What's new
What's new

Pillow bearing massive failure

Overland

Stainless
Joined
Nov 19, 2017
Location
Greenville, SC
My son works with OEM's doing vehicle launches, off-road course design and build, and off-road driver training; freelance.
Anyway, he was asked to build a Teeter-Totter for an off-road course in TN.
So we built it pretty strong, and heavy and used massive pillow block bearings for the pivot.
The main frame weighs about 2,200 lbs and we figured 6,000 lbs for the SUV.
We found some 2.5" dia bearings with capacities of 9,000 lb static and almost 10,000 lb dynamic.
We figured should be fine.
However, even with some heavy shocks (F350 dually rear) it can come down pretty hard as the vehicle moves over it.
We sunk a railroad tie each end for it to come down on, and attached some sections of a skidsteer tire to act as a bump stop. The center of the bearing is about 26" high, and the whole structure is 20' long.
Well one of the bearings failed, as shown in the photo. The top blew off indicating a force lifting the structure.
Obviously there's not much motion and wear here, so we're thinking of another design, rather than a cast iron housing.
I'm thinking of a using a delrin bearing in a heavy steel housing. The shaft is 2.5" dia, and the bearings could be 3.0" long.
The photos are not that good, but will give you an idea of what I'm talking about.
Any other suggestions on bearing material ?
Bob
 

Attachments

  • D2A406D0-.jpg
    D2A406D0-.jpg
    555.1 KB · Views: 272
  • IMG_3798(3).jpg
    IMG_3798(3).jpg
    197.3 KB · Views: 273
My son works with OEM's doing vehicle launches, off-road course design and build, and off-road driver training; freelance.
Anyway, he was asked to build a Teeter-Totter for an off-road course in TN.
So we built it pretty strong, and heavy and used massive pillow block bearings for the pivot.
The main frame weighs about 2,200 lbs and we figured 6,000 lbs for the SUV.
We found some 2.5" dia bearings with capacities of 9,000 lb static and almost 10,000 lb dynamic.
We figured should be fine.
Ouch! Peak shock loads can easily double and triple the static load (8,200 pounds), so bearings should rated for at least 25,000 pounds are needed. Simply picking a better material (like delrin) is not going to cut it. Go big.

Don't offhand know if 2.5" diameter shaft is enough, but the bearing engineering data for bearings will help. For this kind of high-shock service, plain bearings are best. And it's a dusty dirty environment, so oil is going to be a problem. Perhaps Igus, if they go large enough. And use a shaft steel that will not fracture so easily, like the undercarriage of a big construction truck.
 
Last edited:
Plastic would be fine.

That design looks difficult to make while keeping the shafts in alignment. I'd have went with more of a sloppy clevis and a hard 4140 pin with decent surface area and room for grease.

If you want cool factor for the 4x4 crowd you should use a pair of unit bearings off a Jeep or something.
 
Garwood,
You're correct about the alignment, as we figured pillow bearings, no problem.

Yes Joe, we certainly under-estimated the loads, and didn't expect this lifting force.

This is obviously a pretty "agricultural" design, so we're not too worried about wear, as it's only going through about 30 degree total movement.
I'm more concerned now with this wonderful thing called hindsight, about the strength of the bearing housing.
The cross section of each axle is about 4.5 sq inch. so should be good.
Alignment of the bearings and shaft may well be a problem. I wonder if I machine a solid steel spherical housing to replace the cast iron housing. May not be easy on a manual lathe, but I guess pillow bearings have been around a lot longer than CNC lathes.

The first pic shows how the inner end of the axle is supported on the frame member.
 

Attachments

  • IMG_3714.jpg
    IMG_3714.jpg
    75.4 KB · Views: 96
  • IMG_3700(2).jpg
    IMG_3700(2).jpg
    583.8 KB · Views: 93
Last edited:
Think about making your pivot out of one piece of big thick wall pipe. Like 12 or 18” pipe full width of the platform and laying in an appropriate size pillow block. The block could be made from almost anything, even wood.
 
Think about making your pivot out of one piece of big thick wall pipe. Like 12 or 18” pipe full width of the platform and laying in an appropriate size pillow block. The block could be made from almost anything, even wood.
I think wood would be ideal for this.

Think like how old rail car axles used to run in a half bearing. Flip that idea upside down and use wood for the bearing surface against a large pipe.
 
Agree, the area of the case being case cast iron is not much to resist a big uplift when one end slams down,

Looks like the cross-section of the bearing case is about 7/16 so not even close to the strengths of the two bolts.

