engineering calculation needed

Make that horizontal.

Based on the fact that the deflection would then be symmetrical about the horizontal member, so the root of the cantilever would not be rotated.

So what's right then? In my opinion engineering is all about accuracy. My first model may have been fairly inaccurate, but that's the way it goes: you build one model, think about it and discuss it, then build a better one.

We can't include it as we know nothing about the kinds of joints Steve intends to employ in the structure. I assumed that there are pin joints at each end of the beam because this makes the structure statically determinate. We know so little about the structure that this is only assumption one can make. Granted, I did not draw a pin joint. That was an honest mistake.

In which case, why are there big books of formulae published for structural engineers to use? Like Roark's formula book which Ned mentioned? Probably I could have figured out the formula from first principles if I'd chosen to spend an hour or two on it, but I didn't want to. Structural engineers don't begin all their calculations from scratch.

So okay, I am not a structural engineer. I'm someone who took a few structural engineering courses a while back. But I'm the only one who actually tried to help Steve in the first place. Others then joined the discussion, which I feel is a good thing if everyone remains reasonable, but now it seems some traditional Usenet insults are being thrown in.

And no one here has yet claimed to be a professional structural engineer...

Chris

Chris,

Not good enough.

No - engineering is more about managing inaccuracy. Before you can learn to do that, you must become proficient with the type of problem solving that appears in undergraduate texts.

Chris, it was nonsense.

Not in this case. You clearly lack understanding of the basics that go into such a model.

Again, nonsense. You called it a tension member, hence you were assuming frictionless pins and negligable weight. Otherwise known as a two force member, and it's item one in any engineering statics course.

Time wasted if the FBD is incorrect.

Clearly the tables exist (I consult them at times myself), and they exist to save time, not to escape learning. My statement stands.

If you still have interest, get a Schaum's outline on statics and start solving problems. It will eventually come back to you.

Wisely so. A pro won't touch something like this - responsibility w/o compensation. Some time ago, a poster suggested that the OP needs to find a structural engineer, which was very good advice.

Bill

Bill, you have yet to propose a model of your own. You could have proposed a model in the time you spent criticising mine.

An easy insult to offer, but as I asked, where is your model?

Consider my above answer retracted. Your question does not make it clear which member you're talking about, and now you've changed the force you're talking about from vertical to horizontal.

So here we have three guys who know a little about structural engineering having an argument. If we'd worked together we could possibly have been helpful, but I do not believe we are helping Steve now, so I will not continue this discussion.

I do, however, stand by my recommendation that Steve should try to avoid subjecting his 2" x 2" x 0.25" vertical column to a bending moment. I think he should either place one stand-off at the very top (15' level) and one at the level of the horizontal member (14' level), or change the design to something stronger like the "stiff leg" style of crane. People are welcome to agree or disagree with this opinion.

Chris

For the engineers in the group, and those who seem to know of these things .........

If you would like, e mail me at snipped-for-privacy@cox.net, and I will fax or send you a jpeg of my drawing, and you can get it just right.

I DO believe this thing will work, and I am not going to use it to lift things much heavier than a couple of suitcases or a few bags of groceries.

So far, I'm into this about $200, so it is not just a frivolous widget. But, failure can cause heavy damage to my structure. I think if I stay less than 200#, probably more like 100 most of the time, I shouldn't have a problem. I would just like to be sure, and hear from someone who can calculate such things.

Steve

Hi Steve,

As Bugs suggested, I suspect this is the best you'll get from this group. There probably aren't many true structural engineers who hang out here, and those who do may have reasons for not wanting to get involved (ranging from boredom with structures through to liability). There may be other Usenet groups where you'll find better help, but I can't suggest any. There is also the regrettable tendency for people to be non-constructive in their discussions on Usenet.

As has been demonstrated, one of the problems with your proposed structure is that it isn't as simple as it looks. Its behaviour is more complicated and less easy to predict than I thought, and probably than Ned and Bill thought, too. One of my main reasons for suggesting the "stiff leg" type of crane is that the analysis of the structure will be much more straightforward and less open to dispute. I would attack the problem by making some changes to the design (you should still be able to use the materials you've bought) rather than trying to find someone who's a professional structural engineer.

Best wishes,

Chris

Responsibility w/o compensation. Nice try, but I'm not going on the record with a solution.

It is not an insult; it is a professional opinion.

The identity of the member would be clear to an engineer; it would be clear to just about anyone since I used the unique name _you_ gave it. The h/v thing was a typo, no doubt the result of my distraction with trying to find a nice way to put this.

