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

Adjustable Torsion springs

  1. Feb 14, 2017 #1
    Is there anything like adjustable torsion spring, even if it is non-helical spring based (such as say hydraulic or pneumatic). As in you can change their stiffness/spring rate?
    I know that there are so many videos on adjustable springs on garage doors but I think all they do is adjust preload.
    There's a small electric scooter company that claims to have adjustable torsion spring for their rear swing arm suspension. Sorry I can't find the link now, but just wanted to confirm if such a thing can actually exist.
    (I know we can adjust preload on normal suspensions/springs to change ride heights etc).

    Thanks in Advance
  2. jcsd
  3. Feb 14, 2017 #2


    User Avatar
    Science Advisor

    Yes it can be done. Given a creative mechanical engineer, anything is possible.

    Effective spring constant is usually changed by moving the spring attachment point along a lever relative to the fulcrum. For a torsion spring made from a torque tube or rod, changing the active length would change the spring rate.

    If the torsion spring was a coil spring, then inserting a cylindrical rod part way into the spring, or a tube part way over the spring would restrict the length of the active spring and so change the effective spring constant.
  4. Feb 15, 2017 #3
    Effective spring constant is usually changed by moving the spring attachment point along a lever relative to the fulcrum. For a torsion spring made from a torque tube or rod, changing the active length would change the spring rate.
    I understand you can theoretically change stiffness of spring by changing the effective length. Something like how potentiometer work to change resistance in electrics. However, doing that practically seems a bit difficult. Any working examples already being used? Though a torque tube or rod might work, will have to work on it though.
    If the torsion spring was a coil spring, then inserting a cylindrical rod part way into the spring, or a tube part way over the spring would restrict the length of the active spring and so change the effective spring constant.

    I did not understand this. Can you please explain?
  5. Feb 15, 2017 #4


    User Avatar
    Science Advisor

    When a coil spring is used as a torsion spring the coil diameter will increase or reduce as it is axially twisted. By preventing that change in diameter over part of the coil length you reduce the effective length of the spring and so it requires more force for the same movement.

    No. Preload is adjusted. But by moving the anchor points along a lever, the effective rate can be changed without changing the length or preload.

    The answer is yes. You are new to the field and there are many possibilities. Why do you ask the question ?
    If you want a useful example then you need to ask a specific question about an identified application.
  6. Feb 20, 2017 #5
  7. Feb 20, 2017 #6
    Are you familiar with tilting vehicles. The above image is what I plan to make. For the rear wheels , they can have any form of suspension such as swing arms/trailing arm, double wish bone etc. But for tilting the frame/seat vs the rear wheels I need to have pivot point. I want to add a return back action (that is the vehicle straights itself when unloaded). Best case would be to add a torsion spring, that deforms both clockwise and anti-clockwise. In case I don't get a torsion spring of the right size/stiffness I can try adding regular axial spring.
    Here are the problems:
    1) Except your reply to a different thread on PF, I have yet to see a spring that would be in tension and compression. The diagram/mechanism is beautiful. Are there actual bidirectional springs available commercially or is it your design. It does not look too difficult to make but yes getting all fits right will be expensive//difficult for me.
    2) Are torsion springs also unidirectional only?
    3) I need something around 5000 - 6000 Nmm/degree (note the above image is an early concept image that was readily available/clean, the tilting is a different mechanism that also makes sure the rear wheels tilt as well).
    4)Will adding two compression spring (that are already preloaded/compressed) on opposite sides of an arm which is pivoted centrally sort of work similar? This was my initial idea and someone on a different thread on PF also suggested the same to someone else.

    Any tips or recommendation on the above. I need something that is a quick fix, because I will be trying different springs stiffness for adjusting comfort level. An adjustable torsion spring would be great.

    Thank you for your replies, specially about the adjusting of torsion spring trick.
  8. Feb 20, 2017 #7


    User Avatar
    Science Advisor

    Link to previous post;
    1. My design there is a simple demonstration version of what is implemented in hydraulic directional control valves. You need to make your own.
    2. Torsion can be done the same way. See attached picture where a square cam is pinched between two flat leaf springs. The cam profile sets the restoring torque to angle relationship.
    3. Choose the leaf spring thickness, length and cam to set initial torque to move.
    4. Two springs in opposition will not give a firm return position. The deflection against torque line will move straight through zero like any single spring. You need a response where the deflection – torque line has a vertical step through zero.

    Attached Files:

    Last edited: Feb 20, 2017
  9. Feb 21, 2017 #8
    For 2) So the cam will be attached to the moving part and clamps to fixed part? I did not understand the mechanism well
    4) Why is that? Why will they not give ma firm return position? Is it no matter how fine I adjust preload, there will be some off balance or it will be like an inverted pendulum, that it will be static but unstable?
    'The deflection against torque line will move straight through zero like any single spring. You need a response where the deflection – torque line has a vertical step through zero.'
    Can you explain the above in detail. Sorry I am still new to springs and couldn't understand anything. (though I am a mechanical engineer and can hopeful pick things up quickly)
    What about Damping? I need to add some damning or the board will keep overshooting?

    What if I use just one torsion spring? Is it really that bad to rotate in the opposite direction than intended to?

