How to Read a Frameless Servo Motor Torque-Speed Curve Before You Select a Motor

A torque-speed curve is one of the most important documents in a frameless servo motor selection process. It is also one of the easiest documents to misread.

Many projects begin with a simple requirement such as "we need 2 N·m" or "we need 3 N·m peak." That is not enough for a direct-drive system. A motor that can produce the required torque at stall may not sustain it at speed. A motor that looks strong in a peak table may overheat in continuous operation. A motor that works at 48 V may not have enough speed margin at 24 V.

We see this misunderstanding in roughly one out of every three first-time RFQs we receive. A customer sends a one-line requirement—"3 N·m, 48V"—and expects a single model recommendation. Without knowing the speed, duty cycle, thermal environment, and drive current limit, we cannot give a responsible answer. This guide explains exactly why, and how to read a torque-speed curve before selecting a frameless servo motor kit or torque motor.

Start with the operating point, not the maximum number

The first question is not "what is the highest torque?" The first question is:

A useful motor request should include:

1. Continuous torque at speed.
2. Peak torque at speed.
3. Peak duration.
4. Duty cycle.
5. Bus voltage.
6. Ambient temperature or housing thermal condition.
7. Current limit of the drive.

Without these values, the curve can be interpreted too optimistically.

A quick selection workflow

Before looking at a supplier curve, write down the motor duty in a simple table.

Then compare the curve against the table. If the required point falls inside the peak zone only, ask whether the peak duration and recovery time are acceptable. If the required point falls outside the continuous zone but is used repeatedly, request an RMS torque and thermal review.

The visual reading sequence is simple: locate the required operating point, check whether it sits in the continuous or peak region, then confirm whether the boundary is driven by heat, current, or voltage. If the point is close to a boundary, treat the selection as provisional until housing thermal behavior and drive current are confirmed.

Continuous torque vs peak torque

Continuous torque is the torque the motor can sustain without exceeding the defined thermal limit under the stated test condition. For frameless motors, this condition is especially important because the motor does not have its own housing and fan. The customer's housing is usually the heat sink.

Peak torque is short-duration torque. It may be available for acceleration, deceleration, emergency response, or transient load changes. But peak torque is not a continuous operating point.

If a supplier only provides peak torque without continuous torque and test context, the buyer does not yet have enough information for engineering selection.

RMS torque: the bridge between motion profile and heat

For repeated motion, RMS torque is often more useful than peak torque alone. It estimates the heating effect of a changing torque profile.

A simplified RMS torque calculation is:

T_rms = sqrt((T1^2 x t1 + T2^2 x t2 + ... + Tn^2 x tn) / total cycle time)

Example duty cycle:

The simplified RMS torque is:

T_rms = sqrt((3.0^2 x 0.5 + 1.2^2 x 2.0 + 0.8^2 x 5.0 + 2.5^2 x 0.5 + 0.1^2 x 2.0) / 10)
T_rms = sqrt((4.50 + 2.88 + 3.20 + 3.13 + 0.02) / 10)
T_rms = 1.17 N·m

If the curve shows a continuous torque capability above 1.17 N·m under a similar thermal condition, the motor may be a reasonable starting point. If the continuous limit is below this value, the motor may still work only with stronger heat sinking, lower duty, lower ambient temperature, or a larger motor.

This is still a simplified screen. Final validation should include real housing thermal behavior and drive settings.

The following visualization shows the duty-cycle profile from the example above. The colored bars represent real torque demand over one cycle, and the dashed red line is the resulting RMS torque:

Notice that even though the peak torque is 3.0 N\u00b7m (well into the peak zone), the RMS result of 1.17 N\u00b7m sits comfortably below the motor's continuous rating of 1.5 N\u00b7m. This means the motor will likely survive thermally\u2014but only if the housing provides the same thermal path assumed by the supplier's test. If the housing is sealed and compact, you should still apply a 20-30% thermal derating to that continuous line.

Why bus voltage changes the curve

Frameless servo motors are often used with low-voltage DC bus systems such as 24 V or 48 V in robotics, medical devices, and compact automation. A higher bus voltage generally gives more speed margin before voltage limitation becomes dominant. A lower bus voltage may be acceptable for low-speed torque but can run out of speed headroom earlier.

This is why a curve for 48 V should not be blindly applied to a 24 V system.

When comparing curves, check:

1. No-load speed at the stated bus voltage.
2. Torque available at your target speed.
3. Where the curve begins to roll off.
4. Whether the current limit or voltage limit is the real boundary.
5. Whether the winding option matches your controller.

Kt and current: can the drive deliver the requested torque?

Torque constant, usually written as Kt, links current to torque.

