Please Help Me With Structural Analysis!

[wderrick]

Hello all,

I am stumped on how to analyze the effects of a moment-load on a vertical cantilever support. I have attached a PDF of the structure I am analyzing... The load is evenly distributed onto the four green plates. the smaller of the two shafts is a screw jack, the larger is a steel shaft with linear bearings.

Theoretically the goal is to have the axial load supported by the screw, while the moment load is supported by the shaft. The base is welded, anchored, and gusseted flat stock. I know how to do cantilever calculations and all that but am confused by how the moment arm and moment load will affect my calculations.

Please Help!

Wes

(PLEASE NOTE THAT I HAVE HIDDEN MANY ELEMENTS OF THE DESIGN, CORE STRUCTURE REMAINS.)

assembly of new screw design structure.PDF

[PinkertonD]

Hi Wes,

First off, I would add another "nut" to the lower horizontal green member. May not be necessary but just makes me feel better for wear and failure about sharing an offset load.

I hope you are controlling slewing in the design to remove binding side loads on the lead screw. I will assume at this time, that slewing control may be some of the parts you have omitted.

You have two cantilever situations. Consider the first by mentally rotating your design so that the lead screw is horizontal. Since you say that the load is equal on all four pads on the green thing, then that load can be considered as a single constant load which will include the weight of the green thing. That will need to be handled at the top of the lift for acceptable deflection.

Because the lift is vertical, I think you could safely assume that load to be maybe 30% of the actual total load. That's just a guess, but I would also check it out at the full load and deflection as it may not be a great deal of difference in material size etc., for a safe deflection and I'd go with that as a safety factor. However, if that figure came out to about double (say 3" dia.) to that of the 30% (1-1/2" dia.) then I would consider the 30%.

The second is the green thing hanging off the the vertical post. I would ignore any stiffness benefits from the lead screw. You need some safety and wear factors built in and any stiffness the lead screw afforded, would add to that. Always good to have too much instead of too little.

[wderrick]

Ok, I will definitely use my full load calculation... I was worried that depending on the length of my moment arm on the green structure I may have to use a load value that is higher than the actual load.

In more detail than I discussed earlier, this structure must be analyzed for many different types of loads and determining how these loads affect the lead screw can be very difficult to determine. Axial loading is the only form of loading that does not concern me. The moment-Load could be a problem if it causes the lead screw to bind up. Finally I need to design for the event where a sudden large radial load is applied (forklift bumping into green structure or employee yanking on green structure) which could cause trouble if it causes the lead screw to bend!

When you talk about controlling my slewing, is there anything you recommend? Do you need me to attach a pdf of my slew bearing? (I know its a poor picture of it in the current pdf)

[PinkertonD]

Ok, I will definitely use my full load calculation.

I would work both the full load and 30% as it reasonable to assume that not all of the load is truly perpendicular to the vertical column. But as I said, if using the full load is not much difference I'd go with it instead.

When you talk about controlling my slewing, is there anything you recommend?

Personally, I would try for two vertical columns as it then takes care of that.

Finally I need to design for the event where a sudden large radial load is applied (forklift bumping into green structure or employee yanking on green structure) which could cause trouble if it causes the lead screw to bend!

And, that's why I suggested two columns, four linear bearings and the second "nut" for the lead screw. The thing would be intrinsically stable within the basic design.

[wderrick]

Pinkerton, I appreciate your prompt response!

Currently I plan on designing it to lock into position radially with a shear pin. Therefore, if an outside force acts on it the pin would shear and the structure would rotate away freely... do you think that is a viable solution?

It is interesting that you say this. The original design was exactly as you describe it, minus the second "nut". It was a structural channel with the lead screw inside the channel. Then a shaft located on either side of the channel... The design was changed to what you see in the PDF based on the fact that we were concerned about the moment load causing failure to the Turntable. We were looking at this turntable: Mcmaster: 6640K3 (30,000 in-lbs rating). The rating is just fine for my overhung load, which I generously estimated to be no more than 700lbs @ 26inches or 18200 in-lbs. However the design got rejected because of the unknown scenario where an outside force also acts on it (like a forklift). So, if I were to go back to that design, I have one of two options! Option A: I switch to a slew bearing like this Mcmaster: 6651K11 or Option B: I Stick to the same rotational design as you can see in the PDF, and rotate the entire channel around a support shaft.

The last option is to keep the exact same design as in the PDF and control the slewing by bracing it to some thing on the opposite side of the linear shaft.

What do people think?!

[PinkertonD]

Currently I plan on designing it to lock into position radially with a shear pin.

