G Scale Model Train Forum banner
1 - 17 of 17 Posts

·
Premium Member
Joined
·
1,704 Posts
Discussion Starter · #1 ·
I have a problem with a turnout causing some cars to derail. After investigating I can see the reason. As the cars approach the turnout they are coming out of a 9' diameter curve. The turnout is a left hand #6 and cars are coming at it head on.

When the points are thrown for the divergent path, what I see is that since the cars are coming out of a curve the wheel flanges are pushed all the way to the rail where the point is touching the stock rail. The wheels hit the very end of the point and bounce up, causing some to derail. This occurs even though the points are recessed slightly in the rail. These are sunset valley switches. I see now I should not have a curve immediately before a switch, but to modify the layout now is more time that I have.

So I was considering a couple of options and wondering if anyone could say one of these is better or another solution that might work:

1 - grind down the end of the point to be even thinner than it is.
2 - grind down the rail so the point fits inside it further.
3 - add a piece of thin brass just before the point to help guide the wheels around the point.

I see a problem with each of these:
- The first, I may grind the end of the point down too far and cause a problem?
- The second, with the rail ground down, the point will sit inside it better, but now when I throw the point for the straight path it may not go over far enough without modifying the points to be further spread apart?
- The third, not sure if this will work or just push the problem back to just before the point?
 

·
Registered
Joined
·
3,538 Posts
Posted By jimtyp on 07/03/2008 9:46 AM
I have a problem with a turnout causing some cars to derail. After investigating I can see the reason. As the cars approach the turnout they are coming out of a 9' diameter curve. The turnout is a left hand #6 and cars are coming at it head on.
When the points are thrown for the divergent path, what I see is that since the cars are coming out of a curve the wheel flanges are pushed all the way to the rail where the point is touching the stock rail. The wheels hit the very end of the point and bounce up, causing some to derail. This occurs even though the points are recessed slightly in the rail. These are sunset valley switches. I see now I should not have a curve immediately before a switch, but to modify the layout now is more time that I have.
So I was considering a couple of options and wondering if anyone could say one of these is better or another solution that might work:
1 - grind down the end of the point to be even thinner than it is.
2 - grind down the rail so the point fits inside it further.
3 - add a piece of thin brass just before the point to help guide the wheels around the point.
I see a problem with each of these:
- The first, I may grind the end of the point down too far and cause a problem?
- The second, with the rail ground down, the point will sit inside it better, but now when I throw the point for the straight path it may not go over far enough without modifying the points to be further spread apart?
- The third, not sure if this will work or just push the problem back to just before the point?



1. Just use a fine file and take your time. You want to run your fingernail along the rail and not have it "catch" at the point.

2. Not a problem. The intent here is that with the "spring action" of the turnout throw/motor, the point can tuck a little further into the rail so the wheel makes a smoother transition. When the turnout thows the other way, this will have no effect. In other words, it doesn't "shift" everything over toward that side and away from the other.

3. Another way to implement #3 is illustrated below. I did this and it helped get my heavy weights through a turnout. But even this was not 100% and I ultimately put in a 6" piece of straight track on one side of the turnout that totally fixed the problem.

 

·
Premium Member
Joined
·
5,334 Posts
I had a simalar problem a few years back....The source of the derailment was that I had not set the Gauge of my wheel sets proper.

That is when I came of with the idea of a Gauging tool made from a "C" clamp.
 

·
Registered
Joined
·
1,579 Posts
1) What type of turnout?
Manufacturer?

2) ALL the cars, or just selected ones?
If selected cars, what EXACTLY is the back-to-back on the wheelsets?

Step ONE is checking and correcting wheel gauge.
Often, the "fix" is not the point of the frog, but rather the guardrail.
If the wheels are fine, and the guardrail is not holding the wheels AWAY from the point of the frog, that happens.

Several fixes, plastic shim, brass shim inside the guardrail, or whack the guardrail out and install a new metal one, super-glued down, pin-vice drill the ties and push spikes in with super glue (or, add a piece of wood in the hollow under the ties and spike through to that).

Wheels 1.575", + or - about .010". Check gauge (wing rail of frog to guard rail) about 1.535".

TOC
 

·
Premium Member
Joined
·
1,704 Posts
Discussion Starter · #5 ·
Dave,

1) Sunset Valley #6

2) Most cars coming out of the 9' diameter curve hit the end of the point and bounce up, some bounce enough to derail, even at low speed. It's generally the lighter cars, the heavy ones seem to be able to hit the point and stay on track.

