track angles

[leesheridan]

I'm working on an HO layout with two layers. In coming down from layer two to layer one on a 24" radius should the track be banked to the inside and if so, what angle? The grade will be about 2 1/2% to 3%. Also this will be my first time with DCC so is there some way that I can set up a sensor that will inform the computer that the loco has started down the grade and then send a speed change to the decoder?
thanks

 

[NYC_George]

The photo below is the last part of the grade that branches off from the helix and the upper deck. The 3% grade and 22 inch radius is not banked and can handle 50 ft cars freight cars with out a problem. My passenger trains use the main line route down from the helix and wouldn’t be able transverse this radius.

NYC_George

 

[UP2CSX]

Mike, welcome aboard. Banking tracks in our small scale does nothing to help curves on grades and usually does more harm than good. You want the track on as flat a surface as possible for any grade. There are no sensors I'm aware of that will do what you want in DCC. However, some DCC systems come with a type of "cruise control", where you set the speed you want the train to maintain and the controller will apply more voltage up the grade and less down the grade in an attempt maintain a constant speed. It works pretty well uphill but a 3% grade downhill is more than the controller can handle automatically and the train will speed up. We have to do some things just like real railroaders. The best approach is to have the train medium slow before it starts down grade and be ready to throttle back if the speed gets too high. If you run short trains, like ten cars or less, it shouldn't be a problem. Long trains will start to exert more gravitational force to the cars in front of them if the train going too slow or too fast, causing stringlining, which means the train tries to pull taut, like a string. In the process, you get a domino effect of cars derailing. Your goal is to keep the couplers with a little slack between them, not so they are tight against the coupler of the leading car. It will take some thottle adjustment with short trains and a lot of throttle adjustment with long trains. After you do it for a bit, you'll pick up the technique.

 

[LotoQuestions]

Superelevation

Superelevation or banking of curved track - I had always thought was a good thing.

Matter of fact I was going to post asking how much for a particular radius and what technique to achieve it.

Dont think grade enters into it but speed does.

thanks ,,, ken

 

[NYC_George]

I did bank my 2 outside horse shoe curve tracks 2% because I thought it would help trains negotiate the turn. When trains didn’t have a problem with the inner third unbanked 19 inch radius track I didn’t bother banking any more tracks. In the 2005 photo of our train club meeting before I added the second deck the 2% bank causes the cow catcher of Jay’s loco to scrape the track. We had to file it down a bit to get the some clearance. In the photo below the reworked cow catcher just makes it.
I forgot to answer your second question. I replaced the Atlas stock 4 function decoders in the GP40’s with Digitrax DH165A0. I had lots of problems with trains over 32 cars pushing the engines down the 3% helix grade. The Back EMF worked perfectly to solve the problem and can be adjusted as needed.

 

[CSX Robert]

Our trains generally do not travel fast enough for super elevation(banking) of our curves to make a difference(at least not in the smaller scales - HO and smaller - I'm not sure about the larger scales). Many people do super elevate there curves for aesthetic purposes, and if done right, it does look good. Super elevating your curves that are on a grade can cause problems because it will increase the chances of a train traveling uphill to "string-line"(pulling the cars off the inside of the curve).

... Also this will be my first time with DCC so is there some way that I can set up a sensor that will inform the computer that the loco has started down the grade and then send a speed change to the decoder?
thanks

There are some possibilities here. If you consider any of these possibilities, you have to also consider whether or not you have a separate track for uphill and down hill because you would not want to slow the uphill trains. You also have to consider what DCC systems and what decoders they would work with

One solution would be to drop the voltage on the downgrade track using diodes. If the track is used in both directions, you could include a switch to bypass the diodes, but that would not be automatic. To make it automatic, you could use a latching relay to bypass the diodes with a sensor at each end of the grade to trigger the relay(bottom sensor bypasses the diodes for uphill trains, top sensor switches to the diodes for downhill trains).

From here, the solutions get more complicated. Using JMRI and Digitrax transponding, you could have the computer adjust the speed table or forward and/or reverse trim settings of the locomotives on the fly at the start and stop of the grade.

There are a couple of solutions that require passing off control of the locos from the walk around throttle. Again using JMRI, you could have the computer controlling the train from one level to the next and automatically adjusting the throttle setting according to sensors. With Digitrax, you could use a LocoShuttle to control the train between levels and have it automatically adjust the throttle according to the sensors.

Lenz's Asymmetrical DCC will be ideal for this sort of thing, but it is still i it's early stages of availability.

