surging steam locomotive problem solved.

[peteski]

Same here - just gray areas with a do-not-enter signs.  Too bad Loren has problems with uploading photos directly to the Railwire - when it works it is pretty darn foolproof.

[Loren Perry]

Okay, let's try this again using the photo posting methods you guys have taught me.

A couple of months back I asked if anyone had any ideas about how I could fix a brass steam locomotive I had that exhibited a longitudinal surging while operating. It's a brass ATSF 4-8-4 imported by Hallmark and despite a number of good suggestions from RW members, I was never able to smooth out its running qualities, so I just set it aside and moved on to other things.

Yesterday, I decided to revisit the problem and tried again. This time, while running (and surging) through one of my curved tunnels, it derailed and I had to drag it out backwards about six inches. This caused it to throw a traction tire. After I carefully re-installed it and made sure it was on as precisely as possible, I ran the engine again, and Hallelujah! The surge disappeared and it ran flawlessly around the layout while pulling a 12-car heavyweight passenger train.

My theory is the soft traction tire had somehow partially slipped on the driver and bunched up on one side, or possibly even climbed partially out of its groove, making it a few thousandths of an inch thicker at that point. This caused the driver to tend to lift the engine with each revolution, but with all the tungsten ballast I had installed, the engine didn't exhibit any vertical motion that I could detect visually. The act of lifting the heavy engine on one side caused the motor to slow slightly with each revolution, hence the surging effect. The traction tire may have also caused the derailment in the curved tunnel by lifting the drivers just enough to allow the lead driver on the outside of the curve to slip over the railhead.

I never spotted the misaligned traction tire when I was troubleshooting the problem earlier. My attention was focused more on the rods and bearings. So if anyone has a similar surging problem with their steamer, check your traction tires and make sure they are properly seated in the driver's groove and not bunched up on one side.

This photo shows the position of the traction tire in question:




One more tidbit of useful information for steam aficionados - this same 4-8-4 once again started derailing as it entered my curving tunnel; I thought I had fixed it with the traction tire but there was another problem.

Inspecting the track work inside the tunnel opening with a flashlight, I spotted a short section where the rails' curvature isn't as uniform as I had thought. There was a short section where the curve radius increased a bit and then reverted back to a smooth contour. This increase was just enough to cause the big rigid wheelbase engine to climb over the outside rail when under a load. The problem's inconvenient location would require massive amounts of work to correct, so I decided to turn my attention to the engine instead.

My ultimate solution was to take a page out of the real locomotive manufacturers' handbook and make the pilot truck functional, i.e. have it exert a slight amount of lateral push to guide the drivers around the curve.

I did this by making a small flat spring from .003" phosphor bronze stock I had lying around. The spring is a single T-shaped piece with a small hole in the center of the broad middle leg of the "T". The hole is sized to fit closely over the threaded portion of the pilot truck's mounting screw allowing the shoulder of the screw (the slightly larger diameter section immediately below the screw head) to press down hard against it thereby locking it against the driver axle cover plate. The two "wings" of the "T" are located just aft of the screw and are bent up (or actually down in this case) 90 degrees to entrap a small projection that is conveniently part of the articulated "tiller" that joins the pilot truck to the frame. This arrangement causes the phosphor bronze spring to apply gentle pressure to the tiller when it's displaced left or right, guiding the engine frame into the curve. (If your engine does not have this rear projection on the tiller, the spring can easily be modified so it can be turned 180 degrees and fitted to the front portion of the tiller near the screw hole.)

After several tests under load with 12 passenger cars in tow, the engine now sails through this curve without a problem, and at virtually any normal speed. It also operated flawlessly around the rest of the layout and through turnouts of several types including Peco's Code 55 double crossover.

It's a very simple fix involving one small and very simple part with absolutely no surgery on any of the engine's components. It's even reversible in that it can be removed easily if desired.

Here are some photos to show what I did. This first one shows the assembled unit ready to run.




Here are the major elements to the pilot truck centering rig:




And here is a second view from a different angle:




The first part to go in is the centering spring, like this:




Next, the tiller is careful placed on top of the spring with the tiller's rear projection slipped in between the two vertical "wings" in the sheet metal spring toward the rear; both must have their holes perfectly centered over the screw hole:




The final part of the assembly is the insertion of the shoulder screw, seen here - make sure it's tight enough to lock the spring in position so it won't swivel:



Finally, some troubleshooting tips. If the drivers still derail at the severe curve radius, you should increase tension on the pilot truck centering spring by squeezing the vertical wings together so they make solid contact with the tiller extension. This will increase lateral pressure on the pilot truck wheels in the curve, applying more steering force to the lead drivers.

If the pilot truck derails (and probably also the drivers), then loosen the spring tension by gently spreading the vertical wings apart slightly. This will decrease the lateral pressure on the pilot truck wheels in the curve.

[mmagliaro]

Hey!  All the photos worked now!
One tip... you might resize them so that they are only 12" wide at 72 dpi (864 pixels wide).
They will scale more manageably in the browser window.  Right now they are pretty huge.

As for the project, now I see.  So the idea depends on the pilot truck being held in place by a shoulder screw,
so that the shoulder bottoms out and keeps the little "wing" thing firmly in place while the truck can still pivot.
I suppose if the part is made from phosphor bronze, it will keep its shape.  I'm will say I'm a little worried that with a strip that is thin enough to apply the right sort of light pressure (you used .003", right?), it will deform over time, gradually
bending away from the truck arm so that it doesn't apply enough pressure anymore.

[robert3985]

Congrats Loren! PHOTOS!!!

As to the fix for your engine...I would have NEVER thought of doing that.  Amazing that it fixed your kinky track derailment problem...really. 

I'm looking forward many more great photos from you!

