Hypothetical overdrive question

[willowbilly3]

I have posed this on a couple other forums and thought I would shoot it at you guys. My query is why the automakers have overdrive. Please read on before you respond. First of all I want to make sure you understand when I say "overall" ratio I mean the combined transmission and rear end ratio in any given gear. Lets use a manual 5 speed with .80 overdrive with a 3.55 rear end. So with my improved transmission you would have the same overall ratio in each of the 5 speeds but 5th would be direct or 1:1 and we would use a 2.84 rear end (do the math, that's the same final drive rpm as the .80 overdrive and 3.55)
Would it not be more efficient to have a direct power path in top gear instead of the overdrive that has to route the power through 2 gearsets? And also isn't the higher rear end ratio more efficient with less parasitic drag? In an age where automakers will trade their offspring for ounces off the weight or minuscule reductions in drag it would seem they missed something here.

[averagef250]

Direct has absolutely no less parasitic loss than any other gear including an overdrive. All parts in a manual spin all the time regardless of the gear your in. Direct is stronger than any other gear since there is no side loading of the input or mainshaft.

In theory your 2.84 gear works the same as your 3.55 with over, but it doesn't. Acceleration is different and loads on drivetrain parts are different. It will drive and behave worse without overdrive.

A for instance would be the first gen dodge diesel trucks. The 727 auto trucks use 3.07 gears only. Any lower and you won't make freeway speed. The gutbag and A518 auto trucks had 3.55's or 4.10's as options. The trucks with overdrive accelerate much better than the 727 trucks.

Think about converter slip. in a 727 truck, 500 RPM of converter slip is considerably more road speed differance than the same slip with lower gears.

[Hawkrod]

You make good observations but where you do not have enough information is torque versus rear ratio. The numerically lower rear ratios require more power to get them moving which is why they use overdrive with a mid 3.00 ratio. It is a very fine balancing act between producing useful power and maximizing mileage. A car with a 2.50 gear and a 1-1 trans ratio will use more fuel and be slower than a car with a 3.50 gear and overdrive for the same overall final ratio. I realize this may not be intuitive but that is how it works. Also note that the higher the rear is numerically the more likely you are to break something due to torque multiplication (I realize that is not related but my industrial design instructor used to love proving it! LOL). Hawkrod

[willowbilly3]

Dustin, I'd have to call you on that, sure the gears are always spinning but the power is not being transferred through them in direct. I just can't understand how there is no power loss difference to drop the power path through a gearset as opposed to straight through. You even admit to more sideloading, which would mean friction and power loss. My hypothetical transmission comparison is a stickshift so torque converter loss and lockup is not part of this equation.

And Hawk, numerically lower gear ratios take more power to get them moving? I guess the drag racers had that all wrong.

So what I am reading is there is some magic mysterious juju that makes using overdrive and a lower rear end work BETTER than the exact same overall ratios with a 1:1? How does that change how torque is applied to the ground if everything else is the same? My question as posed is the same engine rpm in every gear with both setups, the engine has the same mechanical advantage over the same diameter of tire, the only thing that's changed is that we have moved some of the overall reduction from the rear end into the transmission and eliminated the power path having to drop through the overdrive gears in 5th.


One thing Hawk said I do agree with is the higher numerically rear ends might be weaker. I have always thought the taller gears with the big pinion look like they have much better contact with the ring gear. If you compare a 4.56 with a 2.73 you will see about twice as many pinion teeth making contact with the ring gear. Also there would be less twist on the axles. Of course the corresponding lower transmission gear would put a lot mote twist on the pinion too I suppose. But that is getting off on a bit of a rabbit trail.

[willowbilly3]

Just posted on my thread elsewhere. It seems Richmond Gear has the same theory.

http://www.richmondgear.com/07pdfs/RG23.pdf

[FLATBEDFORD]

I remember reading somewhere that BMW manual shift cars use the willowbilly gearing. I've wondered the same thing myself. In a typical rear drive setup, you would also have lower driveline speed, perhaps making it quieter with less vibration.

[averagef250]

WB, it sounds like you've already made up your own mind about this. Doesn't seam like you quite get the idea that driving is not steady state and there is a real world mechanical advantage to numerically taller gears.

I know you're referring to stick transmissions, but the converter slip is a prime example of the problem.

Really? Does the side loading of the gears at cruising speed and gear tooth face contact friction make for a recognizable parasitic loss? I think you will find this is not the case, any loss here is rediculously minute.

If you went to any OEM automaker and told them they need to redesign thier stick trannies to be more efficient they'd laugh at you.

Today, autos are more efficient than sticks for almost every type of driving. Especially so when you factory in the idot factor and the warranties automakers offer today.

