Terry Smith’s Historical Maine On2 loco and track curving experience, 2013 version 3.
Terry Smith has provided the following historical data from his own tests and experience. It is a more complete version written up for these FAQ’s in October 2013 from additional notes, dating from 1995 to 2004, which were recently re-discovered and both supercedes and finishes his series of previous posts to the MaineOn2 Yahoo! Group on this subject.
Minimum radius: in my On2 “tyro” days (around 1988-90), I did a lot of tests on different locomotives and their curving ability as part of my layout design. My then available space had a restriction close to one corner which meant that being able to use a 36″ minimum radius would be particularly “useful”.
I had attended a number of the early Whitman meets, and in talking with other On2 modellers, I was aware that if I wanted to run the Big Forneys (B&SR/B&H #7 & #8, SR&RL #10) that the advice was to try to use a 42 or 48″ minimum radius.
I set up test radii using Precision Scale flextrack on Sundeala board (our UK favourite for proper track pin retention, believed to be denser than Homasote). I laid out different radii (60″, 48″, 42″, 36″, 33″, 30″) and repeatedly ran each loco forwards, backwards, curving to left and to right, with and without freight cars attached.
Based on these series of tests I selected 36″ as my minimum radius, and then started to build the layout proper with hand-laid track, code 83 rail, Clover House ties, Railcraft spikes plus turnouts using #6 frog angles and constant radius curvature through the point/switch……..and then started having running problems!.
The problems arose with both the Car Works B&SR #7 & #8 derailing fairly regularly on the first point/switch after the particular corner when running in a “clockwise” direction. I was disappointed to say the least. There were no kinks in the curve and it had sort of transitions into it.
After a lot more testing and observations, I found that the problem was that one of the rear drivers was consistently riding up onto the inside rail and ran like this on the plain track until de-railing at the first point/switch. I put this strange behaviour down to some artefact of vertical stiffness/motion problems between the drivers and the rear truck, and possibly not enough weight on the drivers. On one of the loco’s, probably the #8, I packed the springs with shims and reduced but never eliminated the derailing tendency.
The table below lists the summary of experience to that point
Loco type Example Track Experience
Small Forney CB SR&RL #6 Precision Scale flex 36″ rad OK
Big Forney Car Works B&SR #7 Precision Scale flex 36″ rad OK
Big Forney Car Works B&SR #8 Precision Scale flex 36″ rad marginal
Big Forney Car Works B&SR #8 Precision Scale flex 48″ rad OK
Prairie CB SR&RL #16 Precision Scale flex 36″ rad OK
Medium Forney Car Works B&SR #6 Precision Scale flex 48″ rad OK (*)
Medium Forney Car Works WW&FR #7 Precision Scale flex 48″ rad OK (*)
Big Forney Car Works B&SR #8 Precision Scale flex 48″ rad OK (*)
Small Forney CB SR&RL #6 Hand laid 36″ rad OK
Big Forney Car Works B&SR #7 Hand laid 36″ rad marginal
Big Forney Car Works B&SR #8 Hand laid 36″ rad not OK
Prairie CB SR&RL #16 Hand laid 36″ rad OK
Small Forney includes the Putnam & Stowe 0-4-4 locos, the Portland Product 0-4-4 locos and CB SR&RL #6.
Medium Forney includes the more recent 2004 production Car Works B&SR #6 and Car Works WW&FR #7. I have no experience with the CB Eustis Forneys.
(*) Tests done in 2013
Big Forney includes the Carworks B&SR #7, Carworks B&SR #8 and by close similarity the equivalent SR&RL #10.
Prairie includes the CB SR&RL family such as #16, #18 & #24, but does not include The Car Works #23.
Ok; suitable for layout operations.
Marginal; occasional failure, loco either stalled or derailed frequently enough to give raised concerns when operating layout.
Not OK; consistent failure, loco either stalled or derailed frequently enough to give sufficient concerns for it not to be used when operating layout.
What went wrong?
My original thoughts in the 1990’s were that perhaps the original tests were flawed, in that the flex-track was pinned down, but perhaps not in enough places, and it could have retained sufficient lateral “displacement capability” to allow the big Bridgton engines to get round a nominal curve of 36″. By comparison, the hand-laid track on the layout was rigid.
Sometime later (around 2006-2008) I got around to measuring the track gauge at various places with a two point dial calliper.
My handlaid straight track measured .495-.510″ gauge, the transition curve measured .505″ and the 36″ radius curve measured .500″ to .514″, but most measurements were grouped in the .500″ to .502″ range.
In comparison Precision scale flex track measures around .510-.515″, and does not get below .508″ when curved to 36″ radius.
For reference, the current NMRA standard (S-1.2 dated July 2009) for On2 plain track gauge is .500-.522”, and for “Guarded work” (presumably pointwork in my language) (S-3.2 dated February 2010) the gauge at frog is .500-.511″. I believe that Fast Tracks set up their switch building jigs to a nominal track gauge of .510-.512”.
My current conclusion based on the gauge measurements and repeated tests is that the problem with the hand-laid track on the former layout was that it was too tight in gauge for the curve radius, and particularly so for the Big Forneys.
And finally what went right with the pointwork?
The curious thing with all this testing and problems is that I do not recall ever (apart from above exceptions) having problems with any loco’s derailing or having problems with negotiating my pointwork, which initially were based on #6 frog angles and constant curvature through the point/switch.
Inspection has showed that my handlaid pointwork measured .493-.505″ gauge with the two point dial calliper. However, each point was individually built in a home made jig using the same code 83 rail as for the plain track except that I used Clover House PCB ties so that the rail could be soldered to the ties. Each point was tested and adjusted during building on the workbench and also when fitted in place on the layout.
I now suspect that this testing and adjustment regime explains why the locomotives could successfully negotiate the pointwork.
As an experiment, two points were made to larger radii. These measured close to #8 frog angle and had a constant radius of around 48”. They gave much smoother running and operation than the #6’s.
I had tried some modifications to the rear trucks of both my #7 & #8, these may have been successful for #7, but not #8 in terms of consistently getting round a 36″ radius.
When I experimented with a laminated spline roadbed (from Sundeala) I used a 48″ inch radius with a decent transition curve on entry and was surprised at just how well #8 looked and operated.
Rather than spend more time trying to persuade these iconic loco’s round tighter curves, I have chosen to use Precision Scale flex track, a minimum radius of 48” and pointwork built on a Fast Tracks #8 jig for the new layout.
Click here to view the related topic “Locomotive model minimum radius” on another page.
Click here to view the related topic “Coupler offsets and angular mis-alignments on curves” on another page.
Written by Terry Smith 12-10-2013, amended 14 & 17 & 22-10-2013. Updated: 25-May-2016 and 26-08-2016.