Replacement Athearn Bombardier BiLevel Trucks, aka making the train roll

[Dwight in Toronto]

Craigolio, thanks for the warm welcome.  I am indeed in Toronto.  Well, Markham, to be more precise. In fact, I’m only 20 minutes from Rapido’s facility, so I had them scallop the I-beam under The Canadian coaches to improve rollability. It helped a lot, but I’ve since had to do lots of other tweaks.  They roll much better, but still not as smooth and effortless as my Kato coaches, so I remain vocally critical, mainly because of the Rapido price point - premium cost for mediocre performance just rubs me with perpetual annoyance!

In non-covid times I frequent several annual train shows, and would very much like to meet you guys.  My guess is that the Barrie show might be the first opportunity to do so, hopefully February 2022 … fingers crossed!

Within the last three days, I’ve been exploring DecoderPro for the very first time. Question- where do folks find those little loco icons that I see in their JMRI rosters?

[daniel_leavitt2000]

Clearly Rapido found more problems with micro roller bearings than they solves as they went with a solid bearing on the Comet cars.

I wish I could locate a good source for jewel bearings as they have solved the same issues in the watch industry for over a century. Obviously ordering a bearing custom made for this would be insanely expensive, but I wonder if two small jewel surfaces would work. This would essentially replace the wire bearing surface on your trucks with either a flat edge or rounded bearing.

This wouldn't conduct electricity, but I'm more concerned with just getting a sound equipped Kato F40PH to pull 5 or more cars.

[craigolio1]

Clearly Rapido found more problems with micro roller bearings than they solves as they went with a solid bearing on the Comet cars.

I wish I could locate a good source for jewel bearings as they have solved the same issues in the watch industry for over a century. Obviously ordering a bearing custom made for this would be insanely expensive, but I wonder if two small jewel surfaces would work. This would essentially replace the wire bearing surface on your trucks with either a flat edge or rounded bearing.

This wouldn't conduct electricity, but I'm more concerned with just getting a sound equipped Kato F40PH to pull 5 or more cars.

I’m not really the guy to talk to about the engineering side of this.  I’m more the testing and good humour side of things (and that’s a stretch - trying to make it look like I bring more to the table here). @CNR5529 may want to chime in here. He did all of the design on our two test versions.

Craig

[daniel_leavitt2000]

I was looking up bearings and found this:
https://www.microlap.com/products/jewel-bearings/olive-single-cut-bombe/olive-hole-single-cup-bombe-mt8004/

No idea what each one costs, but here you have a small inner shelf that reduces friction along the axle and a tapered end that would reduce friction if the wheel back rubs against it.

[peteski]

I was looking up bearings and found this:
https://www.microlap.com/products/jewel-bearings/olive-single-cut-bombe/olive-hole-single-cup-bombe-mt8004/

No idea what each one costs, but here you have a small inner shelf that reduces friction along the axle and a tapered end that would reduce friction if the wheel back rubs against it.

While not fully enclosed Kato Amfleet truck bearings are pretty close to what you are looking for. I don't think that having the bearing closed (circular instead if U-shape) would give any benefits.  Only to top part of the bearing is really being utilized in those trucks (since it always supports the weight placed over the bearing (the weight of the car).  Also the more precise (tight fitting) the bearing is, the more critical the more perfect the alignment has to be.  Tolerance-wise an N scale truck is nothing like a wristwatch movement.

As I see it the problem is that even at 1mm, the axles are very thick compared to the shafts in wristwatches.  You could probably thin down the Kato bearings to reduce the contact area between the axle and bearing, 

As for the ball-bearing experiment Rapid did, as soon I've seen that design, I knew it would be trouble.  Unless fully sealed, ball bearings are very unforgiving of any debris that will get inside (and train layout is far from clean).  Also, any lubricant applied to the bearing will dampen its ability to roll freely.  Not a good idea for N scale trucks.

