N Scale Truck Reamers?

[SandyEggoJake]

@mmagliaro & @Lemosteam

I see your ...err.... point, re sideframe divot geometry.  Yes, the the Kato (and Bmann) axle point contacts are bowls, not cones.  So I concede,while noting such might rather be an artifact of the realities of punching the contacts, and not practical or optimal geometry.

And I love the test!  Thank you. 

But this bring up a related question.  How the appropriately test improvements?  An incline plane test comparing two trucks of what one might assume to be of different mass is not an ideal test.  Once it starts moving, mo takes over.  I guess we could get laser gates and scales and all... but no.  We can do better. 

Which is why I was also looking at making a N scale version of the Reboxx roll tester.  http://www.reboxx.com/wheelsets.htm#RollTester  (Not for sale, just for my own personal chuckles!)  If I'm following their data correctly it seems to be a test till suppression of motion.  Metric being a simple count of visible direction changes till still?  Not sure. 

The description suggest 2 inches is the sagitta (triple word score!), but doesn't give a cord length.  Such is the missing dimension to determine the arc used in their HO setup.  And that says nothing if such a set up is ideal for n...or even HO.  Likely doesn't matter what the arc is (within reason) (nor does accurate testing set up of a tool some debate if there is any need for!)  But let's play along a bit. 

Any thoughts on a suitable arc to develop if not the optimal test, at least a highly sensitive test with a reasonable footprint?   Off the top of my head, I would guess a 180 degree of curvature might be optimum (ol half pipe)  and a 90 degrees (quarter round) still pretty good and easier to get a realizable release.  But the reboxx version doesn't even seem close to even a quarter round.  And bending of rail could be a practical consideration as well as making the test somewhat relevant.   

Now I've already been patiently schooled on the math behind vertical curve transitions by @GaryHinshaw (my brain still hurts Prof - thank Sir - can I have another!).  While this arc isn't a compound, but a simple vertical curve, wouldn't hurt to get input from our resident guru on all things Newtonian? 


@m3798 

I think you might be over thinking tbis...

Absolutely!  Some of us call it FUN....
 
 

[SandyEggoJake]

Yikes ... and can we turn the post edit feature on again?

[ednadolski]

Which is why I was also looking at making a N scale version of the Reboxx roll tester.  http://www.reboxx.com/wheelsets.htm#RollTester  (Not for sale, just for my own personal chuckles!)  If I'm following their data correctly it seems to be a test till suppression of motion.  Metric being a simple count of visible direction changes till still?  Not sure.

Cool idea, I never would have thought of that  .   Says to use your own track so no reason why it could not be used for N scale AFAICT.   I wonder if the arc is circular, or if there is some mathematical/physics reason to make it parabolic ( @GaryHinshaw  perhaps could elaborate?).

Since gravity is conservative, a lossless (perfect) truck would in principle oscillate forever, reaching the same arc limit on either maximum.   One could compute the energy loss per cycle by measuring the deviation from the maximum.   Or, as you say, just count the cycles before it stops .

Ed

[SandyEggoJake]

That's the spirit! 

And what about adding a transitional drop-in ramp on one end, with a mechanism for optimal release? 

[GaryHinshaw]

Did someone say Newton?

I think a simple circular arc is more than sufficient for a "rolling resistance" meter.  Ed is correct that, for a perfectly frictionless system, a truck released from rest will roll up the other side of the arc to the same height it started from, then oscillate forever.  For a system with friction, counting the number of cycles before it stops would be a very good measure of resistance.

Do the size and shape of the arc matter?  Not much.  The only interesting consideration I can think of is to keep non-inertial g-forces under control.  In other words, the downward force on the truck at the bottom of the arc is more than it would be for a truck rolling down a normal linear grade, and this might contribute more rolling resistance to the truck than you would see in normal operation.

With that preamble, let's get to work and estimate some numbers.  [This is just for fun.  ]

Consider a circular arc of radius R, and a truck released from rest a height h above the bottom of the arc.  We can use conservation of energy (neglecting friction) to compute the speed of the truck at the bottom of the arc.  The total energy of the truck, E, is entirely potential energy when it is at rest,

E = mgh

where m is the truck's mass, g = 9.8 m/sec² is the acceleration of gravity at the Earth's surface, and h is the initial height.  At the bottom of the arc, this potential energy is converted to kinetic energy,

E = ½mv²

where v is the truck's speed at the bottom of the arc.  Equating these two (conservation of energy) gives the speed at the bottom as a function of h,

v² = 2gh

which is independent of m.  (The fact that this does not depend on m led Einstein to develop General Relativity 100 years ago.  Pure genius!)  The centripetal acceleration experienced by the truck at the bottom of the arc (due to its constant change in direction of motion) is

a = v²/R = 2gh/R

So the effective g force on the truck at the bottom of the arc is

g' = g+a = g + 2gh/R = g*(1 + 2h/R)

So as long as 2h ≪ R, the additional g force on the truck at the bottom of the arc will be negligible. 

In other words, just make a big shallow arc, let 'em roll, and start counting.

-gfh

[Lemosteam]

Two issues I have about that test rig, and not knowing the complete dimensions.

1: The starting tangent angle must be great enough to INITIATE rolling, as you could see my ramp at 1.6 degrees was not enough to overcome friction on the Kato truck

2: N scale track does not handle bending in the vertical plane well.  The rails will either pop out of the ties, or, more prone to adversely affect the test by creating additional resistance, will tip the rail head in or out affecting gauge and therefore rolling resistance of the truck.  A 9mm wide groove milled into the curved edge of a 3/4" thick aluminum plate. 

