Machine Balancing Article

From Imrtwiki

Return to Precision Maintenance

How Well should your Machines be Balanced? A big Opportunity

I recently ran a best practice meeting on Precision Rotating Equipment and it was not a surprise that most of the people that turned up were involved with condition monitoring. There would be little dispute of the fact that a key driver for improving equipment reliability is increasing the precision of how machines are designed, manufactured, assembled, installed and repaired. Condition Monitoring people tend to see the final result of the level of precision used and can usually measure lack of precision with the parameters such as Velocity Vibration. One easy opportunity raised during the meeting for improving the precision for rotating equipment was by specifying better Balance Standards for new and overhauled machines. If you don’t fully understand what balance is then do this simple experiment. Find a ceiling fan and observe the level of wobble at full speed, then securely tape a large lump of plasticine or Blue Tac at the end of one of the blades and run up the speed and watch the increase in wobble. Thats a balance problem and if left uncorrected will eventually cause damage. An additional attached mass away from the axis of rotation causes imbalance and is measured by vibration (or wobble).

A few of the organisations attending the meeting were very focused on improving equipment reliability and indicated that they were specifying ‘G-1.0’ Balance Grade on overhauls of equipment like electric motors, pumps, fans and turbomachinery. It is very possible "G-1.0 Balance Grade" means nothing to you, so let me explain. It comes from the dominant international balance quality standard, which is ISO 1940. The table below shows the recommendation from ISO 1940 for different machines.

You will quickly note from the table that fans and pumps are specified into G 6.3 balance grade, which is two levels worse than the G 1.0 grade suggested in the meeting. My cynical view is that the people attending equipment related standards meetings are not paid, so they are filled by experts from equipment suppliers who have an interest in not setting standards too high. G 6.3 balance grade means that imbalance will cause a machines rotating shaft at full speed to vibrate at 6.3mm/sec vibration under specific mounting conditions. This does not mean that the machine when installed in your plant will vibrate at 6.3mm/sec but it could do if it was mounted on reasonably flexible foundations. You can see from the chart at the right from ISO Standard 10816 for installed machine vibration that 6.3mm/sec is at the upper boundary of Unrestricted Operation for flexible foundations. A machine running at 6.3mm/sec would generally be seen as a rough running machine by condition monitoring people. The benefit of specifying a higher grade of balance such as G 2.5 or G 1.0 is that the damage that is related to higher vibration such as lower bearing life is minimised or eliminated. Achieving a better balance specification means lower vibrations when machines are installed in your plant. The mechanisms of how lower velocity vibration levels results in higher reliability and lower maintenance costs is very well established. Higher vibration means higher forces in your machine on on the bearings. Halve the load on a rolling element bearing and you can increase its life by a factor of eight.

Anyone who has been involved with balancing knows that the majority of the time and effort is in the preparation and the initial setting up of the shaft in the balancing machine. Achieving a higher balance grade is usually just a few more balance runs and does not increase to overall balance time by much. It is a huge lost opportunity if the balancing process is stopped at G 6.3 grade. The major benefits occur with balancing to G 2.5 but again as suggested by the meeting attendees, going the extra step to G 1.0 balance grade will still give additional reliability and maintenance cost benefits. Not being able to achieve a G 1.0 balance grade in every situation should not be a deal killer but should raise the question of why it could not be achieved. There are many reasons why there might be problems balancing specific machines but some causes such as metal fatigue cracks, loose motor winding or items loose within the machine need be followed up. So how much extra should you have to pay for specifying a G 1.0 balance grade? As long as there is some flexibility for difficult to balance machines it should not be large and the payback should easily justify the cost. Some organisations have negotiated balancing to G 1.0 grade as a standard when negotiating large supply contracts with motor overhaul shops. One way to specify this might be achieve G 2.5, plus two more balancing runs and to provide details of the level reached.

Balance specifications on new equipment should be considered in the same way. A statement in an equipment specification of ‘balanced to ISO 1940’ should not be treated as satisfactory. If you have an opportunity of specifying a better balance grade then you should do this, unless it can be shown not to be necessary though achieved vibration levels from actual service experience. One issue I have read about on the web is with the world’s procurement organisations consistently buying the lowest priced items. It has been observed with high volume manufacture of rotating equipment that some suppliers are eliminating the step of shaft balancing to minimise their costs. On average machines may be within balance specification due to manufacturing tollerances but natural variation means that some of these machines may suffer from imbalance.

Another issue I have seen discussed is the shape of a velocity vibration spectrum showing a machines running speed harmonics. The classic shape shown at the right has the 1X rotational speed peak substantially higher the other peaks. This is becaue all imbalance vibration occurs at 1X rotational speed frequency (around 950rpm (cpm) in this specific spectrum). The question is why that spectrum pattern should normally be the case. The suggestion is that the 1X is only higher because we accept a high level of imbalance in our machines. If we improve our balance specifications then the 1X prominence in this type of spectrum can be substantially reduced.

Specifying better grade balancing for new and overhauled equipment will not fix all your existing installed machines that are out of balance. There are a range of machine types where in-situ balancing is practical, especially so for many fans. Again the aim should be to achieve the best in-situ balance that is practical to gain increases in machine life. For many plants the biggest issue with reducing vibrations by in-situ balancing is getting the machine offline long enough to get the job done. In-situ balancing can be a time consuming job as it requiring isolation and de-isolation multiple times. Before you embark on in-situ balancing it is important that the vibration you are trying to eliminate is an imbalance and not one of a range of other problems that can show similar vibration characteristics. Some of these are looseness in bearings or supports, poor stiffness of foundation or support structure or a structural resonance. In many organisations the local condtion monitoring personnel have taken on the role of in-situ balancing. This should be encouraged, even if it is only treated as a skills improvement exercise for CM people. Most skilled CM people have had a go at in-situ balancing at some stage.

If a machine goes out of balance there is always a reason for it and it needs to be understood, just in case the vibration is a symptom of something worse than imbalance. The typical reasons for imbalance are dirt/product build-up on fans impellors, erosion/wear/damage of impellors, debris clogging pump impellors, items breaking off a rotating element, items loose inside rotating elements, shaft distortion and shaft speed increases. More dangerous problems to watch for could be fatigue cracking of the shafts, cracking of the impellor blades or bolts/rivets failing on assembled impellors.

Just a quick side comment on balance terminology that has confused me in the past about when you use the word ‘Balance’ and when you use the word ‘Imbalance’. If we are talking about a known balance problem we call it ‘Imbalance’, as in “Has the imbalance got worse with the ID fan?”. If we are talking about balance in general terms or as a specification we call it ‘Balance’, as in “What’s the balance like on that new fan you checked?”.

Article by Peter Todd IMRt Facilitator NSW