Facebook Twitter LinkedIn YouTube Menu Search Arrow Right Arrow Left Arrow Down Arrow Up Home Arrow Next Arrow Previous RSS Icon Calendar Icon Warning Icon

Articles in Industry Publications

Article

Applying the best of repair best practices: Rewind study continues to pay off with important tips

  • June 2008
  • Number of views: 10596
  • Article rating:

Tom Bishop, P.E. 
EASA Technical Support Specialist

There are certain repair processes that can impact the efficiency and reli­ability of electric motors. Prudent repair practices must not increase overall losses, and preferably should maintain or reduce them. In some cases, repairers can also employ the principles applied by the motor designers and further reduce losses and enhance efficiency. Most of the following material is taken from, or based on, the “The Effect of Repair/Rewinding on Motor Efficiency; EASA/AEMT Rewind Study And Good Practice Guide to Maintain Motor Efficiency.” 

Stator core processing and repair 
Concerns about the possibility of core degradation during the rewind process have been expressed since at least the early 1990s. Higher tempera­ture rated core plate insulation mate­rial greatly reduces the possibility of core degradation during the burnout process. However, a best practice approach is to avoid the possibility of core damage no matter what type of core plate is used. 

The key steps to take during the burnout process are to set the burnout temperature to no more than 680° F (360° C), and use a temperature-sensing device attached to the core being processed to control the oven temperature. Further assurance that degradation will not occur is to use an oven equipped with a water suppression system. If an over-temperature condition is detected, the water spray is immediately activated. This method is highly effective because water changing from a liquid to a gas (steam) absorbs a tremendous amount of heat energy; much more than if simply changing tem­perature by absorbing heat energy. That is, water absorbs as much energy in changing from liquid to steam as it would in theoretically increasing temperature by 540° C (970° F). 

Image

Prior to and following the burnout process the core should be tested, as illustrated in Figure 1. The core can either be loop tested (see Tech Note 17) or tested with a commercial type core tester. Both methods are effective. The watts per unit of weight core loss and temperature rise of the core during the test should be compared to each other (pre-and post-burnout process) and to typical limits. Typical limits for core loss are about 4 watts per pound (9 watts per kg) and for temperature rise about 15° C rise (27° F). Further, the watts loss per unit of weight should not increase more than 20% during the process, and best practice would be for neither temperature nor watts loss to increase at all. 

If the core test or visual inspection reveals core damage, the core should be repaired prior to winding. Minor defects such as splayed or flared laminations should be tamped back in place. A technique that is usually effective for flared laminations is to bend the teeth at the end of the slot at the vertical middle. That is, create a bowed effect, with the center bowed away from the core.

Tamping the teeth (by striking with a slight down­ward angle at the top of the teeth) back to the core causes the bowed teeth to act as a clamping mechanism. 

If lamination material has been eroded but the extent of the damage is minor, the laminations can normally be un-stacked in the affected area and restacked after repositioning the laminations to fill in the area that was missing lamination material. Removal of complete laminations should be avoided. As a guide to determining the limit of “minor” missing core ma­terial, it should not exceed 2% of the length of the core, or not affect more than 10% of the number of teeth. If damage is more extensive than these guidelines, best practice action steps would be to replace the damaged laminations with new laminations, restack the core with all new lamina­tions, replace the core, or replace the entire motor. New laminations can be obtained through firms that specialize in laser cut laminations, using a good original lamination as a template. 

Following core repair, always retest the core before proceeding with the rewind. The watts loss and tem­perature rise should both be less than prior to repair of the core damage; and the watts loss and temperature rise levels should be within the typi­cal limits given above. 

Winding practices 
The best practice goals in winding are to maintain or reduce the winding resistance and to maintain or improve the motor performance characteristics. The winding resistance is maintained by using the same size wire area, and the same mean (average) length of turn. Increasing the wire size area, reducing the mean length of turn, or doing both, reduces winding resis­tance. That reduces the stator winding I2R losses as the winding resistance is the “R” in the I2R equa­tion. Reduced losses mean that efficiency increases and heating is reduced, which length­ens the thermal life of the insulation. 

