Question for track hand layers

[narrowminded]

Using the NMRA Gage for point clearance (.043) at the point heal, an 18' point rail (PRR #8 turnout) and a .018" rail head width (c43 rail), anything under 2.5 degrees would work.


Jason

... and I have an honest angle measurement from my #4 fixture that's 2 degrees.  That and some UP specs for prepping switch rails, many different switch #'s using the same rail length, have me thinking they are all the same with sufficient clearance built in and that's all I'm hoping to find out.  I'll know for sure when I get deep into the drawings and think I'll then know why this whole thing works.

Thanks

[jagged ben]

BETTER YET!  I'm being thick skulled over here.   You have two tools.  I'm really trying to measure the angles to see if they are the same.  If you could just unbolt the two fixtures and place about a .8mm to 1mm wire in the groove and set the second tool against it, see if the sides remain parallel.  That would say they're the same, a #5 and a #8... or they're not the same.   I'm guessing they are the same.

Since you're asking about Fast Tracks jigs, I'm just gonna tell you without measuring anything, because I've filed enough different point sizes in my days to know...

They are not the same angle.  It's what Rick told you in post #2.  The difference between a #5 and a #8 may be hard to discern with the naked eye, but the difference between a #4 and a #10 sticks in your mind after you've done a few and learned about the different challenges of each.

Now since you're apparently trying to machine your own points, I might suggest that you stop worrying about what Fast Tracks does and take a look at prototype stuff, as I noticed also being suggested upthread as I somewhat skimmed my way through. Fast Tracks points are not very true to prototype.  I don't know how active Robert (the 'Track Nazi') is these days but if he finds this thread he can probably be counted on to be quite informative about that and possibly other details that will be pertinent.   I myself will need to plead complete ignorance about how the prototype handles points for different size turnouts.

[Chris333]

Robderebel?   

[narrowminded]

Since you're asking about Fast Tracks jigs, I'm just gonna tell you without measuring anything, because I've filed enough different point sizes in my days to know...

They are not the same angle.  It's what Rick told you in post #2.  The difference between a #5 and a #8 may be hard to discern with the naked eye, but the difference between a #4 and a #10 sticks in your mind after you've done a few and learned about the different challenges of each.

Now since you're apparently trying to machine your own points, I might suggest that you stop worrying about what Fast Tracks does and take a look at prototype stuff, as I noticed also being suggested upthread as I somewhat skimmed my way through. Fast Tracks points are not very true to prototype.  I don't know how active Robert (the 'Track Nazi') is these days but if he finds this thread he can probably be counted on to be quite informative about that and possibly other details that will be pertinent.   I myself will need to plead complete ignorance about how the prototype handles points for different size turnouts.

Thanks, Ben.  I actually am talking to @robert3985 and he's the one who forwarded a TON of UP spec stuff.   It was there that I stumbled upon the switch rail lengths and prep that was a constant over a big range of turnout numbers.  What I was hoping to do was to get the fixture(s) for these in the current group of fixtures that I'll be running in a week or two.  As it is, there is no clear answer and I doubt I'll have the time to do all of those drawings to determine the answer for sure.  In the meantime, I'll go ahead with a two degree one but very well may have to go back at it later after I've completed my homework.   It's not the end of the world but there was some wishful thinking that I could get these fixtures done too while I have the machine access.  It's OK, I've got plenty that I will get through.  This is a pretty involved project but I think the results will be well worth it.

BTW, I'm figuring on Code 40 and Code 55, branch and main, ties and details, especially at the prompting of Robert.

Thanks all for your input!

[nkalanaga]

Here is a file that might be of interest.  It's from the Sept.Oct 1994 issue of N-Scale Magazine, and if someone can tell me how to upload a PDF file, it can be printed in actual N scale.  It shows the switch (point) length for various turnouts.


And, in case the actual file can't be uploaded, just the numeric data table:

[peteski]

Nick,
If you use standard reply (not quick reply), you can attach PDF files to your post.
There is "Attachments and other options" link on the left under the text window.

