Reply With QuoteSomething like this is going to require some visuals. Please sketch up and post whatever you can.
Carrying moment load with roller bearings
Hi,
I need to design a bearing assembly that carries a bending moment load (as large as 7000 Nm) on a shaft with roller bearings. Axial load exists but is small compared to bending moment load.
What is the best way to carry this load in the smallest space possible and how do I find the force on each bearing if I use multiple of them.
Thanks in advance.
Edit:
I have added a sketch to be more clear but don't limit yourselves to this configuration.
Untitled.jpg
Last edited by roketman; 08-09-2012 at 04:30 AM. Reason: Sketch added
Something like this is going to require some visuals. Please sketch up and post whatever you can.
I didn't want to limit posters to what I have in mind but I put the sketch anyway.
My first and strongest suggestion would be to try to find one commercially available. There is great value in letting someone else work out all the bugs - and you will have bugs. Trust me.
If you need help finding a manufacturer you could ask here. I know several and I'm sure the other users know even more.
If you must design it from scratch your first concern will be to come up with a rail configuration that will withstand the load, and provide the stiffness and precision you require. Then you can look at bearings.
As far as how to calculate the force, begin by understanding that the rail really never sees a "moment", it only sees direct forces. Offset forces (often called a "force couple") produce the moment. Given the diagram attached, do you know how to calculate the forces? (A hint: only the outer bearings matter. You can eliminate the middle bearings. They contribute nothing to the system.)
rail section.jpg
7000 NM is huge.
I doubt you'll find an off the shelf bearing for this.
I think you may need 2 inline bearings with sufficient space between them to handle the moment couple.
Also, how do you get a pure moment? Is there a force at A distance?
jboggs this application requires minimum frictional torque on the shaft while it is rotating, which is the motive to use roller bearings in the first place. I believe having the rail to carry the moment will create enormous frictional forces.
Having a certain distance between the bearings would definitely help, but I thought having the space occupied with an extra bearing or bearings would increase the load carrying capacity. However that way the force analysis becomes very tricky as the system is statically over-constraint.
What shaft? All I see is a rectangular rail. If this is a rotary application maybe you need a ball spline.
Do a search ... Plain Bearings hydrodynamic... and a search plain bearings hydrostatic.
If you want next to no friction from your bearing and you're talking huge loads this is where you need to look.
Then grab your Machinery's Handbook and go to index for "Plain Bearing Design." It's toward the end of the book somewhere.
I don't know what sort of speeds and I'm too lazy to pick up my book to convert 7000 NM but Zeke or Joe said really heavy so that works for me.
Big heavy shafts spinning fast go hydrodynamic and slow speed have a pump to pressurized the lube and keep the shaft lube film in place to keep it suspended.
If this is huge and would require a special made huge bearing... you'd be looking at HUGE dollars. A spin cast (no sand) piece of bearing brass will be tiny money by comparison.
Good Luck,
Bob
Bob, for future rule-of-thumb, NM to Lbs/Ft is roughly .7 x NMOriginally Posted by RWOLFEJR
Soooo 0.7 x 7000 = roughly 4900-lbs/ft. (Actually 5162.935-lbs/Ft) but the approximation will give you a close ballpark.
Without seeing more diagramatically, I would envisage using more bearings to spread the load. The 7,000nm you mention is a lot of pressure. My previous family business we made spiral steel tube and had dozens of roller bearings along an assembly table to bear the weight of the tube as it came out of the machines. As the tube sizes and weights increases, we added roller bearings along the tables to handle the weight.
Bob, the shaft speed is low at around 25 rpm which would require a pressurized oil type journal bearing and in our case it is not feasible. I am confident that roller bearings are the most suitable.
As chixfashionz mentioned, I will need a multiple of roller bearings and possibly of different kinds. The question is to determine bearing loads for an assembly of more than 2 bearings, from the moment value I have (7000 Nm).
Thanks.
See my post above. The same diagram still applies. If you have more than one bearing base your calculations on the theoretical centerpoint of roller bearing contact. Design your mount so that the load on that point is shared relatively equally among all the bearings associated with that point. You do know how to calculate the force at a point based on a moment value, right?
jboggs, what you say makes sense, however with more than 2 bearings the free free body diagram of the shaft becomes 'statically indeterminate', i.e. I can't solve for 3 forces from 3 bearings using force and moment equality, I need one extra equation. I thought I could relate force magnitudes to local beam deflection to get the extra eqns I need. What do you think about this?
Put as many bearings in there as you want. What I'm saying is to base your calculations on the theoretical center point of contact for a group of bearings and then try to make sure the load at that point is shared equally (as much as possible) by all the bearings in that group.
Again, the closer any single bearing is to the axis of the moment the less value it has. Bearings at the center are a waste of money.
I finished my solution with the elastic deformation approach to solve for multiple bearing loads and the results are in line with what's said above. The center bearing does not carry any load. This concludes the thread. Thanks for all the replies.
