ball bearing races - conductivity

andrewnummelin
Posts: 374
Joined: Mon Jul 21, 2008 9:43 am

ball bearing races - conductivity

Postby andrewnummelin » Tue Aug 24, 2021 4:20 pm

During a discussion on the virtual group https://www.scalefour.org/forum/viewforum.php?f=161 I discovered a problem with high resistance in the ball bearing races of a rolling road, a couple of bearings were effectively fully insulating!
As I have a project where I am considering the use of ball bearing races on rolling stock axles I was concerned about the conductivity as I am intending to use a split chassis approach for current collection where rolling resistance must be kept to a minimum. I therefore got out a set of bearings I had bought for a different project (that never got started) and I tested the conductivity of these. All showed variable figures between a few hundred to a couple of kilo-ohms which would not be appropriate for my planned project.
I was advised by a supplier to try using a contact cleaner followed by electrolube. This helped significantly but certainly did not completely cure the problems with the rolling road or my other ball bearing races.

Has anyone else found similar problems, and if so, how may they be cured?
Regards,

Andrew Nummelin

davebradwell
Posts: 1174
Joined: Fri Jan 18, 2019 3:48 pm

Re: ball bearing races - conductivity

Postby davebradwell » Tue Aug 24, 2021 8:41 pm

Part of a bearing spec - the long number it's ordered by - is for the lubricant so it will be pot-luck what you have in there. I'll assume they're not sealed brgs. Suggest you soak them in solvent to remove lubricant and hope, or fit shorting contacts like electric locos have to avoid spark eroding brgs with the high currents involved.

DaveB

Terry Bendall
Forum Team
Posts: 2420
Joined: Sun Jul 27, 2008 7:46 am

Re: ball bearing races - conductivity

Postby Terry Bendall » Wed Aug 25, 2021 8:41 pm

andrewnummelin wrote: I discovered a problem with high resistance in the ball bearing races of a rolling road, a couple of bearings were effectively fully insulating!


An interesting point. I have a rolling road which is not used very often but there have been occasions when the loco being tested does not run very well so this may well be the cause.

andrewnummelin wrote:As I have a project where I am considering the use of ball bearing races on rolling stock axles I was concerned about the conductivity as I am intending to use a split chassis approach for current collection where rolling resistance must be kept to a minimum.


I have no personal experience of building locos with split chassis but I have often operated those built by Barry Luck who uses this method. They work very smoothly with plain bearings and 1/8 axles. Personally I would have thought that using roller bearings was an unnecessary complication.

Terry Bendall

User avatar
zebedeesknees
Posts: 330
Joined: Mon Jul 12, 2010 5:15 pm

Re: ball bearing races - conductivity

Postby zebedeesknees » Wed Aug 25, 2021 10:05 pm

andrewnummelin wrote:
Has anyone else found similar problems, and if so, how may they be cured?

I have just checked my stock of ball races, 2mm i.d. and 1mm i.d. and those on the Hornby rolling road, and found that the majority read o/c in a static resistance test. In fact the one(!) with the least resistance read 2K and that was on the rolling road.
However, unlike Terry, I have never noticed any hesitation by locos on that rolling road when powering through the races. This suggests that a static resistance check doesn't tell the whole story, and the resistance within the bearing may reduce significantly when running under load.

Not that I would notice now, most of my traction being battery powered, and the rest on the way.

But I have to agree with Terry in that there is unlikely to be any great advantage in using ball races on driven axles, unless using a very low torque drive arrangement. Plain bushes have worked fine for decades, and given a suitable lube, conduct well enough too.
On the other hand, if the mechanicals in fitting ball races aren't too difficult, it might be worth a try.

Ted.
(A purists' purist)

davebradwell
Posts: 1174
Joined: Fri Jan 18, 2019 3:48 pm

Re: ball bearing races - conductivity

Postby davebradwell » Thu Aug 26, 2021 8:27 am

Looks like it would be more appropriate to test conductivity with battery and bulb (not LED) as there's just too much intermittent or light contact. It's worth asking yourselves if such aa rolling road is a reliable way of testing individual pick-ups.

DaveB

Terry Bendall
Forum Team
Posts: 2420
Joined: Sun Jul 27, 2008 7:46 am

Re: ball bearing races - conductivity

Postby Terry Bendall » Thu Aug 26, 2021 8:37 am

davebradwell wrote:Looks like it would be more appropriate to test conductivity with battery and bulb (not LED) as there's just too much intermittent or light contact. It's worth asking yourselves if such aa rolling road is a reliable way of testing individual pick-ups


I would normally use a test meter to check for continuity. My normal use of a rolling road is to test for free running and to help "run in" a motor or drive system. If I want to test pick ups then I put the loco in a cradle of some sort and apply a power lead to the wheels.

