Mickoo's American Modelling Empire

[richard carr]

HI Mick

The SD35 cab is just brilliant, you don't happen to have a spare one do you for my SD35 ?

Richard

 

[mickoo]

HI Mick

The SD35 cab is just brilliant, you don't happen to have a spare one do you for my SD35 ?

Richard

Might have

Need to make it a bit more robust as yet mind, it's a bit sticky plaster and bailing twine at the moment, the nose is fine (all the noses are actually fine, 81, 84, 88 and 112" as well as differing handbrake notches) but the cab sections are not as tuned as I'd like. There's a couple of slots in the top of the lower cabinets to clip the cab into on the Atlas model, I might add corresponding lugs onto my side panels and that might help the slight bow the bottom of the sides seem to constantly want to do.

I've also got to add the WC vent and nose top sand filler fitting to the 3D set up and will work up a MARS light as well. I may also work up some new lower cabinets as the doors and catches on the RTR do stand out as being rather clunky when you sit the 3D model on top, I can then amend the battery box vents to suit the SP engines.

I've been happily banging on with the SP L screen and some guy over on the SP modelers FB page just quietly tipped me off that the SD35 never had the L screen, I knew this of course and luckily have the correct cab front with the wider centre RH screen but not L shaped already drawn up.

Amazingly of all the other hundreds of views over several user groups, none else has noticed my FUBAR....

Do you have a particular road number or period you want, the vents in the cab front and nose side changed on some engines, they should both be 4 slots but some are 3 and some are totally sealed and there are some other changes with the E and R rebuilds in later life. The knuckle duster catch retention clip on the nose also changed with specific models, probably by build date or reworks, my nose has them outboard but I've worked up one with them in board as found on SD40-2's etc.

 

[mickoo]

A question I actually knew the answer to, and Mick beat me to it by several hours!!

Love the work on the cab, Mick. Don't forget to add CF7 Angle cab to your 'eventually' list....

Jonte - great work on that fuel tank!! I bet that is quite unique in modelling circles.

Front, nose and steps are already drawn, it'll need some tweaks to suit the new machines greater accuracy and work out the optimum orientation, then the sides, rear and roof can be added. But that's only when I have solved how to make it all fit as I'm aiming for the Airfix, Tamiya kit component approach, rather than a full blown complete module at the moment.

 

[Jordan]

I've been happily banging on with the SP L screen and some guy over on the SP modelers FB page just quietly tipped me off that the SD35 never had the L screen, I knew this of course

Not following the SP myself, I wouldn't have known.

(ReCF7) Front, nose and steps are already drawn, it'll need some tweaks

My version would like to have the deeper ('dropped'?) Engineer's front windscreen, just to be awkward... (as on my HO version)

 

[mickoo]

Not following the SP myself, I wouldn't have known.



My version would like to have the deeper ('dropped'?) Engineer's front windscreen, just to be awkward... (as on my HO version)
View attachment 173097

Noted

 

[mickoo]

What follows is a rather convoluted way of killing two birds with one stone.

First off a photo, not particularly good photogenically, but fits in later.



3D printers are very good at organic shapes, they're not that good at engineering shapes and US diesels with all their flat surfaces and hard edges are about as difficult as it gets when printing.

The very nature of the technology is it's own downfall, namely the peeling of the translucent film (FEP) off the base of the newly printed layer, the FEP is generally an oblong/square shape in the bottom of the vat and when the plate pulls up the peeling begins at the edges first and ends up in the centre, in crude terms, just before the last piece peels free the FEP is effectively an inverted bowl shape.

This causes the layers to exhibit a curved, warped, bowed effect, you can get around this by adjusting your orientation, extra supports or sacrificial edges as well as adjusting other parameters. Sacrificial edges are probably the most useful and the bowing tends to only occur at the beginning of the object (closest to build plate), the more layers it prints the less the effect becomes. Adding an extension of sacrificial material means the bowing occurs here and not in the part of the model you want to keep straight.

By contrast I've not tried extending the supports the same distance as the sacrificial part, the issue could only be occurring close to the build plate........note to self to test that at some point.

Often if your part exhibits a constant bow or twist you can inverse engineer it in CAD so that when printed it actually ends up straight, I've had to do that for the EMD spartan cab sides/roof section and it worked really well. Often as not, you don't need to inverse the actually object you want to keep, quite often you can inverse engineer a sacrificial plate or object and join it to the problem area; that will often be enough to pull the area straight.

