Scenery at Donohue - Part 3

Grass helps to blend the backdrop and the modelled scene

Work on the club layout as resumed yesterday after a short hiatus. Yet again, the focus in on the Donohue yard.

In the previous session, we applied ballast and let it dry without looking at the results for a few weeks. It a relief to discover it dried correctly without any discoloration.

Looking at picture from the real yard, it became evident that a lot of vegetation took root into the ballast and between ties. In the 2000s, it was no longer that well maintained. Lots of grass also grew along the building.

I started applying grass by making blobs on diluted white glue on the ballast where I wanted it. Using a straw color short static grass, I pinch the fibers between my fingers and dab it into the glue puddle. I see very little merit using a static grass applicator for this step.

The next step is similar, but this time, using greyish dead grass fibers. They had variety and also some shadows. Another way to add texture is sprinkling my usual olive green grounded foam to create smaller plants brought back to life by spring.

Sprinkling dead leaves and ground foam add texture

At that point, the layout is relatively drab and looks like a desert. Honestly, it looks great, however, it's not how early May looks like in Charlevoix. So I need more greenery!

I generally using a Noch Spring grass blend, mixed with a very toned down green from Woodland Scenics (almost mint) plus some straw colored fibers. Since a few months, under the guidance of Chris Mears, I've been adding jute rope cut in 3 to 4mm fibers. They are move texture and more muted colors. I then, once again dab these fibers into my previous work to build up the effect.

Finally, using the same blend, I fire up the static grass applicator and add a subtle coat of fibers to blend everything together. This is generally followed by a subtle sprinkling of crushed dead oak leaves.

The site of a future grade crossing

I must admit I'm quite satisfied with the ground cover. It makes the yard looks much larger and it also divide the space in two very distinct areas: the warehouse and the unloading bays. Another thing that works well is brushing fine limestone dust over the roads with a large soft brush. It really helps to get that dirt road look and the dust settle in quite a permanent way into the rough texture.

Loose dirt over roads really improve realism

On the negative side, I must admit I went overboard with grass in the warehouse area. In real life, this area was paved and covered in crushed stone so trucks could move there. Looking at my work, it looks like an abandoned warehouse, which wasn't my goal. That said, it will be easy to fix by building up new layers of dirt and gravel up to the rail heads and leaving some grass pocking through the surface. This isn't an easy scene because it must be composed as we go forward with it.

Too much grass toward the warehouse...

Monk Subdivision - Correcting The Grades

After thinking about it quite a while, doing some tests and maths, it was time to correct the grade, install tracks and test again. It all happened yesterday and the new results were quite fascinating.

The new grade is about 1.7%, including curvature factor. Some lighter locomotives do experience a slight loss of speed whe leaving the second curve. At this point in time, the drag caused by the cars being all in the curves is at it's maximum. I may try to improve that a little bit, but to be honest, most locomotives do climb the grade without issue and it would be foolish to expect a constant speed on a grade.

Newer tests were performed with the habitual suspects, i.e. a few key steam locomotives and a few diesels. We both tried with passenger and freight trains. Personnaly, it is my belief long passenger cars with 6-axle are more taxing than freight cars. I suspect a 3-axle truck on a curve creates more friction than a few 2-axle trucks. Indeed, for equal lenght and weight, I found out most of my locomotives perform better pulling freight cars than passenger cars. And mind you, my freight cars having been shelf queens for years aren't great runner either.

With that said, here are some experiments:

The 4-6-0 is a smooth and impressive small steamer (credit: Bachmann)

-Bachmann 4-6-0 (with 56" drivers): This locomotive is small yet very powerful for its size. It can indeed pull 1 boxcar and 5 Athearn heavyweight coaches. We also added 3 Atlas Branchline boxcars. It was it's limit. Quite impressive. The drive is also quiet and steady. Definitely, the metal boiler makes a difference.

