My friend and I were discussing this earlier: why don't escalators slow down when people get on? We are not interested in any fancy schmancy computers that might adjust their speed for capacity/traffic flow, we were simply wondering about the most basic possible escalator.

So our thoughts were: If an escalator is running at a constant speed, and then a single person (or more people) stand on it, it is obviously doing more work to lift that mass. Should the speed of the escalator drop? If something is keeping the speed from changing, does that mean the torque of the motor changes? Or does it simply draw more power but keep the same torque? Are we missing something obvious?

  • 1
    $\begingroup$ Not sure this is a physics question, seems to be about engineering $\endgroup$
    – innisfree
    Sep 13, 2018 at 11:35
  • $\begingroup$ @innisfree The answer is dependent on how the escalator is engineered, but you can still discuss the physics for a given "design". $\endgroup$
    – Aaron Stevens
    Sep 13, 2018 at 11:56

3 Answers 3


Should the speed of the escalator drop?

That would be undesirable behavior. A competent engineering team would not allow that to happen.

If something is keeping the speed from changing, does that mean the torque of the motor changes?

The steps form a chain. Weight on the steps is transmitted up the chain to the drive sprocket that keeps the chain moving. Any change in weight on the steps equates to a change in torque on the sprocket's drive shaft. And, that torque must be transmitted through some transmission to the motor shaft.

Or does it simply...keep the same torque?

That would not make sense. See above.

Are we missing something obvious?

No, I don't think so. The system is designed to keep the motor running at a constant speed. For some types of motor, that can be achieved simply by connecting the motor to a constant voltage power supply. For other types of motor, there may be a controller that monitors the speed and adjusts the voltage accordingly. Changes in weight equate to changes in torque which cause changes in the amount of electric current (and therefore, the amount of electric power) drawn by the motor.


The response of the system you describe is of course dependent on the construction. Even if we do not allow "fancy schmancy computers", mechanical RPM-controllers have been around since the steam era.

As a first example we can assume that we have a simple escalator that is powered by an engine with an RPM controller. This engine will likely have a maximum torque. The actual torque excerted by the engine is dependent on the load, and the decrease in engine RPM (i.e. elevator speed) that you are expecting will only occur if the load is greater than the maximum torque of the engine at the set RPM.

Here, the excerted torque - and in normal cases, the consumed energy - of the engine will increase with increasing loads, but the speed will be constant (until the load is greater than the maximum torque).

In a second example we can assume an engine running with a constant energy flow (e.g. a gas throttle) and no RPM controller. In this system the load increase will directly incur a decrease in escalator speed - the work performed by the system is dependent on the inflow of energy to the engine, and as we stated that this is a constant, the work must also be constant. With more weight added, the lifting speed must be lowered.

  • $\begingroup$ FYI: Many electric motors approximately obey a simple law: Shaft speed is proportional to voltage, and torque is proportional to current. If you drive your escalator with an electric motor that obeys that law, and if the motor is connected directly to the power grid (which does a very good job of maintaining constant voltage under normal conditions), then you get a constant speed drive without any need for a fancy controller. $\endgroup$
    – besmirched
    Sep 13, 2018 at 16:57
  • $\begingroup$ OTOH, fancy electronic motor controllers are relatively cheap these days, and they're popping up in all kinds of applications that used to be implemented in a more simple way. (Don't ask me how I know!) $\endgroup$
    – besmirched
    Sep 13, 2018 at 16:58
  • 1
    $\begingroup$ I agree with @besmirched. Escalators use AC motors. The AC grid is held to a constant 60 Hz frequency. If you load an AC motor, it is still "locked" to the grid. That motor will respond by drawing the amps and power needed to stay locked to the power grid's 60 Hz frequency. I've seen this action involving large industrial motors, none of which had fancy electronic motor controllers on them to control their speed. $\endgroup$
    – David White
    Sep 13, 2018 at 18:29
  • $\begingroup$ @DavidWhite You are of course right. If we assume an AC motor, then it is natively RPM controlled via the power grid - but for the sake of presenting two examples, I meant to portray a simpler system. The important bit of the example was that the engine was RPM-controlled, and I am at fault for dragging the details of how into the example. Thank you for clarifying! $\endgroup$
    – Sternerson
    Sep 14, 2018 at 7:21
  • $\begingroup$ Mechanical rpm controllers have been around since Henry VIII commissioned the Royal Society to improve England's fleet of Windmills. A survey of existing mills discovered the flyball governor already in use. It regulated a gate that controlled the inflow of grain to the millwheels, thus adjusting the load to the power available at any instant (and relieving the miller of having to keep his nose to the grindstone to judge feed rate by smell). $\endgroup$
    – Phil Sweet
    Sep 15, 2018 at 1:41

It depends on how your escalator is running.

It seems like you want the simplest thing possible, and it seems to me the simplest thing is an escalator whose mechanism can only apply a single, constant, force/torque (I say force/torque because we aren't being specific about what is actually causing the escalator to move. From now on I will just say force.).

Let's consider before anyone gets on the escalator. It is supplying enough force to move the mass of the escalator parts as well as resist internal frictions that are present (which for our discussion we will assume to be constant friction forces that do not depend on the velocity of the escalator. Having velocity dependent friction could change some things, but I will leave thinking about that to you). Let's assume at this point the stairs are moving at a constant speed.

Now someone steps on. If we want to maintain a constant speed up the escalator, the thing supplying our force would need to increase its "strength" to counter the new "resistance" caused by the person stepping on. It needs to be able to lift the stairs, fight the friction, and now also lift the person. Since our force being supplied is constant, it cannot do this. Therefore the escalator would actually begin to slow down, and then eventually start to move backwards if the person does not get off in time (this could change with a velocity dependent friction force). This would be analogous to me lifting my hand at a constant speed, and then someone puts a mass in my hand while I supply the same upward force to my hand. The net force on the mass will be downwards, so it will have a downward acceleration.

If something is keeping the speed from changing, does that mean the torque of the motor changes? Or does it simply draw more power but keep the same torque?

So let's say our escalator can adjust its applied force/torque accordingly to keep the escalator moving at a constant rate. Then yes, the force would need to increase to keep people moving at the constant rate. Just how if someone puts that mass onto my hand I will need to push harder to keep my hand (and mass) moving at a constant speed. This also means more power is needed too. If we want the escalator to move at a constant rate, but it is doing more work to lift the people up, then it is doing more work in the same amount of time. Therefore, the rate of energy expenditure increases, i.e. more power needs to be supplied.

Therefore, to answer your question title, escalators don't slow down because they are designed to be able to rotate the steps at a constant speed by adjusting the supplied force/torque.

  • $\begingroup$ You're right. I skimmed it. Thought I saw something different from what it actually says. $\endgroup$
    – besmirched
    Sep 13, 2018 at 17:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.