In many of the projects I work on, the engineer is trying to fit a motor into a space that seems too small for the amount of torque, speed, and efficiency they need, almost to the point where it defies the laws of physics.
Advantages of Frameless Motors for Robotics Designers
Expert Interview with Brent Ford, Sales Director | Delta Line North America, Inc.
Tell us about yourself and your role with Delta Line.
Delta Line has been designing and manufacturing motors, gearboxes, drives, and integrated motor solutions for OEM customers for over 40 years. I’m the Sales Director for North America and have been in this role for more than five years. I’ve been in the motion industry for over 20 years and greatly enjoy working directly with customers on their projects across many different high-tech industries throughout North America.
The advancements in robotics over the last several years have been incredibly exciting, especially because robotics now encompasses such a broad range of applications. I’ve had the opportunity to work directly with customers developing everything from relatively traditional automation systems to much more advanced robotic applications, and it’s been interesting to see how quickly the requirements for motors and motion systems continue to evolve.
What advantages do frameless motors offer robotics designers, and what integration responsibilities come with them?
The main advantage of frameless motors is the ability to design a motor into the robotic system in a way that fits the application as well as possible, without many of the tradeoffs that come with a standard housed motor. Most engineers looking to use a frameless motor place a very high value on physical space, a high level of system integration, and strong motor performance.
In many of the projects I work on, the engineer is trying to fit a motor into a space that seems too small for the amount of torque, speed, and efficiency they need, almost to the point where it defies the laws of physics. These motors can be as small as roughly 12 mm in diameter or as large as 250 mm or more, but the challenge is often very similar: get the maximum possible performance from the available package.
The designer is usually also trying to optimize the overall system, not just the motor. That includes power utilization, thermal performance, and how the motor integrates with other components such as gearboxes, bearings, and feedback devices.
The tradeoff is that with a frameless motor, the customer takes on much more responsibility for the final motor integration. The housing, bearings, rotor support, air gap, thermal path, feedback device, electrical connections, and assembly process all become part of the customer’s design. A frameless motor gives the engineer much more freedom, but it also means those surrounding components have a direct impact on how well the motor ultimately performs.
Why are air gap, concentricity and alignment so important when integrating a frameless motor?
With a housed motor, the motor manufacturer takes care of the air gap, concentricity, and alignment through its design and manufacturing processes. The customer primarily needs to select a motor based on its performance characteristics and external physical dimensions.
With a frameless motor, however, the customer is essentially receiving the “guts” of the motor, and it becomes their responsibility to maintain a stable physical relationship between the stator and rotor within their own system. They need to understand how the stator and rotor must be aligned axially, radially, and angularly, and how their own bearings, shafts, housings, and manufacturing tolerances contribute to that alignment.
Maintaining that relationship is critical because variations in the air gap or alignment can affect important motor characteristics such as torque capability, cogging torque, torque ripple, efficiency, vibration, noise, and thermal performance. In more severe cases, poor alignment can even create the risk of physical contact between the rotor and stator.
For that reason, frameless motor manufacturers should clearly provide the key dimensions and tolerances required for the motor to achieve the performance stated on its specification sheet. The customer then has to consider those requirements as part of the complete tolerance stack of their own assembly.
Customers also need to consider the capabilities of their manufacturing processes to make sure they can maintain the same level of consistency and quality at the volumes they expect to build. It may be feasible to successfully integrate a frameless motor when an engineer is closely involved and handholding every part of the assembly process, but can the same result be achieved when a production associate is building hundreds or thousands of units per month? That is something that needs to be considered early in the design.

What options do engineers have for mounting the stator, and how should they choose between them?
Engineers have several different options for mounting a frameless motor stator, and the best method depends heavily on the rest of the system and the stage of development. In my experience, the three main things to consider are the thermal requirements of the system, securely retaining the stator against the motor’s torque, and manufacturability in production.
Some mounting methods should be considered permanent, while others can allow the stator to be removed. Because of that, engineers may choose one method during prototyping and a different method when designing the final production system.
The first main method is to retain the stator using end caps. A frameless motor stator should have tightly controlled dimensions on the lamination stack OD and end faces so that matching precision-machined end caps can locate and clamp the stator from both sides. These end caps may also support bearings and other system components, and in some designs may incorporate the feedback device. This approach can effectively create a motor module that is then integrated into the larger system. Depending on the design, it can also allow the stator to be removed later for service or modification. One potential disadvantage is that the stator may have less direct surface contact with the surrounding structure, which can reduce heat transfer compared with other mounting methods.
