Sequential Transmission Shifter Torque?

[stanifm1]

Hi everyone,

I am a 2nd Year Mechanical Engineer working on a gear shifter (Kliktronic) for Aston University's FS team 2015. Some of you may have seen user ChristianChalliner on the forums, and I am working alongside him on the car.

The query I have is: does anyone know of the dynamic gear shift actuation torque for a Honda CBR600rr transmission? I have measured the torque to be 6.18Nm at rest (6kg on a 105mm arm) with +-14 degrees swing for up/downshift. Is there any way of finding this out without starting the engine up and measuring it on the move? (engine not currently operational).

I will be needing this to order some centring springs for a mock-up jig of the assembly, which the Electronic engineers who are making the shifting Brain need to test their circuitry under loads and different situations such as failed shifts.
I have designed the jig and all that needs now is the springs, for which I am struggling to find a force to find the right ones.

Regards, Michael

Aston Racing 2015
2nd Year Mechanical Engineer

[Jay Lawrence]

I'm pretty sure this has been answered on here already but the gist is that: yes you can work it out, but you need quite a bit of knowledge. You need to know the applied tangential load on the shifting dogs and the coefficient of friction at that interface, plus the ramp angle on your shifter barrel, plus the masses of the components you are trying to move, plus... plus.... etc. It might be a good idea to ask other teams who have done the experiment with a running engine. Alternatively, find someone who has a CBR600RR bike and test it. Given Christian's posts about keeping the car very simple, I'm surprised you are adding this complication.

[Kevin Hayward]

Michael,

I wouldn't put too much hope in getting great development on a rig built to model the engine. It will really only help determine whether there are any major issues with load capacity. In that regards make the rig a little too stiff, and don't expect that it will ever be a great representation of what the real force will be like.

The majority of your testing should be on a running engine. Dyno's are a good start, but remember your driver may be shifting under any combination of throttle, rpm, and rear wheel grip. Your shifting system needs to work under a wide range of conditions without fault. One missed shift in an endurance event will usually outweigh the small time advantage gained over a mechanical system.

I have the great fortune (misfortune) of having been involved (at varying levels) with shifters adapted to motorcycle engines that include hydraulic, solenoid based (actually voice coil based), and pneumatic. For any team, beginner or advanced, I would recommend pneumatic as the way to go. Easy to cobble together something off the shelf, pretty quick, easy on the gearboxes at a variety of pressures, and easy to hook up to something like a motec, without added electronics.

The gains from going the hydraulic or electrical route just aren't significant, especially for all the added headaches. Where-as pneumatic is so straight forward I would be happy to make the argument that it can be easier than getting a decent mechanical linkage working well.

Once basic sizing is done for the actuators (as per the experiments you have already performed) the only testing required for flawless operation was starting at the low end of expected pressure and increasing until shifts worked every time. Main limitation will be 1st to 2nd gear (and back) without hitting neutral. Have driver clutch on downshifts, the button routed through the ECU calls for ignition cut on upshifts. Don't bother having a system to find neutral and make sure you have a good enough engine tune around idle to be able to start reliably with the clutch is in. This has some clutch drag compared to neutral, but nothing a decently tuned engine couldn't start against time after time.

Overall any shifting system like this is best developed on an old car until you can prove it works. Quite simply it is a mistake to try and get it to work first time on an unproven car where your competition is at stake. If you have an available old car that is working then your electronics guys can use that as a guinea pig and test rig (which will be far more accurate than a spring contraption). If you don't then the team might be best served by putting it off as a project for next year.

Kev

[Kevin Hayward]

By the way when I mentioned off the shelf for pneumatics I did not mean going to a company that sells these kits for motorcycle engines. This path is too expensive and doesn't work any better than putting together your own "kit".

What I meant was sourcing the individual components from someone like festo. Solenoid valves, simple two way aluminium pistons, CO2 canisters (Soda stream ones work well), regulator (adjustable first time out), standard air hoses and valves.

It is a very simple pneumatic circuit, and as long as you don't try to over-complicate it (as many of the motorcycle shifting systems do) then it is pretty easy to get working.