[GaryHinshaw]

I really like what you're doing here @narrowminded.  Here are a few overly long comments, probably worth what you paid for them.

• The proto:87 store already offers milled points in a variety of rail codes. They are marketed for use with any # turnout frog.  As long as the points accommodate the longest turnout you intend to build, they will work for any shorter one as well.  It might be desirable to have a range of point angles, but I don't really think it's necessary.
• Their points mate to the closure rail with a heel block that also sets the spacing between the point/closure rail joint and the stock rail.  The points themselves are straight, so the width of the heel block and the length of the point rail determine the angle at which the points meet the stock rail.
• They have a number of templates posted on their site that specify the length of the points for each turnout number.  These can be used to determine the shallowest angle you'd like to accommodate for a given set of points.  For example, looking at their #10 template, the points are spec'ed to be 1.636" long, and the stock-closure rail spacing is set by the heel block is 0.062", so the angle the points make with the stock rail in this design is about 2.2 degrees.
• One of the feature I really like about their points is that they drill a small hole in the inner base of the rail to accept a short bit of wire protruding from the throw-bar.  This gives the point/throw-bar joint some excellent mechanical strength once the joint is soldered.  The photo below shows an example of a more or less finished code 55 installation on my layout.  If you look closely, you can see the wire sticking through the rail base.  The design actually uses 2 throw-bars: the soldered one noted above that gives the point assembly mechanical strength, and a more standard one with etched clips that push the point web tightly against the stock rail.
• Honestly, I think their code 55 points meet all my needs already, and since I'm just about done with my code 55 builds, I'm not really in the market for any more.  However, their code 40 points could be much better, and I would be very excited to try out anything you come up with in this regard.  (The p:87 code 40 points don't have the base hole, and the milling on them seems incomplete, e.g. the outer rail base is not fully planed away.  It would be worth getting a hold of some to study, so you know what not to do.  I'd be happy to send you some samples if you're interested.)
• I would be especially impressed if you could come up with a way to drill a hole in the code 40 base, or if you had a better idea for a robust throw-bar assembly that works with code 40.  I too am leery of just a soldered butt joint, so I have been putting off my code 40 builds until I have a better solution.

Anyway, great stuff!  Please keep us posted!

[Maletrain]

Doooh!  Shouldn't have "turned in my homework early" last night.  In the middle of the night, it came to me that I had conflated "flangeway width" with the gauge line offsets at the heal block.  There is a rail head thickness in there to contend with. 

So, the angles I calculated are correct, but the paragraph on flange ways looking prototypical for scale point rail lengths is off-base.  I had started writing about how the NMRA specs would make it necessary to change the geometry of the switch when scaling it down, but, when I saw the NMRA spec scale to 6.5" and saw the prototype drawing spec of 6.5" at the same location, I incorrectly conflated those two things and deleted that paragraph.  So, I am redoing it below:

The prototype B&O throw distance is only 4.5", which scales to 0.028" in N scale - about the width of the rail head for code 40 rail.  But, the NMRA spec for flangeways comes out to about 0.042".  That is going to require that the space between open point rails and their adjacent stock rails look nearly twice as wide as prototype when our brains use the rail head width as a mental gauge for how wide that space should look.  (Oddly, that is one small aspect that could make code 55 rail look better than code 40 in that area.)  Still, that is better than the roughly 0.060" I just measured in a commercial Atlas code 55 turnout. 

Further, that is going to require that the spacing of the rails at the heel block be wider than scale.  If the model point rail lengths are made to scale length, then they will be changing the geometry of the rail paths through the turnout.  And, they will be changing the point rail angles needed to match the stock rail, so that they are not the same as the prototype angles.

One solution would be to move the hinge location by making the point rails longer, which keeps the angles the same, but requires the location in the turnout where the straight rail meets the curved rail to be farther from the point ends, requiring the curved rails to curve more sharply.  Another solution is to keep the prototype straight point rail length, and make the curved rails curve slightly less.  Or some compromise between the two, to make things look as good as possible with the compromises required.

