Or do what the real railroads often did....

And single track the line through the tunnel.  

Double track tunnels on freight railroads in North America are fairly rare.  There are some.  The old B&O tunnel at Harper's Ferry to name one.  But the B&O had serious clearance issues moving into the diesel era, and their tunnels were a big part of the problem.

Track centers have widened from 12'0" in the 1800s to 13'0" on most mainlines of the 1900s, to 14'0" on many of today's railroads if real estate allows it.  If we can, new construction is on 16' or 20' centers.

So the question becomes what is the story behind your tunnel?  Is it an old bore that you want to bring up to modern standards?  Or a new bore?

Another issue to keep in mind is curves.  I can't think of a single HO tunnel portal casting I've seen that was intended to be put on a curve.  Modern cars are a lot longer than steam era cars, with more overhang on curves.  And of course our model curves are much tighter than real world curves.

And how do you keep two trains from side-swiping in a tight tunnel?  Only let one in at a time.

Scott Chatfield

Raise them?

Its not illegal to be in violation of the NMRA! If the tunnels are not tall enough that higher cars hit the arch of the tunnel it might be possible to raise the portal by installing them on a concrete base or notching this area. Here is my example of my double track mainline narrowing down to one prior to entering the tunnel. This is prototypical as noted above, adds some operating interest, plus I did not have enough rough in the utility closet to allow both tracks to make the turnback curve! 

Clearances

First, RP7.1 is not a "standard", its a "recommended practice", that's why the number starts with "RP".  A standard is a "must do", a recommended practice is a should do (sorta kinda).

As far as the trains hitting in tunnels, not a problem as long as the equipment is close to scale width.  The risk of trains hitting each other is when the track is curved and when the track is curved there are other tables that address clearances on curves and based on radius, increase the recommended track spacing as well as the side clearance dimension.  

You are assuming that S = 2*A, which is not the case.  The track centers do not equal twice the side clearance.

NMRA RP7.1 Clearance

Hi Magnus,

I think the below will show that your perceived error in the NMRA Recommended Practice (RP) isn't there:

The tunnel portal opening width in the Modern Era per NRMA RP7.1 is listed in the right most (third) diagram on the top of page 6 as:

A + S_min + A

(not A + A + A + A : which I believe you are using based on doubling the width of the single track diagrams (first two of three diagrams at the top of pages 3-6)

For HO scale this gives via the table on page 6:

A = 31 mm

S_min = (14ft * 12 * 25.4) / 87.1 = 49.0 mm (just under 2". 14ft taken from the table on page 1)

This gives a total minimal width for a modern era double tunnel portal opening of:

31 + 49 + 31 = 111.0 mm

So yes, your 2 portals you bought at 108 mm are technically 3 mm too narrow.

Lets do this in inches to check potential slop in the NMRA RP:

A = 1.24"

Smin = (14ft * 12) / 87.1 = 1.93"

1.24" = 1.93" + 1.24" =  4.41" width

Bring it back to mm to compare:

4.41" * 25.4 = 112.0 mm

So there is about 1 mm worth of slop in the RP.

Conclusion:

Your 108mm wide tunnel portals are likely going to be okay. The missing 1.5 mm on either end is about double the width of many HO oversized grout lines. Maybe be extra careful ensuring your tracks are properly on their center lines and use of the portal for tangent track only?

Thank you

Thank you all for the great answers. Let me first address some questions:

 No, but I'm assuming S_min

 Yes, and S_min must be equal or greater to A+A to avoid collisions (theoretically), or am I missing something?

And that is exactly where I see the error in RP 7.1. That calculation creates the risk of collision if two trains meet in the tunnel which both need a clearance of, say, A=28mm.

I also think the portals might be OK, but as your calculation shows, it is a narrow, and negative, margin, so I will do a lot of testing with my widest trains. And your advice to be extra careful laying the tracks is well taken.

 Agreed, so the tunnel portals will be on straight tangent track.

Yes, and there will be curves inside the tunnel, but inside the tunnel I have control and will have plenty of space.

 Yes, of course, but I want my layout to follow "Recommended Practice" as far as possible.

Good question - I am modeling the Great Northern Tunnel in Seattle, see picture. Granted, it was built in 1904 with 2 tracks, so it is a bit narrow for use today, but it is still used heavily today for both freight and passenger cars, see

Fun fact: One of my portals is from Chooch, and I believe this Great Northern Tunnel was the prototype for them, they even have the same year (1904). Chooch was based in Seattle, so it would make sense. The marketing picture even has a Northern Pacific locomotive in it. Perhaps someone can confirm this?

Smin

And as several of us have said, that is your fundamental error. 

Smin= 14 ft for the two modern eras. NOT, repeat NOT 2*A.  

Prototype minimum track spacing is 13 ft.  14 ft is wider than that.  An AAR Plate F, the clearance envelope for a high cube car or an auto rack, is 10'8".  That is the clearance envelope for the CAR.  That means there is at least 3 ft of clearance between the cars on adjacent tracks.

Once more, Smin is NOT = 2*A.

Tunnel

You should notice that the tunnel has been modified to accommodate larger cars that exceed AAR Plate C.

The BNSF has ground out two notches at the 10 o'clock and 2 o'clock positions on the portal to allow it to accommodate taller cars.

If you notice on the NMRA clearance diagram there are two dimensions labeled A, one is pure side clearance and the other is to the upper corner of the car clearance envelope.  THAT is why A is greater than 1/2 Smin.    It is larger to accommodate the upper CORNERS of the car near the roof line.

