Railroad Engineering

Track-Laying Questions

Click on the Question Number for answers
1. __ Is it best to use a three-point gauge on curves to widen, or two-point?
2. __ What's the best way to get wood ties level and straight?
3. __ Is there a way to get the rail smooth?
4. __ My spiked code 70 came out all irregular and bumpy... What happened?
5. __ Please send me or point me to correct specs. for track gauge.
6. __ Is this statement based on empirical evidence, or some theory?
7. __ I'm working on the drawings and template to build my own N scale switches.
8. __ Is it OK to use Flex-Track?
9. __ I've weighted all my cars to NMRA but what is three point bolstering..?
10. _ What is the best weight for my rr-cars?
11. _ What about a Kadee spiker for laying rail?
12. _ What's a good source of small track spikes for code 55 rail?
13. _ I can't get my locos to run smooth and stay on the track.
14. _ Will N code 55 work as HOn2-1/2 track?
15. _ What is a frog angle and where is radius measured too?
16. _ What's the difference between regular and DCC friendly turnouts?
17. _ Do Rotor Motor switch machines support a frog polarity control switch?
18. _ Given the choice of plastic or metal, is the metal wheel better ?
19. _ Your Turnout is a little too narrow for the NMRA track gauge to pass through.
20. _ How do you actually make laminated roadbed? Do you need three hands?
21. _ My loco stops on frogs when I run slow. What's wrong?
22. _ Will cork road bed on blue foam, deaden sound?
23. _ Whats the advantage of using non-insulated switch frogs? Are they difficult to wire up
24. _ What do you use for throw bars that holds up?

QUESTION: #1 When handlaying track, is it best to use a three-point gauge on curves to get a slightly wider gauge, or is it better to use two-point gauges to get the exact gauge?
Here's what we've found:
The NMRA standard has you laying your rails over an inch (prototype) wider than the wheel sets. This means you already have enormous space for play. Curves ... straight ... whatever..
Now if you use a three point gauge and widen the rails just a bit in the curves, your just increasing the slop. If your laying your curves sharper than 18 inch radius, then use the three point. But if you are using any radius larger than 18, and you wish to run your trains forward and backward with no length/weight limit, then I have by absolute experience found that you don't want to widen anywhere.
As a matter of fact, you can narrow the track about .005 to .010 through the frogs, and you will get much better performance.

In 1980 at the NMRA convention at Disneyland, I took a portable railroad (HO) set it up and ran 65 car trains backward with three four five units. Some on front.. ... some pushing ... etc. with my back to the display for six (6) days with absolutely no derailments using the narrow track standard to prove once and for all that the wider gauge of the three point gauge just causes derailments. That's all it does. Period.
By the way, I used both Athearn and Atlas units running together with no problems. If the track is the right gauge, you have fewer problems. Please understand that I don't care what gauge you use and I make the two point gauges for those who want them. At 4.95 for four of the gauges plus the time to package and send them, believe me I don't make any money from the gauges.

One last time ....... Don't widen your track anymore than you have to. There's already too much play. Way more than is needed for your curves. (by the way, the same ratio is in N scale as well)


A little insider info: John Allen didn't use a track gauge.
He used a wheel set for gauge and his
trains ran very well.

I have done all the research to prove this both theoretically as well as physically. If you want to reduce your derailments..... this is one way that works well.
Happy Railroading. I hope this helps some of you.

QUESTION: #2 What's the best way to get wood ties level and straight?

First of all, I prefer to spread about 4 or 5 inches of Elmer's glue (yellow aliphatic preferred over white) and then using a #11 exacto as a spear, place the ties individually on the roadbed.
Spear each tie .. place it .. then hold it down with your other hand as you pull the blade out. Now do another 4 or 5 inches. Be careful to place the ties on their flat side. Most ties are wider than they are tall.

I've never had much luck getting the ties flat and down snug using the strip tape method. The ties pull up too much when you try to remove the tape. Also there is no point in placing the ties in a jig to tape them. You might as well place them on the roadbed in the glue. Saves a whole step.