Huh? It is sliding. How's that going to work? Do you mean a shear pin sliding in a groove in the vertical column?

I think I am going to pass on further comment as the design path you seem intent on following presents, to this tired old mind anyway, too many forced compromises and ultimately a poor design overall.

As to anything being nudged by a forklift, hire competent forklift drivers. In 40+ years of Engineering design, I have never been asked to allow for damage from an errant forklift.

I wish you luck with that project.

[wderrick]

Huh? It is sliding. How's that going to work? Do you mean a shear pin sliding in a groove in the vertical column?

No, it would be a plunger pin located on the base connected to the slew bearing.

I think I am going to pass on further comment as the design path you seem intent on following presents, to this tired old mind anyway, too many forced compromises and ultimately a poor design overall.

Ok. Thanks?

As to anything being nudged by a forklift, hire competent forklift drivers. In 40+ years of Engineering design, I have never been asked to allow for damage from an errant forklift.

As such an experienced engineer, I'd think the idea of spending the extra dollar to beef up a design would make sense, if it could prevent the chance of that $100,000 accident, as unlikely as it may seam.

[PinkertonD]

Wes, you are not the first, in fact, you are pretty much following the "normal" path. We are forever getting people asking for design help. There is about 500-years collective knowledge, wisdom and skill here. Kelly, Bob, Joe and Zeke, just to name four off the top of my head, combined with my experience probably amounts to about 200 years.

It amazes me that people come here and ask for design help, but they continue to cling to their original approach even though it may be clouded with flaws. There is no way I (we?) would encourage you with your chosen direction when it is obviously not the best approach for what you are trying to achieve.

While my (and maybe others here) suggestions may not make for the perfect solution to your problem, they will make for a much better design approach. It is for that reason that I say "I will pass," as you keep coming back to trying to make your approach work. It is not about ego, it is about the best possible design to solve the problem on hand.

Given that, I have stated which direction I think the thing should go. If you choose otherwise, which you are perfectly entitled to do, I am not about to keep repeating my opinions, nor am I likely to eventually agree with something I don't believe in.

I bear you no ill will and I sincerely do hope you have success with the project.

[wderrick]

It is interesting that you say this. The original design was exactly as you describe it, minus the second "nut". It was a structural channel with the lead screw inside the channel. Then a shaft located on either side of the channel... The design was changed to what you see in the PDF based on the fact that we were concerned about the moment load causing failure to the Turntable. We were looking at this turntable: Mcmaster: 6640K3 (30,000 in-lbs rating). The rating is just fine for my overhung load, which I generously estimated to be no more than 700lbs @ 26inches or 18200 in-lbs. However the design got rejected because of the unknown scenario where an outside force also acts on it (like a forklift). So, if I were to go back to that design, I have one of two options! Option A: I switch to a slew bearing like this Mcmaster: 6651K11 or Option B: I Stick to the same rotational design as you can see in the PDF, and rotate the entire channel around a support shaft.

Here is what I said this morning. It was me rejecting my current design concept and attempting to revert/discuss the concept YOU had suggested to me... (two shafts instead of one)

[zeke]

".....The design was changed to what you see in the PDF based on the fact that we were concerned about the moment load causing failure to the Turntable. We were looking at this turntable: Mcmaster: 6640K3 (30,000 in-lbs rating). The rating is just fine for my overhung load, which I generously estimated to be no more than 700lbs @ 26inches or 18200 in-lbs. However the design got rejected because of the unknown scenario where an outside force also acts on it (like a forklift). So, if I were to go back to that design, I have one of two options! Option A: I switch to a slew bearing like this Mcmaster: 6651K11....."

Makes no sense. So your company rejected a superior design based on potential failure from an unknown exterior force and you consider a weaker design which is even more vulnerable to that same mystery load.

Your present design as Dave said is very poor and in my opinion not worthy of further consideration. The difficulty you have in the analysis is not only due to the indeterminate nature of the problem, e.g. removing the couple from the offset load to be borne either wholly from the post or a combination of the post and the lead screw. Potential binding along the linear bearing or the jack screw due to the curvature caused by the moment or sidewise forces could be analyzed , but why??

I would strongly suggest you go back to the 2 post design and then ask specific questions if needed.

And furthermore I also never heard of designing for a forklift accident. If you did, how would you determine the impact load??

[jboggs]

A new commenter here. A few thoughts starting at the bottom:

First, draw a vertical line from the CG of the entire mechanism, including payload, down to the floor. The moment on the turntable bearing is the product of the horizontal distance from the center of the turntable to that line and the total weight of the structure. The turntable bearing, and the attachment to the vertical framework must withstand that moment.