There are more theories, work around's and sacrificial designs than UFO sightings in Roswell NM, what works for one doesn't always work for others and is not uniform across all shapes, sizes or orientations, bit of a lottery really.

One theory that seems to stand the test of time and keeps popping up (other than sacrificial edges which actually do work very well and an important design tool in printing straight and sharp objects) are sacrificial walls, bunds. moats, keeps etc etc.

In essence they are additional objects on the build plate in an attempt to even out or counter the peel effect off the FEP or force it to behave more to your command than the usual witchcraft.

The problem item is the sub base cabinets for the Spartan cab, having details on four of the six sides reduces your options for orientation and thus minimizing distortion. Having expended a lot of mental capacity, time and resin it was fast approaching the point where it may not be possible to print it effectively as it stood; inverse engineering the part was impractical due to the surface detail.

Almost at the same time (waiting for the latest failed rendition of the sub base to clean up in IPA) I was browsing one of several FB pages on US layouts and Flickr for inspirational industrial layout scenes.

There comes a point where even the dimmest of folks can join two dots together, right there on both web pages was the solution, probably the most ubiquitous scenery item you can see and find in the US, so common you almost become blind to them, the 'Jersey barrier'.

Right there was the solution to what was normally a throw away sacrificial wall/bund, it does give you straighter edges to the piece you want, but the sacrificial part just feeds the bin monster.

A quick whizz in google found an eye watering selection of designs and drawings, so I picked one that looked most like the ones I'd seen....back to first image....in the US, they exist in nearly all urban areas where rail meets other traffic but also out in the wilds to stop 'Hill Billy's' tearing up and down the ROW with their 4x4's and quads. Though their effectiveness at curtailing that activity seems mute.

A quick doodle in CAD, load the build plate and et viola, nice straight sub base parts and expendable sacrificial walls/bunds that can actually be useful.



They still need the wire joining hoops adding and a bit of distressing but they're certainly more useful on the planned layout than in the bin. Just need to dig around and see if there are any other scenery bits that can double as print wall/bunds and kill two birds with one stone, I do need some dumpsters too, actually, my mind is going mad at all the scenery bits that can be printed and few of them need the accuracy of parts on engines.

 

[SimonT]

I spent most of the post up to the punch line trying to work out how you built a nose cowl on top of a road barrier.
Having only used a Form, I didn't know that the sheets of FEP came away until there was a decent amount of built resin to stop it moving. Presumably there is still a scarificial edge to get around the scalloping between the supports?

 

[JasonD]

Mick, do you have any of your excellent close-ups featuring the handrail stanchions?

My gripe is the dreadful stanchion shape chosen by MTH - what I call Priest's Hat cross-section. Starting with the GP9s the EMD cross-section has been square, flat U, but it's not that simple these days: varying widths, not just bolted to the side of the frame, but continuing below it(?) which gave MTH the excuse to have most of the stanchions punched together with this easier cross-section and joined by a continuous fold inwards.

Jason

 

[mickoo]

I spent most of the post up to the punch line trying to work out how you built a nose cowl on top of a road barrier.
Having only used a Form, I didn't know that the sheets of FEP came away until there was a decent amount of built resin to stop it moving. Presumably there is still a scarificial edge to get around the scalloping between the supports?

Simon, all forms of STL printer require the FEP to peel clear of the item with each layer, even the Form. If it doesn't then the fresh resin can not flood back in ready to be burned for the next layer.

The Form is slightly different in that the FEP actually sags with the weight of resin in the tank, the laser head then transits across the tank and pushes the FEP up to the required level which in combination of the build plate height gives you your layer depth. Because the laser head is quite narrow it only pushes up on a small area so your hydraulic compression forces are much smaller, AKA liquid being squished out of the gap ahead of the laser head. This reduces the dreaded elephants foot and blooming seen on normal printers.

Example, imagine a 100x100 slab printed flat on the bed, on the form the leading edge of the laser under the FEP is only 10 mm wide, therefore your resin compression is only ever 10x100 as the head moves across the FEP, that allows the resin more space to squidgy out. On a normal printer the FEP sits above toughened glass so when the build plate lowers it has to force out resin for the full 100x100 surface area, being quite viscous it can take some time for all the resin to squidge out and stop moving.

