Beginners doing servo selection or designing motion mechanisms have almost all heard an experienced engineer say this: “Your inertia is mismatched; the motor can’t drive it.” A lot of people freeze on the spot: I calculated the torque, and it’s clearly big enough, so how can it not drive the load? What exactly is inertia? You can’t see it or touch it, so how can it make a motor vibrate, miss position, or even trigger an overload alarm?
After a few years of commissioning work, you’ll understand that inertia is essentially the “inertia coefficient” of a rotating object. It isn’t mysterious, but if you treat it only as a concept and don’t apply it in conversion formulas and acceleration/deceleration torque, the drawing you produce is just scrap metal on paper.
1. Core Physical Intuition: The “Mass” of Rotation
In middle-school physics, mass m is the measure of inertia in linear motion. Push it, and F = ma. In rotational motion, moment of inertia J is the “mass” of rotation. Its physical formula is:
where \alpha is angular acceleration (unit: rad/s^2 ). The torque T produced by your motor is used partly to overcome external resistance (friction, cutting force), but most of it goes into J \times \alpha.
Hard conclusion: If the load inertia J doubles, then to achieve the same start-stop response time, the motor output torque must also double. That is why the “constant-speed torque” you calculated may clearly be enough, yet the system alarms as soon as it accelerates at startup: you never calculated the acceleration torque at all.
2. The Three Standard Body Inertias You Must Be Able to Calculate by Hand in Engineering
Don’t memorize calculus. As a mechanical designer, as long as you remember three standard shapes, you can build up 90% of mechanisms:
1. Solid cylinder/disc (motor shaft, timing pulley, rotary table)
2. Hollow cylinder/ring (sleeve fitted on a shaft, large gear rim)
3. Object moving in a straight line (converted through a lead screw or belt)
This is what beginners most often miss! A worktable with mass m, driven by a lead screw with lead P (unit: meters), has the following inertia converted to the motor shaft:
For example: a 50 \, \text{kg} worktable with a lead screw of lead P = 0.02 \, \text{m} (20 mm) has a converted inertia of only:
As you can see, a large-mass linear motion can have a very small inertia after conversion through a lead screw (because the lead is squared in the denominator). Conversely, the larger the lead, the larger the converted load inertia, and the harder the motor has to work.
3. Must-Calculate for Selection: “Total Inertia” at the Motor Shaft and the “Inertia Ratio”
Many newcomers take the load inertia J_L and compare it directly with the motor catalog. That is wrong. You must convert every rotating part back to the motor shaft.
Step 1: Convert the total inertia at the motor shaft, J_{total}
If you add a reducer with reduction ratio i in the middle (the reducer’s own inertia is J_g, and the coupling inertia is J_c ), the formula is:
Note: the reduction ratio i is squared in the denominator! This is why adding a gearbox is a “magic tool for inertia matching”: it can make the load inertia “lighter” by the square of the ratio.
Step 2: Calculate the inertia ratio
Iron rule for engineering judgment (with practical corrections):
- Rigid direct connection (coupling locked directly): the inertia ratio must be strict. Recommended R \le 3, with an upper limit of 5 times. Because there is no cushioning in a direct connection, a large inertia makes the system’s resonance frequency very low, and increasing the gain will cause squealing.
- Belt/timing belt transmission: a belt has elastic damping, so the inertia ratio can be relaxed to R \le 10.
- Planetary reducer with a large reduction ratio: as long as R \le 15, and the load is constant-torque (not frequent forward/reverse operation), most Japanese and domestic servos will run extremely well.
- Special reminder: if you choose a “large-inertia” motor series from Yaskawa or Mitsubishi (whose J_M is itself 3-5 times larger than that of a small-inertia motor of the same specification), then an R of 20 or even 30 is common in the industry.
4. More Critical Than “Inertia Matching” Is the “Acceleration/Deceleration Torque Check”
When an experienced engineer asks you to calculate inertia, the final goal is to calculate the following formula. A motor selection catalog lists two torques: rated torque T_n and peak torque T_{max} (generally 3 times the rated torque).
Your acceleration torque must be less than the peak torque, otherwise the drive will immediately trigger an overload alarm:
where:
- \alpha (angular acceleration)= \frac{rated speed \times 2\pi}{60 \times acceleration time}
A reality check: Many beginners think, “Can’t I just set a longer acceleration/deceleration time?” Yes, but the production-line cycle time won’t allow it. If, because of inertia mismatch, you stretch the start-stop time from 0.1 s to 0.5 s, the machine’s capacity is basically ruined. So the essence of inertia matching is achieving stable, controllable motion with the lowest motor cost while meeting the cycle time requirement (high acceleration).
5. The Physical Essence of Servo Motor Vibration and Squealing (Advanced)
You spent a lot of money on a servo. After installation, it squeals at high frequency during operation, and the motor gets hot enough to fry an egg. You spend an entire afternoon tuning PID filters; it barely stops squealing, but it still has to shake twice before settling into position.
The physical reason behind this is: the system’s natural frequency is fighting the servo response bandwidth. The larger the inertia ratio, the lower the mechanical system’s natural frequency \omega_n:
K is the mechanical connection stiffness (coupling stiffness, belt stiffness). When the load inertia is especially large, the natural frequency drops into the servo drive’s response band, and the two resonate. If you only increase the motor size without changing the connection stiffness (for example, replacing a belt with a lead screw, or replacing a jaw coupling with a diaphragm coupling), even a much larger motor will be useless. It will still resonate.
6. The Three Most Common “Selection Traps” for Beginners
Trap 1: Selecting only by rated torque and ignoring peak torque
Pushing a block of iron at constant speed requires very little torque, but at the instant of starting and stopping, the peak torque is 5-8 times the constant-speed value. If the inertia is underestimated and the calculated acceleration torque exceeds the drive’s limit, the machine will stall as soon as it reaches high speed.
Trap 2: Forgetting to calculate the inertia of the coupling/belt pulley itself
In high-speed, low-inertia applications (such as placement machines and light-load 3C equipment), the inertia of a 50 mm diameter coupling may be larger than that of your load turntable. Remember this during design: every part on the rotating shaft must be checked against the catalog and included in the inertia calculation. Don’t calculate only the final load.
Trap 3: Blindly trusting automatic software calculations without understanding the underlying formulas
SolidWorks can indeed calculate inertia, but what it exports is the inertia about the centroid coordinate system. If you need rotation about the motor shaft (for example, an eccentric fixture), you need the parallel-axis theorem:
(where d is the eccentric distance). The extra inertia caused by eccentricity often far exceeds the object’s own rotational inertia. The software will not remind you of this trap.
7. A Few Practical Words for New Designers
- Calculate acceleration torque first, then look at the inertia ratio. The inertia ratio is a “health indicator”; acceleration torque is the “life-or-death red line.”
- A reducer is not mainly for reducing speed; it is mainly for reducing inertia. If torque is insufficient, use a larger motor. If inertia is mismatched, first see whether you can add a reducer.
- An aluminum turntable has about half the inertia of a steel turntable (due to density differences). In applications that require high response, don’t be stingy with material cost.
- During commissioning, if it is stable at low speed but shakes at high speed, it is basically resonance caused by excessive inertia. Adding a filter treats the symptom; reducing weight, shortening the axial dimension, and switching to a high-stiffness coupling treat the root cause.
In mechanical design, once you understand T=J\alpha, you understand the fundamentals of motion control. Don’t just draw from torque catalogs. Spend more time punching those squared terms into a calculator, and you’ll find that many machine problems were already doomed at the drawing stage.