Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Model vs. prototype adhesion

9238 views
98 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Saturday, September 19, 2020 11:42 AM

Note that as soon as the drivers 'spin uselessly' you are assessing coefficient of sliding friction, with added complications of hydrodynamic lubrication if an appropriate layer of shearable contaminant is present.  The actual thing we are wanting to measure is 'rolling adhesion' which involves asperities in one surface rolling at zero slip into interference with those on a stationary surface; nobody's "engineering" testing models so far seem to be concerned with this, and the double-salient influence of gravitational 'adhesion' vs. motor torque to produce lateral motion should not be lost on you.  What is needed is a peak-reading forcemeter that gives peak drawbar pull just at the instant before the wheels break loose (we aren't concerned about stall in this context).  Rolling stock is a bit more complex: you need a way to brake a wheel rim without putting strain on the axle bearing means, and adjusting the force to read the peak just before slide occurs.

As an aside, trying to assess the moment of slip while reading something like simultaneously reading something like pointer position on a spring scale is nugatory... fortunately in most cases something as simple as a slider moved ahead of a spring pointer would give us rough peak results.

Unless there is a consistent contact patch on both wheel tread and rail much of the variance in conditions far exceeds variance being observed due to frictional effect.  Several references very pointedly take up the fact that static 'friction' of quite a few "lubricants", in the absence of factors causing hydrodynamic lubrication shear, can be higher than clean-surface asperity interaction.  This can obviously be tested for something like the timeless-topic Wahl clipper oil... but only on track properly and reproduceably prepped, and wheels of known and equally reproduceable tread profile and characteristics.

It's going to be fun to see where this leads...

  • Member since
    January 2004
  • From: Canada, eh?
  • 13,375 posts
Posted by doctorwayne on Saturday, September 19, 2020 12:31 PM

SeeYou190
...I used the tender on one of my Oriental Powerhouse Light Mikados. The Powerhouse models have terrible tenders, so the Athearn Genesis tender was an upgrade. The resulting Hybrid model is something I am quite happy with. I hope to find another tender from an Athearn Light Mikado for my last one.

Kevin, I have one of those Athearn tenders, NiB, if you're interested.  You should have my e-mail address (via GERN) if you want to contact me, as my "Conversations" feature appears to be out-of-service.

Wayne

 

  • Member since
    January 2009
  • From: Maryland
  • 12,897 posts
Posted by ATLANTIC CENTRAL on Sunday, September 20, 2020 12:35 PM

Lastspikemike

It may be worth pointing out that our scale steam locomotives aren't.

The torque delivery to the rails is pretty smooth for our electric motor gear driven models. Anyone observing wheel slip on a real steam locomotive would notice the difference even if not understanding why.

I'm pretty sure you do not need to "quarter" the drivers on a model electric motor driven steam locomotive. The parallel would be early diesel powered British locomotives that tied drivers together with external rods. The tricky part about steam locomotives is the drive rod connection to the main driver, not the connecting rod(s). Quartering is only required to accommodate the two cylinder "engine" which our models do not have. Some have been built with three driving cylinders per crank which is even harder to "balance" from a torque delivery perspective.

Rapido drives all three axles on its Royal Hudson, MRR gives a detailed positive review of this model in the October issue.  Clearly, it would not matter whether those drivers were quartered or not, they don't even need a connecting rod.

Another thought relates to scale effects on weight.  Length maybe 1:87 but planform is not and volume is even further from 1:87. You can fit 87 scale locomotives end to end alongside the real one but that won't be enough to fill the volume occupied by the prototype.

Weight does not scale in part because physics doesn't scale in our real world environment, a 1:87 atmosphere would be problematic, for example.

Scale models should really haul @ $ $ because they are pretty heavy compared to prototype. Just as they can reach ridiculous "scale" speeds (which is a very odd  concept in itself, given that speed is speed....) although in that single dimension 1:87 holds true.

The real area of interesting inquiry relates to the coefficient of friction in the model world. Interestingly, steel on steel isn't particularly good for tractive force but it sure works well for rolling resistance!

 

Volume and weight are cube equations....

Most of my locomotives run at scale top speeds that are very close to their prototype top speeds - My DC power supply is limited to 13.8 volts, filtered and regulated before it enters my Aristo wireless throttles.

Now if you run them with those silly MRC power packs at 16-18 volts like Mike Wolf thinks DC power should be...........

