degrees and 240-degrees.

7. Construct a parallelogram whose diagonal length equals the 428-gram correction weight in the direction of 159-degrees with sides in the 120-degree and 240-degree direction, the locations available for attachment of weights. Complete the parallelogram as shown in Figure 6.

Figure 6. Weight Splitting Parellelogram Add: 240-gram to 240 location Add: 460-gram to 120 location 8. In this example 240-grams are added to the 240-degree location and 460-grams are added to the 120degree location. These correction weights must be placed at the same distance from the center of rotation, as was the trial weight from which this solution was derived. In summary, the amplitude measurements can be in acceleration, displacement or velocity in this method. All work equally well for helicopter rotor balancing frequencies. The trial weight locations need not be at 120-degree locations. They can be at any angle location as long as they are at the same gramcentimeters, ounce-inches and generating the same centrifugal force. Drawing the amplitude circles will introduce some small measurement errors whose effect will be insignificant. The circles will not always intersect perfectly but by inspection you can estimate the L vector end point and its length. Any small difference here will also be insignificant.

You might also like