As part of a personal project, a rail for a TV, I would like to ask for your experience and expertise to help me answer the questions I have in the different fields that its realization requires. Knowing that beyond modeling, my goal is to make the whole thing mobile, to manage linear and circular movements, and to simulate forces and resistances. Knowing that I am new to sheet metal modeling, some questions will concern the realization of some shapes.
Where it gets interesting is that I reproduce the model myself from the ones I have in my hands. A model that is not intended to be manufactured but to be used in simulations (and renderings) to verify the viability of the project, so it must comply with the design rules.
I started as best I could to bend sheet metal for the different rails of the slide, as independent parts (I'll come back to this), then come the bearings.
For your information you can find the backstage in cad format on their site: https://web.hettich.com/fr-fr/cad# And agree with @DoubleL , why recalculate a manufacturer's data?
1°) for the simulation part, the balls are of no interest, since you can't put connectors on the balls (and simulation only works in static and dynamics are not interesting in this case.) In addition, it should be noted that no more than four balls will be in contact with the outer rail. Multiple balls are of no interest when there is an overhang greater than 1/2. I am only talking about static simulation here.
The multi-ball only exists because we don't know how much the overhang will be because it's a slide. This is also why as soon as the slide has to come out a lot, double-stroke slides are put in.
2°) for the animation part, the constraints can only be put on the balls and therefore only on the rails or parts of the rails, provided that separation lines are subtly used. In other words, a banal contact constraint is enough.
3°) For rendering with Visualize, you only make static images so the ordinary constraints are enough if you have several positions with several images.
Thank you @sbadenis , unfortunately the models are incomplete and inaccurate. They have no bearings, and some have a missing rail (including the model that concerns me, of course). They are a bit coarse in size. And I can't find the exact model, but I found one on which I should be able to recover the profile (or not, given the errors) of the rails (some dimensions are difficult to measure accurately irl) so it still helps me a little to move forward.
It'll probably be clearer when you see an overall preview, but basically, I also want to test the slide sag when opening. With a parallel mounting to the ground, the end of the outgoing rail will tend towards the ground and lose its horizontality. For a drawer we don't care, we agree, but for a TV... As with a painting, "if it's not straight, it's poop". I want to try to provide a tilt compensation of the mount to have a TV parallel to the ground when it is deployed.
Coming back to bearings, this sag is partly a consequence of the gap between the balls and the parts in contact. The other bearings of this slide are in 2 independent parts, which by the effect of sagging opposes their directions. When pressing on the rail, the high bearing rolls outwards, while the low bearing rolls inwards.
I would like to be able to consider this aspect in the simulation. Sagging by play, by slipping, and by bending. That's almost the most important thing.
If I fixed the slides with 2 screws on a board, yes, but there is no, there will be a steel monorail, pre-drilled (or even pre-tapped) between the slides and the boards and the maximum number of screws.
An alternative could be a pivot on 2 axes but you need a strong one, so a heavy one,... At the end of my arm, I'm not sure. I was going for a simple pivot, on a bearing or not.
It's not a question of doing a destructive test but rather a prototype, especially since with this kind of hardware you may end up being very far from your simulation with a different offset between the 2 sides. I also think that providing a calibration/correction system (especially since we don't know how the system will age, but it won't be for the better) would be wise.
So beyond the preparation of the system and its implementation, the idea is to take advantage of it to improve my use of SolidWorks by learning how to exploit some of its functions/capabilities. And this project happens to require sheet metal, movement, and simulation, 3 aspects of SW that I don't master, so it's ideal. Maybe the passage through SW is not always the most relevant but, at some point, to learn how to use a GPS you have to try to go somewhere, even if you know the way.
You suggested that I take out the toolbox, but apart from the message from @Zozo_mp I don't have much more information about SW, especially the bearing modeling. I should have said that before, I guess.
On the other hand, if you make an ASM such as the image monorailetcoulisse.PNG and provided you do not use moving balls, you can make a simulation with ball slides with full extended.
Concretely, your logs must be blocked and there must be only four per part of the rail. And that uses the equivalent of penetration pitch for contacts with the rail.
I find it very complicated so why not use roller slides instead Roller slides FR6500 - load 50kg - total outlet FULTERER | Bricozor Anyway, whatever the model, these parts have play in the horizontal direction, so the simulation to take everything into account will be complicated and not totally realistic to take into account all the deformations (especially those that look like spills.
So your model must have constraints that simulate these lateral games well.
If you don't want any play, either you don't use inexpensive roller slides but stronger than ball ones, or you switch to industrial
The slides will end up costing you the price of the TV (which is not that heavy even for the very very large ones (65 KG)
Between us, what bothers us is why you want to do this simulation apart from better mastering SW.
They have no bearings, and some have a missing rail (including the model that concerns me, of course). They are a bit coarse in size. And I can't find the exact model, but I found one on which I should be able to recover the profile (or not, given the errors) of the rails (some dimensions are difficult to measure accurately irl) so it still helps me a little to move forward.