I will post my next installment on this subject later today if time allows. The grass needs cutting.

Sheldon

    

  • Member since
    May 2002
  • From: Massachusetts
  • 2,899 posts
Posted by Paul3 on Sunday, September 20, 2020 3:13 PM

Sheldon,
Yep.  A 40' boxcar that weighs in empty at 46,500 lbs. scales down by 87.1 three times for weight.  So 46,500 / 87.1 / 87.1 / 87.1 = 0.070 lbs, or 1.12 oz.  At a Load Limit of 130,500 lbs., the same car would weigh 0.197 lbs., or 3.16 oz.

Likewise, a 259,500 lb. GP9 would be 0.393 lbs. or 6.28 oz.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, September 20, 2020 3:18 PM

Paul3
Likewise, a 259,500 lb. GP9 would be 0.393 lbs. or 6.28 oz.

While you're having fun with physics, don't forget to scale the contact patches, too, using comparable math.  It's not quite circular, but somewhere between the size of a dime and a quarter, and you have to assure it both on the wheeltread and railhead...

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Sunday, September 20, 2020 5:30 PM

ATLANTIC CENTRAL

Now if you run them with those silly MRC power packs at 16-18 volts like Mike Wolf thinks DC power should be...........

I will post my next installment on this subject later today if time allows. The grass needs cutting.

Sheldon 

Sheldon, did you get the lawn cut?   Laugh

Rich

Alton Junction

  • Member since
    January 2014
  • 1,500 posts
Posted by ROBERT PETRICK on Sunday, September 20, 2020 6:07 PM

If we're having fun with physics, have any of our intellectual heavyweights pondered whether scaling down weight and scaling down mass is the same? Does a ten-pound block of steel, a ten-pound block of concrete, a ten-pound block of wood, and a ten-pound bag of feathers weigh the same in HO?

I haven't pondered it, and I also have grass to cut and some cotoneasters to trim.

Robert 

LINK to SNSR Blog


  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, September 20, 2020 6:49 PM

ROBERT PETRICK
If we're having fun with physics, have any of our intellectual heavyweights pondered whether scaling down weight and scaling down mass is the same?

If gravity doesn't change, why would there be a difference... both are intrinsic properties.

Of course we're imagining the full-scale car or locomotive at the same place on the Earth as the model when we "measure" its characteristics, if you need a little more assumption information...

 Does a ten-pound block of steel, a ten-pound block of concrete, a ten-pound block of wood, and a ten-pound bag of feathers weigh the same in HO?

You blew it, son: your line was "do the various 10-pound items fall the same in HO..."

Thanks for introducing me to C. horizontalis.  I had no idea!

  • Member since
    May 2002
  • From: Massachusetts
  • 2,899 posts
Posted by Paul3 on Sunday, September 20, 2020 7:09 PM

Overmod,
Well, the railhead in HO scale has a sharp 90 degree corner and real rail has a rounded top.  The contact patch won't be the same.  That still has nothing to do with how much something weighs.

Robert,
Weight is only a force.  Mass is a function of density and volume, which if multiplied by the force of gravity equals weight.  Because mass is related to volume, to scale it properly you have to reduce it by the scale ratio three times: once for width, once for length, and once more for height.

Think about it.  If you took an 87-foot long autorack and sliced 1/87th off the length making it the correct length for an HO version of the same car, you'd still have a slice that's 10-foot wide and 19-foot high.  You also have to reduce the height by 1/87th and the width by 1/87th to get an HO scale sized car.

Weight is a cube function and to scale it, you have to reduce it by the scale three times.

What doesn't scale is friction or gravity (among other things).

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, September 20, 2020 7:37 PM

Paul3
Well, the railhead in HO scale has a sharp 90 degree corner...

Not if you gleam your track correctly.  One of the most very basic principles stated by the 'inventor' is getting rid of that unprototypical (and undesirable) edge.

Now the point does remain that, in practice, it is very difficult (bordering on practical impossibility) to superfinish or burnish the gauge-corner region so even the neutral running position gives even an order of magnitude of actual contact patch size -- usually erring on larger, rather than smaller, for "practical" adhesion, I think.  But the premise of the thought experiment is to scale the patch proportional to the weight and then see how nickel silver under test would react to what is very close to indenter dimensions...

