Welcome to the Onshape forum! Ask questions and join in the discussions about everything Onshape.
First time visiting? Here are some places to start:- Looking for a certain topic? Check out the categories filter or use Search (upper right).
- Need support? Ask a question to our Community Support category.
- Please submit support tickets for bugs but you can request improvements in the Product Feedback category.
- Be respectful, on topic and if you see a problem, Flag it.
If you would like to contact our Community Manager personally, feel free to send a private message or an email.
How to create this specific mate ?
eric_pascual
Member Posts: 8 EDU
I'm designing a mechanism including a small R/C servo actuating some kind of trigger :
The red hook rotates freely around the bolt at the end of one of its arms and the horn rotates around the servo output axis. No problem for modeling the related mates. But hat I'd like to model is the mate between the horn and the hook, so that the hook moves when the servo horn rotates. I've tried various kind of mates, using faces or edges of the involved parts, but without any success.
I'd be more that grateful to any OnShape wizard who could point me to the right direction… or to say that it is not possible to model such a thing.
Thanks in advance for any comment.
Answers
Hey @eric_pascual, so when the grey part rotates, it in turn rotates the red part? The Tangent mate should be able to do this. It will need some helper shapes to get it working properly since there currently isn't a "Partial tangent" mate which would allow it to be tangent only when it's touching. Feel free to share a link to your document if you need assistance getting the tangent mate to work.
Learn more about the Gospel of Christ ( Here )
CADSharp - We make custom features and integrated Onshape apps! Learn How to FeatureScript Here 🔴
Many thanks Michael for your quick reply.
Here is a link to the document : https://cad.onshape.com/documents/e611f9c9fe382489dc6c6b31/w/aa15001cdbadba9e845ac582/e/bca21fd677662a557f8ba01b?renderMode=0&uiState=6786e3a631048d13ede0ccc0
This is a triggering mechanism I'm adding to a double pendulum I've built for a student working on an experimentation project about the chaos theory.
The goal is to reproduce as exactly as possible identical initial configurations by placing the trigger at fixed positions (thanks to an additional external support) and observe that even then, due to unavoidable infinitesimal differences, the motion of the pendulum for several trials will be completely different. The experiments are recorded with a high speed camera (240 FPS) and a video tracking app (Tracker - https://physlets.org/tracker/) is used to record the trajectory, speed and accelerations of the red and green markers at the end of the arms.
Here is the full view of the trigger :
The yellow part is some sort of piston, used to hold the pendulum arm at the chosen initial positon by placing it in the end fork. Elastic bands attached to the hooks of the piston and the blue base plate pull so that it quickly retracts and releases the pendulum at experiment start. While armed, the piston is kept in place by the red hook. When the grey horn rotates, it pulls the hook up, which makes it release the piston.
This could seem a bit convoluted as a mechanism, and one could ask why the servo horn is not used directly to keep the piston in place. It was the option used in the first version of the mechanism, but it reveals that the constraints applied to the horn by the piston are too important and in a wrong direction, resulting in high chances to have something broken very quickly. In the current configuration, these constraints are all absorbed by the rotation axis of the red part.
I hope that these explanations make my goal easier to understand.
WRT to your remark about a partial tangent which would be active only when the parts are in contact, I'd add that I'm fine with a model where they are always in contact. But even then, I've not been able to model two perpendicular edges (one on the hook and one on the servo horn edges) sliding on each other.
Thanks again for you support.
After having a closer look, you don't even need a tangent mate, you just need to put some mate connectors in well placed spots to allow for a sliding mate paired with a revolution mate. The sliding mate could have limits so that when the actuator gets near the red arm, it starts to hit the slider limit in turn pushing the red arm the correct direction.
Will need access to copy the document for further guidance.
Learn more about the Gospel of Christ ( Here )
CADSharp - We make custom features and integrated Onshape apps! Learn How to FeatureScript Here 🔴
My bad, I was thinking that the sharing link allows writing. I've changed the document to public, so you should be able to play with it now. Let me know if it's ok.
I've tried various configuration of sliding edges but none has worked. But I'm far from an OnShape expert, having used it until now to model simple mechanisms composed of 3D printed parts.
Besides this is no good way to design a trigger (make servo and trigger axis parallel and you will avoid a lot of issues!), a tangent condition won't help here, simply because there is no tangency, only two edges being coincident at one point, and these are not necessarily the same two edges all the time. If we had collision detection, that'd be the way to go, but we haven't.
May I propose using a gear mate between the two rotate mates used for servo arm and trigger rotation? Moving one would rotate the other in a given ratio, then. This cannot simulate the free travel of the servo arm before it touches the trigger, of course, but maybe it could get close enough to demonstrate the action intended.
Got it 😀
I've created 2 dummies (almost transparent small spheres, but they can be of any shape):
Dummies can be hidden for the final rendering.
