Atlas NSC 50' plug door box car - a little twist

[GaryHinshaw]

[Apologies in advance - this is an overly long rant post about a relatively small issue with this car.]

I have a few of these cars and (of course) I have been working on converting them to N-Possible couplers:



I have managed to convert these pretty easily (I'll post it to the N-P Journal thread later) but in the process I noticed something peculiar about my samples.  If I put a pair together on test track everything seems good:



Nice ride height, good car spacing, etc.  But if I happen to swap the position of the cars and repeat the comparison, they don't line up:



   It's not a huge effect, but all 4 of my samples were showing a similar symptom, so it was triggering my OCD.

My initial assumption was that the cars were listing because the floors were not seated properly, but that was not obviously the case.  Maybe the replacement BLMA trucks were loose and allowing the cars to tilt?  No, the trucks are fine.   I pulled the floors off and saw that the steel weights were screwed to the floor and on first blush they all seemed fine.  Now I'm really getting obsessed!  Upon closer inspection, I noticed that the end of the floor and the end of the weight were relatively twisted:




The culprit is the design of the attachment points: there are two posts that the weight screws into, but they are offset in opposite directions from the car axis:



Moreover, the posts do not quite reach the top of the weight, so when the screws are (overly) tightened they torque the floor causing it to twist.  The amount of the twist depends on how overtightened the screws were at the factory.   When the car body is attached the weight fits very snugly between some stiffening ribs and the floor clips into place.  I think the entire body is then getting twisted slightly in the process.

For now, I have removed the weights to see of the floor will relax back to being flat.  It does seem to be better, but still not quite fully true.  Hopefully with a bit more time, it will be fine, and I can reattach the weight without so much pressure.  I will also file down the inner ribs a bit so the weight is not quite so tight.  I'll report back on the verdict.

Otherwise these are very nice cars...

[Bob]

Very good detective work Gary - I don't think I would have figured this out.

[turbowhiz]

I had a problem with my examples of these cars along a similar line.

Its the uncentered weight's interaction with the stiffening ribs inside the shell that caused my issue as I recall. Janky. I think I ended having to tweak the holes in the weight itself in at least one of my examples to get it to fit properly. The factory wheels on these cars are terrible as well. I had a Rapido SP box that had a similar weight/underframe distortion fit fail, so this phenomenon isn't limited to Atlas or these cars.

I did use the factory extended draft gear which is at least correct for the ONR cars, although an extended length N-Possible box would be better. I have requests for those, so things along those lines are in the cards.

As a side note, these NSC boxcars are the prototypically correct base car for the CN track inspection car that MTL made not too long ago. If we could only combine the artwork/design of MTL with these Atlas cars... Oh wait a minute. Not that I'm holding my breath for that in the least.

[nkalanaga]

Would just backing the screws out solve the problem?  I wouldn't think they'd have to be tight, as there's nothing trying to remove the weight.

[Rossford Yard]

[Apologies in advance - this is an overly long rant post about a relatively small issue with this car.]

I appreciate your willingness to admit that this is what most of our (mine included....) posts really amount to!

[Sokramiketes]

Would it be possible to see the plastic floor without the weight installed?  Are those posts really not centered on the same line on the underframe, or do the large holes in the weight allow it to fall past the fins on the post?

[eja]

Why not forget the screws entirely?   

Make sure the weight is flat and glue it to the floor with goo or some other removable adhesive

[GaryHinshaw]

Are those posts really not centered on the same line on the underframe?

Yep.  I can hear you muttering WTF under your breath, just like I was.    I don't have a shot of the underframe without the weight but I can post one tonight.  In the meantime, here is an annotated version of one of the shots above:



The red dashed line is the centreline of the car and the weight, and the posts are indeed offset on either side of the centreline.   There are small fins (the solid red segments) to make sure the weight stays centered, but the torque is coming from the fact that the screw head only engages one side of the weight hole... 

The solution I came up with this morning is to add a washer so the pressure is roughly equal on both side of the weight hole, and not too tight:




I also filed the stiffener ribs inside the body so the weight was not pressed tight against them.  This solved the problem on 3 of my 4 samples:



The 4th sample still has a residual warp in the underframe, but I'm hopeful that it will relax back after a little more time with the weight removed.  We'll see, but for now I'm happy I understand the source of the problem and that 3 of the 4 were easily fixed.

[mu26aeh]

That was a prototypical twist to help modelers get the rocking boxcar down the tracks

[Sokramiketes]

Yep.  I can hear you muttering WTF under your breath, just like I was.   

WTF!