-Bachmann 4-6-2 (with DCC Sound): As I previously said, this locomotive isn't a stellar performer due to being a little bit too light, but with the new grade, it can pull 1 boxcar and 8 heavyweight coaches. Far more than I ever intended to do.

BLI Light Mikados are better than I thought... (credit: BLI)

-BLI 2-8-2: I read a lot of nasty comments about the pulling power of that locomotive. Maybe mine is an exception, but it pulled the entire passenger consist without issue. Better, we then tried to see its limit by building the longest freight train physically permissible on the layout. It meant a 31 cars train, including about 6 cars with Confalone-style weights (8 oz. per car) and maybe 3 with plastic wheels. This is the longest train possible in the staging. It's even too long for Armagh passing track. The BLI Mikado climbed the grade pulling that crazy consist without effort. A beautiful sight... Imagine my relief to see this steam locomotive perform much better than anticipated. I never intended to pull more than 10, maybe 16 cars with Mikados... These locomotives are keepers in my book. That said, I'm not a big fan of Paragon 3 sound. It's fine at low speed, but under effort or faster speed, it's just a non descript annoying noise.

-Proto 1000 F3A and B: This duo of locomotive once again performed flawlessly. One loco isn't enough for the 31 cars train, but two is the sweet spot.

-Proto 2000 GP7: This pair was able to move the entire train, but had difficulty in the last stretch when it became apparent one loco had a split gear issue.

-Atlas C424/RS11: Two of them are enough for the long freight trains, but three are better.

To make a long story short, most steamers will be able to pull their intended payload. Diesels are powerful, but most trains will need multiple units to perform flawlessly. This is expected and desired. I don't regret accumulating a lot of F-units for that purpose and my large fleet of kitbashed RS-18 will come handy.

As for sound, I'm on the fence... It's cool, but most of the time, it's just annoying. I'm getting tired of it to be honest. I see two options in the future: setting sound decoders at low volume levels and maybe equipping DC locomotives with silent decoders. I hardly see any benefit installing costly sound decoders in old Proto locomotives. Also, the real sound that was missing yesterday was wheel squealing. From my experience railfanning trains, hearing the freight car wheels and other sounds is much more pervasive than the locomotives themselves. If I ever go the sound route, I may serious think about adding train ambiant sound with detectors in curves if required rather than equip the entire locomotive fleet.

I'm curious, what are your thoughts about sound? Surprisingly enough, I find among the young model railroaders a tendancy to prefer lower volume sound... and in some times, just plain silence. Are we reaching a level of noise saturation? After all, a silent train gliding over the rails with a gentle "woooshh" sound is a calming sight... almost white noise.

With that, my next challenge is building the Joffre grade (from staging to Armagh through Langlois). It may be more tricky, but now I know the recipe!

Monk Subdivision - Assessing the Grades

A week ago, I wrote a very long blog post about how I could improve the grade on the layout. Lots of numbers and even a spreadsheet. It explains how curves and grades compound together to create a horror show. I postponed its publication because some tests were performed in the last few days to better understand what was really happening. I discussed the matter with friends and people online... Most replies left me perplex, but others did confirm me that compensating curvature to get a constant effective grade was the way to go.

A 4-6-2 struggling with an effective 2.5% grade

The first test was very simple and served to determine the pulling power of my Bachmann 4-6-2 light Pacific on a straight grade. The goal was to find out the maximum grade permissible to haul a decent passenger consist made of one express boxcar and 5 heavyweight coaches. Using wood shims and a 8 feet long straight plank of pine, I built a temporary track. At 1.6% I didn't see any serious loss in power. Sure the train slowed down a little bit, but it wasn't struggling nor exaggerate. About the speed reduction you would expect in real life. 

Testing performance on a straight grade to get useful data

At 1.8%, it's still within acceptable limit if the train pulls 3 very free rolling coaches and a boxcar. At 2%, you break the proverbial camel's back and performance goes down the drain. Serious slowing down and wheel slippage. This is good information because it means the grade + curvature factor should never exceed 1.8%-2% to ensure smaller and lighter steam locomotives can perform adequately. Let's see the data for my layout.

