Dynamic braking for hoists

Follow these directions to a fail-safe system.

By Thomas Barkand and William Helfrich, U.S. Dept. of Labor, Mine Safety and Health Administration

Hoists and elevators can injure or kill. Accidents can occur on counterweighted hoisting systems if the mechanical brake fails while the cage is empty. The counterweight falls; the cage overspeeds and crashes. Direct-current hoist motors prevent this type of accident if equipped with passive electrical braking systems known as dynamic braking. Installing a dynamic brake requires minimal modifications to the control system and modest expense.

Hoists and elevators have safety features to prevent the cage from falling. Safety catches activate if the brakes or wire ropes fail. Safety catches, however, are not normally installed on the counterweight.
Many hoisting systems rely solely on the mechanical brakes to stop the cage in an emergency. Under normal operation, the electrical drive equipment controls the speed of the hoisting system while the mechanical brakes only hold the cage at a stopped position. The frequency with which the mechanical brakes are exercised is minimal when compared to the constant use of the drive equipment. However, in an emergency, the majority of hoists in the United States rely on the mechanical brakes to stop the hoist. The assumption is that the mechanical brakes are 100 percent reliable when the electrical drive becomes inoperative.

History proves that this is not a good assumption. Accidents occurred when the emergency stop button was pushed--an action that defeated the retarding effort of the hoist motor--when the mechanical brakes were inoperative. This allowed the overhauling load to free-fall, with the final speed limited only by inertia and frictional forces. The high-speed crashes at the travel limit cause extensive mechanical damage and fatal injuries.

The direct-current motors on elevators and hoists can prevent the failure because the electrical drive and control system can limit the speed of the falling overhauling load. This electrical source of braking retards the free-fall speed when the mechanical brakes fail.

Dynamic braking exploits the ability of the direct-current drive motor to act as a generator. The motor requires torque and kinetic energy of the falling load to generates electricity that is dissipated as heat in a resistor. The retarding torque limits the speed of the falling overhauling load. The amount of retardation and the final speed of the cage depend on the motor terminal characteristics and the resistance value of the dynamic braking resistor.

Hoist motor performance
The direct current motor circuit has the motor armature in series with the power supply. This power supply is either a generator or SCR bridge that converts line voltage to a variable direct current voltage that controls the speed of the motor. The field of the motor is normally supplied from a separate source--either fixed (constant potential) or variable (field weakening)--that controls the speed of the hoist motor.
A shunt-wound direct current motor can operate as either a motor or generator. It operates as a motor when it produces torque in the same direction as shaft rotation. The motor operates as a generator when the direction of motor torque opposes shaft rotation, as when a load overhauls motor torque, thus reversing shaft rotation. Then, the direct current generator--a.k.a. hoisting motor--converts the energy of the overhauling load into electrical power to be returned to the grid. This regenerative braking actively pushes power into the ac power system instead of dissipating it as heat.

The motor is said to be in the powering mode when motor action is taking place. In the inverting mode, generator action is taking place. If raising the load produces positive shaft rotation, then the four quadrants of operation are defined. The motor torque and direction are directly proportional to the armature current and voltage, respectively.

The constant-speed unbalanced hoisting system operates in quadrants 1 and 4. Quadrant 1 represents motor speed and torque acting in the same direction. Thus, the motor supplies a positive motive force to the load. Quadrant 4 represents the negative direction and positive torque of a motor that is developing a braking force. During deceleration and acceleration, the hoisting system operates in quadrants 2 and 3, respectively.

Constant-speed counterweighted (balanced) hoisting systems operate in four quadrants. When the counterweight is heavier than the load, the hoisting system operates primarily in quadrants 2 and 3. The motor acts as a generator for part of the hoist cycle and as a motor for another other part, depending on the load.

