Bearing/Bushing Design

[Tpintar]

Hello,

I have a design issue with a bearing/bushing. I'm not really certain how to proceed in checking this element.

The attached image shows the schematic. I have three cylindrical bodies: the housing, the pin and the bearing. The bearing surface is equal to 2xDiameter of pin.

The Pin loading is cyclic changing the direction along the z-axis with small offsets inside the xz plane. The load is applied over the pin surface, not as a point load. The load is in the 5000kN area, and the resulting loads are in the 300MPa area.

The bearing must allow the pin to move axially along the y-axis and some 30 degree rotation around y-axis. It is for that reason hydrostatic lubricated with 10bar pressure grease system.

I assume the 10 bar is good for axial movement but when loaded with 5000kN the pin comes into direct contact with bearing.

_________________

Now the question is what is the critical design issue with this kind of element, what is this type of bearing called in English or if anyone could point out some useful reading material.

 

[GregLocock]

um, and roughly what size is this bearing that takes 500 ton loads?

Or to put it another way you are aware that 300 MPa is greater than the yield strength of mild steel?



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Tpintar]

Roughly 400mm ID, 17.5mm THK.

I never mentioned that the construction material is mild steel, and yes I am aware. The Pin, and base elements are not an issue strenghtwise.

 

[Walterke]

start here:

http://www.skf.com/portal/skf/home/products?maincatalogue=1&lang=en&newlink=1_0_1

P.S.: I have no affiliation with SKF, it's just a good informational website.

NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[Tmoose]

10 bar ~ 150 psi or 1 MPa.

I think To hydrostatically support 5000kN (~ 1,125,000 lbs ) would require a 7500 square inch pocket in the bearing shell.

 

[dhengr]

Would you please post your sketch in pdf format, I can’t see it as you posted it. I don’t think you are going to support that magnitude of load or that type of movement on a hydrostatically lubricated bearing. Your bearing is actually a journal bearing isn’t it? I don’t know what its dimensions are, except it sounds like the journal is about 2(pin dia.) long, and the pin axis is your y axis. At best you have a very narrow line load about 2(p.d.) long if the journal bearing can adjust to any deflection (movement) of the pin; and the bearing pressure will most certainly be greater near the root of the pin than at the far cantilevered end. Every time this stops or changes direction, it dwells for a second, and it will tend to squeeze out any lub and be running almost dry. Have you looked at self lubricating (non petrol lubricated) bearings?

 

[unclesyd]

Both of these approaches should be able to carry the load.

I'm a little confused as to the actual load being carried by the assembly, but I'll give it shot anyway. I have run Graphalloy plain bearings under some very high loads with great results. In the past these people have been quite helpful on several difficult applications.
I would contact Graphalloy with your specific requirements.

Again depending on your requirements you may want to look at a Tension Bushing for you application.

http://www.graphalloy.com/index.html

http://www.vogelsangcorp.com/products/tension-bushings/

 

[Tpintar]

@Tmoose
I agree that the numbers do not allow hydrostatic support for full load. The greasing is for axial movement when the pin is unloaded. Once axial movement is completed there is only a relativley small rotational oscillation. So a fully loaded case I would say is dry contact.

@dhengr
- sketch posted in pdf
- journal bearing - in a way yes or a bushing, but the rotational movement is almost negligble.
- your assumptions are correct
- simple beam calculation makes the loads at bearing ends at about 6000kN and 1000kN respectivley.
- no experience with non-lubricated bearings (how would they fair against wear over time)

@unclesyd
- I will have a look at the links, thanks

One question still remaining is how would I go about determining the actual loads, the pressure distribution, lenght of contact etc. All the information I found was based on support of rotational parts. Do I calculate Hertzial line pressure, if so how do I determine the lenght of the bearing/bushing loaded.