After the glue has dried, sand the tops of the ties lightly to get them all even. Hold a ruler on edge across the ties and sight the gap to see where you need to sand. Touch up the tops with stain.

QUESTION: #3 Is there a way to get the rail smooth?

Just like the Prototype Railroads, your track will come out much smoother if you pre-curve or pre-straighten each rail before you lay it.
Also take a file and lightly kiss the corner edges of the end of the rail to get rid of any irregularities. This makes a better joint.

QUESTION: #4 My spiked code 70 came out all irregular and bumpy... What happened?

When your pushing in your spikes, be careful that you don't push them in too tight. The rail will end up roller-coaster..ish. If you get down next to the rail and look at it sideways, you'll see that the rail humps up a little in-between the spikes. Try easing off on the spikes to get the rail level.

Don't over do the tightness of the spikes. As a matter of fact, the spike only needs to be up against the base of the rail to hold it in gauge. Then snugged down to the top of the base without bending the rail down.
Most spike heads are too large and they push against the web of the rail if installed against the base. So turn the spike sideways about 45 degrees to allow the length of the spike head to lay alongside the rail rather than pushing it out of gauge.

QUESTION: #5 Please send me or point me to correct specs. for track gauge.

My problem in answering you is this:
The list for all gauges is quite extensive.
Do you want NMRA standards (which will derail your trains)?
or do you want specifications that work?

You see NMRA had two standards committees. One did the track spacing and the other did the wheel spacing. Now since the wheels on locos and such are very difficult to change it seemed best to change the distance between the rails especially through turnouts to match the wheel gauge provided by NMRA.

By the way the heads of each committee hated each other (jealousy) and never spoke to each other. Most commercial track including turnouts are a bastard compromise between these two dimensions.

What scale do you wish accurate specs for?
I/I'll be happy to answer.

QUESTION: #6 Is this statement based on empirical evidence, or some theory?

Many years ago (1947) I wanted to be a member of an rr-club that only accepted adult members. At 7 yrs/old they let me clean up and sweep and lay track on the O-gauge and On3 but only after I read all the spec sheets that they had in the library from Santa-Fe. When I was taught (by the On3 Guys) to read a micrometer, in typical 7 yr/old fashion I went around checking everything. OK!
The proto specs (I later (at 30 yrs old) got the specs from UP) from Santa-Fe and Union Pacific Show a 4' 8&1/4" maximum through the frog. They lay all track at 4' 8&1/4" and as it spreads to 1/2" go back and regauge.
I noticed then as now that the wheel standards in all gauges as put forth by the NMRA were narrow. Therefore the only way to lay the track (at least through the frog) was narrow. The theory in engineering is that when pushing a chain through a trough the closer the tolerance the easier the push. If the wheel sets are held straight for an inch or so before and after the frog, they simply can not pick.

Rather than go through the attempt to regauge everyone's wheel sets, I opted for gauging to the existing wheel dimension.

Lastly, over forty years and at least one layout in virtually every scale (even TT) I found that derailments were non existent with this tighter gauge through the frog.

There is so much slop in the "NORMAL" NMRA gauging that the wheels can turn slightly sideways and scrub the rail edge looking for flaws.

QUESTION: #7 I'm working on the drawings and template to build my own N scale switches. I have used the table from the civil engineering manual along with the NMRA gauge for frog and heel clearances. Will this be a reliable turnout, or should I follow another path. I want to end up with decent looking and running code 55 and 40 turnouts. Thanks for any suggestions you may offer.

My only suggestion would be that we use .335 to .340 as the gauge through the frog in N with excellent results. Remember the optimum performance occurs when the wheels touch nothing as they traverse the frog. Each wheel set holds the other in alignment as they cross the frog as long as there isn't a lot of lateral play between the flanges and the rail. With too much play, the wheels can turn slightly sideways and scrub the rail looking for flaws.