Second, if that lead screw is as long as you have pictured, you have an EXCELLENT candidate for buckling, especially if it will be rotating with any kind of speed at all. Once it starts whipping, you're toast. You can: (1) redesign it so that the lead screw is in tension rather than compression (that eliminates the buckling tendency and lessens the effects of spinning), or (2) make the screw bigger and stiffer, or (3) replace it altogether with a belt or chain drive doing the lifting (like on a forktruck).

Third, If I understand it correctly, your plan to resist the twisting action of the carriage along the single rail is to use a pin in a slot along the rail. Right? If so, you have no real resistance to twisting at all. You MUST have two vertical guide rails and four linear bearings.

You have two moments to deal with:
> one is seen by the vertical rail. The upper and lower bearings exert equal and opposite forces on the vertical rail, a force couple which creates the moment it must resist as a beam.
> the other is seen by the turntable. It is created by the horizontal distance from the center of the turntable to the CG of the mass. Discussed it above.

As for forktrucks? All bets are off. You aren't building a guard. You're building a mechanism. If you want to protect it from forktrucks that is a separate structure, one that prevents this structure from seeing ANY forktruck related shocks.

You've now heard from three of the four names Dave mentioned, all telling you pretty much the same thing.

[wderrick]

You guys are great. I was quite frustrated yesterday, because the two shaft design that you guys are recommending was my original concept (that I had drawn up two weeks ago). My boss rejected it because of the possibility moment load combined with the possibility of it hitting a fork truck (I know that you guys don't agree with designing for that). That being said, after hearing from pinkerton yesterday regarding slewing, I did analysis on the screw and found that I would pass my yield strength (leading to deformation in the screw) with only a central concentrated load of 230lbs being applied to the center of the screw! (Houston we have a problem).

So, forget about the single shaft design, I met with him yesterday afternoon, and I am back on the two shaft design Assembly.PDF

However I found what Mr. Boggs said about the screw to be alarming, or at least worth looking into. Off the top of my head I can't think of a good way to put the screw in tension... any suggestions? On a side note about that, the two shaft to be used are 1.5" diameter hardened steel, so I feel like the side load on the screw should be verrry minimal (but obliviously buckling isn't fully dependent on that).
^and the screw is 1"-4

Zeke asked how I would analyze the force of a fork truck anyways. Basically I am going to design with the beefiest turntable I can. Then I will take its moment-load rating, subtract off the overhung load, then divide by a moment arm of like 1 meter, and from that get the max force. From there, my plan was to say that I have a 4500kg or 6000kg fork, that would have a collision time of sayyy 1/20s... and use the equation of velocity = Force*Time/Mass to determine the max speed the lift could take... (Whats the point of finding this out?... I am still working on that part haha)

[wderrick]

Mr. Bogg, I needed the idea to sink in... but upon further thinking about what you said, I am going to switch to a chain drive I believe! thanks for the great advice, I got my boss to say: "I can't believe I didn't think of that"

[jboggs]

That's good to hear.

More thoughts on your linear rails: the arrangement you have now requires them to be perfectly parallel, which in turn means the mounting surfaces must be machined that way. You should plan on some shimming, adjustment, etc. In my experience this is a VERY common mistake in application of linear bearings. I suspect you don't really need "linear bearings"; "linear guides" would probably suffice.

One option I have used in similar applications with good results is V-wheels (see Bishop-Wisecarver or Osborn Load Runners). They have several advantages. They are inherently self-cleaning and less susceptible to dirty environments . They do not require round or even precision surfaces. They can run on round, but they can also run on angles or square tubes at an angle. Precise alignment of parallel tracks isn't critical. Its easier to design a structure with the stiffness and moment resistance you will need. They have simpler maintenance and are easier to replace. Check them out.

[PinkertonD]

Good to hear Wes.

Regarding the lead screw approach, I had assumed because of the reduced diameter at the top, that is where the motor would go so the lead screw would be in tension. Given that we are talking about a relatively light load (700-lbs) I would not completely disregard the lead screw yet, as it is a simple system to build. It also has the advantage of self-locking when stopped at any position in the lift, providing you are not using ball-screws. I'd still use two nuts!

With the other guys chiming in with more suggestions and advice, you may be back on track here.

[wderrick]

I don't see V-wheels and other guides having the positioning control I'd like to achieve... especially if it was a chain system, however they could work with a screw setup. Currently the linear rails are planned to be hardened precision shafts with very good OD tolerances but it doesn't seem to have a per foot straightness tolerance. Regardless, I do plan on using "self aligning" bearings and it also allows me to specify for "extra tolerance" which is what they recommend for parallel shafts... Also these linear bearings are self lubricating liners opposed to actual bearings!