 

[dhengr]

Tpintar:
The links that Unclesyd showed are the types of products/materials I was thinking of, I just didn’t have the catalogs sitting on my desk. There are other products out there also. Why is lubrication for axial movement important, but not for rotational movement? Without giving the store away, you would help yourself and save us the game of 20 questions, by providing some more detail on that sketch. In rereading, I did see that you said the pin was about 400mm O.D., that gives some real dimensional proportions, that makes the cantilever from the end of the bushing out to “F” about 300mm. How is the 5000kN force applied out there and what causes the rotation about the y axis (about 30°, under what loadings) and the translation along the y axis (how much and under what smaller loadings). What is the actual structure out at, and around, point “F”? How do you fix that structure to the pin out there? In 300mm of pin length, you can’t grab onto that pin sufficiently to make it a real cantilever, can you? Why isn’t the y axis axial motion and rotation taking place right under “F” on a 300mm long bushing? Then the pin would be a shrink fit into the base and act as a firm cantilever. Is this a crane boom base or some such? With the proper structure around “F” you would have fairly uniform bearing pressure on the 300mm long bushing, not triangular. Then a tight fit btwn. the pin and the bushing will improve the bearing stresses. You’ll get to know Mr. Hertz pretty well before this is over. Maybe the pin should be hardened, chromed and polished, and the bushing be in the structure around the pin.

I can imagine how you got your 1000kN and 6000kN reactions without running the numbers, but in fact, they will be triangular pressure diagrams under that 400mm pin. That size pin will just tend to rotate as a rigid body in its canti. action; thus max. pressure at the bushing edge nearest “F” and going to zero 200-250mm +/- inboard, at the bottom of the pin; then another triangle at the top of the pin with a much lower max. pressure at the left edge of the bushing and moving right to zero in about 50-100mm. Then, why not use a front and back bushing of about those lengths? But, this arrangement might be much tougher to line-up and get the tight fit you want to lower the bearing pressure. The left end of this pin could be a smaller dia. and looser fit to facilitate assembly and still meet bearing stresses.

 

[Tpintar]

@dhengr

I will try to go through all the points you made. And just to note that the current design has been built and is working with cast bronze bushing solution. But, as I am new to the entire project, I would like to get a better understanding of the design and confirm it theoretically.

"Why is lubrication for axial movement important, but not for rotational movement?
I meant the other way around, saying that the current greasing design would support axial movement because of low loads, but since rotation occurs under maximum load the current greasing system would be useless.

The Pin connects two bodies in motion. The forces are a reaction to the motion of the bodies, as is the rotation. The pin moves axially to allow disconnect from other element. Full load and rotation occur only after axial movement is performed. (The system is similar to a door hinge with a retracting pin)

One side of the pin has two equal bushings of total lenght 800mm, other side is 200mm support similar to bushing. Side 1 and side 2 are separated by roughly 20mm clearance (so stress is closer to shear than bending).

Bushing is shrink fit and pin tolerance is set as close as possible to still allow axial movement (axial movement must be present).Pin has greater hardness than other elements.

Triangular force distribution, as well as new sketch of system is in attached file. Due to the position of the 0 stress point it is not very practical to have appropriately sized bushings.
Yes I understand the triangular stress distribution, and the concept of different sized bushings. But the size ratio is impractical due to other construction issues.

 

[unclesyd]

I would be careful not to paint yourself into a corner. i would try to stay with a std sleeve or bushing. The offerings by Spieth can do a very good job. It looks like something like the Tschudin & Heid Linear Rollers might work in your small envelope. If you can ope up the bearing area envelope the clamping sleeve is another alternative.

 

[Tpintar]

@unclesyd

I don't understand which corner you are refering to.
1. Most probably we will be going with the preiously tried solution
2. Envelope (if by this you mean dimension limitations of bearings) can still be changed if found inadequate.

Besides that, I am still looking for a way to determine the relevant stresses in design and how to correlate them to market products.

 

[Tpintar]

OK, a more detailed image is attached including static calc.

I used support centres to determine reactive forces, and calculated pressure distributed on bearing surface. I know this is a gross aproximation but I don't know if it is to great.

Anyway max. force 9900kN, distributed yield p=61N/mm2 - 61MPa, adjusted by 30% safety margin is about 80MPa. This is perfectly reasonable, but I am afraid that I over idealized the solution. FEM results give about 200MPa stress, and I trust this even less then the analytical results.

My questions are:

1. Should I adjust my static model, and if so, how?
2. Should I calculate contact pressure differently (Hertz seemed wrong for clearance fit.)
3. Bearing material is G-CuSn12 / CC483KK, yield point 140MPa, 80HB - how does this relate to max allowed pressure

Would be very grateful for any info.
Thanks

 

[Tpintar]

Any pointers...? I am sorry for bumping the thread but I would really appreciate any pointer on the actual stress estimate...