I think he was well aware of the scaling laws when invoking the timeless topic of weight vs. mass ... at least I sure hope so, as it would make no sense otherwise.  I confess that all you really had to ask him is 'why all objects in a vacuum at the same point fall at the same speed' to get him to actually look at gravitational physics and answer his own question...

  • Member since
    January 2009
  • From: Maryland
  • 12,897 posts
Posted by ATLANTIC CENTRAL on Sunday, September 20, 2020 11:42 PM

The grass is cut, so here goes.

All theory aside, example #2, the Bachmann 2-8-4. Sorry I don't have more photos.

The Bachmann 2-8-4 was one of the first in a series of regular line products from them that are nearly Spectrum quality.

The drive is very conventional, and out of the box the loco is a little light and a little tail heavy. It also suffers from an awkward drawbar attachment with a long pin extending down from from the back of the frame to allow the trailing truck free swing.

Out of the box it pulled 18 of my piggys.

Since my plan was to convert them to 2-8-2's while improving their pulling power, a new tender connection was necessary.

I added weight, a total of 6 oz to the loco:

 

The weight was mostly added in the front, balancing the loco evenly on the drivers. Loco weight was increased from 12oz to 18oz

And 2oz to the tender for better tracking.

 

The original drawbar post was completely removed and a new drawbar was fabricated from brass and is actually attached to the loco where the old trailing truck was attached. Then the new PSC (previously Kemtron) brass Delta trailing truck is attached to the drawbar just behind where it attaches to the loco.

If you look cloesly at the photo you can see the drawbar going up and over the truck and you can see the attachment screw. 

 

Years ago this was a common drawbar design on a number of diecast and brass steam locos. it puts the pulling force in a straight line at coupler level and makes for much better tracking in curves.

Pulling power increased about 40%, and tracking thru turnouts and curves is noticeably smoother.

A little weight and a better drawbar turned this so-so puller into a great puller - about 30 of my 5oz piggys with no problem on level track. I have done five of these like this.

To be clear, I do my pulling tests with about 80 of these:

Kit bashed Athearn 50' piggybacks equiped with Kadee sprung trucks and Intermountain wheelsets - total weight 5.2 oz

Sheldon

 

    

  • Member since
    January 2017
  • From: Southern Florida Gulf Coast
  • 18,255 posts
Posted by SeeYou190 on Monday, September 21, 2020 12:16 AM

ROBERT PETRICK
If we're having fun with physics, have any of our intellectual heavyweights pondered whether scaling down weight and scaling down mass is the same?

Lastspikemike
I assume you were referring to me?

Well done Robert. That is the funniest thing I have seen in a week!

Laugh   Laugh   Laugh   Laugh   Laugh   Laugh   Laugh   Laugh   Laugh   Laugh   

-Kevin

Living the dream.

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Monday, September 21, 2020 5:11 AM

SeeYou190
 
ROBERT PETRICK
If we're having fun with physics, have any of our intellectual heavyweights pondered whether scaling down weight and scaling down mass is the same? 
Lastspikemike
I assume you were referring to me? 

Well done Robert. That is the funniest thing I have seen in a week!

-Kevin 

I like to read through most of the threads that are posted on this forum. But, lately, I find myself unable to enjoy a lot of the technical stuff that is posted that has more to do with prototype engineering than model railroading. I see this "intellectual heavyweight" approach creeping into more and more threads where it doesn't belong IMHO. I would prefer to read less of it than more of it.

Rich

Alton Junction

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, September 21, 2020 6:13 AM

i think it's fair to question if our models can pull an equivalent number of cars as a prototype loco.    if not, how many

my models seems to be proportionally heavier than prototypes when their weight is scaled by 87^3.  and while car weight may similarly scale I think what's relavent is the friction of the cars.

armstrong's chart on bearing friction indicates 6 lb/ton for an MT car at 25 mph or 0.3%.   MT cars would be 3 lb/ton.    while i'm sure some cars are better, i think 2% is typical value for model rolling stock.

so if the friction of the train is 10x more, i think it's reasonable to expect a model locomotive to be correspondingly limited assuming loco adhesion is roughly the same as the proto (~25%) which i believe it is.

 

greg - Philadelphia & Reading / Reading

  • Member since
    January 2014
  • 1,500 posts
Posted by ROBERT PETRICK on Monday, September 21, 2020 8:01 AM

SeeYou190

Well done Robert. That is the funniest thing I have seen in a week!