Terry Bendall

User avatar
Winander
Posts: 847
Joined: Thu Mar 20, 2014 12:19 pm

Re: ball bearing races - conductivity

Postby Winander » Thu Aug 26, 2021 12:33 pm

I googled and found that "standard bearing lubricant is non-conductive" and conductive grease is available. Apparently there is a problem with static electricity building up in assemblies containing bearings and this can cause arcing and pitting on the surface of the bearing. I would speculate that arcing is not a problem for us although it can be present in cars that use 12v, so it could be a thing. Further speculation leads me to assume that any lubricant found by us would be standard, cheap as chips and probably non-conductive.

I was surprised a bearing is not a good conductor and it seems the simple and effective way to test pick ups is Terry's method

Terry Bendall wrote:If I want to test pick ups then I put the loco in a cradle of some sort and apply a power lead to the wheels.


Given the efficacy of the brass bearings used almost universally and recommended here, I wonder if it's worth the trouble to build in a poor pick up arrangement and the topic has been added for discussion at our next virtual meeting.
Richard Hodgson
Organiser Scalefour Virtual Group. Our meeting invitation is here.

davebradwell
Posts: 1174
Joined: Fri Jan 18, 2019 3:48 pm

Re: ball bearing races - conductivity

Postby davebradwell » Thu Aug 26, 2021 5:47 pm

I test individual pick-ups by running loco onto 2 sheets of paper placed on track as wheels must be at correct running height.

You're flogging a dead horse with your ballraces - they're never used singly and as part of an assembly are inevitably end loaded to eliminate any clearance. Here, with single bearings, you're trying to measure the resistance of a loose assembly that is touching occasionally. At least with loco in place there's some loading so better contact. It's looking too closely at something that normally seems to work - somehow.

DaveB

andrewnummelin
Posts: 374
Joined: Mon Jul 21, 2008 9:43 am

Re: ball bearing races - conductivity

Postby andrewnummelin » Mon Aug 30, 2021 11:59 am

Thanks to all for the comments and information.

To expand a little on my thinking, my original thoughts were to help the free running of a loco chassis in the dim and distant past when I still worked in OO and at that time I had no thoughts about doing anything other than phosphor bronze strip wiper pickups. Now I'm working on a couple of steam railmotors that provide some modelling challenges:
- single driver (so no quartering problems at least)
- rather open motor compartment so not easy to hide a conventional gearbox & motor, hence thoughts go towards an axle hung layout (as described on the CLAG website)
- to avoid wiper pickups acting as brakes on trailing wheels, a split frame chassis could be a good idea
- for inside bearings ball races should offer much lower rolling resistance than conventional brass bushes (possibly needed with a tiny motor driving only one axle and a trailing bogie with inside frames)
- in split axle locos I've not seen reports of poor conductivity through the bearings - presumably using conventional light oil lubrication - so I'd expected ball races to perform adequately as well
- horrific valve gear.

So, I'll probably do the following:
1. build basic underframe first
2. build 2 non-powered split frame bogies, one with ball races and the other with simple brass bearings (hopefully a single design that can take either)
3. check free running of the bogies
3. add weight to typical level, test electrical pick-up
4. try to get far enough on to attempt assembling the valve gear the next time I get to a Missenden event. I'm sure I'm going to need help with this!
Regards,

Andrew Nummelin

davebradwell
Posts: 1174
Joined: Fri Jan 18, 2019 3:48 pm

Re: ball bearing races - conductivity

Postby davebradwell » Mon Aug 30, 2021 3:24 pm

Regarding valve gear suggest you just copy the layout of the prototype as in distance from loco centre line. I appreciate the original drg may be lost but would the arrangement on a Rheidol engine work? By copying every joggle and offset it's bound to fit and work sweetly - the original draughtsman's done the hard work - whereas trying to design it on the job from a 4mm scale drg would be a nightmare. I always drill a hole in the frames to co-incide with the expansion link pivot and this helps when assembling the brackets - an idea pinched from the live steamers. If you use forked joints ensure forks are deep enough.

DaveB

Stephan.wintner
Posts: 109
Joined: Sun Mar 15, 2020 11:04 pm

Re: ball bearing races - conductivity

Postby Stephan.wintner » Mon Aug 30, 2021 8:58 pm

Speaking of the brass bushes - I believe much of their friction comes from misalignment of the pair to the axle centerline, and the edge of the journal rubbing on the axle.

Care paid to concentricity, easing the edges (rounding / polishing - not simply deburring or chamfering) and selecting thinner ones (axially) should get the friction down. Polishing the axle may help.

Another source of unwanted friction will come from whatever locates the axle transversely (axially). Ensure that gap is also between smooth faces, round any offending edges, etc.