Reducing the length of the coil extensions is the only method of re­ducing the mean length of turn (MLT – the av­erage length of a single turn of the winding, as depicted in Figure 2) during rewinding. The core length is fixed, thus the only variable is the length of the end turns. The end turn length can be reduced to the point that any further reduction will result in a side force between the coil and the end of the slot. Going beyond that point can result in a winding ground fault due to the coil pulling against the slot cell extension and eventually breaking through it. 

Image

Another consideration with coil extension length is that by reducing it, the surface area exposed to cooling air is also reduced. Although this would rarely be a significant pos­sibility, it should be kept in mind especially when there appears to be an opportunity to significantly reduce the coil extension distance. An example would be the pos­sibility of being able to reduce an approximately 4-inch (100 mm) coil extension to just less than 3-5/8 inches (90 mm). The 10% reduction in exposed length could increase heating due to less heat transfer from coils to cooling air. The effect of a +/- 10% change in MLT for a variety of motor power ratings is illustrated in Table 1. 

Image

Increasing wire area is possible if slot space is available. A benefit of increasing slot fill is that there will be less space between wires, mak­ing varnish penetration and bonding more effective and resulting in better heat transfer as air pockets (voids) are reduced. However, making the wires fit too tightly in the slot can result in damage to the wire insulation as the winding is tamped in place with excessive force; the slot liner can also be damaged. It can also increase the time required to insert the coils. The increased wire area reduces copper (I2R) losses and reduceswinding tem­perature. The effects of these changes are increased efficiency and longer winding thermal life. 

Mechanical repairs 
Replacement bearings should be equivalent to those provided by the motor manufacturer. Selecting an incorrect bearing, such as changing from an open to a sealed bearing, will increase friction losses in the bearing, thus reducing efficiency. Incorrect in­stallation of a bearing—for example, driving it on by pressing against the outer race—can damage the bearing and cause rapid failure. Even a slight amount of damage can result in a noisy bearing. 

Bearings of C-3 internal clearance are the standard for most electric motors. A contact-type sealed bearing can create more friction than a shielded, open or non-contact sealed bear­ing. The increased friction results in a slight drop in efficiency. To avoid degrading efficiency and reducing reliability, it is good practice to remain with the open bearing style installed by the manufacturer. 

Fill the grease reservoir cavity to about one-third to one-half full. Over greasing a bearing, even by a small amount, increases friction losses. This not only reduces ef­ficiency (by 500 watts in one case cited in the EASA/AEMT study); it also causes local overheating, which can seriously reduce bear­ing life. Allow the motor to oper­ate unloaded long enough for the bearing temperature to drop. The drop in temperature indicates that the bearing has expelled excess lubricant and seated itself into a stable position. In essence, this denotes the bearing “break-in” period as shown in Figure 3. 

Image

When application and environment dictate the installation of sealed bearings for reasons of reliability, some increase in bearing temperature and friction losses should be expected. A better alternative is to consider the installation of non-con­tact seals or bearing isolators, which exclude contaminants without causing friction. Some bearing manufacturers also offer non-contact sealed bear­ings. 

Ventilation issues 
Unfortunately, there is little op­portunity to improve efficiency by changing fans or ventilation, except in rare cases where a large increase in wire current capacity is possible, such as when converting from aluminum to copper wire. In such a case the fan size can be reduced if the aluminum wire is replaced with the same size copper wire. Reducing fan size or airflow reduces windage losses at the expense of increased winding heating. The converse also applies; increasing fan size or airflow reduces winding heating at the expense of increased windage losses.

Although we may not have opportunities to reduce losses with ventilation issues, good practices will result in maintaining the original efficiency. 

Installing an incorrect fan, or locating the fan or fan cover in the wrong position (improper clearance between the fan and fan cover), can affect windage. A fan that moves more air, i.e., has higher flow, inher­ently increases windage loss and reduces efficiency. Conversely, a smaller or lower flow fan (see Figure 4) reduces windage but also reduces cooling due to the lower airflow. If a fan has a broken blade or blades, it should be replaced. The miss­ing blade(s) reduce airflow and may increase vibration due to mechanical unbalance. 

Image

Windage varies among fan designs, depending on factors such as diameter, the number and size of blades, mate­rial, and surface finish. The single most important variable is fan diameter. All else being equal, a smaller diameter (D1) trimmed fan moves considerably less air than the larger original diameter (D2), by the ratio: [(D2 / D1)3]; and symmetrical fans of different diameters vary by [(D2 / D1)4]. Thus a propor­tional replacement fan that is 5% larger in diameter compared to the original requires 22% more power to drive the fan. That diverted power is lost power, which reduces motor efficiency. 