-Kevin

Autumn in the Rocky Mountains, fall gardening chores. I was just spreading compost in the flower beds when the thought came to me.

Robert 

LINK to SNSR Blog


  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, September 21, 2020 8:19 AM

richhotrain
I like to read through most of the threads that are posted on this forum. But, lately, I find myself unable to enjoy a lot of the technical stuff that is posted that has more to do with prototype engineering than model railroading.

And you expected what out of a thread titled 'Model vs. prototype adhesion'?

I would prefer to read less of it than more of it.

Then I suggest that you do so, instead of commenting about it in threads that evidently don't interest or concern you.

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Monday, September 21, 2020 9:24 AM

Overmod
 
richhotrain
I like to read through most of the threads that are posted on this forum. But, lately, I find myself unable to enjoy a lot of the technical stuff that is posted that has more to do with prototype engineering than model railroading. 

And you expected what out of a thread titled 'Model vs. prototype adhesion'? 

I would prefer to read less of it than more of it. 

Then I suggest that you do so, instead of commenting about it in threads that evidently don't interest or concern you. 

As I said, I like to read most of the threads that are posted on this forum. Although, I intend to only speak for myself, I may in fact be speaking as well for others. In this case, there appear to be at least three of us who feel the same way considering that I was replying to a reply to a reply.

Rich

Alton Junction

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, September 21, 2020 9:33 AM

richhotrain
In this case, there appear to be at least three of us who feel the same way considering that I was replying to a reply to a reply.

I'm sure it's more than just three, especially considering the diversions from actual discussion of adhesion or its scientific vs. empirical achievement that are occurring in it.

There are plenty of threads I don't much care for, either, but I just don't read them, let alone comment merely about why I don't like them or why they are irrelevant to me personally.  Much of what I write about isn't going to appeal to those who only value experience, for example, rather than theory.  It just so happens that this thread was started to discuss theory and analytic techniques, not experience per se.  That certainly makes it 'not for everyone', but I also don't think it's up to self-appointed forum police to whine that it shouldn't be the way it is because they don't like it.  If you don't like it, or don't understand it, or don't want to understand it, or don't want to contribute positively ... you're right, I'm the last person in the world to tell you you can't express yourself, but please don't expect me to avoid discussing theory just because you don't like it.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, September 21, 2020 9:57 AM

Overmod
There are plenty of threads I don't much care for, either, but I just don't read them

+1

greg - Philadelphia & Reading / Reading

  • Member since
    May 2002
  • From: Massachusetts
  • 2,899 posts
Posted by Paul3 on Monday, September 21, 2020 10:29 AM

gregc,
While friction doesn't scale, we (usually) use a different kind of bearing in our rolling stock models.  The needle point bearing offers a lot less starting friction due to the very small surface area in contact with the rotating axle.

For an example, the Rapido NH 8600-series coaches came with a plain bearing truck.  The axles had blunt ends and fit into brass strips that had holes punched in them.  These cars do not roll well, to be kind.  The next design, the NH parlor, came with a redesigned truck that had needle point axle ends that fit into dimpled brass strips.  These parlor trucks roll extremely well to the point you can't park these cars on any kind of grade or they'll roll away.

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Monday, September 21, 2020 11:38 AM

Overmod
That certainly makes it 'not for everyone', but I also don't think it's up to self-appointed forum police to discuss why it shouldn't be the way it is. 

So I guess that you can do what you are telling me not to do.

Alton Junction

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, September 21, 2020 1:00 PM

Paul3
While friction doesn't scale, we (usually) use a different kind of bearing in our rolling stock models.  The needle point bearing offers a lot less starting friction due to the very small surface area in contact with the rotating axle.

i agree that the needlw point bearings in models is a good solution.

however, when i tested my trucks on an incline, i found a lot of variation (i.e. grade they rolled at).   some seem to roll uphill.   it was a challenge to get others to roll on a 2% grade.

when we measured pull force needed for a train on a layout, we measured ~2%.     this is 10x the value from the Armstrong chart

i wouldn't say the rolling resistance (i.e. friction) of rolling stock doesn't scale.   i'd say the mechanism is different, resulting in different performance.

understanding this will help modelers have realistic expectations.

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, September 21, 2020 1:17 PM

richhotrain
So I guess that you can do what you are telling me not to do.

Looks like it, doesn't it.  Let me edit that a bit to clear up the semantics.