I think you will find the level of knowledge of the "three guys" varies just a bit. Hopefully the OP has figured out that your advice is not to be taken at face value.

Toward taking my leave of this thread, I will address some of your other comments here.

Speak for yourself. You haven't come close to the complex part (searching for the weak link in the chain so to speak). The deformation analysis is at the level of an undergrad homework problem.

More bad advice.

To the group at large, I offer an apology for the nature of this discussion, but I have learned to respect you. To not point out the flaws would be unfair to you, and potentially dangerous to the OP. I trust you will do the same should anyone give me bad advice about clamping or feed rates.

Bill

"Christopher Tidy" wrote

I would attack the

Thanks for your input, Chris. This is going to be mounted at a corner, and needs to swing in through the handrail, so a stiffleg is out because it would hit the building.

I have built a lot of "stuff", and most of it has worked. Anything that really needed to be engineered, I built from plans okayed by an engineer and provided by the builder.

However, as we all know, we do make Rube Goldberg things for ourselves with varying results. One of the major things to have is a positive solution finding attitude. It may take a prototype or six, and some modifications, but I have found that a workable solution to most problems is always found.

That said, there are all different types. I am sure you saw the back yard swing ..... as it was intended ....... as an engineer would look at it ....... etc. Sometimes, it's just like the big bucks fighting to the death while the little forkyhorn runs in and mates with the doe. A lot gets done by us little forkyhorns.

I like what Henry Ford said, "Whether you think you can or can't, you're right."

Intellectualizing gives me a headache and takes too much time. I like getting it done, and then ruminating afterward over some cold beer.

Steve

Bill,

I feel I must enter this thread one more time.

Amid your stream of criticism and confusion between horizontal and vertical, I gave the answer to a different question: the question "Why is there no vertical force shown on the left-hand FBD acting at the junction of the horizontal and vertical members?" I stand by the answer I gave to this question. The answer to your question is very simple. I believe the free body diagram on the left to be complete and correct. No horizontal force is shown acting on the top end of the beam because the upper support provides a reaction against this (see the roller support in my diagram). If you disagree with this, please elaborate.

I agree with your description of the tension member. I don't see what you're complaining about here.

If it's so simple, I still don't understand why you don't go ahead and solve it. From my recollection of undergraduate structures questions, this might have been a difficult second year "bonus" question. The kind which looks simple but isn't. We are also presented with the problem of having to model a vague structure from scratch; we haven't been given an idealised structure on which to perform the mathematics.

Why? Steve is unlikely to hire a structural engineer. Some structures are easier to analyse than others. Some structures are designed so that their behaviour is easier to predict. I've seen older bridges which are built with a roller support at one end. Why? So that the structure is statically determinate and easier to model. Changing a design so that its behaviour is more predictable is a sensible thing to do. If you disagree with this, please explain why.

You seem devoted to accusing me of offering bad advice without explaining why my advice is bad, or attempting to offer better advice of your own. Your refusal to solve what you claim is an easy problem is, to me, suspicious. It's all too convenient. The risk involved in offering Steve a solution is not great. You're not a practising structural engineer, so you needn't worry about your reputation, and if you say that your advice is offered without warranty I don't believe you have any legal worries. The risk doesn't concern me, and it doesn't appear to concern Ned either.

I tried to shed some light on Steve's problem in good faith. Perhaps the model I proposed was somewhat flawed, but you just comdemned it as "nonsense". At no point did you attempt to explain why it was nonsense, or how I might refine the model to give a better answer. You never explain why my advice is "bad". Furthermore, some of your questions and comments barely make sense.

I'm of the opinion that you're attempting to make me look a fool, and I don't appreciate it. An engineer contacted me by e-mail this afternoon and offered some suggestions about how to refine my model. Why didn't you take this approach?

I am also intrigued to learn that you work in an anaesthesiology lab:

Construction and installation of this major project will take place in Spring of 2006. I will keep all appraised. Photos will be available as fabrication progresses, at installation, testing, and finish.

What I intend to do is fully mount everything, and leave the middle standoff out. I will then make a 200# load and measure deflection. I will then install the standoff as a CYA feature. I am going to make a public guess on record that the deflection in the middle of 15' and the lifting point 2'6" off the vertical will be less than one inch. We'll see if I am even close.

Thanks for all the help, input, and insights into areas I had never thought of.

Steve

I get a better idea what you need now. I'm sorry this thread has turned into a flame war about mathematics with which none of use are entirely familiar. It does muddy the waters.

