Calculating a moment from a circular bolt pattern 3

[Idrivetrains]

I am designing a large structure that rotates on a large bearing. In order to order a bearing I have to have the horizontal loads, vertical loads and a moment. I have been designing trusses and transfer towers for conveyors, and so normally getting these loads would be straight forward; but as expected the bolt pattern in circular, and this has me wondering how to do this. I have the horizontal loads, and the axial loads. I have the vertical reactions at the bolt pattern, but how do I us the reactions in determining a moment? These reactions are positive and negative around a neutral axis, and so I know there is a moment in this bearing. My first approach was to calculate the moment based on the individual bolt reactions x their distance (horizontal not radial) from the neutral axis. Is this correct? I thought this would work since I would consider the compression/tension of a square structure the same way, and use that to determine overturning, and other stability issues as well as sizing the anchor bolts. Does this work for a circular bolt pattern as well? The bearing is 15' in diameter, and has 48 evenly spaced bolts. Can someone point me in the correct direction to figure this out?
Thanks

 

[Idrivetrains]

The reactions at the boundary conditions I set up, bear out what you have said. The "bolt" reactions alternate between really high and really low loads with of course compression at the front and tension at the back. Only about half of the bolts are taking the majority of the loads, and the half at the front and back are taking the greatest amount of load. What I also found is that the neutral axis is about 18" behind the geometric center of the circle. I ran the DL case and then multiplied the reactions of the bolts by their distance from the neutral axis, added everything together and got roughly 13500+- k-ft. I found the original moment by putting the entire structure on a single node with all its DOF fixed. When I ran the DL case again the moment reaction was 13700+- k-ft. I figure, since they were close to each other I am on the right track. At this point I have only run unfactored cases to check may balance, and to come up with the forces to give to the bearing people. I have also run single live, wind and seismic unfactored load cases in order to get Not to Exceed loads and send them to the guys designing the foundations. I figured I would get this question answered before I started running the LRFD loads cases to finish the design process. Your explanation really seems reasonable, and I think you right about the delivery of the bearings, although I have not had experience with that yet. It's good I have the weekend to absorb this before getting back to work. I really appreciate your help.

 

[BAretired]

Duh!

BA

 

[Lion06]

Just my two cents, but I think you should be summing moments about the geometric center of the bolt group to get the net moment. This is no different than taking a cut in a member to find internal forces (the forces always act at, or about the centroid of the section). The fact that you have line of zero bolt force that doesn't coincide with the geometric center means that you have axial load in combination with moment. I would analogize to building a P-M curve for a concrete column section. You assume different strain patterns and determine what the axial loads and moments are at those assumed strain conditions. When you do that, however, you are always taking moments about the centroid of the section, not the neutral axis depth.

 

[dhengr]

Idrivetrains:
Half the bolts (24) taking most of the load (that’s 12 in back) sounds about right, and half of those (that’s 6 or 8, by gully that’s what I said abv.) sounds about right too. I’d size those 6 or 8 to take the full load at yield. The 18" N.A. shift doesn’t surprise me either, without your programs, I guess that’s just some sorta Mc/I +/- (-P/A) shifting the N.A. Again, the brg. people will want to know what your structure is above the outer race, (a 15' dia. ring beam?) because you provide the system stiffness, they can’t. What percentage of your moments and load are DL & LL? How do you account for vibration and fatigue? Is there a industry design std., or what stds. do you design to? Where are you located?

If your boss (or another senior engineer) is a good practical design engineer, you should involve them in these questions. There is no reason for a young engineer to feel embarrassed by these kinds of questions when you’ve never seen them before, he should understand this. He knows your equip. much better than any of us do, and can give you much better, and to the point, guidance than we can. Ask him to be your mentor, he knows your operation far more than we do, history, std. details and ways of handling various conditions. He certainly has an interest in your getting it right, and will want you to do well. It’s much better if you can both be looking at the same plans, specs., suppliers drawings, etc., and be able to draw sketches, etc. as your questions come up. This can really be a rewarding working and learning relationship. With your apparent enthusiasm and general understanding of a mighty complex structure, I can’t imagine that he wouldn’t want to help you and work more closely with you. He wants you to do as much as you can on your own, but at the same time he doesn’t want you getting in trouble for lack of an intelligent question or two.

Idrive... a simple line drawing free body diagram, with loads and a few dimensions would meet BA’s needs, I suspect.
BA... There are probably 20 or 30 assembly drawings and sub-assembly drawings of that monster at that particular level, along with several arrangement drawing and piece details. Basically, it’s a 20x30 or 40' box structure, or space frame, about 6 or 8' high; with a 200' conveyor/loader boom hinged to one of the 20' sides. The whole thing rotates on a 15' thrust bearing.