The ability of the motor to return power to the grid is what allows the motor to provide a braking force. Under normal operation, the motor control circuit provides both motive power and braking power. The mechanical brakes are called upon only to provide a very low-speed stop at the top or bottom or, in case of an emergency, to provide complete stopping. The mechanical brakes are called upon very infrequently to completely stop the hoist. However, when they are called upon, they must provide 100 percent of the stopping force. This places a severe burden on the mechanical brakes at a critical time.
Normally, motors do not experience debilitating failures. However, failures to the motor control circuits and the related power supply do occur. The motor will not develop torque to continue motoring or regenerating if the field supply fails, if the power supply for the motor fails or if utility power is lost. This warrants a system that uses the retarding capabilities of the hoist motor.

There are variations of the dynamic-braking control scheme. However, the general philosophy is to allow the motor to dissipate mechanical energy as heat in a resistor. The fundamental idea is to connect a low value (about one ohm) resistor across the motor armature when armature power fails. The motor field can either be excited separately or incorporate a field loss circuit, which connects the field across the motor armature and is, therefore, self-excited. The latter method is favored since it provides protection during complete power loss. The control philosophy is divided into three categories:

  • dynamic braking separately excited,
  • dynamic braking self-excited
  • dynamic braking separately or self-excited.

Separately excited dynamic braking
This is the simplest form of dynamic braking. It assumes the field power supply is unaffected and it provides dynamic braking only when the motor armature supply fails. It requires a separate, normally closed contactor for the dynamic braking resistor. A double-acting contactor provides a normally open contact for the motor loop and a normally closed contact for the dynamic braking resistor. The normally closed contact connects the braking resistor in parallel with the motor armature. The system configured in this method provides braking any time the motor armature power supply fails or loop contactor opens. Separately excited dynamic braking will not function if the motor field supply fails.

Self-excited dynamic braking
Self-excited dynamic braking requires additional contacts that allow the field to be energized from the dynamic braking resistor. Every time the loop contactor opens, it automatically shunts the field across the braking resistor. The hoisting motor supplies the field current, generating into the resistor in parallel with the field. As the hoisting motor generates higher voltages, the field excitation increases and, as a result, the motor armature voltage increases. This method is preferred since the self-excited motor requires no external power to provide dynamic braking, as happens when there is a total power loss and the mechanical brakes fail.

Separately or self-excited dynamic braking
This form of dynamic braking is the most complicated and expensive. However, it provides the most desirable braking by combining the best control characteristics of the first two methods. It requires a device to sense when there is no field current to the motor--usually a field-loss relay held closed by field current that drops out when field current is lost.

By using the field supply when it is available, the hoist is held at a relatively constant speed over the complete hoisting cycle. The hoist does not accelerate into overspeed before dynamic braking develops.

Characteristics of direct current motors and generators
A review of direct current machine characteristics provides a basis for analyzing the performance of the braking circuit configurations. The following equations The summarize voltage and torque relationships for direct current motors and generators:

eg = K * if * W   (Eqn 1)
T = K * if * ia   (Eqn 2)

where
eg = generated voltage (volts)
if = field current (amps)
W = angular velocity (radians per second)
T = torque (newton-meters)
ia = armature current (amps)
K = a proportionality constant

These equations assume the magnetic system of the machine is linear by neglecting magnetic saturation, hysteresis, and magnetic retention in the field poles. The magnetization curve illustrates the effect of saturation neglecting hysteresis and magnetic retention.

The equations are valid where the slope of the line linear. The graph shows that the proportionality constant is a measure of both the torque per ampere of armature current and the generated voltage per ampere of field current.

The constant K can be determined from the manufacturer's torque/armature current curves for motors or generated voltage/current curves (at rated speed) for generators. If these curves are not available, approximate the constant using nameplate information and measured data. Nameplate information for motors typically includes rated horsepower, armature current, armature terminal voltage, speed, field current and field resistance. The nameplate data for generators is similar, except the power rating is given in kilowatts.

