Tilting Trikes Part 4 - CAD model demo tilting / steering mech design for reverse trike velobobile

This video builds on my previous videos and is Part 4 in the series on the subject of designing the tilting and steering mechanism for tilting trikes. In this video, I demonstrate various tilting and steering mechanism designs in CAD (Computer Aided Design) so you can see how well they perform. I am looking particularly at finding a design where the tilting action does not affect the steering action and where Ackerman steering geometry is maintained at all tilt angles.

[@CraigLandsberg-lk1ep]

Was glad to see this 😅 the only thing I could add from my racing experience is that when turning, toe out is good cause it allows the inside wheel to turn more, since the inside wheel does actually have a tighter turn radius to overcome, but I don't have any actual experience with tilting trikes...yet😅 my experience is based on 4 wheeled rally cars with 7 degs of caster and a modest 3 degs of negative camber, but I am building my own tilting trike and have made a cardboard mockup, but definitely haven't considered as many parameters as you have. Good work and would love to see how your geometry translates into a working model, thankyou 😅

[@LeesChannel]

These videos are really great, I appreciate all the time and effort you went through not only thinking of several approaches, but also sharing them and illustrating their operation and how they are altered.

[@patrickfle9172]

Ackermann is over estimated. It's nescessary to follow it for very sharp turns and loses practucal effect in large radii. You might want to consider the orientation of the contact patches when leaning and sterring, as these 'climb' up the front of the wheel and change in direction significantly (orthogonal to the direction of travel for any steer angle when tiltet 90°, as an unrealistic example)

[@MetalMachineShop]

Hi, yes I agree, Ackerman less of an issue for large radius turns. I had thought about the contact patch issue - I think it works with you rather than against you - but I shied away from trying to model it. Thin racing tyres would reduce the effect!

[@koustubhheblikar4693]

Was eagerly waiting for this sir! Awesome work as always. Let's hope the tilting trike works. If it doesn't, we will make it happen

[@MetalMachineShop]

Thanks! I have now started work on the prototype so the next video will be looking at that. Hope your project is coming along well!

[@richardschaefer8605]

Really digging these video's. Hope to see one of these design's put into a working prototype.

[@MetalMachineShop]

Thanks! You will see a prototype soon, hopefully a working one. It is currently being made in my workshop. Stay tuned...

[@richardschaefer8605]

@@MetalMachineShop Yep. Count me in.

[@drumandbassob0007]

@@MetalMachineShop i am just starting to gather parts for a similar design i have lots of bike parts lying about that i will use prob go 26inch at the rear tho and no plans lol. I have a good idear what i want to do but prefer to work in real life than in cad i adjust as i go and problem solve that way ✌

[@MetalMachineShop]

drumandbassob0007 the advantage of your approach is that you actually get on with it rather than think about it for ever without making anything! Good luck, let us know how you get on!

[@benjaminkarlsson6854]

I would love to see the "spot on" design that also has suspension! :)

[@necdetsaltek6790]

This is engineering. Thank you for your given information.

[@benedictgrimm5235]

THIS is sooooo awsome done!!! I learned so much and I tryed to get it in my head for days and you learned it to me in minutes!

[@MetalMachineShop]

Cool, thanks!

[@rupeshpadwalkar3794]

Perfect! Just loved it 😍

[@vech400]

When you have the wheels turned your aiming to have 0 toe, The idea of the Ackerman angle is to turn the inside tyre more than the outside. The inside tyre has to turn on a tighter radius as it turns so effectively you should have toe out. The more steering angle(tighter turn radius)the more toe out effectively you will need. To calculate it correctly, picture you are looking down on the vehicle from above and you draw a line thru the Rear axle. If you select a point(for example 3m) from the center line of the vehicle, (this will be the turning radius)and draw a line from that point thru the center of each front wheel, the 2 triangles you create will have different angles with the inside tyre having a greater angle. So to calculate Ackerman correctly you need to know the wheelbase of the vehicle. A really short wheelbase will need to turn the inside wheel more(effectively greater toe out) to turn the corner than a long wheelbase. It’s a lot easier to explain with a picture but I hope this helps. Just think that if the vehicle did a full 360* turn. The diameter of the circle created by the inside wheel will have a smaller radius than the outside so therefore it has to turn more than the outside wheel so it doesn’t scrub. Don’t think of it as toe out as this is making you think that the wheels will scrub, the wheels actually need to be on different angles so they don’t scrub as they need to turn on different radius circles.

