Summary
Although single-channel Datastick VSA™ Vibration Spectrum Analyzers are not designed for balancing, It is possible to balance some rotating equipment in certain circumstances, using a single-channel Datastick VSA Vibration Spectrum Analyzer.

Details
Sometimes a piece of rotating equipment needs to be balanced, but it is impractical to obtain phase information due to large fan diameter, inaccessible scanner location, or other factors.

On many large fans, it is not practical to install a phase sensor due to air turbulence and the likelihood of the temporary wire getting caught in the fan and the sheer difficulty of mounting the sensor. On the other hand there are applications where the number of motor starts in a day is limited or the difficulty in shutting down the equipment make this method impractical.

Without phase information, it is impossible to use dedicated balancing hardware since the calculations depend on an accurate phase reference. In such instances, it is possible to obtain an acceptable balance using a single-channel Datastick VSA Vibration Spectrum Analyzer and the four-run balancing method.

The four-run method uses total vibration amount only, and involves making a baseline measurement run and three test runs — four runs in all. It is much easier if you draw some pictures as you go along. In fact this balancing method relies more on drawing good pictures than on doing any calculation.

Important Safety Note: The procedure requires that you mount a test mass, and normal safety procedures must be followed. It is vital to ensure that the test mass is mounted safely and securely to avoid injury or machine damage.
It’s also wise to clean the equipment before balancing because dirt buildup can unbalance a machine, and you might get lucky and balance the machine by just by cleaning it. On the other hand, if you balance a dirty machine, it may become unbalanced if a piece of dirt falls off later (or in the middle of the 4 run process!).

It is always important to make sure you have a good stable accelerometer mounting, and of course it’s vital that you do not move the sensor between runs as that would change the data. This method is completely based on the change in vibration amplitude related to the addition of a known mass.

For this example, we see on the VSA handheld that the overall vibration is 2.49 inches per second. This measurement alone is not enough data to give us a picture of the unbalance, so we start by taking a guess as to the position and amount of test mass. The test mass must be great enough to change the overall vibration reading significantly but not so much as to create an unsafe condition. A reasonable rule of thumb is that 1 ounce of test mass per 100 lb of rotor mass is going to put us somewhere close. The point here is that the test mass must make a significant change in the reading but not a large enough change to cause damage, and that can be a wide range!

Let’s say that the rotor mass is 100 lbs. We will use a test mass of 1 ounce, and we will mount it at a radius of 6 inches. This will change the balance by 6 ounce inches. All test runs must be performed with the test mass at the same radius — in this case, 1 ounce at a radius of 6 inches.

The next step is to take a sheet of paper and draw a circle with radius 2.49 inches (or scale by a percentage if needed). This corresponds to the initial overall vibration value of 2.49 inches per second.

Now we mark the circle and the rotor with corresponding angular marks for test points. The test points don't have to be exactly equally spaced but it’s best if they are as close to equally spaced as possible. For simplicity we have divided the circle into three 120-degree portions. Mark three test points on the rotor, at 0º, 120º, and 240º, with each point 6 inches from the center (6-inch test radius).

NOTE: The illustrations don’t show the 6-inch test radius circle, but all test points are located on a 6-inch radius.

Next we add the test mass to the 0º mark, start the motor and let the system stabilize. Then take a new overall vibration measurement. This gives us an overall vibration reading of 3.34 inches per second. We stop the motor and draw an arc on our diagram with a radius of 3.34 inches from the 0º point on the circle.

NOTE: At this point if the amount of overall vibration has not changed significantly by the addition of the test mass, increase the amount of mass or the radius of the mass and repeat the measurement.

Now move the test mass to the next point (240º in this case), start the machine, let it stabilize, and take measurement 3. This gives an overall vibration reading of 2.38 inches per second. Draw a second arc from the 240º point on the circle.

 Finally we repeat with the test mass in the final position (120 degrees in this case). The overall vibration is 2.01 inches per second. Draw the final arc with a radius of 2.01 inches from test the 120º point on the circle.

If the machine is stable, the arcs will intersect at a single point. Use the intersection as your reference point. (If machine is not stable the arcs will not intersect, but instead will form a triangle. In this case, use the center of the triangle as the reference point.)

We see from the data that the intersection point has a radius of 0.81 inches, so our correction mass should be 0.81 x 6 ounce inches = 5 ounce-inches. We also see that the angle is 167.31 degrees from the zero point. This is the angle where we need to make the correction. (The angle of correction will normally be between the two smallest readings.)

Place the correction mass at the167.31 degree point on the rotor. This is a permanent installation and should be equivalent to 5 ounce inches (in other words, the mass X radius should be 5 ounce inches and the radius is to the center of the balance correction mass).

When the correction mass is safely added, a final check measurement is made. The overall vibration should be in the range of 10 – 25% of the original value. It may be necessary to add a smaller additional correction mass to trim the vibration to within limits.

Since we are working in a real world environment the readings are not as precise as we would achieve on a dedicated shop balancing machine but the results are achieved without the requirement to strip the machine down and take it to the repair facility — and the entire operation can be done in a few minutes.

Derek Norfield
Director of Applications
Datastick Systems, Inc.