Viscosity depends on many factors, the two most important ones are the resin it's self and it's temperature, the Form enclosure is heated so it's pretty good at that, other printers often need reptile heaters or the such like. On mine I usually do a quick test, I watch the first layer and as the resin flows up and over the build plate you can see it moving, when the build plate stops I can still see it moving and just count until it stops, takes about 6 seconds.

Obviously you don't want to burn the layer during that time because your burning a moving liquid and that's just going to leave a poor outer surface. The most obvious time this visible is burning the base layers, this leaves the elephants foot witness mark as well as the extended burn times. The extended burn time is basically causing bleed through the resin to adjoining areas and swells the area hardened.

When you get a stay line or wobble or surface defect higher up it'll usually be down to a rapid change in surface area, that rapid change changes the fluid dynamics and you go from a small area being squidged out to a larger one, the resin is still moving so you end up with whats called a bloom layer. It's a double wammy as the larger area then suddenly needs to be peeled which creates a bigger tearing force and thus warping.

You can mitigate a lot of this by adding a time before burn after the plate has lowered, most people don't as they're in a hurry and if your printing organic shapes then it tends to show up less. I was running 6 seconds but for sharp engineering shapes I'm up to 8 second sand considering 10 seconds, yes it adds time but it also sharpens your objects and reduces the effects of blooming.

Not sure what you mean about scalloped supports but below is the set up on the build plate.

I could print the Jersey Barriers direct onto the plate but then the gap underneath for the forklift tines would get no support and turn into a saggy mess.



There are two warping issues with the sub base modules, the open inside is causing the end wall to bow outward at the open edge (inside of locomotive). The solution here is a mass of supports to try and slow that creep, in addition I have extended the wall down by 3 mm to give a sacrificial wall, that wall takes all the extra supports so can be as messy as you like because it's not part of the final model. I suspect some of this creep is fluid hydraulics in the early layers which may benefit from the whole object being moved up out of this zone.

I have a test in the printer now where the same object is jacked up much higher to see if the creep is only in a boundary close to the build plate, it adds an hour to the print but if it solves it, or even reduces it, then it will have longer term benefits. I could also try a bund/wall across the end to try and stop the fluid moving out toward the view point. In fact an L shaped bund in the bottom right corner may force the fluid inward toward the middle of the sub base and bow the wall inward?

The Jersey barriers do not stop that end wall creep, what they do help is the upper layer bow, by spreading the peel force across objects either side and slightly higher then the peel force is reduced and the bow lessened, ideally the Jersey barriers need to be closer but I have FEP marks that create dimples and spots on the build plate so I tend to avoid those areas, A new FEP would solve that but this one got dinged on the third print after replacement.

I have a saved file that marks where these bad spots are and load that first, then I simply add my bits around the bad spots and delete the pillars of doom before slicing the plate.





Hope that makes some sort of sense, probably not

 

[mickoo]

Mick, do you have any of your excellent close-ups featuring the handrail stanchions?

My gripe is the dreadful stanchion shape chosen by MTH - what I call Priest's Hat cross-section. Starting with the GP9s the EMD cross-section has been square, flat U, but it's not that simple these days: varying widths, not just bolted to the side of the frame, but continuing below it(?) which gave MTH the excuse to have most of the stanchions punched together with this easier cross-section and joined by a continuous fold inwards.

Jason

SD40-2

















Note there are two types, fixed with a rolled over top and removable with a bolted clamp....usually around the rad section.

SD70M, note whole engine is bolted clamp type as far as I can see, same for following ACe.



SD70ACe



Earlier catalogs do not have a specific handrail staunchon detail page but the one for later (3rd) generation EMD engines, GP60, SD60 etc does.



I've no reason to believe the figures are not the same (or close to) for older engines, but do be aware there are an awful lot of handrail set ups spread across the EMD 2nd Gen engines GP20-GP50 & SD18-SD50, not just spacing but heights to suit the varying car bodies, you really need to be more specific as to which class you're interested in. I think there's 20 plus pages alone just for front end handrail set ups etc.

 

[Rob R]

Very useful Mick.
Much food for thought.
Thank you.
Edit. You snook another post in there.
I was referring to the printy bit not the handrails - although they are jolly useful as well.

 

[Dog Star]

Mick,

What is the story behind the sticker about ownership?

