Pnuematic Shifting

[mikalo28]

Hi all.
My name is Michael, I'm from the university of QLD Australia. I'm studying Mechatronic engineering.
I'm doing my 4th year thesis project on the Pneumatic Shifting system for the uni's fsae car.
I am searching for articals and academic papers on the topic.

I'm having a very hard time finding anything.
If anyone has come across any articles on the topic could they please let me know.
Or Any other ideas on where to find more information?

Thanks , Michael



More About My Project:

The system will comprise of:
- a High Pressure Air bottle,
- valves
- pneumatic actuator
- gear feedback sensor
- Circuit board.

Up and Down shifts will be initiated by the driver via paddles on the steering wheels.

By programming an Atmega Microcontroller, I hope to achieve the following functions:
- full control of the valves (which will enable half shifts into neutral)
- Auto Up shift (at optimum rpm) (can be switched on and off by the driver)
- Gear position feedback
- neutral finder button

[mikalo28]

Hi all.
My name is Michael, I'm from the university of QLD Australia. I'm studying Mechatronic engineering.
I'm doing my 4th year thesis project on the Pneumatic Shifting system for the uni's fsae car.
I am searching for articals and academic papers on the topic.

I'm having a very hard time finding anything.
If anyone has come across any articles on the topic could they please let me know.
Or Any other ideas on where to find more information?

Thanks , Michael



More About My Project:

The system will comprise of:
- a High Pressure Air bottle,
- valves
- pneumatic actuator
- gear feedback sensor
- Circuit board.

Up and Down shifts will be initiated by the driver via paddles on the steering wheels.

By programming an Atmega Microcontroller, I hope to achieve the following functions:
- full control of the valves (which will enable half shifts into neutral)
- Auto Up shift (at optimum rpm) (can be switched on and off by the driver)
- Gear position feedback
- neutral finder button

[Corey H]

I hate to be the first to say it, but you might need to search a little harder. Paradigm Motorsports supplies Air shifters for some teams, an extremely simple system with loads of data logging to prove that it works when set up correctly.

[TimR]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">gramming an Atmega Microcontroller, I hope to achieve the following functions:
- full control of the valves (which will enable half shifts into neutral)
- Auto Up shift (at optimum rpm) (can be switched on and off by the driver)
- Gear position feedback
- neutral finder b </div></BLOCKQUOTE>

There is a guy on our team who has made a Hydraulic shifter for our car (Sydney uni, Aprilia 550). We've also got a pneumatic one adapted from last years car, but the guy doing it has talked up the greater control hydraulics bring- with pneumatics the valve is on or off but hydraulics can adjust speed of opening (apparently- this is not my thesis). He's got a pretty good controller for it too, with an engine RPM sensor, and he's planning on doing a launch control and auto upshift function.
His one has a netural switch and a clutch button(if you want to engage it seperate to the gearshift, or shift 2 gears for example), and the paddles that shift up and down. And he has a gear position display but I think that was an easy bit.
I don't know what he read but he has put a heap of work into it. I don't think there is any one spot he got his info off, he worked alot out, researched hydraulics by speaking to people and just by trying it.

Just a word of warning- have a backup system- last years sydney uni team thought that the new pneumatic shifter would be so good that there was no need to make a backup manual shifter. Subsequently it didn't work and we had a bad comp.

[Luniz]

Building a basic pneumatic system is fairly simple: What we are using is basically a pneumatic cylinder and a 5/3way electromagnetically actuated valve off the shelf. As a pressure supply we are using a CO2 bottle from a soda maker with a throttle valve to keep the system pressure at about 12 bar. Shifting into neutral is not possible with our system, a possibility to do so would be a second cylinder with half the travel, with which you can shift into neutral from first gear. Auto upshift should be possible via your ECU, most systems are able to give out some sort of blip when the engine reaches a specific rpm. Gear feedback is done via counting up-and downshifts with a syncronisation everytime you change from 1st to 2nd and pass the neutral position, which has a sensor in the engine already.

What I am trying to say is that you might be thinking too complicated from the start on. I'd rather build something reliable and simple in the first place and then try to make it more trick.