Cheerio!
Bob Gilmore

[mmagliaro]

I'm still trying to understand why it works.
When the engine enters a curve, the problem is that the drivers try to climb up and out of the outside rail.
With that retainer in there, when the pilot truck wants to move over to go around the curve, the retainer somewhat prevents it, trying to keep it straight.  Since it cannot go straight (unless it also derails), the only other alternative is for the frame to move more into the curve to stay in straight alignment with the pilot truck.  When it does that, it pulls the drivers into the curve also, so they go around instead of derailing.

That's all well and good, except that there is almost no downward pressure on that pilot truck is there?  I mean, not compared to the big frame weight resting on the drivers.  It can only work, I suppose, if there is enough of a spring above it to get some serious frame weight on that truck.

Loren, is that how it works?

[Loren Perry]

Hey!  All the photos worked now!
One tip... you might resize them so that they are only 12" wide at 72 dpi (864 pixels wide).
They will scale more manageably in the browser window.  Right now they are pretty huge.

As for the project, now I see.  So the idea depends on the pilot truck being held in place by a shoulder screw,
so that the shoulder bottoms out and keeps the little "wing" thing firmly in place while the truck can still pivot.
I suppose if the part is made from phosphor bronze, it will keep its shape.  I'm will say I'm a little worried that with a strip that is thin enough to apply the right sort of light pressure (you used .003", right?), it will deform over time, gradually
bending away from the truck arm so that it doesn't apply enough pressure anymore.

Hi, Max -

I won't know for sure about the durability of the spring until I've run the engine around the layout for a time. It's less than a week old as of now. The way I see it, the limited movement of the pilot truck is so small that the side pressure against the spring only amounts to about 1/64 inch each way or so. That phosphor bronze seems to be pretty stout stuff. Time will tell. Anything stiffer and the pilot truck's wheels with their shallow scale-like flanges will start derailing due to excessive side thrust.

[Loren Perry]

I'm still trying to understand why it works.
When the engine enters a curve, the problem is that the drivers try to climb up and out of the outside rail.
With that retainer in there, when the pilot truck wants to move over to go around the curve, the retainer somewhat prevents it, trying to keep it straight.  Since it cannot go straight (unless it also derails), the only other alternative is for the frame to move more into the curve to stay in straight alignment with the pilot truck.  When it does that, it pulls the drivers into the curve also, so they go around instead of derailing.

That's all well and good, except that there is almost no downward pressure on that pilot truck is there?  I mean, not compared to the big frame weight resting on the drivers.  It can only work, I suppose, if there is enough of a spring above it to get some serious frame weight on that truck.

Loren, is that how it works?

You've got it exactly right. The pilot truck has no spring pressure holding it down. That's why the bronze spring has to be very light on its touch. The engine was derailing at only the slightest provocation (sometimes it stayed on the rails, other times not) so I reasoned that it would only take an extremely light lateral pressure to do the trick. So far it's doing it. Stay tuned.

[nkalanaga]

The design is different, but that's exactly how the prototype pilot truck works.  They have a curved bearing above the truck, with the highest point (furthest from the rails) on the centerline, and as the truck turns into a curve, the bearing lifts the front of the locomotive.  In turn, the weight of the engine then forces the frame to follow the truck, and the weight slides back towards the center of the bearing.  In Loren's case, he's moved the bearing to behind the truck, which is probably a lot easier to work with on a small model.

And, yes, I had no trouble seeing the new photos.

[Loren Perry]

So far this has been the only engine I was having derailment problems with. I'll keep a sharp eye on the other big steamers on my roster when they go through the same track work. Certain engine derailments occurred while going through turnouts but I was able to trace the problem to faulty point alignment on the turnout (i.e.. they were sagging and needed to be shimmed up to the correct level.)

[Loren Perry]

While still trying to perfect my layout and the operational qualities of my larger steam locomotives, I decided that the functional pilot truck on my ATSF 4-8-4 wasn't quite enough to satisfy my standards. Although the pilot truck modification did keep the engine on the track through a rough curve inside a tunnel, it was still struggling and I wanted to improve things further.

It took me a while to finally find it, but there was a virtually undetectable hump in the track at the tunnel portal that was part of the problem. Because of its location, I was never able to see it visually but finally, after watching the locomotive carefully, I was able to localize the problem. Seeing as it was just inside the tunnel portal, repairs would cause a lot of labor intensive restoration if I were to cut away scenery to access it. I decided on another approach: I used a flat file to grind down the railheads slightly to reduce the severity of the hump.

The hump was causing the engine's front end to lift slightly just as it was trying to pass through a slightly tight curve radius. This lifting action allowed the lead drivers' flanges to slip over the railhead and drop down outside the curved rail. The functional pilot truck managed to control this but it was really a band-aid solution to a more involved problem.

The plan was to smooth the radius of this vertical curve so it wouldn't be so abrupt. At the same time, I had to avoid cutting down into the rails so that the wheels' flanges contacted the ties. Luckily I was able to pull this off while using a freight car truck equipped with Micro-Trains "pizza cutter" wheels with deep flanges as a gauge to prevent excessive metal removal. I laid the flat file on the track inside the tunnel and used my fingertips at both ends (working via an open access hatch) to slide the file back and forth to remove excess metal. Next, I used a curved riffler file to reach into the tunnel portal and smooth the inner edges of the freshly filed railheads so the driver flanges wouldn't dig in and climb over the rail. This was followed up with a thorough vacuuming using a special miniature attachment on my shop vac. The engine can now run easily through this problem area even with its pilot truck completely removed, temporarily making it into a 0-8-4 wheel arrangement for track testing purposes.

I'm keeping the pilot truck modifications in place just for extra insurance. But I think I finally have a truly "steam friendly" layout.