A moron can do a heck of a number to a clutch/manual tranny and the rest of vehicles drivetrain and far less than desired mileage if they drive like a moron. A stick requires skill to drive for optimal mileage a new smart auto does not.

If you want to redesign a manual tranny invent a gear-gear CVT. You'd be a millionaire and save millions of barrels of oil every year.

[Hawkrod]

Dustin, I'd have to call you on that, sure the gears are always spinning but the power is not being transferred through them in direct. I just can't understand how there is no power loss difference to drop the power path through a gearset as opposed to straight through. You even admit to more sideloading, which would mean friction and power loss. My hypothetical transmission comparison is a stickshift so torque converter loss and lockup is not part of this equation.

And Hawk, numerically lower gear ratios take more power to get them moving? I guess the drag racers had that all wrong.

So what I am reading is there is some magic mysterious juju that makes using overdrive and a lower rear end work BETTER than the exact same overall ratios with a 1:1? How does that change how torque is applied to the ground if everything else is the same? My question as posed is the same engine rpm in every gear with both setups, the engine has the same mechanical advantage over the same diameter of tire, the only thing that's changed is that we have moved some of the overall reduction from the rear end into the transmission and eliminated the power path having to drop through the overdrive gears in 5th.


One thing Hawk said I do agree with is the higher numerically rear ends might be weaker. I have always thought the taller gears with the big pinion look like they have much better contact with the ring gear. If you compare a 4.56 with a 2.73 you will see about twice as many pinion teeth making contact with the ring gear. Also there would be less twist on the axles. Of course the corresponding lower transmission gear would put a lot mote twist on the pinion too I suppose. But that is getting off on a bit of a rabbit trail.

You misread what I wrote and your response is not correct. I said "The numerically lower rear ratios require more power to get them moving" to which you twisted and responded in such a way as to make it sound as if I had said the opposite and you responded "numerically lower gear ratios take more power to get them moving? I guess the drag racers had that all wrong.". There are no drag racers I have ever heard of that use numerically low gears, maybe you know some losers? All drag cars that I am aware of use numerically high gears such as 4.56 or 5.00 6.00 etc... to use a numerically lower gear set like a 2.00 gear set would mean no matter what you did you would lose. Numerically lower gears require higher torque values to operate. This is why chassis dyno results are affected by ratios. Take the same car with a 2.00 gear set and then test it again with a 3.00 gear set and the dyno will tell you that the engine is producing more horsepower with the 3.00 gear set although there were no other changes. This is a simple basic engineering principle and something you actually learn early in automotive chassis engineering classes in college. Hawkrod

[willowbilly3]

WB, it sounds like you've already made up your own mind about this. Doesn't seam like you quite get the idea that driving is not steady state and there is a real world mechanical advantage to numerically taller gears.

I know you're referring to stick transmissions, but the converter slip is a prime example of the problem.

Really? Does the side loading of the gears at cruising speed and gear tooth face contact friction make for a recognizable parasitic loss? I think you will find this is not the case, any loss here is rediculously minute.

If you went to any OEM automaker and told them they need to redesign thier stick trannies to be more efficient they'd laugh at you.

Today, autos are more efficient than sticks for almost every type of driving. Especially so when you factory in the idot factor and the warranties automakers offer today.

A moron can do a heck of a number to a clutch/manual tranny and the rest of vehicles drivetrain and far less than desired mileage if they drive like a moron. A stick requires skill to drive for optimal mileage a new smart auto does not.

If you want to redesign a manual tranny invent a gear-gear CVT. You'd be a millionaire and save millions of barrels of oil every year.

I haven't made up my mind so much as you didn't offer any proof to change it, or offer any tangible proof to why overdrive and lower rear to accomplish the same overall final ratio is more efficiently than 1:1 top gear and a lower numerical rear ratio to achieve the same top gear overall ratio. You went off on something that had no real connection to what I was theorizing by comparing some Chrysler automatics. This is a case where "Dustin says" just won't answer the question for me. One thing you said I do agree with is loads would be different on driveline components. The driveshaft would probably have a bit more load and the axles a bit less but well within the limits of either under normal use I imagine.

And Hawk, I appologize about that, You were right and I was thinking backward about lower gear, not numerically lower. I seem to be having a hard time getting anyone to understand a simple postulation and get answers to what would be half of the equation, not the whole idea and was apparently chasing my own tail a bit there..
I do understand all about what you just explained but I don't understand why it makes a difference to have a portion of the total reduction moved from the differential to the transmission. If I had a 3:1 rear end and a 4:1 first gear then switched to a 4:1 rear end and a 3:1 first gear it shouldn't take a bit more torque to move the same vehicle at the same rate. It would be a total reduction of 12:1 either way. I am still waiting for an explanation why it wouldn't. I just used those numbers for an easy example.

[fomocoguy]

One thing Hawk said I do agree with is the higher numerically rear ends might be weaker. I have always thought the taller gears with the big pinion look like they have much better contact with the ring gear. If you compare a 4.56 with a 2.73 you will see about twice as many pinion teeth making contact with the ring gear. Also there would be less twist on the axles. Of course the corresponding lower transmission gear would put a lot mote twist on the pinion too I suppose. But that is getting off on a bit of a rabbit trail.

I think you may have your answer right there. Bigger parts will wear less, and with more contact area last longer than small parts, let alone the physical strength of a larger rear gear. This way they can use a tall stout rear gear in, lets say, a one ton truck that will see a lot of abuse, then just use the same identical trans in a half ton pickup to save $$$$. Makes sense to me...

[Hawkrod]

snip

And Hawk, I appologize about that, You were right and I was thinking backward about lower gear, not numerically lower. I seem to be having a hard time getting anyone to understand a simple postulation and get answers to what would be half of the equation, not the whole idea and was apparently chasing my own tail a bit there..
I do understand all about what you just explained but I don't understand why it makes a difference to have a portion of the total reduction moved from the differential to the transmission. If I had a 3:1 rear end and a 4:1 first gear then switched to a 4:1 rear end and a 3:1 first gear it shouldn't take a bit more torque to move the same vehicle at the same rate. It would be a total reduction of 12:1 either way. I am still waiting for an explanation why it wouldn't. I just used those numbers for an easy example.

The difference between having the ratio in the rear end versus the trans is because of the change in direction. Moving the rotaional force from the propeller shaft to the axles requires a 90 degree turn and that is where the load changes. If the engine was mounted side to side and you did not use a hypoid or crown gear to change the direction this would not factor into the equation. The issue is overcoming the load or friction of the tires at the same time as turning the direction of power. Engineers have been working on this specific problem for a long time which is why we have so many front drive cars now. A live rear axle is a power eating device and as I mentioned before, it is a balancing act to find the right ratio to allow power to be transfered and still provide reasonable fuel consumption and durability. Hawkrod

[FLATBEDFORD]

I'm also thinking that if you put the reduction in the differential, the drive shaft itself handles less torque load. Possibly taking some of stress out of a more vulnerable and perhaps weaker link in the drive train. Kinda like of you use a multi part block and fall. With only, say, 100 lbs of force and the single hand line, a person can lift many 100s of pounds depending on how many parts the system has, or what its final drive ratio ends up being.

Here's a page about block and falls;
http://en.wikipedia.org/wiki/Block_and_tackle

[willowbilly3]

That sounds reasonable Hawk. What is the source of the info? Do you have any side by side dyno test results to show the difference between the numerically higher gears and numerically lower? I am interested to know the actual amount of this difference in a chronicled lab setting. the only real difference I can see outright is the added friction and drag from more tooth contact in a lower nemerical ratio.

[Hawkrod]

That sounds reasonable. What is the source of the info? Do you have any side by side dyno test results to show the difference between the numerically higher gears and numerically lower? I am interested to know the actual amount of this difference in a chronicled lab setting. the only real difference I can see outright is the added friction and drag from more tooth contact in a lower nemerical ratio.

If you had only asked me 30 years ago! LOL When I was an automotive engineering student we had tons of that stuff and I am sure you can find it easily enough online but I don't have much left from school other than the diplomas and a couple of binders of special stuff that probably will never get used unless I start teaching again. Here is a great page on Model T dyno work and specifically it has a section on how axle ratio effected it. note that teh effects on this example are exagerated due to the fact that vehicle in question is built with archaic technology, produces relatively little power, and has an unusually high drivetrain friction power loss ratio. The point is really not the vehicle but the dyno results with only an axle change. It is true that most modern chassis dyno's are adjusted for ratio but teh reality is that there is still a difference and it can be very noticable:

http://clubs.hemmings.com/clubsites/MTF ... torque.htm

Also note that you will find people who will tell you that because the tire size and axle ratio are accomodated for in the dyno program that there is no difference but real world testing always proves otherwise. If you do not think so, then next time you have a chance to get a dyno test, take a much larger or smaller pair of tires with you and see what the difference it makes. The tires will not have as a dramatic effect as axle ratio does but it is fun to see what the dyno does. Hawkrod

[FLATBEDFORD]

Hawk,
I think you are missing that we are talking about maintaining the same final drive ratios, but using taller rear gear ratios and shorter gears in the transmission. We all know that you need lower (numerically higher) ratios for performance. The article talks about changing rear end gears, but not changing the transmission gears at the same time. In the article they are changing the final drive ratio. Willowbilly is talking about keeping the same final drive ratio, but using higher (lower numeric) gears in the rear and lower (higher numeric) gears in the transmission, thus maintaining the same final drive ratio, but not having an overdrive top gear, but a 1:1 top gear.