Weight Location for Most Realistic Motion

Discussion of model and prototype wheel/rail interaction.
NickM
Posts: 18
Joined: Wed Jun 19, 2019 7:59 pm

Weight Location for Most Realistic Motion

Postby NickM » Tue Feb 13, 2024 7:44 am

Hi everyone.

My goal is to try to create the most realistic impression of a real locomotive in motion. This is both viewing in person, but even more so when a layout is filmed, which can expose the tiniest movements, appropriate or otherwise. The challenge is to convey a sense of mass - many tons rather than a few hundred grams.

My personal interest is in small end-to-end layouts, so the general context is small/medium sized locos, short trains, slow running, start/stop rather than high speed.

If we assume:

- accurate, well laid track, to P4 standards
- accurately built, free running chassis and wheels
- correctly designed, implemented and tuned CSB suspension
- correct CoG location
- correct ride height
- sufficient loco weight to pull anticipated load
- appropriate gear ratio, motor and electronics

and assuming the above result in reliable running, no derailments, good electrical conductivity etc.

I have two questions.

1 – what is the optimum location of the loco’s weight:

a: concentrated over the CoG?
b: concentrated at the loco's extremities?
c: positioned over the outermost axles
d: evenly distributed?
e: it doesn’t make any difference?

2 – what is the optimum total weight:

a: as light as possible, given the intended load?
b: as heavy as possible?
c: some other ‘optimum’ value?
d: it doesn’t make any difference?

I realise we might not have much scope, especially in relation to the first question, but lets assume we have at least some control. I guess that as we add more weight we have less options where to place it, so there may be some dependency between the questions.

Thanks, in anticipation,
Nick

andrewnummelin
Posts: 373
Joined: Mon Jul 21, 2008 9:43 am

Re: Weight Location for Most Realistic Motion

Postby andrewnummelin » Tue Feb 13, 2024 9:11 am

Not being a mechanical engineer I’m not confident in the following but I suspect:

1 f. - it doesn’t make a significant difference (but it would on the prototype - a matter of dynamics)

2 b. - The only disadvantages from maximum weight I have encountered are the weight of stock boxes and gauge widening on weakly constructed track (seen 6 decades ago with one of my locos on a club layout!). This does assume no difficulty in meeting your specification of suitable motor etc.


I look forward to seeing mathematical/engineering responses rather than just opinions.
Regards,

Andrew Nummelin

User avatar
Noel
Posts: 1970
Joined: Wed Jun 23, 2010 1:04 pm

Re: Weight Location for Most Realistic Motion

Postby Noel » Tue Feb 13, 2024 10:59 am

andrewnummelin wrote:I look forward to seeing mathematical/engineering responses rather than just opinions.

I suspect that these may be in short supply, Andrew. There are rather a lot of non-independent variables [i.e. changing one alters one or more others]. The answers may be different for tank and tender locos, and, for example, a weighted tender will potentially affect the answers. Some of the assumptions depend on missing information - what is the anticipated maximum load? What sort of motor, how much space does it take up and where is it [front, middle or back of the loco or in the tender, if there is one]? How do the loco and tender affect each other dynamically? What space is actually available for weights? What is the drive train from the motor, and what happens when the loco stalls - presumably slipping is preferable to burning out the motor?

So far as question 1 is concerned, the answer to a) depends on where the CoG is - the starting assumption is in the 'correct location', which tells us nothing. Correct for the prototype? Correct for the model? If the latter, how do you define correct? Once this is sorted out b), c) and d) may all be usable answers, although b) may affect the structural integrity of the loco adversely compared to the others.

For question 2, the answers depend on how the weight is distributed; a) seems most reasonable [why put in more weight than necessary?], but how it is distributed matters. Based solely on weight it is difficult to visualise a sensible answer to c).

The last answer to both questions is ruled out by the laws of physics. All the answers assume that the loco is laterally symmetrical, which may be open to question to a limited extent. A bare chassis can be weighted with complete freedom theoretically, but models are seriously constrained by the envelope of the body, and the need to include an electric motor, so that the optima have to be defined in the context of a particular model and what it is expected to do, and may be quite different to those produced by any theoretical mathematical model. The best you can hope for, I think, is a general concensus of opinion amongst those in a position to have one [which I am not].
Regards
Noel

User avatar
Simon_S
Posts: 134
Joined: Wed Aug 06, 2008 7:32 am

Re: Weight Location for Most Realistic Motion

Postby Simon_S » Tue Feb 13, 2024 2:31 pm

Changing the weight distribution as described will alter a model's moment of inertia in the yaw and pitch axis so weight at the extremities is probably what you want for a smooth running express loco, but maybe not for an outside cylinder industrial 0-4-0.

Optimum weight for me would be option (a) just heavy enough. My opinion is that a light but effectively sprung loco looks much more realistic in motion than an unsprung but very heavy one.

Why not take a chassis you know runs well and knock up a test body that you can place different weights on, then film a few variations and see what looks best?

User avatar
Will L
Posts: 2508
Joined: Sun Jul 20, 2008 3:54 pm

Re: Weight Location for Most Realistic Motion

Postby Will L » Tue Feb 13, 2024 3:29 pm

NickM wrote:---I have two questions.

1 – what is the optimum location of the loco’s weight:

2 – what is the optimum total weight:


Hi Nick, nice questions. I do think you are starting from the right place, but then I would wouldn't I.

In the end I think we are driven by practicalities.

The loco does need to be heavy enough, but what that is, that is an interesting question.

Firstly there needs to be sufficient weight on any axle to reduce the possibility of derailment due to minor track imperfections and other outside influences such as buffering forces or even cross winds. The forum may wish to discuss exactly what the minimum axle weights should be but 25gm per axle does seem to be accepted. In the right hands I suspect good performance maybe possible at weights less than that.

To pull a reasonable load you may want to exceed 25 gm on the driving axles. The spread sheet defaults at 100gm per axle as a starting point, but that will give you really significant haulage power, more than you need on a small layout, and It will not even be possible to get that much into a small loco. Experience suggests that loco's with as little as 50gm per axle will shift anything you are likely to want to pull on a small flat layout.

I would expect that to get the sort of steady glide you (and I) want out of a loco does require a reasonable amount of weight or more accurately inertia. The whole point about CSB springs is that they are tuned to suit the weight of the loco which should minimise the difference between different weights, but lets not forget there are other forces at play on our models.
.
Getting the CofG in the right place is important and something I think too many people quietly ignore. And this is not just for CSB fitted locos, but at least a proper CSB design includes knowing where the correct CofG location is. Compensation users should be aware of the implication of their chassis design ioo, which might include some maths, while for rigid locos, only trial and error suggests itself.

In my experience, which is at the small engines end of the game, you don't get much choice about where the weight goes, and its more important to get that right than worry too much about the absolute weight, or any stability issues from exactly where it went. It is true that you do need to keep in mind when adding weight, that putting it at the extremities impacts the CofG location more than putting it near the centre. I suppose in theory hanging a lot of weight at the extremities might have an effect on stability on big/long locos where you get some choice as to where you put it. I doubt it has a significant effect (see the Finney bogie thread) but I'm happy to be contradicted by the guys who do have experience of building that sort of loco.

davebradwell
Posts: 1169
Joined: Fri Jan 18, 2019 3:48 pm

Re: Weight Location for Most Realistic Motion

Postby davebradwell » Tue Feb 13, 2024 5:26 pm

I'm generally following the trend here by saying your questions refer to factors at the lower end of the realism scale - other things will have far greater effects. The CofG should reflect the weight diagram of the prototype and there will be limited ways of achieving this. Carrying wheels should carry weight so they don't bounce about and implementation will eliminate any unrealistic see-through areas. I'd go for heavy, so action of springs is less influenced by the inevitable friction in hornblocks. Most of all, chassis should be accurately constructed so not even a hint of a bind with very little play in bearings. I'll go on - proper 3D motion, minimal clearance in slidebars, etc.

You'll probably need a slight brake on the rear vehicle to keep the couplings tight as my bouncing brake vans have been troubling me for some time. Have a look at Chris Pendlenton's Fine Running thread which didn't go down too well but describes just the things you're after here.

DaveB

NickM
Posts: 18
Joined: Wed Jun 19, 2019 7:59 pm

Re: Weight Location for Most Realistic Motion

Postby NickM » Tue Feb 13, 2024 7:49 pm

Thanks, everyone, for the replies - hugely appreciated and thought provoking.

I was heading towards the experimentation route, as Simon suggests.

I have High-Level 94xx chassis, CSB fitted, using HL gearbox (54/1) and Mashima motor (not sure of exact model). I intend to use this chassis for no other reason than it is assembled, works well and unpainted - so good to experiment with.

I intend to try combinations of 'a' and 'b' for each question - giving 4 test scenarios. I will also try loco alone and with load, and note the point from Dave about braking the rear vehicle slightly to keep the couplings tight.

For minimum weight I will use Will's suggestion, starting point 25 gm per axle, adjusted upwards if necessary, to be able to pull a short train. Short train means six wagons (4 wheel) each weighing around 50 gm. The reason it is so short is because I just have space to do anything else at the moment. It is also a relatively light load for the minimum loco weight option.