I can see that a stiff leg is out of the question. Is your building wooden, brick, stone or what? 200 lb is not an enormous load and I think it might be possible to attach the davit to the building without the long vertical column. If you do a Google image search for "wall crane" and "wall hoist" you'll see pictures of cranes which are attached directly to the structure of a building. Some of these are quite hefty.

As I suggested (but it got lost in the argument somewhere) I think the weakest feature of your design is that the load will be attempting to bend the vertical column. I'm not a structural engineer, but my intuition is generally good and this is fairly clear I think. If you took a long length of box section, clamped one end to the bed of a pick-up and leant on the other end, it would deflect quite a bit. But if you push on the end of the section, as if you were trying to push the pick-up along, it would be much harder to buckle it. So basically you could build something like one of those wall cranes, but have a column below which carries compression only. I did a little ASCII sketch (turn the fixed font on now):

-|\ -|\ | \ | \ | \ | \

-|---\ |---\ | | | | | -| | | | | | | | | _|__________________________|_______ \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

My preference is for the design on the left, as it avoids bending the column. But I do think my original suggestion of an 8" deflection without stand-offs is inaccurate. Given Ned's estimate, plus perhaps a little extra flexibility in the mountings, 1" to 2" seems more reasonable. But that's still a bit bouncy for my liking as flexing can cause problems such as pivots jamming, etc.

I'd be interested to see pictures and drawings. Feel free to post them online or e-mail them to the above address (remove NOSPAM). Good luck and let us know how you get on.

Best wishes,

Chris

Chris,

I don't know how to be any more plain about it: you are wrong. The horizontal reaction is one applicable force, but that does not change the behavior of two-force members. Your FBD is incorrect. It's not "less accurate" - it's wrong.

Bill

The problem with the drawing I provided is that it is waaaaaaaaaaaaaay out of scale. The top triangle is 1 foot x 3 foot x the hypotenuse.

Your idea of putting the second standoff at the base of the triangle won't work, because of the following configuration:

Starting from the base it goes something like this -from ground up, the first six feet is sonotube, next foot is joisting, next eight foot is column hooked directly to top of deck joisting, top foot is beam sitting on top of eight foot column. There's no straight continuous piece in the whole deal.

That is why I wanted the majority of the weight to be transferred vertically to the base of concrete in Sonotube. The standoffs are just to keep the load from swinging as I swing it in over the handrail.

My main concern was how much the 14' vertical would flex under a 200# load.

Steve

I don't doubt you can find *some* case where your model is predictive, but a model that only works in one specific instance isn't very useful.

In your original sketch you made the assumption that the triangle was a rigid body. Considering that the triangle is only 1 foot tall, and the rest of the tube is 14 feet, that seems like a reasonable simplification. In other words, I'm confident that the deflection I calculated is quite accurate, erring on the conservative side, as a result of my using the full 15 feet as the beam's length. In hindsight, it may have been more accurate to treat the beam as 15' long, but figure the slope at 14'.

Of course there is also a vertical compression load in the vertical tube equal in magnitude to the suspended load, but its contribution to deflection will be negligible, so I ignored it.

I think you could solve the problem this way as well, and the result would be within a few percent of the way I figured it. You would still have to calculate the slope of the vertical tube at the attachment point of the horizontal.

Unless we've completely misinterpreted Steve's description, the only thing I can think of that would have a signficant effect on the results of the deflection calculations would be a change in the way the structure is supported, e.g., adding another support near the horizontal member, as you've suggested. In that case the structure would be statically indeterminate, and the problem would become one appropriate for the end of a first sememster structures course instead of one introduced near the beginning.

Ned Simmons

For the record, since I'm an ME, but not a registered structural engineer, and am not privy to all the construction details of this device, I've deliberately avoided making a judgement on its safety.

Why would you assume that? It's a very simple problem - the only thing that makes it appear out of the ordinary is that it's more easily worked by looking at the vertical member's slope rather than the more commonly solved for deflection.

Ned Simmons

Ned: I could post further responses to what you've said, but apart from a few minor points I agree with you. Were we face-to-face we could discuss the merits of different models and likely reach an agreement on a good one fairly quickly, but it's rather hard via Usenet. Usenet does tend to be a little combative.

Bill: I am none the wiser. If you believe my FBD to be incorrect, please explain clearly where you believe I should add or remove forces and why. Better still, draw your own FBD and post it. Then I'll be glad to discuss it with you.