Cheers, Lutz
Raceyard Kiel
Suspension Design

[jrickert]

TimR,
Where are you getting hydraulic pressure from?

[TimR]

The guy is running a 12v pump and reservoir. he is running an accumulator so that the pump is not always on- he estimates that you can get 15-20 shifts from the accumulator before the pump needs to recharge it.
I know this and the hydraulic lines will add some weight but if it works well then the weight is worth it for quick, reliable and controllable shifts. The launch control and quick shifts will make up for the weight gain, in conjunction with increased reliability.
Besides, all our weight saving initiatives are unsprung weight this year- Al. box frame uprights (450g each), CF rear rims, and Al. driveshafts with spring steel couplings (900g per driveshaft). But we'll see how these go- we have proven backups.

[dazz]

I've wondered about using the engine oil pump? Use an accumulator (accu-sump?) to avoid losing engine oil pressure during shifts? Possibly mount the cylinders internally. I haven't done any calcs on what cylinder diameters would be required for typical engine lube pressures, though.
You'd want to be pretty confident in the system too, as a gear system failure could well cause an engine seizure.

[Kirby]

I once thought about hydraulic-electric shifting. I figured a 3rd stage on a dry-sump pump using a separate reservoir of mineral oil or similar and some sort of small accumulator.

Waterloo in 2005 had a manual hydraulic setup, with some fancy custom made cylinders, much love went into the design of said system.

[mikalo28]

I started the project last semester so I have already done the majority of the work.

The circuitboard and program have been designed on breadboard, there are a few minor bugs that I need to fix but apart from that, I just need to draw up the schematics and submit the PCB design.

My superviser is suggesting that I look for more academic articles (for the final thesis report) on the topic instead of just using information from various websites and forums.

So, now im searching for past theses, magazine aricles or other academic papers, not just clips of information from here and there.

So yeh, if you know of anything like this that I could gain access to I would greatly appriciate it.

[Steve Yao]

The Waterloo manual hydraulic shifter was the most impressive thing I remember from that comp.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Kirby:
I once thought about hydraulic-electric shifting. I figured a 3rd stage on a dry-sump pump using a separate reservoir of mineral oil or similar and some sort of small accumulator.

Waterloo in 2005 had a manual hydraulic setup, with some fancy custom made cylinders, much love went into the design of said system. </div></BLOCKQUOTE>

[Kirby]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Steve Yao:
The Waterloo manual hydraulic shifter was the most impressive thing I remember from that comp.
</div></BLOCKQUOTE>

Not to buff up the egos of those involved in Waterloo in '05. But that was a really well executed car. I suppose the results showed.

I spent plenty of time having a look over that car learning about how it was put together. Definitely from a different age, one without templates and roll-hoop clearance rules.

[Fischer]

I have also been searching for more information about pneumatic shifting, but didn't really succed until yet.
Following url may be usefull for you:

http://dspace.mit.edu/handle/1721.1/45820

[mikalo28]

Nice http://fsae.com/groupee_common/emoticons/icon_smile.gif
This is exactly what I was looking for,
Thanks for the link.
I'll be happy to share my work when it is finished.

[Fischer]

Fine!
I am glad I was helpfull. If I find anything else I'll write it here, too.

[Charlie]

Auburn successfully ran a pneumatic shifter 1999-~2005 (still similar I think but I don't know fully) without any circuit board of any kind, the only logic was an ignition cut signal to the ecu whenever a button was pressed. I agree with others, you have way overcomplicated the issue. We even had automatic shift ability with an analog output. Still no separate circuit or logic.

As for papers have you checked out 2004-01-3554, not a ton about the shifting system but some info about application of it.

[ibanezplayer]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Charlie:
As for papers have you checked out 2004-01-3554, not a ton about the shifting system but some info about application of it. </div></BLOCKQUOTE>

You might want to do a quick google of "2004-01-3554" before you buy it http://fsae.com/groupee_common/emoticons/icon_wink.gif

Charlie's a generous guy.