An incorrect fan cover may reduce air flow; an example is where the open­ings in it are smaller than the original. Location of the fan relative to the cover is also important. If the fan is too close to the fan cover, cooling air flow will be reduced. A damaged fan cover may result in reduced air flow, as the air may “leak” through the cracks or become turbulent due to a section that has broken off. Even with the correct fan cover, air flow will be reduced if it is not free from dirt or other material that blocks or restricts the vent open­ings. 

Motor design aspects 
Increasing magnetic flux increases core losses and therefore heating of the windings. The results are reduced ef­ficiency and winding life, and reduced reliability. Reducing the number of turns or changing the coil span or connection can increase magnetic flux. Doing the opposite, e.g., increasing turns, reduces magnetic flux. However, the reduced flux reduces torque capa­bility and typically results in higher current for a given load. The higher current means increased I2R losses, reduced efficiency and increased heating. Thus to maintain efficiency and reliability it is best not to change the magnetic flux level of the wind­ing. All else equal, a slight increase in magnetic flux density is preferable to a slight decrease. That’s because a magnetically stronger design has less slip, reducing the rotor losses. 

Repairers of­ten prefer to use lap windings be­cause all coils are the same. This is acceptable provided that the new winding is chosen such that the flux per pole is not changed. Single-layer lap windings are sometimes used motors, because the coils are easier to insert and no separators are required, thus allowing more room for copper. Double layer lap windings give a better flux distribution in the core than single layer windings, and in no circum­stances should a double layer winding be replaced by a single layer wind­ing. To do so will reduce efficiency. 

Conversely, changing from a single- to a double-layer lap winding may reduce losses and improve efficiency slightly. 

If the stator core is partially or fully restacked, a reduction in the total number of laminations reduces the core iron volume, effectively increasing magnetic flux densities. The higher flux levels increase core losses and heating. Improper restacking, such as by not compressing the core tightly enough, or by over-tightening the core, can lead to increased core and stray load losses. A key to a successful restack is to assure that the original core length is maintained and that all of the removed laminations, or equivalent replace­ments, are installed in the core. 

The rotor I2R losses can be in­creased by reducing the end-ring cross-section or by increasing the resistance of the rotor bars and end-rings. The repair process does not normally affect the rotor resistance, unless the rotor is rebarred. If the rotor is rebarred, it is critically important to have the bars and end ring materials tested to determine, and duplicate, the material resistance (or maintain the opposite characteristic, conductivity.) 

If the rotor surface must be cleaned up by machining, a sharp cutting tool is a necessity. The usual reason for need­ing to machine the core is to correct smearing caused by a stator to rotor core rub. Grinding the rotor surface, or machining the rotor core with a blunt tool or at an incorrect surface speed, can result in smearing the laminations together. The smeared laminations probably will not become hot at running speed due to the low rotor frequency of only a few hertz. How­ever, the warmer core area can create a thermal bow, resulting in vibration and an unequal air gap. 

An unequal air gap can cause circulating cur­rents in the stator wind­ing, resulting in increased I2R losses. Repairs to the stator frame or end bracket rabbet/spigot fits that reduce stator-rotor concentricity increase air gap eccentricity, and can result in circulating currents that increase I2R losses. 

An excessive air gap will increase magnetizing current and also increase I2R losses. Machining the rotor diameter to increase air gap can reduce losses at the expense of power factor; however, too great an increase in air gap will increase losses. This should only be done when the manufacturer’s design air gap tolerance is known to the service center. 

Image

Stray load losses, illustrated in Figure 5, are typically 10-20% of total motor loss. Stray loss can increase if the air gap surfaces of the laminations are smeared together. Stray loss will also be increased if the air gap is un­even (i.e., stator and rotor not concen­tric) and may be increased if a wrong replacement rotor is installed. 

Closing comments
Of the things that affect efficiency, a typical repair only influences the core, winding (I2R), and friction losses. These and other key topics have been addressed in these best practices. Documenting the before and after core loss, comparing winding resistance to the manufacturer’s re­cords, and confirming the bearing type provide assurance to you and to the customer that the motor’s efficiency was maintained during the repair.

Documents to download



Comments are only visible to subscribers.