Table of Contents

Full-Power Run-Up Duration

The run-up duration a Spindle Capability carries: what the number is, the rotor inertia HiNC derives from it for the spindle speed under load, how to measure, convert or look it up, and what the field takes. The torque curves it is read against are on that page.

SpindleRunUpDuration_s (API), the Full-Power Run-Up Duration on the Thermal / Energy node of Spindle Capability Setup, shown while Show Physics Options is on, is the time in seconds the spindle takes from standstill to the top speed of its shortest-duration torque curve while the drive gives that curve's torque all the way, with nothing in the spindle but the rotor: the tool and the holder in the spindle are not added to the inertia (Known limits). HiNC derives the rotor's moment of inertia from it, for the simulation of the spindle speed under load:

\[ J = \frac{t_\text{run-up}}{\displaystyle\int_0^{\omega_\text{top}} \frac{d\omega}{T(\omega)}} \]

in kg·m², where \(T(\omega)\) is the shortest-duration torque curve in N·m, linear between its points, and \(\omega\) the spindle speed in rad/s, from standstill to the curve's top speed \(\omega_\text{top}\). The inertia grows with the duration: the longer the spindle takes to run up, the heavier its rotor, and the more slowly the spindle loses speed under the same overload. How the simulation uses the inertia, and where its reading stops, is Reading the Load Leaves and Known Limits of the Load Leaves; this section is about the number.

  • Measuring it. Run the spindle up from standstill to that top speed with no tool in it, and time it. The time is this duration only where the drive gave the torque of the shortest-duration curve all the way, and two things make a timed run-up another one:
    • A drive commonly accelerates with more than the shortest-duration rating. The source of the shipped Mitsubishi-SJ-VLS15-11FZT-15000rpm.default.SpindleCapability, for one, prints an acceleration and deceleration output of 18.5 kW beside the 11 kW of the 30-minute rating the capability's shortest curve holds (its Note says so). A run-up at such an output is shorter than this duration; convert it with the two torque curves, the shortest-duration one \(T_\text{curve}\) and the one the drive accelerates with \(T_\text{accel}\), each integrated from standstill to the top speed: \(t_\text{run-up} = t_\text{timed} \int d\omega / T_\text{curve} \div \int d\omega / T_\text{accel}\). Where the two hold constant torque up to the same base speed and constant power above it, that is about \(t_\text{timed} \times P_\text{accel} / P_\text{curve}\), 18.5 ÷ 11 ≈ 1.7 for those two ratings. Entered unconverted, the time makes the rotor lighter than it is, and the simulated spindle loses speed faster than the machine's.
    • The controller's acceleration settings can stretch the run-up. The time is then longer than this duration; entered here, it makes the rotor heavier than it is, the simulated spindle loses speed more slowly than the machine's, and the deviation reads too small.
  • Telling a stretched time. Where the spindle's own data sheet prints the rotor's moment of inertia \(J\) (a motor's is not it, as under Looking it up below), \(J \int_0^{\omega_\text{top}} d\omega / T(\omega)\) over the shortest-duration curve is the time the spindle needs at that curve's torque. A timed run-up clearly longer than that was stretched by the controller. Where you cannot tell, take the data sheet's figure; where there is none either, leave the field empty (1 s).
  • Looking it up. A spindle's or spindle motor's data sheet may print a run-up or acceleration time, or the rotor's moment of inertia \(J\); a controller's spindle acceleration time parameter is the setting that paces the spindle, not this duration. A time printed for an acceleration output above the shortest-duration curve converts as above. A \(J\) gives the duration through the same integral, \(t_\text{run-up} = J \int_0^{\omega_\text{top}} d\omega / T(\omega)\); for a curve of constant torque \(T_0\) up to the base speed \(\omega_b\) and constant power \(P\) (in W) above it, that is about \(J\,\bigl(\omega_b / T_0 + (\omega_\text{top}^2 - \omega_b^2) / (2P)\bigr)\). Only the spindle's own data sheet, a motor spindle's or one a spindle maker prints, gives a time or a \(J\) that is this duration's; a motor's figures are not it, and for such a spindle time the machine's own run-up as above:
    • A direct-drive spindle. The motor's \(J\) leaves out the spindle shaft, the drawbar clamping and the coupling. The ratio is 1, so nothing is multiplied, but they have to be added.
    • A built-in spindle whose \(J\) is a motor kit's rotor leaves out the spindle shaft in the same way.
    • A spindle driven through a belt or a gear. Each turning part counts at the spindle multiplied by the square of its own speed over the spindle's: the motor's \(J\), with the pulley or gear on the motor, by the square of the ratio, motor speed over spindle speed, each range's own ratio on a gear-shift spindle, which HiNC simulates with one inertia all the same (Known limits); the spindle shaft, the pulley or gear on it and the drawbar clamping, which turn at the spindle's speed, are added as they are.
  • Left empty, the field reads 1 s (DefaultSpindleRunUpDuration_s (API)), an assumption about no particular spindle; the shipped capabilities give none. Which way that errs is under Known limits. The capability's XML then holds an empty <SpindleRunUpDuration_s /> element, and a capability saved by an earlier version holds none; both read 1 s. A NaN in the capability shows as an empty field as well but is not read as 1 s: it leaves no rotor inertia (Spindle-Inertia--Unreadable); type a duration into the field to replace it.
  • The field takes 0 and anything above, and the capability keeps any value it is given, in its XML, by a script or by a REST client as well. The simulation is checked from 0.01 s up: a shorter duration, a lighter rotor, can lose accuracy where the speed settles on the torque curve: the integration by angle can end off the curve, and the step's power budget can be missed, sooner with few load samples a revolution; and a step's speed under load can fail to compute (Spindle-SpeedUnderLoad--Faulted); one that is not a finite number above zero, 0 included, leaves no rotor inertia (Spindle-Inertia--Unreadable), both under Known limits.
  • The steps already played keep the speed they were played with until they are recomputed, as under Spindle meter power and energy. The exception is the endpoint ratio of a step that cuts and has no stored speed, one played without a rotor inertia among them: it is read against the capability as it is now, infinity or the step's end point where that capability gives the spindle limits but no inertia, NaN where it gives an inertia. So a duration corrected before the recompute turns such a step's infinity into NaN at once (Recomputing a Finished Strip).

See Also

  • Spindle Capability — the torque and power curves the duration is read against, and the ratios and the thermal model they drive