Custom N Loco Frame Casting/Buidling?

[tehachapifan]

Has anyone ever attempted to cast or build in some other way their own loco frames? Any success stories and/or how-to's?

If casting in metal is a bit far-fetched as a DIY project, I wonder if resin might work, where perhaps the inside of the loco shell itself could maybe serve as part of the mold.

Obviously resin would not be conductive, but in a DCC loco the motor has to be isolated anyway. Would have to figure some other way to route power.

Also, weight would be an issue as well with resin, but I'm pondering this for small switchers that would pull only a few cars at a time.

[peteski]

I have not attempted this but I have seen others kitbash frames for steamloco models. Either splice 2 metal frames or add a resin extension to an existing metal frame.

As far as switchers go, the smaller the model is, the more maximizing weight is important. It is not just about its pulling power - it is really about maximizing good electric contact with the rails. Weight is extremely important (even if the small loco just had to pull itself).

But if you want to scratch build, look at how brass models are designed. They don't have the conventional frame halves of the typical plastic model.  They are built up from various photo-etched pieces of sheet brass. Maybe you could do the same (then fill all the hollow spaces with lead.

[superturbine]

Here was an interesting experiment I did a few years ago.  It's a
2-10-2 cast from a 2-10-2 mikado frame I built.  It ran really smooth using only power pick up from the tender.  I suppose it would have been interesting to see if I could have added power pick up on the frame by any number of methods.  My plan was to do lost urethane brass casting..... unfortunately the shrinkage rate was slightly too much.

/>

[mmagliaro]

When I get back to making my all-hand-made 0-6-0 project (the one with the
scratchbuilt driving wheels), the way I plan to do it is to sandwich two heavy brass bars
together, drill them, tap them, put screws in them, and then machine out the slots for the axle bearings
in a single step (so they will match up exactly).
Then take the screws out, and put a thin styrene shim between the halves to electrically isolate them.
There will only be one geared driver, so I will then machine away some clearance between the two
halves to allow for a gear to sit in there on the axle.

I really am not a fan of plastic or resin frames: not for weight, nor for good pickup.  If you don't use brass split frames, you have to pickup through wipers.  And you lose so much weight by not having a metal frame.
The fact that you want to use DCC is not an issue.  Isolate the motor leads from the frame, and you
are all set.  Then you still route power from each frame half to the decoder inputs, and you have
axle-based pickup instead of wipers.

Honestly, if I were going to do it (and later this year, I hope, I will) I would machine it
out of brass.  In fact, even if I had to do it completely with hand tools, I would make it from brass,
and just slog through the process of filing out axle slots by hand.

[mmagliaro]

Here was an interesting experiment I did a few years ago.  It's a
2-10-2 cast from a 2-10-2 mikado frame I built.  It ran really smooth using only power pick up from the tender.  I suppose it would have been interesting to see if I could have added power pick up on the frame by any number of methods.  My plan was to do lost urethane brass casting..... unfortunately the shrinkage rate was slightly too much.

That's cool, Jason!  When you say the shrinkage was too much, does that mean that the axle slots
ended up slightly off, so the rod motion wasn't quite up to snuff?   I would expect problems like that.
A few percent off on a body or a detail part is one thing.  But for rod, driver, and gear alignment, things
really have to be on the mark.  Much more than a few thousandths of an inch off, and you're cooked.
I suppose you could cast it, and then use a caliper, test running, and a good eye to hand-file and shim
the slots to get it right.

I think that the plastic extension thingies we use to turn x-8-x into x-10-x, for example, can work
because it's only one driver being tacked on, and we can shim and fiddle to get it right.
I just don't know about 3, 4, or 5 drivers in a frame being made to work that way.

[garethashenden]

From the "Projects I need to get back to" files  , we can consider the British 2mm Finescale approach to chassis making. Brass wheels on stub axles are used with cut or etched frames. These are then isolated from each other with PCB. Plenty of space between the frames for the dense metal of your choice.

[superturbine]

Hey Max,

The urethane frame was just to insure that the frame worked.  When the frame was converted to brass via lost urethane casting, the shrinkage rate was too great across the length of the frame.  The bearing boxes and other parts were extremely close.  I actually got a brass frame to work but require filing which made it more work to build as compared to a urethane frame addition.

[daniel_leavitt2000]

As much as Shapeways can annoy me, the do have the ability to rapid prototype metal parts. I am pretty sure they use the same laser-metal media fusing that GE uses on their new LEAP jet engines.

If you can design it, it should work.

[victor miranda]

Forumers,

I have tried a to get a commercial casting firm to make
some lost wax frames for me.  I got a promise that they would maintain
dimensions.... and they didn't.  they said,  you have to pay for them and
then we will replace any that are out of spec.
it seems they destroyed my original frames,
they said that is why the originals didn't ship.

I inspected the lot...
so when I said they are all fails for various dimensions
they said... send in your original with the parts...

I have no idea why they quit taking my calls.


If you are going to cast, you will still need to machine. 
with the size of n-scale, machining out a single frame is not hard, just tedious.

pewter is lead free, a reasonable bearing surface and can be cast into RTV silicon.
so the theory is you can make a frame that way.

... do not solder to it.

luck.

victor

[nstars]

When he makes a new chassis my brothers normally machines out a frame using brass.

Marc

[tehachapifan]

Thanks for the replies! All good info but I probably should've specified I'm looking to maybe do diesel switcher frames, which add-in the issue of allowing for swiveling trucks maybe more than a steam switcher.

Regarding machining, say, out of brass, not sure exactly what that means. Sounds like you need some type of "machine(ry)" to do that, unless using a Dremel is considered "machining".

[Spades]

Russ

JnJ and NSN made replacement frames for the Kato/Atlas RS-3 framed GP-9.  John Coots/NSN produced other replacement frames and fuel tanks as well. I thought they were spin casting?  Jim Claudio? of Aztec machines frames for DCC conversions. It has been done, but who has that info?

G

[victor miranda]

Oh! 
a diseasel....

this way will get you most ...
this thread: to read about it if the photos are not clear.

Island Modelworks SW1001s
https://www.therailwire.net/forum/index.php?topic=16540.0

[tehachapifan]

Thanks, Victor! Those pics are great! What I envisioned was way more complicated than that and now I see it doesn't need to be. What exactly did you use for the bottom frame piece(s)?

[victor miranda]

it is printed circuit board.
that one is fiberglass and I recall I said it was 1mm thick.

while I think the copper is well glued to the board,
keep the fact that it is glued in mind when you build

for example, if you want a fuel tank of brass use screws to install it,
soldering it to the PCB will hold, and I would worry...

eta

the white parts are styrene glued to strips for the screws.
the joints are from nwsl 
I think I raided a stash of old kato parts for the pick-up strips
I think Atlas sells them.

after you build a frame you come to understand why cast zamac frames are how
most n-scale locos are done....

build one as an experiment and expect the frame to be lost for a strange oversight
the second will be much better.

victor