Industries / Industrial equipment and services / Industrial services / Motor rewinding
Stripping and rewinding
Stripping and rewinding is the work of removing a failed winding from the steel core, checking that the core is still sound, and winding new coils into it to the original design or a corrected one.
5
Milestones achieved, from 1920 to 2026.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1920
A handbook for the winder
Daniel H. Braymer's "Armature Winding and Motor Repair", published by McGraw-Hill, gathered "practical methods that are used by repairmen and armature winders" into one book. Before a machine was stripped, it told the winder to enter its winding data in a notebook, such as the number of slots and the throw of each coil (how many slots it spans), and to keep a copy as an office record for later repairs; to save one old coil in its original shape as a guide; and to clean the old insulation out of the slots "by scraping and burning with a torch".
Credit
Daniel Harvey Braymer; McGraw-Hill
2001
Winding records go digital
Before stripping, a shop records the winding's design, such as coil pitch, group spacing and internal connections, so the new winding matches. EASA members sent these records in on 4 by 6 inch "winding cards", which EASA stored on microfiche; in 2001 the records moved to a computer database on CD, built into EASA's motor verification and redesign software. By November 2026 EASA counted more than 250,000 three-phase motor windings in the database, contributed by members over decades.
Credit
EASA and its member shops
2014
Cores that would not rewind
A motor's core is a stack of thin steel sheets, each coated so that stray currents cannot flow between them, and core loss is the power the core wastes as heat, measured in watts per pound (W/lb) of core steel. In a June 2014 EASA article, Steve Skenzick of HPS Electrical Apparatus Sales & Service wrote that his service center had begun to see motors built from the late 1990s on that "could not be rewound more than once or twice before core losses increased well beyond acceptable limits", and that some failed within weeks of a rewind when core testing had not caught them, as quoted by Dreisilker Electric Motors in 2017.
Lessontest every core before and after stripping, because a core that survived one rewind may not survive the next.
Credit
Steve Skenzick, HPS Electrical Apparatus Sales & Service
2017
One oven, two outcomes
Dreisilker Electric Motors, whose own method heats the core to about 400 °F and pulls the coils out hydraulically instead of burning them, burned out two stators following EASA's recommendations at a 750 °F oven setting; during the cycle the oven passed 800 °F and its fire suppression went off. By Dreisilker's measurements, core loss in the 100 hp stator rose from 4.678 to 10.101 W/lb with many hot spots, while in the 200 hp stator it fell from 2.473 to 2.028 W/lb with none. EASA's repair standard AR100-2015, as Dreisilker quotes it, says core temperature should not exceed 700 °F (370 °C) for organic and 750 °F (400 °C) for inorganic lamination coatings, and that a burnoff oven should have a water suppression system.
Lessonthe same oven cycle can ruin one core and spare another, so the core test, not the oven setting, decides.
Credit
Leo Dreisilker, Dreisilker Electric Motors
2026
A tighter limit for cores
An EASA article in August 2026 says older core loss testers rated cores "good" or "marginal" at up to 10 W/lb or more, and that core loss should now not exceed 5 W/lb, with no hot spot more than 15 °C above ambient. It warns that a hot spot 12 mm across may not raise the total loss yet can still fail the new winding: in one large alternator rewound in an emergency over a known hot spot, the new winding failed in the same place within an hour.
Lessona core can pass on total watts and still fail on one hot spot, so both must be checked.
Credit
EASA
Watch for
Bare-core designs from shared records
EASA says its Motor Winding Workbench, an online application released in 2026, lets member shops search more than 250,000 winding records by core length, bore diameter, slot count and poles, and adapt a matching winding when the original design was never recorded, for example because another shop stripped the core.
Achieved whena member shop or an outside party publishes test results, such as current and temperature rise, for a set of motors rewound to Workbench bare-core designs.
Claimed by
EASA, November 2026
Open
01
Core loss before and after
A public table of core loss readings taken before and after stripping on several hundred stators from several shops, with each motor's maker, build year, rating, oven temperature and number of earlier rewinds, to test whether newer cores lose more with each burnout.
The briefCredit · unclaimed02
Robot inserts repair coils
A robot arm that learns from a few demonstrations to insert pre-wound coil groups into the slots of a small, de-energized stator it has not seen before, as a winder does by hand in a repair shop, with a filmed tally of how often it seats a coil without damaging the wire or the slot liner.
The briefCredit · unclaimed
Open 01 · the brief
Core loss before and after
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of core loss readings taken before and after stripping on several hundred stators from several shops, with each motor's maker, build year, rating, oven temperature and number of earlier rewinds, to test whether newer cores lose more with each burnout.
Why it is still open
The evidence in print covers a few motors at a time. EASA's 2019 rewind study measured stator core loss on 10 premium-efficiency motors before and after one controlled rewind and found changes from a rise of 14.3% to a fall of 4.2%; Dreisilker tested two stators (2017); Skenzick described one shop's experience (2014). EASA's accreditation checklist, as Dreisilker quotes it, requires that a core be repaired or replaced if its loss rises more than 20% between the tests before and after winding removal, so accredited shops already record both numbers, but web searches and a November 2026 GitHub search for "core loss stator" found no data pooled across shops.
Built with what exists
Core test sheets from accredited shops, scanned or exported from their core loss testers; an AI model that reads scanned forms into fields, with a person checking a sample; DuckDB or a spreadsheet; a simple regression of the change in W/lb on build year, oven temperature and number of rewinds.
First step
Ask two EASA-accredited shops for one year of before-and-after core test sheets with the motor nameplate data, and enter 30 by hand to settle the fields.
Done looks like
At least 500 stators from at least five shops; a second person checks 50 entries against the original sheets; published medians and ranges of the change in W/lb by build decade, and an answer, with its uncertainty, to whether cores built after the late 1990s lose more.
Open 02 · the brief
Robot inserts repair coils
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A robot arm that learns from a few demonstrations to insert pre-wound coil groups into the slots of a small, de-energized stator it has not seen before, as a winder does by hand in a repair shop, with a filmed tally of how often it seats a coil without damaging the wire or the slot liner.
Why it is still open
Alexander Kuehl and Joerg Franke of Friedrich-Alexander-Universität Erlangen-Nürnberg (2015) write that motor production is highly automated for large series but relies on "intensive manual operations at low volumes". Their two-armed robot cell picked coils from a stencil made for each stator and was the cheapest option only between 820 and 2,580 stators a year. A University of Padua team (Stefano Michieletto and colleagues, 2018) described an "automated and flexible coil winding robotic framework". Coil insertion machines sold today are built for factory series, and no published tally of a robot inserting coils into one-off repair stators was found.
Built with what exists
A low-cost robot arm supported by Hugging Face's LeRobot software, which records demonstrations and trains on them (see the Note "No download"); an open robot model that learns a task from a few demonstrations, such as NVIDIA's GR00T 2 (see the Note "Five demonstrations per task") or Runway's Praxis-1 (see the Note "Open robot model weights"); scrap stators of 1 to 5 hp; coils wound on a manual coil winder.
First step
Film a winder inserting coil groups into one scrap stator 50 times with two fixed cameras, and log each demonstration with LeRobot.
Done looks like
Twenty filmed attempts on each of three stators not used in training; a published tally of coils fully seated with the slot liner intact, judged by an experienced winder, and a continuity check of each coil with a multimeter afterward.