I appreciate the explanation of the issues with 3D printing. If I have understood correctly, are some of those issues dependent upon the footprint of the build plate? If a range of build plates are available for a given printer, might choosing a build plate that is close(r) to the area of printing help to reduce some of the issues?

regards, Graham

 

[Big Train James]

I'm curious about the results of the test with the longer supports. That was my very first thought when you mentioned the idea that prints would sort of correct themselves as they got further from the build plate. If such a solution were effective, it would mean that the model itself wouldn't need to be modified, so no sacrificial edge to remove. That being said, the suggestion to include a crease in the sacrificial edge to locate the cut line and the blade is also a good one . If the extended supports don't work, I will try the sacrificial edge.

Also thanks for the screen cap with the barriers and cab parts together. I too was having a bit of trouble figuring out just how things might be arranged, despite understanding the principle.

Lastly, having the file with the defect location markers is another good idea. I also have the little marks on my film, and they translate through to the surface of parts very clearly. I call it the Princess and the Pea effect. They aren't a problem in the middle of a big flat area, they're easily sanded off. However, mine invariably end up across a detail part though, something like a hinge or door latch, those examples usually go in the bin . I did finally figure out where all the blemishes were located, and recorded their position by grid position. But I often forget exactly where they are, and I also often forget to make sure in Chitubox that I'm looking at the build plate from the right side. Having the markers on the plate would might protect me from myself .

 

[Big Train James]

Note there are two types, fixed with a rolled over top and removable with a bolted clamp....usually around the rad section.

I vaguely recall reading or noting somewhere in the past that the rolled versus clamped top on the stanchions were based on the joints between rod lengths, where clamping types were used at the ends of rods and rolled types were used at intermediate positions. I'd have to dig to find that reference though. It doesn't really matter much I suppose.

Anyway, here's some additional information on the stanchion. These are pdf's, so attached at the end. Note that the attached drawings are both specified for an sd35 application, even though one is the clamp style and one is the rolled style. Lengths for other units may vary, and in fact there are multiple lengths of stanchions on any single locomotive. These drawings are really more useful for all other dimensions aside from the length. Missing from my files, at least for now, is a drawing of the clamping piece used with types's A and B, as specified in Mick's catalog capture above.

Jim

 

[mickoo]

I vaguely recall reading or noting somewhere in the past that the rolled versus clamped top on the stanchions were based on the joints between rod lengths, where clamping types were used at the ends of rods and rolled types were used at intermediate positions. I'd have to dig to find that reference though. It doesn't really matter much I suppose.

Anyway, here's some additional information on the stanchion. These are pdf's, so attached at the end. Note that the attached drawings are both specified for an sd35 application, even though one is the clamp style and one is the rolled style. Lengths for other units may vary, and in fact there are multiple lengths of stanchions on any single locomotive. These drawings are really more useful for all other dimensions aside from the length. Missing from my files, at least for now, is a drawing of the clamping piece used with types's A and B, as specified in Mick's catalog capture above.

Jim

Most excellent!

I'm exploring getting some cast here in the UK so those drawings will be perfect

 

[SimonT]

Mick,
thanks for all explanation, all writen in fluent simond.
Formlabs told us on the Form 1 that the tank didn't use FEP. I never bothered to look on the Form3 but I've just pulled the tank out and it is soft enough to be curved when not supported - I guess that's where the 'soft peel' comes from.

The scallops were the uneven sufaces we got between supports. The rest wil need some thinking about.

 

[mickoo]

I'm curious about the results of the test with the longer supports. That was my very first thought when you mentioned the idea that prints would sort of correct themselves as they got further from the build plate. If such a solution were effective, it would mean that the model itself wouldn't need to be modified, so no sacrificial edge to remove. That being said, the suggestion to include a crease in the sacrificial edge to locate the cut line and the blade is also a good one . If the extended supports don't work, I will try the sacrificial edge.

Also thanks for the screen cap with the barriers and cab parts together. I too was having a bit of trouble figuring out just how things might be arranged, despite understanding the principle.

Lastly, having the file with the defect location markers is another good idea. I also have the little marks on my film, and they translate through to the surface of parts very clearly. I call it the Princess and the Pea effect. They aren't a problem in the middle of a big flat area, they're easily sanded off. However, mine invariably end up across a detail part though, something like a hinge or door latch, those examples usually go in the bin . I did finally figure out where all the blemishes were located, and recorded their position by grid position. But I often forget exactly where they are, and I also often forget to make sure in Chitubox that I'm looking at the build plate from the right side. Having the markers on the plate would might protect me from myself .

