N Scale Truck Reamers?

[SandyEggoJake]

I think I've been thinking about these axle pockets all wrong...

Previously, I thought the optimal (minimal friction and lateral runout ) would be the <50 degree axle point nesting with the 60 degree axle pocket, right to the point, with axle length equal to distance to the bottom of a pair of pockets.
 


But practically speaking, warping and production variables makes this design less than desired.  A small amount of gap between the axle points and the axle pocket bottoms is obviously preferable to a truck that is too tight. 



Then @narrowminded stated "The standard has the tread at a 3 degree taper across the tread".  Such reminded me of why that is necessary - at least for real world wheelsets - and why the ideal rail gauge widens on a turn.  Because the axle tilts slight around a curve as the thicker part of the tread, fixed by the flange, rides the outer rail of a curve, and the thinner tread rides the inner rail, making a shorter path, and helping the wheelset navigate the curve.



Thus, in reality, there is some additional benefit from this truck slop, in helping a wheelsets do more roll and less slip.

As to how to test THIS function of an optimal truck, your guess is as good as mine!   Set of S - curves on an arc?   

[C855B]

You don't factor the elasticity built into the sideframes. I wager the reality is closer to your first illustration, and there is no slop. That, I believe, explains at least some of the variability I was encountering in my testing. Once I started bending the sideframes to remove and replace the wheelsets, I realized things weren't perfectly elastic and it wasn't returning to the unmodified state. IOW, I was changing the thrust pressure on the axle points, but that pressure was still springy enough to seat the needles into the apexes of the sideframe cones.

Hence my original postulate about plastic axle warming with subsequent runs - heating would be only at the very tip of the needle, a low mass with little inherent ability to sink away the heat. You know, akin to the old Boy Scout-style friction fire starter.

[Lemosteam]

@SandyEggoJake

I pointed this out on page one, see below.  Your example is missing the spherical radius at the bottom of the cone on the sideframe though.  The sideframe cannot be made without one, the pocket cone would never come to a true point and my guess is the smallest radius in the bottom of the cone might be 0.1mm, that is 0.2mm shown.

Just because there is a proven standard that works for mass produced designs (which FUD is not), it really does not matter how you achieve a single point contact inside the pocket.

Mass production would require a spherical point in the pocket anyway- it cannot be sharp- the plastic requires a radius in the bottom of the cone, likely a very small one, to be manufactured.

In a dynamic condition the axle point will make contact in multiple areas of the cone due to car rocking, slowing, starting and track irregularities, etc, but none more so than the TOP, due to the weight of the car.

The Dremel bit produces a very smooth surface inside the FUD pocket and accuracy to the axle axis is not an issue because I can still clean up the bottom of the cone with the bit at an angle to the sideframe.  With FUD, opening and closing a truck frame can be perilous so I like to leave that to the final operation of inserting the axle.

When used, the bit makes my sideframe designs roll as smoothly as a Kato passenger car truck on electrical pickup stampings.

I have not tried it inside a Mass produced truck as yet.

@C855B that is until the sideframes have been opened up.  There is no way they work axle point on point-in-pocket.  That sharp point pocket cannot be molded.

[peteski]

Guys,
I tried to tell you that much earlier in this thread (here), when we were discussing the rounded shape of the bearing cups. In most (or even  all) trucks (locos and cars) there is slight side-play of the axle so as shown in the above drawings the axle end rides on the side surface (not the rounded bottom) of the bearing cone.That design also results in self-centering of the axle.

[Lemosteam]

Guys,
I tried to tell you that much earlier in this thread (here), when we were discussing the rounded shape of the bearing cups. In most (or even  all) trucks (locos and cars) there is slight side-play of the axle so as shown in the above drawings the axle end rides on the side surface (not the rounded bottom) of the bearing cone.That design also results in self-centering of the axle.

Pete, the cones would have to be the same angle to do what you are saying.  Now the axle point may ride up the angle, but it is still a single point contact.

[peteski]

Pete, the cones would have to be the same angle to do what you are saying.  Now the axle point may ride up the angle, but it is still a single point contact.

I agree - that is why that design is so free rolling (very small point of contact). But the shape of the bottom of the axle cup is irrelevant (since the axle is not pressed against it). I guess we are saying the same thing, but in different terms.

[narrowminded]

I agree - that is why that design is so free rolling (very small point of contact). But the shape of the bottom of the axle cup is irrelevant (since the axle is not pressed against it). I guess we are saying the same thing, but in different terms.

