hydraulic gear change?
Discussion
Just thinking about the guts of my scratch build, decided I cant be doing with organising 101 links n levers to get the gear change around the engine right to the back of the car, but I cant stand the sloppy nature of cable, so am thinking hydraulics instead now.
Obviously with a bit of effort and a handfull of brake cylinders I could cobble something together, just wondering if there are any "pre-prepared" solutions out there that anyone can recommend, preferably that aren't going to cost more than the rest of the car combined!
Cheers
Obviously with a bit of effort and a handfull of brake cylinders I could cobble something together, just wondering if there are any "pre-prepared" solutions out there that anyone can recommend, preferably that aren't going to cost more than the rest of the car combined!
Cheers
If it is a sequential box (eg. motorcycle), then as Mikey_G says, there are off-the-shelf electro-mechanical solutions available using solenoids, but I'm not aware of any for a conventional H-gate box.
Allan Staniforth tried hydraulic remote change with his original Mini-engined 'Terrapin' racer back in the late '60's though, but abandoned it 'cos it was too unreliable. He eventually came up with a rod linkage instead.
Not saying that it can't be done - in theory there is no reason it shouldn't work - but expect a bit of development and a few teething troubles.
>> Edited by Sam_68 on Saturday 22 April 11:08
Allan Staniforth tried hydraulic remote change with his original Mini-engined 'Terrapin' racer back in the late '60's though, but abandoned it 'cos it was too unreliable. He eventually came up with a rod linkage instead.
Not saying that it can't be done - in theory there is no reason it shouldn't work - but expect a bit of development and a few teething troubles.
>> Edited by Sam_68 on Saturday 22 April 11:08
The MR2 and Toyota engined Elise/Exige actually have very good gearchanges. Cable changes are difficult to get right, though, particularly in a mid-engined installation. The Rover engined Elises vary from nasty to just about acceptable, and that's with Lotus' engineering expertise behind them!
Trouble is, their main problem (lack of precision) would seem on the face of it to be even more difficult to resolve with hydraulics.
It's not a sequential, engine and box are a audi v6 bi-turbo (RS4 if I can find one) with the 6 speed 2wd transaxle.
I think cable *can* be OK but the only time I've tried/majority I've seen have been pretty awful.
With the accuracy of the hydraulics, as long as I work the throw's out correctly in CAD then it shouldn't be too tricky - I know thats being heartily optimistic and will require some faffing but my thoughts were, at a very basic level... have 2 brake cylinders either side of each throw, two for fore and aft and two for side to side. A cage fully bearinged cage on the bottom of the gear lever then means that either pair of cylinders will only see direct throws rather than angle. This will give a one-way only requirement on each cylinder rather than trying to use a single one for push and pull. Once the cage is made, as the fluid is totally uncompressable 1cm gear lever movement will equal 1cm at the other end reliably? Well that's my theory anyway!
I think cable *can* be OK but the only time I've tried/majority I've seen have been pretty awful.
With the accuracy of the hydraulics, as long as I work the throw's out correctly in CAD then it shouldn't be too tricky - I know thats being heartily optimistic and will require some faffing but my thoughts were, at a very basic level... have 2 brake cylinders either side of each throw, two for fore and aft and two for side to side. A cage fully bearinged cage on the bottom of the gear lever then means that either pair of cylinders will only see direct throws rather than angle. This will give a one-way only requirement on each cylinder rather than trying to use a single one for push and pull. Once the cage is made, as the fluid is totally uncompressable 1cm gear lever movement will equal 1cm at the other end reliably? Well that's my theory anyway!
What sort of force/travel are you expecting? Brake/clutch hydraulics are designed to carry relatively large forces over a short distance. They're also designed only to apply a load not hold an accurate position - they can afford to be flexible and imprecise and fluid volume changes over time can be ignored. I assume you'll want to minimise the backlash if this thing is going to be manually operated, are you going to have some sort of spring preload to take out the slack? All in all it sounds quite a challenge and I would have thought that a simple mechanical linkage would be far easier to implement and work better.
Theory sounds fine and I agree that twin cylinders side-side and fore-aft would be the way to go, rather than expecting a single cylinder to both push and pull.
Not sure about the detail of why it didn't work for Staniforth: he just mentioned in his book 'High Speed, Low Cost' that the idea had been abandoned in favour or a rod change after extensive effort had failed to get the hydraulic change to function adequately.
Two potential problems spring to mind:
1) Whilst the fluid is incompressible, it does expand as it heats up. This might cause the 'linkage' to go out of alignment.
2) Brakes and clutches don't need give such large movement with accuracy. Brakes barely move (pad clearance distances only), and clutch forks can be moved though biting point, so don't need any real accuracy. The drag of moving 8 hydraulic pistons (4 x master + 4 x slave) and their working fluid might make the change feel heavy or sluggish and the cumulative play in the mechanical parts of the linkage might be enough, in total, to make the change imprecise.