First, eyes can trick you. When you look at a grade curving toward you, an optical illusion is created and it seems the incline is much steeper than it is. I was under the expression that my grade wasn't constant and that it was the chief cause of all my issues. Using a level and a ruler, several measurement were taken. When compiled, a constant 1.6% grade was obtained, which was in the ballpark of what I planned. It's also the Bachmann 4-6-2 theoretical limit to pull a decent consist. A conclusion imposed itself: I didn't take into account the curvature effect on the grade. Maybe I thought it was of little consequence when using 36" curves. Go figure out... When I designed a helix for the first version of this layout, I took it into account.

We know from John Allen and John Armstrong that the effective grade for a curve is Grade + 32/R, R being the curve radius. These results are a rule of thumb obtained by empirical means. They are not a scientific formula, but they give an excellent approximation when dealing with model railroad typical radii.

For Monk subdivision, the radius is 36" (32/36) thus we get an additional 0.9%... Everybody knows that 1% in railroading (the real thing) is a serious operation challenge that can warrant helpers or double heading. No different with model trains even if their pulling power isn't scalable. A mere 1.6% is thus bumped to 2.5%, which is starting to enter Woodland Scenics trainset territory... and for our poor Bachmann 4-6-2, it's far beyond the 2% threshold.

I seriously won't go further with math explanations has I'm terrible with numbers and several modellers have done much more valuable and serious work than me in that department. The layout goal isn't about doing feats of engineering, but rather getting decent and consistent results whatever the quality of locomotives. It's understandable that the current grade works well with quality diesel and steam locomotives. One could elect to only run these excellent models, but this isn't the reason why I designed that layout, otherwise, I would have made different choices. This layout is a display case for my entire collection, thus it must takes into account the poor runners too.

Diesels are versatile compared to steamers...

That leaves us with contemplating various solutions, but let's start by assessing the acceptable grade for curves. Let's say we want a ruling effective grade of 1.8%. It means 1.8% - 0.9% = 0.9%. We should thus try to build something that isn't over 1% in the yard.

The actual numbers are a rise of 1.5" for 8 feet.  The curve length is about 5 feet long thus an elevation of almost an inch at 1.6%. We must drop that a little bit. At a 0.9% grade, it means an elevation of about 1/2". Can it be done? Yes, if I spread the grade over a longer length.

Fast forward a few days and I've done exactly that. I dropped the riser in the yard and curves to get a 0.9% grade and extended the straight grade over the swing gate. Taking curvature into account, both sections now have an effective grade of approximately 1.7%. The Bachmann 4-6-2 can now easily pull a decent consist of one boxcar and 6 poorly rolling Athearn Blue Box heavyweight cars! Now, imagine if I succeed adding a little bit more weight into the boiler and making the car trucks free rolling! We've reached the point I wanted. I'm quite happy to have solved that issue even if it meant tearing down previous work and removing all the tracks that were already installed.

I also relearned a lesson. I wasted one week calculating things instead ot tweaking pieces of wood when the actual work took about 1 hour... Another proof overthink problems solves nothing. That said, I've learned a good share about pulling power, free rolling trucks, grades and operational challenges. These are aspects I never really faced seriously with other layouts in the past. As expected, this layout is not only a project, but a challenge to myself and my skills in model railroading.

On a side note, I found out track on fiberboard are noisy so a cork roadbed will be added to reduce noise transmission. Some tests shown me it was indeed the way to go. I guess trying to be cheap didn't pay off in the end.

Monk Subdivision - Taming the Monster (Grade)

Oh my! Imagine my surprise Wednesday evening when Jérôme spiked a few stretch on track on the main line and requested we test a few locos. Everything looked so great and I didn't see the need for a test, but I knew he wouldn't leave me alone if some locos wouldn't climb the grade with rolling stock in tow... A Bachman GP9 was put on the rails to pull one 40 express baggage boxcar and eight heavyweight coaches... it started well, but once half the train entered the first curve, the speed decreased dramatically until some wheel slippage happened. It wasn't a nice sight... at all. But hey, I blamed it on cheap Bachmann motor.

