What's new
What's new

Pillow bearing massive failure

The pillow block housings are made with ductile cast iron. It will have a tensile strength of 60,000 psi. The bearing housing side walls are .25 inch thick and 1.5 inches wide. The tensile strength of the bearing housing with a static load would be 2 sides x 1.5" x .25" x 60,000 psi =45,000 pounds.

The load on the bearing is the sum of the static load of 8,000 pounds plus the rate of change of the momentum of the truck as it drives on to the ramp. The second ( moving) load is determined by the trucks speed and the angle of the ramp. If the ramp and the approach driveway are both level and at the same height and if the truck is not accelerating or braking as it approaches the ramp the moving load would be zero.

The over load condition was caused by one of the following:

A very large impact load which exceeded the ductile cast iron's Charpy impact strength.
A very large dynamic load caused by the truck accelerating up a ramp.
A misaligned shaft in the bearing.
Keep in mind that the shaft is 12 inches (?) long and the bearing housing is only 1.5 inches wide. A 1000 pound load at the far end of the shaft would produce a bearing housing load of 7000 pounds if the shaft pivots on the inside edge of the bearing. This is why a stiff platform design is important.
 
Overland, I will check my Fafnir flange and pillow bearing collection tomorrow and see if I have any large enough to give you the added load range you might need. Will give you a call soon as I can
 
In short, I ran some very quick calcs and I don't think those bearings have a chance. You have obviously missed the upward inertia aspect. If your budget is on extremes, I recommend dumping the bearings entirely and go loose bore a couple big chucks of steel, and squirt in some grease. You have next to no cyclic expectations here, and no real speed that would warrant a roller bearing design.

Think BIG, think like the pin arrangements in an excavator. Massive forces on nothing but pins/bushings. You can ignore any kind of bushing here because it's just not needed. I think you will fight massive bending force on the pins here but apparently no budget to fix it.
 
That axles is going to fail. The weld is a stress riser and the outer portion could snap off. Or as already mentioned, the axle deflecting in the beam could fatigue a weld and break.

A little late now, but a great place to start would have been a semi trailer axle, complete with hubs. Strong tube, strong bearings, high load ratings.
 
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/
Joe, The 1/2" plate has a hole bored through. and the 2-1/2" bar slips through and is welded both sides. Didn't have quite enough of the solid bar so a piece of schedule 40 was slipped over it to reach the first "long" frame member.
The wood bearing idea is interesting. Thanks.
 
Maybe something I should have outlined earlier.
This is a fairly well controlled situation.
Drivers are allowed to use their own vehicle or one of the facility vehicles.
The instructors have been trained on their roles, and they are in each vehicle.
The idea is to try to balance the vehicle, before moving forward, so hopefully not high speeds, etc.
However there are people involved.
I'm a retired engineer and worked with a good friend who's a retired mechanical design engineer. We didn't do many detailed design calculations, other than the obvious beam calcs. We used our judgement, based on several years combined experience, and tried to go big.
Yes we missed the uplift forces, that destroyed the bearing housing.
I'm open to constructive comment, and appreciate the discussion here.
The vehicle speeds, and number of cycles per day are quite low, so we have confidence.
We'll go for a simple robust solution to the bearing issue, and see what else we can do to take some of the shock loading out by adding bumpstops.
I'll keep y'all informed.
Thanks again.
Bob
 
An old tire laid flat (less rim) is one of cheapest most durable bump stops you can find. Dig it down to give the desired cushion distance, choose different tire sizes for force absorbed.
 
Designing that apparatus from scratch I would have a double pillow block on each side of the pivot arm for each side. Making the blocks out of steel with one inch of steel around a 2 1/2" or what shaft with an oilite or bronze simple shell bearing and with seals made for grease filling. Cotter pins and washers to keep the shaft centered.
Super simple with not needing much high precision, Once set up it would be less costly than anything bought and would last 100 years.
 
Last edited:
The pillow block housings are made with ductile cast iron. It will have a tensile strength of 60,000 psi. The bearing housing side walls are .25 inch thick and 1.5 inches wide. The tensile strength of the bearing housing with a static load would be 2 sides x 1.5" x .25" x 60,000 psi =45,000 pounds.

The load on the bearing is the sum of the static load of 8,000 pounds plus the rate of change of the momentum of the truck as it drives on to the ramp. The second ( moving) load is determined by the trucks speed and the angle of the ramp. If the ramp and the approach driveway are both level and at the same height and if the truck is not accelerating or braking as it approaches the ramp the moving load would be zero.

The over load condition was caused by one of the following:

A very large impact load which exceeded the ductile cast iron's Charpy impact strength.
A very large dynamic load caused by the truck accelerating up a ramp.
A misaligned shaft in the bearing.
Keep in mind that the shaft is 12 inches (?) long and the bearing housing is only 1.5 inches wide. A 1000 pound load at the far end of the shaft would produce a bearing housing load of 7000 pounds if the shaft pivots on the inside edge of the bearing. This is why a stiff platform design is important.
Robert,
Appreciate your analysis.
My thoughts are that the damaging force is probably coming from when the T-T tips forward and stops abruptly with some inertia causing a lifting force. However there is a flaw to this idea in that the truck is not attached to the T-T and any truck inertia would tend to lift the truck off the ramp.
My thoughts are that all the forces when the truck rides up the ramp are compressive, or downward forces on the bearings.
The solid shaft sticks out about 5" from the 1/2" plate support. The inner support of this shaft is about 12" inboard of this plate. The bearing is 2.5" wide according to the specs.
I can't see how compressive stresses when the truck drives up the ramp can cause the damage we are seeing.
The bearing is designed to rotate somewhat within the housing, so I'm having a hard time seeing a significant rotation due to deflection, which could cause the damage we see.
We are told that after the bearing had failed, the shaft would lift up about 3" when the T-T tipped forward and hit the bumpstops.
So I'm not clear where this very powerful lifting force at the bearing is coming from.
Thanks
Bob
 