3) Wheel gauge is correct, checked. I should have explained better, it's not the point of the frog but the point itself, the very end where it sits up against the stock rail. So it's the very beginning of the point, way before the guardrails.
 

·
Premium Member
Joined
·
687 Posts
Jim

You were not quite clear ... is the curve coming into the points of the switch turning the opposite way to the divergining route? In other words a reverse curve without any straight track in between?

If so the only solution is to pull that #6 switch for use elsewhere and back the switch up in the curve with a curved switch.

If the curve is not a reverse curve but is turning the same way as the diverging track in the switch, then you should be able to adjust the points to make it work

Regards ... Doug
 

·
Super Modulator
Joined
·
21,610 Posts
Guys, he said the point, as in the point of the point rails, the points...

Not the frog point!

Jim, normally your #1 is what you need, since the stock rail should already be notched enough... you are correct that notching the stock rail too much will affect operation in the straight through direction.

After being sure your wheels are gauged properly, then check the gauge of the switch at the points.

By the way, please let us know what specifications you use for your wheel gauge and track gauge at the points.

If you really want to get confused, take a look and my TRACK & WHEEL STANDARDS page on my site, it's under TRACK ....

Regards, Greg
 

·
Premium Member
Joined
·
1,871 Posts
The Sunset Valley turnouts have a small notch cut out of the point rails where they meet the stock rail. On a properly working turnout this should slip under the top portion of the stock rail making a very smooth transition. I just looked at a couple of mine and ran my finger over it. I could barely feel the transition. Maybe the turnout in question is not seating properly. Could be the screw that holds the point rail to the throwbar is too tight or maybe too loose. As has been said filing the end of the point rail a little might also cure it.
 

·
Premium Member
Joined
·
1,704 Posts
Discussion Starter · #9 ·
Doug, Greg and Paul, thanks for your input. I’ll try to answer your questions.

Yes I have a 9’ diameter curve that comes into the point rail end of the switch (so head on â€" not trailing â€" to the switch), but it is not a reverse curve â€" that is the curves go the same direction so it is not an S curve so to speak.

I'm using the metal AristoCraft tool for track gauge min/max and the wheel gauge min/max. It's the one that has the levels built in. The switch rails and wheels are in gauge according to that tool. I haven't had the time yet to try #1 above but that is what Sunset Valley said to try first also.

Yes the point rail slips under the stock rail a bit when in position to take the divergent path, but there is still a thin part of the point that sits next to the stock rail and if a wheel/flange is pushed all the way over the to that rail I can see and hear it slightly hit. The cars having issues are the AMS freight cars. Running the heavier AMS passenger cars through the switch didn’t reveal the issue. I did check the passenger cars after having issues with the lighter freight cars and saw they also can slightly hit the end of the point but continue through, I’m assuming because of their weight? The lighter freight cars derail consistently when a flange hits the very end of the point rail. Since the flanges on these cars are not very deep, more prototypical, and the cars are lighter, a slight lift of the flange is enough to push the flange over the rail and derail them.

I would have preferred a switch design so the points fit entirely into the rail (i.e. indented) then no chance for a wheel flange to pick the point. Sunset Valley sort of did this on the straight path side of the point rail. On that side I can see the stock rail is actually indented slightly making it virtually impossible for a wheel flange to hit the end of the point rail on that side.

The saga continues …

I didn’t have time to resolve the wheel flanges hitting the very tip of the point rail so I decided to run the train in the opposite direction. This way the switch points are now at the trailing end as the train comes through the switch. However, switching the train’s running direction now makes my other switch a problem. That switch has a long straight track before the switch so I don't have the issue of the wheel flange hitting the end of the point. However, on this switch, as the wheels come through on the divergent path of the switch I see the wheel flange hit the very tip of the frog. At first I figured the wheels or switch must be out of gauge, how else could the flange hit the frog point? Measured the wheels and found they are right on, although at the max in-gauge width. So then I measured the switch gauge, right on again. I can see now that the gap between the stock rail and guard rail furthest from the frog is the issue. The gap is too large and allows the opposite wheels to move too far into the frog.