And I agree with both of you.   The axle coming to a point and then the play and the angle difference between the pocket and the axle assures the smallest bearing engagement and least friction from both the small surface area and the minimum radius that has no leverage to speak of.  And it's adequate because the load of a car spread over 8 points is just not much.  That minimal bearing surface is adequate for the job. 

All of the units I tested were standard MT components and the metal axles were made with the proper axle length to fit MT trucks.  All had end play, and fit basically identical.  A longer axle will bind at the point unnecessarily.  The only real performance differences, aside from dirty parts which suffered and in relationship to the dirtiness, were between plastic and metal with metal rolling measurably better.  Pizza cutter flanges, once cleaned, didn't change the rolling performance in any measurable way in my tests.

[peteski]

Pizza cutter flanges, once cleaned, didn't change the rolling performance in any measurable way in my tests.

I also can't visualize as to why the flange depth would affect the rolling performance. As I visualize the rail to wheel contact . Depending on the rail-head profile, there will only be either single or possibly 2 points of contact between the rail-head and the wheel thread/flange.  With most rail-head profiles I think there is only a single point of contact.

[narrowminded]

I also can't visualize as to why the flange depth would affect the rolling performance. As I visualize the rail to wheel contact . Depending on the rail-head profile, there will only be either single or possibly 2 points of contact between the rail-head and the wheel thread/flange.  With most rail-head profiles I think there is only a single point of contact.

Agreed.  

As I described previously, because there isn't any real difference in the way those two designs engage the rails I anticipated that they would perform the same so I wasn't surprised when they did.

[ednadolski]

I also can't visualize as to why the flange depth would affect the rolling performance. As I visualize the rail to wheel contact . Depending on the rail-head profile, there will only be either single or possibly 2 points of contact between the rail-head and the wheel thread/flange.  With most rail-head profiles I think there is only a single point of contact.

That might apply for primarily tangent track, but if that were the only use case then there would be little point to making wheels with oversized flanges.  The primary purpose of pizza cutter flanges is to keep cars with truck-mounted couplers on very sharp curves.  Under those conditions there are all sorts of additional truck torques and side loads on the flanges and axle points that add up to additional drag, but you would not see those on a tangent track test.

Ed

[peteski]

That might apply for primarily tangent track, but if that were the only use case then there would be little point to making wheels with oversized flanges.  The primary purpose of pizza cutter flanges is to keep cars with truck-mounted couplers on very sharp curves.  Under those conditions there are all sorts of additional truck torques and side loads on the flanges and axle points that add up to additional drag, but you would not see those on a tangent track test.

Ed

Ok, I see how if a wheel is riding on the inside curve, there could be 2 points  of contact in the flange area. But whether the flange is shallow or deep, makes no difference. Flange inside profile (of any depth flange) is designed not to be parallel to the wheels face or perpendicular to the  tread surface. The rail head also has a rounded profile. That minimizes the contact area between the wheel and the track.

I also do not think that the deep flanges were used to allow the wheel to stay on the track on sharp curves. Deep flanges have been around since the infancy of model RR. I think  there was no complex calculation to figure out that they will track better. It was simply done 1. As a preventative measure for keeping the model on track during extremely high-speed operation on curves (no complex math was involved in determining the depth) and 2. to make easier for kids to put the models on the track.

I don't think the lower part of a deep flange is ever in contact with side of the rail head. If that was happening then the wheel tread would no longer touch the top of the rail head. Maybe that does happen but only during extremely high sped operation on a curve where the wheel on the opposite side comes up off the track.

[ednadolski]

I don't think the lower part of a deep flange is ever in contact with side of the rail head. If that was happening then the wheel tread would no longer touch the top of the rail head. Maybe that does happen but only during extremely high sped operation on a curve where the wheel on the opposite side comes up off the track.

Try pushing a long train with truck-mounted couplers upgrade thru a series of sharp curves.  Speed isn't the issue, it's that the off-center compressive forces in the couplers will induce a rotational torque on each truck.  This adds side load to each wheelset which increases drag.  When the torque gets large enough the wheel treads will start to lose contact with the rail head and ride up against the rail as you say.  This is one of the key issues with truck-mounted couplers, since good operation presumes that the treads are always in contact with the railheads.  'Fixes' such as larger flanges and additional weight have been used for a long time, but they don't address the root cause and introduce other drawbacks.

On curves you will always see the extra drag from truck-mounted couplers because of the rotational torque.  That is always there to some degree even if you aren't about to stringline or accordion the train. Where there is truck torque there is side load, which creates the extra drag.