[@sirriozer6618]

Mükemmel mühendislik !👍👏👏👏🇹🇷

[@skraman69]

As usual superb!!!

[@roman_abelardo]

Excelente trabajo 👍👏👏🤗🇦🇷

[@davidwielemaker]

Amazing work as always!

[@MetalMachineShop]

Thanks! Hoping to bring you a working prototype soon. It’s part built in my workshop at the moment so stay tuned!

[@lukebrown6913]

Hi. The bell crank seems to be the issue...if the bell was circular in shape, pivot in the middle and steering rows fixed at 9 and 3 o'clock this should eradicate toe in/out,?

[@leandromendez1211]

Buenas tardes... le consulto el desplazamiento del ángulo es paralelo al plano teórico, cual seria la distancia ideal de desplazamiento entre asta el punto de apoyo de la rueda? saludos. muy bueno el proyecto

[@leigao6120]

Great!!!

[@nicodanger13]

Tilting vehicles when corners at high speed have the steering almost straight. when this trike steer more it goes very slow and almost vertical, You have to consider that while choising the geometry tune.

[@thecorbies]

Hi. Beautifully explained, and demonstrated, but may I ask what software you use for this please? Sketchup, as far as I know doesn't have this facility. Fusion360 does, but has a steep learning curve, so maybe this software is somewhere in between. Regards Mark in the UK

[@MetalMachineShop]

Hi Mark, thanks for your comment. I used Onshape for this. It takes a bit of learning but their site has excellent tutorials and it’s worth the effort imo. You can get a free subscription too. Definitely recommended.

[@thecorbies]

@@MetalMachineShop Thank you. I shall look into it straight away.

[@keraba1]

Why don't the tie-rods and lower wishbones meet in the middle? It doesn't seem too hard to imagine a joint that could handle overlapping axis'. Would that help the geometry?

[@MetalMachineShop]

Kevin Bailey hi Kevin, there would certainly be some layouts where this would be an advantage. The sort of joint you envisage is possible but not easy to design!

[@VadimR4]

Akerman saw toothed guides, shouldn't they be parallel with the ground (and not perpendicular to the wheels) since the vehicle would be at an angle?

[@MetalMachineShop]

Vadim Romanovich I think you may have a good point there, but I think irrespective of this, the way I have done it still indicates that the wheels are steering correctly with respect to each other - at least approximately - although the centre of the turning circle may be different. I might need to give it some more thought!

[@kiwiwombatman]

Great work !! I don't think steering angles are very important while leaned - like a motorcycle, you use very little steering when leaned over ..

[@MetalMachineShop]

True. It’s more of an issue for slow speed tight turns but it’s good to try to get things as right as possible even if it’s only a theoretical advantage!

[@VEC7ORlt]

Everything you are saying is correct, but somehow I feel that a better bike would be a delta - like velotilt, where tilting and steering is decoupled - not that its better pe se - all the aspects of it are easier to iron out separately. Its a shame the there aren't that many lean steering trikes, some if not all of them are plagued by inherent instabilities....

[@MetalMachineShop]

Good point, there are pros and cons to both configurations. It’s interesting that the commercial 3-wheel scooter/motorbike market has gone down the tadpole route. A question worth asking is - why have three wheels at all and not two? Is the extra complexity of three wheels really worth it. I’m not sure!