 

[csd72]

It really depends if it is on a grout bed or on levelling nuts.

If it is on levelling nuts and there are a good number of bolts (at least 8) then you can model the ring of bolts as an equivalent ring.

Then using stress =M/s and P=stressxarea you get the result that P = 4M/ND.

This may also be a good starting point if you do have a grout bed.

 

[Idrivetrains]

dhengr,
Here is a sheet showing the reactions at the pinned boundary conditions. The axis is 7.5degrees off of the page centerline, but I have marked them so it should be easy to see how the reactions bear out what we have been discussing. Normally our company does conveyors and transfer towers. They are all square, pretty easy, and straight forward, and I have been doing them for four years (however, I was hired here 4 months ago.) This is the first time the company has moved into this type of thing. The company took a look at what the market offered, and thought we could come up with many improvements. That's when this hit my desk. Our senior engineer is retiring, and is spending less and less time in the office, but of course I will get with him and go over this. The only other structural engineer was out of the office last week, and so I figured I would put it out to a place that I have been reading and following for some time. Figured I would get some outside views, ideas, opinions, and perspectives from a different angle than what was available to me right at this moment. I have, and appreciate all that has been put forth. I have sized the bolts based on the few that are most heavily loaded. Since the machine has a rotating requirement of 270 degrees I am going to design all the bolts based on the highest tension/shear shown in the calculations since most will eventually see those forces. The included picture shows only the tension reactions at the nodes. I was able to get the information to the bearing manufacturer using your suggestion of giving them the reactions that the connection would be subject to. Hindsight being what it is I should have done that in the first place, since they know more about their bearings than I do, and should know the capacity of their design. We were able to order the bearing and it was only slightly bigger than my original estimation. Thanks for your help. I will now know where to start and what to figure out first when we go to expand this to a larger conveyor set up. Cheers

 

[Idrivetrains]

csd72,

The base will be on both leveling nuts and 4 inches of non-shrink grout. This is really the only way we can provide a flat enough surface for the bearing. Thanks for your help.

 

[rb1957]

why to the reactions alternate so much ?

 

[dhengr]

Idrivetrains:
I don’t understand the bolt force values either (in kips?), and I’m not sure we are talking about exactly the same thing. I understand the 7.5° or 12" bolt spacing, but not the 7.5° offset, I don’t really understand the two different kinds of bolts and exactly how they interrelate, or why are they are so close together to need the offset. And, I certainly don’t understand the wild variation in bolt forces, from one bolt to the next as you move around the circle. Assuming the back is at 12:00 noon; I would expect the bolts from 11 to 1:00 to take 60% of the tension from the boom moment, and those from 10 to 11 & 1 to 2:00 to take another 20%, give or take. In the front your structure just sees an arc shaped bearing force (compression from the ring beam) and the values are symmetrical about the boom axis. Their brg. sees the concentrated roller loads too. Then I’d superimpose the DL righting moment, and at this point the bolt forces should be pretty symmetrical; and finally I’d superimpose any lateral load moment you want to find a max. bolt force. I would expect that max. will happen at only two approx. symmetrical places on your bolt circle to the top race, but can happen anyplace within the 270° swing arc on the lower race. I think your sketch shows all of this, at once, and I can’t quite digest that so easily, in one step. I want to see that each of these makes common sense, then I’d superimpose them. Furthermore, I want to see that any variations in your structural system stiffness and the ring beam stiffness above the upper race is not radically influencing the bolt forces or loads on the brg., that will not be good for their brg. That ring beam really has to be almost as rigid as your conc. pedestal and grouting. If the upper structures stiffness is significantly influencing bolt forces you may want to study and adjust that.

Try to get that senior engineer to continue to work with you for a while, and not just on this project, but rather, as you learn the ropes. Maybe part time, maybe on call, etc. Even to the point that you get your boss to let you buy him lunch or a beer periodically so you can learn from him and pick is brain on company history and methods. Obviously, that depends on what he has to offer, since you’re all just starting out on this equip. That can be a wonderful interaction and quite helpful to you.

 

[BAretired]

Idrivetrains,

Perhaps you believe that you have explained your problem in a way that everyone and his dog can understand. Well I, for one haven't a clue what the problem is.

When I look at your diagram, I see that every second dot in the upper area of your diagram has a circle drawn around it. Are we supposed to know what that means? I do not.

Starting at the top center, the dot has 30.6 (no circle). Moving clockwise, we have 126 (circled), 29.4 (no circle), 121.3 (circled) and so on.

What do the numbers and circles represent and why the hell haven't you explained in before now?

I believe that the issue you are attempting to solve is relatively simple but I also believe that you owe this forum a proper explanation of your problem. So how about starting now?

BA

 

[csd72]

ditto