In a free-rolling test there is no way to reproduce the rotational torque.  In that sense then the free-rolling test doesn't really represent the performance of the truck under actual operating conditions (unless your layout has no curves ).  That's not a knock on the test, since it's real benefit is to check minimal performance as well as relative improvements (from reaming/cleaning, etc.)


Ed

[DKS]

Try pushing a long train with truck-mounted couplers upgrade thru a series of sharp curves.  Speed isn't the issue, it's that the off-center compressive forces in the couplers will induce a rotational torque on each truck.  This adds side load to each wheelset which increases drag.  When the torque gets large enough the wheel treads will start to lose contact with the rail head and ride up against the rail as you say.  This is one of the key issues with truck-mounted couplers, since good operation presumes that the treads are always in contact with the railheads.  'Fixes' such as larger flanges and additional weight have been used for a long time, but they don't address the root cause and introduce other drawbacks.

On curves you will always see the extra drag from truck-mounted couplers because of the rotational torque.  That is always there to some degree even if you aren't about to stringline or accordion the train. Where there is truck torque there is side load, which creates the extra drag.

In a free-rolling test there is no way to reproduce the rotational torque.  In that sense then the free-rolling test doesn't really represent the performance of the truck under actual operating conditions (unless your layout has no curves ).  That's not a knock on the test, since it's real benefit is to check minimal performance as well as relative improvements (from reaming/cleaning, etc.)

This. And this has long been one of the central points in the truck- versus body-mount argument. Body-mounting avoids many of the problems associated with pushing strings of cars, although body-mounting raises problems of its own, such as sideways forces applied when longer cars move through S-curves (which plague 1:1 trains as well), among others.

[SandyEggoJake]

So I'm in the planning stage for a new portable bench track.

It will be rather short, as it is mostly a section to install running stands and saddles.  But I've made the base wider just to have more stability.  Given the real estate, and desire for additional weight over just a flat board, I'm thinking I'll squeeze more utility out of it by including my own truck test arc on it, on a second track (parallel, behind).

@narrowminded :

I see your length is about 2', but I'm considering a shallower arc (higher radius than your 19.5") and/or considering to have a short horizontal tangent at the nadir.  Not much - perhaps just an inch?  Just enough to eliminate the uncountable microcycles at the end of a test run, thus allowing better result resolution.

But what did you use for that spring release mechanism?  Can't make it out in your video, but I agree it allows for a sweet, repeatable release.  I'm pondering simple solutions - using the guts from a click pen for example -  but thought I would ask what you used. 

Again, thx for the video. 

[narrowminded]

So I'm in the planning stage for a new portable bench track.

It will be rather short, as it is mostly a section to install running stands and saddles.  But I've made the base wider just to have more stability.  Given the real estate, and desire for additional weight over just a flat board, I'm thinking I'll squeeze more utility out of it by including my own truck test arc on it, on a second track (parallel, behind).

@narrowminded :

I see your length is about 2', but I'm considering a shallower arc (higher radius than your 19.5") and/or considering to have a short horizontal tangent at the nadir.  Not much - perhaps just an inch?  Just enough to eliminate the uncountable microcycles at the end of a test run, thus allowing better result resolution.

But what did you use for that spring release mechanism?  Can't make it out in your video, but I agree it allows for a sweet, repeatable release.  I'm pondering simple solutions - using the guts from a click pen for example -  but thought I would ask what you used. 

Again, thx for the video.

I agree and think I may have mentioned that a shallower angle and even shorter length might be more convenient.  It's the simplicity and repeatability of the test that makes it useful.  This really was knocked out of some foam scraps laying about that already had the radius cut so it follows that the trigger release was equally simple.  

It's just a piece of piano wire, I believe .020", bent up at the one end to hook an axle on a 33" wheel from underneath.  A hook that clears the center ties sufficiently to release the axle when depressed in the direction of the ties.  At the other end, an inch or two away, a reverse bend with a fairly long tail left so it can be inserted fairly deep into the foam to afford some structural integrity.  Expect to fiddle with the bend to get it to engage the axle without lifting the truck off the rails but once the right bend is found it will hold that position fine.  Set it up, push down on the center of that spring arm, and you're off to the races!

Edit add: Another reason to stay a little more shallow on the radius or same radius (19.5") but shorter length is to not have the truck at such a steep angle at the start.  This one is almost enough to have the truck tipping off instead of hanging still until the release.  This became evident when I stacked the second inverted truck on top of the first (one with metal wheels one with plastic) to comparison test the metal vs: plastic wheels without changing anything about the configuration or weight.  It was right at the tipping point and took a gentle hand to get it hooked without it tipping forward and off the hook.