The length of a turnout isn't critical, however the best performance for both the prototype and the model is achieved by finding the perfect radius from the tangent (stock rail) to the dissection (frog) of the parallel.

Since frog angles are known (#6 = 1 in 6 angle) etc. and the parallel rail spacing for each gauge. Simple geometry will give you the optimum lengths.

QUESTION: #8 Is it OK to use Flex-Track?

Most flex track is a little bit wide and as such, isn't optimum, but it will work fine under most circumstances. We obviously don't recommend flex turnouts since the inherent "WIDE" gauge generally causes more problems than most of us want to deal with.

On the ends of the joints in flex track (hand laid is better) be sure that you narrow up the last ends just a bit. This tighter gauge where the rail joints are will help to eliminate derailments. Remember to kiss the corners of the end of the rails with a file to remove any unwanted burrs before you lay the track.
Hard to do after the track is laid.

Stay away from section track. It's too wide and all those joints YIKES! Remember the real railroads have gone to ribbon rail to eliminate joints.

QUESTION: #9 I've weighted all my cars to NMRA specifications, I feel that with this weight, all my cars are too heavy. Maybe I am wrong with my conversion to the metric system. But I am thinking of a small Sunset C16 with 10 cars. Hmmmmmmmmm...... Maybe I need a double or triple header.

Whoa! ....... hold on a minute. Please hear me out. I have argued this point for forty years.

1. weight is your enemy ... not your friend (prototype or scale)
2. fix the wiggle instead of throwing lead at it.

Imagine the full size railroads putting more weight in a car because it didn't stay on the track. The NMRA is just plain wrong on this point. Idiotic at best. Treacherous at worst. If someone tells you the models are different than the real thing, you just laugh and move on.
Find a competent engineer to agree with adding weight as a "fix". Won't happen. Now to the problem.

On the prototype, the truck swivel (against body) is free floating not tight. The side bolster on the truck rides on the body bolster to keep the car upright. The springs allow equalization movement to keep all the wheels on the track. This is impractical on the model but we want to simulate this effect.

So on one truck screw, make it loose so it will flop in all directions. On the other end of the car, put two little shims on either side of the car bolster that will ride on the truck bolster. The truck can rock forward and back but not side to side. Leave this screw loose also. Now the two trucks are free to follow the track up-down-whatever and the car rides smooth and stable on the two side bolster shims. No weight needed at all.

If the car still wobbles, look for something (dirt, flashing) on one of the wheels.
I have used this principle for 35 years of HOn3 railroading. I can run plastic unweighted flats at the head end of a twenty car drag. Anywhere in the train I want to. WITH NO WEIGHT AT ALL.

When you get one of these Carter cars, you'll see this bolstering, since I'm milling it into the bottom of the cars.
Please, Please .... no additional weight. Fix the problem for once and for all.

At the very first NG convention in Denver many years ago, I set up a demonstrator railroad, HOn3, and ran 15-20-25 car trains with one or two c-16's for the entire convention with none of the cars weighted. I do the same in Sn3 with this bolstering.

If I've been a bit confusing here, please ask me and I'll try to explain it better. I really want to help you here .... as well as anyone else who's interested.

QUESTION: #10 What is the best weight for my rr-cars?

Thirty years of running on a club layout with track that ranges from superb to wretched, using trains of every conceivable source, of lengths from caboose hops to forty to fifty cars long, with cars from real shorties to modern piggy back and high cube boxes suggests that individual car weights are not really all that important- as a generalization!

What is critical is the trucks, wheels, and that a given train be fairly consistent in its weight distribution.

Bad trucks will put a car on the ground faster than anything else. Trucks that are warped, or in which the wheels are a sloppy fit can derail on flat, straight track. Trucks that don't pivot freely about their center will derail when the track changes direction. Sprung trucks can be a devil if they aren't free to equalize, since (Murphy's Law) they always seem to bind when they are not level.