[@VEC7ORlt]

​@@MetalMachineShop So I did a bit more googling while the comment was up, and it occurred to me that Velotilt, Mosquito, S'trike and Jetrike are essentially solving the same problem, but albeit a bit differently - all the joints and pivots are all around the place but in essence they are trying to achieve positive trail and as steep of a steering angle as possible. What I like about FWD delta - simple drivetrain. Regarding tadpole trikes - I think its safety - a non-leaning tadpole is much harder to roll, also with front wheels doing most/all of the braking - two is better than one. I'm personally fine with two wheels, but three is much easier to start riding on. i.imgur.com/44PV1G7.jpg This is my ride, I love it, but getting GF to ride it - nope. Now I'm thinking about a leaning delta i.imgur.com/kLhIG4p.png Something along these lines.

[@AlbertoDelunesAviernes]

Amazing!! Can you tell what program do you use?

[@MetalMachineShop]

Thanks! It’s Onshape. You can get a free account, highly recommended.

[@haydenseesyou]

would we use a vertical bellcrank like in your last examples if our steering input was perpendicular to the ground? (straight up like a scooter)

[@MetalMachineShop]

That sounds like it would work.

[@unknownarchon8860]

I can imagine it sharp conering with oposite steering on four tilting wheels 😀

[@creativecityis]

thanks for the video, much better understanding of the stearing... I'm also building a leaning trike. going with two rear wheels, current design has the rear wheels doing the castering action to avoid toe in\out... anyone see how this could be a bad idea? (would save me several months of building and testing if someone can point out why/how this wouldnt work) love the vids, didnt even think about the toe in/out until your last video

[@MetalMachineShop]

Thanks! I’d need to know more about your design to comment intelligently. If it’s a delta trike with the rear wheels doing the tilting only and both fixed straight ahead I don’t see any immediate problems - it’s been successfully demonstrated on other designs. Rear wheel steering would potentially be more problematic...

[@creativecityis]

@@MetalMachineShop the rear wheels would act like caster wheels and the front wheel is for steering. I might even work only by leaning... I'm integrating mechanics/pistons to hold it upright until the lever is moved causing one of the wheels to lower. Thought if rear wheels were independently following the directions due to the catering ability this would avoid needing to worry about toe in/out and wheels not steering appropriately for the curved radious

[@DylanDavidVindasLopez]

i do have a question, what if you wanted to control the tilt and the steering at the same time using the handlebars? how whould that work (basicaly if you turn handlebars left, vehicle should tilt left and wheels should also go left)

[@rogermaxwell5946]

Both controlled by one input gives hi lean at slow speeds in tight curves and low lean on fast speeds in large curves. This is opposite to what is actually required, (as in a 2 wheeled pushbike action.). However, for normal road use, the angle of lean to angle of turn can be set to give a workable lean/ turn ratio that gives satisfactory riding control overall..

[@hellopsp180]

I just thought of something Even more simple. with less complexity :D Whats the thought of literally just making a bicycle e.g. 2 wheels 1 front and 1 rear But having 2 stabaliser wheels that are only contacting the ground during "Park" + During leaning turns That way you can exactly keep your lean angles to what you set and never have to deal with this complexity?

[@johnmoncrieff3034]

defeatest!!

[@MetalMachineShop]

Yeah, there's a lot to be said for sticking to just two wheels!

[@YaR0MyR]

In the turn the inner wheel HAS to toe-out as it runs the smaller radius circle. There more, tilting should also provide toe-steer, if the tilting used for the turn performance and not for moving straight forward in lean position. And for ultimate performance (a.e. racing) the tilting should also be lesser for inner wheel.