[JeffreyH]

I disagree regarding overcomplicating the issue. The point of FSAE is to learn. Part of that is designing circuits. I appreciate that you were a pure ME student (by the looks of your paper), but saying "don't bother designing the circuitry to an EE or Mechatronics engineer is like saying "Oh, don't worry about designing the uprights, we'll buy some". You ARE using circuitry in your datalogger, just some that someone else designed and sold you.

[Charlie]

The point is to learn but you shouldn't overcomplicate design in order to learn either. Now I suppose if you want to design a very sophisticated setup for your academic project, knowing it isn't going to improve the car performance, that's a good enough excuse. Just hope the added complexity doesn't cause additional issues.

Our basic design used no circuitry in the ECU except for ignition retard. Any design will have to use that. Additional circuitry is in my experience unnecessary.

[JeffreyH]

I agree, a simple solution is almost always best. The upright analogy wasn't exactly the best either... maybe using the stock exhaust from the bike, say, assuming it fits in the car but isn't ideal.

Regardless, I know I'd rather try to do it myself first and use the ECU as a fallback. Otherwise what's the point of doing FSAE on the electronics side of things... setting up the wiring for an ECU? Pretty boring to be perfectly honest. There's got to be an element of design.

[Charlie]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by JeffreyH:
Regardless, I know I'd rather try to do it myself first and use the ECU as a fallback. Otherwise what's the point of doing FSAE on the electronics side of things... setting up the wiring for an ECU? Pretty boring to be perfectly honest. There's got to be an element of design. </div></BLOCKQUOTE>

Personally I would not want to be working on a system that doesn't add car performance or help ensure reliability. If I didn't think this particular system was beneficial I would not enjoy creating it. I think there are a lot of areas in the electronics group that could be effective and a good use of resources. The rules are pretty open. But the first order of business would be a rock solid reliable electronics package, something that many teams seem unable to accomplish.

[JeffreyH]

True, but you can design your own electronics and make sure they're reliable as well... We design about 4-5 custom electronics modules which live on the car and we've yet to fail to compete based on unreliable electronics, though admittedly there were a few plan Bs which were implemented. Again, expecting your electronics team to just be wire monkeys and implement someone else's system is ridiculous. We're studying to be engineers not technicians. It's always worth trying to design your own system, if it fails make sure you have a plan B and learn from the experience. The exact same thing happens with non-electronics things on FSAE cars also.

Apologies for going off topic a bit - I'm interested to see the results on the UQ car come Decemeber 10th.

[Charlie]

I agree with you that a student designed system is not inherently unreliable; in fact a system that someone knows well can be more reliable due to better understanding of what can happen. However unnecessary complexity is never a good thing. If your team runs a relatively modern ECU they do not need extra circuitry to use pneumatic shifting.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by JeffreyH:
Again, expecting your electronics team to just be wire monkeys and implement someone else's system is ridiculous. </div></BLOCKQUOTE>

I never said anything of the sort; but on my team I expected everyone to work together and have the same goal. Having teams go off on projects for their own entertainment or skill building is not constructive for the goals of the team. Like I said I think electronics group can take on projects that are skill building AND make a better car. This example is not one of them as basically you don't need the complexity to achieve the same performance. Go find something else to improve and there are no shortages of that given the FSAE rules being so open.

I guess we have a difference in philosophy, and you are not alone. Many people like you use FSAE as a reason to experiment and learn to better themselves. They are not so worried if they are contributing to a better car or not. I see it as a test for real-world engineering, where everything should be done for a reason and people need to work together to achieve a goal. I think that is why OEMs value FSAE experience, because it is viewed as not a classroom experiment but a real-life test of cooperation and understanding of priorities.

[Fischer]

Hi,

our team will finally try to use a pneumatic shift, but we have problems when it comes to calculate the capacity of the bottle.
As the gas leaves the bottle, it suffers an expansions from 6MPa 0.6Mpa, so it's temperature will fall. But how can we calculate this difference of temperature?
We have been advised to search the temperature at which the entropy will be the same then it was when beeing in the bottle (and pressureof 6bar). Doing this, the temperature would drop until about 200K.

Does anyone know if this result is correct? The consumption of CO2 would be very big.

Thanks

[jd74914]

I don't think that this can be considered perfectly isentropic expansion because the change is very large and hence non-reversible. Isentropic expansion is really only valid for small changes (ie: small expansions by diffusers).