Then again this could be printed in FUD for, cough, 52 dollars, LOL.

[Rasputen]

I bought some 1/8 inch countersinks from McMaster Carr, but I am concerned  that the tip isn't going to go far enough into the truck to ensure that any problem would be removed.  To test this, I cut an MT truck right through one of the axle center lines, and with an eye loupe I can compare how the standard wheelset point rides versus where the tip of my countersink is going to go.  It looks to me like you need the non-cutting diameter of the countersink to be about .008 inch or less in order to be sure that any bur would be cleaned off.  M A Ford shows that their 1/8 countersink has a max non-cutting diameter of .030, which doesn't sound like much.  However, cut a truck in half, hold a .030 drill bit up to it, and you will see that there is a lot of the bore left that won't be touched.  The countersinks I got from McMaster Carr have a tip of about .025 diameter on them which is still way too large, in my opinion.

[peteski]

I bought some 1/8 inch countersinks from McMaster Carr, but I am concerned  that the tip isn't going to go far enough into the truck to ensure that any problem would be removed.  To test this, I cut an MT truck right through one of the axle center lines, and with an eye loupe I can compare how the standard wheelset point rides versus where the tip of my countersink is going to go.  It looks to me like you need the non-cutting diameter of the countersink to be about .008 inch or less in order to be sure that any bur would be cleaned off.  M A Ford shows that their 1/8 countersink has a max non-cutting diameter of .030, which doesn't sound like much.  However, cut a truck in half, hold a .030 drill bit up to it, and you will see that there is a lot of the bore left that won't be touched.  The countersinks I got from McMaster Carr have a tip of about .025 diameter on them which is still way too large, in my opinion.

Hmm, the single-flute countersink I have seems to have the cutting surface going down to the very tip of the cutter.  Besides, we really don't care if the bottom of the hole is pointed or rounded.  I have used that type of countersink to drill a conical hole in plastic, so the tip is sharp enough to start a hole.

[brill27mcb]

Two issues I have about that test rig, and not knowing the complete dimensions.

1: The starting tangent angle must be great enough to INITIATE rolling, as you could see my ramp at 1.6 degrees was not enough to overcome friction on the Kato truck

2: N scale track does not handle bending in the vertical plane well.  The rails will either pop out of the ties, or, more prone to adversely affect the test by creating additional resistance, will tip the rail head in or out affecting gauge and therefore rolling resistance of the truck.  A 9mm wide groove milled into the curved edge of a 3/4" thick aluminum plate. 

Then again this could be printed in FUD for, cough, 52 dollars, LOL.

For tests I did a few years ago, I simply used the Kato double-track viaduct incline support piers and the standard 248mm straight viaduct track sections. I made a ramp plus a flat run-out stretch at the bottom. I could compare cars side by side on it. Best of all, it can be taken down and later easily set up again in exactly the same configuration and angle of slope.

Rich K.

[SandyEggoJake]

Thx @GaryHinshaw!

OK... so shallow is more sensitive? 

Not so worried about the g-force as such would be comparable and constant at test site, with such a test if using the same trucks.  Weight removed by the reamer would be negligible.  With think on factor to see to avoid would be wheel slip, so yes, I guess somewhat shallow to ensure axle roll, not slide. 

(Besides... gravity is a myth; Earth sucks.)

[SandyEggoJake]

@Lemosteam

The idea of a single channel does remove track issues.  But instead of doing it in FUD (wha... not FXD?) I'm thinking to spark up my band saw and curve jig to cut a smooth arc in MDF, stabilize and sand before topping with a track.  But riffing on your idea, and the issues of vertical curves on track, I guess I could laminate strips of thin plate metal against both sides of a 9mm centered gauge on the arc surface.

And yet, I'm wondering if not using track loses the test some real (fake) world applicability... or at least credibility.

Seems I'm out of uninstalled code 80 atlas flex track at this tic... wondering what the max vertical curvature such would take before rail start to part from the track or otherwise twists and impacts gauge. 

[Lemosteam]

@Lemosteam

The idea of a single channel does remove track issues.  But instead of doing it in FUD (wha... not FXD?) I'm thinking to spark up my band saw and curve jig to cut a smooth arc in MDF, stabilize and sand before topping with a track.  But riffing on your idea, and the issues of vertical curves on track, I guess I could laminate strips of thin plate metal against both sides of a 9mm centered gauge on the arc surface.

And yet, I'm wondering if not using track loses the test some real (fake) world applicability... or at least credibility.

Seems I'm out of uninstalled code 80 atlas flex track at this tic... wondering what the max vertical curvature such would take before rail start to part from the track or otherwise twists and impacts gauge.

FUD was a joke and FXD will not fit in the printing tray.

My suggestion was aluminum. Where's @narrowminded when you need him? 

[narrowminded]

My suggestion was aluminum. Where's @narrowminded when you need him? 

Knocking something together right now.  This will get rid of a few scraps.

Should be done in 6-8 weeks.   Or maybe later tonight.

[narrowminded]

And here you go!

[Lemosteam]

And here you go!

/>

That is decidedly not aluminum!      Looks like my fear of rails coming off ties was unfounded too.  Did you test an NMRA gauge for the length of the track to see if the rails squeezed due to the vertical curvature?