The guard and stock rail across from the frog butt up against each other at their base, but this still allows for a significant gap between the tops of the rails where the wheels/flanges ride. I remembered Doug once said guard rails should be about 2 flange widths from the stock rail. Now I see how significant that is. A better design for the switch would be to reduce the base of the guard rail before butting it up again the stock rail. I assume the easiest remedy at this point is to do what Todd did above, which is add some brass strip or something to the side of the guard rail reducing the gap between it and the stock rail. This will keep the wheel flange closer to the stock rail and in turn keep the opposite wheel flange from hitting the tip/point of the frog.

I can see why some folks call switches a necessary evil :-( The thing is there are easy fixes for these issues, so I don’t understand why manufacturers don’t implement them?
 

·
Premium Member
Joined
·
687 Posts
Jim

I don't have an SV switch to look at as I use Llagas Creek. But, imho, the rails on both sides should be notched just enough to recess the switch point fully and allow for a smooth transition into the switch. Since uou do not have an S curve, if your wheels are gauged correctly, and the track at the switch is in gauge and side to side level, this can be made to work reliably. You seem to have checked everything else - is it possible to file slightly the notch in the offending rail to deepen it and recess the point fully?

Guard rails at the frog are essential to keep flanges from hunting the frog and derailing. unfortunately, the lack of a standard for wheel profile makes the chackgauge standard invalid. The traditional LGB elephant fat flanges led to a very sloppy checkgauge - and that sloppiness is troublesome with semi scale wheelsets. Tightening the checkgauge with a strip of brass or maybe styrene will work (provided that noone shows up with LGB flanges that are too fat) as will relaying the guardrail (provided that the rail base is not so wide as to prevent tightening the checkgauge).

Regards ... Doug
 

·
Premium Member
Joined
·
1,704 Posts
Discussion Starter · #12 ·
Doug, Yes, I believe the point can fully be recessed without too much trouble, that was my option #2 above. The problem is then the switch may not go far enough back the other way, so I could have problems on the straight path. Unless I modify the points to have a further spread, as my switch stand is set for a specific point travel length.

Paul, guess I just got lucky :-( Glad to hear all is well with yours.
 

·
Registered
Joined
·
170 Posts
I just went through this with the very same layout ( 8' curve into a left switch), with a #6 USA switch. The point does not indent into the rail (bottom flange of the rail only) - the tip is very fine and if tight against the rail, no problem. However, I found that the switch solenoid was prone to let the point drift away from the rail slightly, especially when the actuator assembly was hot from the sun. Tweaking the action of the actuator spring (and painting the actuator covers aluminum) fixed the problem with the point not staying close to the rail, but it would still pick wheel sets. I found that if I VERY SLIGHTLY bent the tip toward the rail, then the problem disappeared. Did I say VERY SLIGHTLY? My thinking is now, when the flange hits the point, the point is at a very slight angle to the flange, causing the flange to press the point against the rail, rather than picking it. I've had to do this on the odd N gauge switch too.
 

·
Premium Member
Joined
·
687 Posts
i just checked the Llagas Creek switches I use and all have significant indents to receive the points against the stock rails. I use manual switch throws that have "unlimited" travel so I have never had the precision problem that is referred to here ... this is achieved by either using a spring arrangement between the switch throw bar and the throw mechanism or by using something like a choke cable. I cannot comment on automated switches as I have never used them.

In the absence of being unable to increase the throw of the points, then Paul's solution seems to be a sensible approach.

Regards ... Doug
 

·
Premium Member
Joined
·
1,704 Posts
Discussion Starter · #15 ·
Thanks to all for their input. Finally got the issues resolved and the trains are running through the switches with no problems; not one derailment in 2 days of almost non-stop running :) I ended up filing down the very end of the point rail that sits next to the stock rail for the divergent path. Now the wheel flanges have almost nothing to hit :) The point on the straight path rail was already hidden well so no need to do anything there. Also, put in some extra brass strip on the guardrail next to the stock rail for both the straight and divergent paths, as posted in the picture above by Toddalin. This keeps the wheels close to the outside stock rail, thus ensuring the wheel flanges using the frog don't pick the point of the frog.
 

·
Premium Member
Joined
·
1,871 Posts
Jim,
There is one more thing that you can do with the Sunset Valley turnouts when your turnout control does not have quite enough throw. There is really more space then needed between the point rail and stock rail on the open side. I simply remove one of the screws,drill a new pilot hole right next to the old one and resecure the point rail. This will reduce the amount of throw needed by about 1/8". Sometime makes all the difference in getting a ground throw to work good.
 
1 - 17 of 17 Posts
Top