[@xaytana]

The leaning version of Ackermann angles, for tilting vehicles, would more than likely affect steering Ackermann. You'd have to find the sweet spot between the two where they cooperate nicely, which is probably too much work for a single person creating a leaning velomobile. Plus Ackermann geometry is very well defined in how it functions, the leaning equivalent is not, if the leaning equivalent has even been properly researched at all. Nobody really knows if differential lean angles will have better performance, or if the difference in performance is worth the extra cost of design and parts, until it's properly researched, and it probably won't be properly researched until there is a motorsport that adopts leaning trikes and quads. In theory it should be better, but there's no realistic counterpart to that theory. This also comes before talking about tires, deflection, and contact patch. The scope of a mechanism that both steers and tilts an axle goes beyond that of bike and motorcycle -like designs, and thus doesn't operate in the same way. Both of these also come before how you implement suspension into this kind of design, which would affect lean angles during compression and extension, and thus would also affect steering angles. There's also the issue of incorporating powered axles into these, for tilting quads, for rwd delta trikes, and fwd tadpole trikes. I think differential lean angles is the last concern when discussing any of this.

[@MetalMachineShop]

The more you think about it, the more complex it becomes. Certainly too much for my small brain. I’m just hoping that these issues don’t make too much practical difference.

[@xaytana]

@@MetalMachineShop I know it's beyond the scope of your velomobile project, but I wonder how viable a disconnected steering system would be. Rather than have steering be linked to a singular moving piece, or a chain of pieces originating to a single part, what if each wheel was controlled individually? Like if two single-ended steering racks were used. It'd require a computer to control it, along with a table of each steering angle - arm length ratio for each lean angle, along with a way to accurately measure tilt. In theory, a system like this would provide correct Ackermann geometry for every combination of steering and tilt angles, and could be easily adapted to differential tilt systems. Differential tilt could be done similarly, adjustable length upper control arms controlled by a computer that's programmed with all of the correct geometry. Both systems are fairly simple in theory, just a lot of math, and it should be fairly simple to construct. Almost like a steer-by-wire and tilt-by-wire mechanism, except both sides of each system are discretely controlled. The first issue is the idea of a powered axle, as between steering and tilting, you'd have extreme angles that no universal joint could realistically handle. Unless someone develops a spherical gear system, which I doubt will ever actually happen as spherical geometry isn't a nice thing to deal with, hub motors are probably the only viable solution. Next issue, and the big issue here, is suspension on differential tilting. Sure you could program a computer to adjust the system to compensate, but the issue is if the computer would be fast enough to do so, and if the actual actuation method is fast enough to do so. Maybe add a hydraulic cylinder to work with the suspension's hydraulics, where when the suspension compresses, the tilt arm extends, inversely with suspension extension, to keep correct differential lean geometry correct, and this should reduce any lag the electromechanical system would have. But again, this is just a theoretical concept that should work on paper, but seems like it wouldn't be too difficult to implement in an actual vehicle. Suspension is definitely the most difficult puzzle piece in all of this to figure out. I just wish a motorsport division would get interested in leaning trikes and quads, just so we'd all have actual teams of engineers figure out the difficult parts for us, with the added bonus of ridiculous over-engineered systems that would never go into most road and track vehicles.

[@MetalMachineShop]

You probably could do a lot of this stuff in principle - or even in practice - but the question would be whether it gave you any advantage for the huge increase in cost and complexity. For a small, human-powered vehicle, lightness is everything and a simple passive tilt mechanism should be all you need. It’s hard to justify going from two wheels to three other than as a novelty and an intellectual exercise! Computer controlled steering wheels could be done relatively easily, but you would want a high degree of reliability to be safe, plus a continuous power source. Flat batteries would not be helpful!

[@denerpardo8528]

what CAD software was used?

[@MetalMachineShop]

Onshape.

[@usertogo]

@8:05 no good if wheel and suspension elements penetrate each other...

[@MetalMachineShop]

Haha true! You can get away with it in CAD but the real thing will need some adjustments to get everything to fit properly without clashing.

[@Pushyhog]

deplorable watching! Can ya educate us a lil’ on rear tilt??,,, skid and traction in turns. Channel is: pushyhog velosRus.com