I've simply marked where the damage is and then made a Chitubox file with the pillars of doom on it, simply imported a column from Autocad, placed accordingly, rinse and repeat then saved.

I knew exactly which components failed and when so I reloaded that slice file and placed the pillars of doom in the corresponding places and deleted the rest of the print. But I do get your point about upside down, left right front or back, it's a bit of a mind game that one.

One thing with printing on the flat IKEA style is that occasionally you'll get a bubble right in the middle of your final layers, it's affecting quite a few on the FB page and no one seems to know why, sometimes it'll all print fine, other times you get a bubble/hole. Another niggle is that any wrinkle in your FEP, crease or otherwise non smooth surface shows up on your final layers, I always make sure there is no white sludge on the FEP before a print as even that can have an adverse effect on the final surface smoothness. Princess and the pea is a good analogy.

Back to the sub base and some more detail/observations and trials.

The bow only occurs in two corners seen here in red, the top surface can also exhibit a slight bow or waviness occasionally, normally I'd run with 2.5 or 3 mm walls and that helps a lot, my problem here is the cast lumps and bumps on the Atlas chassis, yes I could hack them off but if anyone else wanted these units then they'd also have to do the same; it would be nice if that sort of hacking could be avoided.



The previous attempts to resolve the bow are shown above, an extension plate (light blue outlines) with an inverse engineered angle to it, it kind of works and solves most issues to be fair. Buttress supports (red dots) to try and stop the wall moving, they help but are not a total solution.

Internal bracing marked with a white X, usually this works well and solved the outward bow on the cab noses at the base, in this case they sort of worked but no total solution. All three gives you the best outcome but never quite perfect and still requires some hand tools to clean up.

The big slab in the bottom serves a dual purpose, it stops the bottom of the side wall from bowing out and is thick enough to act as a securing point with screws from underneath through the cast chassis block. I'll drill those later using the chassis as a template as can't be bothered to measure (guess) where to put holes in the CAD drawing and then spend ages doing test prints to align them.

Conversely the closed side has perfectly straight edges (green lines) and deduction suggests it's the full side that's giving it support. So on to Mk VI, despite all the previous tests and wastage the bin monster is still not sated!



On this one I've kind of blended the Jersey barrier concept into the actual object, logically is should now cause the end wall to perform in the bad corner the same as the good corner.

I've added my usual cut line trough around the outside so I have a guide when cutting the wall off post curing; that will still leave a thin lip along that edge which I can trim back to give a good snug fit with the nose and hood section to the rear. The same goes for the base, the extended sacrificial skirt (green) is not full thickness, you just sand back to where it does become full thickness to obtain the correct height.

I've also increased my first layer height for 3.5 mm to 6 mm so see if that helps as a intermediate test (below) indicated good results. To be fair I think the full wall on the rear will solve most of the problems so the only way to really test the extra height issue is to go back to a failed print and adjust accordingly; being lazy and not really caring (at this moment) what solves the problem I probably won't do that. The extra resin to dissect what the issue solution was is not worth it and frankly I simply do not have the machine time right now to keep testing.

I did do an intermediate rough test (Mk V) with the addition of a partial back wall and the results were promising, hence my comments above about the current one (Mk VI) probably working pretty well.



Both parts had the old style internal brace and buttress supports on the extended edge, but no cut line to take the rear wall off nor designed particularly gracefully for cleaning up afterward. Interestingly the shorter sub base was extended to 6 mm and there is virtually no ill effects on each corner, yet the 3.5 mm still showed very small deformation on the inside lower corners, possibly because the rear wall is not full length (resolved on Mk VI). The deformation did work up past the sacrificial skirt onto the main body but was significantly better than all before.

 

[mickoo]

Mick,
thanks for all explanation, all writen in fluent simond.
Formlabs told us on the Form 1 that the tank didn't use FEP. I never bothered to look on the Form3 but I've just pulled the tank out and it is soft enough to be curved when not supported - I guess that's where the 'soft peel' comes from.

The scallops were the uneven sufaces we got between supports. The rest wil need some thinking about.

Y'all welcome.

Form are probably technically correct, FEP is a term given to a particular material and Form probably do not use that material, but the principle is the same, it's a translucent material that stretches to peel off the cured resin. The bottom of Form tanks are very slack/flexible when compared to a LED based machine whose FEP film is drum tight.