    What if I use 2 springs, in series, one coiling clockwise other anti clockwise. Put them in tight cylinder such that they can't expand outwards but can shrink inwards (is that the correct way to use spring, if not, we can do the opposite, by putting them over a shaft so that they can't shrink but can expand).
    So in clockwise motion one spring will be redundant and other will work, vice versa.

    Thank you so much for your reply
    PS: I can at least try your first suggestion of a bidirectional spring and put it over an arm for rotational leverage.
  10. Feb 21, 2017 #9


    User Avatar
    Science Advisor

  11. Feb 21, 2017 #10


    User Avatar
    Science Advisor

    Attached is the original picture from wikipedia, edited to show the special response in blue.
    As I understand it, you want the blue line response which has an accurate zero return position.
    Any normal springs in series or parallel will have the red line response with a soft zero return position.

    If you use two separate springs for the two separate directions, you must calibrate the break-point force and the spring rate to make the diagram symmetrical. That would be difficult for springs with different diameters.

    Attached Files:

  12. Feb 22, 2017 #11
    Thank you for your reply. Can you please tell me why exactly a torsion coil spring cannot be used in both directions? Check image spring 4. There are so many guidelines that suggest not to use torsion coil spring in uncoiling direction. What happens if we do? Both short term and long term usage.

    Assuming Torsion spring is not an option here is what I plan to do.
    Look at image spring 1: The springs are installed in preloaded compressed condition. But once attached properly, preload will be released (the nut stopping it will be loosened). Hence each spring will try to reach its free length. However it will still be preloaded/compressed due to each spring trying to push the other. When the vehicle is tilted (clockwise as shown in image spring 2), one spring will compress even more, other will decompress (limited upto amount initial preloaded compression)

    Will the above work? Challenge will be to balance them well at the time of initial release of preload.

    Check image spring 3: Just use a bidirectional spring (same as you have mentioned in different thread). Difficult to manufacture but should work in principle.

    Thank you so much for your guidance.

    Attached Files:

  13. Feb 22, 2017 #12
    On image spring 3: An airspring can also be used I guess (similar to one normally used on the boot/rear door of a hatchback/SUV/miniwan
  14. Feb 22, 2017 #13


    User Avatar
    Science Advisor

    I believe the problem is one of “column stability”, a long thin wire will be more stable in tension than in compression as it would kink.

    Your spring1 method will not have a step in force at the zero position. The handle of the scooter is a vertical pendulum. Without the stable zero position, it will be floppy and always fall obviously, to one side of zero.

    If you find a few pairs of springs or rubber bands to play with, you will quickly develop an understanding of the limitations and possibilities of springs in series.

    Spring3 would work. But to fit it within the body you need to mount the mechanism horizontally with a very short vertical lever arm. With the short lever arm and only one spring, the high side forces on the shaft will be asymmetrical, so it will wear faster. Extending the design to symmetrical balanced forces, quickly approaches the conceptual square cam between two leaf springs. Optimising that square cam will define the force versus displacement response. To reduce friction and lubrication requirements, the square cam could be made from two standard ball bearings, one on each side of the shaft, between the leaf springs.

    Two air springs would need to be mounted next to each other in a subtractive mode. Again it becomes two close fulcrums on a short lever with high forces.

    There is the possibility that the two board parts could slide apart along the pin, pulled close together by a spring. The profile of the junction of the two parts where they meet will give the stability at zero. If the design has the scissor or shears shown, it could injure a rider or bend the shaft. By having a tension spring, the gap could widen to reduce injury and damage possibilities.
  15. Apr 1, 2017 #14
    Hi Baluncore,

    I actually head changed the tilt mechanism a bit. Though concept of having a spring remained the same.

    For your info, the bidirectional spring by itself did work.
    But in my system it did not work as intended.

    Firstly you were right about having a step in force at zero position. That was partially solved by adding some preload in the spring.
    Secondly, even though one could clearly see that the system works in both directions, the response was sadly not the same in both directions. Low fulcrum length, asymmetric fulcrum, could be causes.
    The spring also started showing hysteresis (no one told me a spring has to be hardened, not even the spring manufacturer). The spring started deforming permanently (I don't know if that was typical plastic deformation similar to say a steel scale/ruler being bent permanently).

    Anyway, thank you so much for your advice. I am trying to upload pictures but for some reason I can't right now. I will most them later.

    Above is mechanism in the video is quite similar to what I have made.
  16. Apr 1, 2017 #15
    Here is a rough sketch. Don't know if you can understand it.

    Attached Files:

  17. Mar 21, 2018 #16
    I'd like it if you could expand on limiting the expansion of the diameter of a coiled torsion spring. In either a garage door application, or say like a pull starter for a small engine. Could you adjust the amount of torque by how many lays of the coil you limit?
  18. Mar 23, 2018 #17


    User Avatar
    Science Advisor

    Welcome to PF.
    Yes. But you can only make it harder and with less movement.
    There are many parameters you might adjust. Study spring design and you will find parameters that can be changed.

    A coiled torsion spring would be designed to become smaller in diameter as it absorbed energy. Once it is smaller you could slide an external sleeve onto the outside of the spring. Alternatively you could make a cylindrical block that could be wound or screwed inside the coil when it was relaxed. That would reduce the effective coil turns count.

    A pull start on a small engine usually uses a flat clock spring. Reducing the number of turns could be done by clamping, but I see no advantage in doing that. Again it would harden the spring and reduce the movement.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted