Mike Howell
EASA Technical Support Specialist
Relevant standards including IEC 60085 and IEEE 1 have similar definitions for electrical insulating materials (EIM) and electrical insulation systems (EIS). To summarize, EIM are materials suitable for separating conducting parts at different voltages, and EIS are insulating structures containing one or more of these materials.
As with any system, there is an interaction between the materials used, and the insulation system developers take great care to ensure that this interaction does not lead to undesirable outcomes. For example, it is possible for two materials (EIM) classified individually at thermal class H (180°C) to have thermal endurance in a system (EIS) limited to thermal class F (155°C). Far worse outcomes could exist if material compatibility is an issue. At the service center level, our resources are generally insufficient for these types of insulation system development activities. For this reason, two approaches often seen are (1) relying on a third party (e.g., resin manufacturer) to provide a qualified insulation system bill of materials, or (2) applying commonly used materials based on their individual ratings. The first approach is strongly recommended, and the second approach can lead to disaster.
Mitigating Risk
Generally, industry defines risk as the effect of uncertainty on objectives (ISO 31000). An effect in this definition is any deviation from an expected outcome, and effects can be positive or negative. Some examples of negative effects relative to areas where many service centers would establish objectives are injuries to personnel (safety), product failure (quality), late completion (delivery) and over-budget completion (cost). Ideally, successful service centers engage in as many low-risk, high-reward activities as possible.
Any activity can be placed in one of the four categories shown in Figure 1. Engaging in “A” activities (low risk / high reward) are most desirable while “B” activities (low risk / low reward) are acceptable. “C” activities (high risk / high reward) are a gamble, and “D” activities (high risk / low reward) should never intentionally be undertaken. Numerically, where any organization draws the line between high and low can vary based on their business practices, but this tool does not have to be quantitative. Sometimes, it is helpful just to think about where you would intuitively place an opportunity on that figure.
One tool commonly used in the automotive industry for reducing risk associated with implementing new products or product changes is Advanced Product Quality Planning (APQP). A simplified APQP timing chart is shown in Figure 2, and while this article cannot cover APQP in detail, this figure can help demonstrate one way that an insulation system change might be controlled. Note that prototypes (samples) are typically produced to begin validation of the change. The term prototype is used loosely here. For insulation system changes, this might include sealed-tube testing, thermal cycling, voltage endurance testing or a variety of other accelerated life tests where appropriate. If the change appears to be successful, a pilot rewind should be performed, and some criteria should be established for evaluating the pilot rewind before moving into production. This could range anywhere from final electrical testing up to a certain time in-service depending on the complexity of the changes and the risk involved with proceeding. Anyone interested in learning more about APQP should visit the Automotive Industry Action Group’s website (
aiag.org).
Case in Point
In the late 1990s, a service center received an order to rewind two 500 hp (375 kW), 4 pole, 6.6 kV stators for a marine application. The global VPI insulation system chosen had decades of field experience and development and coincidentally was the same insulation system used by the OEM. Within one month of service, both stators experienced ground failures.
Initial speculation was that during failure analysis, a problem with the VPI process would be discovered. But these were routine stator rewinds for the service center, and their VPI process was well-controlled. Both failures occurred near the end of the core, and in each case, there was visual electrical damage to the stress grading material. While removing the materials down to the conductor, strange pockets of a light-colored substance with powder-paste consistency were located. It was clear that the failure occurred through those pockets back to the interface between the stress grading and slot corona protection material to ground.
Rewind the clock to about two months before the coils were manufactured. A purchasing agent in the coil manufacturing facility was working with one of their suppliers on ways to cut costs on “commodity” type items. They changed an adhesive tape used for finishing layers of insulation, and the replacement they chose was not much different than a typical office adhesive tape. The previous tape was an electrical grade product with adhesive having an appropriate thermal rating. Some quick testing confirmed that the inexpensive replacement tape’s adhesive would become light-colored with a powder-paste consistency after a typical winding cure bake cycle. So, this tape should have never been introduced into the insulation system, but at that time, “sacrifice” materials were not part of the bill of materials.
To make things worse, space and mass are critical in marine applications, and this is often reflected in design. The coil extensions in this stator winding were very short, and the scarf joint between the machine-taped portion of the ground insulation and the hand-taped portion was located partly under the stress-grading material. This was a bad idea to begin with, and it was made much worse because of the adhesive tape used to terminate each tape layer (see Figure 3). It was found that different electrical testing procedures would have identified the problem before delivery, but they would not have prevented the rewinds. The root cause was insufficient material control that allowed substandard materials to be introduced into the winding.
Control Change
In general, change should be intentional and well-controlled, and this certainly applies to insulation system changes. Spend time thinking about the potential effects of any change (good or bad) and formalize this process when appropriate. Industry research has shown that end-users consider EASA service centers to be trusted partners, and we should seek that type of relationship with our suppliers. That is, lean heavily on the insulation system development work undertaken by the material manufacturers and take advantage of their qualified electrical insulation systems. This will reduce your risk and increase your customers’ reliability – low risk/high reward.
AVAILABLE IN SPANISH
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print