If you have a print with a large surface area you can hear it fop back quite clearly when it pulls clear and releases. As time goes on the film stretches so you do really need to compensate with your lift height, for a new FEP I run 7 mm lift height then 8 mm once it's run in a few weeks. If your lift height is not high enough then it won't fully peel clear and you'll get failed prints. Some community guys run 10 mm but then their FEP is months and months old and running 23 hrs a day on their print farms.

Obviously the larger the build area the higher you need to go, the figures above are for the Saturn range, the Mars range has a smaller FEP so you probably don't need to lift as high and I suspect the much bigger Jupiters need even more lift clearance.

The scallops I call surface slurry, yes it does still build up on surfaces parallel with the build plate but I've found the Siraya Tech resin is much better at reducing it, I tried some Elegoo 8K resin and it's like being back with the Form, it gives lovely smooth surfaces but isn't good at sharp engineering edges and has a thicker layer of slurry on the undersides. I have to add a disclaimer as I didn't run exposure tests just cherry picked what others had on the web but I doubt I'm out by more than 0.3 seconds.

Siraya tech fast navy grey surface has a particular feel to it, a harder more dense feeling, the 8K felt softer, the sort of difference between Bakelite and Nylon, it's a tactile feel that's hard to describe. Siraya tech do a much harder resin called build if I recall rightly, I've not tried it but am following a modeler from Europe who is trying it with some 3D coach bogies, it's not an easy material to work with but apparently very durable and strong and gives good details.

On the sub bases the slurry is on the inside and not visible in the end, on the cab walls it's also on the inside but on full flat faces so you just sand it smooth. One thing that does seem to help is a lot of smaller supports, a whole forest of supports with tiny joints seems to work much much better for this application.

 

[mickoo]

Warley was a good excuse to take stock of all things I'm currently out of touch with, the wallet was saved by the lack of Swiss stock on sale but took a bit of a beating with US HO.

They've come a long way from Athern blue box models I'm used to and what staggered me the most was the insane cheapness viz quality of product. Even then I didn't manage to pick up one of the supposed best 4 mm GEVOs on the market from Rivet Counter.

I expunged some shitty O gauge RTR models and replaced them with these smaller HO models, they'll sit there with the other two I have retained from near 25 years ago.

The KATO snoot SD40-2 was from a friend unloading his collection and I now finally have a UP fast forty of my own, something near impossible in O gauge and even if you can find a brass import this one was 1/45th the price. Even then the doors will be stamped, grab handles half etched and generally poor build quality unless you get one of the better ones.



The KATO BNSF AC44 was a must have after visits to Transcon2 on previous vacations, I do prefer the bark of the Dash 9 but H2 cigar band logo BNSF ticks all the boxes as well. Both of these have handrails to fit in the bottom of the box.



Finally an Intermountain ES44AC, a true GEVO, it came with DCC and sound so I was curious to hear what HO sound is all about, I struggle to engage in O sound if I'm honest so I'm not holding much hope of HO ticking many boxes, but we'll see.





I didn't realise until after I'd bought it that the front handrails were broken so will need to try and source some replacement parts at some point.

To be honest I'm struggling with the poor quality of O gauge RTR and the play-ability aspect challenge it presents due to it's size, despite the space available (of which many people can only wish and dream about so I do apologise for being glib about it) it only really lends itself to a switching layout; plus, for my US stuff I want lots of detail and I'm realising O may be too small for that. A G1 GEVO would be better suited to scratch the uber detail itch.

Once those thought processes played out it is becoming clear O gauge may not suit my detail itch or play-ability itch and when you reach that point the way forward is suddenly wide open with a vast range of options. One other advantage (for me) in thew smaller scales is more real estate and more terrain, something I've always been keen on, whilst an N gauge West Colton hump yard would be nice, it's too much track for my tastes.

 

[Jordan]

The word that comes to my mind is "compromise"... and how much each individual modeller is prepared to accept. I know the compromises I accept to stick with O Scale - mainly tight curves, short trains and a dearth of certain locomotive types and modern rolling stock. I also stick with standard O rather than P:48, otherwise everything I have would be in bits (as opposed to just some of it) & I'd never get to running anything!!

The length of US stock is a space challenge in any scale - even steam era 40ft cars are longer than UK or EU stock of the same era.