Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What is an inertia load?

  1. Sep 16, 2014 #1
    I'm reading an old engineering book about aircraft Structures from the 1950s and I don't recognize the term inertia load. Can anyone give me a definition of it?

    I'm going to take a guess based on what i do know. Inertia is a body's resistance to change in speed and direction. So when some force were to try to make ti change speed or direction, there then exists some force of resistance in the body. And that force that resists change in speed and velocity is the inertia load.

    That's really my best guess since the terminology is unfamiliar to me and I have not seen it in any of high school level and intro physics textbook. The word inertia alone yes but not as inertia load.

    Those older textbooks are giving me a hard time with unfamiliar terminology, some of which seems outdated. I've searched for "inertia loads" in Google but I find it being discussed in some papers but can't get a definition for it.
     
    Last edited: Sep 16, 2014
  2. jcsd
  3. Sep 16, 2014 #2

    AlephZero

    User Avatar
    Science Advisor
    Homework Helper

    That's the basic idea. It is often used in the context of the distribution of internal forces (stresses) through the body when it accelerates (i.e. changes velocity) rather than just considering the resultant force on the whole body.

    In more modern books on mechanics you might find the term "body forces" used instead, though that can include other effects like electromagnetic forces.
     
  4. Sep 16, 2014 #3

    rcgldr

    User Avatar
    Homework Helper

    It could mean this: One way to measure torque and power output of an engine is to have the engine accelerate something like a big flywheel, which would be the inertia load. The angular inertia of the flywheel is known, then by measuring angular acceleration, the torque versus angular velocity of the engine can be calculated, and the power versus angular velocity equals the calculated torque times angular velocity during a run.

    There are dyno's for automobiles that use the same principle. The driven wheels drive a large drum, and the angular inertia of the drum times the angular velocity of the drum, divided by the radius equals the force generated at the rear tires. This force times the speed of the tires equals the rear tire power output. If the engine rpms are also sampled, then the overall gearing and engine torque can be calculated.
     
  5. Sep 17, 2014 #4
    I am involved with industrial robotics and deal with this all the time. I spend a lot of time trying to get folks to understand the concept.

    Consider this. Websearch for an image of an automotive spot welding robot. The robot carries a big, honkin' spot welding gun on its tool flange. The gun has a certain mass, and in gravity that manifests itself as a weight. The robot must also position the gun and does this by rotating Joint5 (wrist) and Joint6 (tool flange). So it must provide torque (via servo motor) to those axes to rotate that mass. But the resistance to motion that the torque must overcome is determined by the inertia of the gun ****about the respective axis****. The inertia is largely a function of the geometry of the gun relative to the axis of rotation. Look up the mass moment of inertia formulas for any of the typical "primitives" shapes.

    The robot has specs of "payload" which most people can understand. Robot specs also (usually) specify mass moment of inertia limits about Joint5 & Joint6 axes. That defines torque loads applied to the joint axes. Most people don't have a clue about that. Many times the machine designer screws up the robot application because they paid attention to mass/payload but not inertia.

    It's the old physics lecture experiment: put the student in the rotatable instructor chair in front of the class. Give the student a weight to put into each hand. Tell student to hold the weights close to chest, and spin the chair. Chair spins rapidly. Tell student to extend arms out from body during rotation. Rate of rotation slows. That's conservation of momentum, function of inertia.
     
  6. Sep 18, 2014 #5
    I'll try to understand that.
    Regarding inertia limits, let's say that a robotic arm is attached to a tool at one end and a motor at the other end. Once the motor starts to move, it would also get the arm and thus the tool to start moving. But then let's say that the robot needs to slow the arm and tool down. If the arm and tool is moving too fast such that when the motor stops, the arm and tool still have inertia that try to keep them moving but since the motor has stopped but the arm and tool are still moving, if the inertia limit is exceed, the moving arm jams up the motor which has stopped.

    Is that a correct usage?

    If the inertia limit were not exceeded, the stopped motor could handle the torque generated by the arm and tool which still want to move due to inertia.

    I just want to check that I'm understanding this correctly.
     
  7. Sep 20, 2014 #6
    Sort of. Not really.

    If you have a linear link, tool on one end and attached to motor on other end, pivots about motor shaft: then drive motor to rotate shaft about its end.

    The link will rotate about one end, the tool flies through space. If the motor shaft rotation is stopped suddenly and locked in rotation, then you still have the rotational inertia of the link + tool mass moment of inertia. It has to go somewhere (conservation of momentum, and all that). So where does it go? Possible outcomes: (1) The link deflects elastically or plastically. If elastic, then you have a pendulum-type affair. (2) The applied torque to the motor shaft due to rotational inertia tries to keep the motor turning. Either the motor's brakes (if it has any) screech through some amount of rotation or the motor has an external torque applied to & it keeps turning a bit. Hopefully there is some motor torque resistance. If the motor is a servo, then you have the servo driver attempt to move the rotational load back to it's setup position by applying negative torque. If the amount of torque required to move the rotational inertia back to position is too much, then the servo driver will probably fault out.

    I helped a customer with an application once. He put a large aluminum disk on the joint6 tool flange of a robot. The MMI was too great for the robot's torque capacity for Joint6. Command the Joint6 to move to position, and the disk would move, overshoot, then oscillate about the target position because the servo motor did not have sufficient torque to control it. Try to move it too fast and the driver would fault out because it exceeded the error specification trying to achieve that target. I had to get the customer to redesign their tool...they were very embarrassed and the designer lost his job because he didn't have a clue.
     
  8. Sep 30, 2014 #7

    OldEngr63

    User Avatar
    Gold Member

    AlephZero, this is not entirely correct. Body forces include things such a gravitational loads, and that is not an inertial load.

    Inertial loads are the M*a terms in an equation (or system of equations) of motion. They look like forces from a dimensional perspective, but they do not obey the third law.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook