Gearbox Forward-vs-Reverse Efficiency: How Much Difference to Expect?

With the right gear set, it will be the same on both directions.

You will need to determine the cut of the gears first.

The manufacturer will be able to supply this information, during the design phase, if you ask them instead of trying to do unnecessary calculations yourself.

It's all in the design...

http://www.railway-technical.com/trains/rolling-stock-index-l/bogies.html

But, the African Queen was a boat.

Thanks for all the responses. I'm no closer to an answer to my question though. It is not practical to expect a gearbox manufacturer to provide a lot of internal design details regarding any their hundreds or thousands of standard product combinations, especially to a remote buyer of a small fractional hp item!

My role lies in advising a small company which is executing an order for some equipment where these drives are required. The question is more out of my own curiosity and for discussion purposes. There seems to be a lot of information on efficiency and back-driving of worm gears, but I was hoping to tap into something readily available regarding commonly-used helical or bevel-helical reducers having 3 or 4 stages.

There may be information which can be teased out of results for back-to-back testing of gearboxes -- the total losses recorded will presumably be greater than the losses for a single unit operating as speed reducer for the same power transfer.

I suspect that with high ratios the unit may not even start in reverse direction from a cold condition, but that is not a requirement at all. I am essentially trying to determine the stopping distance, which can vary depending on whether the car is empty or partly/fully loaded. The mechanical response is beyond the control of the plc circuit where the ramp-down time from running frequency can be set. Instead of experimenting with the settings at the customer's site during final commisioning, it will help if I have some idea of what surprises to expect. I don't wish to introduce any complications of load-sensing or feed-back control -- only the stopping position should be within a predictable zone of a few mm after a stop signal is triggered by a limit switch. Slow deceleration/acceleration is necessary to avoid any jerking of the fragile load.

All this can be considered food for thought/discussion, and further responses are welcome. Thanks.

Yeah good luck with that, too many constantly changing variables....

http://www.machinedesign.com/archive/second-look-gearbox-efficiencies

Correction - missing word: the third para in my previous post #6 should read:

"There may be information which can be teased out of results for back-to-back testing of gearboxes -- the total losses recorded will presumably be greater than TWICE the losses for a single unit operating as speed reducer for the same power transfer."

@ T sq. Your question as posted is impossible to answer correctly.

How fast is your "slow Speed"

What weights are involved? some freight Trains run to thousands of tonnes.

What do you consider a controlled rate of deceleration, stopping a Train weighing 100 tons will be a lot different to one weighing 10000 tons?

Will the Train be running on level tracks or is there a decline/ incline invoved?

Also you mention potential damage to a fragile load, what kind of load on a railway truck would be subject to damage through braking efforts. Any load would normally be packaged and secured to prevent movement and damage through transport and handling and Trains do not usually stop so suddenly.

...also I would add that this gear/motor mechanism should be used in tandem with a mechanical braking system....imo

http://www.knorr-bremse.hu/en/railvehicles/downloadservices/productflyers/brake_systems/brake_systems_1.jsp

https://en.wikipedia.org/wiki/Regenerative_brake

https://library.e.abb.com/public/b53e94802c75e541c125789a00290729/Railvolution02-11-Traction-transformer-Kiruna.pdf

https://en.wikipedia.org/wiki/Railway_brake

180 Watts!? That's 10% smaller than the motor in my household coffee grinder. These cars better not have a mass more than a few tens of kg, even with VFD and a 480:1 ratio. Any gear system with a ratio that high is going to have considerable friction.

Here are some thoughts:

There are 3 basic categories of gears: Meshing gears with parallel axes, meshing gears with intersecting axes, and meshing gears with axes neither parallel nor intersecting.

If mating gears are in the same plane (axles parallel), then the gear teeth should mainly push and not slide sideways. Mating parts of the teeth will be moving parallel in the same plane when engaged. (Spur gears)

https://en.wikipedia.org/wiki/Gear_train

With non-parallel axes, there are several types of gears:

http://www.meadinfo.org/2008/11/gear-efficiency-spur-helical-bevel-worm.html

The teeth of the bevel gear (intersecting axes) can be moving parallel to the main gear teeth and there should be no slippage. The hypoid gear can have some sideways slippage, and the worm gear a lot of slippage.

I'm thinking that slippage (sliding friction) will make a gear train to not be equally efficient in both directions. When two surfaces are sliding against each other, the amount of frictional force is proportional to the normal force. The amount of friction is determined by the direction of the driving gear tooth (line of action) with respect to the mating surface. If this is different for the two gears, then there will be more friction in one direction than the other.