[Lemosteam]

I wonder if one more iteration of the PB wire trick could work.  What if, from the center of the truck, the wire was fully sprung resting atop the axle with no direct hard contact to the truck frame, other than a groove for the axle to ride in vertically.  This would leave only four single point contacts of friction to the wire to both axles.  The only other source of friction would be the groove, which could be designed as a blunt knife edge in the plan view to minimize contact.  Different diameter wire could be experimented with to determine the most appropriate spring rate for the weight of the car.

[daniel_leavitt2000]

Pete:
The bearings I linked to are a stock item and not custom made, so they might be attainable. I bet if the inside diameter of the jewel was larger than the axle, this would only have two points of contact across the entire axle and both would be a minimized jewel surface. Polishing the surface of the axle would also help greatly.

John:
I actually did this exact experiment with the new Bachmann Amfleets when they came out. It DID reduce friction, but not enough to make it on par with the Kato Amfleets. Even with a knife edge frame (in my case Delern, but this would be Shapeways plastic), there appears to be a lot of friction on the plastic-metal interface. I am convinced that we have hit a wall with what we can do with design, and now we need to focus on materials.

One of the things I noticed with the new Comets is that bearing appears to be machined and the bearing surface on the wheels is dull. That makes sense for a production model as this is the cheapest way to produce the cars. But both surfaces should be lapped and polished to a mirror shine. I have several polishing products used for watches - diamond paste in several sizes to remove scratches from sapphire glass and Cap Cod Cloth to remove scratches and polish steel watch cases. I think both might work here.

In addition, I think polishing the wheel tread may help as this is another point of friction we have not addressed yet.

I'm not expecting big improvements with any of these. But if we have a 15% reduction in rolling resistance with one and a 10% reduction with another idea, well that would equal another car on the train.

[CNR5529]

The latest results from testing were a quick and dirty attempt at installing the Rapido LRC wheelsets and ball bearings in the Rev 1 (bushing) truck. It worked remarkably well, and was a direct fit in the truck frame. The big advantage I noted here was that where the bushing and wire truck versions were very free rolling on their own, they suffered from increased rolling resistance with the weight of the carbody, while the ball bearing truck seemed to decrease rolling resistance with the added weight. Bearings similar to the ones Rapido used are fairly cheap to source too, case in point:

https://www.aliexpress.com/item/32519049244.html?spm=a2g0o.productlist.0.0.50c01f72TM7xxv&algo_pvid=2738780c-8a1c-4443-9abe-cd34f397f04b&algo_exp_id=2738780c-8a1c-4443-9abe-cd34f397f04b-50&pdp_ext_f=%7B%22sku_id%22%3A%2257448762734%22%7D

The problem people ran into with LRC cars not rolling was actually a problem with the Rapido truck frames and pickups being bent, and the bearings ending up being out of alignment. When I used those wheelsets in my straight/square rev 1 BL trucks, they ran perfectly so this should not be a problem.

Now for the bad news: the Athearn wheelsets do not have the same size axle diameter as the Rapido cars. Rapido went with a 1mm axle dia meaning they could source cheap bearings, while Athearn used a 1.2mm diameter (cant you guys just adopt the metric system already?  ). This means we either need to source custom bearings, or get custom wheelsets to replace the Athearn wheels... both solutions are beyond my hobby budget for cars I basically have no need for.

So, TRW collective, here is the mission: Find a cheap ball bearing with a 1.2mm ID and approximately 3mm OD (the truck can be tailored to suite slightly different OD), or source inexpensive replacement wheels/disc brakes with 1mm axles.

I wonder if one more iteration of the PB wire trick could work.  What if, from the center of the truck, the wire was fully sprung resting atop the axle with no direct hard contact to the truck frame, other than a groove for the axle to ride in vertically.  This would leave only four single point contacts of friction to the wire to both axles.  The only other source of friction would be the groove, which could be designed as a blunt knife edge in the plan view to minimize contact.  Different diameter wire could be experimented with to determine the most appropriate spring rate for the weight of the car.