We engineers usually load ball bearings so they carry a touch of axial load, thus locating the shaft both radially and axially. Not sure that's practicable here, but if so that would offer the lowest friction.

Iaian Rice pointed out that having a round compensation beam resting on an axle gave very low friction - perhaps that gives you ideas ?

Stephan

andrewnummelin
Posts: 374
Joined: Mon Jul 21, 2008 9:43 am

Re: ball bearing races - conductivity

Postby andrewnummelin » Fri Sep 03, 2021 9:52 am

Dave & Stephan,

Sorry for the delay in responding to your posts they contain very useful comments and suggestions. I hope to be doing some experiments in the next month or so. I’ll aim to report after the Missenden weekend.
Fortunately I have copies of the original blueprints of the underframe and power unit of the rail motor, but i’m not going to try to scale down some of the clearances shown... Great shame that a GA of the body does not appear to have survived - only rough diagrams.
Last edited by andrewnummelin on Sat Sep 04, 2021 7:15 am, edited 1 time in total.
Regards,

Andrew Nummelin

davebradwell
Posts: 1174
Joined: Fri Jan 18, 2019 3:48 pm

Re: ball bearing races - conductivity

Postby davebradwell » Fri Sep 03, 2021 10:18 am

Final thought (possibly) - file/machine face of wheels flush with tyre and, if yoy can, reduce tyre to scale 1.85 width. With our 2mm wide tyres and boss on front you're already losing the clearance battle. The drg is a big help here.

You don't have to scale the clearances but components might have to be "rounded down" to the nearest material thicknes below scale size. A notoriously tricky bit is getting suffficient swing on the combination lever while the 2 forks at the top wrap round each other.

DaveB

best33
Posts: 61
Joined: Thu May 12, 2011 12:05 pm

Re: ball bearing races - conductivity

Postby best33 » Mon Oct 18, 2021 7:37 am

Damage to bearings from electric currents is well known in industry from circulating currents between the static and rotating structures. We therefore have to take great care to ensure bearings have no current passing through them.

In simple terms you can think of a bearing electrically as a capacitor and resistor in parallel. The rolling element and the tracks are the charge carrying surfaces and the lubricant is the dielectric. As the bearing rotates it pushes a wave of oil ahead of the rolling element and the tracks which creates a localised pressure keeping the two apart. The thickness of the oil film and the area will change the capacitance. As load is applied to the bearing the oil film is squeezed out becoming thinner. At the same time the oil film on the unloaded side becomes thicker. So the capacitance varies as the bearing rotates. In a rolling road scenario the load only comes from the weight of the model so the gaps are constantly changing.

Ball bearings are manufactured with radial and axial internal clearance. If there is no axial load or pre-load the bearing will be located only on the balls in the direct loaded direction. This could be only 1 ball when static possibly more when rotating. The application of an axial pre-load to take up the axial clearance will mean all of the balls are in contact with the track which will lower the electrical resistance. The problems come when the oil film totally insulates the ball from the track which can lead to discharge sparks when the capacitor breaks down which will locally melt the surface of the metal at the point of discharge. At a low level this will give a frosted appearance to the ball and track and at worst a pitted surface which will affect the smooth running of the bearing. At relatively low DC currents this probably won’t be too much of an issue but the lack of pre-load will give a variable resistance as the bearing rotates. Conventional Split axle pickup with plain bearings relies on the contact between the bearing and axle. This will always be a single point contact. When the model is static the point of contact will be at top dead centre and when the model is running the contact point will be at a resultant vector between the vertical force and the tractive force. If there is any oil in the bearing it will act to insulate the gap. We probably get away with it on models because the of the low rotating speeds and relative slop in the system.

Personally for rolling roads I think it is better to provide power to the model directly rather than rely on the contact through rolling road bearings.

Mark Humphrys

User avatar
zebedeesknees
Posts: 330
Joined: Mon Jul 12, 2010 5:15 pm

Re: ball bearing races - conductivity

Postby zebedeesknees » Mon Oct 18, 2021 3:12 pm

best33 wrote:Damage to bearings from electric currents is well known in industry from circulating currents between the static and rotating structures. We therefore have to take great care to ensure bearings have no current passing through them.


Excellent info <snipped> thank you Mark.

Personally for rolling roads I think it is better to provide power to the model directly rather than rely on the contact through rolling road bearings.

Mark Humphrys

Not just for rolling roads, the split axle idea also draws current through the loco bearings. It is still somewhat surprising to me, reading the postings on here from members who struggle with track power wiring, pickup reliability and short circuits 'somewhere', that so many modellers still seem to believe that the traditional way is somehow easier than fitting a battery and a radio receiver in their locomotives.
Ted.
(A purists' purist)


Return to “Chassis and Suspensions”

Who is online

Users browsing this forum: ClaudeBot and 0 guests