Taking the worm gear as the extreme example, if it is driven by the worm, the line of action (direction of driving tooth) is at a small angle to the mating surface. If it were driven by the gear, the line of action would be almost perpendicular to the mating surface. (In this case, the frictional force exceeds the force imparted to the worm, making it impossible to drive.)

So, my thinking is that if your transmission should be about equally efficient if it consists of nothing but spur gears and bevel gears.

Obviously I haven't stated the problem clearly enough! Most of the details do not concern the basic question which I had asked, but I should try to address the doubts of those who have taken the effort to respond meaningfully.

The car(s) in question are not running on a regular railroad track, but in a factory environment with very limited movement range of a few meters. They only need to be moved in and out of a firing kiln for ceramics. There is also a transverse trolley which can move the car itself piggy-back to another parallel location. The mechanisms by which these are achieved are not relevant to this discussion. The loaded car weight would be about 10t, and the trolley self weight around 3t. There are different combinations of speeds and loads involved, in the range of 1.5 to 6 m/min. With a fully loaded car the acceleration/deceleration may have to be of the order of 0.1 m/s².

There are existing designs of such equipment with rather different sizes and loads, but details are not available. In one that I know of the drive is through an encoder motor and central computer control for the movement and precise positioning. That would be overkill for the present requirement which can be achieved using standard available products.

I don't expect anyone in this forum to try solving the whole design problem for me! I merely wanted some feedback on whether the reverse efficiency of a high-ratio helical gearbox is substantially lower than in the forward step-down direction. My post #8 may provide some ideas, but I have no practical exposure to back-to-back testing. Maybe someone in this forum has some relevant data on test results which can provide clues.

Thanks for all the responses so far.

I guess it helps to be there...

OK, you're backing a heavy load out of the oven so another vehicle can move it to another oven, and it has to stop at the right place to get picked up. You have a limit switch to shut off the motor and you need to know how far it will coast before it stops. Is this accurate?

If your gearbox has a very high reduction ratio, it may very well use a worm gear drive. In this case, you do not want to reverse drive it with the inertia of your load or something will break. Maybe the vendor of the gearbox could tell you if it can be reverse driven.

IMHO, it would be much safer to have a position sensor to control power to the driving motor to bring it to a stop at the desired location.

A little high-school physics would be a good starting point here.

F=m*A; 10,000kg * 0.1=1,000 N (Force required to accelerate 10T @ 0.1 m/s², ignoring friction).

6m/min= 0.1 m/s

P=F*V; 1,000N * 0.1m/s =100 Watts (Power required to continue that acceleration @ 0.1 m/s , again ignoring friction).

So IF your friction forces total less than 800 N, the 180 W motor should theoretically be able to do the job...

I agree with Rixter's analysis of the gearing possibilities.

BUT! You say you are moving 10T of car plus ceramic in and out of a kiln. If the load is placed in the kiln while the kiln is cold, and not removed from the kiln until it has cooled to near ambient temperature, then these speeds and accelerations may be acceptable. A car that can hold 7T of ceramic must be around 2 m long, so it must move more than that distance to get in and out of the kiln. At an acceleration of 0.1 m/s², it will take a full minute for the unit to reach a speed of 6 m/s, during which time it will have travelled a distance of 3 m (Just outside the kiln). If the kiln is any where near hot, that's enough time to transfer a lot of heat into the motor, gearbox, etc. That heat could easily destroy the motor.

I say you'd better plan on a MUCH larger motor! It is true that motors are less efficient when working well below full load, but this motor will only operate intermittently, so efficiency is presumably of little concern. With a VFD, a larger motor can be controlled at whatever speed/acceleration is required, but your 180 W motor can barely achieve the 0.1 m/s² acceleration. If you then decide you need to accelerate faster, you have no option but to replace the motor with a larger one.

As we now know the purpose of the machinery involved and the fact it will be subject to extremely high temperatures (Ceramic Kilns operate around 600°C to 2000°C) so no electric Motor /drivetrain will take that heat so you need a remote power plant and a Transmission System.

The wagon should be a heavy duty steel construction with solid axles / wheels. The powerplant should be located away from the heat source so you really need a Transmission System like wire and pulleys or chain drive to actually move the wagon in and out of the Kiln. I have visited several Ceramic production facilities and they use similar Systems.

That way you can use the motor as a brake in both directions.

I think everyone is getting too involved in the design of the equipment! The mechanical equipment is interesting in itself, but already done and being manufactured, with little scope for any change. Just to clarify matters, there is no drive on the car which is a simple bogie with refractory bed, around 2m square. There is a pusher-puller arrangement mounted on the trolley which can reach out by over 2m to engage the axle of the car through a pawl mechanism to move the car to/from the kiln, or even to the opposite side, at more than one location where rails are provided. The firing cycle may last a few days, and there is no exposure of the machinery to kiln temperatures. But yes, the car or trolley movement is necessarily a long slow process.