While that is an innovative solution to provide a sprung suspension (which would potentially mitigate derailment issues if there were any), I am not sure I understand how this would be a better solution than the Rev 2 truck with the straight wires. In the rev 2 truck, all the vertical load was already in single point contact with one wire at each wheel, and there was only a single point of contact at each wheel in the longitudinal direction as well.

[Lemosteam]

One thing I have learned about FUD is that it is very much like glass with respect to hardness, so the knife edge would be similar to the jewel bearing that Daniel posted.  Also since the wire contact is never hard against another surface , the sprung wire will mitigate any variation in the axle position or diameter and under the weight of the car will always be in contact  The travel limiter button I show, would only be contacted under severe loads (more than the car weighs) essentially haven a fully floating axle with one or two points of contact at a time depending on the direction of travel. 

This is the difference from REV 2, which the wire is fully supported on both ends as it passes through the axle pocket and provides no deflection whatsoever. Rigid=suceptible to variation, i.e good on one side, more friction on the other.

Wondering if my idea provides the jewel bearing contact simulation, while removing the stiffness and rigidity from the contact wire.

[peteski]

... while Athearn used a 1.2mm diameter (cant you guys just adopt the metric system already?  ).

LOL!  1.2mm is as metric as they can get.  1.2mm = 0.047" so it is not a converted Imperial size.  While it is  not a whole number, it is just as metric as 1.5mm.

Early N scale manufacturers in Germany used 1.2mm shafts and axles in some of their models even before American manufacturers started offering domestically produced N scale models..  Yes, 1.2mm is a pain to deal with (since it is  not as popular as 1mm, or 1.5mm), but it "metric system". 

[peteski]

I agree Daniel: even with minimizing and polishing the contact surfaces to reduce the friction in these internal-bearings trucks in N scale, it is likely a losing battle. But there is nothing wrong with trying. I just have a chuckle that at some point we might have N scale models with 8 (ruby?) jewels in them (like fine watches).   

[CNR5529]

LOL!  1.2mm is as metric as they can get.  1.2mm = 0.047" so it is not a converted Imperial size.  While it is  not a whole number, it is just as metric as 1.5mm.

To be fair, 0.047" is also 3/64" rounded off to the nearest thou... so I feel like it is meant to be a metric conversion of a fractional inch value... Either way, it is a difficult diameter to find cheap bearings for. 

But I do agree, fleets of ruby/sapphire bushing trucks would be amusing!

[Lemosteam]

Beg to differ.

3/64 =  .047 x 25.4 = 1.1938, when rounded is 1.19, where rounding to the significant metric digit equivalent to thousandths of an inch (0.000 in. vs. 0.00mm)

So if you measured the diameter in three places it should average to 1.19, not 1.2 if it were 3/64" material.

Less than hairs I know, but 1.2mm material is a standard nominal metric wire diameter.

[Dwight in Toronto]

All this renewed discussion of poor rolling performance inspired to pull out my N scale Athearn F59PHI loco and all six Bombardier bi-level coaches (GO train livery), and try them on my test oval of sectional Kato track (12 3/4” radius curves).

I had bought the six cars (two 3-packs) quite some years ago, and picked up the loco maybe one year later.  It was disappointing to discover how dismal the rollability of the coaches was … to find that the loco couldn’t pull any more than three cars on my layout’s 1.5% grades was genuinely maddening. As a result, the set-up barely/rarely got any use.

So, as I said, I just revisited my GO train, only to find that the single loco could barely manage four coaches on that flat, level, simple test oval of Kato track.  All these years, I have resisted trying a tiny blip of LaBelle 108 oil on the axles, but at this point, figured I had nothing to lose.  It helped marginally - I added a fifth coach, and the loco could almost make it around the bends.  I temporarily stuck a small weight on the cab and managed to get the five cars around the oval.  But six?  Forget it.

But it did get me wondering - did anyone ever manage to improve things by adding some weight to the loco?  And if so, how much weight, where, and with what material(s)?  I don’t think I saw any reference to this approach, anywhere in this entire thread.  Just curious!

[craigolio1]

There isn’t much room in the loco for additional weight. Although in my case I added a complete lighting package that took up space for wiring and such.

Craig