Then there are all the myriad of wheel problems. I have seen some nice cast wheels that had a casting gate on the edge of the flange. These little warts would catch switch points, even rail head on curves and derail. But they were almost imperceptible to the unaided eye.

Wheels that are not at right angles to the axle and parallel to each other don't track well. Wheels with straight, deep flanges and sharp flanges are more prone to find track flaws and derail.

But overall, a good set of trucks with decent wheels will track better through almost any trackage, while perfect track will not help bad trucks.

One of the biggest problems is Athearn trucks. Nearly all of them have the bearing holes too deep for the length of the axles, and many of the third party wheels have axles even shorter than stock Athearn. This lets the car sit low on the axles (this causes coupler height problems) and also lets the car rock sideways, shifting the center of gravity to one side and this in turn can cause cars to just fall off the track on curves, particularly if they are super-elevated.

A great many trucks come with either wheels that are off center or where the metal axles are ground off center. Many of the wheel sets on MDC and Athearn have the latter problem. This makes the car wobble badly but of itself doesn't seem to affect tracking very much.

Weight is important, and the formula relating length to weight is useful, but if you have decent track, and the wheels/axles/trucks are good and properly installed, it isn't that important on most layouts.

I have rarely added weight to cars beyond what was supplied and can think of few cases where operation required I fix that. Those gorgeous Detail Associates GS gons are one exception- they really are too light to run in a train, even at the end. Some wood cars have the same problem.

The old NMRA standard was arrived at by testing and does work, and will work well if you have a wide range of rolling stock that needs to be freely mixed. But most modern (in the market sense) equipment will work with less weight because most of us tend to have most of our equipment pretty much from one era. I have recent plastic hoppers mixed with old cast Zamak hoppers, and I'm sure a few others have similar situations. But we're an exception.

If you are having tracking problems, added weight may indeed cause some improvement by forcing poor trucks to better conform to the track. But replacing the trucks might well do the same thing without adding weight.

Try, experimenting and use what works.

QUESTION: #11 I was looking over some of the information on your web page, and I'm curious how you view the kadee spiker? As a note of comparison, Russ Simpson specifically says to use a kadee spiker voids the warranty.

I don't use it. It bends the rail down at the spikes since it hits so hard. If it's set light.. then some of the spikes are left sticking up. So we have warped rail and sticking up spikes. Not good for operation. If you'll accept the unrealistic look of Kadee spikes then just use flex-track. Much better operating trackage.

QUESTION: #12 Could you suggest a good source of small track spikes for code 55 rail? I have tried micro-engineering small spikes, but the head's too large..

Actually the best is to make your own. In the evening sitting 'round the ole' tv or where-ever. Take some 15 thousandths spring wire (.015)(usually in the K&S rack at your local hobby shop).

1. Bend a 90° angle 1/4 inch from the end.
2. put your diag cutter next to the bend away from the 1/4 inch
3. snip ..... one spike.
4. repeat .... they go real fast and you'll have great looking spikes
and you made them yourself. (I've made tens of thousands of them)
They push in just like regular spikes.

Make them a little longer if your roadbed is soft.

QUESTION: #13 I have a Sunset C-16, a Grant Line Box Cab, and a MDC 2 truck Shay which I can't make run right.
It doesn't matter how great the rest of the layout looks if I can't run trains.


You've just touched on the main culprit in Model Railroading. I don't have time for an in depth answer at the moment, but I'll get you started here.

Your right that it makes no sense to do all the work of a pike if it won't run .... and believe me .... for one reason or another .... most don't run.

There are many reasons for a loco not running smoothly ... regardless of the motor or gear combo ... but the main reason is that the crank pin centers are off, not matched, in the drivers, and they create a slight bind in the mechanism.
When these pins are off, all you can do is round out the holes in the side-rods a little and "hope" that the additional play will allow the drivers to turn a little smoother. I have built my own drivers on occasion because the play was unacceptable. Also check the quartering of the drivers. This is almost always a little off. (further explanation later if you want) It is best to check this with one of NorthWest Shortlines quartering fixtures. I have one modified just for HOn3 drivers.

