Last update January 2, 2015
This project is my Christmas present to myself.
For years I have taken turnouts for granted and it turns out that was a big mistake. I have always thought that Atlas was the King of HO track until recently. My current layout is fast approaching 25 years old and for the last few years I have been experiencing problems with my track work. The combination of 25 year old track and newer locomotives have developed problems in one particular area on my layout. Several years ago I swapped out two twenty year old Atlas Code 100 turnouts and that ended up being only a temporary fix.
I recently added some diaphragms between my A & B E7s and that has made the problem worse. I thought the derailing problems was my Atlas Code 100 turnouts again. For some unknown reason I had thought that the Atlas 850 Remote turnouts were 18" radius and figured that was my problem. Come to find out the 850 measures approximately 22".
It is rather difficult to measure a turnout, they are not a smooth continuous contour. The Atlas 850 is larger than a 18" radius but smaller than a 24" radius. Measuring a turnouts outside curved rail from the moving rail point to frog with the concave distance at the center is the closest way to come out correct as if it was a true smooth continuous fashion.
When the Atlas turnouts are installed correctly (correct radius) they work very good, I error-ed on my original design because I took the Atlas turnout radius for granted. That has turned out to be a major mistake. My design miss aligned the flex track to the Atlas turnouts thinking they were an 18" radius.
The Peco web site has very specific dimensions on all of their turnouts so that will be my new direction. I really like the Peco snap spring point lock design much better than the Atlas. It's a better idea
The Peco turnouts are very close to a smooth continuous curves and calculate correctly like an Atlas curved track section.
I also like the Peco moving rail hinge point, better contact area.
A down side to using Peco is their switch machine, it requires a 2¼" clearance hole under the points. The Peco turnouts work so much better than the Atlas it's worth the extra trouble. Actually it's easier to cut a 2¼" hole than it is to mount the Atlas Under Table Switch Machine, just more sawdust from using a hole saw.
The top view of the Peco rail points look very nice, the throw pin or rod is flush with the throw arm.
My problem turned out to be at the grade/radius transition. I currently have an Atlas 22" turnout at the bottom of my 30" radius 3½% grade transition. All my locomotives will clear the Atlas turnout at a reduced speed.
When I built my layout back in 1990 I didn't want problems so it is bolted and double screwed with lots of Elmer's Carpenters glue. To make things even worse the problem area is in a mountain so removing the old track work is going to be a real undertaking.
My plan is to totally remove a 38" section of the 30" radius track and it's support structure then replace it with a new section. The new section will be built on ¼" thick plywood cut to fit the new 24" radius of the Atlas Code 100 Flex Track and the two Peco SL-92 Code 100 24" radius turnouts to make everything a smooth 24" curve. The first SL-92 will be the turnout from the mainline into my station the second SL-92 will be the turnout that selects one of the two station tracks.
The solid lines on my drawing is viewable track and the dashed lines are tracks hidden from view inside the mountain areas. The thin red line is the upper level track, I made it thin so it didn't totally block the view of the lower level track. The black dashed rectangular box is the problem area, the solid black lines in the problem area is the plywood cut to fit the new radius. As you can see the new curved section will be a smooth 24" radius.
I'm going to build the support structure a bit different this time. I think I have come up with a better way to make grade transitions. This will be my first attempt using this type of construction.
The bottom portion of the drawing above shows the dimensions of an angled wedge cut from a 1" x 2" x 48" pine furring strip. That will keep the grade at precisely 2%. So that the board will bend to conform to the 24" radius I'll slot the inside edge with ½' deep vertical cuts with my table saw every 3" so that the board will bend easy. I will use ½" number 6 wood screws through the plywood into the furring strip to secure them together. If every thing works as planned I'll use some Elmer's Carpenters glue for added strength.
After I worked at removing the 25 year plus old track, roadbed and support structure I was almost totally done in for the day. That was a lot more work than I expected. The demolition took 4½ hours.
After cleanup the mainline down grade into the new curve wasn't exactly what I expected so that afternoon I removed the support columns leaving the ¼" x 4" plywood roadbed for the 60" down grade. I will replace the column type supports with a contoured 1" x 4".
I worked on my CAD for several hours and came up with a new grade profile shown in the drawing below. The contour ended up at 3/16" in 24".
Next is cutting out the new support from 1" x 6".
The new curved section is almost ready to be glued in place.
It lines up perfect with the mainline and the turnout into my yard and station.
New Years Day 2015
The above picture shows the finished track work. The track is Atlas Code 100 Flex held down on to Midwest Cork Roadbed with track nails.
The picture above is a view looking up the grade to the tunnel portal. Looking at the drawing of my layout earlier in this post the red mainline is the outside track in the picture above, the green mainline is the inside track leading into the new curve.
Both of the mainline tracks are on a 3% grade and enter the tunnel portal at the same level then they split directions inside the tunnel. The outside track continues up at the 3% grade to the upper level and the inside track turns into a 3% down grade back to the lower level.
To test my new curve section I ran a complement of many types of locomotives. I ran my most touchy locomotives first. I have two Bowser GS4 mechanisms with Bachmann shells, they are super touchy because of their long driver wheelbase. They ran fine so that told me all three grade transitions were on the money. Next came my Rivarossi Articulateds and they had no problems either. Then troubles popped up with my 6 axle diesels, Athearn PA1s & SD40-2s, Proto 2000 E7s, E8s & E9s all derailed randomly at the Peco turnouts.
After three days of frustration dealing with the 6 axle diesels derailing at the Peco turnouts I finally found out what the problem was. The wheels would just nick the Peco frogs even though they all checked perfect in the NMRA Standards Gauge. The fix was to decrease the flange width by .010" on every axle.
I have a total of 18 Atlas code 83 turnouts and the .010" modification did not have any effect on them. I spent over 8 hours running my entire locomotive fleet, some haven't been run for many years so it was an exciting day to see them in operation again.
Watching my 9 car Daylight Passenger pulled by two Athearn SD40-2 frames with Cary E7 shells running at a constant scale speed of 100 MPH without any problems is a first. At this point I declare my new transition/curve fix a success.
My Christmas present to myself turned out to be more work than I anticipated but it fixed my problem curve so it was well worth it.