  • Member since
    November 2013
  • 618 posts
Posted by DAVID FORTNEY on Monday, September 21, 2020 2:27 PM

I have read this entire post and to be honest I have learned very little and was a complete waste of my time.

It seems you guys are more interested in discussing the physics of real vs model then actually building or running your railroad. 

I model only modern and if I find a diesel that don't pull well I don't waste my time or expense trying to figure out why, I sell it and move on. 

Sorry for the rant,

Dave

  • Member since
    January 2004
  • From: Canada, eh?
  • 13,375 posts
Posted by doctorwayne on Monday, September 21, 2020 3:31 PM

Lastspikemike
...The load imposed on the model drawbar isn't related to the drawbar force developed, which is the topic at hand. The concept for the thread originates with the idea that some model locomotives seem to have less drawbar force than they oughta.

DAVID FORTNEY
It seems you guys are more interested in discussing the physics of real vs model then actually building or running your railroad.

Well, the discussion of physics, etc., etc. didn't really do much to alter the facts that about the best we'll get from a locomotive, real or model, is a drawbar force roughly equal to about 25% of the loco's weight.

DAVID FORTNEY
...I model only modern and if I find a diesel that don't pull well I don't waste my time or expense trying to figure out why, I sell it and move on...

That's certainly one option, but the origin of the thread was a reference to improving the pulling power of our model locomotives, hence the somewhat divergent inputs, most of them interesting in their own right, at least to some of us.

Wayne

 

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, September 21, 2020 3:46 PM

DAVID FORTNEY
I model only modern and if I find a diesel that don't pull well I don't waste my time or expense trying to figure out why, I sell it and move on. 

wow

greg - Philadelphia & Reading / Reading

  • Member since
    May 2014
  • From: Pennsylvania
  • 1,154 posts
Posted by Trainman440 on Monday, September 21, 2020 4:21 PM

DAVID FORTNEY

I model only modern and if I find a diesel that don't pull well I don't waste my time or expense trying to figure out why, I sell it and move on.  

Ah, may the gods of RTR shine upon you!

Charles

 

PS but in all seriousness, you must go through a ton of engines...

 

---------------------------------------------------------------------------------------------------------------

Modeling the PRR & NYC in HO

Youtube Channel: www.youtube.com/@trainman440

Instagram (where I share projects!): https://www.instagram.com/trainman440

  • Member since
    November 2013
  • 618 posts
Posted by DAVID FORTNEY on Monday, September 21, 2020 5:10 PM

Guys, guys, what I posted is not what I really do but I did to get the responses from a couple of you guys that are just plain silly come on Charlie your response is like it is from a 5 year old. I have no understanding of the entire thread. There are some things you can't scale down even if the numbers say you can. For instance the low end of sound, you cannot move enough air to get the rumble of a diesel engine. From the size of our speakers in ho.

Maybe one day when I'm dead and buried, we will but for now, NO. 

Excuse me if I offended anyone.

Dave

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, September 21, 2020 6:16 PM

i just hope some of these less than serious comments don't get the thread locked like the last two i was interested in

greg - Philadelphia & Reading / Reading

  • Member since
    January 2009
  • From: Maryland
  • 12,897 posts
Posted by ATLANTIC CENTRAL on Monday, September 21, 2020 6:23 PM

Paul3

gregc,
While friction doesn't scale, we (usually) use a different kind of bearing in our rolling stock models.  The needle point bearing offers a lot less starting friction due to the very small surface area in contact with the rotating axle.

For an example, the Rapido NH 8600-series coaches came with a plain bearing truck.  The axles had blunt ends and fit into brass strips that had holes punched in them.  These cars do not roll well, to be kind.  The next design, the NH parlor, came with a redesigned truck that had needle point axle ends that fit into dimpled brass strips.  These parlor trucks roll extremely well to the point you can't park these cars on any kind of grade or they'll roll away.

 

Let's just keep in mind, that while it is different from the prototype, our "needle point" wheelsets do not ride on the point, they ride on the top of the cone that makes the needle point, no matter how small that contact patch is.

Considerable testing brought me to the Kadee sprung truck/Intermountain wheelset, drop of light oil combination as being the most free rolling setup.

Some plastic trucks equaled it, but none could consistantly beat it. And the benefits of equalization are lost with the rigid plastic trucks.

Sheldon

    

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!