Also check the gear mesh. If the gears, worm and gear, are too close, this will cause binding since the gears are NEVER perfectly round. They are always a bit egg shaped. Again NWSL's gears are much better than most.
Don't make the mistake of thinking that changing the gears and motor will fix a bad running locomotive. You need to fix the running gear first and then check the performance of the motor-gear combination.

You'll need calipers to check the spacing of the side rod holes and the frame axle holes, as these need to be the same. They're usually off just a little. Let me know how you do with this stuff and I'll answer any questions you have.

QUESTION: #14 Can the code 55 N scale rail be used to simulate HOn2 1/2?

Code 55 is code 55 no matter what "scale". In other words code 55 is exactly .055 tall. The scale doesn't matter. I use code 55 in Sn3 and it does just fine.

In S code 55 represents 55lb. rail
In HO code 55 represents 75lb. rail
In N code 55 represents 130lb. rail
but in each case the rail is still .055 (fifty five thousandths) tall. The "code" is simply the height of the rail which in all cases is the same.

Can it be used... ... of course.

QUESTION: #15 I am looking for info on the basics of tracks and switches. For example what the heck is a frog point?. Also when looking at curves that are 18" radius (HO) where is radius measured too: Center of track, inside or what?

A frog point is the place in a track switch (called a turnout) that the two closure rails cross each other. It's number such as 4 6 8 etc. is the angle at which they cross each other. A number 6, as an example, is six to one. That means it measures one inch apart at a point six inches away from the frog.

Radius of curves is determined as the center line of the track.

QUESTION: #16 What's the difference between regular and DCC-friendly turnouts?

DCC turnouts have a gap that separates the frog rails from the point rails. That way the point rails are the same polarity as the stock rail that they sit beside. This insures that the back side of any wheels, drivers, etc. cannot touch the point and cause a short. This also means that the modeler must provide a switch or contacts that will switch the polarity of the frog to match the alignment of the points. This can be done with switch machine contacts or with contacts in the control switch itself on the panel.

With a regular turnout, the frog and points are connected and the points switch the polarity to the frog depending on which stock rail they touch. This eliminates a lot of extra wiring to the turnout if you use the points to control polarity.

QUESTION: #17 Do your Rotor Motor switch machines support a frog polarity control switch? Looking at the pictures on your web page (which by the way are excellent!), it doesn't look like there's a place to hook up the switch.

We use a double pole double throw (DPDT) switch to control the machine and the frog. One single pole double throw contact in the DPDT controls the machine and the other controls the frog. Over the years, our experience has been that contacts in the switch are much more dependable. The switch on the Rotor Motor simply turns it off. It is not suitable for frog polarity.

...this is a serious question...are metal wheel sets a lot better than the standard Athearn ?...are they drop-in replacements for the Athearn wheels ?...given the choice of plastic or metal, is the metal better ? J.D.

If I had a dime for every time this question is asked, I'd be able to afford not only the metal wheels I need, but Plano walkways too. Get rid of plastic wheels. They get dirty because they attract the dirt. It's not just that dirt sticks to them, they really pull it out of the air. One of our club members was lubing trucks with powdered graphite, and the graphite was flying to the wheel instead of puffing into the journal from the needle.
Plastic wheels on metal axles can come out of the box out of gage, and even if in gage when new, can shift. Kadee wheels, which are metal on Delrin axles seem to be made so they can't shift- in 30 odd years of using them I've had one set come wrong, two sets drift. If you want wheels suitable for cars in the post Korean war era, ribbed back wheel are illegal, so that bit of detail is not proper. That cut-off date may be even earlier- I don't have my notes on that handy. Kadee makes 33" wheels in both styles. So far, limited tests suggest that the Intermountain wheels are the way to go. I've got a couple of hundred sets coming to put in some covered hoppers for a unit train, will put still more in other stuff for blind testing.
I know- the cars /come/ with the d#@*& plastic stuff, but spend a buck or so per car and improve their tracking, and watch as you have to clean the track less when you get rid of all of them.