Equation 3 shows the relationship between generated voltage and terminal voltage. The minus signs define the voltage drop for motors and the plus signs define the voltage drop for generators:

eg = Vg ± ia * Ra ± 2  (Eqn 3)
where
Vg = terminal voltage (volts)
Ra = armature resistance (ohms)
2 = voltage drop across the brushes.

The armature resistance is usually not shown on the nameplate, but it is needed for an accurate K value. The armature resistance can be obtained from the manufacturer or from measurement.

The armature resistance varies nonlinearly from 0.15 to 0.08 ohm for 50- to 200-hp motors. This low-resistance can be accurately determined with a Kelvin bridge. Since this instrument is not usually available, find the resistance by measuring the armature voltage and current during a locked-rotor stall-current test and applying Ohm's law.

Calculate the proportionality constant after finding the armature resistance by substituting Equation 3 into Equation 1 and solving for K:

K = (Vr ± iar * Ra ± 2) / (ifr * Wr) (Eqn 4)
Where:
Vr = rated armature voltage (volts)
iar = rated armature current (amps)
ifr = rated field current (amps)
Wr = rated angular velocity (radians per second)

Error is introduced if the rated field current is in the saturated region of the magnetization curve. However, this error should be small since the field excitation currents are typically not rated at a high degree of saturation. Once K is known, use the motor characteristics to analyze the performance of the hoisting system under dynamic braking conditions.

Dynamic braking performance
When dynamic braking is activated, the hoist motor operates as a generator. Generators are classified by the means used to provide excitation for the field windings--separately excited or self-excited. The dynamic braking system logic may require both methods of excitation to be used for different situations.
Power generation characteristics of generators are typically analyzed at rated speed. This greatly simplifies load design calculations. However, designing dynamic braking performance requires analysis of power generation characteristics from 125 to 25 percent of rated speed. This adds an additional variable in the system design calculations.

A dynamic braking system should lower the maximum overhauling load safely. However, the minimum load should not be decelerated faster than 16 ft/sec/sec (0.5 g) when the mechanical and dynamic brakes are activated simultaneously. This compound braking occurs if the dynamic braking logic always backs up the mechanical brake. Therefore, if the mechanical brake is operative during an emergency stop, the sum of the dynamic and mechanical braking forces produce the braking deceleration.

Size the power rating of the dynamic braking resistor to dissipate the power generated when lowering the maximum overhauling load through the full length of its travel under steady-state conditions. In addition, the dynamic braking resistor must dissipate the energy stored in the inertia of the hoisting system and the transient electrical power generated during deceleration and switching conditions.
There are two ways to size the dynamic braking resistor with respect to the speed, load, and power requirements for steady-state conditions based on the method of field excitation. The transient response will be addressed below.

Separately excited dynamic braking
This braking system is one in which the source of the field current is external to the machine. The following equations describe the armature terminal characteristics in the linear range of the magnetization curve where, for steady-state conditions, field current and armature current are constant:

Va = K * ifs - Ra * ia (Eqn 5)
ia = (K * if * W) / (Ra + Rdb) (Eqn 6)
where:
Va = ?????
ifs = steady-state field current (amps)
Rdb = value of braking resistor (ohms)

AC motors typically drive a generator at rated speed. However, the dynamic braking design requires the final dynamic braking speed of the motor/generator to be less than the rated speed. In addition, the final dynamic braking speed depends on a variable load. Therefore the dynamic braking system must operate over a speed range necessary to generate retarding torque from every load condition.  For a given field current, the generated voltage is directly proportional to the speed.

An increase in load current, armature circuit resistance drops the terminal voltage of a separately excited generator linearly. At high armature currents, an additional drop in terminal voltage may occur. This is a result of a reduction in the air-gap flux caused by the non-uniform saturating effect of the cross-magnetizing armature reaction.

The cross-magnetizing saturation effect can be minimized if the magneto motive force of the field winding exerts predominate control on the air-gap flux and, thus, avoids the condition of weak-field and strong armature magneto motive force. The tendency toward distortion of the air-gap flux distribution also occurs when the field excitation remains substantially constant while the armature current exceeds rated values during heavy loads or transient conditions. However, this effect can often be neglected for armature currents less than the rated value.