 

[lmackattack]

on my tight radius curves I placed plastic shims under the outside rail. just to give some of the 6 axel and large steam that extra help to get around safely. I have found that just a small. 030" shim did alot to reduce/elemanate derailments. However as Robert said, it will also reduce the amount of cars you can pull because it will allow them to string line a little more easy. I only pull 10-15 car trains so it did not effect me and as others have said you may get alot more than 15 cars before you start to have trouble. the grade is the real issue as gravity will not be working with you in both ways that you dont want them too. I have about a 2.5 grade with a 22" smimmed radius thats about 210 deg curve. I have pulled 40+ cars with 2 engines without much trouble. I do tend to see that light cars need to be at the back of the train after about 20 cars. going down grade is not an issue for me.



Trent

 

[DaWolf85]

on my tight radius curves I placed plastic shims under the outside rail. just to give some of the 6 axel and large steam that extra help to get around safely. I have found that just a small. 030" shim did alot to reduce/elemanate derailments. However as Robert said, it will also reduce the amount of cars you can pull because it will allow them to string line a little more easy. I only pull 10-15 car trains so it did not effect me and as others have said you may get alot more than 15 cars before you start to have trouble. the grade is the real issue as gravity will not be working with you in both ways that you dont want them too. I have about a 2.5 grade with a 22" smimmed radius thats about 210 deg curve. I have pulled 40+ cars with 2 engines without much trouble. I do tend to see that light cars need to be at the back of the train after about 20 cars. going down grade is not an issue for me.

I hear masking tape works well too, when stacked on top of itself, it can give a nice banking.

 

[Jon NFL]

I had a curved section that was causing a RS27 to derail every time it went over it but only in one direction and then only going forward, not in reverse. After adding to my baldness and correcting a very slight joint alignment problem, repositioning the pickup wires on the front truck that seemed to prevent a free swivel movement, it was better. The engine would pass at slow speed but not full. I walked away and came back to it the next morning. All progress had been lost – even at a crawl it jumped the track at the same point. The section was too restricted to get a level in there. The idea of a pendulum level started formulating within the old baldhead. Out of a piece of 1/8 PVC, 2 pieces of piano wire (one bent and glued to the longer one), a piece of brass tube inserted through the PVC and a little filing the bottom for alignment, the level was made. The pix show that the track was dipped to the outside just a little – but enough to cause the derailment. I pried up the track and forced some ballast under the outside edges. Problem solved, even at full throttle.

 

[UP2CSX]

Jon, that's a very novel idea to solve a really common problem. Most curves, assuming they are really level and the joints are smooth will wirk fine, Determining that they're level is the hard part, and your curve level indicator is a neat trick.

 

[DaWolf85]

Jon, that's a very novel idea to solve a really common problem. Most curves, assuming they are really level and the joints are smooth will wirk fine, Determining that they're level is the hard part, and your curve level indicator is a neat trick.

Yeah, I'm gonna make me one of those handy little things.
This is why I use this forum, for ideas like that.

 

[Larry]

I elevated a bunch of my long curves about 5 years ago w/a business card thickness about 1/4" from the outside edge of the ties. It worked fine on most engines & most long trains. Not on long trains w/lite cars in the middle. Also had problems w/long diesels derailing. The only reason I elevated them was, friends said it would look better.
Well, after about a month of a lot of problems I flattened all of the curves back down to level. Solved all of the problems I was having. Some of the curves were on a grade & some were on level area's. Still had problems on both of the elevated sections.
You will still have some problems w/lite cars in the middle elevated or not.
This layout I didn't try to elevate anything, although people still ask me why I don't elevate the curves like the real trains & I tell them it's not worth the problems.

 

[rem37411]

Hi, is a 3 percent grade too much for an N scale train? Probably a little longer train like 12-15 of the short coal cars?

Ron

 

[UP2CSX]

Ron, 3% is getting steep, especially if any of the grade is on a curve, since flange friction increases the actual tractive effort needed to overcome the grade. N scale makes it a little trickier yet because N scale engines don't have the mass of larger scales and tend to slip more on grades. With a single engine, a train length of 8-10 cars is probably realistic. A 15 car train will most likely require two engines to pull the grade.

 

[Western Star]

Also this will be my first time with DCC so is there some way that I can set up a sensor that will inform the computer that the loco has started down the grade and then send a speed change to the decoder?

Yes there are many ways one could do that. The easiest might be to purchase decoders with "back EMF" function and enable/configure that property.

Personally I don't use that option because that is the job of an engineer - to throttle back before hitting a down grade.

 

[rem37411]

Ron, 3% is getting steep, especially if any of the grade is on a curve, since flange friction increases the actual tractive effort needed to overcome the grade. N scale makes it a little trickier yet because N scale engines don't have the mass of larger scales and tend to slip more on grades. With a single engine, a train length of 8-10 cars is probably realistic. A 15 car train will most likely require two engines to pull the grade.

That is actually kind of what I wanted to hear. I want to run a two engine train anyway. I think I will settle for middle of the road and go 2.5% grade. Thank you for the input though. I am SO itching to start laying rail!

Ron