I recently installed four of your HOn3 code 55 turnouts in my layout. When I made a test run with my engine I found that it always stopped going through the turnout. After checking to be sure that there was no electrical problem I check the gauge and found it to be a little to narrow for the NMRA track gauge to pass through without binding. I took the turnout to my local train shop (Bruces Train Shop) and everyone agreed that the turnout appeared to be too narrow in gauge. Have you had any other problems with this turnout?? Please advise. I purchased 10 of these and all seem to have the same problem. I would appreciate hearing from you right away.

Hi. Did you check your wheels to see if they are a bit wide? Also don't use the NMRA track end of the gauge. Turn it over and use the end that says flangeways. The track end of the NMRA gauge is WAY too wide. It's a scale inch to an inch and a half (prototype) too wide. The flangeway end of the gauge matches the wheel set gauge on that piece of metal. If your wheel sets are gauged to NMRA specks, they should run just fine through the turnouts. I'm sorry you're encountering this problem. We deliberately make the turnouts to fit the flangeway/wheel dimensions. Unfortunately not all manufacturers are careful to gauge their wheel sets properly.

The reason we make the turnouts to the narrower dimension, is that they do not derail when the tolerances are right. They do when it's too sloppy .... such as the track dimension on the NMRA gauge.
Also please be assured that we will happily regauge the turnouts ... if you want them wider. But I would respectfully urge you to check the wheel sets. Be sure they fit all the way in the gauge ... This will result in low maintenance derailment free operation.

Handlayed Track How do you build the splines. Do you take a standard 1x4 and rip it to make the strips? How thick do you make them? I tried it with a scrap I had laying around. I tried a thickness of about 3/16 but it seemed to be too thick to get a small radius curve. How many pieces did you laminate to create the roadbed? Is it necessary to make it at least as wide as the ties? I guess I'm looking for info on the actual construction of the roadbed from someone who has done it and done it as well as you have.

I cut the splines from 1x (1x4, 1x6, 1x12) whatever was cheap at the lumber yard. Set your saw a little more than an eighth inch. 3/16ths is OK for 28 inch radius and up. I use five splines each about 1/8th. The roadbed comes out around 3/4 or so wide. You don't have to make the roadbed any wider than the rails. In fact I recommend a bit smaller than the ties since your scenery will look more natural with the ends of the ties slightly exposed. What gauge/scale are you building? That makes a difference.
You said to build the curves first. Do you have a jig or do you just bend the first and glue the second to it? It seems like one of those three handed jobs. By any chance do you know of a good article on the subject so you don't have to answer all these basic questions; maybe just an occasional more technical one?

I don't use a jig for the curves. I put a few nails in the framework on the curve line to bend the lams around. Then lay out all five pieces of lam flat on the bench. Spread glue on one side of four of the lams. place three glued lams on the fourth one. Then place the fifth unglued lam on the top.
Remember to stagger (stair step) the lams. Put a clamp on the center of the lams. Pick them all up and bend them in place all together around the nail form you put on the benchwork. Put an outside nail at each end of the lams to hold them in place. Now place your clamps all around the curve to squeeze the lams together all the way. Once the clamps are in place, you can wait for the glue to dry or go do the next curve, if you have more clamps.
I put the clamps about 6 inches apart. So make sure you have plenty.
I usually can't wait for the glue so I wrap masking tape four times around the lams next to the clamp. Get it tight. Then remove the clamps. The tape will hold the lams if you do it this way.

One more point. Most 180 degree turns will take about 12 or so feet boards. So if you use 8 foot lengths for your lams, they will come up short on the curves. I strongly suggest you use twelve or even 14 foot lams if you can. These make forming the curves much, much easier.

It seems that at slower speeds (such as a crawl through a switch yard) my DCC locomotive(s) will occasionally lose power when one of the trucks is directly over the frog.
The switches are Atlas and the locos are KATO C44-9W's so it shouldn't be a question of the trucks being too short, should it? They didn't have a problem before DCC.