I do not understand why my motor would put the screw in tension? The plan was to use a screw mounted sprocket to chain to sprocket on top of worm gear face mounted motor setup.

Pinkerton, the reason why I am so willing to switch over to chain system is because I already was talking about using a 112" long screw... too long I know. The linear shafts are 96" or 108" depending on what I end up going with. A 96" gives me a stroke length of around 70" if I recall correctly, but will end up slightly smaller because of my limit switches.

If I were to use the chain, I plan on using a motor with a electromechanical break... am I correct to assume that a worm gear wouldn't suffice as a break?

[jboggs]

Exactly how precise do you need this motion to be?

Worm gear boxes are usually self-locking, depending on the actual ratio. So they can be very good safety brakes. Check with your supplier to be sure. On the other hand they aren't very precise in actual stopping position without a brake.

Speaking of precision, I don't see a lot of it in your framework. Looks to me like a single thin web I-beam as the main support. You aren't going to get much precision out of that over a 100" span. If you need the level of precision that linear bearings are usually associated with you're going to have to fabricate a long stiff weldment and machine it on a very long mill.

Dave, you are the machinist. Do you agree?

(And whatever precision you have in there will be absolutely gone after ANY contact with a fork truck!)

[PinkertonD]

Joe, you are waaaaay too cheeky. Loved that last comment.

If Wes is hell bent on using linear bushings, and I have no issue with that if he is after accuracy, I'd go with open-sided ones so he can add supports along the travel rather than 8' of linear shaft supported at only the top and bottom. With a bushing-version of these...

http://www.grainger.com/Grainger/THO...Bearings-3DWE9

...he can have supports every 12" and reduce the shaft size significantly but still get accuracy and rigidity as the load is not that great. Also he can then get back to shims instead of machining the entire support structure first. Although, running it through something like a Blanchard might also be a good thing and save a lot of shimming.

Wes, the motor would not be in tension but the lifting effort by the lead screw would be and thus no buckling problem. Moot anyway, a chain drive is a good solution too. There is rarely only a single "best" idea for stuff, so go with what you feel will do the job as long as the overall design makes sense.

{edit}
As JB suggests, a worm drive should be OK for self-locking depending on ratio. BUT -- make sure it does not have an internal over-run clutch! Don't ask!!!

[wderrick]

I would love to talk about the framework. Keep in mind that the frame work you are looking at was intended for the screw (no axial load on top, only a moment load). What you see in the picture is 1/2" thick flat stock that is tapped and bolted together (bosses idea). I would like to switch it to a structural I beam of some sort, or maybe a long flanged channel... but really I'd like to hear what you guys think would be an ideal structure, for the chain lift or for the screw lift (I am definitely leaning towards chain). However I will say that I like the "channel shaped" structure because it gives me a convenient chain safety guard, and also a convenient/sexy way to mount my sprocket shafts.

Currently the machine Power design I am leaning towards is as follows, #80 lifting chain with load bearing sprocket on top of structure on same shaft with another sprocket, that sprocket would then have something like a 5:1 ratio that would go to another shaft which also has a worm gear on it (10:1 possibly). The worm would be driven by my 3 phase brake equipped motor.

As far as using open bearing with the supports every 12" or whatever. That was originally what I wanted to do, but I would have to modify the mount of my linear bearings by 90 degrees in order to use those... this is why I chose closed shaft bearings... keep in mind that I did do bending analysis on those shafts and it is not an issue (thats why they are soo beefy though). So! if you think that those open bearings are the way to go for alignment reasons, do you have a structural recommendation for those?

haha although hell bent maybe a little strong choice of words, yes its true I am a fan of linear bushings

[PinkertonD]

haha although hell bent maybe a little strong choice of words, yes its true I am a fan of linear bushings

I only mentioned that as they are the most expensive and for or a 700-lb load and chain lifting, it seems a waste of money. With good design and correct adjustment, "V" or "U" rollers can afford pretty close to the same accuracy and wear life. Since we don't know what this is, nor do we care, or have to know, it is hard to say if linear bearings are over-kill or not, but it appears they may be. But again, that is entirely up to you. There is no right or wrong with either lifting and guiding method if you have the money.

As to further design assistance discussions, I think we have got you started and I am happy to help tweak. If this is your job to design stuff for a salary then I suggest you get to it. As I said before there is rarely only one correct design for anything and there are a myriad of ways to do what you want to do.

Sooo, figure out what is the best way for your design. If you have questions about technical aspects then by all means ask again, but for now I think I will just watch with interest.