The 1100 lb half of the teeter totter that is going up as the vehicle slams down into the stop could have something to do with it.
Must be an awful lot of G to generate the forces Robert is talking about.
It's obviously coming from somewhere.
Never anticipated this level of force in the "up" direction.
 
Not knowing how you mounted the shocks , you know that auto shocks are biased to approx.20% compression and 80% rebound and are valved for way higher velocities than you have. You should have used some cheap ag hydraulic cylinders with flow controls and a small accumulator.

Many years ago one of our mechs was tasked with making a roll tipper. We use 60" 1-2 ton paper rolls.
He used 3" pillow bocks and hyd cyls to control it. It was used to tip the rolls from flat 90 deg to vertical & vise versa. I missed the first test and it ripped both the high doller pillow blocks apart.

And what soon became a ritual I got tasked with making it work. They make roll clamp trucks (which we have several of now) but at the time they didn't want to spend the money.
 
I'm a retired engineer and worked with a good friend who's a retired mechanical design engineer. We didn't do many detailed design calculations, other than the obvious beam calcs. We used our judgement, based on several years combined experience, and tried to go big.
Yes we missed the uplift forces, that destroyed the bearing housing.
I'm open to constructive comment, and appreciate the discussion here.
The vehicle speeds, and number of cycles per day are quite low, so we have confidence.
We'll go for a simple robust solution to the bearing issue, and see what else we can do to take some of the shock loading out by adding bumpstops.
I'll keep y'all informed.
Thanks again.
Bob
I mean, I probably don't care, but are you blowing smoke here? Or were you possible more in electrical or software? I only ask because I needed about 5min with my calculator to figure out the inertia problems. The fact that you "missed" the uplift components means nothing was calculated. That could not be missed.

I am not calling you to the floor. I am trying to understand if there was any math involved?? those axle shafts are asking for a real bad day. They need to be in double shear. Sure, an axle tube might work, but that is a poor design for this application! Why? Because it is a gross overkill for an application where loads can be managed with smarter solutions.

I mean, let me put stupid numbers out there. say you have two vehicles. One 6000#, one 5000#, both in the same position on this thing. What is the acceleration and deceleration of that 1000# difference?

How would I design it? I would assume there is no other vehicle on it at all so the entire 6000# is at the very top, and she comes down like a brick! I mean, the radial arc and XY forces are pretty easy to figure, but maybe I don't understand the budget and time investment.

I just want to put thoughts out there. Let's assume this 6000# weight is coming down at damn near full gravity acceleration, where do you think the other side is going? UP. Everyone talking about this energy absorption for the car, but how does that other side decelerate? through what load path?
 
Last edited:
I think a standard pillow block housing would be grey cast iron (that's what my RHP and NTN catalogues say). NTN in NZ used to offer S.G. Iron and cast steel options, but not in their later catalogues.

FWIW, I have attached a diagram from an old RHP catalogue, it's a guide to their housings steady (not shock) load ratings in various directions. It seems like a 4 bolt flange would be better than the pillow block, but considering the risk and work required, I think a steel pillow block would be better?

A couple of things I have picked up re-reading this thread:

-The bearings are mounted to posts which are set in concrete and therefore still on-site. So ideally any 'fix' needs to be easy to fit i.e. bolt on, without changing the pivot height.

-One of Overland's photos shows a stub axle welded to a single plate inside the channel, however that is not the finished job. The inner ends of the axles (or their pipe extensions) were later welded to the floor beams.

Blocks of steel with an oversize hole for the shaft sounds easiest. Take grease, a round file for the bolt holes and some shims :)

Blocks of steel machined for the existing spherical OD bearing sounds good (presuming the bearing size is adequate). I am not sure how you would do this on a manual machine, though a hydraulic copy would do it.

To get more complicated - you can get 'cartridge' housings for your self-aligning bearings. These are cast iron housings with spherical ID and parallel OD. So you could fit them into a home made steel housing. However, they will probably raise the pivot height a little (1/4" - 3/8"?). For example (using an NTN catalogue), if your existing 2.5" bearings have an OD of 4.7244", the cartridge OD would be 5.5118". Cartridges might move, bearings might not be up to it, daft idea?

RHP Bearing Housing Loads.jpg
 
Last edited:
The failure is tensile loading of the housing. This is most likely due to the bearing element being forced against the bore of the housing in a prying manner due to extreme moment loading resulting from shaft deflection.

As the shaft load increases, the ends of the shaft are deflected upwards as the neutral axis of the shaft approaches a U or V shape and the bearing unit twists in the housing creating a tensile load . . . no inertia involved, just plain old deflection induced stress.

This is most commonly accounted for with spherical mounted pillow block bearings, but even these have limits of allowed misalignment after which the housing fails as shown.
 








 
Back
Top