One point to check. Most trucks do not allow all the wheels to touch the track at the same time. This makes electrical pickup spotty at best. When you assume the longer trucks will get through a frog, your assuming that all the wheels are touching for pickup. They usually are not.
Athearns are notorious for this problem, but any stiff frame truck has the same potential problem. When the truck three corners, then only one wheel is touching on one of the sides. If that one is over the insulated part of the plastic frog ...... then nada ...... zip ...... shes'a'no-go.
We make solid rail frogs with full pickup because of this and other problems. We also make point-stock-common DCC friendly turnouts to improve pickup.

Would you put cork Nscale road bed on the blue foam to deaden sound?
If you can't get door skin in your part of the country, what would be the closest.

If you intend to ballast your roadbed and use wood glue then the sound gets transmitted around the cork anyway. Everybody who says the foam is or isn't quiet, is only judging by unballasted roadway. I've tried all the corks and stuff 50 years worth I've seen cork, homosote, upson board, celutex, tar (AMI) etc . etc ... They all make noise unless you use matte medium ..(latex glue) to fasten the ballast. Latex contact cement to attach the track to "whatever" and then matte medium for the ballast, is the quietest that I've ever heard or seen. Most people build with plywood, and that is the noisiest. Sounds like a big piano sound board... cork or no cork. Plywood has to be cut no bigger than the track itself or excess noise is the result.
Glued ballast attached to hardshell scenery, is also a large sound board. The newer plaster gauzes do a better job of "killing" the sound.

Whats the advantage of using non-insulated switch frogs? They seem much more difficult to wire up.

All switch frogs must be isolated (insulated) to be wired for electrical operation. With metal frogs (all metal) the gaps are provided by the modeler and can be put where the modeler deems appropriate. Also the frog can and will conduct current to the wheels thus eliminating "stall spots".
With plastic frogs, they look atrocious (ugly) and the power routing and gaps are pre-determined by the manufacturer. They can easily have "stall spots" since the plastic does NOT conduct electricity to the wheels. Plastic frog turnouts are more toy like than metal frogs. Plastic frogs wear out. If you run trains, you'll probably have to replace the turnouts in a few years.

If you are interested in operation, then you should consider non-insulated switches and learn to wire the frog power routing the way you want it.

In regard to question #23, how do you wire switches?

There are many fine books in your local hobby shop about wiring. Lynn Wescott comes to mind. These books are easy to understand and you really need to get one as a reference. That said, I'll try to explain without pictures.
Turnouts on your main line only need one gap. That gap is the mainline rail that comes into the frog.

\ frog Mainline
\ \
\ \--------------
\ siding

A blank siding does not need a gap. So most siding turnouts only need the one gap. Now if the turnout comes back to the mainline, you still only need the one gap in the mainline and put one more gap in the siding rail that goes to the frog to keep the track from shorting when the turnout is thrown.

\ gap /
\ \ gap / /
\ \---------||---------------/ /
\ /
This is a passing siding and will get two gaps. Pick one end or the other and put one gap on each rail that comes from the frog.(two gaps total)

This method of gapping means the track only runs the train IF the turnout is thrown to that track. So a train can sit on the siding not moving while another train goes by on the main. Then when the turnout is thrown to the siding, that train will run. This is called "POWER ROUTING"

Whilst on the subject of turnouts, Shinohara or others, I would like to know the most common method is for making the tie rod which connects the tips of the two switch blades such they are mechanically connected yet provides electrically separated. I do know some people solder the tips to copperclad and then cut the cladding between the tips to provide the electrical separation but the problem with this is that the solder joints are usually so solid that they do not allow the blades to pivot properly. Not only this, the strip of copperclad tends to be enormously over-scale width thereby detracting from its appearance. Does anyone have any other ideas?