For maximum weight, my inclination is, with the loco held and unable to move, to find the point at which the motor stalls, and reduce the weight by some value (10%?) to ensure the wheels always slip before the motor stalls. I'm assuming this is going to be better for the motor so represents a reasonable maximum weight.

For CoG position I'm assuming the position provided by the CSB calculations to be correct (for the model). CSB wire diameter will be selected dependent on weight.

In terms of test track, I will probably keep it simple initially. I'm thinking two variations of straight section and curve, the first as flat as I can achieve and the second with some deliberate small undulations. Points will come later.

I realise an experiment is only ever going to be an approximation, but it should at least answer if, for this size of loco, these factors have any observable (by the camera) difference. I completely take on board Will's comment that ultimately we are constrained by practicalities. Mind you, the panniers do have quite a lot of space to play with ;)

Please jump in with suggestions, or things I've missed or misunderstood.

Nick.

davebradwell
Posts: 1169
Joined: Fri Jan 18, 2019 3:48 pm

Re: Weight Location for Most Realistic Motion

Postby davebradwell » Tue Feb 13, 2024 8:06 pm

How are you defining motor stall? Most controllers put out far more than 12v and with feedback it may never stall with all the lead you can put in.

DaveB

User avatar
Will L
Posts: 2508
Joined: Sun Jul 20, 2008 3:54 pm

Re: Weight Location for Most Realistic Motion

Postby Will L » Tue Feb 13, 2024 8:28 pm

It is right that the loco must be able to spin it wheels when it can't move forward, I always test by running them against a buffer stop. Whether there is any point is establishing the weight needed to get the loco to stall I'm not so sure. Beware that stiffness in the chassis can absorb much more of the motors power output than pulling a train and any tendency for the motor to overheat is a likely sign the chassis is a bit stiff.

davebradwell
Posts: 1169
Joined: Fri Jan 18, 2019 3:48 pm

Re: Weight Location for Most Realistic Motion

Postby davebradwell » Tue Feb 13, 2024 8:39 pm

Hang on a minute - doesn't this chassis have a rigid axle or have you sprung it?

I've always been amazed, Will, at the amount of lead I can put in a loco and still it will spin its wheels even with 30:1. The motor is getting far more than 12v of course, so it's not a safety measure at all as motor will eventually overheat. It's why our group switched to Canon 1833 for Pacifics.

DaveB

petermeyer
Posts: 305
Joined: Tue Oct 01, 2013 10:06 am

Re: Weight Location for Most Realistic Motion

Postby petermeyer » Wed Feb 14, 2024 7:19 am

davebradwell wrote: Most controllers put out far more than 12v
DaveB


Mine doesn’t and neither does my DCC version!

User avatar
jim s-w
Posts: 2183
Joined: Wed Jul 30, 2008 5:56 pm

Re: Weight Location for Most Realistic Motion

Postby jim s-w » Wed Feb 14, 2024 8:32 am

davebradwell wrote:Hang on a minute - doesn't this chassis have a rigid axle .

DaveB


All the axles are in sliding bearings

94xx-chassis (1).jpg
94xx-chassis (1).jpg (62.93 KiB) Viewed 518 times


Jim
Jim Smith-Wright

http://www.p4newstreet.com

Over thinking often leads to under doing!

davebradwell
Posts: 1169
Joined: Fri Jan 18, 2019 3:48 pm

Re: Weight Location for Most Realistic Motion

Postby davebradwell » Wed Feb 14, 2024 9:19 am

That's a relief, although it was always possible it had been modified - Nick said he spends a lot of time on his chassis.

I'm surprised there's a dcc system that emits only 12v, although I can't measure mine accurately as I don't have a true rms meter. Yes I could put together a circuit to do this but my mini-NCE system isn't adjustable. It doesn't really matter and I was just pointing out the difficulty of defining loco weight by ability to slip wheels. Yes it sounds a good principle but even my little J72 will move well over 30 wagons before it slips. None of us are likely to leave a loco slipping for 1/2 hr until the motor burns out. Yes there are obscure circumstances. in which such can happen.

DaveB

NickM
Posts: 18
Joined: Wed Jun 19, 2019 7:59 pm

Re: Weight Location for Most Realistic Motion

Postby NickM » Wed Feb 14, 2024 9:36 am

The chassis can be built rigid, compensated, sprung or CSB.

I have built it CSB using High Level hornblocks and tags. I'm using the first 'Pannier' example in the CLAG article: http://www.clag.org.uk/beam-annex3.html#worked-examples for the CSB pivot points.

Adding CSB to the chassis requires minor modifications to the frame spacers to accommodate the CSB wires (guitar strings) . This article: http://www.clag.org.uk/pannier-csb.html is for a different pannier chassis, but the fitting of CSB and associated modifications are very similar.

Here is a pic of the chassis under construction:

94xx Chassis 01.jpg
94xx Chassis 01.jpg (122.08 KiB) Viewed 492 times


In case you are wondering, the square-section tubing is there to mount the dummy springs - which are removable to allow the axles to drop out for painting, maintenance etc. The CSB layout is designed so that the (CSB) springs avoid these mounting points.

Regarding motor stall, using a Gaugemaster Model D for power, I have taken a few measurements, with the motor remove from the gearbox.

These are values measured with the controller set to MAX power.

Motor unloaded:

Voltage = 16.7 V
Current = 56 mA
Power = 0.935 W

Loaded (applying light friction to worm gear using fingers and a soft cloth):

Voltage = 15.3 V
Current = 120 mA
Power = 1.836 W

Stalled (locking the worm gear with fingers and soft cloth):

Voltage - 10 V
Current = 620 mA
Power = 6.2 W

When stalled, the motor gets hot quite quickly.

The 'loaded' cases is obviously very unscientific - but it does give an indication of power consumption under 'normal' operation.

This suggest that this particular controller is limiting current, since the voltage drops significantly under maximum load, i.e. stalled. Even under current limit, I suspect this is not healthy for the motor for any significant period of time. Definitely to be avoided (allow wheels to slip with loco held against buffers).

Other controllers could behave differently, but I would expect (hope?) that any good controller, DC or DCC would have current limiting - hope I didn't open a can of worms!

Nick

petermeyer
Posts: 305
Joined: Tue Oct 01, 2013 10:06 am

Re: Weight Location for Most Realistic Motion

Postby petermeyer » Wed Feb 14, 2024 10:39 am

I was measuring voltage at the rails.

davebradwell
Posts: 1169
Joined: Fri Jan 18, 2019 3:48 pm

Re: Weight Location for Most Realistic Motion

Postby davebradwell » Wed Feb 14, 2024 11:00 am

Gaugemaster claim a 1.25 amp output so I would expect unit to deliver this before tripping out to zerowhen overloaded. Looks like output is just sagging under lesser load. They're a bit vague on what red LED does when it trips.

All motors will heat up when running and a decent motor should be able to run at 100C but I've never seen this clarified. There's various insulation classes for ac motors but our motors don't claim to comply. Motor data sheet will give max current but you don't say what motor is.

I think this is a distraction but does suggest you can check chassis for binding by monitoring current drawn at very low speed. An analogue meter might be better for this if they exist any more.

DaveB

NickM
Posts: 18
Joined: Wed Jun 19, 2019 7:59 pm

Re: Weight Location for Most Realistic Motion

Postby NickM » Wed Feb 14, 2024 11:55 am

As this is a little off-topic - I have created a separate one: "Controller Behaviour Under Different Load Conditions" in the "Electrics" section.

There is some subtle behaviour going on with the controller during motor stall - might be interesting if you are technically inclined.

Nick

User avatar
Hardwicke
Posts: 1506
Joined: Sun Sep 14, 2008 10:25 pm

Re: Weight Location for Most Realistic Motion

Postby Hardwicke » Wed Feb 14, 2024 1:10 pm

Watching real trains rather than models gives the answer. There is if I remember a section in the Digest on vehicle dynamics.
Real trains move smoothly.
Most model trains do not, usually caused by the gear/motor combination, electrical connections and suspension or lack of.
Builder of Forge Mill Sidings (BR(M)), Kirkcliffe Coking Plant (BR(E)), Swanage (BR (S)) and Heaby (LMS/MR). Acquired Thorneywood (GNR). Still trying to "Keep the Balance".

wakefield
Posts: 21
Joined: Sat Aug 14, 2010 3:11 pm

Re: Weight Location for Most Realistic Motion

Postby wakefield » Wed Feb 14, 2024 1:38 pm

Dave is spot on with his comment re analog meters. When you have experience of using one you will find that interpreting the way the needle moves can tell you a lot.
I spent 40 years of my working life with one in my tool kit and do not like digital meters at all.
Personal preference of course.
Mike