I would go with a 2.5 steel shaft with a pressed bronze or oil lite shell bearing on the shaft, and then fabricate a steel split box that would contain the bearing with the base thickness and the sides and top 1" thick..so being a custom machined simple greased bearing.

likely 5/8"+ thickness would be enough
 
Last edited:
It's in a pretty remote spot about 300 miles away, and also don't have the budget for re-engineering.
I thought about using a four bolt flange bearings with some sort of angle plate to bolt to the original supports, but that would raise the center height about 2 inches, and mess up the existing shock mounts.
The supports are 12" I-beam set in a pretty massive block of concrete.
I really have to find a bearing that will work the same dimensions as the pillow block.
I could make a simple steel clevis, as Garwood suggested. End float is not an issue as it really can't go anywhere.

Buck, the alignment of the shafts my not be that good, as discussed above.
Bob
 
Pillow block bearings have spherical seats for the installed ball, roller, or sleeve bearing. The spherical seat has limited travel and is intended to allow some misalignment of the supported shaft.

Your combined shaft and platform design has very low bending stiffness. This allows the shaft to bend and exceed the limited travel of the bearing.

The pillow block failed when the bearing reached the end of travel of its spherical seat. The single axle design suggested previously is one solution to the problem. The existing design can be repaired by welding a end plate to the axle and then welding this assembly to the channel section of the frame. The shaft would then be supported at two points 12" (?) apart. Once this is done the channel section needs to be closed off by welding a plate to the open side. You now have a stiff beam with a stiff connection to the shaft.
 
Last edited:
The alignment of the stub axles concerns me, as it never was an issue in the original design. So any "bearing" I use would need to be pretty sloppy. And the longer the length, the more sloppy it would need to be. The stub axles were put in pace with tape measure, squares, etc. so are probably not too badly aligned, but not good enough for solidly mounted "tight" bearings.

Robert, are you suggesting there was so much bending in the axle that the rotation of the bearing within the cast iron housing was such that it blew out the top of the housing ?
I find that hard to understand, as the main axle is made of W12 x 22 about 7 feet long.
The stub axles pass through the 1/2" thick end-plates and are then welded to the outermost "long" beams.
With a total load of about 8,000 lbs, I'm not seeing that much rotation in the bearing housing, in my mind.
Right now we've found the weakest link, the top of the cast iron bearing housing not standing up to the shock vertical load.
Just need to "move" that weakest link somewhere else, and hope we don't find the new one !

I'm liking Garwood's extra heavy pillow blocks, but $500 each.
I'm also like the simple block of steel with the hole in it, and lots of grease !
Bob
 
Fab a 1" thick steel top-bar, The hold down of high quality threaded rod,
A pair of nuts on the same bearing design, and a pair of nuts on top of the bar.

would likely get to near the strength of the bolts.
 
Set the bar on top of the same design pillow block with it having the sane bolt hole pattern as the pillow block bearing has.

Instead of the hex head bolts, a threaded rod would go into the structure with a nut under the structure, then a nut on the bearing ..and then another nut above the 1" rod that sets on top of the bearing case.

So the bearing, and 6 nuts(with lock washers), and two threaded rods and the bar on top of the bearing.
 
One thought is to machine up a a solid steel housing for the existing spherical pillow block bearings.
Any thoughts on how I'd bore out the housing on a manual lathe ?

But I do like the simplicity of the big block with a hole in it.
Bob
 
The added top -bar would be perhaps 1" x 1" x 6" with it having two bolt holes.
drill press or lathe with 4jaw to make the bar.

or 1,5 x 1.5 x6(whatever) IF we have a mechanical engineer here he might figure out the added strength. One would like it to be as strong as the lift strength of the bolts...and then hope the welds are as strong.
ALRO steel bar $20 each perhaps. same day. perhaps $130 for the parts + shipping to the site...
Or go you are overdue for a fishing trip:)

Strap if I-beam might be OK.

 
Last edited:
Garwood,
You're correct about the alignment, as we figured pillow bearings, no problem.

Yes Joe, we certainly under-estimated the loads, and didn't expect this lifting force.

This is obviously a pretty "agricultural" design, so we're not too worried about wear, as it's only going through about 30 degree total movement.
I'm more concerned now with this wonderful thing called hindsight, about the strength of the bearing housing.
The cross section of each axle is about 4.5 sq inch. so should be good.
Alignment of the bearings and shaft may well be a problem. I wonder if I machine a solid steel spherical housing to replace the cast iron housing. May not be easy on a manual lathe, but I guess pillow bearings have been around a lot longer than CNC lathes.

The first pic shows how the inner end of the axle is supported on the frame member.

The picture on the right also bothers me. I'd fear that the garter weld will fail from shock twisting. A second plate welded to both cross-beam and pivot shaft should prevent that.

A wax-impregnated wooden bearing seat should work, and would not be at all fussy about alignment or dirt. One could do the alignment with a laser beam or two and some paper.

One can buy rock maple bearings of this kind, such as Woodex.

https://woodexbearing.com/
 








 
Back
Top