I've ended creating dummy parts because I could find a way to create mate connectors "in the air" (such a choice is not proposed when creating a connector). Is there a way to avoid them ?
BTW I didn't succeed in fully creating the real mechanism, in which the actuator is not always in contact with the hook, but grab it only in the upper part of its motion. You've mentioned the use of mate limits to achieve such a result, but I failed having it work. Could you tell me more about this point please ?
Thanks Martin for your suggestion.
It was my first idea but I've rejected it because I flet it as some kind of of cheating 😉. We are not simulating the real physics but visually faking it (there is not a real gear ratio between both rotations, because of the linear motion of the contact point in between).
As explained in a later post, I've ended up modeling the contact point so that the physics is respected.
What I miss at this time is modeling the part of the motion where the parts are not in contact. But it's more an intellectual challenge than a real need, because what was important is to check the "active" part of the mechanism.
The part where there is no contact between the parts is a bit similar to the geneva wheel setup dicussed in this forum earlier. Suggestions were, to create some kind of 2D animation panel in 3D and apply tangency to dummy-objects moving along curved surfaces in that panel. In another step, the vertical movement of these dummy objects would be connected to the movement of a part in the model by means of a gear or rack/pinion ratio mate. That way, editing the slopes of the guide curve in that panel would control the movement of a part, and any straight horizontal part of the guide would mean no movement. John McClaray explaned that in this posting, where he animated rubic's cube.
It is getting pretty messy pretty soon. In your case, though, only two sliders on one pretty simple 2-level curve each would do: One for the trigger and one for the piston.
And then comes the double pendulum …. ;0)
Then there's also this:
Thanks everybody for your suggestions.
I've watched with very much interest the animation system videos and found them amazing. The base idea is very smart too, and it reminds me of the animation systems available in applications such as Blender and alike, where you create tracks and key frames for animating various parameters of the scene (bodies DOFs, visual aspect,…).
However, applied to my case (as for the Geneva wheel BTW), it's animation and not mechanical simulation. My initial goal was to simulate the real mechanism, in order to check that it is behaving as I imagine and not to animate it as an illustration.
Gotta agree with @martin_kopplow though…
Having the servo axis parallel to the trigger axis would probably work better (not talking about CAD but "real-life"). I would do something along these lines:
Something like this will hold the plunger steady until a sudden tipping point where the trigger will get pushed out (when corner of plunger reaches radius of trigger). The wider slot in the trigger in that area allows the trigger to freely rotate at this point.
The release will be the same regardless of how fast the servo is moving.
Yes, in the case Eric prposed, you'd easily overvome the friction while the servo motor spins up, and then also eliminate backlash of the plunger on the trigger. One could even slightly modify the slot shape so that there is a "loading" state where the trigger can still move freely and then a "locked" position with the plunger in the 'loaded' state but the trigger immobilized by the servo.
Thanks Eric and Martin for your comments.
I agree that it could work too, but not so well in reality. Because of the profiles of the faces in contact, the mechanism would not be auto-locked when the piston is armed, unless the servo maintains it in place. The pression the piston applies to the cam (due to the force applied to it by the springs pulling it) makes it turns towards its "open" position. I had to rework my parts to take into account such phenomenons that I've overlooked in the initial naive design.
The detail view hereafter shows the working profile for the faces in interaction:
As you may have guessed, the normal at the contact point must always pass "above" the pivot access at rest, so that the torque resulting from the piston pushing leftwards tends to rotate the trigger CCW, thus locking it in place without needing any spring or alike. I've tested that this configuration works reliably even upside-down, where the gravity tends to move the trigger in its unlocked position.
WRT the servo rotation axis orientation, I agree that a 90 degrees variant would work too. But in my case, it would have required a more complex part in terms of geometry and thus more constraints on printing it in 3D then. I do my best during the design to avoid supports at printing time and also to take into account the lower strength of the Z axis compared to the XY plane. In addition, it would have resulted in a less compact system, with the back of the servo protruding more than in my current configuration, and thus being more exposed.
Another benefit of the "open" shape of the upper part of the trigger is that it can be easily actuated by hand, needing only a very slight push to the upper corner for unlocking, hence creating less perturbations to the experiment itself. To put the whole things in context, the goal of the contraption is to reproduce as closely as possible the same drop conditions of the pendulum, to demonstrate than even if the initial conditions seem identical, the inevitable microscopic differences lead to completely different trajectories, because of the underlying chaos introduced by the configuration of the pendulum. This is BTW why I've included the servo as a remote actuator avoiding any physical contact with the operator.
I'm very grateful to all of you for your comments and suggestions. Analyzing other approaches to a given problem is always exciting and fruitful. Thanks again for your time and efforts contributing to the discussion.