Chuck Yung
EASA Senior Technical Support Specialist
One of the most briskly debated issues in our industry is the comparison – and procedures for – vacuum pressure impregnation (VPI) versus dip & bake. For this article, I have expanded the discussion to include trickle epoxy and B-stage coils. Service centers that have a VPI tank will quickly point out the many benefits of VPI, such as better sealing of the windings and improved heat transfer from the winding conductors to the frame for enhanced heat dissipation.
Form and random windings have two distinctly different issues. For the form-wound machine, resin penetration is the biggest concern – giving a clear advantage to a VPI process. For random windings, the concern is retention of the resin.
Let’s start with random-wound motors, which represent the majority of rewinds by our industry. With unconstrained magnet wires through the slots and coil extensions, it should be obvious that varnish can easily seep in between the wires and into the slots. The longer the slot, the more concern we should have about varnish finding its way into the middle of the slots.
Tip: Dip the stator with the bore vertically, as that makes it easier for air to escape the slots so that varnish can fill the voids. Just make sure the core is not loose in the frame.
Preheat the Stator
An important step is to preheat the stator before resin treatment. This is not only to warm the windings so the resin flows better, it’s also critical to “gas off” (evaporate} the wire lubes that are used on magnet wire during manufacture, so the resin will bond. Failure to preheat random windings before resin treatment will result in resin “beading up” like water on the surface of a waxed car. The recommended preheat temperature varies by resin manufacturer and product, typically in the range of 200°F to 275°F (95°C to 135°C), with the stator being allowed to cool to the recommended process temperature for your resin (typically between 120°F and 175°F / 50°C – 80°C).
Tip: Don’t get in a hurry and use fans to cool the stator more quickly while preparing to VPI it. That cools the frame surface to the desired temperature, but heat within the core transfers to the surface and may overheat the resin.
Keeping Resin in the Slots
It should also be obvious that keeping the resin in the slots is the challenge. Just look at the amount of resin that drains from the slots during the bake cycle. The solution to resin retention in a random wound stator is to dip & bake the stator first using the VPI resin, and then to VPI process it. That may sound backwards, but the dip resin penetrates the slots and (mostly) fills gaps between wires. As the dipped stator cures, some of the resin drains out because there is little in the way of tape to seal the slots. Partially cure the resin (one suggestion is that it should be tacky to the touch), and then cool the stator to the appropriate temperature for your VPI resin.
The reasons for partially curing the first treatment are that the second treatment will bond better to the surface, and to save energy and total baking time. See
Figure 1. By VPI processing the stator after it’s been dipped & baked, there are fewer voids for the VPI resin to penetrate and fill. Retention is considerably better than the conventional methodology of “VPI followed by dip & bake.” The manufacturer I once worked for used this method for random-wound 2300-volt stators. That was key to our success with 2300-volt random windings. Those of you who remember the rush to random winding higher voltage machines remember that most were unsuccessful.
Trickle Epoxy Method
An alternative to VPI or dip & bake for random windings is the trickle epoxy method. For cores up to approximately 12” (300 mm) length, the trickle method is in many ways superior to traditional dip & bake. During my years of work in a service center, we did quite a bit of testing to maximize random winding life for motors and alternators used on offshore oil rigs. We tried everything: fully lacing the end turns, lacing over a felt wrap, encapsulation, multiple VPI treatments, etc. Trickle epoxy when properly done outperformed every combination of VPI and dip & bake processes. Eventually, we were able to run a cutaway motor submerged in an aquarium for a trade show. Our sales force was ecstatic.
The method of processing is key to obtaining good results with the trickle method. Upend the stator so the bore is vertical with the connection end up. Energize the stator with approximately 10% of rated voltage and heat the windings to 130°F (55°C) or so; then slowly pour the trickle epoxy onto the winding extension. A steady stream approximately 1/8” (3 mm) diameter works well. The goal is to “trickle” (hence the name) the epoxy so that it flows into and through the slots, as well as covering the connection end. Energizing the windings accomplishes two things:
- It heats the windings so the epoxy flows more freely, improving penetration.
- The windings vibrate at twice line frequency, which facilitates movement of the epoxy through the slots.
Once the coil extension is thoroughly coated, turn off the power and flip the stator so the connection end is down, reenergize the stator, and follow the same procedure to coat the other end.
An additional benefit of the trickle epoxy treatment is that it reduces process time as compared to the time required to oven cure the dipped or VPI processed stator.
For form-wound motors, VPI has the clear advantage over a dipped & baked coil. See
Figure 2. An oft-quoted claim is that VPI processing a form wound stator results in a 10°C (18°F) lower temperature rise than the same motor that is only dipped and baked. Although primarily used for machines rated 11 kV and higher, on-site rewinds, or stators too large for the VPI tank, a B-stage coil is an alternative to VPI.
Most coil vendors can supply VPI coils or dip & bake coils. Most, but not all, VPI tapes are a dry mica tape and rely on resin penetration to saturate the tape to obtain the full dielectric strength. There is not universal agreement over the use of film backing on mica tapes for VPI. One manufacturer has never deviated from their preference for the film-backed mica tapes, while others argue that it inhibits VPI penetration.
The VPI Advantage
Regarding VPI, the recommendation is to pull a vacuum to 1-3 Torr (1-3 mm of Hg) and hold for two hours. Transfer in the resin and maintain vacuum of < 3 Torr (<3 mm of Hg) for a minimum of 2 hours. Pressurize to at least 90 psi (620 kPa) and hold the pressure cycle for a minimum of 1 hour per kV of rating. For example, a 4 kV winding should receive a minimum of 4 hours pressure cycle. That’s a far cry from the old days when a VPI process was likely to mean “1-hour dry vacuum, 1-hour wet vacuum, and 1-hour pressure.”
Tip: When VPI processing an armature, monitor resin transfer and briefly open the pressure valve as the resin level approaches the risers. A few seconds of dry air or nitrogen will dissipate the foam so you can tell the actual resin level. Resin level should never reach the risers.
The “Bumping” Process
Europeans often take the VPI process a step further than most in North America, with a step they refer to as “bumping.” The rationale behind bumping is that, once full pressure is obtained, entrained air bubbles are compressed but unable to escape. The thinking is that releasing the pressure gives trapped air a chance to expand and escape. Then full pressure is reimposed.
Verify Penetration of Resin
Rather than processing a sacrificial coil with the stator, curing it and cutting it open, it’s far better to unwrap the wet coil to verify complete penetration of the resin. The stator should be left in the VPI tank while the coil is inspected in case additional process time is required.
For machines rated 7 kV and higher, VPI penetration using conventional resin is not always reliable. Those who regularly rewind machines rated around 11 kV and higher often rely on coils where the tape contains the catalyst for the resin. This allows the tank resin to be much thinner to ensure adequate penetration of the tapes. The thinner the resin, the greater the chances of VPI resin penetrating all those layers of tape.
An alternative to VPI – especially for higher voltage machines – is the B-stage coil, which uses resin-rich tapes during the coil-making process. We used platens to form the straight section of each coil to the required size for the slots. We imbedded heaters in the platens to cure the tapes only in the straight section. For our larger on-site rewinds, that left the coil extensions flexible for handling.
Tip: For those in the coil business, use cartridge heaters in the platens, and liberally smear the heaters with heat sink compound to extend the heater life.
Differences in Varnish, Resin
Hopefully this article has given you some insights into different treatment methods for windings. There are a few other environment-related tips worth sharing. Although many of us use the terms varnish and resin interchangeably, they are distinctly different. Varnish includes a solvent, making viscosity adjustment a matter of adding a solvent (e.g., Xylene) to a dip tank of varnish to thin it. Varnish contains approximately 50% solids – meaning it contains a lot of solvent, so varnish is strictly for dip tanks. Trying to draw a vacuum on a tank of varnish would remove solvent and be an environmental mess, as well as being an explosion and fire risk. Resin is solventless, “100% solids,” and suitable for VPI use. But adjusting the chemistry is more complicated than just adding a drum of solvent. Whether varnish or resin, it’s important to regularly send in a sample to your vendor for evaluation.
More tips: As more regulatory agencies ask us to report the amount of volatiles and carbon released into the atmosphere, it’s helpful to weigh each stator before & after burnout. For billing and inventory control, weigh them before and after the VPI / dip & bake process. You will develop a better understanding of how much resin is used per rewind. It looks like there is a renewed push for governments to impose carbon taxes, so we all need a clear understanding of how much carbon we are releasing to the atmosphere.
ANSI/EASA AR100
More information on this topic can be found in ANSI/EASA AR100
Print