There must be lots of such systems in operation, but I have seen only one which had central computer control for all plant operations. We are convinced that ordinary geared motor drives will serve our purpose, and with variable frequency control we should be able to set the proper frequencies and ramp-up ramp-down times at least by trial and error at the time of commissioning. My interest as a mechanical design engineer is to arrive at some tentative values by calculation beforehand. If needed the speeds can be reduced. The response of motors to VFD control seems a bit tricky but manageable. Even the gearbox reverse efficiency may not be relevant at all, but since the question popped up during a deceleration check I thought I'd put it out in this forum.

I'm still hoping to get some useful feedback on my basic question.

Because of the variations in gearbox design and the lack of manufacturer performance data, the only way you will be able to make an estimate is to either disassemble and evaluate the gear drive yourself or set one up and test it. Any other assumptions you make are just spitting into the wind.

Quite. And what does efficiency matter at <...0.18kW...> alongside a hearth (rhetorical question - NNTR)?

"Rolling Scenarios

Dynamic rolling is simulated. The operator downloads the mill setup, starts the mill and initiates charging billets from the reheat furnace. The simulator displays motor speeds and currents as the billet threads through the mill. The operator applies cascade speed trims to any of the stands and observes the changes to motor loads due to push-pull interstand forces. He is forced to react to events as if encountered in the normal operation of a mill."

http://seseng.com/automation-products.html

I guess it's best to wind up this discussion. Though I'm no closer to an answer regarding my original question, a number of interesting points have been thrown up, and my thanks to all those who have participated.

Some comments, possibly off-topic:

My interest in such topics often tends to be academic. I don't subscribe to the attitude that (for instance) if the efficiency of a small 0.18kW drive is irrelevant in proportion to the larger problem it should be ignored. In my career (I'm past 70 now) I have found it stimulating to seek out the whys and hows of technical problems, however trivial they may seem. Colleagues invariably found my attention to detail quite exasperating, but my experience has been that the devil is nearly always in the details. An understanding of how gear losses occur may be unnecessary in the present case, but could prove useful in some other situation which could be handled with more confidence, especially if time or resources are limited, which is very often the case in real world engineering. One gains more confidence by learning from mistakes (one's own and others') if we take the trouble to analyse the reasons honestly, even if it turns out to be a design error or oversight.

With regard to the problem on hand, I vaguely recall something I read about gear tooth profiles being designed differently for step-down and step-up gears. Probably it had to do with how tooth friction is distributed between arc of approach and arc of recess along the line of action. Intuitively I have the feeling that it is like the difference between pushing a stick along the ground and dragging it behind you -- though the change in friction effects may be numerically negligible in both situations, considering the forces and angles involved. I think what I read was that for a speed step-up drive the tooth profiles and correction factors should be selected with more of the action after the pitch point. Whether efficiency is a criterion or not (rather than strength or lubrication issues) I just don't know, as I have never tried designing or analysing a step-up drive during my career of dealing with essentially one-off designs. That's why I thought there may be useful data in results of back-to-back testing which is routinely done by manufacturers of serial gearboxes, but I have no access to such information.

More specifically, my doubt was regarding the deceleration of the drive when switched off, and the load inertia may be trying to drive the motor for a brief while (depending on the inertia distribution and losses along the drive train) when it will be acting as a step-up drive of very high ratio! Under such conditions, I would like to estimate the stopping time and distance using meaningful data, however tentative or inaccurate, rather than pure guesswork which is always possible. I had tried to state that in the original post, most of the discussion veered off into other design aspects.

I'm also pretty hopeless at dealing with computers and internet, as I have remained rather stubbornly old-fashioned in trying to work things out from first principles wherever possible, rather than believe whatever results are thrown out by mouse-clicks. I still keep a slide rule handy though it's seldom used now. Obviously it has left me rather isolated, with almost no one I can engage in a meaningful technical discussion! It is also partly the reason I discontinued participating in the CR4 challenge questions a few years ago -- simply could not keep up the pace. Also there was a change in the type of challenges. There used to be a lot of interesting discussions about mechanical topics like bouncing balls, weighing of coins, mirror reflections, ..., with insightful (also humourous) inputs from the likes of Fyz, Kris, Del the Cat, and many others.

My apologies for the long-winded rant (assuming anyone had the patience to actually read through it!). I'll look in again to see if there are any further responses, but otherwise thanks again to all those who have contributed to this thread.