Best wishes,

Chris

Okay, got it. I thought the Sonotube was something you were going to install, rather than an existing part of your house. So you have what we in Britain would call a "balcony" (an outdoor platform on the first floor) and you want to lift stuff onto it? From your description I can see the problem of having a stand-off in the position I suggested, and I can see why you don't want to change the design.

Ned's suggestion of a deflection of an inch or so at the winch attachment point probably isn't far off. But I would be more concerned about the deflection in the middle of the vertical column: if it's close to the wall, it might jam. Similarly the bending of the column might cause the top and bottom bearings to jam. And of course you might just want to lift more than 200 lbs sometimes!

If it were me, my gut reaction would be to use something stouter than 2" x 2" x 0.25" box section in order to resist the bending. I would likely use something like 4" x 0.25" round tube. Some might say this is overkill, but it gives a good margin of safety without doing about calculations, allows you to lift heavy stuff without worrying about it, and I think round would look neat. I know a place where I could get some

Best wishes,

Chris

After all this hullabaloo, I will mention that I am a retired registered professional engineer. I try to make constructive comments, but will not attempt to suggest any design when all the facts are not given. If the loading is 200#, there are many models of pickup cranes available that are guaranteed for 1/2 ton or more. Buy one of them. If you want to use the material you bought, I would suggest an open web steel joist construction to fit the dimensions of your ?boat? plain tubing will deflect just as discussed.

____________________________________ !____________________________________! \ \ ! \ \ / /! ! \ \ ! \ \ / / ! ! \ \ ! !\ \ / / ! ! \ \ ! ! \ \ / / ! ! \ \ ! ! \ \ / / ! !

Well, you get the general idea. Properly designed & executed, the joist structure would pick up a Volkswagen. Look at a commercial construction crane for inspiration. Bugs

Chris,

Please note that I did describe the problems in each case. However, I will try this from another angle.

Going all the way back to Newton's Laws and how to apply them, a free body diagram is a diagram of a free body. A free body is one that has been isolated from its surroundings, with all interacting bodies represented by the forces and moments they exert _on the free body_. Some people treat force fields (magnetic, electric, gravitational) separately, but it's the same deal: didn't draw the earth? Put in the weight of the object facing toward the center of the earth (down for this type of work), and add support reactions.

Never show internal forces on an FBD.

One of the most insightful things I have ever heard on this front came the late John Wesley Hoover: "take it to pieces". Sadly you can't hear him bellow that, but do your best to imagine it (add a slight southern US accent for character). Pick any one or more bodies of a system, draw them and represent each missing item by the forces/moments it exerts on the system. Again, never show forces or moments internal to the system. Feel free to cut any member; when you do, add the forces/moments the part you removed exerts on the part you kept. I've repeated myself a bit, but it deserves repeating.

You cannot correctly reason about mechanics w/o a full understanding of this stuff. Sometimes you will get lucky and miss something that does not completely change the behavior; there is also risk of missing something that leads to dangerously incorrect results.

For drill, grab any statics book or relevant Schaum's outline and tackle the example problems; follow along on one or two, then compare your independent solution with theirs. In particular, look for problems suitable to the method of sections; that tends to pound on the rules, and is very useful for getting at specific information and/or reducing the coupling of equations one would suffer with a naive decomposition into the simplest possible FBDs.

From there, review technical beam theory. The FBD of a section of bent beam and the cut used to get at shear stresses across a vertical section through a beam will further reinforce how this works. Likewise, computer shear and bending moment by cutting the beam is good reinforcement. On the latter, pay particular attention to distributed loads: they can be reduced to resultants, but only _after_ cutting the beam. Doing otherwise leaves the support reactions unchanged but alters the internal shear and bending moment.

Re the top/left diagram we discussed, the "tension member" must be represented by its tension in the direction of the member. You can treat that as horizontal and vertical compontents, or as magnitude and direction. I prefer the latter because it helps in scanning for errors.

I suspect you drew an FBD of the pin supporting the load. Either way, start there, summing forces to zero vertically and then horizontally. Out of that you will get the tension in the "tension member" and the horizontally directed compressive force in the horizontal member.

Then draw an FBD of the beam itself, representing the members above by the forces they exert on the beam. It is the diagonal tension member where you have problems. Remember, for a weightless member with smooth pins, the tension/compression acts along the line between the pins. If its weight is significant, then draw a free body, including its weight and unknown horizontal and vertical components of reactions the pins apply to the member. To start, ignore the weight of the tension member and treat it as a two-force member.

Bill