The throw bar width should be the same as your ties. That way it represents one of the ties and once painted disappears into the roadbed along with the other ties. Have you done this? I think you'll find that they DO disappear.
Second thing is to reinforce the copper clad (pc board) with a strip of brass from point to point on top of the pc tie. This strengthens the solder joints. I have some that are 45 years old .... still doing fine.

Third, if you leave the closure point rails at least three inches long, not spiked down, they will pivot very nicely. They bend minimally along the whole length and thus put very little pressure on the point-tie rod-joints. This installation, if done with care, will not only out last any other type of connector, but will give you smooth flowing, uninterrupted no-joint point rails, eliminating power flow problems (no joints) and eliminating one more source of derailments.

Check out the point throw-bar on any Railway Engineering turnout (yes I'm prejudiced) (I own the Company) and you'll see all of the above

One last thing. I have a plastic injection machine and could make any kind of throw-bar I wish. I have a full machine shop ..... yet I choose to make my products this way ...... because they LAST.
I respectfully hope you will re-think your opposition to PC throw-bars. I think you'll see that you prefer them.

Some notes about actual narrow gauge curves

It's notes the curvature that 2-8-0's could get around was 147 feet (40 degrees). On the main line of the D&RGW were wider curves. How much wider?
The answer is mostly 24 degrees or less - but the Marshall Pass main line had the sharpest curves, with 25 degrees 30 minutes maximum curvature between Marshall Pass and Shavano. This translates to 224 feet radius.

In S scale, this would be a 42 inch radius curve; in HO scale, approximately 31 inch radius. So the HOn3 models of K class locos made by PSC are operating on prototypically sharp curves when they run on 30 inch radius curves. Less than that is pushing the envelope.
More common maximum curvature was 24 degrees (238.75 feet), or 20 degrees (286.5 feet).
24 degree curves scale out at 44.75 inch radius in S scale, or 32.93 inches in H-O scale.
20 degree curves scale out to 53.7 inch radius in S scale, or 39.51 inch radius in H-O scale.
For O scalers, 25 degrees 30 minutes = 56 inch radius, 24 degrees = 59 inch radius, and 20 degrees = 71.5 inch radius.

I looked at the profiles for the D&RGW and the RGS. Maximum degree of curvature is shown for the various portions of the system.

D&RGW - Salida to Gunnison
25 degrees 30 minutes - Between Marshall Pass and Shavano.
24 degrees - between Shirley and Marshall Pass, and between Shavano and Buxton.

D&RGW - Gunnison to Montrose
20 degrees - Between Gunnison and Cedar Creek.

D&RGW - Gunnison to Crested Butte
16 degrees

D&RGW - Poncha Jct. to Monarch
24 degrees

D&RGW - Mears Jct. to Alamosa
20 degrees - Between Mears Jct. and Poncha Pass.

D&RGW - Alamosa to Chama
20 degrees - Between Big Horn and Lobato.

D&RGW - Chama to Durango
24 degrees - Between Monero and Amargo.

D&RGW - Durango to Silverton
24 degrees - Between Hermosa and Cascade

RGS - Ridgway to Rico
24 degrees - Between Lizard Head and Ames

RGS - Rico to Durango
20 degrees - Between Millwood and Glencoe

RGS - Vance Jct. to Pandora
24 degrees - Between Vance Jct. and Keystone.

Notice that the curvature would not have precluded use of the K-28, K-36, and K-37 class locomotives on the RGS - let alone the Silverton branch, if bridges, rail weight, and clearance problems were resolved. What I come away with is the realization that most model railroads are trying to run big locos on curves that are way too tight to be prototypical. Mine included - the 24 inch radius curves I have been able to squeeze my K-37 and K-36 class around are considerably sharper than the 25 degree 30 minute curvature on Marshall Pass. Model railroading is all a matter of compromises, as we fussed, cussed, and discussed on the HOn3 list. Most of us are living with running trains around much sharper curves than the prototype. Anyone for modeling the Uintah? (it had REAL sharp curves 12" in HOn3)