Under normal operating conditions, the hoisting machine also operates as a motor. Figure 8 shows the terminal characteristic for reverse (motoring) armature current. High values of armature current reduce the flux and generated voltage as a result of the saturating effect of the cross-magnetizing armature reaction.

Selecting the proper value for dynamic braking resistance requires knowing the needed retarding torque at the desired braking speed. The required braking torque varies from no-load to full-load. The worst-case is a full load. This provides the maximum inertia and the greatest imbalance because the counterweight typically weighs 40 to 45 percent of the rated full-load capacity. Using a counterweight of 40 to 45 percent provides the most power-efficient operations since the hoisting system typically is used for light loads. Therefore, if the dynamic braking system is designed to provide the full-load retarding torque (worst case), then lighter load conditions will have sufficient braking action.

Torque is directly proportional to the armature current by the field current and the proportionality constant. For separately excited systems, the field current is constant. The required braking torque is simply the full-load unbalanced weight applied at the drum radius through the drive transmission ratio. This steady-state dynamic braking torque is plotted as a vertical line (idb) on the terminal characteristic curve since it is a function of the armature current.

Another design consideration is the desired dynamic braking speed. Dynamic braking does not stop the load, but only limits the final speed to a designed value. Since the dynamic braking system is operating under the condition of mechanical brake failure, the overhauling load will not come to rest until it reaches the bottom buffer that reduces the impact of the overhauling load. These buffers decelerate a load traveling at 115 percent of rated speed at an average deceleration of not more than 32 ft/sec/sec (1 g). Internationally, deceleration rates less than 16 ft/sec/sec (0.5 g) are accepted as safe. Therefore, the dynamic braking speed should be no more than 50 percent of rated speed to limit the deceleration rate to less than 0.5 g when the load strikes the buffers.

After establishing the final braking speed, select the corresponding speed curve from Figure 8. Draw a load line (Rdb) through the intersection of the torque and speed curve. This resistance load line defines the possible operating points of the dynamic braking system. The initial dynamic braking effort produced when the hoisting system is traveling at rated speed is about twice as large as the steady-state retarding torque as shown by the initial dynamic braking torque line (io) in Figure 8. This initial braking torque rapidly decelerates the hoisting system along the load line Rdb to the steady-state operating point.

Decreasing the resistance of the dynamic braking resistor decreases final braking speeds and provides a greater margin of safety. However, systems designed to provide the required retarding torque at very low speeds inherently provide excessive compound braking at rated speed. Reducing the value of the braking resistance to provide the required steady-state retarding torque at 25 percent of rated speed delivers triple the braking torque when the cage is traveling at rated speed. This braking torque, when combined with the mechanical braking force, may produce an excessive deceleration rate (greater than 0.5 g).

In general, a resistor with a power rating equal to the drive motor horsepower provides sufficient power dissipation for dynamic braking applications. A power rating calculated in this manner typically incorporates a safety factor.

The separately excited dynamic braking system requires an external power source. If a power system interruption or failure of the drive power system occurs, the separately excited dynamic braking system provides no retarding torque.

Self-excited dynamic braking system
A self-excited dynamic braking system generates retarding torque without having an external power source. Therefore, in the event of mechanical brake and power system failure, the braking system provides sufficient retarding effort to safely lower the overhauling load.

Up to this point, the effect of hysteresis and the magnetic retention of the field poles have been neglected to simplify the analysis. However, the operation of a self-excited dynamic braking system depends on the residual magnetism in the field pole iron after the field winding is de-energized.

At rated speed or slower, the voltage generated is directly proportional to the speed. Under load and overspeed conditions, armature reaction may render a generated voltage that is less than proportional to the speed. This effect is shown graphically as a reduction in generated voltage during maximum armature current (near the knee of the saturation region) for the 125-percent speed magnetization curve. The saturating effect of the armature reaction is a factor that limits the braking current and torque during voltage build-up under transient conditions.