The second main method is a shrink or interference fit. This is typically used when the customer can heat the surrounding housing enough to allow the stator to be installed and then retained as the housing cools. When the tolerances are designed correctly, this can provide both a very secure mechanical fit and a large contact area for transferring heat from the stator into the housing. This would normally be considered a permanent installation. It is especially important to understand the materials used for both the stator and the housing, because their coefficients of thermal expansion can affect the fit over the operating temperature range. If this is not considered correctly, the fit may become too loose or excessive stresses may be introduced into the stator or housing.
The third main method is adhesive bonding. This is generally one of the simpler methods to implement, but it can provide excellent results when the interface is designed correctly. The stator is installed into the housing with a controlled clearance or bond gap, and adhesive is introduced between the stator and housing. The housing or stator can also include grooves, reservoirs, or other features to help distribute the adhesive consistently. The adhesive has to do more than simply hold the stator in place, it also becomes part of the thermal path between the stator and the surrounding structure, so selecting the correct adhesive and controlling the bond line are important. This is normally considered a permanent installation and can work especially well with complex housings or production processes where a controlled bonding operation is easier to implement than a shrink fit.
Ultimately, the correct mounting method is not just a mechanical decision. It should be selected based on the complete system, including torque, temperature, housing material, serviceability, assembly process, and expected production volume.
How do housing design, material selection and thermal management affect motor performance?
With a frameless motor, you primarily want to make sure the stator is securely integrated into the system and that you can remove the heat generated by the motor as efficiently as possible. The customer’s housing effectively becomes part of the motor’s thermal system, so its design can have a significant impact on how much continuous performance the motor can deliver.
We’ve seen customers with very effective liquid-cooled housings around the stator operate a motor at high current levels for extended periods that would not be possible with passive cooling alone. On most frameless motor datasheets, the manufacturer will state the size and material of the heatsink or mounting structure used when establishing the motor’s continuous torque and current ratings. If the customer’s system has a different housing size, material, contact area, or cooling method, it may remove heat either better or worse than the test setup, and the motor’s usable continuous performance can change accordingly.
The engineer also needs to think carefully about how the motor will actually be used (i.e. their motion profile). Will it produce high torque for only a few seconds followed by a long dwell period? Will it operate continuously? How much heat will build up during each cycle, and how much time is available to remove that heat? There may also be other motors, drives, gearboxes, or components adding heat into the same structure.
This is something I see customers sometimes underestimate early in a project. The initial focus is often on fitting the required torque and speed into a very small package, but as the system develops, thermal management can become just as important as the motor’s electromagnetic performance. Ultimately, the motor, housing, cooling method, duty cycle, and surrounding components all need to be considered as one thermal system.

How should bearings, feedback devices and electrical connections be considered as part of the actuator design?
Having the ability to optimize bearings, feedback devices, and electrical connections is a major benefit of using a frameless motor. It allows the engineer to avoid many of the constraints that come with a standard off-the-shelf housed motor and instead choose where these components should be located and which components make the most sense for the application.
For example, an actuator may benefit from having different feedback devices located on the motor side and the output side of the system, depending on what the designer is trying to measure or control. In other cases, redundancy or functional safety requirements may drive the use of multiple unique feedback devices. Bearings can also be selected and positioned around the actual loads, stiffness requirements, and mechanical architecture of the actuator rather than simply accepting the bearing arrangement built into a standard motor.
The electrical interface can be optimized in the same way. Connectors, phase pins, lead wires, or solder pads can be incorporated directly into the frameless motor PCB so that the motor interfaces more efficiently with the customer’s controller board or wiring architecture. On compact robotic systems, these connection methods can have a significant impact on packaging, assembly, and manufacturability.
Customers have very different requirements for redundancy, safety, serviceability, packaging, and production. The freedom of design that comes with a frameless motor can be a major advantage in all of these areas, but these components should ideally be considered early in the actuator design rather than added after the motor and mechanical structure have already been finalized.
What should designers consider when attaching the rotor and transmitting torque into the machine?
Much like mounting the stator into the customer’s system, there are a few primary methods for attaching a frameless motor rotor. These generally include mechanical retention with an end cap or screw, an interference or press fit, and adhesive bonding. Regardless of the method, the rotor needs to remain concentric, securely transmit the required torque, and maintain its position throughout the operating speed and temperature range.
One of the major advantages of a frameless motor, though, is that the rotor does not have to be mounted to a traditional motor shaft. It can instead be integrated directly onto fairly complex components within the customer’s machine. We’ve seen customers mount frameless rotors directly onto linear screws, pulley systems, intricately machined shafts, pump rotor assemblies, and the input structures of high-ratio gearheads.
This gives the engineer much more freedom to think about the best way to transmit torque into the application. Instead of mounting a housed motor and then using a traditional coupling to connect its shaft to the driven component, the frameless rotor can often be incorporated directly into that component or assembly.
That can eliminate unnecessary interfaces and potentially reduce overall package length, part count, inertia, compliance, and assembly complexity.
How can a frameless motor be customized to better match a robotic application?
Some frameless motor manufacturers, including Delta Line, give engineers quite a few levers to pull when it comes to customizing the motor around the application. There are generally easier modifications and more complex ones. The more complex customizations can involve meaningful NRE costs, longer development and manufacturing lead times, and potentially minimum production-volume requirements.
One of the relatively straightforward customizations is designing a stator winding around the torque, speed, voltage, current, and efficiency requirements of the customer’s application. This is primarily accomplished by changing parameters such as the copper wire size and number of turns around the stator teeth. Rather than simply accepting the electrical characteristics of the closest standard winding, the goal is to better match the motor to the customer’s available bus voltage, current capability, and required operating points.
Another relatively straightforward customization is the wiring and electrical connection of the motor. This may be as simple as specifying a particular wire length or adding a customer-selected connector to the phase or Hall-sensor wires. In other cases, it becomes more involved, such as incorporating connectors, phase pins, or soldering pads directly onto the motor PCB. Hall sensor type and position, as well as thermistor type and temperature-sensing configuration, can also be customized to better match the customer’s control and safety requirements. The objective is usually to make the electrical integration and production assembly of the frameless motor as simple as possible within the customer’s system.
A more involved customization is changing the active stack length of the stator and rotor. This effectively changes the axial length of the motor. With the same basic electromagnetic design, torque capability will generally increase approximately with active stack length, although other motor characteristics also change and need to be evaluated. An engineer may request a custom stack length because the standard motor does not use all of the available axial space, or because only a modest increase in torque is needed to meet the application requirement. This type of change will normally require the motor manufacturer to develop new winding, assembly, and potting fixtures.
Customers may also request physical modifications to the rotor or stator to improve integration with their system. Examples include adding chamfers, mounting features, slots, threaded inserts, or flanges, or modifying certain rotor dimensions such as the back-iron geometry. The difficulty and cost of these changes vary considerably depending on the motor construction and exactly what needs to be modified. In many cases, though, a relatively small mechanical change to the motor can eliminate an adapter, additional component, or difficult assembly operation in the customer’s system.
There are also opportunities to modify the magnetic design. Depending on the application, that can include magnet material or grade, magnet geometry, and potentially the air gap. For example, magnet shape can be optimized to influence characteristics such as cogging torque and torque ripple, which can be particularly important in robotic joints where smooth low-speed motion and precise control are critical.
A more complex, but potentially very beneficial, customization is changing the outside diameter of the stator lamination. This starts to become much closer to developing a new motor rather than modifying an existing one, and not every frameless motor manufacturer is set up to support it economically.
For example, a customer may determine that a 102 mm OD motor does not quite provide the performance they need, while a 115 mm motor is physically too large for the available space. They may instead ask us to develop something around 110 mm OD with performance specifically optimized for their application.
The electromagnetic design itself can often be developed relatively quickly, but manufacturing the new motor is the more significant undertaking. During development and lower quantities, the custom laminations may be laser cut. If the program progresses to higher production volumes, dedicated stamping tooling can then be developed so the laminations can be manufactured efficiently on production stamping equipment. Additional tooling may also be required for winding, assembly, fixturing, and potting.
There can be meaningful NRE cost and development time associated with that level of customization, so it generally makes the most sense for an application with sufficient long-term production volume. At higher volumes, however, the resulting unit cost can potentially approach that of a standard frameless motor while giving the customer a motor that is much better optimized around their mechanical envelope and performance requirements.
Ultimately, the goal of customization is not to make the motor different for the sake of being different, it is to remove compromises from the customer’s system and make the motor fit the application as closely as possible.
What integration decisions should robotics engineers address early to avoid problems later?
Some of the biggest integration decisions robotics engineers should address early when using a frameless motor are thermal management, stator and rotor mounting for both prototype and production builds, feedback-device requirements, electrical connections, the customer’s own assembly capabilities and alignment tolerances, and the overall power and efficiency goals of the system.
It is also important to have realistic goals for cost, development timeline, expected production volume, and the level of customization that makes sense for the program. In my experience, the best results usually come when the motor, mechanical system, electronics, and manufacturing approach are all considered together from the beginning.
Visit the Delta Line Frameless Motor Installation & Operation Manual
Brent Ford is Sales Director for Delta Line North America and has held this role for more than five years. With over 20 years of experience in the motion industry, he works directly with OEM customers across North America to develop motor and motion solutions for robotics, automation and other high-tech applications. His work includes supporting customers with motor selection, customization, mechanical and electrical integration, testing, and the transition from prototype development to volume production.
The content & opinions in this article are the author’s and do not necessarily represent the views of RoboticsTomorrow
Featured Product
