[Beowulf] flange clearance on sliding rails

Jim Lux James.P.Lux at jpl.nasa.gov
Wed Oct 19 20:00:46 PDT 2005


At 04:13 PM 10/19/2005, David Mathog wrote:
>We have two racks in the machine room.  They are both generic
>(ie, not from a major manufacturer like Dell, Sun, or IBM.)
>Neither one of them will hold a set of sliding rails properly.


This is a real common problem (TM) with sliding rails in general.


A couple things to check..
Just to get terminology defined:
The part attached to the equipment I'll call the sliding part
The part attached to the rack, I'll call the fixed part.
The x axis is across the front of the rack
The y axis is back through the rack
The z axis is vertical
Now, some dimensions to check...

In theory, the front panel of the equipment should be <19" wide
In theory, the equipment width should be <17" wide (that is, there should 
be an inch from the outside of the box to the end of the flange).
In theory, there should be 17.75" clearance between the inner edges of the 
mounting flanges
In theory, the center to center distance for the holes should be 18.3125" 
(that is, 18 5/16")
In theory, the holes are 10-32, and spaced 5/8"-5/8"-1/2" repeating  You'll 
also see racks with 1/4" or so clearance holes drilled in this pattern, and 
you put "speed clips" at the appropriate places or backing nuts (ick!).

OK.. here's the practical experience..

The fixed part has a right angle at the end which is supposed to mount to 
the vertical rails. It mounts on the INSIDE of the rail (i.e. from the 
front of the rack, the fixed part is entirely behind the plane of the 
vertical rails.

Sometimes, the vertical rail has its bend (moving from the xz plane to the 
yz plane) too close to the inside.  You get out your trusty rotozip, 
dremel, or hacksaw and shorten the little tabs sticking out from the fixed 
part of the slides.

Some of these things have weird looking spacers or a sort of combination 
nut (a piece of metal 1/8" thick, about 3/8" by 4", with 4 tapped holes in 
the appropriate spacing).  If your rack has tapped holes, and you want to 
use the funky nut, you have to drill out the holes in the rack to clear the 
10-32 hardware.  If you have clearance holes, you use the 4 holed thing as 
a sort of clamp to hold the slides in place (since the slide prongs will 
typically have slots that are somewhat bigger than the hardware).

In general, stuff that's mounted on slides doesn't push in as far as stuff 
that's mounted flush (because either the bracketry gets in the way, or the 
screw heads stick out, etc.

Once you've got the stuff on the slides, then you have to slide it in and 
see if the holes in the faceplate line up with the holes in the rack, so 
you can secure the equipment into the rack.  If they don't line up, then 
you have to move the brackets (having racks with removable sides helps a 
lot, because you can then reach the screws holding the fixed part of the 
slides.

Some blocks of wood, wedges, and a rubber mallet are very useful in getting 
the "fine" adjustment.

pray that your slides don't have too much junk in the back.  Sometimes they 
have an adjustable sliding thing on the fixed part to hold the back end of 
the slides. The screws that hold that in place often get in the way of 
internal rack structure.

As you can tell, this is why "rack and stack" is hardly as easy as one has 
been led to believe.  It's also why some brands of racks are a lot easier 
to work with than others.  An extra half inch of clearance makes a world of 
difference when trying to snake that screwdriver in along side something.

We do a lot of stuff where we use a pair of pieces of 2x2 inch angle (1/16" 
steel or 1/8" aluminum) that act as shelves along the side (the edges of 
the equipment chassis sits on the little shelf on each side. (using the 
formerly described coordinate system, one leg of the angle is in the +z 
direction, the other in the +x, with the length in the y direction.  The 
other side has one leg in +z, and one leg in -x.  There's a hole drilled in 
the angle (the +z leg) that matches with a hole in the flange of the rails, 
or in an added piece of strut inside the rack. (Unistrut or equivalent, is 
very nice for this.. you mount 4 pieces, 2 on each side, the height of the 
rack.  Then, you bolt your slide/shelves to the unistrut.


>When I set mine up I pushed the forward vertical rails
>all the way to the front of the case.  That didn't cause any problems
>with the flanges on the nodes which are permanently mounted.  However
>the flanges on the ends of rails hit a structural member on
>the front/outside of the case.  So the rails are displaced
>inwards a couple of millimeters  andthe resulting space is
>about 4 mm too narrow to hold a case.

Dremel tools are your friend.


>The other rack, from another group, has the same problem on one
>side.  For that one (square holes with the little clamp in nuts)
>there's absolutely nothing to adjust.  What's most infuriating about
>this case is that there's almost 2 inches of free space on the other
>side, all they had to do was to make the narrower rail a tiny bit
>wider, and the wider one a tiny bit narrower.
>
>In both cases the mounting flanges are only about 2 mm too wide.
>I've seen this now with rails from two different sources.
>
>Is this a common problem?
>
>Does somebody make sliding rails which have flanges that are
>just a bit narrower?
>
>At the moment we only need to mount one set of rails so I
>guess we'll have to take it over to the metal shop and have
>them cut or grind 2 mm off of the flange on one side.
>
>Thanks,
>
>David Mathog
>mathog at caltech.edu
>Manager, Sequence Analysis Facility, Biology Division, Caltech
>_______________________________________________
>Beowulf mailing list, Beowulf at beowulf.org
>To change your subscription (digest mode or unsubscribe) visit 
>http://www.beowulf.org/mailman/listinfo/beowulf

James Lux, P.E.
Spacecraft Radio Frequency Subsystems Group
Flight Communications Systems Section
Jet Propulsion Laboratory, Mail Stop 161-213
4800 Oak Grove Drive
Pasadena CA 91109
tel: (818)354-2075
fax: (818)393-6875




More information about the Beowulf mailing list