A Bachmann 4-6-2 struggling to exit the staging yard

Then, it was a Proto GP9... it climbed the hill, slightly slowed down but performed quite well! Hurray, it was just because of a crappy loco. Then the Proto F3, a brick of metal, had serious slow down... but the C-Liner performed OK... Bachmann Light Pacific stalled and slipped entering the first curve and the IHC Mogul was even worse... unable to pull a 4 cars train. Something was happening... locomotives didn't perform well, except heavy diesels which required multiple units. I started to panic to be honest as I'm accustomed to flat layouts with no serious pulling issues. Trevor Marshall's warning against grade and Lance Mindheim poor opinion of them were ringing in my head... Did I made a mistake? Jérôme didn't care that much about that. It wasn't an issue for him, but an opportunity to run locomotives for "real" by controlling the throttle. I knew he was right, but it didn't appeased my fear at all!

The next evening, after a day of doubts, I went downstairs and performed a few other tests to try to understand. I knew I didn't take into account how curves affected my maximum 1.6% grade, transforming it in a 2.7% grade. And I knew my 2.1%, not yet built, would go over 3%. Steam locomotives are finicky and whimsical, it was a disaster that could trash my dream to run the 1950s. I had to test the other locomotives, many of which had poor reviews online when running on 2% grades.

I started with a good old friend, a Bachmann Spectrum 2-8-0 weighing a lot. It ran perfectly and climbed the grade with very little effort, pulling the heavy train... I had a sigh of relief since I absolutely needed that locomotive to perform well. Then, I tried a Spectrum 2-10-2 which generally is considered light and underpowered. It was going to be another disaster yet, it performed quite well, with a little bit more of speed reduction than the Consolidation. It was a good sign since I need this loco to pull long trains.

Then came a Bachmann Chinese 2-10-2 QJ. Performance was average at best because the loco was a little bit too light. With maybe 2 oz, it would be a great puller... but it didn't matter too much. Finally, I tried a BLI Mikado. Once again, performance was relatively satisfying. I feel this loco could benefit a little bit more weight, but it should be fine pulling a 10-car freight train. These results made my day as they proved a good chunk of the fleet was fit to purpose, albeit with some extra tuning.

However, it was clear I would have to rework the grades and make the vertical transition somewhere else than the curve in the yard. At 1% grade + a .0.9% curve factor, it was a 2% grade that was too much... But more on that in a later post.

After the test, I was curious to know what was the pulling power of real similar CNR locomotives. As you may know, there aren't any reliable formulas to equate tractive effort to tons pulled by a locomotive. However, I found a rule of thumb by Al Kurg explained by someone nicknamed Akarmani on MRH forums. The rule isn't that reliable, but he discovered it overlapped several data obtained by different means. I wasn't finicky about it, just wanting an average number, something in the ballpark to see if the model locomotives were completely bogus or close to their real counterpart.

Here's an excerpt of Akarmani's formula taken from MRH forums:

"So far the only "rules of thumb" I have so far is (1) average 6lbs per ton to over come starting friction and average 3 lbs per ton once moving and (2) average 20 lbs per ton per % grade.  

Example: a 2-8-0 with a TE of 42000 lbs would be able to pull (42000/6lbs per ton) 7000 tons on level dry track. A little less that 140 (50 ton) cars after subtracting the weight of the engine and tender, plus the caboose.  On a 1 percent grade the total weight would reduce to (42000/26lbs per ton)  1615  tons.  This is very close to Dave's information concerning an RDG 2-8-0 pulling 1500-1700 tones up 3/4% to 1% grade. It is also close to what I calculated using Davis formula. I came up with 4.26 lbs per ton with the train going 1 MPH on level straight track.  4.26 lbs per ton is between 6 lbs and 3 lbs per ton."

Using this forumula, I could be able to derive what a "real" locomotive on the real Monk Subdivision could pull, then do the same with the layout conditions and see if the number of cars was close to what I saw when performing my test. Mike Confalone did something quite similar when dealing with diesel locomotive tonnage rating VS their horsepower. Here are the preliminary results:

As you can see, the "Layout" numbers are quite close to what I have seen on the real layout. By fine tuning my locomotives and tweaking the grades a little bit, it can be achieved. Keep in mind, that most steam era freight trains on the layout are 10 cars long, with longer ones reaching about 16 cars. Wayfreights with small steamers are generally about 6 cars long. As for passenger trains, the longest ones will have 8 cars. The Light Pacific will generally pull short ones, as it happened on Monk Subdivision, meaning it will pull 3-4 cars, which was fine when I tested it.

The table also gives me idea about which locomotives will need to be improved by adding some weight. IHC 2-10-2 will certainly need some because it's a little bit underpowered. Fortunately, there is still a lot of space in the boiler to do that. Spectrum 2-10-2 is almost alright out of the box. BLI Mikado is at its limits. The tender is lightweight and tracks badly. Also, the locomotive need more weight and I know many people did that to make them perform reliably. Bachmann Pacific, out of the box is too light. It seriously needs some extra weight under the boiler. Domes will be filled with lead at a minimum. The IHC Mogul will also need to be beefed up. I wish to rekitbash it and it will be a good opportunity to make it perform better.

In a next post, I will cover the options available to tackle this issue, namely the grade themselves. Fortunately, I have several options that will help a lot without rebuilding too much of the layout. Once again, using splines on open frame was the right choice for this specific layout.

Monk Subdivision - Changing My Work Ethic

People that know me well will probably tell you I'm lacking in the department of constant effort. I will generally pick up a project, work on it like crazy without counting hours then let it rot until a new shiny thing appear. While I will circle back to that project someday, it doesn't make for a very satisfying work ethic as thing looks promising at first then die off after a while. The strange thing is that I always cycle through my different passions and hobbies in the same way.

Gluing splines is simple, yet a relaxing process...

 All my life, I've followed that manic-depressive way... to the point close friends can tell exactly the moment I'll start working on something and when I will stop... abruptly. It seems that to fight that tendency and get things done, I've developed a habit of trying to get everything done as fast as possible before I lose interest. It generally means accumulating all the stuff and material I need to get the job done and focus only on that thing. For small projects, that intensive way can yield impressive results. Generally, I will be able to work on three projects in a row until I got both tired and, understandably, burned out!

For Monk Subdivision, I tried that recipe once again, believing that having some friends to help me would help the project to reach a reasonable level of completion from which I could no longer back off... Well, so untrue! You can't count on others to fill the gaps when their schedule isn't as flexible as yours... and many things must be done at slow pace to control both quality and learn new techniques. That can't be done on the fly.

Little by little, day by day, the layout comes to life

About two months ago, I started to get angry at the lack of progress. I just wanted to run train on what is a twice around. I didn't care about prototype since it's not the end goal and I explained it on this blog. Meanwhile, I was terrorized to try new techniques like MDF spline roadbed or motorize turnout controls. Worst, I couldn't decide on a staging yard plan. Talking with friends helped to rationalize the track plan and made hard decisions, but things weren't moving because nobody was building the layout for me and I quickly discovered you can't build splines in one day... Something had to change... and it was me.

I counted the number of hours I could lose behind the computer screen trying to make the track plan perfect and tweak the scenery... it was astounding. I could have built the layout completely just using these dozens of hours lost because I thought moving pixels could give real answers. You may not know, but I've reached track plan #92 for Monk Subdivision... we are talking about 6 visible turnouts, a mainline and a passing siding. Then, I counted all these wasted hours... 20 minutes here watching an online video before supper, another 30 minutes here because I felt it was not long enough to start working on the layout. At the end of the week, these numbers were mind numbing. I'm not even a social media junky nor own a cellphone yet I was a slave of the web, i.e., I filled every moments I thought "too short" with useless passivity. Then I got really angry and stopped doing these things as much as I could.

Soldering a few turnouts after work is another small step

First, I discovered my days felt longer when I was not behind a screen. I seriously think time goes faster when consuming data than doing real things. My personal experience seems to indicate times runs about thrice faster. For any adult in the room here, we all know time goes faster as we age... well, now add that 3x factor and you're about to see you life pass by you at light speed.

Second, I saw someone dismantling a huge collection recently. It was accumulated over 30 years for the big dream layout after retirement... Health issues, unexpected moving and other reasons destroyed that dream... The layout should have been built 30 years ago... smaller, clumsier, but improved upon.

Splines make sense when new scenes start to take shape

Third, I remembered Mike Confalone's clumsy but authentic words, that he was unskilled, never planned correctly what he had to do and preferred to make mistakes, correct then and at least move forward. Anybody remotely knowledgeable about Mike's work knows he's accomplished in what he does. If a perfectionist like me who draws precise plans can't do the same, something is wrong. And from that point, I started to think I was better to cut wood, make mistakes and correct them than sit in my armchair.

But that meant a change in my ethics. If you make room for making mistakes, you must also accept things will take time and won't be efficient. Working in intense sessions, expecting perfect results, won't cut it. I knew a lot of unskilled modellers did build more than decent layout, so why should I be afraid.

From that point on, I started to work on the layout in bouts of 30 minutes here, 20 minutes there, sometimes the best part of an evening or a Saturday morning. I had no time limit, just using what was required to reach decent goals and feel the few minutes invested brought me to somewhere meaningful.

Little is needed to make the magic happens...

And I made a lot of mistakes... I first screwed the staging yard plywood directly on the frame only yo discover later it was too low and hard to access. It was frustrating, but I put down the tools, went to do something else, knowing next time I would unscrew everything and raise it to the correct height. Later, when I installed wooden risers for the MDF spline, I found out I didn't know what I was doing. They were too short due to miscalculation and wobbly. Once again, it wasn't fun, but calmly I assessed the issue and worked on it the next day. Surprisingly enough, fixing mistakes was faster than I thought and it helped me to learn better ways to do this peculiar job.

Another thing I changed is that when I meet a road block and felt I no longer knew how to fix an issue, I didn't walk away. I took some time to access what was wrong. In the past, I would have left the room, dejected, trying to overplan and overthing the problem in hope to find a magic answer. This time, if I knew I had still some time available, I would do something else to make sure the project wasn't stalled. It could be installing fiberboard in the staging yard, screwing a few risers or gluing some MDFC splines. The goal was to stay active and make sure the project was moving forward. If tasks were to hard to figure out in one shot, I simply broke them in very, very small parts that could be completed fast so I could claim victory at the end of the day.

Closer to the original vision from sketches

I've been working that way for two months now and I'm surprised how things have progressed on all front. Changing my work ethic made it possible at a more healthy and satisfying pace. Also, I know I will be able to run trains before I retire. If the layout end up not being satisfying, no problem... I will have learned much more building this one. I may hate Tortoise switch machines, but at least I will have wired some, including dreaded live frogs. Not so long ago, each time I would have hit a serious wall, it would have been a good reason to switch my attention to another project. This is no longer the case.

Working on a layout is a leap of faith. You plan as best as you can and wish it will work. Like any plans, reality hits and you have to adjust. The moment you know that and that your objectives are clear, there is no longer stress moving forward and adjusting as you go. I'll be honest though, I've had more than one moment in the last two months when I felt issues I met were quite though to figure out. Doubts take control and abandoning the mermaid call to work on another project can be very strong. I'm glad to have soldiered on because I know track laying will soon happen and maybe, just maybe, trains will start to run on Monk Subdivision before summer.