I may be wrong, but I believe this process should be considered adiabatic since the expansion happens so quickly that, at least theoretically, no heat is transferred.

That said, when using some adiabatic compression formulas: T2 = T1 * (P2/P1)^((k-1)/k) where k is the ratio of specific heat yields a ridiculous answer. Assuming P1=6000kPaA,P2=600kPaA, T1=293K and k=1.31, I'm getting T2=170K which is definitely wrong.

In the end the error comes down to the fact that CO2 is a real gas and the formula derivation assumes to to be ideal, but I have no idea how to elegantly take that into account. You could simply do a bunch of iterations while changing the ratio of specific heats.

Jim D.
UCONN Racing

[tinkeringIdiot]

CO2 is very definitely not an ideal gas. There was a prof at my school that would throw things at people that assumed as such on thermo homework.

[jd74914]

Nope. CO2 is about as far from ideal as you can get.

[Chris Allbee]

jd# hit it on the head. Iterate. I suggest a spreadsheet or some matlab action.

[Fischer]

I have also read somewhere what jd74914 says. But instead of using the ideal gas formula, I thought I could use a table with properties (pressure, temperature, u, s, h) of CO2. The one I have is not good enough, as not all values appear. But by extrapolating I also reach a result close to 200K. I thought I was misunderstanding something, but if you get the same, then I don’t know where the mistake could be.

Chris Allbee, how can I use Matlab to get the T of CO2???

[TimR]

Why are you using CO2? Why not use compressed air? Thats what the previous sydney uni teams have done

[aerowerks]

The reason for using CO2 rather than compressed air is that you can keep CO2 liquid and thus have greater amount to work with.

[AJoshi]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by aerowerks:
The reason for using CO2 rather than compressed air is that you can keep CO2 liquid and thus have greater amount to work with. </div></BLOCKQUOTE>

I think that "having a greater amount to work with" is really not a good justification for using CO2. That's like saying we will use a 5 gallon fuel tank instead of a 2 gallon because it gives us more to work with. What you need to consider is the downside of using CO2 (canister weight, inherently more dangerous, special fittings required, handling procedures, potential liability, special filling stations, etc) to the potential upside of having more capacity/pressure. IMO, unless you really need the ~1800 psi that liquid CO2 provides (assuming you're using a standard 9 or 20 oz paintball size tank), there is really no advantage to CO2 as opposed to compressed air to pull or push a shift cable. If you really need that kind of pressure, why not use Nitrogen (3-5kpsi), since after all, more is better?

EDIT: grammar

[Fischer]

But, as far as I know, most teams use CO2 rather than compressed air. Honestly, we are taking CO2 because it is what we have mostly seen. It may be more expensive and difficult to design, but we don't have any air compressor, so we would rely on shops or other supplier anyway.

[Chapo]

We have used compressed air in the past. We have a paint ball container on the car and we fill it from a SCUBA tank. The SCUBA tank gets filled at the diving place and there is a fairly common adapter that you can get to go from the SCUBA tank to the paint ball tank. Really quite a simple system and the SCUBA tank lasts for a long time before it needs to be refilled.
This year we are giving completely electric shifting a go, pingle shifter internals in an Al case, as it saves a fair amount of weight over the airshift we were using, and there is one less consumable on the car.

[m6paiva]

Im a first year FSAE member designing a pneumatic shifter also.

Ive heard all the talk about needing special hosing and fittings for c02, but cant find any answers to why. I understand the pressure requirements, but with a tank mounted regulator, why cant you use a normal air hose and fittings?

[Fischer]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">t needing special hosing and fittings for c02, but cant find any answers to why. I understand the pressure requirements, but with a tank mounted regulator, why cant you use a normal a </div></BLOCKQUOTE>

I think there is no problem in using normal air fittings al long as it meets the requirements. But I don't know more than you, I am in the same situation than you.

What I don't know is why nobody uses an air compressor. Ok, it weights about 2.5 kg but this is just 1.5 kg more than an air bottle. We are considering this option, although we have to take care about the battery. But teams without this problems could use it.
I can't find the reason why nobody does so...