The voltage build-up develops sequentially. If the hoisting machine is stopped and no power is available when the mechanical brakes fail, the dynamic braking system must self-excite to generate braking torque. When the brakes fail, the weight of the overhauling load accelerates the hoisting machine at a rate initially limited by the inertia of the hoisting system and frictional forces. The armature rotating through the residual magnetic field of the permanently magnetized field poles generates a small voltage. This small voltage is imposed across the field winding, which creates a small field current. The braking system should be designed so that the flux the field winding ampere-turns produce adds to the residual flux. This positive feedback produces progressively greater voltages, field currents, and retarding torque. The field, armature, and dynamic-braking load resistance of the circuit and the driving mechanical torque govern the final value of armature voltage and field current.

A field-resistance line below and parallel to the linear portion of the magnetization curve, as shown for the 50-percent rated speed curve, implies a very underdamped system response. The system oscillates around the required torque and final speed. The operating point is stable at the point the field resistance line intersects the magnetization curve well into the saturated region.

The steady-state operating point is governed by the following equations:
Vdb = eg - Ra * ia (Eqn 7)
Vdb = Rf * if    (Eqn 8)
Where:
Vdb = steady-state dynamic braking voltage (volts)
Rf = ?????

The steady-state dynamic braking voltage for a given armature current is such that the field current produces a generated voltage (from the magnetization curve), which is Ra times ia greater than Vdb. The armature current, which corresponds to any value of dynamic braking voltage, can be found by dividing the vertical distance between the magnetization and field-resistance curves by Ra. The dynamic braking load current is:
idb = ia - if (Eqn 9)
where:
idb = dynamic braking current (amps)

Note that the maximum steady-state dynamic braking current is limited by the maximum vertical separation between the magnetization and field-resistance curves. The saturation effect of armature reaction limits the maximum dynamic braking current for the 125-percent rated speed curve.

Generator for various percentages of rated speed
The dynamic braking voltage intercept corresponds to no-load terminal voltage (Rdb = omega OMEGA, no retarding torque). The dynamic braking current intercept occurs when the armature terminals are shorted (Rdb = 0 OMEGA; residual retarding torque). The maximum retarding torque occurs for the operating point at the peak of the dynamic braking current. Optimum dynamic retarding effort is not always desirable when the effects of compound braking are considered.

The retarding torque is directly proportional to the field flux. Self-excited dynamic braking systems are usually operated in the saturated region of the magnetization curve. This corresponds to the upper linear portion of the terminal characteristic curve. In this highly saturated region, the field flux is relatively constant over a range of load currents and speeds. Therefore, the field flux can be assumed to be constant for the purpose of calculating retarding torque from Equation 2.

The retarding torque is also proportional to the armature current--the sum of the field and load currents. If the field current is small compared to the load current, it can be neglected with minimal error. Therefore, the armature current equals the load current. The assumption of constant field flux and negligible field current with respect to the load current allows the required steady-state retarding torque line (idb) to be drawn directly on the load-current axis as shown above for the separately excited system.

Select the desired speed curve (typically 50 percent) and draw a load line through the intersection of the required torque and speed curve. The slope of the load line is the approximate value of the dynamic braking resistor (in ohms).

The simplifying assumptions introduce error into the calculation of the resistor load line. The assumption of negligible field current with respect to the load current may produce a significant error for systems operating in magnetic saturation. The error is directly proportional to the ratio of the load resistance to field resistance.

The armature current is the sum of the field current and load current. Neglecting the field current drain from the armature circuit results in a larger retarding torque than indicated by the vertical torque line (idb) drawn on the terminal characteristic curve. Thus, the final dynamic braking speed will be slower than calculated for the dynamic braking resistance. Therefore, the predicted operating point for self-excited dynamic braking system should only be used as a general design guideline.

Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments