Friday, October 26, 2018

Rolling Stock Standards

With track laying and wiring completed, the layout is now ready for regular operation under JMRI Operations management, or so we could think. Jumping to the next step (paint and ballast) is  enticing, but the real challenge now is to improve the rolling stock performance. Last year many cars were tested (wheel gauge and coupler height) but many new cars found their way in the collection and certainly don’t meet our standards. Meanwhile, we have settled on a weight standard that better suits our needs.

Why the fleet now standing at 115 cars, my new Excel spreadsheet roster is quite useful in tracking them and monitor which ones must be improved.

I consider three types of mechanical improvements: car weight, coupler type/height and metal wheel/gauge.  Weathering, correct paint scheme and customization/kitbashing can also be added to this list though they don't directly relate with operation. They improve the realism and immersion, which at the end of the day is our goal since we are replicating a real railway. Many cars only require renumbering or relettering, some must be repainted and a very few ones must be kitbashed. Weathering on all rolling stock is definitely a goal but far to be a priority at this time.

That brings me to the heated topic of car weight. I’ve read a lot about it and came to the conclusion I didn’t care about what most people advocate for because my interaction with trains isn’t similar to most of them. We aren't a club, we don't run excessively long train with large motive power consist, we don't have grades nor tunnels nor helixes. The last three elements are among things I consider vastly overrated in this hobby, but this is my own personal opinion based on my goals in this hobby. Your mileage may vary and I certainly would say "go for it" if these things do matters to you.

That said, once again it’s a matter of finding the right solution to the right problem. Someone running very long train on a multi-deck layout with certainly push for lighter cars while another person with a smaller layout (or no helix) could go with heavier cars. Mike Confalone proved that many years ago and I don't see the need for a debate. You must also take in account that steam locomotives are generally poor pullers, particularly smaller ones while diesel are relatively good pullers. Era and locale will have an impact too since older trains were far shorter than modern ones by several factor. Finally, depending on electronics on board (decoder, sound, keep alive), a locomotive can lose a lot of pulling power too. It's surprising how these gadgets can eat up space and remove weight from a locomotive!

In our case, we tried Mike Confalone recipe a few years ago. It means most cars weighed about 8 oz to 11 oz. This is quite a lot. I liked it a lot and it was perfectly fine, but I must admit it was a little bit hard on our locomotives when pulling trains on the peninsula. The 24 inches radius there was a killer and I expect the new 28 inches radius won't make that much of a difference. Since this area is now a yard, this limitation can quickly become a liability.

But in fact, the real problem is that heavy cars are a little bit harder to handle and store with our drawer system. Many cars must rest on their side with their fragile ladders supporting quite a lot of weight. Not ideal in the long run for fragile cars such as tanks. Also, most people are so accustomed to light cars they have a tendency to not grasp them firmly enough. While not accident happened so far, I'm pretty sure it was bound to happen.

Thus we decided to find a middle ground between NMRA standards and Mike Confalone’s one. Using Excel, I created various options including an altered NMRA formula. The classic NMRA formula is 1 oz. per car plus ½ oz./inch of car length. I tried using 2 oz., then 3 oz. I also replaced the ½ oz. with ¾ oz. or 1 oz. By coincidence, I came close to Joe Fugate’s simple formula of 1 oz./inch.

At the end of the day, I felt Joe’s formula was interesting by its simplicity; however, you quickly end up with high numbers with longer cars. Given most longer cars on our layout are flat cars and gondolas, this is physically impossible to implement with normal means. A 10 oz. centerbeam flat car isn't realistic at all if your loads are removeable.

Knowing that, I decided to determine what would be the optimal weight for a standard 50ft car. I wanted a car that was sufficiently heavy to not roll easily during coupling while being light enough to not crush details when stored. The 6.5 oz. to 7 oz. range seems to be right. It fitted Joe Fugate’s formula but also the altered NMRA formula if using 3 oz. instead of 1 oz.

At the end of the day, I settled on the altered 3 oz. NMRA formula. It makes shorter cars a little bit heavier, which is a good thing, while keeping longer cars within a more realistic range. On the positive side, it means less lead is required to weight a car. Given the price of lead wheel weights at auto parts stores, this is much more economical. Also, when handling the car, they feel heavy, but not to the point you feel it's a brick (which was the case before).

So far, the entire 50ft boxcar fleet have an average weight of 6.5 oz. Now is time to standardize other car types, including very tricky closed cars such as kaolin tank cars and covered hoppers. I may not reach my goal with centerbeam flat cars, but I'll try to go as close as I can.


  1. A real 50' IPD-era boxcar weighs ~30 tons empty and ~110 tons fully-loaded. While consistent car weighting schemes improve model operation, I wonder if we're not missing an opportunity by creating variable rolling stock weights to simulate loaded and empty cars? I doubt we could triple the weight of the car between the simulated loaded and empty states, but some difference would surely be apparent, even in model form. Maybe the relative difference is only 1.5x instead of 3x, but it would force modelers to consider placing empty cars at the end of the train (a prototype consideration).

    1. You bring forward a very legitimate observation. While it could be difficult to implement with certain enclosed cars (covered hopper, tank cars, boxcars), it would be quite doable with open cars. In fact, this is partly why I don't think flat cars need to match perfectly the weight formula as long as it is heavy enough to run correctly. The removable load take care of the rest to reach the total weight wanted. As you said, more than once we see real railroads block lighter cars at the end of a train and this can certainly implemented when operating our own models. It add another layer of realism by understanding another very pragmatic aspect of operations. As a kid, I recall gondolas, flatcars and empty open hoppers were always at the end of train on Murray Bay subdivision and it is quite likely to happen too on the layout. Your suggestion of a relative difference of 1.5x seems right to me. Enough difference to be noticeable yet keeping the car's empty weight high enough for improved performance. I'm curious to weight an empty and a loaded Walthers bulkhead flatcar.