Chris Pendlenton
Posts: 52
Joined: Wed Dec 30, 2009 7:14 pm

Re: Weight Location for Most Realistic Motion

Postby Chris Pendlenton » Thu Feb 15, 2024 10:46 pm

It will always help the impression of mass if the engine has it, so for me its as much lead as I can get in the boiler and distributed evenly over the driving wheelbase. I'm no more scientific than that. But to complement that mass and reach the illusion you seek its all about steadiness of progress which in turn is about the usual things- round wheels, axle, rod, and crankpin throw centres and quartering all harmonious. Springing, unlike rigid or compensation smooths out bumps at imperfect rail joints, weight bearing carrying wheels limits yawing and well made motors, transmission and gearboxes reduce distracting noise. Good current collection another key factor.
Have to say though, if unhelpfully, the engine on my roster that conveys mass on the move most completely is "Hal O' The Wynd" with its Dynadrive propulsion. As the Engineer said "there was no way of stopping it..."
Chris P

User avatar
Hardwicke
Posts: 1506
Joined: Sun Sep 14, 2008 10:25 pm

Re: Weight Location for Most Realistic Motion

Postby Hardwicke » Fri Feb 16, 2024 11:19 am

Oh I'd forgotten I have a Dynadrive in a box with an MOK 47...
Builder of Forge Mill Sidings (BR(M)), Kirkcliffe Coking Plant (BR(E)), Swanage (BR (S)) and Heaby (LMS/MR). Acquired Thorneywood (GNR). Still trying to "Keep the Balance".

User avatar
BryanJohnson
Posts: 181
Joined: Wed Jul 23, 2008 7:45 pm

Re: Weight Location for Most Realistic Motion

Postby BryanJohnson » Fri Feb 16, 2024 12:17 pm

With all this talk of meters etc., could someone please show how best to incorporate a multi-meter into the wiring for a DC test track / rolling road to show the current variations as the loco turns and what setting on the meter is likely to be required.
I think I know the answer to the first part, but am less certain on the second, especially with the suitability of the 2 multi-meters I have.
An even better alternative would be to incorporate a permanent meter into my test track / rolling road. eBay has a range of analogue ammeters available. One supplier has 1A, 2A, 3A and 5A versions available; which would be the most suitable for our purposes? I'm working on the basis that we are more interested in seeing how the meter changes as the wheels go round rather than the absolute accuracy of a reading.
Thanks in anticipation.
Bryan

User avatar
Winander
Posts: 836
Joined: Thu Mar 20, 2014 12:19 pm

Re: Weight Location for Most Realistic Motion

Postby Winander » Fri Feb 16, 2024 2:45 pm

Bryan,
I got a 500 mA one similar to this that is available from a UK supplier, https://www.ebay.co.uk/itm/354982940359 You can get them for half the price from China. I also got a 15V meter. I was shocked to see the date I bought them!
Richard Hodgson
Organiser Scalefour Virtual Group. Our meeting invitation is here.

Porcy Mane
Posts: 127
Joined: Wed Apr 11, 2012 4:55 pm

Re: Weight Location for Most Realistic Motion

Postby Porcy Mane » Fri Feb 16, 2024 3:37 pm

I find those ebay meters quite handy. Like the poster above I chose 15v and 1 amp to achieve maximum deflection of the needles. Seeing the needles bounce helps ID tight spots in the chassis.

Voltmeters wired in parallel and Ammeters in series. Important not to connect the meters on the track side of the direction change over switch so that the needles don't deflect backwards against any stops following direction changes. I modified (cut) a PCB track on a Gaugemaster 100 to achieve this.

Excuse my infant school drawing abilities..

Volt&Amps-TT.jpg


TestT-DC-Side.jpg


I wasn't too bothered about the accuracy of the Ebay meters ( although they are reasonably accurate when comparing measurements to my Flukes*). I find the deflection more important. The deflection Is easier to see on an analogue meter than digital, and if its total accuracy I'm after I have banana plugs incorporated into the wiring to connect up a Fluke. The fluke has both digital and analogue display plus I can record results into exel in real time.

Should any of you be insomniacs watching this part of my test setup may be of assistance.

https://www.youtube.com/watch?v=eYrlu072Y3A

*American manufacturer of virtually bullet proof and accurate Test and calibration equipment.

P
Last edited by Porcy Mane on Fri Feb 16, 2024 7:50 pm, edited 1 time in total.

Enigma
Posts: 533
Joined: Tue Aug 13, 2013 2:49 pm

Re: Weight Location for Most Realistic Motion

Postby Enigma » Fri Feb 16, 2024 4:57 pm

Chris Pendlenton wrote: As the Engineer said "there was no way of stopping it..."Chris P


Did he manage to tell you before he died.....................? ;)


Return to “Vehicle dynamics and trackholding”

Who is online

Users browsing this forum: No registered users and 1 guest