Required phase current = Required torque / Kt

Example:

In this example, the mechanical torque target may be reasonable, but the selected drive cannot deliver enough current margin. The buyer should either choose a different winding, a stronger drive, a lower peak torque requirement, or a larger motor package.

The same check should be done for continuous current. A motor can have enough peak current capacity while still exceeding the drive or thermal continuous current limit.

Thermal assumptions matter more for frameless motors

A housed servo motor may include a known thermal package. A frameless motor depends heavily on the customer's structure. The same stator and rotor can perform differently depending on housing material, contact area, potting method, airflow, duty cycle, and ambient temperature.

For this reason, the torque-speed curve should state the test condition. Examples include:

1. Ambient temperature.
2. Natural convection or fixed heat sink.
3. Winding option.
4. Housing contact assumption.
5. Continuous temperature limit.
6. Peak duration.

If your mechanism has a weak thermal path, derate the curve before committing to the motor.

Simple thermal derating example

Assume a supplier curve shows:

If the buyer's housing has less contact area, higher ambient temperature, or no airflow, using 1.5 N·m as the production continuous point is risky.

A conservative early screen might derate by 20-35% until better thermal evidence exists:

This is not a replacement for test data. It is a buyer-side risk control before locking a model. If the application needs the full 1.5 N·m continuously, request a thermal test or a housing-specific validation plan.

Duty cycle: the hidden selection variable

Two applications can have the same peak torque and completely different motor requirements.

Example:

This is why the buyer should share motion profile, not only peak torque.

Common mistakes when reading curves

Mistake 1: selecting by stall torque alone

Stall torque may be relevant, but most applications move. The motor must deliver torque at the required speed. We had a medical device customer select a 90mm OD motor based on its 4.0 N\u00b7m stall torque. Their operating speed was 1200 RPM at 24V. At that speed, the voltage roll-off had already cut available torque to 1.8 N\u00b7m. The motor was physically too large for the application it could actually serve.

Mistake 2: ignoring housing heat transfer

If the stator is poorly coupled to the housing, continuous torque will fall\u2014sometimes dramatically. A compact sealed robot joint with set-screw mounting can lose 40-50% of the catalog continuous torque due to the air gap acting as a thermal insulator. We have measured housing surface temperatures of 85\u00b0C on joints where the catalog predicted 55\u00b0C, simply because the stator was not press-fitted.

Mistake 3: treating peak torque as a production duty point

Peak torque is temporary. A customer designing a 6-axis cobot once used the 2-second peak torque value for the shoulder joint's holding torque at full extension. The motor overheated within 90 seconds because the "hold" phase was continuous, not transient. Always ask for allowed peak duration and recovery conditions.

Mistake 4: comparing curves from different voltage classes

A 48V curve and a 24V curve from the same motor look dramatically different above 500 RPM. At 24V, the no-load speed may drop by 40-50% compared to 48V. If your target operating speed is in the upper half of the 48V curve, switching to 24V may make that speed physically unreachable regardless of torque.

Mistake 5: forgetting drive current limits

Even if the motor can theoretically produce the torque, the drive must deliver the current. A common trap: a motor with Kt = 0.15 N\u00b7m/A needs 20A to produce 3.0 N\u00b7m peak. If your servo drive is rated at 15A peak, you will never reach that torque. The curve shows motor capability; you must separately verify controller capability.

Buyer review table for a submitted curve

When a supplier sends a torque-speed curve, review it against these questions.

Decision matrix: when to accept, derate, or redesign

Use this decision matrix before moving from curve review to sample purchase.

For procurement, this table is also useful because it turns an engineering curve into a buying decision. It makes clear whether the supplier needs to provide a better curve, a larger model, a different winding, or a thermal validation plan.

What to send when requesting a torque-speed curve

To get a useful curve, send a compact test alignment note:

1. Target bus voltage.
2. Controller current limit.
3. Continuous torque-speed point.
4. Peak torque-speed point and peak duration.
5. Duty cycle or motion sequence.
6. Ambient temperature and housing condition.
7. Cooling method or sealed enclosure constraints.

For early evaluation, a baseline curve can help narrow options. For final selection, request a project-matched curve or validation plan.

Where Frameless Servo fits in the process

Frameless Servo provides baseline torque-speed resources for first-round sizing and can align project-specific curve requests during RFQ. The goal is to prevent a common failure mode: a motor looks correct in a table, but misses the real duty point after assembly.

Useful next steps:

1. Review the Datasheet Library for baseline technical assets.
2. Compare product families in Products.
3. Send target torque-speed, voltage, and duty-cycle information through Contact / RFQ.