Still, the worst that can happen is that you find it doesn't work and have to engineer a cable or rod alternative, so why not give it a go?
edited to add: like wot GreenV8S just said more clearly in half as many words!
>> Edited by Sam_68 on Saturday 22 April 20:42
Not sure about the detail of why it didn't work for Staniforth: he just mentioned in his book 'High Speed, Low Cost' that the idea had been abandoned in favour or a rod change after extensive effort had failed to get the hydraulic change to function adequately.
Two potential problems spring to mind:
1) Whilst the fluid is incompressible, it does expand as it heats up. This might cause the 'linkage' to go out of alignment.
2) Brakes and clutches don't need give such large movement with accuracy. Brakes barely move (pad clearance distances only), and clutch forks can be moved though biting point, so don't need any real accuracy. The drag of moving 8 hydraulic pistons (4 x master + 4 x slave) and their working fluid might make the change feel heavy or sluggish and the cumulative play in the mechanical parts of the linkage might be enough, in total, to make the change imprecise.
Still, the worst that can happen is that you find it doesn't work and have to engineer a cable or rod alternative, so why not give it a go?
edited to add: like wot GreenV8S just said more clearly in half as many words!
>> Edited by Sam_68 on Saturday 22 April 20:42
GreenV8S said:
What sort of force/travel are you expecting? Brake/clutch hydraulics are designed to carry relatively large forces over a short distance. They're also designed only to apply a load not hold an accurate position - they can afford to be flexible and imprecise and fluid volume changes over time can be ignored. I assume you'll want to minimise the backlash if this thing is going to be manually operated, are you going to have some sort of spring preload to take out the slack? All in all it sounds quite a challenge and I would have thought that a simple mechanical linkage would be far easier to implement and work better.
I was going to address the necessary force/travel issues by choice of piston diameter at either end once I have accurate figures of what's needed - this is something I've done before with hydraulics and quite easy to work with.
Spring pre-load... unsure yet, that's one of the things to work out as I go along
I think I'm going to try a mock up with some old cylinders if I can find which box they are in! Backlash should be virtually zero with a push-push system though. fluid volume should just be a case of sorting the resevoir out adequately I think? I agree a simple mechanical linkage would probably be easier, but I'm not convinced on working better. At the end of the day if the hydraulics turns out to be cack I'll have to go to mechanical though
Sam_68 said:
it does expand as it heats up....
The drag of moving 8 hydraulic pistons (4 x master + 4 x slave) and their working fluid might make the change feel heavy or sluggish and the cumulative play in the mechanical parts of the linkage might be enough, in total, to make the change imprecise...
Some good point there actually!
GreenV8S said:
fluid volume will change over time though - mainly with temperature but also with water absorption etc. You need some way to accomodate this.
yeah this is a good point. Obviously the reservoir would take of the majority of the problems but the absorbtion/temp changes could be an issue.
Interestingly, i've just been out and measured the throws on a gearbox outside - gate through at the long end of the actuator lever is only about 1/4 in so not a problem there!
Davi said:
yeah this is a good point. Obviously the reservoir would take of the majority of the problems but the absorbtion/temp changes could be an issue.
Um, if I've understood what you've got in mind that won't work, you have to have all the circuits permanently closed off from the reservoir (constant volume operation) in order for a push/push system to work.
If hydraulicswere going to give a better outcome regardless of cost, labour, etc then Porsche would have used them on the Carrera GT - instead it went for an upsate of the cable system with some form of flat metal band I believe. No other maker of mid-engined cars uses hydraulics unless in robotised forms, and I believe even the great Gordon Murray, reputedly "obsessed" with a "rifle-bolt" gear change quality, chose cables for his F1. Your system, while interesting, sounds heavy. Brake cylinders are built for much higher loads, and are you going to have 8 hydraulic lines to route and bleed and trouble shoot? Use the opportunity to add lightness somehow...
the boxter S or any 996 has cable shift I think, so go ta a garage and pretend you want to buy..... or a PH meet as these are excellent times to test fit a performance car... I had a guy who wanted to buy a Boxter ask me to sit in it at a PH meet earlier this year. No real PH will turn you down.
Alternativelly, I have the porsche cables for my GTT but you'd have to come to me to try them out.... you'd be welcome if your local tho.
Basically they are 5mm solid resonably flexible bars in a cable sheeth.. so they have no play push or pull.
Alternativelly, I have the porsche cables for my GTT but you'd have to come to me to try them out.... you'd be welcome if your local tho.
Basically they are 5mm solid resonably flexible bars in a cable sheeth.. so they have no play push or pull.
Vacuum? Cruise control servos and a vac reservoir off an Esprit? Whichever way you go, you could use limit switches to tell the control circuit when the ram/ servo is fully home, then if there's any creep it'd be self-regulating.
Esprits incidentally use a combination of cable and rod; they are usually OK until the link bushes wear or the cables start to stiffen up.
Esprits incidentally use a combination of cable and rod; they are usually OK until the link bushes wear or the cables start to stiffen up.
Gassing Station | Kit Cars | Top of Page | What's New | My Stuff