Calculating worm drive thrust force

[Quarryscapes]

Potentially putting a very expensive Maxon into a 16mm NG loco and would like to get the maths done before selling out for something that will potentially wear out fast. So I need to calculate the axial force along the motor shaft. Assuming that it's run to it's stall torque of 9.44mNm through a 0.4MOD Worm Set of 30:1 ratio, what sort of force is that going to be putting along the motor shaft?

 

I have found various formulae online, but they require a lot of data about worm gears that I just don't know! 

[AndyID]

Potentially putting a very expensive Maxon into a 16mm NG loco and would like to get the maths done before selling out for something that will potentially wear out fast. So I need to calculate the axial force along the motor shaft. Assuming that it's run to it's stall torque of 9.44mNm through a 0.4MOD Worm Set of 30:1 ratio, what sort of force is that going to be putting along the motor shaft?

 

I have found various formulae online, but they require a lot of data about worm gears that I just don't know!

 

I think you are worried that the motor's thrust bearings will wear out too quickly? If so, the stall end thrust does not have much to do with it and calculating the actual wear rate would require a lot more data than you have there.

 

If there is space you might solve the problem by putting thrust bearings at the worm and driving it via a universal of sorts. That will eliminate end thrust on the motor.

[RAF96]

Any way you could set up a representative gear train and measure the observed end thrust at the given torque.

It would obviously require some test kit capable of measuring down to those small values. I have no idea what such kit is available and if solid state or direct reading.

Rob

[peterfgf]

I think you are worried that the motor's thrust bearings will wear out too quickly? If so, the stall end thrust does not have much to do with it and calculating the actual wear rate would require a lot more data than you have there.

 

If there is space you might solve the problem by putting thrust bearings at the worm and driving it via a universal of sorts. That will eliminate end thrust on the motor.

A couple of thoughts:

 

1.  A shaft shouldn't have more than one axial location point (i.e.thrust bearing) unless there is a flexible coupling capable of accommodating axial movement between the two thrust blocks.  For example, the shaft in between the bearings might expand thermally and wreck both bearings.

 

2. If the motor is as expensive as you say, then surely the manufacturer will have used a high quality, precision bearing commensurate with the overall cost of the motor.  Good bearings, especially roller bearings, are expensive and you get what you pay for.  Why not ask the manufacture what they have fitted and what the expected life (number of rotations) versus load is?

 

Peterfgf

Edited April 17, 2018 by peterfgf

[brigo]

2. If the motor is as expensive as you say, then surely the manufacturer will have used a high quality, precision bearing commensurate with the overall cost of the motor.  Good bearings, especially roller bearings, are expensive and you get what you pay for.  Why not ask the manufacture what they have fitted and what the expected life (number of rotations) versus load is?

 

Peterfgf

 

The cost of a motor does not necessarily relate to how it is designed to be used. The Maxon motor is of a coreless design and most manufacturers of this type state they are not designed for axial loads. In most cases the end load in one direction is taken by the retaining circlip rubbing on the face of a sleeve bearing.

 

Brian

[34theletterbetweenB&D]

I would make a very 'wet finger' simple estimate of the running maximum axial force on the motor shaft with 30:1 gears, at between one tenth and one fifteenth of the maximum drawbar force, an easily measured figure.  

 

If you already have a model of the same tractive capacity you are aiming at, then the drawbar force it exerts is a useable value for this purpose. From there it is simple action and reaction, and if we assume the loco will require much the same force to move itself as it exerts at the drawbar, a fifteenth of the drawbar force must act axially on the motor shaft. That figure will increase with losses in the drive train.

[PatB]

A very rough, rule of thumb type number would be possible by taking the motor stall torque (taking this to be the maximum torque exerted by the motor), dividing by the radius of the worm (to get the lateral force at the periphery of the worm) and multiplying by the worm and wheel reduction ratio (because you are effectively using this lateral force to drive a wedge with a slope determined by this ratio). Using the figures you give, and assuming a worm diameter of 6 mm, I get (9.44/6)*30 = 47.2 mN. Because this method is very approximate, I'd then add a factor of safety of at least 2, so in this case I'd assume 100 mN end-thrust to be taken care of as a result of the motor driving the worm.

 

However, something that must also be taken into account is the potential effect of a heavy train and a falling gradient, which could easily put much greater forces into the drivetrain than the motor could manage on its own.

[brigo]

Is the drive arrangement simply a motor with a worm on it, driving a worm wheel on one of the axles ? If so, would not the greatest force be when the loco is unpowered and being pushed along the track. The force on the motor would be that required to push the loco (weight x coefficient of friction) times the ratio of driving wheel diameter to the worm wheel diameter.

 

Brian

[AndyID]

Maxon brushed motors seem to be available with ball bearings or sintered bronze bearings. I don't think there should be a problem if the motor has ball-bearings.

[Jol Wilkinson]

Ball bearings aren't necessarily designed to deal with end loads. The may be offered in Maxon motors for greater reliability, handling greatest axial loads, etc.

 

If in doubt over end thrust capacity of the motor bearings, I would suggest, as others have, using a gearbox with integral thrust bearing capability. Mike Sharman used to produce a machined brass casing gearbox and the original Exactoscale range had a built up version with pressed housing components, but neither are now available other than S/H occasionally. Are others available?

 

The High Level motor mounts, at least the couple I have built, include an optional outer bearing support for the motor shaft. Used with some spacing washers it could be possible to restrict the end float of the shaft and take up the load with the bearing.

[AndyID]

Ball bearings aren't necessarily designed to deal with end loads. The may be offered in Maxon motors for greater reliability, handling greatest axial loads, etc.

If in doubt over end thrust capacity of the motor bearings, I would suggest, as others have, using a gearbox with integral thrust bearing capability. Mike Sharman used to produce a machined brass casing gearbox and the original Exactoscale range had a built up version with pressed housing components, but neither are now available other than S/H occasionally. Are others available?

The High Level motor mounts, at least the couple I have built, include an optional outer bearing support for the motor shaft. Used with some spacing washers it could be possible to restrict the end float of the shaft and take up the load with the bearing.

The static and dynamic axial loads are available at the Maxon website.

[DCB]

Proper model loco motors used to have ball bearings to take the end thrust, the problem is they seem to have ceased production with GR Wrenn.

I would agree that a separate gearbox and some form of sliding coupling is probably the best idea if you must use a worm, however  would dsagree about a single thrust bearing.  A pair of thrust faces is fine as long as there is some end float even when hot

 

Personally I would seek to avoid a worm if I possibly could.

[Quarryscapes]

A very rough, rule of thumb type number would be possible by taking the motor stall torque (taking this to be the maximum torque exerted by the motor), dividing by the radius of the worm (to get the lateral force at the periphery of the worm) and multiplying by the worm and wheel reduction ratio (because you are effectively using this lateral force to drive a wedge with a slope determined by this ratio). Using the figures you give, and assuming a worm diameter of 6 mm, I get (9.44/6)*30 = 47.2 mN. Because this method is very approximate, I'd then add a factor of safety of at least 2, so in this case I'd assume 100 mN end-thrust to be taken care of as a result of the motor driving the worm.

 

However, something that must also be taken into account is the potential effect of a heavy train and a falling gradient, which could easily put much greater forces into the drivetrain than the motor could manage on its own.

 

Puts it well within the capabilities of the motor which is good for 2.2N dynamic axial load, but would be pushing the sleeve bearing model with its 0.8N rating.