The major restriction on the tunnel now would be speed, they wouldn't want the cars to sway or bounce going through the tunnel because the clearance in the upper corners is a matter of inches, not feet.

NMRA RP7.1 Continued

As I mentioned in my first response and others re-iterated, the value of S_min does not equal 2*A for tangent track and as such are NOT interchangeable in your clearance calculation.

The NMRA RP 7.1 only covers the recommended practice for tangent (straight) track. For clearances on a curve you have to look at the table on page 2 of NMRA RP 7.2. For adjusting tunnel (portal) clearances on a curve you use NMRA RP 7.5 instead.

Looking at RP 7.5, the adjusted values for A (shown as A_w) and S_min (shown as S_i) are calculated based on the given formulas which uses your track curvature as input.

Structure Gauges and Kinematic Envelopes

Wow, this led me down a deep rabbit hole. 

 Ok, I got that. But now I wonder why. After some time with Google I learned that:

1. Structure gauges (e.g. tunnel) are separate from Loading gauges (vehicles)
2. Kinematic envelope is the max a Loading gauge can wobble due to speed, aerodynamics, wear, failing equipment, superelevation, etc.
3. Structure gauges are derived from and encapsulates, but is not the same as, Kinematic envelopes.
4. In addition to the Kinematic envelope, Structure gauge add additional margin for construction tolerances, walkways, fire and safety equipment, signs, new wiring, etc.

And you don't need that extra margin from 4) between tracks as you need on the outside of tracks. So it makes sense that:

S_min <  2*A

S_min should really be derived from the Kinematic envelope K_e, worst case 2*K_e - that would be the worst possible case if both trains malfunction and wobble max against each other right when they meet, which is very low probability, but theoretically could happen.

A way to lower the Kinematic envelope is, as some of you pointed out, to lower the speed. But it can also be lowered by for example well maintained and regularly checked equipment.

So just as BNSF for sure have rules of speed etc. in this tunnel, I will incorporate that in my operation planning.

Modern Era S_min is 14' as pointed out in the NMRA RP. But I have seen several references that needs to increase for several reasons. For example, for high-speed train main tracks, the proposal is to increase it to 20' based on experiences with Shinkansen.

Yes, I noticed that, but it changed (disappeared) in the Modern Era diagram and was replaced by R_C.

Yep, that makes sense, but as I mentioned I'm not so worried about the curvature inside the tunnel as I have much more flexibility there.

 Oh, yeah! That is awesome! I did not see that earlier. I have to do the same. Thank you for pointing that out.

Thank you all for your input. I learned a lot today.

Answer

To conclude, here is the answer to the original question for future readers of this thread.

Q: Is the width of a 108mm/4.25" HO double portal too narrow according to the NMRA Recommended Practice 7.1?

A:  Depends on the era, see calculations below:

• No for Old-Time, Classic and Early Modern Era
• Yes for Modern Era, but only 3mm/0.16"

Just like the prototype, older tunnels out of the clearance envelope can be still used, but under certain rules; most common to limit the speed.

Thanks to all for the help, and especially Yaron for providing the correct equations.

Don't get me started...

Hi,

when talking about "kinematic envelopes" - there was an issue with the underground system in the "big city"... They had received new trains and were in the process of certifying them. They did the usual test with blocks of styrofoam tacked to the sides of the new, shiny train running along the tunnels. Which was fine.

Until the local certification authority engineer decided they wanted a full calculation of the kinematic envelopes for every bit of the tunnel. As far as I know, they have the train certified on one of the six lines. At least when I was still commuting regularly two years ago - they might have a second line done by now.

This whole disaster happened when the underground system desperately needed lots of new trains and there were ~20 new trains at the manufacturer's plant... With the new trains at least running on one line, that takes away enough pressure from the pool of trains to keep the system running...

Anyway, this might help you with the tunnel portals: tack some blocks of foam on the models and run them through the tunnels to see where they hit

Honestly, I think you're fine with pushing your longest cars through the narrow spots.

Have fun and stay healthy

ps: this would be an opportunity to run some gauntlet track through the tunnel

There Are Way Too Many Engineers Here

And not excluding myself (ME), but man alive, it's a hobby. Just make it work and look good. 

I should perhaps explain:

I went to tech and E-school, and first jobs after graduation, you work problems by the book solution. Because that's what you've been trained for. The hobby equivalent would be articles and videos. Mentors smile, because they've been there. After a couple years, you learn the job (hobby). This is likely true for any field. 

After a while you see what works and what doesn't. You get to a point where the presented problem is either familiar, or can be dealt with by experience and training. Same with a hobby. My opinion is influenced by desire for operations in a realistic environment, but that's not everyone's jam. Just sayin'; don't overthink the problem. 

Speaking of tight clearances...

Thailand's 

Model follows prototype

Turns out I will need to cut out corners like the prototype did.

Re model follows prototype

Hi Magnus,

Or you could jack up the portal by adding another concrete-layer to the portal foundations, provided that your portals aren't glued in. Many railroads effectively did this by digging out the tunnel floor, thus increasing the tunnel height.

Tunnel Clearance

All you need to do is shim your tunnel portal about an 1/8” and that should provide sufficient clearance.

Raising

Thanks for your comments. Yes, I will raise the portal a bit, but I will still do the cut in the portal like the prototype did. Nothing is glued yet.