Unless Mark wants to go through a development exercise because he doesn't think the Fast Tracks geometry looks right, it is probably best to use Fast Tracks geometry, whatever it is.  And, that would allow for the future use of Marks 3-D switch tie prints with Fast Track tools as an option to having Mark make the whole switch or machine all of the rail parts for switch kits. 

[wcfn100]

  Another solution is to keep the prototype straight point rail length, and make the curved rails curve slightly less.

That's the way I approach it when drawing turnouts.  For something like a #8, the difference between a point rail at 1° and 2.5° creates a diverting radius difference from ~21" to ~28" respectively.  Since most modelers use turnouts that are too sharp, the larger radius is a good way to compromise IMO.


Jason

[narrowminded]

I really like what you're doing here @narrowminded.  Here are a few overly long comments, probably worth what you paid for them.

• The proto:87 store already offers milled points in a variety of rail codes. They are marketed for use with any # turnout frog.  As long as the points accommodate the longest turnout you intend to build, they will work for any shorter one as well.  It might be desirable to have a range of point angles, but I don't really think it's necessary.

Thanks, Gary.  And THIS is exactly what I found from prototype drawings, admittedly to my surprise, being new to this, and when contemplating the relationship between these parts in service started to realize that there really shouldn't be a need for varying angles of the actual point section of the switch rail, at least over many turnout sizes.  This is because the relationship of these specific components when isolated from the rest of the angles of the dangles, don't really change.  The rest of the turnout is where all of that wonderfulness occurs, at least in a prototype which is the goal of this effort, to the extent possible in scale.  I don't have the balance of the drawings done but at this early point I don't see a reason why this relationship can't be accomplished.  My rude awakening may still come, but later.

BTW, I often reference being new to this and it's true, new to making turnouts and many of these modeling techniques.  What I'm not new to is making parts and machines accurately and very functional.  That understanding also leads me to admiring what Proto:87 stores has done in making those milled points.  It's not easy, what they're doing.  And the cost is there too, totally understandable with some inkling of what that takes, especially in code 40.  With that said, I am not planning on offering that detail as a standard component from both a cost standpoint but especially the ruggedness.  That method has the rail foot attachment, regardless of how you accomplish it, pretty delicate.  If it's really wanted I expect it could be done, as an option at an additional cost, or just purchasing the parts from Proto:87.

• Their points mate to the closure rail with a heel block that also sets the spacing between the point/closure rail joint and the stock rail.  The points themselves are straight, so the width of the heel block and the length of the point rail determine the angle at which the points meet the stock rail.

EXACTLY! This is what I found on prototype drawings and why I started this thread, thinking I might be able to sneak the point rail prep fixture in to this batch of fixtures even though I haven't seen this gory detail in a scale drawing yet, the ultimate confirmation.   The 2 deg measurement of that angle on my only point rail tool (a Fast Tracks #4) was accomplished using the three point measurement method, utilizing a milling machine table for the accurate travel distance and a dial indicator for the third measurement.  It's accurate and concurs with what the calculations show plus some windage for tolerance.

• They have a number of templates posted on their site that specify the length of the points for each turnout number.  These can be used to determine the shallowest angle you'd like to accommodate for a given set of points.  For example, looking at their #10 template, the points are spec'ed to be 1.636" long, and the stock-closure rail spacing is set by the heel block is 0.062", so the angle the points make with the stock rail in this design is about 2.2 degrees.

One thing I want to make perfectly clear is that this will NOT be a copy of anybody's existing product.  Yes, for reference in conversation and as part of my own research I will occasionally look at what others have done, it'd be silly not to, but my end result will be based on prototypes so any similarities that occur will likely be based on the fidelity of an existing product's details to prototype design, not because somebody copied from somebody.  At it's core this is already a bit different in the approach and level of detail as a standard, hopefully all to the good.   Many of the details, throwbars for example, will be pretty original in the model world (not in the proto world) if development allows and it all works as envisioned.

• One of the feature I really like about their points is that they drill a small hole in the inner base of the rail to accept a short bit of wire protruding from the throw-bar.  This gives the point/throw-bar joint some excellent mechanical strength once the joint is soldered.  The photo below shows an example of a more or less finished code 55 installation on my layout.  If you look closely, you can see the wire sticking through the rail base.  The design actually uses 2 throw-bars: the soldered one noted above that gives the point assembly mechanical strength, and a more standard one with etched clips that push the point web tightly against the stock rail.

I hear you on the room for drilling a hole in the Code 40 rail foot, just by available rail foot space.  It's more than 25% narrower in that area than Code 55.  And the prototype doesn't drill the rail foot but bolts a bracket to the rail web that extends down and out toward track center where it's drilled for the attachment.  That's closer to what I'm envisioning for my pieces although our oversized wheel flanges won't allow for anything coming down over the rail foot, especially in code 40.  Code 55, maybe, but not 40. 

But that brings us to the real problem with Code 40, wheel flange clearance.  The wheel flanges we use are bigger in width and especially in diameter.  If a wire were to be placed or have to pass directly below the inside edge of the rail head, with code 40, it doesn't leave sufficient wheel flange diameter clearance even in the best of circumstances.  At this point, with fixturing and testing currently in development, I still intend to use drilled holes but they will be incorporated into the piece that attaches to the rail, guiding it under the stock rail for a pretty robust design while still maintaining some adherence to a prototype appearance.  There may be a couple of choices on the tie bar appearance but that's not fully determined yet. 

• Honestly, I think their code 55 points meet all my needs already, and since I'm just about done with my code 55 builds, I'm not really in the market for any more.  However, their code 40 points could be much better, and I would be very excited to try out anything you come up with in this regard.  (The p:87 code 40 points don't have the base hole, and the milling on them seems incomplete, e.g. the outer rail base is not fully planed away.  It would be worth getting a hold of some to study, so you know what not to do.  I'd be happy to send you some samples if you're interested.)
• I would be especially impressed if you could come up with a way to drill a hole in the code 40 base, or if you had a better idea for a robust throw-bar assembly that works with code 40.  I too am leery of just a soldered butt joint, so I have been putting off my code 40 builds until I have a better solution.

Anyway, great stuff!  Please keep us posted!

Thanks, Gary.   I would very much like to keep in touch with you and hope that I will have something worthy of your layout in the not too distant future.    In the meantime, I assume your preference would be for a branchline style switch as opposed to mainline.  This will effect tie sizes, some spacing, and other finer details.  I also wouldn't mind being able to use you as a source for information and opinion especially where compromises may be required.  @robert3985 has already been VERY helpful in that regard.   Thanks, Robert.

[CNscale]

BETTER YET!  I'm being thick skulled over here.   You have two tools.  I'm really trying to measure the angles to see if they are the same.  If you could just unbolt the two fixtures and place about a .8mm to 1mm wire in the groove and set the second tool against it, see if the sides remain parallel.  That would say they're the same, a #5 and a #8... or they're not the same.   I'm guessing they are the same.

Lots of conversation here that I haven't had time to read yet, so I'm not sure if this is still helpful, or even what you're looking for, but here's what I've got: I set down a length of code 40 rail and placed the fixtures on top with the rail in the groove. For the points groove, the fixtures appear to be parallel, or at least very close. Just out of curiosity I tried the same thing for the frog grooves and they're clearly not parallel. (note that the glare on the rail makes it appear to not be straight, hence the graph paper)

[narrowminded]

Lots of conversation here that I haven't had time to read yet, so I'm not sure if this is still helpful, or even what you're looking for, but here's what I've got: I set down a length of code 40 rail and placed the fixtures on top with the rail in the groove. For the points groove, the fixtures appear to be parallel, or at least very close. Just out of curiosity I tried the same thing for the frog grooves and they're clearly not parallel. (note that the glare on the rail makes it appear to not be straight, hence the graph paper)

Thank you so much!    That's what I had begun to suspect, that the point rail angles were the same at least in many fixtures if not all.  A better way to check them might be as I described, set a piece of wire that sits just above the face in the groove and then lay the other one face down on the first one.  If the outside edges of the fixture are parallel you're in! 

