Ways to prevent slippage of a structure under testing

[MarleyDH]

So I have a structure that I have to test, or more so a "model" structure.

It is in the shape of a T, but rotated on its side, so that the stem of the T is a cantilever beam. An actuator will be used to load the structure at the free tip of the stem.

My task is to design the supports of the structure such that it can be loaded in such a manner. I will be attaching bars/cables to the top to provide restraint, and will have a base plate at the bottom.

Herein lies the problem. The base plate rests on a concrete floor. I am not allowed to alter that concrete floor in any way, nor are there holes in the floor close enough to bolt the structure down.

So how do I prevent the base plate from slipping? Because if it slips, the results will be inaccurate and there is the risk of someone getting injured. There is also no way to fix the top of the column to something rigid.

 

[pantaz]

Size -- is it 6 inches high, 6 feet high?
Materials used for the structure?
Forces -- Is the cantilever being loaded with 100 grams, or 100 kilograms? Is it pushing/pulling up, down, sideways?

 

[MarleyDH]

The structure is made of structural grade steel. In our code its 350W steel.

Its 1m high and the cantilever is 1.5 m long. Its going to be loaded by a hydraulic actuator which is capable of producing a point load in its current configuration of 2500 kN, but I hardly doubt we will reach that. Its a vertical force going down.

 

[pantaz]

What are the limits of the baseplate design? If the baseplate can extend all the way under the cantilevered beam, then it should be easy to counteract the actuator's force.

Also, if the column is supposed to remain static, I would add bracing, something like one of these:

Note: The compression style bracing in top sketch is not attached to the beam -- it would be a pair of braces on either side of the structure.

 

[MarleyDH]

Thank you for the input.

There is no real limit to the base plate. I did note however that the lab floor had regular fixing holes in a set of 4 spaced evenly along the length of the building. The sets occurred in pairs. So to draw up a diagram using ASCII, it would look something like this

+ = bolt hole.

+ + + +
+ + + +

+ + + +
+ + + +

Where 2 sets of bolts are used for the A-Frame assembly which has the acutator on and the other bolts are free. I didn't have a tape measure with me so I couldn't measure the distance between the holes.

That pattern repeats itself for roughly 10 metres, before it stops at our Wave Tanks.

In fact one of my designs had the base plate extending in the opposite direction, but the problem I forsaw was the the possibility of that base plate buckling.

Also, thank you for the idea about the compression bracing, I totally forgot to even consider compression. Perhaps its because compression is that much more dangerous than tension - buckling AND yielding/fracture, as opposed to just yielding/fracture.

Ideally the entire system has to remain static, with only small deformations happening (less than say 20 mm at the maximum point).

 

[pantaz]

... Perhaps its because compression is that much more dangerous than tension - buckling AND yielding/fracture, as opposed to just yielding/fracture. ...

Buckling doesn't necessarily lead to fracture. Quite often, partial buckling can serve as an indicator of failure, while still maintaining some support. Failure of a tension member tends to be sudden and complete.

 

[MarleyDH]

Oh no no, don't get me confused. I know buckling doesn't always lead to fracture, its just compression members usually fail in one of two ways, whereas tension members I find only really fail in one.

But then again, my exposure to failure is extremely limited as I am still an undergrad. Just one more year to go.

 

[nvn]

Compression members usually fail in one of two ways -- buckling AND yielding/fracture, as opposed to just yielding/fracture [in tension members].

Well said. And this is why the second diagram in post 4 is more efficient, stronger, and simpler, except that the column base plate does not need to be large, and the tension diagonal member preferably should be slanted at an angle of 45 or preferably 30 deg, measured from the ground, if you can afford the space. Tension members tend to fail gradually. Of course I am referring to your stated material. The tension diagonal will not slip if you use adequate bolt installation torque and enough bolts.

 

[MarleyDH]

So an update:

After going back and forth with my supervisor. This is what my design has come to for the support rig assembly.

Support Rig Assembly:

Total Assembly

Its pretty much spot on, but, its lacking a few things.

1: I need to provide the bracing to prevent sway at the top. That is no problem to do.

2: I have to reduce the length of the support rig. This is the issue I'm having as the loading rig sits on an A-Frame between the two "beams"

The A-Frame

The A-Frame is made from the biggest H-Sections that my country produces. That is a 305x305x158 H. The base plate that each of the H-Sections is welded to is a 500 x 500 x 30 plate. Which then to prevent the base plates from warping severely has gusset plates to transfer loads to 4 x M40 8.8 bolts per plate. To connect the two columns they have the biggest I-Section we have, which is a 533x210x122 I. The span of the I-Section is roughly 1400 mm.

The issue

I have to marry the support rig with the A-frame assembly by making the support rig shorter. I'm pretty much at a loss at this point right now.

Any help will be appreciated.

 

[nvn]

MarleyDH: I am currently having some trouble understanding your structure, because you currently show only one view, presumably a top view (?). Normally, you show two (or three) views (top, front, elevation, side, end views). And the dimension between floor bolt hole groups, vertically on the page, is currently not shown. Also, I thought you had eight bolt holes per floor bolt hole group. Or did your ASCII diagram in post 5 malfunction? What is the distance between adjacent bolt holes in the eight-bolt-hole group? Also, I am not sure if it makes sense to use massive I-beams, but then rotate them to their weak direction. I would think you might want to utilize their strong direction, instead. But we currently cannot visualize your structure too well, until you show a dimensioned front/elevation view.

 

[MarleyDH]

Correct it is a top view. My apologies, when I was drawing them for my supervisor, the other views were not needed. The floor layout is square. So all intrabolt group distances are all 460 mm and interbolt group distances are 1380 mm, with an uncertainty of 5 mm.

The use of the I or H section on its weak axis was purely for mounting the structure onto the rig. It was going to be layed onto the floor, however I have noted with the removal of the other C200X75 section, it can be used for the mid-section instead. (See diagram below)

The revision shown above is revision 3. My current revision (4) looks like so:

With a few extra pieces of information as shown here:

 

[Studiot]

If you can't fix it to the floor, why not tie it to nearby walls and/or columns?

 

[nvn]

MarleyDH: Would you be able to post a front view, and a side/end view, of your structure?

Are you sure you have M40 bolts? That would be very uncommon. Are you certain you do not have M39 bolts, or more commonly, M36 or M42 bolts? You should be able to measure the bolt outside diameter accurately. And measure the axial length of, say, ten threads, to compute the thread pitch accurately. Do you have M39 x 4, or M39 x 3 bolts (the second number is thread pitch, in mm). Or do you have M36 x 4, M36 x 3, M42 x 4.5, or M42 x 3?

 

[MarleyDH]

My apologies, one of the lab technicians thought I was asking about the hole diameter and he rounded up to 40 mm (from 38 mm)

When the lab was set up, they still used English units, so the holes are 1.5" holes, which is roughly 38 mm. Therefore the bolts are probably M36's, with the holes being drilled to 38 mm. The pitch I am unable to tell you.

I'll try and get an isometric up.

Right, this is my 3rd Revision of the idea, and I've produced an isometric for it. Which is analogous to the other one posted earlier.

Top View:

Isometric:

http://i.imgur.com/cCf4p.png

<< Image removed because it was too big. Please use the link above to view the image. >>

 

[nvn]

MarleyDH: It is doubtful the bolts are M36, because an M36 could strip the tapped hole. Get an actual bolt that will be used. Ensure it fits easily, but snugly, into a clean, undamaged floor bolt hole, turning it five or more full rotations. Once you have proven that it is an exact fit for the tapped hole, remove the bolt, and accurately measure, in mm, the axial length of ten threads on the bolt. Post the axial length of ten threads, in mm. Also, tell us what marks and/or characters appear on the bolt head.

And what is the name of the steel of the floor tapped holes, in your code? It could be as low as 250 or 235 MPa tensile yield strength steel (?), if tapped directly into structural steel. Or is it a threaded insert, mounted in the floor? What is the nominal pull-out strength of the insert? Are the tapped hole threads greased, or unlubricated?

Thanks for labeling the views (such as Top view, Front view, etc.), so we know which direction we are looking.

It will be interesting to see what structure you attach to your support rig, when you post views of your structure.

By the way, can you reduce the width of your images in post 14? You can reduce the width at, e.g., imageshack.us. Please do not post images wider than 640 pixels within forum threads. If any image is wider than 640 pixels, just post a plain text link (url) to the image, instead of embedding the image within your post.