Kev

[Jay Lawrence]

The system Kev mentions is pretty much what we had in 2007. It was simple enough and worked very well on an old car, but then played up at comp on the new car, and there were a bunch of missed shifts etc. Also, you will have to package extra stuff, and the cylinders will need refilling, and 1-N-2-N-1 shifting + clutch actuation is a pain. If you're going to go down this road, I'd suggest modifying your shifter barrel so that you block out 1st, and your gearbox becomes N-2-3-4 (you shouldn't need 5th or 6th either, so can save some rotating mass there).

[Kevin Hayward]

Jay,

The ECU guys just had a system that skipped neutral everytime, could not use the system to get the car into neutral. Worked great. Cylinder refills at Kmart (or equivalent) being only sodastream bottles. Clutch actuation was switched to pushing down a hand lever rather than gripping anything. Very easy to actuate. Just a quick bump during downshifts. The cars shift without the clutch on downshifts but its not as nice.

Kev

[mech5496]

Michael,

I will sound a bit like Z (no idea if this is bad or good), but knowing your teams' background from Christian, you would be better off with a simple mechanical linkage; as in matter of fact 90% of the teams should use those. I will give you some reasons:

Way simpler-> Less weight, less parts to design/build, less parts to (potentially) break
(Way) cheaper
Probably as fast as a pneumatic/electronic one

Note that I come from a team that was obsessed with electrical/pneumatic systems, and we have tested almost everything available, including a cordless drill motor. In the end:
a) a tested car will be a faster car
b) a working mechanical linkage will be faster than a non-working electric/pneumatic shifter

[stanifm1]

*As it currently stands*

From what I hear from last year, the shifter was successful, despite the number display not working (it was just hard-wired into paddles on the wheel instead). We have a team of dedicated electronics guys who have the system working. So I'm not keen to negate out all their work (they've made a ton of progress).

I suppose if the worst comes they can just keep their
Gear display and rev display. At the moment they have their project interfaced with both ECU and Kliktronic control box. They can just ditch the Klik if it's too much hassle.

I can still make the test rig as it's for their benefit really as I think their project is assessed.

Today is crunch day- The chassis is getting sorted so I'm sure Christian or I can update once today is over.

[Jay Lawrence]

hmmm sounds like you have a management problem. There should be no 'electronics team' per se; there is your team and a bunch of members who contribute. There is a lot for sparkies to do that is much higher on the list than getting a shifter working: reliable engine management, data logging (especially vehicle dynamics (wheel speed, steering angle, etc.)), design preparation, etc.

Incoming anecdote: we have some brilliant sparkies on the Wollongong team, but unfortunately when I was there some of them spent huge amounts of time on a customised driver display. It kinda worked on it's own (like your shifter sub-project), but in the car it was completely meaningless because all the driver has time to notice is a big f*ck-off STOP THE CAR OR IT WILL EXPLODE light.

Anyway, as suggested before I think your best bet (without a running engine) is to go and see another team with the RR engine and do the test there.

[stanifm1]

I should elaborate, they are working on a complete ECU - So dashboard -revs and warnings lights etc- mounted on wheel/dash. I have made a bracket to test with the potentiometer they provided - it's surprisingly overkill for the use but that's what the management gave them to work with. I suggested acquiring a smaller method of measruing when the Shifter actuates but they wish to continue with the current setup. The bracket I have made has been received well, but need improvement.

It basically modifies the stock Kliktronic bracket to include the potentiometer without changing the mounting bolts. However a problem with the stock bracket is the need to completely disassemble the kliktronic to remove it from the car, so I have been suggested to make a better way of mounting it to the car. Appologies for lack of pics, Currently not accessible to CAD.

[ChristianChalliner]

Yeah just to add to this a little,

We already have the kit (bought kliktronic) which we've had for a few years. The electronics team is dedicated to that because it is strictly speaking not an FSAE project but a marked project for their course so they are confined specifically to that project, no opportunity to get them to wire the car, plus, it's of little interest to them since most of it is basic compared to the requirements of their project.

The reason Michael asks this is mostly my fault, I remember reading somewhere that someone on this forum said that it took something like 10 times the static force to change gear under high load, unfortunately, I can't find that post anymore and I also might have misread it.

Thanks,

Christian

P.S Big F'off warning light is planned :)

Michael, I like the bracket you made it's a neat way to modify the old one, just wish they hadn't insisted on using that enormous potentiometer!