The frog angles would clearly not be the same but they were also easy to find sources for, prototype being one.   I think I'm on to this.   It's been a torturous path that was instigated by the UP prototype drawings from @robert3985 .  I'm pretty comfortable now that I understand it and can run the 2 deg fixture without wasting my time.   Thanks again.  It would be interesting to see others I guess but I really don't think we're going to find anything different.

One reason a longer switch rail taper might be needed is if the point rail radius is carried beyond where it's supposed to be.  A careful design with all of the rails being guided by the tie plates to their exact proper location along the length won't allow that.  I'm starting to feel all warm and fuzzy over here.

[GaryHinshaw]

In the meantime, I assume your preference would be for a branchline style switch as opposed to mainline.  This will effect tie sizes, some spacing, and other finer details. 

Yes, branchline specs for me, definitely.  I'm very excited to see how your ideas for code 40 throw-bars pan out.

[nkalanaga]

Peteski:  Thank you!  I never use the regular reply form, as a habit from years ago, when I was on dial-up.  At the time, on the Atlas forum, the quick-reply really was quicker.  I learned the codes for URL and IMG, then just typed them myself.  But that means that, when I do want something special, I don't know it's available!

So, hopefully, here is the PDF.  As has been noted by others, the designs don't scale perfectly, because of our over-wide flangeways.  But it is interesting that multiple turnouts use the same point length, which could make model design easier, even if we have to change that length.

[ Guests cannot view attachments ]

[ Guests cannot view attachments ]

[Maletrain]

I don't understand why several posters keep saying that the point rails are the same length on the switches.  The posted image from the "Engineering and Maintenance of way Cyclopedia" clearly says that point rails ("switch rails") are 11'0" for #6, 16'6" for #s 7, 8 and 10, 22'0" for #11, and 30'0" for #s 16 and 20.  My similar drawings from the B&O show 11'0" for #s 4 and 5, but 13'0" for #6, 15' for #7, 16'6" for #s 8 and 10, 24' for #16, and 30'0" for #20.

So, there is a clear trend for the point rails to be longer for the higher number switches, which means that the point rail angles are clearly smaller for the higher number switches.  But, it is also clear that there are ranges where the same point rail length is used for several adjacent switch numbers, and that those ranges vary from one railroad to another - for instance the Cyclopedia showing 11' for #6 while the B&O shows 13'0".  And, the Cyclopedia shows 16'6" for #7s while the B&O shows 15'0".  So, there is obviously a range of point rail lengths and thus point angles, for the same switch number on various railroads.  And, one railroad may have longer point rails than another railroad for one switch number and shorter for another switch number.

So, I think that Mark is on the right track to use a single point rail length for a family of N scale switches that I would characterize as "medium", say 15' for #s 6, 7 and 8.  But, I think those will start looking too sort around #8, and that #10s and larger will need longer point rails to look right.  And, if he is going to make #4s or #5s, a point rail about 11" would look more prototypical and make the curve between the hinge and frog less sharp.

Point rail angles on the models are a somewhat different matter.  Our models probably don't really need the tight match to the stock rails that the prototype needs for the purpose of supporting weight - although with code 40 rail, I am not sure of that.  And maybe there is also a side-force issue with code 40 rail that needs to be considered to make sure that the point rails don't get pushed out of gauge when something like an x-8-x or x-10-x steamer goes through.  But, if not needed for strength in the models, then it might be reasonable to use smaller point angles (that can fit larger switch numbers) on shorter switches, if that is a benefit to the manufacturing process.

Has anybody here made code 40 N scale turnouts, before?  I see that Fast Tracks has jigs for them.